u.k. and global wine markets by 2025, and implications of ... · u.k. and global wine markets by...

31
U.K. and Global Wine Markets by 2025, and Implications of Brexit* Kym Anderson a and Glyn Wittwer b Abstract The United Kingdom has accounted for a major share of the worlds wine imports for centuries, and wine accounts for more than one-third of U.K. alcohol consumption. It is therefore not sur- prising that suppliers of those imports and U.K. wine consumers, producers, traders, distributors, and retailers are focusing on what the United Kingdoms planned withdrawal from the European Union (Brexit) might mean for them. In this paper, a model of the worlds wine markets is used to project those markets to 2025 without, and then with, the occurrence of Brexit. The Brexit sce- narios involve adjustment not just to U.K. and EU27 (the countries remaining in the European Union) bilateral tariffs but also to assumed changes to the United Kingdoms income growth and currency. The relative importance of each of these three components of the initial shock are reported, as are impacts on bilateral wine-trade values and volumes for still and sparkling wines. The results suggest that the impact outside the United Kingdom will be minor compared with other developments in the worlds wine markets. Inside the United Kingdom, however, the effect of Brexit on incomes and the British pound are likely to have nontrivial initial impacts on the domestic wine market and to be far more consequential than the direct impact of changes in bilateral tariffs. (JEL Classications: F15, F14, F13) Keywords: Brexit, global wine market modeling, preferential trading agreements. I. Introduction The United Kingdoms planned withdrawal from the European Union (Brexit) will affect markets for many products, including wine. Very little wine is produced in the *The authors are grateful for helpful comments from referees and from L. Alan Winters, Jancis Robinson, Tamara Roberts, and other participants in a seminar at Chatham House in London on May 19, 2018. We also thank the U.K. Trade Policy Observatory of the University of Sussex and the Royal Institute for International Affairs for hosting the seminar. a Kym Anderson, Wine Economics Research Centre, School of Economics, University of Adelaide, Adelaide, SA 5005, Australia; and Australian National University, Canberra, ACT 2600; e-mail: kym. [email protected] (corresponding author). b Glyn Wittwer, Professorial Fellow, Centre of Policy Studies, Victoria University, PO Box 14428, Melbourne, Vic. 8001, Australia; e-mail: [email protected]. Journal of Wine Economics, Volume 12, Number 3, 2017, Pages 221251 doi:10.1017/jwe.2017.19 © American Association of Wine Economists, 2017

Upload: others

Post on 15-Jun-2020

44 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: U.K. and Global Wine Markets by 2025, and Implications of ... · U.K. and Global Wine Markets by 2025, and Implications of Brexit* Kym Andersona and Glyn Wittwerb Abstract The United

UK and Global Wine Markets by 2025 and Implicationsof Brexit

Kym Anderson a and Glyn Wittwer b

Abstract

The United Kingdom has accounted for a major share of the worldrsquos wine imports for centuriesand wine accounts for more than one-third of UK alcohol consumption It is therefore not sur-prising that suppliers of those imports and UK wine consumers producers traders distributorsand retailers are focusing on what the United Kingdomrsquos plannedwithdrawal from the EuropeanUnion (Brexit) might mean for them In this paper a model of the worldrsquos wine markets is used toproject those markets to 2025 without and then with the occurrence of Brexit The Brexit sce-narios involve adjustment not just to UK and EU27 (the countries remaining in the EuropeanUnion) bilateral tariffs but also to assumed changes to the UnitedKingdomrsquos income growth andcurrency The relative importance of each of these three components of the initial shock arereported as are impacts on bilateral wine-trade values and volumes for still and sparklingwines The results suggest that the impact outside the United Kingdom will be minor comparedwith other developments in the worldrsquos wine markets Inside the United Kingdom however theeffect of Brexit on incomes and the British pound are likely to have nontrivial initial impacts onthe domestic wine market and to be far more consequential than the direct impact of changes inbilateral tariffs (JEL Classifications F15 F14 F13)

Keywords Brexit global wine market modeling preferential trading agreements

I Introduction

The United Kingdomrsquos planned withdrawal from the European Union (Brexit) willaffect markets for many products including wine Very little wine is produced in the

The authors are grateful for helpful comments from referees and from L AlanWinters Jancis RobinsonTamara Roberts and other participants in a seminar at Chatham House in London on May 19 2018 Wealso thank the UK Trade Policy Observatory of the University of Sussex and the Royal Institute forInternational Affairs for hosting the seminaraKym Anderson Wine Economics Research Centre School of Economics University of AdelaideAdelaide SA 5005 Australia and Australian National University Canberra ACT 2600 e-mail kymandersonadelaideeduau (corresponding author)bGlyn Wittwer Professorial Fellow Centre of Policy Studies Victoria University PO Box 14428Melbourne Vic 8001 Australia e-mail GlynWittwervueduau

Journal of Wine Economics Volume 12 Number 3 2017 Pages 221ndash251doi101017jwe201719

copy American Association of Wine Economists 2017

United Kingdom (although the volume is now five times what it was in the 1980s)and wine has accounted on average for just 05 of UK merchandise imports sinceWorld War I Over the past six decades though winersquos share of UK alcohol con-sumption has steadily risen from 5 to more than one-third so wine traders distrib-utors and retailers as well as consumers are concerned about Brexitrsquos potentialimpact Among wine producers and consumers outside the United KingdomBrexit is also attracting considerable attention because the United Kingdom hasalways accounted for a major share of the worldrsquos wine imports

To examine how wine markets might be affected by the United Kingdomrsquos exitfrom the European Union it is necessary to look beyond just the immediatetrade-reducing and trade-diverting effects of altering bilateral import tariffs thatare the focus of the standard comparative static economic theory of (withdrawalfrom) customs unions Because the process of exiting establishing new tradingarrangements and adjusting to altered incentives is expected to spread over manyyears and initially to slow the growth of UK incomes and to devalue the poundone needs to begin with a projection of how wine markets would have lookedwithout Brexit in several years and then show how that projected baseline mightchange under various Brexit scenarios and a replacement trade agreementbetween the United Kingdom and the EU27 We do that using a model of theworldrsquos wine markets projected to 2025

This paper begins by briefly reviewing the United Kingdomrsquos historic and presentroles in global wine markets It then summarizes what trade theory would lead us toexpect for a country leaving a customs union1 A model of the worldrsquos wine marketsis then outlined along with a description of the way in which the model projects andof how that projection can be altered to simulate the effects of Brexit and subsequentbilateral trade agreements on the United Kingdom and other countries The modelrsquosresulting prospective changes to grape and wine markets by 2025 for a baseline caseare then summarized followed by results for a range of alternative adjustments fol-lowing Brexit The final section draws out implications of the findings for wine

1As of mid-2017 it was not yet certain whether the United Kingdom would leave the single market oralternately form a UKndashEU27 customs union but both options would require the United Kingdom toretain the European Unionrsquos tariff policy and the former would also require the United Kingdom to con-tinue to allow freedom of movement of labor and to remain under the European Court of Justice ndash none ofwhich Brexiteers want We therefore ignore these two possibilities and following Rollo et al (2016) andSmith (2017) assume that the United Kingdom will commit to the current EU tariff schedule at theWorld Trade Organization in the first instance and then seek a free trade agreement (FTA) with theEU27 Presumably other trading partners would want to wait and see what that FTA looked likebefore signing bilateral FTAs of their own with the United Kingdom Deep FTAs normally takeseveral years to negotiate and get ratified by the relevant parliaments even when a sufficient number ofexperienced negotiators is on each side of the table At present the United Kingdom has very few suchnegotiators having delegated that task to the European Commission (EC) in Brussels during the past45 years of its EU membership

222 UK and Global Wine Markets by 2025 and Implications of Brexit

markets and their participants in the United Kingdom and abroad within andoutside the European Union

II Historical Backdrop Wine in the United Kingdom and the UnitedKingdom in Global Wine Markets

Wine has rarely accounted for more than 2 of UK merchandise imports and sinceWorld War I it has averaged just 05 Also prior to 1950 wine accounted for lessthan 4 of the volume of alcohol consumed in the United Kingdom (except brieflyin the 1920s) and not much more as a share of alcohol expenditure Over the past sixdecades though winersquos share of UK alcohol consumption has steadily risen and itnow exceeds one-third in volume and value terms (Figure 1)

The United Kingdom is a key player in wine-trade circles as it has alwaysaccounted for a major ndash and often the largest ndash share of the worldrsquos wine importsPrior to 1960 the United Kingdomrsquos share of the value of world imports of wineroughly matched its share for all merchandise more than 20 in the 19th centurybut steadily declining to around 10 by 1960 Since then the two shares havediverged with the all-goods share falling to around 4 while the wine shareclimbed back above 20 by the beginning of the twenty-first century The UnitedKingdomrsquos share of the volume of global wine imports has always been below itsvalue share but those two shares have converged since 1960 and are now botharound 14 (Figure 2)

These two trends are summarized in Figure 3 The ratio of the United Kingdomrsquosshares of world imports of wine to that of all goods rose from 1 to 4 between 1960and 2000 and the ratio of the United Kingdomrsquos average import price to the worldaverage has come down from between 3 and 4 pre-1960 (when the United Kingdomwas mostly importing relatively expensive wines from Bordeaux and Champagne) toabout 12 by 1980 and is now close to 1

Associated with the change in the average quality of UK wine imports are dra-matic changes in the importance of different wine-exporting countries to UKimports and in UK shares in the wine exports of those countries

Those recent shares are very different from what they were in 1995 when the NewWorld was just beginning to expand its wine exports But such changes are notunprecedented Indeed they changed considerably not only because of the changingglobal shares of the various wine-exporting countries but also because of changes inthe United Kingdomrsquos preferential trading arrangements such as the 1703 MethuenTreaty with Portugal the 1860 Cobden-Chevalier Treaty with France and the 1932Ottawa Agreement with Commonwealth countries (Tables 1 and 2)

In the middle rows of Table 1 the 2010ndash2014 shares of exporters in the interna-tional market are shown below their shares in the UK markets For some countries

Kym Anderson and Glyn Wittwer 223

those two sets of shares are similar (France Italy Germany Chile) while for othersthey are very different The latter is most noticeable for the former colonies ofAustralia New Zealand and South Africa whose shares in the United Kingdomare more than twice their shares in the rest of the world Both shares for theEU27 exceed two-thirds Many commentators expect the EU27 shares to dropand shares of eg Australia and New Zealand to rise as a consequence of Brexit

The United Kingdomrsquos recent importance to producers in wine-exporting coun-tries is clear from Figure 4 For seven key suppliers the United Kingdom accountedin 2010ndash2014 for more than one-sixth of their wine-export earnings and for three ofthem (Australia the United States and New Zealand) the United Kingdom was amarket for more than one-third of their volume of wine exports

Recent import duties and other taxes affecting the consumer prices of alcohol inthe United Kingdom are summarized in Table 3 expressed per liter of beverageImport and excise duties on wines vary according to their alcohol content Aboutone-third of UK wine imports arrive in bulk perhaps half of which contain lessthan 13 alcohol and one-tenth of imports are sparkling so the volume-weightedaverage import duty is 13 pence per liter This figure contrasts with the volume-weighted average excise tax on wine which is 297 pence per liter To that increasein the wholesale price is added perhaps a 25 retail margin for off-trade sales andwell over a 100 margin for many restaurant sales before the 20 value-addedtax (VAT) is added So the average import price in 2013ndash2015 of 222 pence per

Figure 1

Winersquos Shares of UK Merchandise Import Value and of Volume and Value of UK AlcoholConsumptiona 1800 to 2015 ()

Fig1-Colou

ron

line

BW

inprint

a Reliable wine consumption volume data are not available for the 1930s and 1940s nor are value of alcohol consumption data pre-1955Sources Compiled from data in Anderson and Pinilla (2017) and Holmes and Anderson (2017a b)

224 UK and Global Wine Markets by 2025 and Implications of Brexit

liter is escalated to 800 pence for off-trade sales and more than 1300 pence for on-trade sales (which are about one-fifth of the total sales volume in the UnitedKingdom) The share of that latter retail price that is due to tariffs on wineimports is thus just 1 The change in the United Kingdomrsquos import trade regimefrom imposing such tariffs on wines currently imported free of duty from theEuropean Union Chile and South Africa is therefore likely to be very minorHowever as becomes clear below the effect of Brexit on wine sales involves farmore than just the trade-reducing and trade-diverting effects of altering bilateralimport tariffs

III The Economics of Leaving a Customs Union

The standard theory of customs unions (Viner 1950) focuses on the fact that whencountries join a union and impose a common external tariff on imports from non-union countries net trade can be created (depending on the height of the commonexternal tariff relative to the previous national tariffs) but trade can also be diver-sified (because of the preference to producers within the union) When a country

Figure 2

UK Shares of Value of World Merchandise Imports and of Value and Volume of World WineImports 1850 to 2015 ( 3-Year Averages to Year Shown)

Fig2-Colou

ron

line

BW

inprint

Source Compiled from data in Anderson and Pinilla (2017)

Kym Anderson and Glyn Wittwer 225

leaves a union the reverse happens because the leaving countryrsquos tariffs now apply toits imports from union countries as well as from the rest of the world Hence importsfrom union countries fall because other things equal these countriesrsquo preferentialaccess to the leaving country no longer applies That is also the case for countriesthat enjoyed an FTA with the union Of significance to wine markets are theEuropean Unionrsquos FTAs with Chile and South Africa

The size of that trade-diverting impact of leaving the union on wine depends onthe external tariff imposed on wine imports not only by the union but also by theleaving country Some have suggested that the United Kingdom should becomethe Hong Kong of Europe and move immediately to free trade on all productsOthers have suggested that this move would impose huge structural changes onthe UK economy that society would not tolerate ndash at least not without majorcompensation packages But all agree that a new trade policy that sets most-favored-nation (MFN) tariff rates is needed before the United Kingdom can beginto negotiate new preferential trading arrangements with the EU27 its FTA partnerssuch as Chile and South Africa or other countries Rollo Borchert Dawar Holmesand Winters (2016) suggest that the most practical trade policy for the UnitedKingdom to adopt at the outset is the European Unionrsquos tariff schedules previouslyagreed to at the WTO In all but one of the Brexit scenarios examined below we

Figure 3

UK Price Relative to World Price of Wine Imports and UK Wine-Import Intensitya 1950 to2015 ()

Fig3-Colou

ron

line

BW

inprint

a Import intensity is defined as the United Kingdomrsquos share of the value of global wine imports divided by the United Kingdomrsquos share of theworldrsquos total merchandise importsSource Compiled from data in Anderson and Pinilla (2017)

226 UK and Global Wine Markets by 2025 and Implications of Brexit

Table1

Shares

ofUKW

ineIm

portsfrom

TodayrsquosKey

Wine-Exp

ortin

gCountries1

675to

2014

andProjected

to2025

with

outandwith

Brexit()

Volum

eFrance

Spain

Portugal

Italy

Germany

South

Africa

Australia

United

States

Chile

New

Zealand

Argentin

aOther

countries

Total

1675ndash1696

2542

231

90

00

00

00

100

1697ndash1862

526

491

32

00

00

014

100

1863ndash1919

2626

220

00

20

00

025

100

1920ndash1940

1218

320

05

140

00

020

100

1995

volume

3210

318

143

73

21

07

100

value

439

413

102

83

21

05

100

2010ndash2014

volume

159

117

48

2110

84

12

100

value

358

215

44

115

66

12

100

Exp

ortersrsquoshareof

world

wineexports2010ndash2014

volume

1520

322

44

74

72

39

100

value

309

319

42

64

53

312

100

2025

projected

noBrexit

100

volume

188

188

26

207

45

45

65

43

64

17

20

92

100

value

323

100

30

218

44

23

58

44

54

41

20

45

100

2025

projected

with

Brexit

volume

190

188

27

208

43

44

66

42

63

17

19

93

100

value

311

96

29

210

41

24

59

44

55

41

20

70

100

SourcesCom

piledfrom

data

inAnd

ersonan

dPinilla

(201

7)to

1940

UnitedNations

COMTRADEh

ttpscomtrad

eun

orgdata

for19

95ndash2

014

andau

thorsrsquomod

elresults

for20

25

Kym Anderson and Glyn Wittwer 227

assume that Rollo et alrsquos suggestion will be the new UK trade policy commitmentto WTO members that subsequent negotiations for preferential arrangements willtake years and that any consequent agreements with the European Union andothers will be gradually implemented after that

The impact of leaving a customs union on wine markets does not only come fromtariff changes Also relevant are any effects of leaving on real UK incomes and thevalue of the British pound If the United Kingdom were to move immediately to freetrade on all products (the Hong Kong option) its per-capita income could eventuallyrise but only after considerable adjustment Should the United Kingdom insteadcommit to the current EU tariff schedule at the WTO in the first instance as weassume below then its per-capita income growth rate and the British poundrsquosexchange rates almost certainly will be lower for some time ndash at least until newtrade agreement negotiations with the EU27 and others are sufficiently advanced

Table 2Taxes on British Wine Imports by Source 1660ndash1862 (British Pounds per Kiloliter)

France Germany Spain Portugal South Africa

1660ndash1665 7 9 8 81666ndash1684 7 9 8 81685ndash1691 14 20 19 181692ndash1695 22 20 19 181696 47 20 19 181697ndash1702 51 25 23 221703 52 27 24 231704ndash1744 55 31 26 251745ndash1762 63 35 30 291763ndash1777 71 39 34 331778 79 43 38 371779 84 41 40 391780ndash1781 92 49 44 431782ndash1785 96 51 47 46 441786 65 51 37 37 371787ndash1794 47 51 32 32 371795 78 64 51 51 571796ndash1797 108 92 71 71 771798 111 96 73 73 791799ndash1801 107 92 71 71 771802 112 97 74 74 801803 131 109 87 87 871804 142 117 95 95 951805ndash1824 144 119 96 96 961825ndash1830 78 50 50 50 251831ndash1859 58 58 58 58 291860 32 32 32 32 321861 16 21 21 21 211862 11 26 26 26 26

Source Summarized from Ludington (2013 Table A1)

228 UK and Global Wine Markets by 2025 and Implications of Brexit

Figure 4

UK Shares in Wine Exports of Key Wine-Exporting Countries 2010ndash2014 ()

Fig4-BW

onlin

eBW

inprint

Source Compiled from data in Anderson and Pinilla (2017)

Table 3Import Duties Excise Duties and VAT Affecting Consumer Prices of Wine and Other Alcohol

in the United Kingdom April 1 2017

VAT 20

MFN import duties on wine pound per liter-bottled still wine lt13 alc 0114-bottled still wine 13ndash15 alc 0134-bottled still wine 15ndash18 alc 0162-bulk still wine lt13 alc 0086-bulk still wine 13ndash15 alc 0105-bulk still wine 15ndash18 alc 0134-sparkling wine 0278Weighted averagea 0130

Excise duties on alcohol pound per liter-still wine lt15 alc 2887-still wine 15ndash22 alc 3848-sparkling wine 55ndash85 alc 2795-sparkling wine 85ndash15 alc 3697-spirits (assumed 40 alc) 11551-beer (assumed 5 alc) 0954

VAT on alcohol 20

a Assumes that one-third of UK wine imports arrives in bulk that half has less than 13 alcohol and that one-tenth of imports is sparkling

Sources HM Revenue httpswwwgovukgovernmentpublicationsalcohol-duty-rate-changes for excise duties and VAT accessed April 92017 and Wine Australia (2015) for import duties converted at the 2016 average exchange rate of pound0740634 per euro

Kym Anderson and Glyn Wittwer 229

as to restore investor and consumer confidence in the United Kingdomrsquos economyThe size of those impacts is uncertain of course (see Baldwin 2016 and especiallyCampos 2016) so we consider a range of possibilities below

Those assumed adverse macroeconomic effects will add to the initial impact ofaltered wine tariffs on aggregate wine consumption in the United Kingdom andhence on its bilateral trades in wine They will make the loss of sales to theUnited Kingdom by EU (and Chilean and South African) suppliers greater thanwould otherwise be the case And they will reduce the likelihood that other countriesrsquosales of wine in the United Kingdom will be higher than in the baseline Indeed themacroeconomic effects could outweigh the trade-diverting effects so that even coun-tries that are currently discriminated against by the EU28rsquos wine-trade policy may beworse off because of Brexit

IV Global Wine Markets Model and Database

Our model of the worldrsquos wine markets first published by Wittwer Berger andAnderson (2003) and revised by Anderson and Wittwer (2013) is summarized inthe Appendix It disaggregates wine markets into four types namely nonpremiumcommercial-premium and superpremium still wines and sparkling wines2 There aretwo types of grapes premium and nonpremium Nonpremium wine uses nonpre-mium grapes exclusively superpremium wines use premium grapes exclusivelyand commercial-premium and sparkling wines use both types of grapes to varyingextents across countries The world is divided into 44 individual nations and 7 com-posite geographic regions that capture all other countries

The modelrsquos database is calibrated to 2014 based on the comprehensive wine-market volume and value data and trade and excise-tax data provided inAnderson and Pinilla (2017) and in Anderson Nelgen and Pinilla (2017) It is pro-jected assuming that aggregate national consumption population and realexchange rates change between 2014 and 2025 to the extent shown in AppendixTable 13 The Brexit alternatives to that baseline also are projected to 2025

Concerning preferences we assume a continued considerable swing toward allwine types in China and a swing away from nonpremium wines in all other countriesuntil 2025

In our baseline scenario grape- and wine-industry total-factor productivity areassumed to grow at 1 per year everywhere while grape- and wine-industry

2Commercial-premium still wines are defined byAnderson Nelgen and Pinilla (2017) to cost between US$250 and US$750 per liter pretax at a countryrsquos border or wholesale3The real exchange rate changes over the projection period are the changes expected in the nominal valueof country irsquos currency relative to the US dollar times the expected ratio of the GDP deflator for theUnited States versus that for country i

230 UK and Global Wine Markets by 2025 and Implications of Brexit

capital is assumed to grow net of depreciation at 15 per year in China but zero else-where (consistent with the almost-zero growth in global wine production and con-sumption over the past two decades)

Two alternative scenarios are considered (ldquolargerdquo and ldquosmallrdquo) for the initialimpact of Brexit to provide a range of results We assume that because of theUnited Kingdomrsquos decision to exit the European Union the rate of UK realgross domestic product (GDP) growth is only one-third or two-thirds as fast overthe projection period (09 or 18 per year instead of 26) and the Britishpound will be 20 or 10 lower in real terms than in our modelrsquos core baselineprojection4 In both alternative scenarios we assume that the United Kingdomapplies the European Unionrsquos external tariffs on wine from March 30 2019 atthe end of the two-year period following the United Kingdomrsquos formal triggeringof Article 50 (see Table 3)

We then consider a subsequent impact of Brexit presumed to result from negoti-ating signing implementing and responding to an FTA with the EU27 by 2025(Chile and South Africa currently have preferential access to EU wine marketsbut we continue to assume in this scenario that the United Kingdom does not imple-ment new bilateral FTAs with them or others in our time frame)5 This subsequentscenario assumes the British pound returns to what it would have been in 2025 in theabsence of Brexit and that real incomes regain two-thirds of the value they wouldhave had without Brexit as compared with our ldquosmallrdquo initial Brexit scenario

This global model has supply-and-demand equations and hence quantitiesprices and price elasticities for each of the grape and wine products and for asingle composite of all other products in each country Income elasticities ofdemand also exist for each final product Grapes are assumed to not be traded inter-nationally but other products are exported and imported Each market is assumed tohave been in equilibrium before any shock and to find a new market-clearingoutcome following any exogenously introduced shock

4The nominal price of the British pound in US dollars in the fortnight following the Brexit vote on June23 2016 dropped 13 to US$130 and a year later the pound sat at the same rate having dropped to anaverage of $124 (an 18 devaluation) between October 2016 and April 2017 Our choice of a low of 10and a high of 20 aims to capture future possible rates while uncertainties remain The average real wagein the United Kingdom fell in the first half of 2017 and projected real GDP growth during 2018ndash2020 hasbeen revised down to 17 in the United Kingdomrsquos latest budget (HM Treasury 2017)5South Africa currently has duty-free access to the EU28 for just 50 megaliters (ML) of wine per yearbeyond which the MFN tariff rate applies South Africa currently exports around 320 ML to theEuropean Union one-third of which initially goes to the United Kingdom How that quota of 50 MLis divided between the United Kingdom and the EU27 is subject to future negotiation (Rollo et al2016 Swinbank 2017) In the following discussion we assume none of it is accepted by the UnitedKingdom so that all South African wine imports pay the United Kingdomrsquos MFN tariff under Brexitbut altering that assumption makes very little difference to our results for any countryrsquos trade exceptSouth Africarsquos

Kym Anderson and Glyn Wittwer 231

V Projecting Global Wine Markets to 2025

Global wine production and exports are projected in the baseline from 2014 to 2025consistent with past trends The modelrsquos global volume of production (and consump-tion) rises little over that 11-year period (9) made up of a 6 decline in nonpre-mium wine and a one-sixth rise in commercial-premium and superpremium wineIn real (2014 US$) value though global wine output and consumption increaseby about 50 in total and 60 in the two premium categories The internationaltrade projections are similar although a little larger with the share of global wineproduction exported (equivalent to the share of global consumption imported)rising 2 percentage points between 2014 and 2025

The baseline projection does not greatly alter the 2014 shares of various countriesin global wine production apart from China because we assume vineyard expansionthere is faster than elsewhere6 In value terms this assumption means that Chinamoves from fifth to fourth by 2025 behind France the United States and ItalySpain remains barely ahead of Australia and they and Germany take the nextthree places (Figure 5(a)) In total wine-production volume terms China movesfrom sixth to fifth place and Argentina drops from fifth to eighth (and fromeighth to ninth in value terms)

When their products are subdivided into fine wines (superpremium still plus spark-ing) commercial-premium wines and nonpremium wines France and the UnitedStates retain the highest two places on the global ladder for fine-wine productionand Spain and Italy retain the top two places for nonpremium wine As for commer-cial-premium wine production (defined to cost between US$250 and US$750 perliter pretax at a countryrsquos wholesale level or national border) Italy retains the topranking over our projections period but at least in terms of value China challengesFrance for second place

The country rankings by projected value of total wine consumption change some-what more than those for production by 2025 with China taking second place afterthe United States ahead of France and Germany and then the United Kingdomslightly overtaking Italy to slip into fifth place (Figure 5(b)) The United StatesFrance and Germany retain the top three rankings for consuming fine wine butCanada slightly overtakes Italy for fourth place in terms of value at least In thecase of commercial-wine consumption China strengthens its number-one positionahead of the United States and the United Kingdom does likewise vis-agrave-visGermany for third place

As for the projected changes in consumption volumes China is projected to dom-inate the increase in aggregate although the United States is projected to lead the

6 In fact Chinarsquos wine production fell steadily between 2012 and 2016 by a total of one-sixth so Chinarsquoswine imports may grow faster in practice than in this baseline projection

232 UK and Global Wine Markets by 2025 and Implications of Brexit

Figure 5

Value of Wine Production and Consumption in Key Countries 2014 and Projected Baseline for2025 (2014 US$ Million at WineryWholesale Pretax Prices)

Fig5-BW

onlin

eBW

inprint

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 233

increase in consumption of fine wine In Western Europe and in the SouthernHemispherersquos New World countries fine wines are projected to substitute for com-mercial wines (defined as the sum of commercial-premium and nonpremiumwines) with almost no change in total wine consumption Sub-Saharan Africa isthe next region that is projected to take off with its growth accounting for morethan one-third of the rest of the worldrsquos increase in volume consumed

Those differences in production versus consumption rankings are reflected ininternational trade Figure 6 shows that France Italy and Spain remain the threedominant exporters of wine in aggregate value but the rankings of the next fewchange with Australia being slightly ahead of Chile and the United StatesGermany and New Zealand being nearly tied for sixth place in value termsFrance and Italy are even more dominant in fine-wine exports and remain so by2025 while Italy outranks France in the commercial-premium export categoryand Spain outranks Italy Australia and Chile in the nonpremium export class

Among the importers the United States and the United Kingdom are projected tocontinue to hold the first two places in 2025 in value terms but China moves intothird place slightly ahead of Germany followed well behind by Canada HongKong Belgium-Luxembourg the Netherlands and Japan (Figure 6(b)) OtherAfrica (excluding South Africa) is projected to experience the largest increase inimports among all the other regions followed by Other Asia which becomes asbig as Germany in value terms (Figure 7(a)) In terms of total volume of wineimports Germany and the United Kingdom held the top two shares in 2014 butby 2025 the United Kingdom is projected to be well ahead of Germany (Figure 7(b)) However this projection ignores the effects of Brexit to which we now turn

VI How Might Wine Markets Be Affected by the United Kingdomrsquos Exitfrom the European Union

As mentioned earlier for our two alternative scenarios to capture the initial effects ofBrexit (ldquolargerdquo and ldquosmallrdquo) we assume that following the United Kingdomrsquos exitfrom the European Union the United Kingdomrsquos rate of economic growth wouldbe only one-third or two-thirds as fast for the period to 2025 the British poundwould be 20 or 10 lower in real terms than in our modelrsquos baseline projectionand the United Kingdom would apply the European Unionrsquos external tariff onwine to imports from EU member countries (as part of establishing MFN ratesvia the WTO to then start new bilateral FTA negotiations) In these initial scenarioswe assume the United Kingdom does not implement any new FTAs particularlywith the EU27 Chile and South Africa Following the discussion of those initialresults we present the results of a subsequent scenario that includes the implemen-tation of an FTA between the United Kingdom and the EU27 We assume these aretwo discrete steps with no agreed arrangements to smooth the transition betweenthem

234 UK and Global Wine Markets by 2025 and Implications of Brexit

Figure 6

Value of Wine Exports and Imports Key Wine-Trading Countries 2014 and Projected Baselinefor 2025 (2014 US$ million)

Fig6-BW

onlin

eBW

inprint

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 235

Figure 7

National Shares of Global Wine Import Value and Volume 2014 and Projected Baselinefor 2025 ()

Fig7-BW

onlin

eBW

inprint

Source Authorsrsquo model results

236 UK and Global Wine Markets by 2025 and Implications of Brexit

A Initial Impact of Brexit

Because the British pound dropped by one-sixth against the US dollar in the fourmonths following the Brexit vote in June 2016 and because the UnitedKingdomrsquos average real wage fell 1 between November 2016 and April 2017(according to the latest available data see Office of National Statistics 2017) weviewed our ldquolargerdquo scenario as more likely soon after the Brexit vote But wecompare those results with ones from our ldquosmallrdquo initial-impact scenario becauseit has become more likely following the June 2017 general election that ldquosofterrdquooptions will be considered by the new government Generally the results areabout half the size in the ldquosmallrdquo scenario with the exception of the bilateraltrade effects To show the sensitivity of results to our assumptions we point outthe differences when the ldquosmallrdquo results are not close to half the results shown forthe ldquolargerdquo scenario

Before turning to those results we first present a ldquofree-traderdquo scenario whichassumes the United Kingdom chooses the radical option of becoming the HongKong of Europe Unlikely as this scenario is the results provide assurance thatour global wine model generates the standard types of effects of exiting a customsunion In this ldquofree-traderdquo scenario the United Kingdom is assumed to move tozero tariffs on all wine imports Thus the current preferences on EU Chileanand South African wine imports disappear as all other wine exporters enjoy thesame free access to UK wine markets We assume in this scenario that Brexit hasno impact on the United Kingdomrsquos real income or consumption growth rates by2025 and that by then the British pound has returned to its 2014 value relative toother currencies (similar to the real exchange rate assumed in our baseline scenarioto 2025 see Appendix Table A1)7 Under these assumptions EU Chilean andSouth African wine exports to the United Kingdom in 2025 are 003 lower thanin the no-Brexit baseline and UK imports from other wine exporters are 025higher but the United Kingdomrsquos total wine imports are hardly any different(because local wine prices fall by only 05 and incomes are unchanged) In this sce-nario UK consumers expand their volume of consumption by just 05 while theEU27 exports more and other countries export less to the rest of the world such thatworld wine exports are almost unchanged In short all these effects are in the direc-tion that comparative static-customs-union theory would predict but they are smallbecause we assume the United Kingdomrsquos real income and currency are the same asin our baseline for 2025 as described in the previous section

In the ldquolargerdquo alternate initial scenario involving Brexit as compared with thebaseline scenario to 2025 the consumer price of wine in 2025 is 22 higher in the

7If this scenario were seen as a credible long-run trade policy it would encourage more investment thatwould eventually raise UK incomes but not without major structural changes that would initiallydisrupt the economy Implicitly we assume in our free-trade scenario that these two opposite impactson real UK incomes and the British pound exactly offset each other by 2025

Kym Anderson and Glyn Wittwer 237

United Kingdom in local currency terms (20 because of real depreciation of theBritish pound 4 because of new tariffs on EU Chilean and South Africanwines and ndash2 because of slower UK income growth) The volume of UKwine consumption is 28 lower 16 because of slower UK economic growth7 because of real depreciation of the British pound and 5 because of newtariffs Superpremium still-wine sales are the most affected dropping by two-fifths while sparkling and commercial-premium wines drop a bit less than one-quarter Because the average price rises by more than the decline in volume soldthe aggregate value of UK sales even in local currency terms falls under thisldquolargerdquo Brexit scenario Under the ldquosmallrdquo Brexit scenario the consumer price ofwine in 2025 is 11 higher in the United Kingdom and its volume of wine consump-tion is 17 lower

The volume of projected UK imports in 2025 is 427 ML or nearly one-quarterless in the ldquolargerdquo scenario than in the baseline scenario comprising 58ML less spar-kling wine 31 ML less superpremium still wine and 339 ML less commercial-premium wine World imports are lower by just 239 ML because imports byother countries are 189 ML higher in response to the lower international wineprices in this scenario In value terms UK imports are $175 billion (or 27)lower in 2025 because of ldquolargerdquo Brexit $113 billion because of lower incomes$038 billion because of the fall in the British pound and $014 billion because ofthe rise in wine-import tariffs (Table 4) These aggregate trade impacts are a littlemore than half this size under the ldquosmallrdquo Brexit scenario

Despite the import levels falling because of raised import tariffs domestic con-sumption of all three quality categories of UK-produced wine is lower with thanwithout Brexit because of shrunken demand for all wines resulting from loweredUK incomes and raised local prices because of devaluation of the British poundThe British poundrsquos devaluation does make it easier for the United Kingdom tosell wines abroad though The countryrsquos exports are 7 ML or nearly 5 higherin 2025 in the ldquolargerdquo Brexit scenario and UK production is 3 higher ThoseUK exports (or re-exports of imported bulk wine after it is bottled in the UnitedKingdom) that go to EU27 countries are reduced though because of the tariffnow imposed at the new EU border

Without Brexit the United Kingdomrsquos shares of global wine imports are slightlyhigher in volume terms in 2025 than in 2010ndash2015 but 2 percentage points lower invalue terms thanks to East Asiarsquos expanding demand for imports of premium winesWith ldquolargerdquo Brexit however that value share is an additional 2 percentage pointslower and the volume share is almost 5 points lower (Figure 8) The net effects ofthese impacts on global trade are shown in Figure 9 Most of the initial tradeeffect of Brexit is a large decline in net imports of wine by the United Kingdomwith very little offsetting positive effect on trade in the rest of the world Theldquosmallrdquo Brexit numbers are a bit more than half of those for ldquolargerdquo Brexit

238 UK and Global Wine Markets by 2025 and Implications of Brexit

Table4

Difference

in2025

Projected

VolumeandValue

ofWineIm

portsb

ytheUnitedKingdom

andtheResto

fthe

World

asaConsequence

oftheInitial

BrexitS

hock

(MLand2014US$

Million

ldquoLargerdquoSc

enario)

Volum

e(M

L)

Value

(US$

million)

NP+CPa

SuperPrb

Sparkling

TOTAL

NP+CPa

SuperPrb

Sparkling

TOTAL

ΔUK

impo

rtsdu

eto

Low

erincomes

minus198

minus20

minus29

minus247

58minus644

minus253

minus234

minus1131

65Low

erpo

und

minus70

minus10

minus14

minus93

22minus248

minus127

minus102

minus476

27Highertariffs

minus71

minus1minus16

minus87

20minus110

minus8minus2

4minus143

8TOTA

Lminus339

minus31

minus58

minus427

100

minus1001

minus388

minus360

minus1750

100

difffrom

base

2332

3325

2432

3227

of

totalc

uts

797

14100

5722

21100

ΔROW

netim

ports

143

2125

189

230

143

minus181

192

ΔWORLD

TRADE

minus195

minus10

minus34

minus239

minus763

minus246

minus543

minus1552

aNon

prem

ium

plus

commercial-premium

still

wines

bSu

perpremium

still

wines

SourceA

utho

rsrsquomod

elresults

Kym Anderson and Glyn Wittwer 239

The aggregate effect of ldquolargerdquo Brexit on the market shares of various wine-export-ing countries in the United Kingdom is almost indiscernible even with one decimalpoint The projected 2025 shares are quite different from the actual 2014 shares forseveral countries They are much smaller in 2025 for South Africa Australia andNew Zealand (and the United States in volume terms) and are much larger involume for Spain and in value for Italy These results occur because wine-exportingcountries benefit differentially from the varying rates of growth in net importdemand for wine in non-UK countries over this projection period The most impor-tant projected changes are the increase in the real value of annual wine importsbetween 2014 and 2025 by China (200 or $3 billion) Other Asia (110 or $22billion) and Africa (270 or $16 billion) More than half of Australiarsquos increasein annual exports from 2014 to 2025 go to Asia and more than half of SouthAfricarsquos increase in exports go to Other Africa

Table 5 reveals that ldquolargerdquo Brexit lowers EU Chilean and South African wineexports by 150 ML or US$12 billion in the case of the European Union with someof their exports diverted from the United Kingdom to the EU27 and other marketsin competition with New World exporters Although the United States Australiaand Argentina sell only a little less to the United Kingdom they also sell less toother countries For Chile and South Africa which lose their preferential access toUK (but not to EU27) markets in this Brexit scenario some exports are redirected

Figure 8

UK Shares of World Wine Imports 2010ndash2015 and Projected to 2025 without and with theldquoLargerdquo Initial Brexit Shock ()

Fig8-BW

onlin

eBW

inprint

Sources Anderson and Pinilla (2017) and authorsrsquo model results

240 UK and Global Wine Markets by 2025 and Implications of Brexit

from the United Kingdom to EU27 countries ndash but again they export less overallGlobal wine trade in 2025 is 240 ML (19) or $18 billion (35) less under thisldquolargerdquo Brexit scenario The percentage by which wine exportersrsquo trade shrinks isgreater for values than for volumes because of changes in relative prices of differ-ent-quality wines Those differences are shown in the numbers in parentheses inTable 5

Three other points are worth making about Table 5 One is that Australia sellsslightly more to the United Kingdom in the ldquosmallrdquo Brexit scenario rather thanslightly less as in the ldquolargerdquo Brexit case Evidently the negative income and price(devaluation) effects do not more than offset the positive trade-diverting effect onAustralian exports to the United Kingdom of removing preferences in the ldquosmallrdquoscenario Second New Zealand sells slightly more to non-UK countries underBrexit despite greater competition from the EU27 Chile and South Africa Thisanomaly is due to changes in the relative prices of different qualities of wine inglobal wine markets bearing in mind that New Zealand has the worldrsquos highestaverage price for still-wine exports And third the value (but not the volume) ofexports of ldquoOtherrdquo countries to markets other than the United Kingdom arehigher under Brexit This result too is due to changes in the relative prices of differ-ent qualities of wine in global wine markets

Figure 9

Difference in 2025 Wine Import Volumes and Values as a Result of the ldquoLargerdquo Initial BrexitShock (ML and US$ Million in 2014 US dollars)

Fig9-BW

onlin

eBW

inprint

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 241

B Subsequent Impact of Brexit from a UK-EU27 FTA

The next-most-likely step in the Brexit process is for the United Kingdom to nego-tiate a new trade arrangement with the EU27 We therefore assume that a UK-EU27FTAwith free bilateral wine trade is implemented and adjusted to by 2025 and thatprogress toward that end occurs soon enough that the adverse macroeconomicshocks from the initial impact of uncertainty over the Brexit process are confinedto those assumed in the ldquosmallrdquo scenario outlined above In this subsequent scenariowe assume that the British pound returns to the value it would have reached in 2025in the absence of Brexit (ie reversing the 10 devaluation assumed in the ldquosmallrdquoinitial Brexit scenario) and that real incomes in the United Kingdom are 8 higher

Table 5Difference in 2025 BilateralWine Import Volumes and Values fromKey Exporters by the United

Kingdom and the Rest of the World (RoW) as a Result of Initial Brexit Shock(ML and 2014US$ Million)a

(a) ldquolargerdquo scenario

Volume (ML) Value (2014US$ million)

UK RoW WORLD () UK RoW WORLD ()

EU27 minus287 136 minus150 (ndash17) minus1187 minus5 minus1192 (ndash31)Chile minus59 35 minus25 (ndash30) minus169 31 minus138 (ndash48)South Africa minus53 35 minus18 (ndash32) minus105 20 minus85 (ndash67)USA minus7 minus6 minus13 (ndash24) minus75 minus40 minus115 (ndash50)Australia minus4 minus3 minus7 (ndash09) minus25 minus65 minus90 (ndash30)Argentina minus3 minus9 minus12 (ndash48) minus16 minus39 minus55 (ndash52)New Zealand minus11 9 minus2 (ndash09) minus162 71 minus91 (ndash43)Others minus2 minus10 minus12 (ndash02) minus11 minus52 minus63 (ndash44)WORLD minus427 187 minus240 (ndash19) minus1750 minus79 minus1829 (ndash35)

(a) ldquosmallrdquo scenario

Volume (ML) Value (2014US$ million)

UK RoW WORLD () UK RoW WORLD ()

EU27 minus178 82 minus96 (ndash12) minus692 minus43 minus736 (ndash19)Chile minus46 28 minus18 (ndash24) minus128 36 minus91 (ndash32)South Africa minus43 29 minus14 (ndash42) minus82 23 minus59 (ndash47)USA 1 minus6 minus5 (ndash11) minus23 minus28 minus51 (ndash22)Australia 5 minus10 minus5 (ndash06) 19 minus56 minus38 (ndash13)Argentina 0 minus6 minus6 (ndash26) minus3 minus25 minus29 (ndash27)New Zealand minus5 4 minus1 (ndash06) minus80 34 minus46 (ndash22)Others 0 minus9 minus9 (ndash01) minus1 minus33 minus34 (ndash24)WORLD minus266 112 minus154 (ndash13) minus991 minus92 minus1083 (ndash21)

a Numbers in parentheses are the percentage difference between the Brexit and baseline scenarios for 2025 projected wine-import volumes orvalues by source

Source Authorsrsquo model results

242 UK and Global Wine Markets by 2025 and Implications of Brexit

in 2025 than in the ldquosmallrdquo scenario ndash that is they regain two-thirds of the value theywould have reached without Brexit-related uncertainty

This subsequent development in the Brexit process reverses most of but not allthe initial effects of Brexit by 2025 because of our assumption that the lost growthin the initial years of uncertainty following the Brexit vote are only partly recoveredby 2025 following the implementation of a UK-EU27 FTA Moreover the longer ittakes before this FTA is finalized and implemented the longer the estimated initialadverse effects persist and the larger the cumulative cost of Brexit to UK wine con-sumers and to grape and wine producers in wine-exporting countries

Table 6 summarizes the subsequent trade effects for 2025 It suggests that all butone-tenth of the loss in value of world trade in wine from the initial ldquosmallrdquo impact isrestored most of it because of a smaller reduction in the United Kingdomrsquos wineimports following the FTA Most of that improved outcome is because of recoveredimports from the EU27 commensurate with the latterrsquos high share of UK importsEven though Chile and South Africa are assumed in this scenario to have not yetsigned an FTA with the United Kingdom they export slightly more to the UnitedKingdom (and even more to all other countries) than in the initial ldquosmallrdquo scenario

Within the United Kingdom this FTA brings down the local currency consumerprice of wine by 9 largely offsetting the 11 rise in the initial ldquosmallrdquo Brexit sce-nario and it raises the volume consumed in the United Kingdom by 18 fully off-setting the 17 fall in the initial ldquosmallrdquo Brexit scenario

Table 6Difference in 2025 Bilateral Wine-Import Volumes and Values from Key Exporters by the

United Kingdom and the Rest of the World (RoW) as a Result of Implementing a UK-EU27FTA Following Initial ldquoSmallrdquo Brexit Shock (ML and 2014US$ Million Difference Relative to

Initial ldquoSmallrdquo Brexit Scenario)a

Volume (ML) Value (2014US$ million)

UK RoW WORLD (ldquosmallrdquo)a UK RoW WORLD (ldquosmallrdquo)a

EU27 212 minus112 100 (minus96) 750 0 750 (minus806)Chile 1 3 4 (minus18) 3 40 43 (minus92)South Africa 3 2 5 (minus14) 3 18 21 (minus60)USA 3 3 6 (minus5) 26 21 47 (minus50)Australia 1 1 2 (minus5) 8 38 46 (minus33)Argentina 1 4 5 (minus6) 4 20 24 (minus28)New Zealand 4 minus3 1 (minus1) 61 minus25 36 (minus46)Others 1 6 7 (minus9) 5 minus46 minus41 (92)WORLD 226 minus96 130 (minus154) 860 66 926 (minus1022)

a Numbers in parentheses are the world trade differences between the ldquosmallrdquo initial Brexit scenario and the baseline scenario copied fromcolumns 3 and 7 of Table 5(b)

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 243

In short this subsequent step in the Brexit process can be expected to restore muchof the initial adverse effects in the United Kingdom of the Brexit vote by 2025 Thewine exporters in the EU27 Chile and South Africa also lose less in this scenarioas compared with the initial ldquosmallrdquo scenario while other key exporters sell nearlyas much in 2025 in this scenario as they are projected to in the baseline ndash that isthey are beneficiaries of the reduced discrimination in global wine markets in thiscase where Chile and South Africa no longer enjoy preferential access to the UKmarket (and even though EU27 producers are again allowed to do so) Keep inmind though that although the estimated losses in the initial scenarios are reportedjust for 2025 they are expected to be felt throughout the preceding years of uncertaintyTherefore the longer it is before uncertainty abates and the United Kingdomrsquos currentpreferential trading arrangements are replaced by newones such as a UK-EU27 FTAthe higher the cumulative cost of the Brexit vote to the wine trade

VII Caveats and Conclusions

The above Brexit simulations are just a few of many scenarios that could be modeledObvious additional ones could also assume that FTAs are reached between theUnited Kingdom and other trading partners including Chile and South Africaand such countries as Australia and New Zealand Some of these other countriesare already exploring the ECrsquos invitation (see European Commission 2015) to nego-tiate an FTA with what will be the EU27 The sequence in which FTAs are signedand the speed with which they are implemented will matter (as was also the casewith the sequential signing over the past decade of bilateral FTAs with NortheastAsian countries by Chile Australia and New Zealand see Anderson andWittwer 2015) Even if the United Kingdom were able to sign additional bilateralFTAs and begin implementing them before 2025 it would have little effect on theabove results (because wine tariffs are such a minor contributor to them) unlessthose FTAs were to accelerate the United Kingdomrsquos economic growth and therise in the British poundrsquos value before 2025

We assume above that no changes are made to alcohol excise duties in the UnitedKingdom following Brexit even though they are scheduled to be progressively raisedwith inflation and may be raised even more for wine relative to spirits to offset theopposite effects of Brexit on those two domestic industries Nor do we make anyallowance for UK increases in consumer wine prices that may be needed to coverthe higher cost of clearing customs on imports from EU27 countries (which maybe more or less offset by increased smuggling and duty-free purchasing)Importers of fine wines including individuals who buy Bordeaux en primeur maywell reduce their demand because of concerns that delays on docks will affect thequality of their wines or because of greater currency uncertainty

To summarize Brexit is costly initially to UK consumers of wine (and of manyother tradable products) because the domestic retail price in local currency tax-

244 UK and Global Wine Markets by 2025 and Implications of Brexit

inclusive terms is 22 higher than otherwise in 2025 and the volume of wine con-sumed domestically is 28 lower in the ldquolargerdquo Brexit scenario (or 11 higher pricesand 17 lower quantities in the ldquosmallrdquo case) by 2025 unless a UK-EU27 FTA isdeveloped before then Even if such an FTA does get signed ratified by all 28 par-liaments and implemented by 2025 the slower income growth in the interim meansa smaller UK wine market in 2025 than would otherwise have been the case Thevolume reduction is a blow to many participants in UK wine-bottling transportingstoring wholesaling and retailing businesses in addition to restaurants and pubsVery little of that initial impact occurs because of higher import tariffs most impor-tant is the assumed fall in UK real incomes

The small but growing number of local UK vignerons (who supply less than 05of all domestic-wine sales and less than 4 of sparkling-wine sales) is initially pro-jected to sell less wine domestically because of reduced demand and to export only alittle extra abroad while the British pound is devalued Overall their production isonly 3 higher in 2025 even in the ldquolargerdquo Brexit scenario (not taking intoaccount the fact that their casual labor is likely to be more expensive under Brexitbecause of tighter restrictions on immigration) and their production diminishes ifand when a UK-EU27 FTA is in place

Even with a UK-EU27 FTA in place EU27 wine exporters export US$56 millionless wine in 2025 thanks to Brexit Chile and South Africa export $88 million lesswine and wine exports from the rest of the world are little different in aggregatealthough slightly larger from Australia

Clearly there will be great uncertainly for some time over the possible policy out-comes to flow from Brexit and their consequent sequential impacts on UK house-hold disposable incomes foreign exchange rates and bilateral wine tariffsMeanwhile the above projections under explicit assumptions provide some idea ofhow wine markets might be affected by the most-likely first two stages of theBrexit process (agreeing on a new tariff schedule at the WTO and agreeing to andimplementing a UK-EU27 FTA) In particular they make clear that nontrivialinitial impacts could affect the domestic wine market impacts that are likely to belarger than just the direct impact of changes in bilateral tariffs If the UnitedKingdom succeeds in getting countries to agree to transition arrangements thatdelay the changes in tariffs until new FTAs are signed and ratified the initialeffects will be less dramatic than in our first scenarios but it remains to be seenwhether any such agreements can be reached In any event the net effect of Brexiton the welfare of the worldrsquos consumers and producers of wine as a whole will benegative not just initially but permanently unless new trade policy commitmentsby the United Kingdom with major wine-exporting countries are sufficiently moreliberal than current arrangements

Kym Anderson and Glyn Wittwer 245

Appendix Key Equations in the Global Wine Markets Model

Amodel of the worldrsquos wine markets was first published byWittwer et al (2003) andmuch improved by Anderson and Wittwer (2013) Several significant enhancementshave been made to that original model Wine types have been disaggregated intomore types ndash namely nonpremium (including generic bulk) commercial-premiumand superpremium still wines and sparkling wines8 As in the original modelthere are two types of grapes premium and nonpremium Nonpremium wines usenonpremium grapes exclusively superpremium wines use mainly premium grapesand commercial-premium and sparkling wines use both types of grapes In theregional dimension the number of countries and country groups has expandedfrom 10 in the original model to 51 44 individual nations and 7 compositeregions The modelrsquos database is calibrated to 2014 for this paper based on datain Anderson et al (2017) trade data for which are downloaded from httpscomtradeunorg The modelrsquos supply-and-demand equations are based on the orig-inal ORANI modelrsquos theory (see Dixon Parmenter Sutton and Vincent 1982) andthe model is implemented using GEMPACK software (Harrison Horridge Jerieand Pearson 2014)

An enhancement of importance to the present study is the inclusion of exchange-rate variables in the model which allow a distinction between price impacts asobserved in local currency units and those observed in 2014 US dollars

In the model the grape and wine sectors minimize costs of intermediate inputssubject to weak constant elasticity of substitution (CES) substitutability betweeninputs By assumption no intermediate inputs are imported from other countries9Hence

Xcid frac14 f ethX1id CESfrac12P

cid=P1id $THORN eth1THORN

P1id X1id frac14X

c

Xcid P

cid eth2THORN

whereXcid is the quantity demanded of commodity c by grape or wine industry i in

region d Pcid is the corresponding price and X1id and P1id are the respective inter-

mediate composite quantities and prices

Two primary factors are employed in the sector labor (the quantity of which isendogenous with perfectly elastic supply) and capital Capital is usually treated asexogenous in quantity with rates of return bearing all the adjustment in thevarious scenarios This reflects the fact that grapes (a perennial crop) and wine-

8Commercial-premium still wines are defined by Anderson and Nelgen (2011) to be those priced betweenUS$250 and $750 per litre pre-tax at a countryrsquos border or wholesale9An exception concerns bulk wine imports used mainly in the bottling of commercial premium wine in theUK USA Canada Japan and Germany

246 UK and Global Wine Markets by 2025 and Implications of Brexit

plant capacity adjust slowly to market signals

Lid frac14 f ethFid CESfrac12Wid=PFid $THORN eth3THORN

Kid frac14 f ethFid CESethRid=PFid $THORN eth4THORN

PFid Fid frac14 Lid Wid thorn Kid Rid eth5THORN

Grape and wine producers are assumed to minimize costs subject to CES substitu-tion between capital and labor Equations (3) to (5) show primary factor demandsfor the labor composite Lid and capital Kid subject to a composite factor demandFid by industry i in region d The factor prices are Wid for labor Rid for capitalrentals and PFid for composite factor prices

Appendix Table A1Cumulative Consumption and Population Growth Rates and Changes in the Real Exchange Rate

(RER) Relative to the US Dollar 2014 to 2025 without Brexit ()

Aggregateconsumption Poprsquon RER

Aggregateconsumption Poprsquon RER

France 18 4 minus11 Australia 35 11 minus17Italy 11 2 minus9 New Zealand 32 9 minus26Portugal 14 0 minus9 Canada 27 8 minus18Spain 26 8 minus9 United States 31 8 0Austria 19 4 minus7 Argentina 7 10 109Belgium 20 7 minus9 Brazil 16 8 minus29Denmark 22 2 minus9 Chile 55 8 minus2Finland 21 3 minus7 Mexico 42 12 minus8Germany 14 minus2 minus11 Uruguay 45 3 1Greece 22 minus1 minus14 Other Latin America 60 10 minus5Ireland 42 12 minus9 South Africa 36 12 minus1Netherlands 21 4 minus9 Turkey 50 8 20Sweden 24 9 minus13 North Africa 53 11 0Switzerland 18 8 minus6 Other Africa 109 18 84United Kingdom 32 6 1 Middle East 52 18 minus12Other West Europe 21 10 minus1 China 79 3 5Bulgaria 41 minus7 7 Hong Kong 42 3 2Croatia 20 minus2 minus1 India 134 13 17Georgia 35 0 23 Japan 11 minus3 minus24Hungary 25 minus3 minus11 Korea 38 1 minus9Moldova 49 minus11 13 Malaysia 62 15 minus16Romania 45 minus4 22 Philippines 75 18 7Russia 18 minus2 minus8 Singapore 44 21 minus22Ukraine 22 minus5 14 Taiwan 29 1 minus13Other East Europe 40 minus5 48 Thailand 47 3 minus9

Other Asia 99 10 10

Source Authorsrsquo compilation from projections by various international agencies and from global economy-wide modeling by Anderson andStrutt (2016)

Kym Anderson and Glyn Wittwer 247

The composite factor demand Fid is proportional to total output Qid subject to aprimary factor using technology Aid Hence

Fid frac14 Qid Aid eth6THORN

The perfectly competitive zero-pure-profit condition is that total revenue valued atthe output price P0s

i multiplied by Qid equals the total production cost

P0si Qid frac14

X

cPcid X1cid thorn

X

oWo

id Loid thorn Rid Kid eth7THORN

Household demands follow a linear expenditure system in each region We reducethe optimizing problem for household consumption of each commodity subjectto a budget constraint to equations describing subsistence and discretionarydemands Aggregate subsistence expenditure WSUBd depends only on consumerprices P3cd for each commodity and the number of households N as per-capita sub-sistence quantities XSUBcd subject to given preferences are constant

WSUBd frac14X

c

P3cd XSUBcd Nd eth8THORN

Discretionary expenditures for each commodity (the left-hand side of equation (9))are equal to the marginal budget share (βcd) of aggregate discretionary expenditureThis aggregate is the bracketed term on the right-hand side of equation (9) whereW3TOTd is aggregate nominal expenditure

P3cdethX3cd XSUBcd NdTHORN frac14 βcdethW3TOTd WSUBdTHORN eth9THORN

Because real aggregate consumption is usually exogenous in our partial equilibriumsimulations the linear expenditure system determines the consumption shares ofindividual final commodities (ie the five wine types plus a composite of all otherconsumption items) driven by changes in relative prices as faced by domestic con-sumers The income elasticity of demand for each commodity is equal to the mar-ginal budget share divided by the expenditure share This number varies from 05for nonpremium wine to 17 for superpremium still wine The income elasticity ofdemand for other consumption is very close to 10 because wine accounts for anaverage of only 03 of aggregate expenditures globally and no more than 11in any country (Anderson and Nelgen 2011 Table 166)

A feature of our revised model of world wine markets is the inclusion of nominalexchange rates These appear directly in the equation linking retail prices (P3scd ) toproducer prices by country of origin (P0s

c ) where c denotes the wine type

P3 scd frac14 P0s

cfd

fsTtarcd T

taxcd thorn Pm

cd eth10THORN

248 UK and Global Wine Markets by 2025 and Implications of Brexit

The exchange rates in the consuming and producing regions are fd and fs respec-tively expressed as local currency units per $US Ttar

cd is the power of the tariff in theconsuming region and Ttax

cd is the power of the domestic consumption (or excise) taxprior to any generic value-added or goods-and-services tax Pm

cd is the price of marginm assumed to be locally supplied nontradable and therefore unaffected by theexchange rate

A given level of consumption for wine type c (X3cd) is satisfied using theArmington (1969) assumption in which wines from different countries of originare imperfectly substitutable First domestic wine is imperfectly substitutable witha composite of imports

X3sscd frac14 f ethX3cd CESethP3sscd=P3cdTHORNTHORN ss frac14 domestic imports eth11THORN

and then imports by origin (X3scd ) are determined in a second CES equation

X3scd frac14 f ethX3ssfrac14importscd CESethP3scd=P3

ssfrac14importscd THORNTHORN eth12THORN

The model enables us to show how changes in international competitiveness affectthe worldrsquos wine markets A crucial part of this exercise is explaining how pricesdetermined outside the grape and wine markets influence these markets Becausethe model is in partial equilibrium to depict the impacts of changes in internationalcompetitiveness outside price changes need to be imposed as shocks on the modelThe price of intermediate inputs other than grapes and wine shown in equations (1)and (2) is set equal to the price of GDP (Pg

d ) multiplied by a shifter Fcd

Pcid frac14 Fc

dPgd eth13THORN

If no specific price observations are available the shifter Fcd remains exogenous and

unshocked with the change in price being determined by a shock to the price ofGDP If observations are available for specific input price movements the shifterFc

dbecomes endogenous with Pc

id now exogenous and shocked

Wid frac14 Fwd P

gd eth14THORN

Wage rates are treated similarly In equation (14) if the wage shifter Fwd is exogenous

changes in wage rates Wid are determined by changes in the price of GDP If wagerate data are available Fw

d becomes endogenous andwage rates are shocked directly

Pmcd frac14 Fm

d Pgd eth15THORN

The prices of trade and transport margins are also determined by the price of GDP ifthe shifter Fm

d in equation (15) is exogenous

Changes in international competitiveness depend on changes in relative pricelevels and changes in nominal exchange rates In equation (16) fR

s denotes real

Kym Anderson and Glyn Wittwer 249

exchange-rate movements relative to the US dollar

fRs frac14 Pg

s =frac12PgUSA

fs$ eth16THORN

In equation (16) the nominal exchange rate for the United States is alwaysunchanged because nominal and real exchange rates are expressed relative to theUS currency

Changes in international market conditions may have impacts in one direction onproducer prices as expressed in US dollars P0s

i and potentially in the opposite direc-tion in local currency units Hence we calculate real producer pricesP0s

iloc in localcurrency terms (ie the price most relevant to domestic producers)

P0siloc frac14 P0s

ifs=P

gs eth17THORN

To obtain real price changes in local currency terms we convert US dollar prices(P3 s

cd for the source-specific price and P3cd for the source-composite price) to reallocal currency prices (P3 s

cdloc and P3cdloc) using the consumer price index (CPI)(Pc

d ) as the deflator

P3 scdloc frac14 P3 s

cdfd=P

cd eth18THORN

and

P3cdloc frac14 P3cdfd=Pcd eth19THORN

References

Anderson K and Nelgen S (2011) Global Wine Markets 1961 to 2009 A StatisticalCompendium Adelaide University of Adelaide Press Also available as an e-book atwwwadelaideeduaupresstitlesglobal-wine

Anderson K Nelgen S and Pinilla V (2017) Global Wine Markets 1860 to 2016 AStatistical Compendium Adelaide University of Adelaide Press (forthcoming) Also tobe available as an e-book at wwwadelaideeduaupress

Anderson K and Pinilla V (with the assistance of A J Holmes) (2017)Annual Database ofGlobal Wine Markets 1835 to 2016 Wine Economics Research Centre University ofAdelaide to be posted in October at wwwadelaideeduauwine-econdatabases

Anderson K and Strutt A (2016) Impacts of Asiarsquos rise on African and Latin Americantrade Projections to 2030 World Economy 39(2) 172ndash194

Anderson K andWittwer G (2013) Modeling global wine markets to 2018 Exchange ratestaste changes and Chinarsquos import growth Journal of Wine Economics 8(2) 131ndash158

Anderson K and Wittwer G (2015) Asiarsquos evolving role in global wine markets ChinaEconomic Review 35 1ndash14

250 UK and Global Wine Markets by 2025 and Implications of Brexit

Armington P A (1969) A theory of demand for products distinguished by place of produc-tion IMF Staff Papers 16 159ndash178

Baldwin R E (ed) (2016) Brexit Beckons Thinking Ahead by Leading Economists LondonCentre for Economic Policy Research

Campos N F (2016) Lousy experts Looking back at the ex ante estimates of the costs ofBrexit In R E Baldwin (ed) Brexit Beckons Thinking Ahead by Leading Economists(ch 3) London Centre for Economic Policy Research

Dixon P Parmenter B Sutton J and Vincent D (1982) ORANI A Multisectoral Modelof the Australian Economy Contributions to Economic Analysis 142 Amsterdam North-Holland

European Commission (2015) Trade for All Towards a More Responsible Trade andInvestment Policy Brussels European Commission

Harrison J Horridge M Jerie M and Pearson K (2014) GEMPACK ManualMelbourne GEMPACK Software

HM Treasury (2017) Spring Budget 2017 London HM TreasuryHolmes A J and Anderson K (2017a)Annual Database of National Beverage Consumption

Volumes and Expenditures 1950 to 2015 Wine Economics Research Centre University ofAdelaide posted at wwwadelaideeduauwine-econdatabases

Holmes A J and Anderson K (2017b) Convergence in National Alcohol ConsumptionPatterns New Global Indicators Journal of Wine Economics 12(2) 117ndash148

Ludington C C (2013) The Politics of Wine in Britain A New Cultural History BasingstokeUK Palgrave Macmillan

Office of National Statistics (2017) Analysis of real earnings London Office of NationalStatistics Available at httpswwwonsgovukemploymentandlabourmarketpeopleinworkearningsandworkinghoursarticlessupplementaryanalysisofaverageweeklyearningslatest

Rollo J Borchert I Dawar K Holmes P and Winters L A (2016) The World TradeOrganisation A safety net for a post-Brexit UK trade policy Briefing Paper 1UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-1-final-1pdf

Smith A (2017) Brexit Hard truths and hard choices UKTPO blog 19 June Available athttpsblogssussexacukuktpo20170619brexit-hard-truths-and-hard-choices

Swinbank A (2017) World trade rules and the policy options for British agriculture post-Brexit Briefing Paper 7 UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-7pdf

Viner J (1950) The Customs Union Issue New York Carnegie Endowment for InternationalPeace

Wine Australia (2015) Export Market Guide European Union Adelaide Wine AustraliaWittwer G Berger N and Anderson K (2003) A model of the worldrsquos wine markets

Economic Modelling 20(3) 487ndash506

Kym Anderson and Glyn Wittwer 251

Page 2: U.K. and Global Wine Markets by 2025, and Implications of ... · U.K. and Global Wine Markets by 2025, and Implications of Brexit* Kym Andersona and Glyn Wittwerb Abstract The United

United Kingdom (although the volume is now five times what it was in the 1980s)and wine has accounted on average for just 05 of UK merchandise imports sinceWorld War I Over the past six decades though winersquos share of UK alcohol con-sumption has steadily risen from 5 to more than one-third so wine traders distrib-utors and retailers as well as consumers are concerned about Brexitrsquos potentialimpact Among wine producers and consumers outside the United KingdomBrexit is also attracting considerable attention because the United Kingdom hasalways accounted for a major share of the worldrsquos wine imports

To examine how wine markets might be affected by the United Kingdomrsquos exitfrom the European Union it is necessary to look beyond just the immediatetrade-reducing and trade-diverting effects of altering bilateral import tariffs thatare the focus of the standard comparative static economic theory of (withdrawalfrom) customs unions Because the process of exiting establishing new tradingarrangements and adjusting to altered incentives is expected to spread over manyyears and initially to slow the growth of UK incomes and to devalue the poundone needs to begin with a projection of how wine markets would have lookedwithout Brexit in several years and then show how that projected baseline mightchange under various Brexit scenarios and a replacement trade agreementbetween the United Kingdom and the EU27 We do that using a model of theworldrsquos wine markets projected to 2025

This paper begins by briefly reviewing the United Kingdomrsquos historic and presentroles in global wine markets It then summarizes what trade theory would lead us toexpect for a country leaving a customs union1 A model of the worldrsquos wine marketsis then outlined along with a description of the way in which the model projects andof how that projection can be altered to simulate the effects of Brexit and subsequentbilateral trade agreements on the United Kingdom and other countries The modelrsquosresulting prospective changes to grape and wine markets by 2025 for a baseline caseare then summarized followed by results for a range of alternative adjustments fol-lowing Brexit The final section draws out implications of the findings for wine

1As of mid-2017 it was not yet certain whether the United Kingdom would leave the single market oralternately form a UKndashEU27 customs union but both options would require the United Kingdom toretain the European Unionrsquos tariff policy and the former would also require the United Kingdom to con-tinue to allow freedom of movement of labor and to remain under the European Court of Justice ndash none ofwhich Brexiteers want We therefore ignore these two possibilities and following Rollo et al (2016) andSmith (2017) assume that the United Kingdom will commit to the current EU tariff schedule at theWorld Trade Organization in the first instance and then seek a free trade agreement (FTA) with theEU27 Presumably other trading partners would want to wait and see what that FTA looked likebefore signing bilateral FTAs of their own with the United Kingdom Deep FTAs normally takeseveral years to negotiate and get ratified by the relevant parliaments even when a sufficient number ofexperienced negotiators is on each side of the table At present the United Kingdom has very few suchnegotiators having delegated that task to the European Commission (EC) in Brussels during the past45 years of its EU membership

222 UK and Global Wine Markets by 2025 and Implications of Brexit

markets and their participants in the United Kingdom and abroad within andoutside the European Union

II Historical Backdrop Wine in the United Kingdom and the UnitedKingdom in Global Wine Markets

Wine has rarely accounted for more than 2 of UK merchandise imports and sinceWorld War I it has averaged just 05 Also prior to 1950 wine accounted for lessthan 4 of the volume of alcohol consumed in the United Kingdom (except brieflyin the 1920s) and not much more as a share of alcohol expenditure Over the past sixdecades though winersquos share of UK alcohol consumption has steadily risen and itnow exceeds one-third in volume and value terms (Figure 1)

The United Kingdom is a key player in wine-trade circles as it has alwaysaccounted for a major ndash and often the largest ndash share of the worldrsquos wine importsPrior to 1960 the United Kingdomrsquos share of the value of world imports of wineroughly matched its share for all merchandise more than 20 in the 19th centurybut steadily declining to around 10 by 1960 Since then the two shares havediverged with the all-goods share falling to around 4 while the wine shareclimbed back above 20 by the beginning of the twenty-first century The UnitedKingdomrsquos share of the volume of global wine imports has always been below itsvalue share but those two shares have converged since 1960 and are now botharound 14 (Figure 2)

These two trends are summarized in Figure 3 The ratio of the United Kingdomrsquosshares of world imports of wine to that of all goods rose from 1 to 4 between 1960and 2000 and the ratio of the United Kingdomrsquos average import price to the worldaverage has come down from between 3 and 4 pre-1960 (when the United Kingdomwas mostly importing relatively expensive wines from Bordeaux and Champagne) toabout 12 by 1980 and is now close to 1

Associated with the change in the average quality of UK wine imports are dra-matic changes in the importance of different wine-exporting countries to UKimports and in UK shares in the wine exports of those countries

Those recent shares are very different from what they were in 1995 when the NewWorld was just beginning to expand its wine exports But such changes are notunprecedented Indeed they changed considerably not only because of the changingglobal shares of the various wine-exporting countries but also because of changes inthe United Kingdomrsquos preferential trading arrangements such as the 1703 MethuenTreaty with Portugal the 1860 Cobden-Chevalier Treaty with France and the 1932Ottawa Agreement with Commonwealth countries (Tables 1 and 2)

In the middle rows of Table 1 the 2010ndash2014 shares of exporters in the interna-tional market are shown below their shares in the UK markets For some countries

Kym Anderson and Glyn Wittwer 223

those two sets of shares are similar (France Italy Germany Chile) while for othersthey are very different The latter is most noticeable for the former colonies ofAustralia New Zealand and South Africa whose shares in the United Kingdomare more than twice their shares in the rest of the world Both shares for theEU27 exceed two-thirds Many commentators expect the EU27 shares to dropand shares of eg Australia and New Zealand to rise as a consequence of Brexit

The United Kingdomrsquos recent importance to producers in wine-exporting coun-tries is clear from Figure 4 For seven key suppliers the United Kingdom accountedin 2010ndash2014 for more than one-sixth of their wine-export earnings and for three ofthem (Australia the United States and New Zealand) the United Kingdom was amarket for more than one-third of their volume of wine exports

Recent import duties and other taxes affecting the consumer prices of alcohol inthe United Kingdom are summarized in Table 3 expressed per liter of beverageImport and excise duties on wines vary according to their alcohol content Aboutone-third of UK wine imports arrive in bulk perhaps half of which contain lessthan 13 alcohol and one-tenth of imports are sparkling so the volume-weightedaverage import duty is 13 pence per liter This figure contrasts with the volume-weighted average excise tax on wine which is 297 pence per liter To that increasein the wholesale price is added perhaps a 25 retail margin for off-trade sales andwell over a 100 margin for many restaurant sales before the 20 value-addedtax (VAT) is added So the average import price in 2013ndash2015 of 222 pence per

Figure 1

Winersquos Shares of UK Merchandise Import Value and of Volume and Value of UK AlcoholConsumptiona 1800 to 2015 ()

Fig1-Colou

ron

line

BW

inprint

a Reliable wine consumption volume data are not available for the 1930s and 1940s nor are value of alcohol consumption data pre-1955Sources Compiled from data in Anderson and Pinilla (2017) and Holmes and Anderson (2017a b)

224 UK and Global Wine Markets by 2025 and Implications of Brexit

liter is escalated to 800 pence for off-trade sales and more than 1300 pence for on-trade sales (which are about one-fifth of the total sales volume in the UnitedKingdom) The share of that latter retail price that is due to tariffs on wineimports is thus just 1 The change in the United Kingdomrsquos import trade regimefrom imposing such tariffs on wines currently imported free of duty from theEuropean Union Chile and South Africa is therefore likely to be very minorHowever as becomes clear below the effect of Brexit on wine sales involves farmore than just the trade-reducing and trade-diverting effects of altering bilateralimport tariffs

III The Economics of Leaving a Customs Union

The standard theory of customs unions (Viner 1950) focuses on the fact that whencountries join a union and impose a common external tariff on imports from non-union countries net trade can be created (depending on the height of the commonexternal tariff relative to the previous national tariffs) but trade can also be diver-sified (because of the preference to producers within the union) When a country

Figure 2

UK Shares of Value of World Merchandise Imports and of Value and Volume of World WineImports 1850 to 2015 ( 3-Year Averages to Year Shown)

Fig2-Colou

ron

line

BW

inprint

Source Compiled from data in Anderson and Pinilla (2017)

Kym Anderson and Glyn Wittwer 225

leaves a union the reverse happens because the leaving countryrsquos tariffs now apply toits imports from union countries as well as from the rest of the world Hence importsfrom union countries fall because other things equal these countriesrsquo preferentialaccess to the leaving country no longer applies That is also the case for countriesthat enjoyed an FTA with the union Of significance to wine markets are theEuropean Unionrsquos FTAs with Chile and South Africa

The size of that trade-diverting impact of leaving the union on wine depends onthe external tariff imposed on wine imports not only by the union but also by theleaving country Some have suggested that the United Kingdom should becomethe Hong Kong of Europe and move immediately to free trade on all productsOthers have suggested that this move would impose huge structural changes onthe UK economy that society would not tolerate ndash at least not without majorcompensation packages But all agree that a new trade policy that sets most-favored-nation (MFN) tariff rates is needed before the United Kingdom can beginto negotiate new preferential trading arrangements with the EU27 its FTA partnerssuch as Chile and South Africa or other countries Rollo Borchert Dawar Holmesand Winters (2016) suggest that the most practical trade policy for the UnitedKingdom to adopt at the outset is the European Unionrsquos tariff schedules previouslyagreed to at the WTO In all but one of the Brexit scenarios examined below we

Figure 3

UK Price Relative to World Price of Wine Imports and UK Wine-Import Intensitya 1950 to2015 ()

Fig3-Colou

ron

line

BW

inprint

a Import intensity is defined as the United Kingdomrsquos share of the value of global wine imports divided by the United Kingdomrsquos share of theworldrsquos total merchandise importsSource Compiled from data in Anderson and Pinilla (2017)

226 UK and Global Wine Markets by 2025 and Implications of Brexit

Table1

Shares

ofUKW

ineIm

portsfrom

TodayrsquosKey

Wine-Exp

ortin

gCountries1

675to

2014

andProjected

to2025

with

outandwith

Brexit()

Volum

eFrance

Spain

Portugal

Italy

Germany

South

Africa

Australia

United

States

Chile

New

Zealand

Argentin

aOther

countries

Total

1675ndash1696

2542

231

90

00

00

00

100

1697ndash1862

526

491

32

00

00

014

100

1863ndash1919

2626

220

00

20

00

025

100

1920ndash1940

1218

320

05

140

00

020

100

1995

volume

3210

318

143

73

21

07

100

value

439

413

102

83

21

05

100

2010ndash2014

volume

159

117

48

2110

84

12

100

value

358

215

44

115

66

12

100

Exp

ortersrsquoshareof

world

wineexports2010ndash2014

volume

1520

322

44

74

72

39

100

value

309

319

42

64

53

312

100

2025

projected

noBrexit

100

volume

188

188

26

207

45

45

65

43

64

17

20

92

100

value

323

100

30

218

44

23

58

44

54

41

20

45

100

2025

projected

with

Brexit

volume

190

188

27

208

43

44

66

42

63

17

19

93

100

value

311

96

29

210

41

24

59

44

55

41

20

70

100

SourcesCom

piledfrom

data

inAnd

ersonan

dPinilla

(201

7)to

1940

UnitedNations

COMTRADEh

ttpscomtrad

eun

orgdata

for19

95ndash2

014

andau

thorsrsquomod

elresults

for20

25

Kym Anderson and Glyn Wittwer 227

assume that Rollo et alrsquos suggestion will be the new UK trade policy commitmentto WTO members that subsequent negotiations for preferential arrangements willtake years and that any consequent agreements with the European Union andothers will be gradually implemented after that

The impact of leaving a customs union on wine markets does not only come fromtariff changes Also relevant are any effects of leaving on real UK incomes and thevalue of the British pound If the United Kingdom were to move immediately to freetrade on all products (the Hong Kong option) its per-capita income could eventuallyrise but only after considerable adjustment Should the United Kingdom insteadcommit to the current EU tariff schedule at the WTO in the first instance as weassume below then its per-capita income growth rate and the British poundrsquosexchange rates almost certainly will be lower for some time ndash at least until newtrade agreement negotiations with the EU27 and others are sufficiently advanced

Table 2Taxes on British Wine Imports by Source 1660ndash1862 (British Pounds per Kiloliter)

France Germany Spain Portugal South Africa

1660ndash1665 7 9 8 81666ndash1684 7 9 8 81685ndash1691 14 20 19 181692ndash1695 22 20 19 181696 47 20 19 181697ndash1702 51 25 23 221703 52 27 24 231704ndash1744 55 31 26 251745ndash1762 63 35 30 291763ndash1777 71 39 34 331778 79 43 38 371779 84 41 40 391780ndash1781 92 49 44 431782ndash1785 96 51 47 46 441786 65 51 37 37 371787ndash1794 47 51 32 32 371795 78 64 51 51 571796ndash1797 108 92 71 71 771798 111 96 73 73 791799ndash1801 107 92 71 71 771802 112 97 74 74 801803 131 109 87 87 871804 142 117 95 95 951805ndash1824 144 119 96 96 961825ndash1830 78 50 50 50 251831ndash1859 58 58 58 58 291860 32 32 32 32 321861 16 21 21 21 211862 11 26 26 26 26

Source Summarized from Ludington (2013 Table A1)

228 UK and Global Wine Markets by 2025 and Implications of Brexit

Figure 4

UK Shares in Wine Exports of Key Wine-Exporting Countries 2010ndash2014 ()

Fig4-BW

onlin

eBW

inprint

Source Compiled from data in Anderson and Pinilla (2017)

Table 3Import Duties Excise Duties and VAT Affecting Consumer Prices of Wine and Other Alcohol

in the United Kingdom April 1 2017

VAT 20

MFN import duties on wine pound per liter-bottled still wine lt13 alc 0114-bottled still wine 13ndash15 alc 0134-bottled still wine 15ndash18 alc 0162-bulk still wine lt13 alc 0086-bulk still wine 13ndash15 alc 0105-bulk still wine 15ndash18 alc 0134-sparkling wine 0278Weighted averagea 0130

Excise duties on alcohol pound per liter-still wine lt15 alc 2887-still wine 15ndash22 alc 3848-sparkling wine 55ndash85 alc 2795-sparkling wine 85ndash15 alc 3697-spirits (assumed 40 alc) 11551-beer (assumed 5 alc) 0954

VAT on alcohol 20

a Assumes that one-third of UK wine imports arrives in bulk that half has less than 13 alcohol and that one-tenth of imports is sparkling

Sources HM Revenue httpswwwgovukgovernmentpublicationsalcohol-duty-rate-changes for excise duties and VAT accessed April 92017 and Wine Australia (2015) for import duties converted at the 2016 average exchange rate of pound0740634 per euro

Kym Anderson and Glyn Wittwer 229

as to restore investor and consumer confidence in the United Kingdomrsquos economyThe size of those impacts is uncertain of course (see Baldwin 2016 and especiallyCampos 2016) so we consider a range of possibilities below

Those assumed adverse macroeconomic effects will add to the initial impact ofaltered wine tariffs on aggregate wine consumption in the United Kingdom andhence on its bilateral trades in wine They will make the loss of sales to theUnited Kingdom by EU (and Chilean and South African) suppliers greater thanwould otherwise be the case And they will reduce the likelihood that other countriesrsquosales of wine in the United Kingdom will be higher than in the baseline Indeed themacroeconomic effects could outweigh the trade-diverting effects so that even coun-tries that are currently discriminated against by the EU28rsquos wine-trade policy may beworse off because of Brexit

IV Global Wine Markets Model and Database

Our model of the worldrsquos wine markets first published by Wittwer Berger andAnderson (2003) and revised by Anderson and Wittwer (2013) is summarized inthe Appendix It disaggregates wine markets into four types namely nonpremiumcommercial-premium and superpremium still wines and sparkling wines2 There aretwo types of grapes premium and nonpremium Nonpremium wine uses nonpre-mium grapes exclusively superpremium wines use premium grapes exclusivelyand commercial-premium and sparkling wines use both types of grapes to varyingextents across countries The world is divided into 44 individual nations and 7 com-posite geographic regions that capture all other countries

The modelrsquos database is calibrated to 2014 based on the comprehensive wine-market volume and value data and trade and excise-tax data provided inAnderson and Pinilla (2017) and in Anderson Nelgen and Pinilla (2017) It is pro-jected assuming that aggregate national consumption population and realexchange rates change between 2014 and 2025 to the extent shown in AppendixTable 13 The Brexit alternatives to that baseline also are projected to 2025

Concerning preferences we assume a continued considerable swing toward allwine types in China and a swing away from nonpremium wines in all other countriesuntil 2025

In our baseline scenario grape- and wine-industry total-factor productivity areassumed to grow at 1 per year everywhere while grape- and wine-industry

2Commercial-premium still wines are defined byAnderson Nelgen and Pinilla (2017) to cost between US$250 and US$750 per liter pretax at a countryrsquos border or wholesale3The real exchange rate changes over the projection period are the changes expected in the nominal valueof country irsquos currency relative to the US dollar times the expected ratio of the GDP deflator for theUnited States versus that for country i

230 UK and Global Wine Markets by 2025 and Implications of Brexit

capital is assumed to grow net of depreciation at 15 per year in China but zero else-where (consistent with the almost-zero growth in global wine production and con-sumption over the past two decades)

Two alternative scenarios are considered (ldquolargerdquo and ldquosmallrdquo) for the initialimpact of Brexit to provide a range of results We assume that because of theUnited Kingdomrsquos decision to exit the European Union the rate of UK realgross domestic product (GDP) growth is only one-third or two-thirds as fast overthe projection period (09 or 18 per year instead of 26) and the Britishpound will be 20 or 10 lower in real terms than in our modelrsquos core baselineprojection4 In both alternative scenarios we assume that the United Kingdomapplies the European Unionrsquos external tariffs on wine from March 30 2019 atthe end of the two-year period following the United Kingdomrsquos formal triggeringof Article 50 (see Table 3)

We then consider a subsequent impact of Brexit presumed to result from negoti-ating signing implementing and responding to an FTA with the EU27 by 2025(Chile and South Africa currently have preferential access to EU wine marketsbut we continue to assume in this scenario that the United Kingdom does not imple-ment new bilateral FTAs with them or others in our time frame)5 This subsequentscenario assumes the British pound returns to what it would have been in 2025 in theabsence of Brexit and that real incomes regain two-thirds of the value they wouldhave had without Brexit as compared with our ldquosmallrdquo initial Brexit scenario

This global model has supply-and-demand equations and hence quantitiesprices and price elasticities for each of the grape and wine products and for asingle composite of all other products in each country Income elasticities ofdemand also exist for each final product Grapes are assumed to not be traded inter-nationally but other products are exported and imported Each market is assumed tohave been in equilibrium before any shock and to find a new market-clearingoutcome following any exogenously introduced shock

4The nominal price of the British pound in US dollars in the fortnight following the Brexit vote on June23 2016 dropped 13 to US$130 and a year later the pound sat at the same rate having dropped to anaverage of $124 (an 18 devaluation) between October 2016 and April 2017 Our choice of a low of 10and a high of 20 aims to capture future possible rates while uncertainties remain The average real wagein the United Kingdom fell in the first half of 2017 and projected real GDP growth during 2018ndash2020 hasbeen revised down to 17 in the United Kingdomrsquos latest budget (HM Treasury 2017)5South Africa currently has duty-free access to the EU28 for just 50 megaliters (ML) of wine per yearbeyond which the MFN tariff rate applies South Africa currently exports around 320 ML to theEuropean Union one-third of which initially goes to the United Kingdom How that quota of 50 MLis divided between the United Kingdom and the EU27 is subject to future negotiation (Rollo et al2016 Swinbank 2017) In the following discussion we assume none of it is accepted by the UnitedKingdom so that all South African wine imports pay the United Kingdomrsquos MFN tariff under Brexitbut altering that assumption makes very little difference to our results for any countryrsquos trade exceptSouth Africarsquos

Kym Anderson and Glyn Wittwer 231

V Projecting Global Wine Markets to 2025

Global wine production and exports are projected in the baseline from 2014 to 2025consistent with past trends The modelrsquos global volume of production (and consump-tion) rises little over that 11-year period (9) made up of a 6 decline in nonpre-mium wine and a one-sixth rise in commercial-premium and superpremium wineIn real (2014 US$) value though global wine output and consumption increaseby about 50 in total and 60 in the two premium categories The internationaltrade projections are similar although a little larger with the share of global wineproduction exported (equivalent to the share of global consumption imported)rising 2 percentage points between 2014 and 2025

The baseline projection does not greatly alter the 2014 shares of various countriesin global wine production apart from China because we assume vineyard expansionthere is faster than elsewhere6 In value terms this assumption means that Chinamoves from fifth to fourth by 2025 behind France the United States and ItalySpain remains barely ahead of Australia and they and Germany take the nextthree places (Figure 5(a)) In total wine-production volume terms China movesfrom sixth to fifth place and Argentina drops from fifth to eighth (and fromeighth to ninth in value terms)

When their products are subdivided into fine wines (superpremium still plus spark-ing) commercial-premium wines and nonpremium wines France and the UnitedStates retain the highest two places on the global ladder for fine-wine productionand Spain and Italy retain the top two places for nonpremium wine As for commer-cial-premium wine production (defined to cost between US$250 and US$750 perliter pretax at a countryrsquos wholesale level or national border) Italy retains the topranking over our projections period but at least in terms of value China challengesFrance for second place

The country rankings by projected value of total wine consumption change some-what more than those for production by 2025 with China taking second place afterthe United States ahead of France and Germany and then the United Kingdomslightly overtaking Italy to slip into fifth place (Figure 5(b)) The United StatesFrance and Germany retain the top three rankings for consuming fine wine butCanada slightly overtakes Italy for fourth place in terms of value at least In thecase of commercial-wine consumption China strengthens its number-one positionahead of the United States and the United Kingdom does likewise vis-agrave-visGermany for third place

As for the projected changes in consumption volumes China is projected to dom-inate the increase in aggregate although the United States is projected to lead the

6 In fact Chinarsquos wine production fell steadily between 2012 and 2016 by a total of one-sixth so Chinarsquoswine imports may grow faster in practice than in this baseline projection

232 UK and Global Wine Markets by 2025 and Implications of Brexit

Figure 5

Value of Wine Production and Consumption in Key Countries 2014 and Projected Baseline for2025 (2014 US$ Million at WineryWholesale Pretax Prices)

Fig5-BW

onlin

eBW

inprint

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 233

increase in consumption of fine wine In Western Europe and in the SouthernHemispherersquos New World countries fine wines are projected to substitute for com-mercial wines (defined as the sum of commercial-premium and nonpremiumwines) with almost no change in total wine consumption Sub-Saharan Africa isthe next region that is projected to take off with its growth accounting for morethan one-third of the rest of the worldrsquos increase in volume consumed

Those differences in production versus consumption rankings are reflected ininternational trade Figure 6 shows that France Italy and Spain remain the threedominant exporters of wine in aggregate value but the rankings of the next fewchange with Australia being slightly ahead of Chile and the United StatesGermany and New Zealand being nearly tied for sixth place in value termsFrance and Italy are even more dominant in fine-wine exports and remain so by2025 while Italy outranks France in the commercial-premium export categoryand Spain outranks Italy Australia and Chile in the nonpremium export class

Among the importers the United States and the United Kingdom are projected tocontinue to hold the first two places in 2025 in value terms but China moves intothird place slightly ahead of Germany followed well behind by Canada HongKong Belgium-Luxembourg the Netherlands and Japan (Figure 6(b)) OtherAfrica (excluding South Africa) is projected to experience the largest increase inimports among all the other regions followed by Other Asia which becomes asbig as Germany in value terms (Figure 7(a)) In terms of total volume of wineimports Germany and the United Kingdom held the top two shares in 2014 butby 2025 the United Kingdom is projected to be well ahead of Germany (Figure 7(b)) However this projection ignores the effects of Brexit to which we now turn

VI How Might Wine Markets Be Affected by the United Kingdomrsquos Exitfrom the European Union

As mentioned earlier for our two alternative scenarios to capture the initial effects ofBrexit (ldquolargerdquo and ldquosmallrdquo) we assume that following the United Kingdomrsquos exitfrom the European Union the United Kingdomrsquos rate of economic growth wouldbe only one-third or two-thirds as fast for the period to 2025 the British poundwould be 20 or 10 lower in real terms than in our modelrsquos baseline projectionand the United Kingdom would apply the European Unionrsquos external tariff onwine to imports from EU member countries (as part of establishing MFN ratesvia the WTO to then start new bilateral FTA negotiations) In these initial scenarioswe assume the United Kingdom does not implement any new FTAs particularlywith the EU27 Chile and South Africa Following the discussion of those initialresults we present the results of a subsequent scenario that includes the implemen-tation of an FTA between the United Kingdom and the EU27 We assume these aretwo discrete steps with no agreed arrangements to smooth the transition betweenthem

234 UK and Global Wine Markets by 2025 and Implications of Brexit

Figure 6

Value of Wine Exports and Imports Key Wine-Trading Countries 2014 and Projected Baselinefor 2025 (2014 US$ million)

Fig6-BW

onlin

eBW

inprint

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 235

Figure 7

National Shares of Global Wine Import Value and Volume 2014 and Projected Baselinefor 2025 ()

Fig7-BW

onlin

eBW

inprint

Source Authorsrsquo model results

236 UK and Global Wine Markets by 2025 and Implications of Brexit

A Initial Impact of Brexit

Because the British pound dropped by one-sixth against the US dollar in the fourmonths following the Brexit vote in June 2016 and because the UnitedKingdomrsquos average real wage fell 1 between November 2016 and April 2017(according to the latest available data see Office of National Statistics 2017) weviewed our ldquolargerdquo scenario as more likely soon after the Brexit vote But wecompare those results with ones from our ldquosmallrdquo initial-impact scenario becauseit has become more likely following the June 2017 general election that ldquosofterrdquooptions will be considered by the new government Generally the results areabout half the size in the ldquosmallrdquo scenario with the exception of the bilateraltrade effects To show the sensitivity of results to our assumptions we point outthe differences when the ldquosmallrdquo results are not close to half the results shown forthe ldquolargerdquo scenario

Before turning to those results we first present a ldquofree-traderdquo scenario whichassumes the United Kingdom chooses the radical option of becoming the HongKong of Europe Unlikely as this scenario is the results provide assurance thatour global wine model generates the standard types of effects of exiting a customsunion In this ldquofree-traderdquo scenario the United Kingdom is assumed to move tozero tariffs on all wine imports Thus the current preferences on EU Chileanand South African wine imports disappear as all other wine exporters enjoy thesame free access to UK wine markets We assume in this scenario that Brexit hasno impact on the United Kingdomrsquos real income or consumption growth rates by2025 and that by then the British pound has returned to its 2014 value relative toother currencies (similar to the real exchange rate assumed in our baseline scenarioto 2025 see Appendix Table A1)7 Under these assumptions EU Chilean andSouth African wine exports to the United Kingdom in 2025 are 003 lower thanin the no-Brexit baseline and UK imports from other wine exporters are 025higher but the United Kingdomrsquos total wine imports are hardly any different(because local wine prices fall by only 05 and incomes are unchanged) In this sce-nario UK consumers expand their volume of consumption by just 05 while theEU27 exports more and other countries export less to the rest of the world such thatworld wine exports are almost unchanged In short all these effects are in the direc-tion that comparative static-customs-union theory would predict but they are smallbecause we assume the United Kingdomrsquos real income and currency are the same asin our baseline for 2025 as described in the previous section

In the ldquolargerdquo alternate initial scenario involving Brexit as compared with thebaseline scenario to 2025 the consumer price of wine in 2025 is 22 higher in the

7If this scenario were seen as a credible long-run trade policy it would encourage more investment thatwould eventually raise UK incomes but not without major structural changes that would initiallydisrupt the economy Implicitly we assume in our free-trade scenario that these two opposite impactson real UK incomes and the British pound exactly offset each other by 2025

Kym Anderson and Glyn Wittwer 237

United Kingdom in local currency terms (20 because of real depreciation of theBritish pound 4 because of new tariffs on EU Chilean and South Africanwines and ndash2 because of slower UK income growth) The volume of UKwine consumption is 28 lower 16 because of slower UK economic growth7 because of real depreciation of the British pound and 5 because of newtariffs Superpremium still-wine sales are the most affected dropping by two-fifths while sparkling and commercial-premium wines drop a bit less than one-quarter Because the average price rises by more than the decline in volume soldthe aggregate value of UK sales even in local currency terms falls under thisldquolargerdquo Brexit scenario Under the ldquosmallrdquo Brexit scenario the consumer price ofwine in 2025 is 11 higher in the United Kingdom and its volume of wine consump-tion is 17 lower

The volume of projected UK imports in 2025 is 427 ML or nearly one-quarterless in the ldquolargerdquo scenario than in the baseline scenario comprising 58ML less spar-kling wine 31 ML less superpremium still wine and 339 ML less commercial-premium wine World imports are lower by just 239 ML because imports byother countries are 189 ML higher in response to the lower international wineprices in this scenario In value terms UK imports are $175 billion (or 27)lower in 2025 because of ldquolargerdquo Brexit $113 billion because of lower incomes$038 billion because of the fall in the British pound and $014 billion because ofthe rise in wine-import tariffs (Table 4) These aggregate trade impacts are a littlemore than half this size under the ldquosmallrdquo Brexit scenario

Despite the import levels falling because of raised import tariffs domestic con-sumption of all three quality categories of UK-produced wine is lower with thanwithout Brexit because of shrunken demand for all wines resulting from loweredUK incomes and raised local prices because of devaluation of the British poundThe British poundrsquos devaluation does make it easier for the United Kingdom tosell wines abroad though The countryrsquos exports are 7 ML or nearly 5 higherin 2025 in the ldquolargerdquo Brexit scenario and UK production is 3 higher ThoseUK exports (or re-exports of imported bulk wine after it is bottled in the UnitedKingdom) that go to EU27 countries are reduced though because of the tariffnow imposed at the new EU border

Without Brexit the United Kingdomrsquos shares of global wine imports are slightlyhigher in volume terms in 2025 than in 2010ndash2015 but 2 percentage points lower invalue terms thanks to East Asiarsquos expanding demand for imports of premium winesWith ldquolargerdquo Brexit however that value share is an additional 2 percentage pointslower and the volume share is almost 5 points lower (Figure 8) The net effects ofthese impacts on global trade are shown in Figure 9 Most of the initial tradeeffect of Brexit is a large decline in net imports of wine by the United Kingdomwith very little offsetting positive effect on trade in the rest of the world Theldquosmallrdquo Brexit numbers are a bit more than half of those for ldquolargerdquo Brexit

238 UK and Global Wine Markets by 2025 and Implications of Brexit

Table4

Difference

in2025

Projected

VolumeandValue

ofWineIm

portsb

ytheUnitedKingdom

andtheResto

fthe

World

asaConsequence

oftheInitial

BrexitS

hock

(MLand2014US$

Million

ldquoLargerdquoSc

enario)

Volum

e(M

L)

Value

(US$

million)

NP+CPa

SuperPrb

Sparkling

TOTAL

NP+CPa

SuperPrb

Sparkling

TOTAL

ΔUK

impo

rtsdu

eto

Low

erincomes

minus198

minus20

minus29

minus247

58minus644

minus253

minus234

minus1131

65Low

erpo

und

minus70

minus10

minus14

minus93

22minus248

minus127

minus102

minus476

27Highertariffs

minus71

minus1minus16

minus87

20minus110

minus8minus2

4minus143

8TOTA

Lminus339

minus31

minus58

minus427

100

minus1001

minus388

minus360

minus1750

100

difffrom

base

2332

3325

2432

3227

of

totalc

uts

797

14100

5722

21100

ΔROW

netim

ports

143

2125

189

230

143

minus181

192

ΔWORLD

TRADE

minus195

minus10

minus34

minus239

minus763

minus246

minus543

minus1552

aNon

prem

ium

plus

commercial-premium

still

wines

bSu

perpremium

still

wines

SourceA

utho

rsrsquomod

elresults

Kym Anderson and Glyn Wittwer 239

The aggregate effect of ldquolargerdquo Brexit on the market shares of various wine-export-ing countries in the United Kingdom is almost indiscernible even with one decimalpoint The projected 2025 shares are quite different from the actual 2014 shares forseveral countries They are much smaller in 2025 for South Africa Australia andNew Zealand (and the United States in volume terms) and are much larger involume for Spain and in value for Italy These results occur because wine-exportingcountries benefit differentially from the varying rates of growth in net importdemand for wine in non-UK countries over this projection period The most impor-tant projected changes are the increase in the real value of annual wine importsbetween 2014 and 2025 by China (200 or $3 billion) Other Asia (110 or $22billion) and Africa (270 or $16 billion) More than half of Australiarsquos increasein annual exports from 2014 to 2025 go to Asia and more than half of SouthAfricarsquos increase in exports go to Other Africa

Table 5 reveals that ldquolargerdquo Brexit lowers EU Chilean and South African wineexports by 150 ML or US$12 billion in the case of the European Union with someof their exports diverted from the United Kingdom to the EU27 and other marketsin competition with New World exporters Although the United States Australiaand Argentina sell only a little less to the United Kingdom they also sell less toother countries For Chile and South Africa which lose their preferential access toUK (but not to EU27) markets in this Brexit scenario some exports are redirected

Figure 8

UK Shares of World Wine Imports 2010ndash2015 and Projected to 2025 without and with theldquoLargerdquo Initial Brexit Shock ()

Fig8-BW

onlin

eBW

inprint

Sources Anderson and Pinilla (2017) and authorsrsquo model results

240 UK and Global Wine Markets by 2025 and Implications of Brexit

from the United Kingdom to EU27 countries ndash but again they export less overallGlobal wine trade in 2025 is 240 ML (19) or $18 billion (35) less under thisldquolargerdquo Brexit scenario The percentage by which wine exportersrsquo trade shrinks isgreater for values than for volumes because of changes in relative prices of differ-ent-quality wines Those differences are shown in the numbers in parentheses inTable 5

Three other points are worth making about Table 5 One is that Australia sellsslightly more to the United Kingdom in the ldquosmallrdquo Brexit scenario rather thanslightly less as in the ldquolargerdquo Brexit case Evidently the negative income and price(devaluation) effects do not more than offset the positive trade-diverting effect onAustralian exports to the United Kingdom of removing preferences in the ldquosmallrdquoscenario Second New Zealand sells slightly more to non-UK countries underBrexit despite greater competition from the EU27 Chile and South Africa Thisanomaly is due to changes in the relative prices of different qualities of wine inglobal wine markets bearing in mind that New Zealand has the worldrsquos highestaverage price for still-wine exports And third the value (but not the volume) ofexports of ldquoOtherrdquo countries to markets other than the United Kingdom arehigher under Brexit This result too is due to changes in the relative prices of differ-ent qualities of wine in global wine markets

Figure 9

Difference in 2025 Wine Import Volumes and Values as a Result of the ldquoLargerdquo Initial BrexitShock (ML and US$ Million in 2014 US dollars)

Fig9-BW

onlin

eBW

inprint

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 241

B Subsequent Impact of Brexit from a UK-EU27 FTA

The next-most-likely step in the Brexit process is for the United Kingdom to nego-tiate a new trade arrangement with the EU27 We therefore assume that a UK-EU27FTAwith free bilateral wine trade is implemented and adjusted to by 2025 and thatprogress toward that end occurs soon enough that the adverse macroeconomicshocks from the initial impact of uncertainty over the Brexit process are confinedto those assumed in the ldquosmallrdquo scenario outlined above In this subsequent scenariowe assume that the British pound returns to the value it would have reached in 2025in the absence of Brexit (ie reversing the 10 devaluation assumed in the ldquosmallrdquoinitial Brexit scenario) and that real incomes in the United Kingdom are 8 higher

Table 5Difference in 2025 BilateralWine Import Volumes and Values fromKey Exporters by the United

Kingdom and the Rest of the World (RoW) as a Result of Initial Brexit Shock(ML and 2014US$ Million)a

(a) ldquolargerdquo scenario

Volume (ML) Value (2014US$ million)

UK RoW WORLD () UK RoW WORLD ()

EU27 minus287 136 minus150 (ndash17) minus1187 minus5 minus1192 (ndash31)Chile minus59 35 minus25 (ndash30) minus169 31 minus138 (ndash48)South Africa minus53 35 minus18 (ndash32) minus105 20 minus85 (ndash67)USA minus7 minus6 minus13 (ndash24) minus75 minus40 minus115 (ndash50)Australia minus4 minus3 minus7 (ndash09) minus25 minus65 minus90 (ndash30)Argentina minus3 minus9 minus12 (ndash48) minus16 minus39 minus55 (ndash52)New Zealand minus11 9 minus2 (ndash09) minus162 71 minus91 (ndash43)Others minus2 minus10 minus12 (ndash02) minus11 minus52 minus63 (ndash44)WORLD minus427 187 minus240 (ndash19) minus1750 minus79 minus1829 (ndash35)

(a) ldquosmallrdquo scenario

Volume (ML) Value (2014US$ million)

UK RoW WORLD () UK RoW WORLD ()

EU27 minus178 82 minus96 (ndash12) minus692 minus43 minus736 (ndash19)Chile minus46 28 minus18 (ndash24) minus128 36 minus91 (ndash32)South Africa minus43 29 minus14 (ndash42) minus82 23 minus59 (ndash47)USA 1 minus6 minus5 (ndash11) minus23 minus28 minus51 (ndash22)Australia 5 minus10 minus5 (ndash06) 19 minus56 minus38 (ndash13)Argentina 0 minus6 minus6 (ndash26) minus3 minus25 minus29 (ndash27)New Zealand minus5 4 minus1 (ndash06) minus80 34 minus46 (ndash22)Others 0 minus9 minus9 (ndash01) minus1 minus33 minus34 (ndash24)WORLD minus266 112 minus154 (ndash13) minus991 minus92 minus1083 (ndash21)

a Numbers in parentheses are the percentage difference between the Brexit and baseline scenarios for 2025 projected wine-import volumes orvalues by source

Source Authorsrsquo model results

242 UK and Global Wine Markets by 2025 and Implications of Brexit

in 2025 than in the ldquosmallrdquo scenario ndash that is they regain two-thirds of the value theywould have reached without Brexit-related uncertainty

This subsequent development in the Brexit process reverses most of but not allthe initial effects of Brexit by 2025 because of our assumption that the lost growthin the initial years of uncertainty following the Brexit vote are only partly recoveredby 2025 following the implementation of a UK-EU27 FTA Moreover the longer ittakes before this FTA is finalized and implemented the longer the estimated initialadverse effects persist and the larger the cumulative cost of Brexit to UK wine con-sumers and to grape and wine producers in wine-exporting countries

Table 6 summarizes the subsequent trade effects for 2025 It suggests that all butone-tenth of the loss in value of world trade in wine from the initial ldquosmallrdquo impact isrestored most of it because of a smaller reduction in the United Kingdomrsquos wineimports following the FTA Most of that improved outcome is because of recoveredimports from the EU27 commensurate with the latterrsquos high share of UK importsEven though Chile and South Africa are assumed in this scenario to have not yetsigned an FTA with the United Kingdom they export slightly more to the UnitedKingdom (and even more to all other countries) than in the initial ldquosmallrdquo scenario

Within the United Kingdom this FTA brings down the local currency consumerprice of wine by 9 largely offsetting the 11 rise in the initial ldquosmallrdquo Brexit sce-nario and it raises the volume consumed in the United Kingdom by 18 fully off-setting the 17 fall in the initial ldquosmallrdquo Brexit scenario

Table 6Difference in 2025 Bilateral Wine-Import Volumes and Values from Key Exporters by the

United Kingdom and the Rest of the World (RoW) as a Result of Implementing a UK-EU27FTA Following Initial ldquoSmallrdquo Brexit Shock (ML and 2014US$ Million Difference Relative to

Initial ldquoSmallrdquo Brexit Scenario)a

Volume (ML) Value (2014US$ million)

UK RoW WORLD (ldquosmallrdquo)a UK RoW WORLD (ldquosmallrdquo)a

EU27 212 minus112 100 (minus96) 750 0 750 (minus806)Chile 1 3 4 (minus18) 3 40 43 (minus92)South Africa 3 2 5 (minus14) 3 18 21 (minus60)USA 3 3 6 (minus5) 26 21 47 (minus50)Australia 1 1 2 (minus5) 8 38 46 (minus33)Argentina 1 4 5 (minus6) 4 20 24 (minus28)New Zealand 4 minus3 1 (minus1) 61 minus25 36 (minus46)Others 1 6 7 (minus9) 5 minus46 minus41 (92)WORLD 226 minus96 130 (minus154) 860 66 926 (minus1022)

a Numbers in parentheses are the world trade differences between the ldquosmallrdquo initial Brexit scenario and the baseline scenario copied fromcolumns 3 and 7 of Table 5(b)

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 243

In short this subsequent step in the Brexit process can be expected to restore muchof the initial adverse effects in the United Kingdom of the Brexit vote by 2025 Thewine exporters in the EU27 Chile and South Africa also lose less in this scenarioas compared with the initial ldquosmallrdquo scenario while other key exporters sell nearlyas much in 2025 in this scenario as they are projected to in the baseline ndash that isthey are beneficiaries of the reduced discrimination in global wine markets in thiscase where Chile and South Africa no longer enjoy preferential access to the UKmarket (and even though EU27 producers are again allowed to do so) Keep inmind though that although the estimated losses in the initial scenarios are reportedjust for 2025 they are expected to be felt throughout the preceding years of uncertaintyTherefore the longer it is before uncertainty abates and the United Kingdomrsquos currentpreferential trading arrangements are replaced by newones such as a UK-EU27 FTAthe higher the cumulative cost of the Brexit vote to the wine trade

VII Caveats and Conclusions

The above Brexit simulations are just a few of many scenarios that could be modeledObvious additional ones could also assume that FTAs are reached between theUnited Kingdom and other trading partners including Chile and South Africaand such countries as Australia and New Zealand Some of these other countriesare already exploring the ECrsquos invitation (see European Commission 2015) to nego-tiate an FTA with what will be the EU27 The sequence in which FTAs are signedand the speed with which they are implemented will matter (as was also the casewith the sequential signing over the past decade of bilateral FTAs with NortheastAsian countries by Chile Australia and New Zealand see Anderson andWittwer 2015) Even if the United Kingdom were able to sign additional bilateralFTAs and begin implementing them before 2025 it would have little effect on theabove results (because wine tariffs are such a minor contributor to them) unlessthose FTAs were to accelerate the United Kingdomrsquos economic growth and therise in the British poundrsquos value before 2025

We assume above that no changes are made to alcohol excise duties in the UnitedKingdom following Brexit even though they are scheduled to be progressively raisedwith inflation and may be raised even more for wine relative to spirits to offset theopposite effects of Brexit on those two domestic industries Nor do we make anyallowance for UK increases in consumer wine prices that may be needed to coverthe higher cost of clearing customs on imports from EU27 countries (which maybe more or less offset by increased smuggling and duty-free purchasing)Importers of fine wines including individuals who buy Bordeaux en primeur maywell reduce their demand because of concerns that delays on docks will affect thequality of their wines or because of greater currency uncertainty

To summarize Brexit is costly initially to UK consumers of wine (and of manyother tradable products) because the domestic retail price in local currency tax-

244 UK and Global Wine Markets by 2025 and Implications of Brexit

inclusive terms is 22 higher than otherwise in 2025 and the volume of wine con-sumed domestically is 28 lower in the ldquolargerdquo Brexit scenario (or 11 higher pricesand 17 lower quantities in the ldquosmallrdquo case) by 2025 unless a UK-EU27 FTA isdeveloped before then Even if such an FTA does get signed ratified by all 28 par-liaments and implemented by 2025 the slower income growth in the interim meansa smaller UK wine market in 2025 than would otherwise have been the case Thevolume reduction is a blow to many participants in UK wine-bottling transportingstoring wholesaling and retailing businesses in addition to restaurants and pubsVery little of that initial impact occurs because of higher import tariffs most impor-tant is the assumed fall in UK real incomes

The small but growing number of local UK vignerons (who supply less than 05of all domestic-wine sales and less than 4 of sparkling-wine sales) is initially pro-jected to sell less wine domestically because of reduced demand and to export only alittle extra abroad while the British pound is devalued Overall their production isonly 3 higher in 2025 even in the ldquolargerdquo Brexit scenario (not taking intoaccount the fact that their casual labor is likely to be more expensive under Brexitbecause of tighter restrictions on immigration) and their production diminishes ifand when a UK-EU27 FTA is in place

Even with a UK-EU27 FTA in place EU27 wine exporters export US$56 millionless wine in 2025 thanks to Brexit Chile and South Africa export $88 million lesswine and wine exports from the rest of the world are little different in aggregatealthough slightly larger from Australia

Clearly there will be great uncertainly for some time over the possible policy out-comes to flow from Brexit and their consequent sequential impacts on UK house-hold disposable incomes foreign exchange rates and bilateral wine tariffsMeanwhile the above projections under explicit assumptions provide some idea ofhow wine markets might be affected by the most-likely first two stages of theBrexit process (agreeing on a new tariff schedule at the WTO and agreeing to andimplementing a UK-EU27 FTA) In particular they make clear that nontrivialinitial impacts could affect the domestic wine market impacts that are likely to belarger than just the direct impact of changes in bilateral tariffs If the UnitedKingdom succeeds in getting countries to agree to transition arrangements thatdelay the changes in tariffs until new FTAs are signed and ratified the initialeffects will be less dramatic than in our first scenarios but it remains to be seenwhether any such agreements can be reached In any event the net effect of Brexiton the welfare of the worldrsquos consumers and producers of wine as a whole will benegative not just initially but permanently unless new trade policy commitmentsby the United Kingdom with major wine-exporting countries are sufficiently moreliberal than current arrangements

Kym Anderson and Glyn Wittwer 245

Appendix Key Equations in the Global Wine Markets Model

Amodel of the worldrsquos wine markets was first published byWittwer et al (2003) andmuch improved by Anderson and Wittwer (2013) Several significant enhancementshave been made to that original model Wine types have been disaggregated intomore types ndash namely nonpremium (including generic bulk) commercial-premiumand superpremium still wines and sparkling wines8 As in the original modelthere are two types of grapes premium and nonpremium Nonpremium wines usenonpremium grapes exclusively superpremium wines use mainly premium grapesand commercial-premium and sparkling wines use both types of grapes In theregional dimension the number of countries and country groups has expandedfrom 10 in the original model to 51 44 individual nations and 7 compositeregions The modelrsquos database is calibrated to 2014 for this paper based on datain Anderson et al (2017) trade data for which are downloaded from httpscomtradeunorg The modelrsquos supply-and-demand equations are based on the orig-inal ORANI modelrsquos theory (see Dixon Parmenter Sutton and Vincent 1982) andthe model is implemented using GEMPACK software (Harrison Horridge Jerieand Pearson 2014)

An enhancement of importance to the present study is the inclusion of exchange-rate variables in the model which allow a distinction between price impacts asobserved in local currency units and those observed in 2014 US dollars

In the model the grape and wine sectors minimize costs of intermediate inputssubject to weak constant elasticity of substitution (CES) substitutability betweeninputs By assumption no intermediate inputs are imported from other countries9Hence

Xcid frac14 f ethX1id CESfrac12P

cid=P1id $THORN eth1THORN

P1id X1id frac14X

c

Xcid P

cid eth2THORN

whereXcid is the quantity demanded of commodity c by grape or wine industry i in

region d Pcid is the corresponding price and X1id and P1id are the respective inter-

mediate composite quantities and prices

Two primary factors are employed in the sector labor (the quantity of which isendogenous with perfectly elastic supply) and capital Capital is usually treated asexogenous in quantity with rates of return bearing all the adjustment in thevarious scenarios This reflects the fact that grapes (a perennial crop) and wine-

8Commercial-premium still wines are defined by Anderson and Nelgen (2011) to be those priced betweenUS$250 and $750 per litre pre-tax at a countryrsquos border or wholesale9An exception concerns bulk wine imports used mainly in the bottling of commercial premium wine in theUK USA Canada Japan and Germany

246 UK and Global Wine Markets by 2025 and Implications of Brexit

plant capacity adjust slowly to market signals

Lid frac14 f ethFid CESfrac12Wid=PFid $THORN eth3THORN

Kid frac14 f ethFid CESethRid=PFid $THORN eth4THORN

PFid Fid frac14 Lid Wid thorn Kid Rid eth5THORN

Grape and wine producers are assumed to minimize costs subject to CES substitu-tion between capital and labor Equations (3) to (5) show primary factor demandsfor the labor composite Lid and capital Kid subject to a composite factor demandFid by industry i in region d The factor prices are Wid for labor Rid for capitalrentals and PFid for composite factor prices

Appendix Table A1Cumulative Consumption and Population Growth Rates and Changes in the Real Exchange Rate

(RER) Relative to the US Dollar 2014 to 2025 without Brexit ()

Aggregateconsumption Poprsquon RER

Aggregateconsumption Poprsquon RER

France 18 4 minus11 Australia 35 11 minus17Italy 11 2 minus9 New Zealand 32 9 minus26Portugal 14 0 minus9 Canada 27 8 minus18Spain 26 8 minus9 United States 31 8 0Austria 19 4 minus7 Argentina 7 10 109Belgium 20 7 minus9 Brazil 16 8 minus29Denmark 22 2 minus9 Chile 55 8 minus2Finland 21 3 minus7 Mexico 42 12 minus8Germany 14 minus2 minus11 Uruguay 45 3 1Greece 22 minus1 minus14 Other Latin America 60 10 minus5Ireland 42 12 minus9 South Africa 36 12 minus1Netherlands 21 4 minus9 Turkey 50 8 20Sweden 24 9 minus13 North Africa 53 11 0Switzerland 18 8 minus6 Other Africa 109 18 84United Kingdom 32 6 1 Middle East 52 18 minus12Other West Europe 21 10 minus1 China 79 3 5Bulgaria 41 minus7 7 Hong Kong 42 3 2Croatia 20 minus2 minus1 India 134 13 17Georgia 35 0 23 Japan 11 minus3 minus24Hungary 25 minus3 minus11 Korea 38 1 minus9Moldova 49 minus11 13 Malaysia 62 15 minus16Romania 45 minus4 22 Philippines 75 18 7Russia 18 minus2 minus8 Singapore 44 21 minus22Ukraine 22 minus5 14 Taiwan 29 1 minus13Other East Europe 40 minus5 48 Thailand 47 3 minus9

Other Asia 99 10 10

Source Authorsrsquo compilation from projections by various international agencies and from global economy-wide modeling by Anderson andStrutt (2016)

Kym Anderson and Glyn Wittwer 247

The composite factor demand Fid is proportional to total output Qid subject to aprimary factor using technology Aid Hence

Fid frac14 Qid Aid eth6THORN

The perfectly competitive zero-pure-profit condition is that total revenue valued atthe output price P0s

i multiplied by Qid equals the total production cost

P0si Qid frac14

X

cPcid X1cid thorn

X

oWo

id Loid thorn Rid Kid eth7THORN

Household demands follow a linear expenditure system in each region We reducethe optimizing problem for household consumption of each commodity subjectto a budget constraint to equations describing subsistence and discretionarydemands Aggregate subsistence expenditure WSUBd depends only on consumerprices P3cd for each commodity and the number of households N as per-capita sub-sistence quantities XSUBcd subject to given preferences are constant

WSUBd frac14X

c

P3cd XSUBcd Nd eth8THORN

Discretionary expenditures for each commodity (the left-hand side of equation (9))are equal to the marginal budget share (βcd) of aggregate discretionary expenditureThis aggregate is the bracketed term on the right-hand side of equation (9) whereW3TOTd is aggregate nominal expenditure

P3cdethX3cd XSUBcd NdTHORN frac14 βcdethW3TOTd WSUBdTHORN eth9THORN

Because real aggregate consumption is usually exogenous in our partial equilibriumsimulations the linear expenditure system determines the consumption shares ofindividual final commodities (ie the five wine types plus a composite of all otherconsumption items) driven by changes in relative prices as faced by domestic con-sumers The income elasticity of demand for each commodity is equal to the mar-ginal budget share divided by the expenditure share This number varies from 05for nonpremium wine to 17 for superpremium still wine The income elasticity ofdemand for other consumption is very close to 10 because wine accounts for anaverage of only 03 of aggregate expenditures globally and no more than 11in any country (Anderson and Nelgen 2011 Table 166)

A feature of our revised model of world wine markets is the inclusion of nominalexchange rates These appear directly in the equation linking retail prices (P3scd ) toproducer prices by country of origin (P0s

c ) where c denotes the wine type

P3 scd frac14 P0s

cfd

fsTtarcd T

taxcd thorn Pm

cd eth10THORN

248 UK and Global Wine Markets by 2025 and Implications of Brexit

The exchange rates in the consuming and producing regions are fd and fs respec-tively expressed as local currency units per $US Ttar

cd is the power of the tariff in theconsuming region and Ttax

cd is the power of the domestic consumption (or excise) taxprior to any generic value-added or goods-and-services tax Pm

cd is the price of marginm assumed to be locally supplied nontradable and therefore unaffected by theexchange rate

A given level of consumption for wine type c (X3cd) is satisfied using theArmington (1969) assumption in which wines from different countries of originare imperfectly substitutable First domestic wine is imperfectly substitutable witha composite of imports

X3sscd frac14 f ethX3cd CESethP3sscd=P3cdTHORNTHORN ss frac14 domestic imports eth11THORN

and then imports by origin (X3scd ) are determined in a second CES equation

X3scd frac14 f ethX3ssfrac14importscd CESethP3scd=P3

ssfrac14importscd THORNTHORN eth12THORN

The model enables us to show how changes in international competitiveness affectthe worldrsquos wine markets A crucial part of this exercise is explaining how pricesdetermined outside the grape and wine markets influence these markets Becausethe model is in partial equilibrium to depict the impacts of changes in internationalcompetitiveness outside price changes need to be imposed as shocks on the modelThe price of intermediate inputs other than grapes and wine shown in equations (1)and (2) is set equal to the price of GDP (Pg

d ) multiplied by a shifter Fcd

Pcid frac14 Fc

dPgd eth13THORN

If no specific price observations are available the shifter Fcd remains exogenous and

unshocked with the change in price being determined by a shock to the price ofGDP If observations are available for specific input price movements the shifterFc

dbecomes endogenous with Pc

id now exogenous and shocked

Wid frac14 Fwd P

gd eth14THORN

Wage rates are treated similarly In equation (14) if the wage shifter Fwd is exogenous

changes in wage rates Wid are determined by changes in the price of GDP If wagerate data are available Fw

d becomes endogenous andwage rates are shocked directly

Pmcd frac14 Fm

d Pgd eth15THORN

The prices of trade and transport margins are also determined by the price of GDP ifthe shifter Fm

d in equation (15) is exogenous

Changes in international competitiveness depend on changes in relative pricelevels and changes in nominal exchange rates In equation (16) fR

s denotes real

Kym Anderson and Glyn Wittwer 249

exchange-rate movements relative to the US dollar

fRs frac14 Pg

s =frac12PgUSA

fs$ eth16THORN

In equation (16) the nominal exchange rate for the United States is alwaysunchanged because nominal and real exchange rates are expressed relative to theUS currency

Changes in international market conditions may have impacts in one direction onproducer prices as expressed in US dollars P0s

i and potentially in the opposite direc-tion in local currency units Hence we calculate real producer pricesP0s

iloc in localcurrency terms (ie the price most relevant to domestic producers)

P0siloc frac14 P0s

ifs=P

gs eth17THORN

To obtain real price changes in local currency terms we convert US dollar prices(P3 s

cd for the source-specific price and P3cd for the source-composite price) to reallocal currency prices (P3 s

cdloc and P3cdloc) using the consumer price index (CPI)(Pc

d ) as the deflator

P3 scdloc frac14 P3 s

cdfd=P

cd eth18THORN

and

P3cdloc frac14 P3cdfd=Pcd eth19THORN

References

Anderson K and Nelgen S (2011) Global Wine Markets 1961 to 2009 A StatisticalCompendium Adelaide University of Adelaide Press Also available as an e-book atwwwadelaideeduaupresstitlesglobal-wine

Anderson K Nelgen S and Pinilla V (2017) Global Wine Markets 1860 to 2016 AStatistical Compendium Adelaide University of Adelaide Press (forthcoming) Also tobe available as an e-book at wwwadelaideeduaupress

Anderson K and Pinilla V (with the assistance of A J Holmes) (2017)Annual Database ofGlobal Wine Markets 1835 to 2016 Wine Economics Research Centre University ofAdelaide to be posted in October at wwwadelaideeduauwine-econdatabases

Anderson K and Strutt A (2016) Impacts of Asiarsquos rise on African and Latin Americantrade Projections to 2030 World Economy 39(2) 172ndash194

Anderson K andWittwer G (2013) Modeling global wine markets to 2018 Exchange ratestaste changes and Chinarsquos import growth Journal of Wine Economics 8(2) 131ndash158

Anderson K and Wittwer G (2015) Asiarsquos evolving role in global wine markets ChinaEconomic Review 35 1ndash14

250 UK and Global Wine Markets by 2025 and Implications of Brexit

Armington P A (1969) A theory of demand for products distinguished by place of produc-tion IMF Staff Papers 16 159ndash178

Baldwin R E (ed) (2016) Brexit Beckons Thinking Ahead by Leading Economists LondonCentre for Economic Policy Research

Campos N F (2016) Lousy experts Looking back at the ex ante estimates of the costs ofBrexit In R E Baldwin (ed) Brexit Beckons Thinking Ahead by Leading Economists(ch 3) London Centre for Economic Policy Research

Dixon P Parmenter B Sutton J and Vincent D (1982) ORANI A Multisectoral Modelof the Australian Economy Contributions to Economic Analysis 142 Amsterdam North-Holland

European Commission (2015) Trade for All Towards a More Responsible Trade andInvestment Policy Brussels European Commission

Harrison J Horridge M Jerie M and Pearson K (2014) GEMPACK ManualMelbourne GEMPACK Software

HM Treasury (2017) Spring Budget 2017 London HM TreasuryHolmes A J and Anderson K (2017a)Annual Database of National Beverage Consumption

Volumes and Expenditures 1950 to 2015 Wine Economics Research Centre University ofAdelaide posted at wwwadelaideeduauwine-econdatabases

Holmes A J and Anderson K (2017b) Convergence in National Alcohol ConsumptionPatterns New Global Indicators Journal of Wine Economics 12(2) 117ndash148

Ludington C C (2013) The Politics of Wine in Britain A New Cultural History BasingstokeUK Palgrave Macmillan

Office of National Statistics (2017) Analysis of real earnings London Office of NationalStatistics Available at httpswwwonsgovukemploymentandlabourmarketpeopleinworkearningsandworkinghoursarticlessupplementaryanalysisofaverageweeklyearningslatest

Rollo J Borchert I Dawar K Holmes P and Winters L A (2016) The World TradeOrganisation A safety net for a post-Brexit UK trade policy Briefing Paper 1UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-1-final-1pdf

Smith A (2017) Brexit Hard truths and hard choices UKTPO blog 19 June Available athttpsblogssussexacukuktpo20170619brexit-hard-truths-and-hard-choices

Swinbank A (2017) World trade rules and the policy options for British agriculture post-Brexit Briefing Paper 7 UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-7pdf

Viner J (1950) The Customs Union Issue New York Carnegie Endowment for InternationalPeace

Wine Australia (2015) Export Market Guide European Union Adelaide Wine AustraliaWittwer G Berger N and Anderson K (2003) A model of the worldrsquos wine markets

Economic Modelling 20(3) 487ndash506

Kym Anderson and Glyn Wittwer 251

Page 3: U.K. and Global Wine Markets by 2025, and Implications of ... · U.K. and Global Wine Markets by 2025, and Implications of Brexit* Kym Andersona and Glyn Wittwerb Abstract The United

markets and their participants in the United Kingdom and abroad within andoutside the European Union

II Historical Backdrop Wine in the United Kingdom and the UnitedKingdom in Global Wine Markets

Wine has rarely accounted for more than 2 of UK merchandise imports and sinceWorld War I it has averaged just 05 Also prior to 1950 wine accounted for lessthan 4 of the volume of alcohol consumed in the United Kingdom (except brieflyin the 1920s) and not much more as a share of alcohol expenditure Over the past sixdecades though winersquos share of UK alcohol consumption has steadily risen and itnow exceeds one-third in volume and value terms (Figure 1)

The United Kingdom is a key player in wine-trade circles as it has alwaysaccounted for a major ndash and often the largest ndash share of the worldrsquos wine importsPrior to 1960 the United Kingdomrsquos share of the value of world imports of wineroughly matched its share for all merchandise more than 20 in the 19th centurybut steadily declining to around 10 by 1960 Since then the two shares havediverged with the all-goods share falling to around 4 while the wine shareclimbed back above 20 by the beginning of the twenty-first century The UnitedKingdomrsquos share of the volume of global wine imports has always been below itsvalue share but those two shares have converged since 1960 and are now botharound 14 (Figure 2)

These two trends are summarized in Figure 3 The ratio of the United Kingdomrsquosshares of world imports of wine to that of all goods rose from 1 to 4 between 1960and 2000 and the ratio of the United Kingdomrsquos average import price to the worldaverage has come down from between 3 and 4 pre-1960 (when the United Kingdomwas mostly importing relatively expensive wines from Bordeaux and Champagne) toabout 12 by 1980 and is now close to 1

Associated with the change in the average quality of UK wine imports are dra-matic changes in the importance of different wine-exporting countries to UKimports and in UK shares in the wine exports of those countries

Those recent shares are very different from what they were in 1995 when the NewWorld was just beginning to expand its wine exports But such changes are notunprecedented Indeed they changed considerably not only because of the changingglobal shares of the various wine-exporting countries but also because of changes inthe United Kingdomrsquos preferential trading arrangements such as the 1703 MethuenTreaty with Portugal the 1860 Cobden-Chevalier Treaty with France and the 1932Ottawa Agreement with Commonwealth countries (Tables 1 and 2)

In the middle rows of Table 1 the 2010ndash2014 shares of exporters in the interna-tional market are shown below their shares in the UK markets For some countries

Kym Anderson and Glyn Wittwer 223

those two sets of shares are similar (France Italy Germany Chile) while for othersthey are very different The latter is most noticeable for the former colonies ofAustralia New Zealand and South Africa whose shares in the United Kingdomare more than twice their shares in the rest of the world Both shares for theEU27 exceed two-thirds Many commentators expect the EU27 shares to dropand shares of eg Australia and New Zealand to rise as a consequence of Brexit

The United Kingdomrsquos recent importance to producers in wine-exporting coun-tries is clear from Figure 4 For seven key suppliers the United Kingdom accountedin 2010ndash2014 for more than one-sixth of their wine-export earnings and for three ofthem (Australia the United States and New Zealand) the United Kingdom was amarket for more than one-third of their volume of wine exports

Recent import duties and other taxes affecting the consumer prices of alcohol inthe United Kingdom are summarized in Table 3 expressed per liter of beverageImport and excise duties on wines vary according to their alcohol content Aboutone-third of UK wine imports arrive in bulk perhaps half of which contain lessthan 13 alcohol and one-tenth of imports are sparkling so the volume-weightedaverage import duty is 13 pence per liter This figure contrasts with the volume-weighted average excise tax on wine which is 297 pence per liter To that increasein the wholesale price is added perhaps a 25 retail margin for off-trade sales andwell over a 100 margin for many restaurant sales before the 20 value-addedtax (VAT) is added So the average import price in 2013ndash2015 of 222 pence per

Figure 1

Winersquos Shares of UK Merchandise Import Value and of Volume and Value of UK AlcoholConsumptiona 1800 to 2015 ()

Fig1-Colou

ron

line

BW

inprint

a Reliable wine consumption volume data are not available for the 1930s and 1940s nor are value of alcohol consumption data pre-1955Sources Compiled from data in Anderson and Pinilla (2017) and Holmes and Anderson (2017a b)

224 UK and Global Wine Markets by 2025 and Implications of Brexit

liter is escalated to 800 pence for off-trade sales and more than 1300 pence for on-trade sales (which are about one-fifth of the total sales volume in the UnitedKingdom) The share of that latter retail price that is due to tariffs on wineimports is thus just 1 The change in the United Kingdomrsquos import trade regimefrom imposing such tariffs on wines currently imported free of duty from theEuropean Union Chile and South Africa is therefore likely to be very minorHowever as becomes clear below the effect of Brexit on wine sales involves farmore than just the trade-reducing and trade-diverting effects of altering bilateralimport tariffs

III The Economics of Leaving a Customs Union

The standard theory of customs unions (Viner 1950) focuses on the fact that whencountries join a union and impose a common external tariff on imports from non-union countries net trade can be created (depending on the height of the commonexternal tariff relative to the previous national tariffs) but trade can also be diver-sified (because of the preference to producers within the union) When a country

Figure 2

UK Shares of Value of World Merchandise Imports and of Value and Volume of World WineImports 1850 to 2015 ( 3-Year Averages to Year Shown)

Fig2-Colou

ron

line

BW

inprint

Source Compiled from data in Anderson and Pinilla (2017)

Kym Anderson and Glyn Wittwer 225

leaves a union the reverse happens because the leaving countryrsquos tariffs now apply toits imports from union countries as well as from the rest of the world Hence importsfrom union countries fall because other things equal these countriesrsquo preferentialaccess to the leaving country no longer applies That is also the case for countriesthat enjoyed an FTA with the union Of significance to wine markets are theEuropean Unionrsquos FTAs with Chile and South Africa

The size of that trade-diverting impact of leaving the union on wine depends onthe external tariff imposed on wine imports not only by the union but also by theleaving country Some have suggested that the United Kingdom should becomethe Hong Kong of Europe and move immediately to free trade on all productsOthers have suggested that this move would impose huge structural changes onthe UK economy that society would not tolerate ndash at least not without majorcompensation packages But all agree that a new trade policy that sets most-favored-nation (MFN) tariff rates is needed before the United Kingdom can beginto negotiate new preferential trading arrangements with the EU27 its FTA partnerssuch as Chile and South Africa or other countries Rollo Borchert Dawar Holmesand Winters (2016) suggest that the most practical trade policy for the UnitedKingdom to adopt at the outset is the European Unionrsquos tariff schedules previouslyagreed to at the WTO In all but one of the Brexit scenarios examined below we

Figure 3

UK Price Relative to World Price of Wine Imports and UK Wine-Import Intensitya 1950 to2015 ()

Fig3-Colou

ron

line

BW

inprint

a Import intensity is defined as the United Kingdomrsquos share of the value of global wine imports divided by the United Kingdomrsquos share of theworldrsquos total merchandise importsSource Compiled from data in Anderson and Pinilla (2017)

226 UK and Global Wine Markets by 2025 and Implications of Brexit

Table1

Shares

ofUKW

ineIm

portsfrom

TodayrsquosKey

Wine-Exp

ortin

gCountries1

675to

2014

andProjected

to2025

with

outandwith

Brexit()

Volum

eFrance

Spain

Portugal

Italy

Germany

South

Africa

Australia

United

States

Chile

New

Zealand

Argentin

aOther

countries

Total

1675ndash1696

2542

231

90

00

00

00

100

1697ndash1862

526

491

32

00

00

014

100

1863ndash1919

2626

220

00

20

00

025

100

1920ndash1940

1218

320

05

140

00

020

100

1995

volume

3210

318

143

73

21

07

100

value

439

413

102

83

21

05

100

2010ndash2014

volume

159

117

48

2110

84

12

100

value

358

215

44

115

66

12

100

Exp

ortersrsquoshareof

world

wineexports2010ndash2014

volume

1520

322

44

74

72

39

100

value

309

319

42

64

53

312

100

2025

projected

noBrexit

100

volume

188

188

26

207

45

45

65

43

64

17

20

92

100

value

323

100

30

218

44

23

58

44

54

41

20

45

100

2025

projected

with

Brexit

volume

190

188

27

208

43

44

66

42

63

17

19

93

100

value

311

96

29

210

41

24

59

44

55

41

20

70

100

SourcesCom

piledfrom

data

inAnd

ersonan

dPinilla

(201

7)to

1940

UnitedNations

COMTRADEh

ttpscomtrad

eun

orgdata

for19

95ndash2

014

andau

thorsrsquomod

elresults

for20

25

Kym Anderson and Glyn Wittwer 227

assume that Rollo et alrsquos suggestion will be the new UK trade policy commitmentto WTO members that subsequent negotiations for preferential arrangements willtake years and that any consequent agreements with the European Union andothers will be gradually implemented after that

The impact of leaving a customs union on wine markets does not only come fromtariff changes Also relevant are any effects of leaving on real UK incomes and thevalue of the British pound If the United Kingdom were to move immediately to freetrade on all products (the Hong Kong option) its per-capita income could eventuallyrise but only after considerable adjustment Should the United Kingdom insteadcommit to the current EU tariff schedule at the WTO in the first instance as weassume below then its per-capita income growth rate and the British poundrsquosexchange rates almost certainly will be lower for some time ndash at least until newtrade agreement negotiations with the EU27 and others are sufficiently advanced

Table 2Taxes on British Wine Imports by Source 1660ndash1862 (British Pounds per Kiloliter)

France Germany Spain Portugal South Africa

1660ndash1665 7 9 8 81666ndash1684 7 9 8 81685ndash1691 14 20 19 181692ndash1695 22 20 19 181696 47 20 19 181697ndash1702 51 25 23 221703 52 27 24 231704ndash1744 55 31 26 251745ndash1762 63 35 30 291763ndash1777 71 39 34 331778 79 43 38 371779 84 41 40 391780ndash1781 92 49 44 431782ndash1785 96 51 47 46 441786 65 51 37 37 371787ndash1794 47 51 32 32 371795 78 64 51 51 571796ndash1797 108 92 71 71 771798 111 96 73 73 791799ndash1801 107 92 71 71 771802 112 97 74 74 801803 131 109 87 87 871804 142 117 95 95 951805ndash1824 144 119 96 96 961825ndash1830 78 50 50 50 251831ndash1859 58 58 58 58 291860 32 32 32 32 321861 16 21 21 21 211862 11 26 26 26 26

Source Summarized from Ludington (2013 Table A1)

228 UK and Global Wine Markets by 2025 and Implications of Brexit

Figure 4

UK Shares in Wine Exports of Key Wine-Exporting Countries 2010ndash2014 ()

Fig4-BW

onlin

eBW

inprint

Source Compiled from data in Anderson and Pinilla (2017)

Table 3Import Duties Excise Duties and VAT Affecting Consumer Prices of Wine and Other Alcohol

in the United Kingdom April 1 2017

VAT 20

MFN import duties on wine pound per liter-bottled still wine lt13 alc 0114-bottled still wine 13ndash15 alc 0134-bottled still wine 15ndash18 alc 0162-bulk still wine lt13 alc 0086-bulk still wine 13ndash15 alc 0105-bulk still wine 15ndash18 alc 0134-sparkling wine 0278Weighted averagea 0130

Excise duties on alcohol pound per liter-still wine lt15 alc 2887-still wine 15ndash22 alc 3848-sparkling wine 55ndash85 alc 2795-sparkling wine 85ndash15 alc 3697-spirits (assumed 40 alc) 11551-beer (assumed 5 alc) 0954

VAT on alcohol 20

a Assumes that one-third of UK wine imports arrives in bulk that half has less than 13 alcohol and that one-tenth of imports is sparkling

Sources HM Revenue httpswwwgovukgovernmentpublicationsalcohol-duty-rate-changes for excise duties and VAT accessed April 92017 and Wine Australia (2015) for import duties converted at the 2016 average exchange rate of pound0740634 per euro

Kym Anderson and Glyn Wittwer 229

as to restore investor and consumer confidence in the United Kingdomrsquos economyThe size of those impacts is uncertain of course (see Baldwin 2016 and especiallyCampos 2016) so we consider a range of possibilities below

Those assumed adverse macroeconomic effects will add to the initial impact ofaltered wine tariffs on aggregate wine consumption in the United Kingdom andhence on its bilateral trades in wine They will make the loss of sales to theUnited Kingdom by EU (and Chilean and South African) suppliers greater thanwould otherwise be the case And they will reduce the likelihood that other countriesrsquosales of wine in the United Kingdom will be higher than in the baseline Indeed themacroeconomic effects could outweigh the trade-diverting effects so that even coun-tries that are currently discriminated against by the EU28rsquos wine-trade policy may beworse off because of Brexit

IV Global Wine Markets Model and Database

Our model of the worldrsquos wine markets first published by Wittwer Berger andAnderson (2003) and revised by Anderson and Wittwer (2013) is summarized inthe Appendix It disaggregates wine markets into four types namely nonpremiumcommercial-premium and superpremium still wines and sparkling wines2 There aretwo types of grapes premium and nonpremium Nonpremium wine uses nonpre-mium grapes exclusively superpremium wines use premium grapes exclusivelyand commercial-premium and sparkling wines use both types of grapes to varyingextents across countries The world is divided into 44 individual nations and 7 com-posite geographic regions that capture all other countries

The modelrsquos database is calibrated to 2014 based on the comprehensive wine-market volume and value data and trade and excise-tax data provided inAnderson and Pinilla (2017) and in Anderson Nelgen and Pinilla (2017) It is pro-jected assuming that aggregate national consumption population and realexchange rates change between 2014 and 2025 to the extent shown in AppendixTable 13 The Brexit alternatives to that baseline also are projected to 2025

Concerning preferences we assume a continued considerable swing toward allwine types in China and a swing away from nonpremium wines in all other countriesuntil 2025

In our baseline scenario grape- and wine-industry total-factor productivity areassumed to grow at 1 per year everywhere while grape- and wine-industry

2Commercial-premium still wines are defined byAnderson Nelgen and Pinilla (2017) to cost between US$250 and US$750 per liter pretax at a countryrsquos border or wholesale3The real exchange rate changes over the projection period are the changes expected in the nominal valueof country irsquos currency relative to the US dollar times the expected ratio of the GDP deflator for theUnited States versus that for country i

230 UK and Global Wine Markets by 2025 and Implications of Brexit

capital is assumed to grow net of depreciation at 15 per year in China but zero else-where (consistent with the almost-zero growth in global wine production and con-sumption over the past two decades)

Two alternative scenarios are considered (ldquolargerdquo and ldquosmallrdquo) for the initialimpact of Brexit to provide a range of results We assume that because of theUnited Kingdomrsquos decision to exit the European Union the rate of UK realgross domestic product (GDP) growth is only one-third or two-thirds as fast overthe projection period (09 or 18 per year instead of 26) and the Britishpound will be 20 or 10 lower in real terms than in our modelrsquos core baselineprojection4 In both alternative scenarios we assume that the United Kingdomapplies the European Unionrsquos external tariffs on wine from March 30 2019 atthe end of the two-year period following the United Kingdomrsquos formal triggeringof Article 50 (see Table 3)

We then consider a subsequent impact of Brexit presumed to result from negoti-ating signing implementing and responding to an FTA with the EU27 by 2025(Chile and South Africa currently have preferential access to EU wine marketsbut we continue to assume in this scenario that the United Kingdom does not imple-ment new bilateral FTAs with them or others in our time frame)5 This subsequentscenario assumes the British pound returns to what it would have been in 2025 in theabsence of Brexit and that real incomes regain two-thirds of the value they wouldhave had without Brexit as compared with our ldquosmallrdquo initial Brexit scenario

This global model has supply-and-demand equations and hence quantitiesprices and price elasticities for each of the grape and wine products and for asingle composite of all other products in each country Income elasticities ofdemand also exist for each final product Grapes are assumed to not be traded inter-nationally but other products are exported and imported Each market is assumed tohave been in equilibrium before any shock and to find a new market-clearingoutcome following any exogenously introduced shock

4The nominal price of the British pound in US dollars in the fortnight following the Brexit vote on June23 2016 dropped 13 to US$130 and a year later the pound sat at the same rate having dropped to anaverage of $124 (an 18 devaluation) between October 2016 and April 2017 Our choice of a low of 10and a high of 20 aims to capture future possible rates while uncertainties remain The average real wagein the United Kingdom fell in the first half of 2017 and projected real GDP growth during 2018ndash2020 hasbeen revised down to 17 in the United Kingdomrsquos latest budget (HM Treasury 2017)5South Africa currently has duty-free access to the EU28 for just 50 megaliters (ML) of wine per yearbeyond which the MFN tariff rate applies South Africa currently exports around 320 ML to theEuropean Union one-third of which initially goes to the United Kingdom How that quota of 50 MLis divided between the United Kingdom and the EU27 is subject to future negotiation (Rollo et al2016 Swinbank 2017) In the following discussion we assume none of it is accepted by the UnitedKingdom so that all South African wine imports pay the United Kingdomrsquos MFN tariff under Brexitbut altering that assumption makes very little difference to our results for any countryrsquos trade exceptSouth Africarsquos

Kym Anderson and Glyn Wittwer 231

V Projecting Global Wine Markets to 2025

Global wine production and exports are projected in the baseline from 2014 to 2025consistent with past trends The modelrsquos global volume of production (and consump-tion) rises little over that 11-year period (9) made up of a 6 decline in nonpre-mium wine and a one-sixth rise in commercial-premium and superpremium wineIn real (2014 US$) value though global wine output and consumption increaseby about 50 in total and 60 in the two premium categories The internationaltrade projections are similar although a little larger with the share of global wineproduction exported (equivalent to the share of global consumption imported)rising 2 percentage points between 2014 and 2025

The baseline projection does not greatly alter the 2014 shares of various countriesin global wine production apart from China because we assume vineyard expansionthere is faster than elsewhere6 In value terms this assumption means that Chinamoves from fifth to fourth by 2025 behind France the United States and ItalySpain remains barely ahead of Australia and they and Germany take the nextthree places (Figure 5(a)) In total wine-production volume terms China movesfrom sixth to fifth place and Argentina drops from fifth to eighth (and fromeighth to ninth in value terms)

When their products are subdivided into fine wines (superpremium still plus spark-ing) commercial-premium wines and nonpremium wines France and the UnitedStates retain the highest two places on the global ladder for fine-wine productionand Spain and Italy retain the top two places for nonpremium wine As for commer-cial-premium wine production (defined to cost between US$250 and US$750 perliter pretax at a countryrsquos wholesale level or national border) Italy retains the topranking over our projections period but at least in terms of value China challengesFrance for second place

The country rankings by projected value of total wine consumption change some-what more than those for production by 2025 with China taking second place afterthe United States ahead of France and Germany and then the United Kingdomslightly overtaking Italy to slip into fifth place (Figure 5(b)) The United StatesFrance and Germany retain the top three rankings for consuming fine wine butCanada slightly overtakes Italy for fourth place in terms of value at least In thecase of commercial-wine consumption China strengthens its number-one positionahead of the United States and the United Kingdom does likewise vis-agrave-visGermany for third place

As for the projected changes in consumption volumes China is projected to dom-inate the increase in aggregate although the United States is projected to lead the

6 In fact Chinarsquos wine production fell steadily between 2012 and 2016 by a total of one-sixth so Chinarsquoswine imports may grow faster in practice than in this baseline projection

232 UK and Global Wine Markets by 2025 and Implications of Brexit

Figure 5

Value of Wine Production and Consumption in Key Countries 2014 and Projected Baseline for2025 (2014 US$ Million at WineryWholesale Pretax Prices)

Fig5-BW

onlin

eBW

inprint

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 233

increase in consumption of fine wine In Western Europe and in the SouthernHemispherersquos New World countries fine wines are projected to substitute for com-mercial wines (defined as the sum of commercial-premium and nonpremiumwines) with almost no change in total wine consumption Sub-Saharan Africa isthe next region that is projected to take off with its growth accounting for morethan one-third of the rest of the worldrsquos increase in volume consumed

Those differences in production versus consumption rankings are reflected ininternational trade Figure 6 shows that France Italy and Spain remain the threedominant exporters of wine in aggregate value but the rankings of the next fewchange with Australia being slightly ahead of Chile and the United StatesGermany and New Zealand being nearly tied for sixth place in value termsFrance and Italy are even more dominant in fine-wine exports and remain so by2025 while Italy outranks France in the commercial-premium export categoryand Spain outranks Italy Australia and Chile in the nonpremium export class

Among the importers the United States and the United Kingdom are projected tocontinue to hold the first two places in 2025 in value terms but China moves intothird place slightly ahead of Germany followed well behind by Canada HongKong Belgium-Luxembourg the Netherlands and Japan (Figure 6(b)) OtherAfrica (excluding South Africa) is projected to experience the largest increase inimports among all the other regions followed by Other Asia which becomes asbig as Germany in value terms (Figure 7(a)) In terms of total volume of wineimports Germany and the United Kingdom held the top two shares in 2014 butby 2025 the United Kingdom is projected to be well ahead of Germany (Figure 7(b)) However this projection ignores the effects of Brexit to which we now turn

VI How Might Wine Markets Be Affected by the United Kingdomrsquos Exitfrom the European Union

As mentioned earlier for our two alternative scenarios to capture the initial effects ofBrexit (ldquolargerdquo and ldquosmallrdquo) we assume that following the United Kingdomrsquos exitfrom the European Union the United Kingdomrsquos rate of economic growth wouldbe only one-third or two-thirds as fast for the period to 2025 the British poundwould be 20 or 10 lower in real terms than in our modelrsquos baseline projectionand the United Kingdom would apply the European Unionrsquos external tariff onwine to imports from EU member countries (as part of establishing MFN ratesvia the WTO to then start new bilateral FTA negotiations) In these initial scenarioswe assume the United Kingdom does not implement any new FTAs particularlywith the EU27 Chile and South Africa Following the discussion of those initialresults we present the results of a subsequent scenario that includes the implemen-tation of an FTA between the United Kingdom and the EU27 We assume these aretwo discrete steps with no agreed arrangements to smooth the transition betweenthem

234 UK and Global Wine Markets by 2025 and Implications of Brexit

Figure 6

Value of Wine Exports and Imports Key Wine-Trading Countries 2014 and Projected Baselinefor 2025 (2014 US$ million)

Fig6-BW

onlin

eBW

inprint

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 235

Figure 7

National Shares of Global Wine Import Value and Volume 2014 and Projected Baselinefor 2025 ()

Fig7-BW

onlin

eBW

inprint

Source Authorsrsquo model results

236 UK and Global Wine Markets by 2025 and Implications of Brexit

A Initial Impact of Brexit

Because the British pound dropped by one-sixth against the US dollar in the fourmonths following the Brexit vote in June 2016 and because the UnitedKingdomrsquos average real wage fell 1 between November 2016 and April 2017(according to the latest available data see Office of National Statistics 2017) weviewed our ldquolargerdquo scenario as more likely soon after the Brexit vote But wecompare those results with ones from our ldquosmallrdquo initial-impact scenario becauseit has become more likely following the June 2017 general election that ldquosofterrdquooptions will be considered by the new government Generally the results areabout half the size in the ldquosmallrdquo scenario with the exception of the bilateraltrade effects To show the sensitivity of results to our assumptions we point outthe differences when the ldquosmallrdquo results are not close to half the results shown forthe ldquolargerdquo scenario

Before turning to those results we first present a ldquofree-traderdquo scenario whichassumes the United Kingdom chooses the radical option of becoming the HongKong of Europe Unlikely as this scenario is the results provide assurance thatour global wine model generates the standard types of effects of exiting a customsunion In this ldquofree-traderdquo scenario the United Kingdom is assumed to move tozero tariffs on all wine imports Thus the current preferences on EU Chileanand South African wine imports disappear as all other wine exporters enjoy thesame free access to UK wine markets We assume in this scenario that Brexit hasno impact on the United Kingdomrsquos real income or consumption growth rates by2025 and that by then the British pound has returned to its 2014 value relative toother currencies (similar to the real exchange rate assumed in our baseline scenarioto 2025 see Appendix Table A1)7 Under these assumptions EU Chilean andSouth African wine exports to the United Kingdom in 2025 are 003 lower thanin the no-Brexit baseline and UK imports from other wine exporters are 025higher but the United Kingdomrsquos total wine imports are hardly any different(because local wine prices fall by only 05 and incomes are unchanged) In this sce-nario UK consumers expand their volume of consumption by just 05 while theEU27 exports more and other countries export less to the rest of the world such thatworld wine exports are almost unchanged In short all these effects are in the direc-tion that comparative static-customs-union theory would predict but they are smallbecause we assume the United Kingdomrsquos real income and currency are the same asin our baseline for 2025 as described in the previous section

In the ldquolargerdquo alternate initial scenario involving Brexit as compared with thebaseline scenario to 2025 the consumer price of wine in 2025 is 22 higher in the

7If this scenario were seen as a credible long-run trade policy it would encourage more investment thatwould eventually raise UK incomes but not without major structural changes that would initiallydisrupt the economy Implicitly we assume in our free-trade scenario that these two opposite impactson real UK incomes and the British pound exactly offset each other by 2025

Kym Anderson and Glyn Wittwer 237

United Kingdom in local currency terms (20 because of real depreciation of theBritish pound 4 because of new tariffs on EU Chilean and South Africanwines and ndash2 because of slower UK income growth) The volume of UKwine consumption is 28 lower 16 because of slower UK economic growth7 because of real depreciation of the British pound and 5 because of newtariffs Superpremium still-wine sales are the most affected dropping by two-fifths while sparkling and commercial-premium wines drop a bit less than one-quarter Because the average price rises by more than the decline in volume soldthe aggregate value of UK sales even in local currency terms falls under thisldquolargerdquo Brexit scenario Under the ldquosmallrdquo Brexit scenario the consumer price ofwine in 2025 is 11 higher in the United Kingdom and its volume of wine consump-tion is 17 lower

The volume of projected UK imports in 2025 is 427 ML or nearly one-quarterless in the ldquolargerdquo scenario than in the baseline scenario comprising 58ML less spar-kling wine 31 ML less superpremium still wine and 339 ML less commercial-premium wine World imports are lower by just 239 ML because imports byother countries are 189 ML higher in response to the lower international wineprices in this scenario In value terms UK imports are $175 billion (or 27)lower in 2025 because of ldquolargerdquo Brexit $113 billion because of lower incomes$038 billion because of the fall in the British pound and $014 billion because ofthe rise in wine-import tariffs (Table 4) These aggregate trade impacts are a littlemore than half this size under the ldquosmallrdquo Brexit scenario

Despite the import levels falling because of raised import tariffs domestic con-sumption of all three quality categories of UK-produced wine is lower with thanwithout Brexit because of shrunken demand for all wines resulting from loweredUK incomes and raised local prices because of devaluation of the British poundThe British poundrsquos devaluation does make it easier for the United Kingdom tosell wines abroad though The countryrsquos exports are 7 ML or nearly 5 higherin 2025 in the ldquolargerdquo Brexit scenario and UK production is 3 higher ThoseUK exports (or re-exports of imported bulk wine after it is bottled in the UnitedKingdom) that go to EU27 countries are reduced though because of the tariffnow imposed at the new EU border

Without Brexit the United Kingdomrsquos shares of global wine imports are slightlyhigher in volume terms in 2025 than in 2010ndash2015 but 2 percentage points lower invalue terms thanks to East Asiarsquos expanding demand for imports of premium winesWith ldquolargerdquo Brexit however that value share is an additional 2 percentage pointslower and the volume share is almost 5 points lower (Figure 8) The net effects ofthese impacts on global trade are shown in Figure 9 Most of the initial tradeeffect of Brexit is a large decline in net imports of wine by the United Kingdomwith very little offsetting positive effect on trade in the rest of the world Theldquosmallrdquo Brexit numbers are a bit more than half of those for ldquolargerdquo Brexit

238 UK and Global Wine Markets by 2025 and Implications of Brexit

Table4

Difference

in2025

Projected

VolumeandValue

ofWineIm

portsb

ytheUnitedKingdom

andtheResto

fthe

World

asaConsequence

oftheInitial

BrexitS

hock

(MLand2014US$

Million

ldquoLargerdquoSc

enario)

Volum

e(M

L)

Value

(US$

million)

NP+CPa

SuperPrb

Sparkling

TOTAL

NP+CPa

SuperPrb

Sparkling

TOTAL

ΔUK

impo

rtsdu

eto

Low

erincomes

minus198

minus20

minus29

minus247

58minus644

minus253

minus234

minus1131

65Low

erpo

und

minus70

minus10

minus14

minus93

22minus248

minus127

minus102

minus476

27Highertariffs

minus71

minus1minus16

minus87

20minus110

minus8minus2

4minus143

8TOTA

Lminus339

minus31

minus58

minus427

100

minus1001

minus388

minus360

minus1750

100

difffrom

base

2332

3325

2432

3227

of

totalc

uts

797

14100

5722

21100

ΔROW

netim

ports

143

2125

189

230

143

minus181

192

ΔWORLD

TRADE

minus195

minus10

minus34

minus239

minus763

minus246

minus543

minus1552

aNon

prem

ium

plus

commercial-premium

still

wines

bSu

perpremium

still

wines

SourceA

utho

rsrsquomod

elresults

Kym Anderson and Glyn Wittwer 239

The aggregate effect of ldquolargerdquo Brexit on the market shares of various wine-export-ing countries in the United Kingdom is almost indiscernible even with one decimalpoint The projected 2025 shares are quite different from the actual 2014 shares forseveral countries They are much smaller in 2025 for South Africa Australia andNew Zealand (and the United States in volume terms) and are much larger involume for Spain and in value for Italy These results occur because wine-exportingcountries benefit differentially from the varying rates of growth in net importdemand for wine in non-UK countries over this projection period The most impor-tant projected changes are the increase in the real value of annual wine importsbetween 2014 and 2025 by China (200 or $3 billion) Other Asia (110 or $22billion) and Africa (270 or $16 billion) More than half of Australiarsquos increasein annual exports from 2014 to 2025 go to Asia and more than half of SouthAfricarsquos increase in exports go to Other Africa

Table 5 reveals that ldquolargerdquo Brexit lowers EU Chilean and South African wineexports by 150 ML or US$12 billion in the case of the European Union with someof their exports diverted from the United Kingdom to the EU27 and other marketsin competition with New World exporters Although the United States Australiaand Argentina sell only a little less to the United Kingdom they also sell less toother countries For Chile and South Africa which lose their preferential access toUK (but not to EU27) markets in this Brexit scenario some exports are redirected

Figure 8

UK Shares of World Wine Imports 2010ndash2015 and Projected to 2025 without and with theldquoLargerdquo Initial Brexit Shock ()

Fig8-BW

onlin

eBW

inprint

Sources Anderson and Pinilla (2017) and authorsrsquo model results

240 UK and Global Wine Markets by 2025 and Implications of Brexit

from the United Kingdom to EU27 countries ndash but again they export less overallGlobal wine trade in 2025 is 240 ML (19) or $18 billion (35) less under thisldquolargerdquo Brexit scenario The percentage by which wine exportersrsquo trade shrinks isgreater for values than for volumes because of changes in relative prices of differ-ent-quality wines Those differences are shown in the numbers in parentheses inTable 5

Three other points are worth making about Table 5 One is that Australia sellsslightly more to the United Kingdom in the ldquosmallrdquo Brexit scenario rather thanslightly less as in the ldquolargerdquo Brexit case Evidently the negative income and price(devaluation) effects do not more than offset the positive trade-diverting effect onAustralian exports to the United Kingdom of removing preferences in the ldquosmallrdquoscenario Second New Zealand sells slightly more to non-UK countries underBrexit despite greater competition from the EU27 Chile and South Africa Thisanomaly is due to changes in the relative prices of different qualities of wine inglobal wine markets bearing in mind that New Zealand has the worldrsquos highestaverage price for still-wine exports And third the value (but not the volume) ofexports of ldquoOtherrdquo countries to markets other than the United Kingdom arehigher under Brexit This result too is due to changes in the relative prices of differ-ent qualities of wine in global wine markets

Figure 9

Difference in 2025 Wine Import Volumes and Values as a Result of the ldquoLargerdquo Initial BrexitShock (ML and US$ Million in 2014 US dollars)

Fig9-BW

onlin

eBW

inprint

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 241

B Subsequent Impact of Brexit from a UK-EU27 FTA

The next-most-likely step in the Brexit process is for the United Kingdom to nego-tiate a new trade arrangement with the EU27 We therefore assume that a UK-EU27FTAwith free bilateral wine trade is implemented and adjusted to by 2025 and thatprogress toward that end occurs soon enough that the adverse macroeconomicshocks from the initial impact of uncertainty over the Brexit process are confinedto those assumed in the ldquosmallrdquo scenario outlined above In this subsequent scenariowe assume that the British pound returns to the value it would have reached in 2025in the absence of Brexit (ie reversing the 10 devaluation assumed in the ldquosmallrdquoinitial Brexit scenario) and that real incomes in the United Kingdom are 8 higher

Table 5Difference in 2025 BilateralWine Import Volumes and Values fromKey Exporters by the United

Kingdom and the Rest of the World (RoW) as a Result of Initial Brexit Shock(ML and 2014US$ Million)a

(a) ldquolargerdquo scenario

Volume (ML) Value (2014US$ million)

UK RoW WORLD () UK RoW WORLD ()

EU27 minus287 136 minus150 (ndash17) minus1187 minus5 minus1192 (ndash31)Chile minus59 35 minus25 (ndash30) minus169 31 minus138 (ndash48)South Africa minus53 35 minus18 (ndash32) minus105 20 minus85 (ndash67)USA minus7 minus6 minus13 (ndash24) minus75 minus40 minus115 (ndash50)Australia minus4 minus3 minus7 (ndash09) minus25 minus65 minus90 (ndash30)Argentina minus3 minus9 minus12 (ndash48) minus16 minus39 minus55 (ndash52)New Zealand minus11 9 minus2 (ndash09) minus162 71 minus91 (ndash43)Others minus2 minus10 minus12 (ndash02) minus11 minus52 minus63 (ndash44)WORLD minus427 187 minus240 (ndash19) minus1750 minus79 minus1829 (ndash35)

(a) ldquosmallrdquo scenario

Volume (ML) Value (2014US$ million)

UK RoW WORLD () UK RoW WORLD ()

EU27 minus178 82 minus96 (ndash12) minus692 minus43 minus736 (ndash19)Chile minus46 28 minus18 (ndash24) minus128 36 minus91 (ndash32)South Africa minus43 29 minus14 (ndash42) minus82 23 minus59 (ndash47)USA 1 minus6 minus5 (ndash11) minus23 minus28 minus51 (ndash22)Australia 5 minus10 minus5 (ndash06) 19 minus56 minus38 (ndash13)Argentina 0 minus6 minus6 (ndash26) minus3 minus25 minus29 (ndash27)New Zealand minus5 4 minus1 (ndash06) minus80 34 minus46 (ndash22)Others 0 minus9 minus9 (ndash01) minus1 minus33 minus34 (ndash24)WORLD minus266 112 minus154 (ndash13) minus991 minus92 minus1083 (ndash21)

a Numbers in parentheses are the percentage difference between the Brexit and baseline scenarios for 2025 projected wine-import volumes orvalues by source

Source Authorsrsquo model results

242 UK and Global Wine Markets by 2025 and Implications of Brexit

in 2025 than in the ldquosmallrdquo scenario ndash that is they regain two-thirds of the value theywould have reached without Brexit-related uncertainty

This subsequent development in the Brexit process reverses most of but not allthe initial effects of Brexit by 2025 because of our assumption that the lost growthin the initial years of uncertainty following the Brexit vote are only partly recoveredby 2025 following the implementation of a UK-EU27 FTA Moreover the longer ittakes before this FTA is finalized and implemented the longer the estimated initialadverse effects persist and the larger the cumulative cost of Brexit to UK wine con-sumers and to grape and wine producers in wine-exporting countries

Table 6 summarizes the subsequent trade effects for 2025 It suggests that all butone-tenth of the loss in value of world trade in wine from the initial ldquosmallrdquo impact isrestored most of it because of a smaller reduction in the United Kingdomrsquos wineimports following the FTA Most of that improved outcome is because of recoveredimports from the EU27 commensurate with the latterrsquos high share of UK importsEven though Chile and South Africa are assumed in this scenario to have not yetsigned an FTA with the United Kingdom they export slightly more to the UnitedKingdom (and even more to all other countries) than in the initial ldquosmallrdquo scenario

Within the United Kingdom this FTA brings down the local currency consumerprice of wine by 9 largely offsetting the 11 rise in the initial ldquosmallrdquo Brexit sce-nario and it raises the volume consumed in the United Kingdom by 18 fully off-setting the 17 fall in the initial ldquosmallrdquo Brexit scenario

Table 6Difference in 2025 Bilateral Wine-Import Volumes and Values from Key Exporters by the

United Kingdom and the Rest of the World (RoW) as a Result of Implementing a UK-EU27FTA Following Initial ldquoSmallrdquo Brexit Shock (ML and 2014US$ Million Difference Relative to

Initial ldquoSmallrdquo Brexit Scenario)a

Volume (ML) Value (2014US$ million)

UK RoW WORLD (ldquosmallrdquo)a UK RoW WORLD (ldquosmallrdquo)a

EU27 212 minus112 100 (minus96) 750 0 750 (minus806)Chile 1 3 4 (minus18) 3 40 43 (minus92)South Africa 3 2 5 (minus14) 3 18 21 (minus60)USA 3 3 6 (minus5) 26 21 47 (minus50)Australia 1 1 2 (minus5) 8 38 46 (minus33)Argentina 1 4 5 (minus6) 4 20 24 (minus28)New Zealand 4 minus3 1 (minus1) 61 minus25 36 (minus46)Others 1 6 7 (minus9) 5 minus46 minus41 (92)WORLD 226 minus96 130 (minus154) 860 66 926 (minus1022)

a Numbers in parentheses are the world trade differences between the ldquosmallrdquo initial Brexit scenario and the baseline scenario copied fromcolumns 3 and 7 of Table 5(b)

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 243

In short this subsequent step in the Brexit process can be expected to restore muchof the initial adverse effects in the United Kingdom of the Brexit vote by 2025 Thewine exporters in the EU27 Chile and South Africa also lose less in this scenarioas compared with the initial ldquosmallrdquo scenario while other key exporters sell nearlyas much in 2025 in this scenario as they are projected to in the baseline ndash that isthey are beneficiaries of the reduced discrimination in global wine markets in thiscase where Chile and South Africa no longer enjoy preferential access to the UKmarket (and even though EU27 producers are again allowed to do so) Keep inmind though that although the estimated losses in the initial scenarios are reportedjust for 2025 they are expected to be felt throughout the preceding years of uncertaintyTherefore the longer it is before uncertainty abates and the United Kingdomrsquos currentpreferential trading arrangements are replaced by newones such as a UK-EU27 FTAthe higher the cumulative cost of the Brexit vote to the wine trade

VII Caveats and Conclusions

The above Brexit simulations are just a few of many scenarios that could be modeledObvious additional ones could also assume that FTAs are reached between theUnited Kingdom and other trading partners including Chile and South Africaand such countries as Australia and New Zealand Some of these other countriesare already exploring the ECrsquos invitation (see European Commission 2015) to nego-tiate an FTA with what will be the EU27 The sequence in which FTAs are signedand the speed with which they are implemented will matter (as was also the casewith the sequential signing over the past decade of bilateral FTAs with NortheastAsian countries by Chile Australia and New Zealand see Anderson andWittwer 2015) Even if the United Kingdom were able to sign additional bilateralFTAs and begin implementing them before 2025 it would have little effect on theabove results (because wine tariffs are such a minor contributor to them) unlessthose FTAs were to accelerate the United Kingdomrsquos economic growth and therise in the British poundrsquos value before 2025

We assume above that no changes are made to alcohol excise duties in the UnitedKingdom following Brexit even though they are scheduled to be progressively raisedwith inflation and may be raised even more for wine relative to spirits to offset theopposite effects of Brexit on those two domestic industries Nor do we make anyallowance for UK increases in consumer wine prices that may be needed to coverthe higher cost of clearing customs on imports from EU27 countries (which maybe more or less offset by increased smuggling and duty-free purchasing)Importers of fine wines including individuals who buy Bordeaux en primeur maywell reduce their demand because of concerns that delays on docks will affect thequality of their wines or because of greater currency uncertainty

To summarize Brexit is costly initially to UK consumers of wine (and of manyother tradable products) because the domestic retail price in local currency tax-

244 UK and Global Wine Markets by 2025 and Implications of Brexit

inclusive terms is 22 higher than otherwise in 2025 and the volume of wine con-sumed domestically is 28 lower in the ldquolargerdquo Brexit scenario (or 11 higher pricesand 17 lower quantities in the ldquosmallrdquo case) by 2025 unless a UK-EU27 FTA isdeveloped before then Even if such an FTA does get signed ratified by all 28 par-liaments and implemented by 2025 the slower income growth in the interim meansa smaller UK wine market in 2025 than would otherwise have been the case Thevolume reduction is a blow to many participants in UK wine-bottling transportingstoring wholesaling and retailing businesses in addition to restaurants and pubsVery little of that initial impact occurs because of higher import tariffs most impor-tant is the assumed fall in UK real incomes

The small but growing number of local UK vignerons (who supply less than 05of all domestic-wine sales and less than 4 of sparkling-wine sales) is initially pro-jected to sell less wine domestically because of reduced demand and to export only alittle extra abroad while the British pound is devalued Overall their production isonly 3 higher in 2025 even in the ldquolargerdquo Brexit scenario (not taking intoaccount the fact that their casual labor is likely to be more expensive under Brexitbecause of tighter restrictions on immigration) and their production diminishes ifand when a UK-EU27 FTA is in place

Even with a UK-EU27 FTA in place EU27 wine exporters export US$56 millionless wine in 2025 thanks to Brexit Chile and South Africa export $88 million lesswine and wine exports from the rest of the world are little different in aggregatealthough slightly larger from Australia

Clearly there will be great uncertainly for some time over the possible policy out-comes to flow from Brexit and their consequent sequential impacts on UK house-hold disposable incomes foreign exchange rates and bilateral wine tariffsMeanwhile the above projections under explicit assumptions provide some idea ofhow wine markets might be affected by the most-likely first two stages of theBrexit process (agreeing on a new tariff schedule at the WTO and agreeing to andimplementing a UK-EU27 FTA) In particular they make clear that nontrivialinitial impacts could affect the domestic wine market impacts that are likely to belarger than just the direct impact of changes in bilateral tariffs If the UnitedKingdom succeeds in getting countries to agree to transition arrangements thatdelay the changes in tariffs until new FTAs are signed and ratified the initialeffects will be less dramatic than in our first scenarios but it remains to be seenwhether any such agreements can be reached In any event the net effect of Brexiton the welfare of the worldrsquos consumers and producers of wine as a whole will benegative not just initially but permanently unless new trade policy commitmentsby the United Kingdom with major wine-exporting countries are sufficiently moreliberal than current arrangements

Kym Anderson and Glyn Wittwer 245

Appendix Key Equations in the Global Wine Markets Model

Amodel of the worldrsquos wine markets was first published byWittwer et al (2003) andmuch improved by Anderson and Wittwer (2013) Several significant enhancementshave been made to that original model Wine types have been disaggregated intomore types ndash namely nonpremium (including generic bulk) commercial-premiumand superpremium still wines and sparkling wines8 As in the original modelthere are two types of grapes premium and nonpremium Nonpremium wines usenonpremium grapes exclusively superpremium wines use mainly premium grapesand commercial-premium and sparkling wines use both types of grapes In theregional dimension the number of countries and country groups has expandedfrom 10 in the original model to 51 44 individual nations and 7 compositeregions The modelrsquos database is calibrated to 2014 for this paper based on datain Anderson et al (2017) trade data for which are downloaded from httpscomtradeunorg The modelrsquos supply-and-demand equations are based on the orig-inal ORANI modelrsquos theory (see Dixon Parmenter Sutton and Vincent 1982) andthe model is implemented using GEMPACK software (Harrison Horridge Jerieand Pearson 2014)

An enhancement of importance to the present study is the inclusion of exchange-rate variables in the model which allow a distinction between price impacts asobserved in local currency units and those observed in 2014 US dollars

In the model the grape and wine sectors minimize costs of intermediate inputssubject to weak constant elasticity of substitution (CES) substitutability betweeninputs By assumption no intermediate inputs are imported from other countries9Hence

Xcid frac14 f ethX1id CESfrac12P

cid=P1id $THORN eth1THORN

P1id X1id frac14X

c

Xcid P

cid eth2THORN

whereXcid is the quantity demanded of commodity c by grape or wine industry i in

region d Pcid is the corresponding price and X1id and P1id are the respective inter-

mediate composite quantities and prices

Two primary factors are employed in the sector labor (the quantity of which isendogenous with perfectly elastic supply) and capital Capital is usually treated asexogenous in quantity with rates of return bearing all the adjustment in thevarious scenarios This reflects the fact that grapes (a perennial crop) and wine-

8Commercial-premium still wines are defined by Anderson and Nelgen (2011) to be those priced betweenUS$250 and $750 per litre pre-tax at a countryrsquos border or wholesale9An exception concerns bulk wine imports used mainly in the bottling of commercial premium wine in theUK USA Canada Japan and Germany

246 UK and Global Wine Markets by 2025 and Implications of Brexit

plant capacity adjust slowly to market signals

Lid frac14 f ethFid CESfrac12Wid=PFid $THORN eth3THORN

Kid frac14 f ethFid CESethRid=PFid $THORN eth4THORN

PFid Fid frac14 Lid Wid thorn Kid Rid eth5THORN

Grape and wine producers are assumed to minimize costs subject to CES substitu-tion between capital and labor Equations (3) to (5) show primary factor demandsfor the labor composite Lid and capital Kid subject to a composite factor demandFid by industry i in region d The factor prices are Wid for labor Rid for capitalrentals and PFid for composite factor prices

Appendix Table A1Cumulative Consumption and Population Growth Rates and Changes in the Real Exchange Rate

(RER) Relative to the US Dollar 2014 to 2025 without Brexit ()

Aggregateconsumption Poprsquon RER

Aggregateconsumption Poprsquon RER

France 18 4 minus11 Australia 35 11 minus17Italy 11 2 minus9 New Zealand 32 9 minus26Portugal 14 0 minus9 Canada 27 8 minus18Spain 26 8 minus9 United States 31 8 0Austria 19 4 minus7 Argentina 7 10 109Belgium 20 7 minus9 Brazil 16 8 minus29Denmark 22 2 minus9 Chile 55 8 minus2Finland 21 3 minus7 Mexico 42 12 minus8Germany 14 minus2 minus11 Uruguay 45 3 1Greece 22 minus1 minus14 Other Latin America 60 10 minus5Ireland 42 12 minus9 South Africa 36 12 minus1Netherlands 21 4 minus9 Turkey 50 8 20Sweden 24 9 minus13 North Africa 53 11 0Switzerland 18 8 minus6 Other Africa 109 18 84United Kingdom 32 6 1 Middle East 52 18 minus12Other West Europe 21 10 minus1 China 79 3 5Bulgaria 41 minus7 7 Hong Kong 42 3 2Croatia 20 minus2 minus1 India 134 13 17Georgia 35 0 23 Japan 11 minus3 minus24Hungary 25 minus3 minus11 Korea 38 1 minus9Moldova 49 minus11 13 Malaysia 62 15 minus16Romania 45 minus4 22 Philippines 75 18 7Russia 18 minus2 minus8 Singapore 44 21 minus22Ukraine 22 minus5 14 Taiwan 29 1 minus13Other East Europe 40 minus5 48 Thailand 47 3 minus9

Other Asia 99 10 10

Source Authorsrsquo compilation from projections by various international agencies and from global economy-wide modeling by Anderson andStrutt (2016)

Kym Anderson and Glyn Wittwer 247

The composite factor demand Fid is proportional to total output Qid subject to aprimary factor using technology Aid Hence

Fid frac14 Qid Aid eth6THORN

The perfectly competitive zero-pure-profit condition is that total revenue valued atthe output price P0s

i multiplied by Qid equals the total production cost

P0si Qid frac14

X

cPcid X1cid thorn

X

oWo

id Loid thorn Rid Kid eth7THORN

Household demands follow a linear expenditure system in each region We reducethe optimizing problem for household consumption of each commodity subjectto a budget constraint to equations describing subsistence and discretionarydemands Aggregate subsistence expenditure WSUBd depends only on consumerprices P3cd for each commodity and the number of households N as per-capita sub-sistence quantities XSUBcd subject to given preferences are constant

WSUBd frac14X

c

P3cd XSUBcd Nd eth8THORN

Discretionary expenditures for each commodity (the left-hand side of equation (9))are equal to the marginal budget share (βcd) of aggregate discretionary expenditureThis aggregate is the bracketed term on the right-hand side of equation (9) whereW3TOTd is aggregate nominal expenditure

P3cdethX3cd XSUBcd NdTHORN frac14 βcdethW3TOTd WSUBdTHORN eth9THORN

Because real aggregate consumption is usually exogenous in our partial equilibriumsimulations the linear expenditure system determines the consumption shares ofindividual final commodities (ie the five wine types plus a composite of all otherconsumption items) driven by changes in relative prices as faced by domestic con-sumers The income elasticity of demand for each commodity is equal to the mar-ginal budget share divided by the expenditure share This number varies from 05for nonpremium wine to 17 for superpremium still wine The income elasticity ofdemand for other consumption is very close to 10 because wine accounts for anaverage of only 03 of aggregate expenditures globally and no more than 11in any country (Anderson and Nelgen 2011 Table 166)

A feature of our revised model of world wine markets is the inclusion of nominalexchange rates These appear directly in the equation linking retail prices (P3scd ) toproducer prices by country of origin (P0s

c ) where c denotes the wine type

P3 scd frac14 P0s

cfd

fsTtarcd T

taxcd thorn Pm

cd eth10THORN

248 UK and Global Wine Markets by 2025 and Implications of Brexit

The exchange rates in the consuming and producing regions are fd and fs respec-tively expressed as local currency units per $US Ttar

cd is the power of the tariff in theconsuming region and Ttax

cd is the power of the domestic consumption (or excise) taxprior to any generic value-added or goods-and-services tax Pm

cd is the price of marginm assumed to be locally supplied nontradable and therefore unaffected by theexchange rate

A given level of consumption for wine type c (X3cd) is satisfied using theArmington (1969) assumption in which wines from different countries of originare imperfectly substitutable First domestic wine is imperfectly substitutable witha composite of imports

X3sscd frac14 f ethX3cd CESethP3sscd=P3cdTHORNTHORN ss frac14 domestic imports eth11THORN

and then imports by origin (X3scd ) are determined in a second CES equation

X3scd frac14 f ethX3ssfrac14importscd CESethP3scd=P3

ssfrac14importscd THORNTHORN eth12THORN

The model enables us to show how changes in international competitiveness affectthe worldrsquos wine markets A crucial part of this exercise is explaining how pricesdetermined outside the grape and wine markets influence these markets Becausethe model is in partial equilibrium to depict the impacts of changes in internationalcompetitiveness outside price changes need to be imposed as shocks on the modelThe price of intermediate inputs other than grapes and wine shown in equations (1)and (2) is set equal to the price of GDP (Pg

d ) multiplied by a shifter Fcd

Pcid frac14 Fc

dPgd eth13THORN

If no specific price observations are available the shifter Fcd remains exogenous and

unshocked with the change in price being determined by a shock to the price ofGDP If observations are available for specific input price movements the shifterFc

dbecomes endogenous with Pc

id now exogenous and shocked

Wid frac14 Fwd P

gd eth14THORN

Wage rates are treated similarly In equation (14) if the wage shifter Fwd is exogenous

changes in wage rates Wid are determined by changes in the price of GDP If wagerate data are available Fw

d becomes endogenous andwage rates are shocked directly

Pmcd frac14 Fm

d Pgd eth15THORN

The prices of trade and transport margins are also determined by the price of GDP ifthe shifter Fm

d in equation (15) is exogenous

Changes in international competitiveness depend on changes in relative pricelevels and changes in nominal exchange rates In equation (16) fR

s denotes real

Kym Anderson and Glyn Wittwer 249

exchange-rate movements relative to the US dollar

fRs frac14 Pg

s =frac12PgUSA

fs$ eth16THORN

In equation (16) the nominal exchange rate for the United States is alwaysunchanged because nominal and real exchange rates are expressed relative to theUS currency

Changes in international market conditions may have impacts in one direction onproducer prices as expressed in US dollars P0s

i and potentially in the opposite direc-tion in local currency units Hence we calculate real producer pricesP0s

iloc in localcurrency terms (ie the price most relevant to domestic producers)

P0siloc frac14 P0s

ifs=P

gs eth17THORN

To obtain real price changes in local currency terms we convert US dollar prices(P3 s

cd for the source-specific price and P3cd for the source-composite price) to reallocal currency prices (P3 s

cdloc and P3cdloc) using the consumer price index (CPI)(Pc

d ) as the deflator

P3 scdloc frac14 P3 s

cdfd=P

cd eth18THORN

and

P3cdloc frac14 P3cdfd=Pcd eth19THORN

References

Anderson K and Nelgen S (2011) Global Wine Markets 1961 to 2009 A StatisticalCompendium Adelaide University of Adelaide Press Also available as an e-book atwwwadelaideeduaupresstitlesglobal-wine

Anderson K Nelgen S and Pinilla V (2017) Global Wine Markets 1860 to 2016 AStatistical Compendium Adelaide University of Adelaide Press (forthcoming) Also tobe available as an e-book at wwwadelaideeduaupress

Anderson K and Pinilla V (with the assistance of A J Holmes) (2017)Annual Database ofGlobal Wine Markets 1835 to 2016 Wine Economics Research Centre University ofAdelaide to be posted in October at wwwadelaideeduauwine-econdatabases

Anderson K and Strutt A (2016) Impacts of Asiarsquos rise on African and Latin Americantrade Projections to 2030 World Economy 39(2) 172ndash194

Anderson K andWittwer G (2013) Modeling global wine markets to 2018 Exchange ratestaste changes and Chinarsquos import growth Journal of Wine Economics 8(2) 131ndash158

Anderson K and Wittwer G (2015) Asiarsquos evolving role in global wine markets ChinaEconomic Review 35 1ndash14

250 UK and Global Wine Markets by 2025 and Implications of Brexit

Armington P A (1969) A theory of demand for products distinguished by place of produc-tion IMF Staff Papers 16 159ndash178

Baldwin R E (ed) (2016) Brexit Beckons Thinking Ahead by Leading Economists LondonCentre for Economic Policy Research

Campos N F (2016) Lousy experts Looking back at the ex ante estimates of the costs ofBrexit In R E Baldwin (ed) Brexit Beckons Thinking Ahead by Leading Economists(ch 3) London Centre for Economic Policy Research

Dixon P Parmenter B Sutton J and Vincent D (1982) ORANI A Multisectoral Modelof the Australian Economy Contributions to Economic Analysis 142 Amsterdam North-Holland

European Commission (2015) Trade for All Towards a More Responsible Trade andInvestment Policy Brussels European Commission

Harrison J Horridge M Jerie M and Pearson K (2014) GEMPACK ManualMelbourne GEMPACK Software

HM Treasury (2017) Spring Budget 2017 London HM TreasuryHolmes A J and Anderson K (2017a)Annual Database of National Beverage Consumption

Volumes and Expenditures 1950 to 2015 Wine Economics Research Centre University ofAdelaide posted at wwwadelaideeduauwine-econdatabases

Holmes A J and Anderson K (2017b) Convergence in National Alcohol ConsumptionPatterns New Global Indicators Journal of Wine Economics 12(2) 117ndash148

Ludington C C (2013) The Politics of Wine in Britain A New Cultural History BasingstokeUK Palgrave Macmillan

Office of National Statistics (2017) Analysis of real earnings London Office of NationalStatistics Available at httpswwwonsgovukemploymentandlabourmarketpeopleinworkearningsandworkinghoursarticlessupplementaryanalysisofaverageweeklyearningslatest

Rollo J Borchert I Dawar K Holmes P and Winters L A (2016) The World TradeOrganisation A safety net for a post-Brexit UK trade policy Briefing Paper 1UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-1-final-1pdf

Smith A (2017) Brexit Hard truths and hard choices UKTPO blog 19 June Available athttpsblogssussexacukuktpo20170619brexit-hard-truths-and-hard-choices

Swinbank A (2017) World trade rules and the policy options for British agriculture post-Brexit Briefing Paper 7 UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-7pdf

Viner J (1950) The Customs Union Issue New York Carnegie Endowment for InternationalPeace

Wine Australia (2015) Export Market Guide European Union Adelaide Wine AustraliaWittwer G Berger N and Anderson K (2003) A model of the worldrsquos wine markets

Economic Modelling 20(3) 487ndash506

Kym Anderson and Glyn Wittwer 251

Page 4: U.K. and Global Wine Markets by 2025, and Implications of ... · U.K. and Global Wine Markets by 2025, and Implications of Brexit* Kym Andersona and Glyn Wittwerb Abstract The United

those two sets of shares are similar (France Italy Germany Chile) while for othersthey are very different The latter is most noticeable for the former colonies ofAustralia New Zealand and South Africa whose shares in the United Kingdomare more than twice their shares in the rest of the world Both shares for theEU27 exceed two-thirds Many commentators expect the EU27 shares to dropand shares of eg Australia and New Zealand to rise as a consequence of Brexit

The United Kingdomrsquos recent importance to producers in wine-exporting coun-tries is clear from Figure 4 For seven key suppliers the United Kingdom accountedin 2010ndash2014 for more than one-sixth of their wine-export earnings and for three ofthem (Australia the United States and New Zealand) the United Kingdom was amarket for more than one-third of their volume of wine exports

Recent import duties and other taxes affecting the consumer prices of alcohol inthe United Kingdom are summarized in Table 3 expressed per liter of beverageImport and excise duties on wines vary according to their alcohol content Aboutone-third of UK wine imports arrive in bulk perhaps half of which contain lessthan 13 alcohol and one-tenth of imports are sparkling so the volume-weightedaverage import duty is 13 pence per liter This figure contrasts with the volume-weighted average excise tax on wine which is 297 pence per liter To that increasein the wholesale price is added perhaps a 25 retail margin for off-trade sales andwell over a 100 margin for many restaurant sales before the 20 value-addedtax (VAT) is added So the average import price in 2013ndash2015 of 222 pence per

Figure 1

Winersquos Shares of UK Merchandise Import Value and of Volume and Value of UK AlcoholConsumptiona 1800 to 2015 ()

Fig1-Colou

ron

line

BW

inprint

a Reliable wine consumption volume data are not available for the 1930s and 1940s nor are value of alcohol consumption data pre-1955Sources Compiled from data in Anderson and Pinilla (2017) and Holmes and Anderson (2017a b)

224 UK and Global Wine Markets by 2025 and Implications of Brexit

liter is escalated to 800 pence for off-trade sales and more than 1300 pence for on-trade sales (which are about one-fifth of the total sales volume in the UnitedKingdom) The share of that latter retail price that is due to tariffs on wineimports is thus just 1 The change in the United Kingdomrsquos import trade regimefrom imposing such tariffs on wines currently imported free of duty from theEuropean Union Chile and South Africa is therefore likely to be very minorHowever as becomes clear below the effect of Brexit on wine sales involves farmore than just the trade-reducing and trade-diverting effects of altering bilateralimport tariffs

III The Economics of Leaving a Customs Union

The standard theory of customs unions (Viner 1950) focuses on the fact that whencountries join a union and impose a common external tariff on imports from non-union countries net trade can be created (depending on the height of the commonexternal tariff relative to the previous national tariffs) but trade can also be diver-sified (because of the preference to producers within the union) When a country

Figure 2

UK Shares of Value of World Merchandise Imports and of Value and Volume of World WineImports 1850 to 2015 ( 3-Year Averages to Year Shown)

Fig2-Colou

ron

line

BW

inprint

Source Compiled from data in Anderson and Pinilla (2017)

Kym Anderson and Glyn Wittwer 225

leaves a union the reverse happens because the leaving countryrsquos tariffs now apply toits imports from union countries as well as from the rest of the world Hence importsfrom union countries fall because other things equal these countriesrsquo preferentialaccess to the leaving country no longer applies That is also the case for countriesthat enjoyed an FTA with the union Of significance to wine markets are theEuropean Unionrsquos FTAs with Chile and South Africa

The size of that trade-diverting impact of leaving the union on wine depends onthe external tariff imposed on wine imports not only by the union but also by theleaving country Some have suggested that the United Kingdom should becomethe Hong Kong of Europe and move immediately to free trade on all productsOthers have suggested that this move would impose huge structural changes onthe UK economy that society would not tolerate ndash at least not without majorcompensation packages But all agree that a new trade policy that sets most-favored-nation (MFN) tariff rates is needed before the United Kingdom can beginto negotiate new preferential trading arrangements with the EU27 its FTA partnerssuch as Chile and South Africa or other countries Rollo Borchert Dawar Holmesand Winters (2016) suggest that the most practical trade policy for the UnitedKingdom to adopt at the outset is the European Unionrsquos tariff schedules previouslyagreed to at the WTO In all but one of the Brexit scenarios examined below we

Figure 3

UK Price Relative to World Price of Wine Imports and UK Wine-Import Intensitya 1950 to2015 ()

Fig3-Colou

ron

line

BW

inprint

a Import intensity is defined as the United Kingdomrsquos share of the value of global wine imports divided by the United Kingdomrsquos share of theworldrsquos total merchandise importsSource Compiled from data in Anderson and Pinilla (2017)

226 UK and Global Wine Markets by 2025 and Implications of Brexit

Table1

Shares

ofUKW

ineIm

portsfrom

TodayrsquosKey

Wine-Exp

ortin

gCountries1

675to

2014

andProjected

to2025

with

outandwith

Brexit()

Volum

eFrance

Spain

Portugal

Italy

Germany

South

Africa

Australia

United

States

Chile

New

Zealand

Argentin

aOther

countries

Total

1675ndash1696

2542

231

90

00

00

00

100

1697ndash1862

526

491

32

00

00

014

100

1863ndash1919

2626

220

00

20

00

025

100

1920ndash1940

1218

320

05

140

00

020

100

1995

volume

3210

318

143

73

21

07

100

value

439

413

102

83

21

05

100

2010ndash2014

volume

159

117

48

2110

84

12

100

value

358

215

44

115

66

12

100

Exp

ortersrsquoshareof

world

wineexports2010ndash2014

volume

1520

322

44

74

72

39

100

value

309

319

42

64

53

312

100

2025

projected

noBrexit

100

volume

188

188

26

207

45

45

65

43

64

17

20

92

100

value

323

100

30

218

44

23

58

44

54

41

20

45

100

2025

projected

with

Brexit

volume

190

188

27

208

43

44

66

42

63

17

19

93

100

value

311

96

29

210

41

24

59

44

55

41

20

70

100

SourcesCom

piledfrom

data

inAnd

ersonan

dPinilla

(201

7)to

1940

UnitedNations

COMTRADEh

ttpscomtrad

eun

orgdata

for19

95ndash2

014

andau

thorsrsquomod

elresults

for20

25

Kym Anderson and Glyn Wittwer 227

assume that Rollo et alrsquos suggestion will be the new UK trade policy commitmentto WTO members that subsequent negotiations for preferential arrangements willtake years and that any consequent agreements with the European Union andothers will be gradually implemented after that

The impact of leaving a customs union on wine markets does not only come fromtariff changes Also relevant are any effects of leaving on real UK incomes and thevalue of the British pound If the United Kingdom were to move immediately to freetrade on all products (the Hong Kong option) its per-capita income could eventuallyrise but only after considerable adjustment Should the United Kingdom insteadcommit to the current EU tariff schedule at the WTO in the first instance as weassume below then its per-capita income growth rate and the British poundrsquosexchange rates almost certainly will be lower for some time ndash at least until newtrade agreement negotiations with the EU27 and others are sufficiently advanced

Table 2Taxes on British Wine Imports by Source 1660ndash1862 (British Pounds per Kiloliter)

France Germany Spain Portugal South Africa

1660ndash1665 7 9 8 81666ndash1684 7 9 8 81685ndash1691 14 20 19 181692ndash1695 22 20 19 181696 47 20 19 181697ndash1702 51 25 23 221703 52 27 24 231704ndash1744 55 31 26 251745ndash1762 63 35 30 291763ndash1777 71 39 34 331778 79 43 38 371779 84 41 40 391780ndash1781 92 49 44 431782ndash1785 96 51 47 46 441786 65 51 37 37 371787ndash1794 47 51 32 32 371795 78 64 51 51 571796ndash1797 108 92 71 71 771798 111 96 73 73 791799ndash1801 107 92 71 71 771802 112 97 74 74 801803 131 109 87 87 871804 142 117 95 95 951805ndash1824 144 119 96 96 961825ndash1830 78 50 50 50 251831ndash1859 58 58 58 58 291860 32 32 32 32 321861 16 21 21 21 211862 11 26 26 26 26

Source Summarized from Ludington (2013 Table A1)

228 UK and Global Wine Markets by 2025 and Implications of Brexit

Figure 4

UK Shares in Wine Exports of Key Wine-Exporting Countries 2010ndash2014 ()

Fig4-BW

onlin

eBW

inprint

Source Compiled from data in Anderson and Pinilla (2017)

Table 3Import Duties Excise Duties and VAT Affecting Consumer Prices of Wine and Other Alcohol

in the United Kingdom April 1 2017

VAT 20

MFN import duties on wine pound per liter-bottled still wine lt13 alc 0114-bottled still wine 13ndash15 alc 0134-bottled still wine 15ndash18 alc 0162-bulk still wine lt13 alc 0086-bulk still wine 13ndash15 alc 0105-bulk still wine 15ndash18 alc 0134-sparkling wine 0278Weighted averagea 0130

Excise duties on alcohol pound per liter-still wine lt15 alc 2887-still wine 15ndash22 alc 3848-sparkling wine 55ndash85 alc 2795-sparkling wine 85ndash15 alc 3697-spirits (assumed 40 alc) 11551-beer (assumed 5 alc) 0954

VAT on alcohol 20

a Assumes that one-third of UK wine imports arrives in bulk that half has less than 13 alcohol and that one-tenth of imports is sparkling

Sources HM Revenue httpswwwgovukgovernmentpublicationsalcohol-duty-rate-changes for excise duties and VAT accessed April 92017 and Wine Australia (2015) for import duties converted at the 2016 average exchange rate of pound0740634 per euro

Kym Anderson and Glyn Wittwer 229

as to restore investor and consumer confidence in the United Kingdomrsquos economyThe size of those impacts is uncertain of course (see Baldwin 2016 and especiallyCampos 2016) so we consider a range of possibilities below

Those assumed adverse macroeconomic effects will add to the initial impact ofaltered wine tariffs on aggregate wine consumption in the United Kingdom andhence on its bilateral trades in wine They will make the loss of sales to theUnited Kingdom by EU (and Chilean and South African) suppliers greater thanwould otherwise be the case And they will reduce the likelihood that other countriesrsquosales of wine in the United Kingdom will be higher than in the baseline Indeed themacroeconomic effects could outweigh the trade-diverting effects so that even coun-tries that are currently discriminated against by the EU28rsquos wine-trade policy may beworse off because of Brexit

IV Global Wine Markets Model and Database

Our model of the worldrsquos wine markets first published by Wittwer Berger andAnderson (2003) and revised by Anderson and Wittwer (2013) is summarized inthe Appendix It disaggregates wine markets into four types namely nonpremiumcommercial-premium and superpremium still wines and sparkling wines2 There aretwo types of grapes premium and nonpremium Nonpremium wine uses nonpre-mium grapes exclusively superpremium wines use premium grapes exclusivelyand commercial-premium and sparkling wines use both types of grapes to varyingextents across countries The world is divided into 44 individual nations and 7 com-posite geographic regions that capture all other countries

The modelrsquos database is calibrated to 2014 based on the comprehensive wine-market volume and value data and trade and excise-tax data provided inAnderson and Pinilla (2017) and in Anderson Nelgen and Pinilla (2017) It is pro-jected assuming that aggregate national consumption population and realexchange rates change between 2014 and 2025 to the extent shown in AppendixTable 13 The Brexit alternatives to that baseline also are projected to 2025

Concerning preferences we assume a continued considerable swing toward allwine types in China and a swing away from nonpremium wines in all other countriesuntil 2025

In our baseline scenario grape- and wine-industry total-factor productivity areassumed to grow at 1 per year everywhere while grape- and wine-industry

2Commercial-premium still wines are defined byAnderson Nelgen and Pinilla (2017) to cost between US$250 and US$750 per liter pretax at a countryrsquos border or wholesale3The real exchange rate changes over the projection period are the changes expected in the nominal valueof country irsquos currency relative to the US dollar times the expected ratio of the GDP deflator for theUnited States versus that for country i

230 UK and Global Wine Markets by 2025 and Implications of Brexit

capital is assumed to grow net of depreciation at 15 per year in China but zero else-where (consistent with the almost-zero growth in global wine production and con-sumption over the past two decades)

Two alternative scenarios are considered (ldquolargerdquo and ldquosmallrdquo) for the initialimpact of Brexit to provide a range of results We assume that because of theUnited Kingdomrsquos decision to exit the European Union the rate of UK realgross domestic product (GDP) growth is only one-third or two-thirds as fast overthe projection period (09 or 18 per year instead of 26) and the Britishpound will be 20 or 10 lower in real terms than in our modelrsquos core baselineprojection4 In both alternative scenarios we assume that the United Kingdomapplies the European Unionrsquos external tariffs on wine from March 30 2019 atthe end of the two-year period following the United Kingdomrsquos formal triggeringof Article 50 (see Table 3)

We then consider a subsequent impact of Brexit presumed to result from negoti-ating signing implementing and responding to an FTA with the EU27 by 2025(Chile and South Africa currently have preferential access to EU wine marketsbut we continue to assume in this scenario that the United Kingdom does not imple-ment new bilateral FTAs with them or others in our time frame)5 This subsequentscenario assumes the British pound returns to what it would have been in 2025 in theabsence of Brexit and that real incomes regain two-thirds of the value they wouldhave had without Brexit as compared with our ldquosmallrdquo initial Brexit scenario

This global model has supply-and-demand equations and hence quantitiesprices and price elasticities for each of the grape and wine products and for asingle composite of all other products in each country Income elasticities ofdemand also exist for each final product Grapes are assumed to not be traded inter-nationally but other products are exported and imported Each market is assumed tohave been in equilibrium before any shock and to find a new market-clearingoutcome following any exogenously introduced shock

4The nominal price of the British pound in US dollars in the fortnight following the Brexit vote on June23 2016 dropped 13 to US$130 and a year later the pound sat at the same rate having dropped to anaverage of $124 (an 18 devaluation) between October 2016 and April 2017 Our choice of a low of 10and a high of 20 aims to capture future possible rates while uncertainties remain The average real wagein the United Kingdom fell in the first half of 2017 and projected real GDP growth during 2018ndash2020 hasbeen revised down to 17 in the United Kingdomrsquos latest budget (HM Treasury 2017)5South Africa currently has duty-free access to the EU28 for just 50 megaliters (ML) of wine per yearbeyond which the MFN tariff rate applies South Africa currently exports around 320 ML to theEuropean Union one-third of which initially goes to the United Kingdom How that quota of 50 MLis divided between the United Kingdom and the EU27 is subject to future negotiation (Rollo et al2016 Swinbank 2017) In the following discussion we assume none of it is accepted by the UnitedKingdom so that all South African wine imports pay the United Kingdomrsquos MFN tariff under Brexitbut altering that assumption makes very little difference to our results for any countryrsquos trade exceptSouth Africarsquos

Kym Anderson and Glyn Wittwer 231

V Projecting Global Wine Markets to 2025

Global wine production and exports are projected in the baseline from 2014 to 2025consistent with past trends The modelrsquos global volume of production (and consump-tion) rises little over that 11-year period (9) made up of a 6 decline in nonpre-mium wine and a one-sixth rise in commercial-premium and superpremium wineIn real (2014 US$) value though global wine output and consumption increaseby about 50 in total and 60 in the two premium categories The internationaltrade projections are similar although a little larger with the share of global wineproduction exported (equivalent to the share of global consumption imported)rising 2 percentage points between 2014 and 2025

The baseline projection does not greatly alter the 2014 shares of various countriesin global wine production apart from China because we assume vineyard expansionthere is faster than elsewhere6 In value terms this assumption means that Chinamoves from fifth to fourth by 2025 behind France the United States and ItalySpain remains barely ahead of Australia and they and Germany take the nextthree places (Figure 5(a)) In total wine-production volume terms China movesfrom sixth to fifth place and Argentina drops from fifth to eighth (and fromeighth to ninth in value terms)

When their products are subdivided into fine wines (superpremium still plus spark-ing) commercial-premium wines and nonpremium wines France and the UnitedStates retain the highest two places on the global ladder for fine-wine productionand Spain and Italy retain the top two places for nonpremium wine As for commer-cial-premium wine production (defined to cost between US$250 and US$750 perliter pretax at a countryrsquos wholesale level or national border) Italy retains the topranking over our projections period but at least in terms of value China challengesFrance for second place

The country rankings by projected value of total wine consumption change some-what more than those for production by 2025 with China taking second place afterthe United States ahead of France and Germany and then the United Kingdomslightly overtaking Italy to slip into fifth place (Figure 5(b)) The United StatesFrance and Germany retain the top three rankings for consuming fine wine butCanada slightly overtakes Italy for fourth place in terms of value at least In thecase of commercial-wine consumption China strengthens its number-one positionahead of the United States and the United Kingdom does likewise vis-agrave-visGermany for third place

As for the projected changes in consumption volumes China is projected to dom-inate the increase in aggregate although the United States is projected to lead the

6 In fact Chinarsquos wine production fell steadily between 2012 and 2016 by a total of one-sixth so Chinarsquoswine imports may grow faster in practice than in this baseline projection

232 UK and Global Wine Markets by 2025 and Implications of Brexit

Figure 5

Value of Wine Production and Consumption in Key Countries 2014 and Projected Baseline for2025 (2014 US$ Million at WineryWholesale Pretax Prices)

Fig5-BW

onlin

eBW

inprint

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 233

increase in consumption of fine wine In Western Europe and in the SouthernHemispherersquos New World countries fine wines are projected to substitute for com-mercial wines (defined as the sum of commercial-premium and nonpremiumwines) with almost no change in total wine consumption Sub-Saharan Africa isthe next region that is projected to take off with its growth accounting for morethan one-third of the rest of the worldrsquos increase in volume consumed

Those differences in production versus consumption rankings are reflected ininternational trade Figure 6 shows that France Italy and Spain remain the threedominant exporters of wine in aggregate value but the rankings of the next fewchange with Australia being slightly ahead of Chile and the United StatesGermany and New Zealand being nearly tied for sixth place in value termsFrance and Italy are even more dominant in fine-wine exports and remain so by2025 while Italy outranks France in the commercial-premium export categoryand Spain outranks Italy Australia and Chile in the nonpremium export class

Among the importers the United States and the United Kingdom are projected tocontinue to hold the first two places in 2025 in value terms but China moves intothird place slightly ahead of Germany followed well behind by Canada HongKong Belgium-Luxembourg the Netherlands and Japan (Figure 6(b)) OtherAfrica (excluding South Africa) is projected to experience the largest increase inimports among all the other regions followed by Other Asia which becomes asbig as Germany in value terms (Figure 7(a)) In terms of total volume of wineimports Germany and the United Kingdom held the top two shares in 2014 butby 2025 the United Kingdom is projected to be well ahead of Germany (Figure 7(b)) However this projection ignores the effects of Brexit to which we now turn

VI How Might Wine Markets Be Affected by the United Kingdomrsquos Exitfrom the European Union

As mentioned earlier for our two alternative scenarios to capture the initial effects ofBrexit (ldquolargerdquo and ldquosmallrdquo) we assume that following the United Kingdomrsquos exitfrom the European Union the United Kingdomrsquos rate of economic growth wouldbe only one-third or two-thirds as fast for the period to 2025 the British poundwould be 20 or 10 lower in real terms than in our modelrsquos baseline projectionand the United Kingdom would apply the European Unionrsquos external tariff onwine to imports from EU member countries (as part of establishing MFN ratesvia the WTO to then start new bilateral FTA negotiations) In these initial scenarioswe assume the United Kingdom does not implement any new FTAs particularlywith the EU27 Chile and South Africa Following the discussion of those initialresults we present the results of a subsequent scenario that includes the implemen-tation of an FTA between the United Kingdom and the EU27 We assume these aretwo discrete steps with no agreed arrangements to smooth the transition betweenthem

234 UK and Global Wine Markets by 2025 and Implications of Brexit

Figure 6

Value of Wine Exports and Imports Key Wine-Trading Countries 2014 and Projected Baselinefor 2025 (2014 US$ million)

Fig6-BW

onlin

eBW

inprint

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 235

Figure 7

National Shares of Global Wine Import Value and Volume 2014 and Projected Baselinefor 2025 ()

Fig7-BW

onlin

eBW

inprint

Source Authorsrsquo model results

236 UK and Global Wine Markets by 2025 and Implications of Brexit

A Initial Impact of Brexit

Because the British pound dropped by one-sixth against the US dollar in the fourmonths following the Brexit vote in June 2016 and because the UnitedKingdomrsquos average real wage fell 1 between November 2016 and April 2017(according to the latest available data see Office of National Statistics 2017) weviewed our ldquolargerdquo scenario as more likely soon after the Brexit vote But wecompare those results with ones from our ldquosmallrdquo initial-impact scenario becauseit has become more likely following the June 2017 general election that ldquosofterrdquooptions will be considered by the new government Generally the results areabout half the size in the ldquosmallrdquo scenario with the exception of the bilateraltrade effects To show the sensitivity of results to our assumptions we point outthe differences when the ldquosmallrdquo results are not close to half the results shown forthe ldquolargerdquo scenario

Before turning to those results we first present a ldquofree-traderdquo scenario whichassumes the United Kingdom chooses the radical option of becoming the HongKong of Europe Unlikely as this scenario is the results provide assurance thatour global wine model generates the standard types of effects of exiting a customsunion In this ldquofree-traderdquo scenario the United Kingdom is assumed to move tozero tariffs on all wine imports Thus the current preferences on EU Chileanand South African wine imports disappear as all other wine exporters enjoy thesame free access to UK wine markets We assume in this scenario that Brexit hasno impact on the United Kingdomrsquos real income or consumption growth rates by2025 and that by then the British pound has returned to its 2014 value relative toother currencies (similar to the real exchange rate assumed in our baseline scenarioto 2025 see Appendix Table A1)7 Under these assumptions EU Chilean andSouth African wine exports to the United Kingdom in 2025 are 003 lower thanin the no-Brexit baseline and UK imports from other wine exporters are 025higher but the United Kingdomrsquos total wine imports are hardly any different(because local wine prices fall by only 05 and incomes are unchanged) In this sce-nario UK consumers expand their volume of consumption by just 05 while theEU27 exports more and other countries export less to the rest of the world such thatworld wine exports are almost unchanged In short all these effects are in the direc-tion that comparative static-customs-union theory would predict but they are smallbecause we assume the United Kingdomrsquos real income and currency are the same asin our baseline for 2025 as described in the previous section

In the ldquolargerdquo alternate initial scenario involving Brexit as compared with thebaseline scenario to 2025 the consumer price of wine in 2025 is 22 higher in the

7If this scenario were seen as a credible long-run trade policy it would encourage more investment thatwould eventually raise UK incomes but not without major structural changes that would initiallydisrupt the economy Implicitly we assume in our free-trade scenario that these two opposite impactson real UK incomes and the British pound exactly offset each other by 2025

Kym Anderson and Glyn Wittwer 237

United Kingdom in local currency terms (20 because of real depreciation of theBritish pound 4 because of new tariffs on EU Chilean and South Africanwines and ndash2 because of slower UK income growth) The volume of UKwine consumption is 28 lower 16 because of slower UK economic growth7 because of real depreciation of the British pound and 5 because of newtariffs Superpremium still-wine sales are the most affected dropping by two-fifths while sparkling and commercial-premium wines drop a bit less than one-quarter Because the average price rises by more than the decline in volume soldthe aggregate value of UK sales even in local currency terms falls under thisldquolargerdquo Brexit scenario Under the ldquosmallrdquo Brexit scenario the consumer price ofwine in 2025 is 11 higher in the United Kingdom and its volume of wine consump-tion is 17 lower

The volume of projected UK imports in 2025 is 427 ML or nearly one-quarterless in the ldquolargerdquo scenario than in the baseline scenario comprising 58ML less spar-kling wine 31 ML less superpremium still wine and 339 ML less commercial-premium wine World imports are lower by just 239 ML because imports byother countries are 189 ML higher in response to the lower international wineprices in this scenario In value terms UK imports are $175 billion (or 27)lower in 2025 because of ldquolargerdquo Brexit $113 billion because of lower incomes$038 billion because of the fall in the British pound and $014 billion because ofthe rise in wine-import tariffs (Table 4) These aggregate trade impacts are a littlemore than half this size under the ldquosmallrdquo Brexit scenario

Despite the import levels falling because of raised import tariffs domestic con-sumption of all three quality categories of UK-produced wine is lower with thanwithout Brexit because of shrunken demand for all wines resulting from loweredUK incomes and raised local prices because of devaluation of the British poundThe British poundrsquos devaluation does make it easier for the United Kingdom tosell wines abroad though The countryrsquos exports are 7 ML or nearly 5 higherin 2025 in the ldquolargerdquo Brexit scenario and UK production is 3 higher ThoseUK exports (or re-exports of imported bulk wine after it is bottled in the UnitedKingdom) that go to EU27 countries are reduced though because of the tariffnow imposed at the new EU border

Without Brexit the United Kingdomrsquos shares of global wine imports are slightlyhigher in volume terms in 2025 than in 2010ndash2015 but 2 percentage points lower invalue terms thanks to East Asiarsquos expanding demand for imports of premium winesWith ldquolargerdquo Brexit however that value share is an additional 2 percentage pointslower and the volume share is almost 5 points lower (Figure 8) The net effects ofthese impacts on global trade are shown in Figure 9 Most of the initial tradeeffect of Brexit is a large decline in net imports of wine by the United Kingdomwith very little offsetting positive effect on trade in the rest of the world Theldquosmallrdquo Brexit numbers are a bit more than half of those for ldquolargerdquo Brexit

238 UK and Global Wine Markets by 2025 and Implications of Brexit

Table4

Difference

in2025

Projected

VolumeandValue

ofWineIm

portsb

ytheUnitedKingdom

andtheResto

fthe

World

asaConsequence

oftheInitial

BrexitS

hock

(MLand2014US$

Million

ldquoLargerdquoSc

enario)

Volum

e(M

L)

Value

(US$

million)

NP+CPa

SuperPrb

Sparkling

TOTAL

NP+CPa

SuperPrb

Sparkling

TOTAL

ΔUK

impo

rtsdu

eto

Low

erincomes

minus198

minus20

minus29

minus247

58minus644

minus253

minus234

minus1131

65Low

erpo

und

minus70

minus10

minus14

minus93

22minus248

minus127

minus102

minus476

27Highertariffs

minus71

minus1minus16

minus87

20minus110

minus8minus2

4minus143

8TOTA

Lminus339

minus31

minus58

minus427

100

minus1001

minus388

minus360

minus1750

100

difffrom

base

2332

3325

2432

3227

of

totalc

uts

797

14100

5722

21100

ΔROW

netim

ports

143

2125

189

230

143

minus181

192

ΔWORLD

TRADE

minus195

minus10

minus34

minus239

minus763

minus246

minus543

minus1552

aNon

prem

ium

plus

commercial-premium

still

wines

bSu

perpremium

still

wines

SourceA

utho

rsrsquomod

elresults

Kym Anderson and Glyn Wittwer 239

The aggregate effect of ldquolargerdquo Brexit on the market shares of various wine-export-ing countries in the United Kingdom is almost indiscernible even with one decimalpoint The projected 2025 shares are quite different from the actual 2014 shares forseveral countries They are much smaller in 2025 for South Africa Australia andNew Zealand (and the United States in volume terms) and are much larger involume for Spain and in value for Italy These results occur because wine-exportingcountries benefit differentially from the varying rates of growth in net importdemand for wine in non-UK countries over this projection period The most impor-tant projected changes are the increase in the real value of annual wine importsbetween 2014 and 2025 by China (200 or $3 billion) Other Asia (110 or $22billion) and Africa (270 or $16 billion) More than half of Australiarsquos increasein annual exports from 2014 to 2025 go to Asia and more than half of SouthAfricarsquos increase in exports go to Other Africa

Table 5 reveals that ldquolargerdquo Brexit lowers EU Chilean and South African wineexports by 150 ML or US$12 billion in the case of the European Union with someof their exports diverted from the United Kingdom to the EU27 and other marketsin competition with New World exporters Although the United States Australiaand Argentina sell only a little less to the United Kingdom they also sell less toother countries For Chile and South Africa which lose their preferential access toUK (but not to EU27) markets in this Brexit scenario some exports are redirected

Figure 8

UK Shares of World Wine Imports 2010ndash2015 and Projected to 2025 without and with theldquoLargerdquo Initial Brexit Shock ()

Fig8-BW

onlin

eBW

inprint

Sources Anderson and Pinilla (2017) and authorsrsquo model results

240 UK and Global Wine Markets by 2025 and Implications of Brexit

from the United Kingdom to EU27 countries ndash but again they export less overallGlobal wine trade in 2025 is 240 ML (19) or $18 billion (35) less under thisldquolargerdquo Brexit scenario The percentage by which wine exportersrsquo trade shrinks isgreater for values than for volumes because of changes in relative prices of differ-ent-quality wines Those differences are shown in the numbers in parentheses inTable 5

Three other points are worth making about Table 5 One is that Australia sellsslightly more to the United Kingdom in the ldquosmallrdquo Brexit scenario rather thanslightly less as in the ldquolargerdquo Brexit case Evidently the negative income and price(devaluation) effects do not more than offset the positive trade-diverting effect onAustralian exports to the United Kingdom of removing preferences in the ldquosmallrdquoscenario Second New Zealand sells slightly more to non-UK countries underBrexit despite greater competition from the EU27 Chile and South Africa Thisanomaly is due to changes in the relative prices of different qualities of wine inglobal wine markets bearing in mind that New Zealand has the worldrsquos highestaverage price for still-wine exports And third the value (but not the volume) ofexports of ldquoOtherrdquo countries to markets other than the United Kingdom arehigher under Brexit This result too is due to changes in the relative prices of differ-ent qualities of wine in global wine markets

Figure 9

Difference in 2025 Wine Import Volumes and Values as a Result of the ldquoLargerdquo Initial BrexitShock (ML and US$ Million in 2014 US dollars)

Fig9-BW

onlin

eBW

inprint

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 241

B Subsequent Impact of Brexit from a UK-EU27 FTA

The next-most-likely step in the Brexit process is for the United Kingdom to nego-tiate a new trade arrangement with the EU27 We therefore assume that a UK-EU27FTAwith free bilateral wine trade is implemented and adjusted to by 2025 and thatprogress toward that end occurs soon enough that the adverse macroeconomicshocks from the initial impact of uncertainty over the Brexit process are confinedto those assumed in the ldquosmallrdquo scenario outlined above In this subsequent scenariowe assume that the British pound returns to the value it would have reached in 2025in the absence of Brexit (ie reversing the 10 devaluation assumed in the ldquosmallrdquoinitial Brexit scenario) and that real incomes in the United Kingdom are 8 higher

Table 5Difference in 2025 BilateralWine Import Volumes and Values fromKey Exporters by the United

Kingdom and the Rest of the World (RoW) as a Result of Initial Brexit Shock(ML and 2014US$ Million)a

(a) ldquolargerdquo scenario

Volume (ML) Value (2014US$ million)

UK RoW WORLD () UK RoW WORLD ()

EU27 minus287 136 minus150 (ndash17) minus1187 minus5 minus1192 (ndash31)Chile minus59 35 minus25 (ndash30) minus169 31 minus138 (ndash48)South Africa minus53 35 minus18 (ndash32) minus105 20 minus85 (ndash67)USA minus7 minus6 minus13 (ndash24) minus75 minus40 minus115 (ndash50)Australia minus4 minus3 minus7 (ndash09) minus25 minus65 minus90 (ndash30)Argentina minus3 minus9 minus12 (ndash48) minus16 minus39 minus55 (ndash52)New Zealand minus11 9 minus2 (ndash09) minus162 71 minus91 (ndash43)Others minus2 minus10 minus12 (ndash02) minus11 minus52 minus63 (ndash44)WORLD minus427 187 minus240 (ndash19) minus1750 minus79 minus1829 (ndash35)

(a) ldquosmallrdquo scenario

Volume (ML) Value (2014US$ million)

UK RoW WORLD () UK RoW WORLD ()

EU27 minus178 82 minus96 (ndash12) minus692 minus43 minus736 (ndash19)Chile minus46 28 minus18 (ndash24) minus128 36 minus91 (ndash32)South Africa minus43 29 minus14 (ndash42) minus82 23 minus59 (ndash47)USA 1 minus6 minus5 (ndash11) minus23 minus28 minus51 (ndash22)Australia 5 minus10 minus5 (ndash06) 19 minus56 minus38 (ndash13)Argentina 0 minus6 minus6 (ndash26) minus3 minus25 minus29 (ndash27)New Zealand minus5 4 minus1 (ndash06) minus80 34 minus46 (ndash22)Others 0 minus9 minus9 (ndash01) minus1 minus33 minus34 (ndash24)WORLD minus266 112 minus154 (ndash13) minus991 minus92 minus1083 (ndash21)

a Numbers in parentheses are the percentage difference between the Brexit and baseline scenarios for 2025 projected wine-import volumes orvalues by source

Source Authorsrsquo model results

242 UK and Global Wine Markets by 2025 and Implications of Brexit

in 2025 than in the ldquosmallrdquo scenario ndash that is they regain two-thirds of the value theywould have reached without Brexit-related uncertainty

This subsequent development in the Brexit process reverses most of but not allthe initial effects of Brexit by 2025 because of our assumption that the lost growthin the initial years of uncertainty following the Brexit vote are only partly recoveredby 2025 following the implementation of a UK-EU27 FTA Moreover the longer ittakes before this FTA is finalized and implemented the longer the estimated initialadverse effects persist and the larger the cumulative cost of Brexit to UK wine con-sumers and to grape and wine producers in wine-exporting countries

Table 6 summarizes the subsequent trade effects for 2025 It suggests that all butone-tenth of the loss in value of world trade in wine from the initial ldquosmallrdquo impact isrestored most of it because of a smaller reduction in the United Kingdomrsquos wineimports following the FTA Most of that improved outcome is because of recoveredimports from the EU27 commensurate with the latterrsquos high share of UK importsEven though Chile and South Africa are assumed in this scenario to have not yetsigned an FTA with the United Kingdom they export slightly more to the UnitedKingdom (and even more to all other countries) than in the initial ldquosmallrdquo scenario

Within the United Kingdom this FTA brings down the local currency consumerprice of wine by 9 largely offsetting the 11 rise in the initial ldquosmallrdquo Brexit sce-nario and it raises the volume consumed in the United Kingdom by 18 fully off-setting the 17 fall in the initial ldquosmallrdquo Brexit scenario

Table 6Difference in 2025 Bilateral Wine-Import Volumes and Values from Key Exporters by the

United Kingdom and the Rest of the World (RoW) as a Result of Implementing a UK-EU27FTA Following Initial ldquoSmallrdquo Brexit Shock (ML and 2014US$ Million Difference Relative to

Initial ldquoSmallrdquo Brexit Scenario)a

Volume (ML) Value (2014US$ million)

UK RoW WORLD (ldquosmallrdquo)a UK RoW WORLD (ldquosmallrdquo)a

EU27 212 minus112 100 (minus96) 750 0 750 (minus806)Chile 1 3 4 (minus18) 3 40 43 (minus92)South Africa 3 2 5 (minus14) 3 18 21 (minus60)USA 3 3 6 (minus5) 26 21 47 (minus50)Australia 1 1 2 (minus5) 8 38 46 (minus33)Argentina 1 4 5 (minus6) 4 20 24 (minus28)New Zealand 4 minus3 1 (minus1) 61 minus25 36 (minus46)Others 1 6 7 (minus9) 5 minus46 minus41 (92)WORLD 226 minus96 130 (minus154) 860 66 926 (minus1022)

a Numbers in parentheses are the world trade differences between the ldquosmallrdquo initial Brexit scenario and the baseline scenario copied fromcolumns 3 and 7 of Table 5(b)

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 243

In short this subsequent step in the Brexit process can be expected to restore muchof the initial adverse effects in the United Kingdom of the Brexit vote by 2025 Thewine exporters in the EU27 Chile and South Africa also lose less in this scenarioas compared with the initial ldquosmallrdquo scenario while other key exporters sell nearlyas much in 2025 in this scenario as they are projected to in the baseline ndash that isthey are beneficiaries of the reduced discrimination in global wine markets in thiscase where Chile and South Africa no longer enjoy preferential access to the UKmarket (and even though EU27 producers are again allowed to do so) Keep inmind though that although the estimated losses in the initial scenarios are reportedjust for 2025 they are expected to be felt throughout the preceding years of uncertaintyTherefore the longer it is before uncertainty abates and the United Kingdomrsquos currentpreferential trading arrangements are replaced by newones such as a UK-EU27 FTAthe higher the cumulative cost of the Brexit vote to the wine trade

VII Caveats and Conclusions

The above Brexit simulations are just a few of many scenarios that could be modeledObvious additional ones could also assume that FTAs are reached between theUnited Kingdom and other trading partners including Chile and South Africaand such countries as Australia and New Zealand Some of these other countriesare already exploring the ECrsquos invitation (see European Commission 2015) to nego-tiate an FTA with what will be the EU27 The sequence in which FTAs are signedand the speed with which they are implemented will matter (as was also the casewith the sequential signing over the past decade of bilateral FTAs with NortheastAsian countries by Chile Australia and New Zealand see Anderson andWittwer 2015) Even if the United Kingdom were able to sign additional bilateralFTAs and begin implementing them before 2025 it would have little effect on theabove results (because wine tariffs are such a minor contributor to them) unlessthose FTAs were to accelerate the United Kingdomrsquos economic growth and therise in the British poundrsquos value before 2025

We assume above that no changes are made to alcohol excise duties in the UnitedKingdom following Brexit even though they are scheduled to be progressively raisedwith inflation and may be raised even more for wine relative to spirits to offset theopposite effects of Brexit on those two domestic industries Nor do we make anyallowance for UK increases in consumer wine prices that may be needed to coverthe higher cost of clearing customs on imports from EU27 countries (which maybe more or less offset by increased smuggling and duty-free purchasing)Importers of fine wines including individuals who buy Bordeaux en primeur maywell reduce their demand because of concerns that delays on docks will affect thequality of their wines or because of greater currency uncertainty

To summarize Brexit is costly initially to UK consumers of wine (and of manyother tradable products) because the domestic retail price in local currency tax-

244 UK and Global Wine Markets by 2025 and Implications of Brexit

inclusive terms is 22 higher than otherwise in 2025 and the volume of wine con-sumed domestically is 28 lower in the ldquolargerdquo Brexit scenario (or 11 higher pricesand 17 lower quantities in the ldquosmallrdquo case) by 2025 unless a UK-EU27 FTA isdeveloped before then Even if such an FTA does get signed ratified by all 28 par-liaments and implemented by 2025 the slower income growth in the interim meansa smaller UK wine market in 2025 than would otherwise have been the case Thevolume reduction is a blow to many participants in UK wine-bottling transportingstoring wholesaling and retailing businesses in addition to restaurants and pubsVery little of that initial impact occurs because of higher import tariffs most impor-tant is the assumed fall in UK real incomes

The small but growing number of local UK vignerons (who supply less than 05of all domestic-wine sales and less than 4 of sparkling-wine sales) is initially pro-jected to sell less wine domestically because of reduced demand and to export only alittle extra abroad while the British pound is devalued Overall their production isonly 3 higher in 2025 even in the ldquolargerdquo Brexit scenario (not taking intoaccount the fact that their casual labor is likely to be more expensive under Brexitbecause of tighter restrictions on immigration) and their production diminishes ifand when a UK-EU27 FTA is in place

Even with a UK-EU27 FTA in place EU27 wine exporters export US$56 millionless wine in 2025 thanks to Brexit Chile and South Africa export $88 million lesswine and wine exports from the rest of the world are little different in aggregatealthough slightly larger from Australia

Clearly there will be great uncertainly for some time over the possible policy out-comes to flow from Brexit and their consequent sequential impacts on UK house-hold disposable incomes foreign exchange rates and bilateral wine tariffsMeanwhile the above projections under explicit assumptions provide some idea ofhow wine markets might be affected by the most-likely first two stages of theBrexit process (agreeing on a new tariff schedule at the WTO and agreeing to andimplementing a UK-EU27 FTA) In particular they make clear that nontrivialinitial impacts could affect the domestic wine market impacts that are likely to belarger than just the direct impact of changes in bilateral tariffs If the UnitedKingdom succeeds in getting countries to agree to transition arrangements thatdelay the changes in tariffs until new FTAs are signed and ratified the initialeffects will be less dramatic than in our first scenarios but it remains to be seenwhether any such agreements can be reached In any event the net effect of Brexiton the welfare of the worldrsquos consumers and producers of wine as a whole will benegative not just initially but permanently unless new trade policy commitmentsby the United Kingdom with major wine-exporting countries are sufficiently moreliberal than current arrangements

Kym Anderson and Glyn Wittwer 245

Appendix Key Equations in the Global Wine Markets Model

Amodel of the worldrsquos wine markets was first published byWittwer et al (2003) andmuch improved by Anderson and Wittwer (2013) Several significant enhancementshave been made to that original model Wine types have been disaggregated intomore types ndash namely nonpremium (including generic bulk) commercial-premiumand superpremium still wines and sparkling wines8 As in the original modelthere are two types of grapes premium and nonpremium Nonpremium wines usenonpremium grapes exclusively superpremium wines use mainly premium grapesand commercial-premium and sparkling wines use both types of grapes In theregional dimension the number of countries and country groups has expandedfrom 10 in the original model to 51 44 individual nations and 7 compositeregions The modelrsquos database is calibrated to 2014 for this paper based on datain Anderson et al (2017) trade data for which are downloaded from httpscomtradeunorg The modelrsquos supply-and-demand equations are based on the orig-inal ORANI modelrsquos theory (see Dixon Parmenter Sutton and Vincent 1982) andthe model is implemented using GEMPACK software (Harrison Horridge Jerieand Pearson 2014)

An enhancement of importance to the present study is the inclusion of exchange-rate variables in the model which allow a distinction between price impacts asobserved in local currency units and those observed in 2014 US dollars

In the model the grape and wine sectors minimize costs of intermediate inputssubject to weak constant elasticity of substitution (CES) substitutability betweeninputs By assumption no intermediate inputs are imported from other countries9Hence

Xcid frac14 f ethX1id CESfrac12P

cid=P1id $THORN eth1THORN

P1id X1id frac14X

c

Xcid P

cid eth2THORN

whereXcid is the quantity demanded of commodity c by grape or wine industry i in

region d Pcid is the corresponding price and X1id and P1id are the respective inter-

mediate composite quantities and prices

Two primary factors are employed in the sector labor (the quantity of which isendogenous with perfectly elastic supply) and capital Capital is usually treated asexogenous in quantity with rates of return bearing all the adjustment in thevarious scenarios This reflects the fact that grapes (a perennial crop) and wine-

8Commercial-premium still wines are defined by Anderson and Nelgen (2011) to be those priced betweenUS$250 and $750 per litre pre-tax at a countryrsquos border or wholesale9An exception concerns bulk wine imports used mainly in the bottling of commercial premium wine in theUK USA Canada Japan and Germany

246 UK and Global Wine Markets by 2025 and Implications of Brexit

plant capacity adjust slowly to market signals

Lid frac14 f ethFid CESfrac12Wid=PFid $THORN eth3THORN

Kid frac14 f ethFid CESethRid=PFid $THORN eth4THORN

PFid Fid frac14 Lid Wid thorn Kid Rid eth5THORN

Grape and wine producers are assumed to minimize costs subject to CES substitu-tion between capital and labor Equations (3) to (5) show primary factor demandsfor the labor composite Lid and capital Kid subject to a composite factor demandFid by industry i in region d The factor prices are Wid for labor Rid for capitalrentals and PFid for composite factor prices

Appendix Table A1Cumulative Consumption and Population Growth Rates and Changes in the Real Exchange Rate

(RER) Relative to the US Dollar 2014 to 2025 without Brexit ()

Aggregateconsumption Poprsquon RER

Aggregateconsumption Poprsquon RER

France 18 4 minus11 Australia 35 11 minus17Italy 11 2 minus9 New Zealand 32 9 minus26Portugal 14 0 minus9 Canada 27 8 minus18Spain 26 8 minus9 United States 31 8 0Austria 19 4 minus7 Argentina 7 10 109Belgium 20 7 minus9 Brazil 16 8 minus29Denmark 22 2 minus9 Chile 55 8 minus2Finland 21 3 minus7 Mexico 42 12 minus8Germany 14 minus2 minus11 Uruguay 45 3 1Greece 22 minus1 minus14 Other Latin America 60 10 minus5Ireland 42 12 minus9 South Africa 36 12 minus1Netherlands 21 4 minus9 Turkey 50 8 20Sweden 24 9 minus13 North Africa 53 11 0Switzerland 18 8 minus6 Other Africa 109 18 84United Kingdom 32 6 1 Middle East 52 18 minus12Other West Europe 21 10 minus1 China 79 3 5Bulgaria 41 minus7 7 Hong Kong 42 3 2Croatia 20 minus2 minus1 India 134 13 17Georgia 35 0 23 Japan 11 minus3 minus24Hungary 25 minus3 minus11 Korea 38 1 minus9Moldova 49 minus11 13 Malaysia 62 15 minus16Romania 45 minus4 22 Philippines 75 18 7Russia 18 minus2 minus8 Singapore 44 21 minus22Ukraine 22 minus5 14 Taiwan 29 1 minus13Other East Europe 40 minus5 48 Thailand 47 3 minus9

Other Asia 99 10 10

Source Authorsrsquo compilation from projections by various international agencies and from global economy-wide modeling by Anderson andStrutt (2016)

Kym Anderson and Glyn Wittwer 247

The composite factor demand Fid is proportional to total output Qid subject to aprimary factor using technology Aid Hence

Fid frac14 Qid Aid eth6THORN

The perfectly competitive zero-pure-profit condition is that total revenue valued atthe output price P0s

i multiplied by Qid equals the total production cost

P0si Qid frac14

X

cPcid X1cid thorn

X

oWo

id Loid thorn Rid Kid eth7THORN

Household demands follow a linear expenditure system in each region We reducethe optimizing problem for household consumption of each commodity subjectto a budget constraint to equations describing subsistence and discretionarydemands Aggregate subsistence expenditure WSUBd depends only on consumerprices P3cd for each commodity and the number of households N as per-capita sub-sistence quantities XSUBcd subject to given preferences are constant

WSUBd frac14X

c

P3cd XSUBcd Nd eth8THORN

Discretionary expenditures for each commodity (the left-hand side of equation (9))are equal to the marginal budget share (βcd) of aggregate discretionary expenditureThis aggregate is the bracketed term on the right-hand side of equation (9) whereW3TOTd is aggregate nominal expenditure

P3cdethX3cd XSUBcd NdTHORN frac14 βcdethW3TOTd WSUBdTHORN eth9THORN

Because real aggregate consumption is usually exogenous in our partial equilibriumsimulations the linear expenditure system determines the consumption shares ofindividual final commodities (ie the five wine types plus a composite of all otherconsumption items) driven by changes in relative prices as faced by domestic con-sumers The income elasticity of demand for each commodity is equal to the mar-ginal budget share divided by the expenditure share This number varies from 05for nonpremium wine to 17 for superpremium still wine The income elasticity ofdemand for other consumption is very close to 10 because wine accounts for anaverage of only 03 of aggregate expenditures globally and no more than 11in any country (Anderson and Nelgen 2011 Table 166)

A feature of our revised model of world wine markets is the inclusion of nominalexchange rates These appear directly in the equation linking retail prices (P3scd ) toproducer prices by country of origin (P0s

c ) where c denotes the wine type

P3 scd frac14 P0s

cfd

fsTtarcd T

taxcd thorn Pm

cd eth10THORN

248 UK and Global Wine Markets by 2025 and Implications of Brexit

The exchange rates in the consuming and producing regions are fd and fs respec-tively expressed as local currency units per $US Ttar

cd is the power of the tariff in theconsuming region and Ttax

cd is the power of the domestic consumption (or excise) taxprior to any generic value-added or goods-and-services tax Pm

cd is the price of marginm assumed to be locally supplied nontradable and therefore unaffected by theexchange rate

A given level of consumption for wine type c (X3cd) is satisfied using theArmington (1969) assumption in which wines from different countries of originare imperfectly substitutable First domestic wine is imperfectly substitutable witha composite of imports

X3sscd frac14 f ethX3cd CESethP3sscd=P3cdTHORNTHORN ss frac14 domestic imports eth11THORN

and then imports by origin (X3scd ) are determined in a second CES equation

X3scd frac14 f ethX3ssfrac14importscd CESethP3scd=P3

ssfrac14importscd THORNTHORN eth12THORN

The model enables us to show how changes in international competitiveness affectthe worldrsquos wine markets A crucial part of this exercise is explaining how pricesdetermined outside the grape and wine markets influence these markets Becausethe model is in partial equilibrium to depict the impacts of changes in internationalcompetitiveness outside price changes need to be imposed as shocks on the modelThe price of intermediate inputs other than grapes and wine shown in equations (1)and (2) is set equal to the price of GDP (Pg

d ) multiplied by a shifter Fcd

Pcid frac14 Fc

dPgd eth13THORN

If no specific price observations are available the shifter Fcd remains exogenous and

unshocked with the change in price being determined by a shock to the price ofGDP If observations are available for specific input price movements the shifterFc

dbecomes endogenous with Pc

id now exogenous and shocked

Wid frac14 Fwd P

gd eth14THORN

Wage rates are treated similarly In equation (14) if the wage shifter Fwd is exogenous

changes in wage rates Wid are determined by changes in the price of GDP If wagerate data are available Fw

d becomes endogenous andwage rates are shocked directly

Pmcd frac14 Fm

d Pgd eth15THORN

The prices of trade and transport margins are also determined by the price of GDP ifthe shifter Fm

d in equation (15) is exogenous

Changes in international competitiveness depend on changes in relative pricelevels and changes in nominal exchange rates In equation (16) fR

s denotes real

Kym Anderson and Glyn Wittwer 249

exchange-rate movements relative to the US dollar

fRs frac14 Pg

s =frac12PgUSA

fs$ eth16THORN

In equation (16) the nominal exchange rate for the United States is alwaysunchanged because nominal and real exchange rates are expressed relative to theUS currency

Changes in international market conditions may have impacts in one direction onproducer prices as expressed in US dollars P0s

i and potentially in the opposite direc-tion in local currency units Hence we calculate real producer pricesP0s

iloc in localcurrency terms (ie the price most relevant to domestic producers)

P0siloc frac14 P0s

ifs=P

gs eth17THORN

To obtain real price changes in local currency terms we convert US dollar prices(P3 s

cd for the source-specific price and P3cd for the source-composite price) to reallocal currency prices (P3 s

cdloc and P3cdloc) using the consumer price index (CPI)(Pc

d ) as the deflator

P3 scdloc frac14 P3 s

cdfd=P

cd eth18THORN

and

P3cdloc frac14 P3cdfd=Pcd eth19THORN

References

Anderson K and Nelgen S (2011) Global Wine Markets 1961 to 2009 A StatisticalCompendium Adelaide University of Adelaide Press Also available as an e-book atwwwadelaideeduaupresstitlesglobal-wine

Anderson K Nelgen S and Pinilla V (2017) Global Wine Markets 1860 to 2016 AStatistical Compendium Adelaide University of Adelaide Press (forthcoming) Also tobe available as an e-book at wwwadelaideeduaupress

Anderson K and Pinilla V (with the assistance of A J Holmes) (2017)Annual Database ofGlobal Wine Markets 1835 to 2016 Wine Economics Research Centre University ofAdelaide to be posted in October at wwwadelaideeduauwine-econdatabases

Anderson K and Strutt A (2016) Impacts of Asiarsquos rise on African and Latin Americantrade Projections to 2030 World Economy 39(2) 172ndash194

Anderson K andWittwer G (2013) Modeling global wine markets to 2018 Exchange ratestaste changes and Chinarsquos import growth Journal of Wine Economics 8(2) 131ndash158

Anderson K and Wittwer G (2015) Asiarsquos evolving role in global wine markets ChinaEconomic Review 35 1ndash14

250 UK and Global Wine Markets by 2025 and Implications of Brexit

Armington P A (1969) A theory of demand for products distinguished by place of produc-tion IMF Staff Papers 16 159ndash178

Baldwin R E (ed) (2016) Brexit Beckons Thinking Ahead by Leading Economists LondonCentre for Economic Policy Research

Campos N F (2016) Lousy experts Looking back at the ex ante estimates of the costs ofBrexit In R E Baldwin (ed) Brexit Beckons Thinking Ahead by Leading Economists(ch 3) London Centre for Economic Policy Research

Dixon P Parmenter B Sutton J and Vincent D (1982) ORANI A Multisectoral Modelof the Australian Economy Contributions to Economic Analysis 142 Amsterdam North-Holland

European Commission (2015) Trade for All Towards a More Responsible Trade andInvestment Policy Brussels European Commission

Harrison J Horridge M Jerie M and Pearson K (2014) GEMPACK ManualMelbourne GEMPACK Software

HM Treasury (2017) Spring Budget 2017 London HM TreasuryHolmes A J and Anderson K (2017a)Annual Database of National Beverage Consumption

Volumes and Expenditures 1950 to 2015 Wine Economics Research Centre University ofAdelaide posted at wwwadelaideeduauwine-econdatabases

Holmes A J and Anderson K (2017b) Convergence in National Alcohol ConsumptionPatterns New Global Indicators Journal of Wine Economics 12(2) 117ndash148

Ludington C C (2013) The Politics of Wine in Britain A New Cultural History BasingstokeUK Palgrave Macmillan

Office of National Statistics (2017) Analysis of real earnings London Office of NationalStatistics Available at httpswwwonsgovukemploymentandlabourmarketpeopleinworkearningsandworkinghoursarticlessupplementaryanalysisofaverageweeklyearningslatest

Rollo J Borchert I Dawar K Holmes P and Winters L A (2016) The World TradeOrganisation A safety net for a post-Brexit UK trade policy Briefing Paper 1UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-1-final-1pdf

Smith A (2017) Brexit Hard truths and hard choices UKTPO blog 19 June Available athttpsblogssussexacukuktpo20170619brexit-hard-truths-and-hard-choices

Swinbank A (2017) World trade rules and the policy options for British agriculture post-Brexit Briefing Paper 7 UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-7pdf

Viner J (1950) The Customs Union Issue New York Carnegie Endowment for InternationalPeace

Wine Australia (2015) Export Market Guide European Union Adelaide Wine AustraliaWittwer G Berger N and Anderson K (2003) A model of the worldrsquos wine markets

Economic Modelling 20(3) 487ndash506

Kym Anderson and Glyn Wittwer 251

Page 5: U.K. and Global Wine Markets by 2025, and Implications of ... · U.K. and Global Wine Markets by 2025, and Implications of Brexit* Kym Andersona and Glyn Wittwerb Abstract The United

liter is escalated to 800 pence for off-trade sales and more than 1300 pence for on-trade sales (which are about one-fifth of the total sales volume in the UnitedKingdom) The share of that latter retail price that is due to tariffs on wineimports is thus just 1 The change in the United Kingdomrsquos import trade regimefrom imposing such tariffs on wines currently imported free of duty from theEuropean Union Chile and South Africa is therefore likely to be very minorHowever as becomes clear below the effect of Brexit on wine sales involves farmore than just the trade-reducing and trade-diverting effects of altering bilateralimport tariffs

III The Economics of Leaving a Customs Union

The standard theory of customs unions (Viner 1950) focuses on the fact that whencountries join a union and impose a common external tariff on imports from non-union countries net trade can be created (depending on the height of the commonexternal tariff relative to the previous national tariffs) but trade can also be diver-sified (because of the preference to producers within the union) When a country

Figure 2

UK Shares of Value of World Merchandise Imports and of Value and Volume of World WineImports 1850 to 2015 ( 3-Year Averages to Year Shown)

Fig2-Colou

ron

line

BW

inprint

Source Compiled from data in Anderson and Pinilla (2017)

Kym Anderson and Glyn Wittwer 225

leaves a union the reverse happens because the leaving countryrsquos tariffs now apply toits imports from union countries as well as from the rest of the world Hence importsfrom union countries fall because other things equal these countriesrsquo preferentialaccess to the leaving country no longer applies That is also the case for countriesthat enjoyed an FTA with the union Of significance to wine markets are theEuropean Unionrsquos FTAs with Chile and South Africa

The size of that trade-diverting impact of leaving the union on wine depends onthe external tariff imposed on wine imports not only by the union but also by theleaving country Some have suggested that the United Kingdom should becomethe Hong Kong of Europe and move immediately to free trade on all productsOthers have suggested that this move would impose huge structural changes onthe UK economy that society would not tolerate ndash at least not without majorcompensation packages But all agree that a new trade policy that sets most-favored-nation (MFN) tariff rates is needed before the United Kingdom can beginto negotiate new preferential trading arrangements with the EU27 its FTA partnerssuch as Chile and South Africa or other countries Rollo Borchert Dawar Holmesand Winters (2016) suggest that the most practical trade policy for the UnitedKingdom to adopt at the outset is the European Unionrsquos tariff schedules previouslyagreed to at the WTO In all but one of the Brexit scenarios examined below we

Figure 3

UK Price Relative to World Price of Wine Imports and UK Wine-Import Intensitya 1950 to2015 ()

Fig3-Colou

ron

line

BW

inprint

a Import intensity is defined as the United Kingdomrsquos share of the value of global wine imports divided by the United Kingdomrsquos share of theworldrsquos total merchandise importsSource Compiled from data in Anderson and Pinilla (2017)

226 UK and Global Wine Markets by 2025 and Implications of Brexit

Table1

Shares

ofUKW

ineIm

portsfrom

TodayrsquosKey

Wine-Exp

ortin

gCountries1

675to

2014

andProjected

to2025

with

outandwith

Brexit()

Volum

eFrance

Spain

Portugal

Italy

Germany

South

Africa

Australia

United

States

Chile

New

Zealand

Argentin

aOther

countries

Total

1675ndash1696

2542

231

90

00

00

00

100

1697ndash1862

526

491

32

00

00

014

100

1863ndash1919

2626

220

00

20

00

025

100

1920ndash1940

1218

320

05

140

00

020

100

1995

volume

3210

318

143

73

21

07

100

value

439

413

102

83

21

05

100

2010ndash2014

volume

159

117

48

2110

84

12

100

value

358

215

44

115

66

12

100

Exp

ortersrsquoshareof

world

wineexports2010ndash2014

volume

1520

322

44

74

72

39

100

value

309

319

42

64

53

312

100

2025

projected

noBrexit

100

volume

188

188

26

207

45

45

65

43

64

17

20

92

100

value

323

100

30

218

44

23

58

44

54

41

20

45

100

2025

projected

with

Brexit

volume

190

188

27

208

43

44

66

42

63

17

19

93

100

value

311

96

29

210

41

24

59

44

55

41

20

70

100

SourcesCom

piledfrom

data

inAnd

ersonan

dPinilla

(201

7)to

1940

UnitedNations

COMTRADEh

ttpscomtrad

eun

orgdata

for19

95ndash2

014

andau

thorsrsquomod

elresults

for20

25

Kym Anderson and Glyn Wittwer 227

assume that Rollo et alrsquos suggestion will be the new UK trade policy commitmentto WTO members that subsequent negotiations for preferential arrangements willtake years and that any consequent agreements with the European Union andothers will be gradually implemented after that

The impact of leaving a customs union on wine markets does not only come fromtariff changes Also relevant are any effects of leaving on real UK incomes and thevalue of the British pound If the United Kingdom were to move immediately to freetrade on all products (the Hong Kong option) its per-capita income could eventuallyrise but only after considerable adjustment Should the United Kingdom insteadcommit to the current EU tariff schedule at the WTO in the first instance as weassume below then its per-capita income growth rate and the British poundrsquosexchange rates almost certainly will be lower for some time ndash at least until newtrade agreement negotiations with the EU27 and others are sufficiently advanced

Table 2Taxes on British Wine Imports by Source 1660ndash1862 (British Pounds per Kiloliter)

France Germany Spain Portugal South Africa

1660ndash1665 7 9 8 81666ndash1684 7 9 8 81685ndash1691 14 20 19 181692ndash1695 22 20 19 181696 47 20 19 181697ndash1702 51 25 23 221703 52 27 24 231704ndash1744 55 31 26 251745ndash1762 63 35 30 291763ndash1777 71 39 34 331778 79 43 38 371779 84 41 40 391780ndash1781 92 49 44 431782ndash1785 96 51 47 46 441786 65 51 37 37 371787ndash1794 47 51 32 32 371795 78 64 51 51 571796ndash1797 108 92 71 71 771798 111 96 73 73 791799ndash1801 107 92 71 71 771802 112 97 74 74 801803 131 109 87 87 871804 142 117 95 95 951805ndash1824 144 119 96 96 961825ndash1830 78 50 50 50 251831ndash1859 58 58 58 58 291860 32 32 32 32 321861 16 21 21 21 211862 11 26 26 26 26

Source Summarized from Ludington (2013 Table A1)

228 UK and Global Wine Markets by 2025 and Implications of Brexit

Figure 4

UK Shares in Wine Exports of Key Wine-Exporting Countries 2010ndash2014 ()

Fig4-BW

onlin

eBW

inprint

Source Compiled from data in Anderson and Pinilla (2017)

Table 3Import Duties Excise Duties and VAT Affecting Consumer Prices of Wine and Other Alcohol

in the United Kingdom April 1 2017

VAT 20

MFN import duties on wine pound per liter-bottled still wine lt13 alc 0114-bottled still wine 13ndash15 alc 0134-bottled still wine 15ndash18 alc 0162-bulk still wine lt13 alc 0086-bulk still wine 13ndash15 alc 0105-bulk still wine 15ndash18 alc 0134-sparkling wine 0278Weighted averagea 0130

Excise duties on alcohol pound per liter-still wine lt15 alc 2887-still wine 15ndash22 alc 3848-sparkling wine 55ndash85 alc 2795-sparkling wine 85ndash15 alc 3697-spirits (assumed 40 alc) 11551-beer (assumed 5 alc) 0954

VAT on alcohol 20

a Assumes that one-third of UK wine imports arrives in bulk that half has less than 13 alcohol and that one-tenth of imports is sparkling

Sources HM Revenue httpswwwgovukgovernmentpublicationsalcohol-duty-rate-changes for excise duties and VAT accessed April 92017 and Wine Australia (2015) for import duties converted at the 2016 average exchange rate of pound0740634 per euro

Kym Anderson and Glyn Wittwer 229

as to restore investor and consumer confidence in the United Kingdomrsquos economyThe size of those impacts is uncertain of course (see Baldwin 2016 and especiallyCampos 2016) so we consider a range of possibilities below

Those assumed adverse macroeconomic effects will add to the initial impact ofaltered wine tariffs on aggregate wine consumption in the United Kingdom andhence on its bilateral trades in wine They will make the loss of sales to theUnited Kingdom by EU (and Chilean and South African) suppliers greater thanwould otherwise be the case And they will reduce the likelihood that other countriesrsquosales of wine in the United Kingdom will be higher than in the baseline Indeed themacroeconomic effects could outweigh the trade-diverting effects so that even coun-tries that are currently discriminated against by the EU28rsquos wine-trade policy may beworse off because of Brexit

IV Global Wine Markets Model and Database

Our model of the worldrsquos wine markets first published by Wittwer Berger andAnderson (2003) and revised by Anderson and Wittwer (2013) is summarized inthe Appendix It disaggregates wine markets into four types namely nonpremiumcommercial-premium and superpremium still wines and sparkling wines2 There aretwo types of grapes premium and nonpremium Nonpremium wine uses nonpre-mium grapes exclusively superpremium wines use premium grapes exclusivelyand commercial-premium and sparkling wines use both types of grapes to varyingextents across countries The world is divided into 44 individual nations and 7 com-posite geographic regions that capture all other countries

The modelrsquos database is calibrated to 2014 based on the comprehensive wine-market volume and value data and trade and excise-tax data provided inAnderson and Pinilla (2017) and in Anderson Nelgen and Pinilla (2017) It is pro-jected assuming that aggregate national consumption population and realexchange rates change between 2014 and 2025 to the extent shown in AppendixTable 13 The Brexit alternatives to that baseline also are projected to 2025

Concerning preferences we assume a continued considerable swing toward allwine types in China and a swing away from nonpremium wines in all other countriesuntil 2025

In our baseline scenario grape- and wine-industry total-factor productivity areassumed to grow at 1 per year everywhere while grape- and wine-industry

2Commercial-premium still wines are defined byAnderson Nelgen and Pinilla (2017) to cost between US$250 and US$750 per liter pretax at a countryrsquos border or wholesale3The real exchange rate changes over the projection period are the changes expected in the nominal valueof country irsquos currency relative to the US dollar times the expected ratio of the GDP deflator for theUnited States versus that for country i

230 UK and Global Wine Markets by 2025 and Implications of Brexit

capital is assumed to grow net of depreciation at 15 per year in China but zero else-where (consistent with the almost-zero growth in global wine production and con-sumption over the past two decades)

Two alternative scenarios are considered (ldquolargerdquo and ldquosmallrdquo) for the initialimpact of Brexit to provide a range of results We assume that because of theUnited Kingdomrsquos decision to exit the European Union the rate of UK realgross domestic product (GDP) growth is only one-third or two-thirds as fast overthe projection period (09 or 18 per year instead of 26) and the Britishpound will be 20 or 10 lower in real terms than in our modelrsquos core baselineprojection4 In both alternative scenarios we assume that the United Kingdomapplies the European Unionrsquos external tariffs on wine from March 30 2019 atthe end of the two-year period following the United Kingdomrsquos formal triggeringof Article 50 (see Table 3)

We then consider a subsequent impact of Brexit presumed to result from negoti-ating signing implementing and responding to an FTA with the EU27 by 2025(Chile and South Africa currently have preferential access to EU wine marketsbut we continue to assume in this scenario that the United Kingdom does not imple-ment new bilateral FTAs with them or others in our time frame)5 This subsequentscenario assumes the British pound returns to what it would have been in 2025 in theabsence of Brexit and that real incomes regain two-thirds of the value they wouldhave had without Brexit as compared with our ldquosmallrdquo initial Brexit scenario

This global model has supply-and-demand equations and hence quantitiesprices and price elasticities for each of the grape and wine products and for asingle composite of all other products in each country Income elasticities ofdemand also exist for each final product Grapes are assumed to not be traded inter-nationally but other products are exported and imported Each market is assumed tohave been in equilibrium before any shock and to find a new market-clearingoutcome following any exogenously introduced shock

4The nominal price of the British pound in US dollars in the fortnight following the Brexit vote on June23 2016 dropped 13 to US$130 and a year later the pound sat at the same rate having dropped to anaverage of $124 (an 18 devaluation) between October 2016 and April 2017 Our choice of a low of 10and a high of 20 aims to capture future possible rates while uncertainties remain The average real wagein the United Kingdom fell in the first half of 2017 and projected real GDP growth during 2018ndash2020 hasbeen revised down to 17 in the United Kingdomrsquos latest budget (HM Treasury 2017)5South Africa currently has duty-free access to the EU28 for just 50 megaliters (ML) of wine per yearbeyond which the MFN tariff rate applies South Africa currently exports around 320 ML to theEuropean Union one-third of which initially goes to the United Kingdom How that quota of 50 MLis divided between the United Kingdom and the EU27 is subject to future negotiation (Rollo et al2016 Swinbank 2017) In the following discussion we assume none of it is accepted by the UnitedKingdom so that all South African wine imports pay the United Kingdomrsquos MFN tariff under Brexitbut altering that assumption makes very little difference to our results for any countryrsquos trade exceptSouth Africarsquos

Kym Anderson and Glyn Wittwer 231

V Projecting Global Wine Markets to 2025

Global wine production and exports are projected in the baseline from 2014 to 2025consistent with past trends The modelrsquos global volume of production (and consump-tion) rises little over that 11-year period (9) made up of a 6 decline in nonpre-mium wine and a one-sixth rise in commercial-premium and superpremium wineIn real (2014 US$) value though global wine output and consumption increaseby about 50 in total and 60 in the two premium categories The internationaltrade projections are similar although a little larger with the share of global wineproduction exported (equivalent to the share of global consumption imported)rising 2 percentage points between 2014 and 2025

The baseline projection does not greatly alter the 2014 shares of various countriesin global wine production apart from China because we assume vineyard expansionthere is faster than elsewhere6 In value terms this assumption means that Chinamoves from fifth to fourth by 2025 behind France the United States and ItalySpain remains barely ahead of Australia and they and Germany take the nextthree places (Figure 5(a)) In total wine-production volume terms China movesfrom sixth to fifth place and Argentina drops from fifth to eighth (and fromeighth to ninth in value terms)

When their products are subdivided into fine wines (superpremium still plus spark-ing) commercial-premium wines and nonpremium wines France and the UnitedStates retain the highest two places on the global ladder for fine-wine productionand Spain and Italy retain the top two places for nonpremium wine As for commer-cial-premium wine production (defined to cost between US$250 and US$750 perliter pretax at a countryrsquos wholesale level or national border) Italy retains the topranking over our projections period but at least in terms of value China challengesFrance for second place

The country rankings by projected value of total wine consumption change some-what more than those for production by 2025 with China taking second place afterthe United States ahead of France and Germany and then the United Kingdomslightly overtaking Italy to slip into fifth place (Figure 5(b)) The United StatesFrance and Germany retain the top three rankings for consuming fine wine butCanada slightly overtakes Italy for fourth place in terms of value at least In thecase of commercial-wine consumption China strengthens its number-one positionahead of the United States and the United Kingdom does likewise vis-agrave-visGermany for third place

As for the projected changes in consumption volumes China is projected to dom-inate the increase in aggregate although the United States is projected to lead the

6 In fact Chinarsquos wine production fell steadily between 2012 and 2016 by a total of one-sixth so Chinarsquoswine imports may grow faster in practice than in this baseline projection

232 UK and Global Wine Markets by 2025 and Implications of Brexit

Figure 5

Value of Wine Production and Consumption in Key Countries 2014 and Projected Baseline for2025 (2014 US$ Million at WineryWholesale Pretax Prices)

Fig5-BW

onlin

eBW

inprint

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 233

increase in consumption of fine wine In Western Europe and in the SouthernHemispherersquos New World countries fine wines are projected to substitute for com-mercial wines (defined as the sum of commercial-premium and nonpremiumwines) with almost no change in total wine consumption Sub-Saharan Africa isthe next region that is projected to take off with its growth accounting for morethan one-third of the rest of the worldrsquos increase in volume consumed

Those differences in production versus consumption rankings are reflected ininternational trade Figure 6 shows that France Italy and Spain remain the threedominant exporters of wine in aggregate value but the rankings of the next fewchange with Australia being slightly ahead of Chile and the United StatesGermany and New Zealand being nearly tied for sixth place in value termsFrance and Italy are even more dominant in fine-wine exports and remain so by2025 while Italy outranks France in the commercial-premium export categoryand Spain outranks Italy Australia and Chile in the nonpremium export class

Among the importers the United States and the United Kingdom are projected tocontinue to hold the first two places in 2025 in value terms but China moves intothird place slightly ahead of Germany followed well behind by Canada HongKong Belgium-Luxembourg the Netherlands and Japan (Figure 6(b)) OtherAfrica (excluding South Africa) is projected to experience the largest increase inimports among all the other regions followed by Other Asia which becomes asbig as Germany in value terms (Figure 7(a)) In terms of total volume of wineimports Germany and the United Kingdom held the top two shares in 2014 butby 2025 the United Kingdom is projected to be well ahead of Germany (Figure 7(b)) However this projection ignores the effects of Brexit to which we now turn

VI How Might Wine Markets Be Affected by the United Kingdomrsquos Exitfrom the European Union

As mentioned earlier for our two alternative scenarios to capture the initial effects ofBrexit (ldquolargerdquo and ldquosmallrdquo) we assume that following the United Kingdomrsquos exitfrom the European Union the United Kingdomrsquos rate of economic growth wouldbe only one-third or two-thirds as fast for the period to 2025 the British poundwould be 20 or 10 lower in real terms than in our modelrsquos baseline projectionand the United Kingdom would apply the European Unionrsquos external tariff onwine to imports from EU member countries (as part of establishing MFN ratesvia the WTO to then start new bilateral FTA negotiations) In these initial scenarioswe assume the United Kingdom does not implement any new FTAs particularlywith the EU27 Chile and South Africa Following the discussion of those initialresults we present the results of a subsequent scenario that includes the implemen-tation of an FTA between the United Kingdom and the EU27 We assume these aretwo discrete steps with no agreed arrangements to smooth the transition betweenthem

234 UK and Global Wine Markets by 2025 and Implications of Brexit

Figure 6

Value of Wine Exports and Imports Key Wine-Trading Countries 2014 and Projected Baselinefor 2025 (2014 US$ million)

Fig6-BW

onlin

eBW

inprint

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 235

Figure 7

National Shares of Global Wine Import Value and Volume 2014 and Projected Baselinefor 2025 ()

Fig7-BW

onlin

eBW

inprint

Source Authorsrsquo model results

236 UK and Global Wine Markets by 2025 and Implications of Brexit

A Initial Impact of Brexit

Because the British pound dropped by one-sixth against the US dollar in the fourmonths following the Brexit vote in June 2016 and because the UnitedKingdomrsquos average real wage fell 1 between November 2016 and April 2017(according to the latest available data see Office of National Statistics 2017) weviewed our ldquolargerdquo scenario as more likely soon after the Brexit vote But wecompare those results with ones from our ldquosmallrdquo initial-impact scenario becauseit has become more likely following the June 2017 general election that ldquosofterrdquooptions will be considered by the new government Generally the results areabout half the size in the ldquosmallrdquo scenario with the exception of the bilateraltrade effects To show the sensitivity of results to our assumptions we point outthe differences when the ldquosmallrdquo results are not close to half the results shown forthe ldquolargerdquo scenario

Before turning to those results we first present a ldquofree-traderdquo scenario whichassumes the United Kingdom chooses the radical option of becoming the HongKong of Europe Unlikely as this scenario is the results provide assurance thatour global wine model generates the standard types of effects of exiting a customsunion In this ldquofree-traderdquo scenario the United Kingdom is assumed to move tozero tariffs on all wine imports Thus the current preferences on EU Chileanand South African wine imports disappear as all other wine exporters enjoy thesame free access to UK wine markets We assume in this scenario that Brexit hasno impact on the United Kingdomrsquos real income or consumption growth rates by2025 and that by then the British pound has returned to its 2014 value relative toother currencies (similar to the real exchange rate assumed in our baseline scenarioto 2025 see Appendix Table A1)7 Under these assumptions EU Chilean andSouth African wine exports to the United Kingdom in 2025 are 003 lower thanin the no-Brexit baseline and UK imports from other wine exporters are 025higher but the United Kingdomrsquos total wine imports are hardly any different(because local wine prices fall by only 05 and incomes are unchanged) In this sce-nario UK consumers expand their volume of consumption by just 05 while theEU27 exports more and other countries export less to the rest of the world such thatworld wine exports are almost unchanged In short all these effects are in the direc-tion that comparative static-customs-union theory would predict but they are smallbecause we assume the United Kingdomrsquos real income and currency are the same asin our baseline for 2025 as described in the previous section

In the ldquolargerdquo alternate initial scenario involving Brexit as compared with thebaseline scenario to 2025 the consumer price of wine in 2025 is 22 higher in the

7If this scenario were seen as a credible long-run trade policy it would encourage more investment thatwould eventually raise UK incomes but not without major structural changes that would initiallydisrupt the economy Implicitly we assume in our free-trade scenario that these two opposite impactson real UK incomes and the British pound exactly offset each other by 2025

Kym Anderson and Glyn Wittwer 237

United Kingdom in local currency terms (20 because of real depreciation of theBritish pound 4 because of new tariffs on EU Chilean and South Africanwines and ndash2 because of slower UK income growth) The volume of UKwine consumption is 28 lower 16 because of slower UK economic growth7 because of real depreciation of the British pound and 5 because of newtariffs Superpremium still-wine sales are the most affected dropping by two-fifths while sparkling and commercial-premium wines drop a bit less than one-quarter Because the average price rises by more than the decline in volume soldthe aggregate value of UK sales even in local currency terms falls under thisldquolargerdquo Brexit scenario Under the ldquosmallrdquo Brexit scenario the consumer price ofwine in 2025 is 11 higher in the United Kingdom and its volume of wine consump-tion is 17 lower

The volume of projected UK imports in 2025 is 427 ML or nearly one-quarterless in the ldquolargerdquo scenario than in the baseline scenario comprising 58ML less spar-kling wine 31 ML less superpremium still wine and 339 ML less commercial-premium wine World imports are lower by just 239 ML because imports byother countries are 189 ML higher in response to the lower international wineprices in this scenario In value terms UK imports are $175 billion (or 27)lower in 2025 because of ldquolargerdquo Brexit $113 billion because of lower incomes$038 billion because of the fall in the British pound and $014 billion because ofthe rise in wine-import tariffs (Table 4) These aggregate trade impacts are a littlemore than half this size under the ldquosmallrdquo Brexit scenario

Despite the import levels falling because of raised import tariffs domestic con-sumption of all three quality categories of UK-produced wine is lower with thanwithout Brexit because of shrunken demand for all wines resulting from loweredUK incomes and raised local prices because of devaluation of the British poundThe British poundrsquos devaluation does make it easier for the United Kingdom tosell wines abroad though The countryrsquos exports are 7 ML or nearly 5 higherin 2025 in the ldquolargerdquo Brexit scenario and UK production is 3 higher ThoseUK exports (or re-exports of imported bulk wine after it is bottled in the UnitedKingdom) that go to EU27 countries are reduced though because of the tariffnow imposed at the new EU border

Without Brexit the United Kingdomrsquos shares of global wine imports are slightlyhigher in volume terms in 2025 than in 2010ndash2015 but 2 percentage points lower invalue terms thanks to East Asiarsquos expanding demand for imports of premium winesWith ldquolargerdquo Brexit however that value share is an additional 2 percentage pointslower and the volume share is almost 5 points lower (Figure 8) The net effects ofthese impacts on global trade are shown in Figure 9 Most of the initial tradeeffect of Brexit is a large decline in net imports of wine by the United Kingdomwith very little offsetting positive effect on trade in the rest of the world Theldquosmallrdquo Brexit numbers are a bit more than half of those for ldquolargerdquo Brexit

238 UK and Global Wine Markets by 2025 and Implications of Brexit

Table4

Difference

in2025

Projected

VolumeandValue

ofWineIm

portsb

ytheUnitedKingdom

andtheResto

fthe

World

asaConsequence

oftheInitial

BrexitS

hock

(MLand2014US$

Million

ldquoLargerdquoSc

enario)

Volum

e(M

L)

Value

(US$

million)

NP+CPa

SuperPrb

Sparkling

TOTAL

NP+CPa

SuperPrb

Sparkling

TOTAL

ΔUK

impo

rtsdu

eto

Low

erincomes

minus198

minus20

minus29

minus247

58minus644

minus253

minus234

minus1131

65Low

erpo

und

minus70

minus10

minus14

minus93

22minus248

minus127

minus102

minus476

27Highertariffs

minus71

minus1minus16

minus87

20minus110

minus8minus2

4minus143

8TOTA

Lminus339

minus31

minus58

minus427

100

minus1001

minus388

minus360

minus1750

100

difffrom

base

2332

3325

2432

3227

of

totalc

uts

797

14100

5722

21100

ΔROW

netim

ports

143

2125

189

230

143

minus181

192

ΔWORLD

TRADE

minus195

minus10

minus34

minus239

minus763

minus246

minus543

minus1552

aNon

prem

ium

plus

commercial-premium

still

wines

bSu

perpremium

still

wines

SourceA

utho

rsrsquomod

elresults

Kym Anderson and Glyn Wittwer 239

The aggregate effect of ldquolargerdquo Brexit on the market shares of various wine-export-ing countries in the United Kingdom is almost indiscernible even with one decimalpoint The projected 2025 shares are quite different from the actual 2014 shares forseveral countries They are much smaller in 2025 for South Africa Australia andNew Zealand (and the United States in volume terms) and are much larger involume for Spain and in value for Italy These results occur because wine-exportingcountries benefit differentially from the varying rates of growth in net importdemand for wine in non-UK countries over this projection period The most impor-tant projected changes are the increase in the real value of annual wine importsbetween 2014 and 2025 by China (200 or $3 billion) Other Asia (110 or $22billion) and Africa (270 or $16 billion) More than half of Australiarsquos increasein annual exports from 2014 to 2025 go to Asia and more than half of SouthAfricarsquos increase in exports go to Other Africa

Table 5 reveals that ldquolargerdquo Brexit lowers EU Chilean and South African wineexports by 150 ML or US$12 billion in the case of the European Union with someof their exports diverted from the United Kingdom to the EU27 and other marketsin competition with New World exporters Although the United States Australiaand Argentina sell only a little less to the United Kingdom they also sell less toother countries For Chile and South Africa which lose their preferential access toUK (but not to EU27) markets in this Brexit scenario some exports are redirected

Figure 8

UK Shares of World Wine Imports 2010ndash2015 and Projected to 2025 without and with theldquoLargerdquo Initial Brexit Shock ()

Fig8-BW

onlin

eBW

inprint

Sources Anderson and Pinilla (2017) and authorsrsquo model results

240 UK and Global Wine Markets by 2025 and Implications of Brexit

from the United Kingdom to EU27 countries ndash but again they export less overallGlobal wine trade in 2025 is 240 ML (19) or $18 billion (35) less under thisldquolargerdquo Brexit scenario The percentage by which wine exportersrsquo trade shrinks isgreater for values than for volumes because of changes in relative prices of differ-ent-quality wines Those differences are shown in the numbers in parentheses inTable 5

Three other points are worth making about Table 5 One is that Australia sellsslightly more to the United Kingdom in the ldquosmallrdquo Brexit scenario rather thanslightly less as in the ldquolargerdquo Brexit case Evidently the negative income and price(devaluation) effects do not more than offset the positive trade-diverting effect onAustralian exports to the United Kingdom of removing preferences in the ldquosmallrdquoscenario Second New Zealand sells slightly more to non-UK countries underBrexit despite greater competition from the EU27 Chile and South Africa Thisanomaly is due to changes in the relative prices of different qualities of wine inglobal wine markets bearing in mind that New Zealand has the worldrsquos highestaverage price for still-wine exports And third the value (but not the volume) ofexports of ldquoOtherrdquo countries to markets other than the United Kingdom arehigher under Brexit This result too is due to changes in the relative prices of differ-ent qualities of wine in global wine markets

Figure 9

Difference in 2025 Wine Import Volumes and Values as a Result of the ldquoLargerdquo Initial BrexitShock (ML and US$ Million in 2014 US dollars)

Fig9-BW

onlin

eBW

inprint

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 241

B Subsequent Impact of Brexit from a UK-EU27 FTA

The next-most-likely step in the Brexit process is for the United Kingdom to nego-tiate a new trade arrangement with the EU27 We therefore assume that a UK-EU27FTAwith free bilateral wine trade is implemented and adjusted to by 2025 and thatprogress toward that end occurs soon enough that the adverse macroeconomicshocks from the initial impact of uncertainty over the Brexit process are confinedto those assumed in the ldquosmallrdquo scenario outlined above In this subsequent scenariowe assume that the British pound returns to the value it would have reached in 2025in the absence of Brexit (ie reversing the 10 devaluation assumed in the ldquosmallrdquoinitial Brexit scenario) and that real incomes in the United Kingdom are 8 higher

Table 5Difference in 2025 BilateralWine Import Volumes and Values fromKey Exporters by the United

Kingdom and the Rest of the World (RoW) as a Result of Initial Brexit Shock(ML and 2014US$ Million)a

(a) ldquolargerdquo scenario

Volume (ML) Value (2014US$ million)

UK RoW WORLD () UK RoW WORLD ()

EU27 minus287 136 minus150 (ndash17) minus1187 minus5 minus1192 (ndash31)Chile minus59 35 minus25 (ndash30) minus169 31 minus138 (ndash48)South Africa minus53 35 minus18 (ndash32) minus105 20 minus85 (ndash67)USA minus7 minus6 minus13 (ndash24) minus75 minus40 minus115 (ndash50)Australia minus4 minus3 minus7 (ndash09) minus25 minus65 minus90 (ndash30)Argentina minus3 minus9 minus12 (ndash48) minus16 minus39 minus55 (ndash52)New Zealand minus11 9 minus2 (ndash09) minus162 71 minus91 (ndash43)Others minus2 minus10 minus12 (ndash02) minus11 minus52 minus63 (ndash44)WORLD minus427 187 minus240 (ndash19) minus1750 minus79 minus1829 (ndash35)

(a) ldquosmallrdquo scenario

Volume (ML) Value (2014US$ million)

UK RoW WORLD () UK RoW WORLD ()

EU27 minus178 82 minus96 (ndash12) minus692 minus43 minus736 (ndash19)Chile minus46 28 minus18 (ndash24) minus128 36 minus91 (ndash32)South Africa minus43 29 minus14 (ndash42) minus82 23 minus59 (ndash47)USA 1 minus6 minus5 (ndash11) minus23 minus28 minus51 (ndash22)Australia 5 minus10 minus5 (ndash06) 19 minus56 minus38 (ndash13)Argentina 0 minus6 minus6 (ndash26) minus3 minus25 minus29 (ndash27)New Zealand minus5 4 minus1 (ndash06) minus80 34 minus46 (ndash22)Others 0 minus9 minus9 (ndash01) minus1 minus33 minus34 (ndash24)WORLD minus266 112 minus154 (ndash13) minus991 minus92 minus1083 (ndash21)

a Numbers in parentheses are the percentage difference between the Brexit and baseline scenarios for 2025 projected wine-import volumes orvalues by source

Source Authorsrsquo model results

242 UK and Global Wine Markets by 2025 and Implications of Brexit

in 2025 than in the ldquosmallrdquo scenario ndash that is they regain two-thirds of the value theywould have reached without Brexit-related uncertainty

This subsequent development in the Brexit process reverses most of but not allthe initial effects of Brexit by 2025 because of our assumption that the lost growthin the initial years of uncertainty following the Brexit vote are only partly recoveredby 2025 following the implementation of a UK-EU27 FTA Moreover the longer ittakes before this FTA is finalized and implemented the longer the estimated initialadverse effects persist and the larger the cumulative cost of Brexit to UK wine con-sumers and to grape and wine producers in wine-exporting countries

Table 6 summarizes the subsequent trade effects for 2025 It suggests that all butone-tenth of the loss in value of world trade in wine from the initial ldquosmallrdquo impact isrestored most of it because of a smaller reduction in the United Kingdomrsquos wineimports following the FTA Most of that improved outcome is because of recoveredimports from the EU27 commensurate with the latterrsquos high share of UK importsEven though Chile and South Africa are assumed in this scenario to have not yetsigned an FTA with the United Kingdom they export slightly more to the UnitedKingdom (and even more to all other countries) than in the initial ldquosmallrdquo scenario

Within the United Kingdom this FTA brings down the local currency consumerprice of wine by 9 largely offsetting the 11 rise in the initial ldquosmallrdquo Brexit sce-nario and it raises the volume consumed in the United Kingdom by 18 fully off-setting the 17 fall in the initial ldquosmallrdquo Brexit scenario

Table 6Difference in 2025 Bilateral Wine-Import Volumes and Values from Key Exporters by the

United Kingdom and the Rest of the World (RoW) as a Result of Implementing a UK-EU27FTA Following Initial ldquoSmallrdquo Brexit Shock (ML and 2014US$ Million Difference Relative to

Initial ldquoSmallrdquo Brexit Scenario)a

Volume (ML) Value (2014US$ million)

UK RoW WORLD (ldquosmallrdquo)a UK RoW WORLD (ldquosmallrdquo)a

EU27 212 minus112 100 (minus96) 750 0 750 (minus806)Chile 1 3 4 (minus18) 3 40 43 (minus92)South Africa 3 2 5 (minus14) 3 18 21 (minus60)USA 3 3 6 (minus5) 26 21 47 (minus50)Australia 1 1 2 (minus5) 8 38 46 (minus33)Argentina 1 4 5 (minus6) 4 20 24 (minus28)New Zealand 4 minus3 1 (minus1) 61 minus25 36 (minus46)Others 1 6 7 (minus9) 5 minus46 minus41 (92)WORLD 226 minus96 130 (minus154) 860 66 926 (minus1022)

a Numbers in parentheses are the world trade differences between the ldquosmallrdquo initial Brexit scenario and the baseline scenario copied fromcolumns 3 and 7 of Table 5(b)

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 243

In short this subsequent step in the Brexit process can be expected to restore muchof the initial adverse effects in the United Kingdom of the Brexit vote by 2025 Thewine exporters in the EU27 Chile and South Africa also lose less in this scenarioas compared with the initial ldquosmallrdquo scenario while other key exporters sell nearlyas much in 2025 in this scenario as they are projected to in the baseline ndash that isthey are beneficiaries of the reduced discrimination in global wine markets in thiscase where Chile and South Africa no longer enjoy preferential access to the UKmarket (and even though EU27 producers are again allowed to do so) Keep inmind though that although the estimated losses in the initial scenarios are reportedjust for 2025 they are expected to be felt throughout the preceding years of uncertaintyTherefore the longer it is before uncertainty abates and the United Kingdomrsquos currentpreferential trading arrangements are replaced by newones such as a UK-EU27 FTAthe higher the cumulative cost of the Brexit vote to the wine trade

VII Caveats and Conclusions

The above Brexit simulations are just a few of many scenarios that could be modeledObvious additional ones could also assume that FTAs are reached between theUnited Kingdom and other trading partners including Chile and South Africaand such countries as Australia and New Zealand Some of these other countriesare already exploring the ECrsquos invitation (see European Commission 2015) to nego-tiate an FTA with what will be the EU27 The sequence in which FTAs are signedand the speed with which they are implemented will matter (as was also the casewith the sequential signing over the past decade of bilateral FTAs with NortheastAsian countries by Chile Australia and New Zealand see Anderson andWittwer 2015) Even if the United Kingdom were able to sign additional bilateralFTAs and begin implementing them before 2025 it would have little effect on theabove results (because wine tariffs are such a minor contributor to them) unlessthose FTAs were to accelerate the United Kingdomrsquos economic growth and therise in the British poundrsquos value before 2025

We assume above that no changes are made to alcohol excise duties in the UnitedKingdom following Brexit even though they are scheduled to be progressively raisedwith inflation and may be raised even more for wine relative to spirits to offset theopposite effects of Brexit on those two domestic industries Nor do we make anyallowance for UK increases in consumer wine prices that may be needed to coverthe higher cost of clearing customs on imports from EU27 countries (which maybe more or less offset by increased smuggling and duty-free purchasing)Importers of fine wines including individuals who buy Bordeaux en primeur maywell reduce their demand because of concerns that delays on docks will affect thequality of their wines or because of greater currency uncertainty

To summarize Brexit is costly initially to UK consumers of wine (and of manyother tradable products) because the domestic retail price in local currency tax-

244 UK and Global Wine Markets by 2025 and Implications of Brexit

inclusive terms is 22 higher than otherwise in 2025 and the volume of wine con-sumed domestically is 28 lower in the ldquolargerdquo Brexit scenario (or 11 higher pricesand 17 lower quantities in the ldquosmallrdquo case) by 2025 unless a UK-EU27 FTA isdeveloped before then Even if such an FTA does get signed ratified by all 28 par-liaments and implemented by 2025 the slower income growth in the interim meansa smaller UK wine market in 2025 than would otherwise have been the case Thevolume reduction is a blow to many participants in UK wine-bottling transportingstoring wholesaling and retailing businesses in addition to restaurants and pubsVery little of that initial impact occurs because of higher import tariffs most impor-tant is the assumed fall in UK real incomes

The small but growing number of local UK vignerons (who supply less than 05of all domestic-wine sales and less than 4 of sparkling-wine sales) is initially pro-jected to sell less wine domestically because of reduced demand and to export only alittle extra abroad while the British pound is devalued Overall their production isonly 3 higher in 2025 even in the ldquolargerdquo Brexit scenario (not taking intoaccount the fact that their casual labor is likely to be more expensive under Brexitbecause of tighter restrictions on immigration) and their production diminishes ifand when a UK-EU27 FTA is in place

Even with a UK-EU27 FTA in place EU27 wine exporters export US$56 millionless wine in 2025 thanks to Brexit Chile and South Africa export $88 million lesswine and wine exports from the rest of the world are little different in aggregatealthough slightly larger from Australia

Clearly there will be great uncertainly for some time over the possible policy out-comes to flow from Brexit and their consequent sequential impacts on UK house-hold disposable incomes foreign exchange rates and bilateral wine tariffsMeanwhile the above projections under explicit assumptions provide some idea ofhow wine markets might be affected by the most-likely first two stages of theBrexit process (agreeing on a new tariff schedule at the WTO and agreeing to andimplementing a UK-EU27 FTA) In particular they make clear that nontrivialinitial impacts could affect the domestic wine market impacts that are likely to belarger than just the direct impact of changes in bilateral tariffs If the UnitedKingdom succeeds in getting countries to agree to transition arrangements thatdelay the changes in tariffs until new FTAs are signed and ratified the initialeffects will be less dramatic than in our first scenarios but it remains to be seenwhether any such agreements can be reached In any event the net effect of Brexiton the welfare of the worldrsquos consumers and producers of wine as a whole will benegative not just initially but permanently unless new trade policy commitmentsby the United Kingdom with major wine-exporting countries are sufficiently moreliberal than current arrangements

Kym Anderson and Glyn Wittwer 245

Appendix Key Equations in the Global Wine Markets Model

Amodel of the worldrsquos wine markets was first published byWittwer et al (2003) andmuch improved by Anderson and Wittwer (2013) Several significant enhancementshave been made to that original model Wine types have been disaggregated intomore types ndash namely nonpremium (including generic bulk) commercial-premiumand superpremium still wines and sparkling wines8 As in the original modelthere are two types of grapes premium and nonpremium Nonpremium wines usenonpremium grapes exclusively superpremium wines use mainly premium grapesand commercial-premium and sparkling wines use both types of grapes In theregional dimension the number of countries and country groups has expandedfrom 10 in the original model to 51 44 individual nations and 7 compositeregions The modelrsquos database is calibrated to 2014 for this paper based on datain Anderson et al (2017) trade data for which are downloaded from httpscomtradeunorg The modelrsquos supply-and-demand equations are based on the orig-inal ORANI modelrsquos theory (see Dixon Parmenter Sutton and Vincent 1982) andthe model is implemented using GEMPACK software (Harrison Horridge Jerieand Pearson 2014)

An enhancement of importance to the present study is the inclusion of exchange-rate variables in the model which allow a distinction between price impacts asobserved in local currency units and those observed in 2014 US dollars

In the model the grape and wine sectors minimize costs of intermediate inputssubject to weak constant elasticity of substitution (CES) substitutability betweeninputs By assumption no intermediate inputs are imported from other countries9Hence

Xcid frac14 f ethX1id CESfrac12P

cid=P1id $THORN eth1THORN

P1id X1id frac14X

c

Xcid P

cid eth2THORN

whereXcid is the quantity demanded of commodity c by grape or wine industry i in

region d Pcid is the corresponding price and X1id and P1id are the respective inter-

mediate composite quantities and prices

Two primary factors are employed in the sector labor (the quantity of which isendogenous with perfectly elastic supply) and capital Capital is usually treated asexogenous in quantity with rates of return bearing all the adjustment in thevarious scenarios This reflects the fact that grapes (a perennial crop) and wine-

8Commercial-premium still wines are defined by Anderson and Nelgen (2011) to be those priced betweenUS$250 and $750 per litre pre-tax at a countryrsquos border or wholesale9An exception concerns bulk wine imports used mainly in the bottling of commercial premium wine in theUK USA Canada Japan and Germany

246 UK and Global Wine Markets by 2025 and Implications of Brexit

plant capacity adjust slowly to market signals

Lid frac14 f ethFid CESfrac12Wid=PFid $THORN eth3THORN

Kid frac14 f ethFid CESethRid=PFid $THORN eth4THORN

PFid Fid frac14 Lid Wid thorn Kid Rid eth5THORN

Grape and wine producers are assumed to minimize costs subject to CES substitu-tion between capital and labor Equations (3) to (5) show primary factor demandsfor the labor composite Lid and capital Kid subject to a composite factor demandFid by industry i in region d The factor prices are Wid for labor Rid for capitalrentals and PFid for composite factor prices

Appendix Table A1Cumulative Consumption and Population Growth Rates and Changes in the Real Exchange Rate

(RER) Relative to the US Dollar 2014 to 2025 without Brexit ()

Aggregateconsumption Poprsquon RER

Aggregateconsumption Poprsquon RER

France 18 4 minus11 Australia 35 11 minus17Italy 11 2 minus9 New Zealand 32 9 minus26Portugal 14 0 minus9 Canada 27 8 minus18Spain 26 8 minus9 United States 31 8 0Austria 19 4 minus7 Argentina 7 10 109Belgium 20 7 minus9 Brazil 16 8 minus29Denmark 22 2 minus9 Chile 55 8 minus2Finland 21 3 minus7 Mexico 42 12 minus8Germany 14 minus2 minus11 Uruguay 45 3 1Greece 22 minus1 minus14 Other Latin America 60 10 minus5Ireland 42 12 minus9 South Africa 36 12 minus1Netherlands 21 4 minus9 Turkey 50 8 20Sweden 24 9 minus13 North Africa 53 11 0Switzerland 18 8 minus6 Other Africa 109 18 84United Kingdom 32 6 1 Middle East 52 18 minus12Other West Europe 21 10 minus1 China 79 3 5Bulgaria 41 minus7 7 Hong Kong 42 3 2Croatia 20 minus2 minus1 India 134 13 17Georgia 35 0 23 Japan 11 minus3 minus24Hungary 25 minus3 minus11 Korea 38 1 minus9Moldova 49 minus11 13 Malaysia 62 15 minus16Romania 45 minus4 22 Philippines 75 18 7Russia 18 minus2 minus8 Singapore 44 21 minus22Ukraine 22 minus5 14 Taiwan 29 1 minus13Other East Europe 40 minus5 48 Thailand 47 3 minus9

Other Asia 99 10 10

Source Authorsrsquo compilation from projections by various international agencies and from global economy-wide modeling by Anderson andStrutt (2016)

Kym Anderson and Glyn Wittwer 247

The composite factor demand Fid is proportional to total output Qid subject to aprimary factor using technology Aid Hence

Fid frac14 Qid Aid eth6THORN

The perfectly competitive zero-pure-profit condition is that total revenue valued atthe output price P0s

i multiplied by Qid equals the total production cost

P0si Qid frac14

X

cPcid X1cid thorn

X

oWo

id Loid thorn Rid Kid eth7THORN

Household demands follow a linear expenditure system in each region We reducethe optimizing problem for household consumption of each commodity subjectto a budget constraint to equations describing subsistence and discretionarydemands Aggregate subsistence expenditure WSUBd depends only on consumerprices P3cd for each commodity and the number of households N as per-capita sub-sistence quantities XSUBcd subject to given preferences are constant

WSUBd frac14X

c

P3cd XSUBcd Nd eth8THORN

Discretionary expenditures for each commodity (the left-hand side of equation (9))are equal to the marginal budget share (βcd) of aggregate discretionary expenditureThis aggregate is the bracketed term on the right-hand side of equation (9) whereW3TOTd is aggregate nominal expenditure

P3cdethX3cd XSUBcd NdTHORN frac14 βcdethW3TOTd WSUBdTHORN eth9THORN

Because real aggregate consumption is usually exogenous in our partial equilibriumsimulations the linear expenditure system determines the consumption shares ofindividual final commodities (ie the five wine types plus a composite of all otherconsumption items) driven by changes in relative prices as faced by domestic con-sumers The income elasticity of demand for each commodity is equal to the mar-ginal budget share divided by the expenditure share This number varies from 05for nonpremium wine to 17 for superpremium still wine The income elasticity ofdemand for other consumption is very close to 10 because wine accounts for anaverage of only 03 of aggregate expenditures globally and no more than 11in any country (Anderson and Nelgen 2011 Table 166)

A feature of our revised model of world wine markets is the inclusion of nominalexchange rates These appear directly in the equation linking retail prices (P3scd ) toproducer prices by country of origin (P0s

c ) where c denotes the wine type

P3 scd frac14 P0s

cfd

fsTtarcd T

taxcd thorn Pm

cd eth10THORN

248 UK and Global Wine Markets by 2025 and Implications of Brexit

The exchange rates in the consuming and producing regions are fd and fs respec-tively expressed as local currency units per $US Ttar

cd is the power of the tariff in theconsuming region and Ttax

cd is the power of the domestic consumption (or excise) taxprior to any generic value-added or goods-and-services tax Pm

cd is the price of marginm assumed to be locally supplied nontradable and therefore unaffected by theexchange rate

A given level of consumption for wine type c (X3cd) is satisfied using theArmington (1969) assumption in which wines from different countries of originare imperfectly substitutable First domestic wine is imperfectly substitutable witha composite of imports

X3sscd frac14 f ethX3cd CESethP3sscd=P3cdTHORNTHORN ss frac14 domestic imports eth11THORN

and then imports by origin (X3scd ) are determined in a second CES equation

X3scd frac14 f ethX3ssfrac14importscd CESethP3scd=P3

ssfrac14importscd THORNTHORN eth12THORN

The model enables us to show how changes in international competitiveness affectthe worldrsquos wine markets A crucial part of this exercise is explaining how pricesdetermined outside the grape and wine markets influence these markets Becausethe model is in partial equilibrium to depict the impacts of changes in internationalcompetitiveness outside price changes need to be imposed as shocks on the modelThe price of intermediate inputs other than grapes and wine shown in equations (1)and (2) is set equal to the price of GDP (Pg

d ) multiplied by a shifter Fcd

Pcid frac14 Fc

dPgd eth13THORN

If no specific price observations are available the shifter Fcd remains exogenous and

unshocked with the change in price being determined by a shock to the price ofGDP If observations are available for specific input price movements the shifterFc

dbecomes endogenous with Pc

id now exogenous and shocked

Wid frac14 Fwd P

gd eth14THORN

Wage rates are treated similarly In equation (14) if the wage shifter Fwd is exogenous

changes in wage rates Wid are determined by changes in the price of GDP If wagerate data are available Fw

d becomes endogenous andwage rates are shocked directly

Pmcd frac14 Fm

d Pgd eth15THORN

The prices of trade and transport margins are also determined by the price of GDP ifthe shifter Fm

d in equation (15) is exogenous

Changes in international competitiveness depend on changes in relative pricelevels and changes in nominal exchange rates In equation (16) fR

s denotes real

Kym Anderson and Glyn Wittwer 249

exchange-rate movements relative to the US dollar

fRs frac14 Pg

s =frac12PgUSA

fs$ eth16THORN

In equation (16) the nominal exchange rate for the United States is alwaysunchanged because nominal and real exchange rates are expressed relative to theUS currency

Changes in international market conditions may have impacts in one direction onproducer prices as expressed in US dollars P0s

i and potentially in the opposite direc-tion in local currency units Hence we calculate real producer pricesP0s

iloc in localcurrency terms (ie the price most relevant to domestic producers)

P0siloc frac14 P0s

ifs=P

gs eth17THORN

To obtain real price changes in local currency terms we convert US dollar prices(P3 s

cd for the source-specific price and P3cd for the source-composite price) to reallocal currency prices (P3 s

cdloc and P3cdloc) using the consumer price index (CPI)(Pc

d ) as the deflator

P3 scdloc frac14 P3 s

cdfd=P

cd eth18THORN

and

P3cdloc frac14 P3cdfd=Pcd eth19THORN

References

Anderson K and Nelgen S (2011) Global Wine Markets 1961 to 2009 A StatisticalCompendium Adelaide University of Adelaide Press Also available as an e-book atwwwadelaideeduaupresstitlesglobal-wine

Anderson K Nelgen S and Pinilla V (2017) Global Wine Markets 1860 to 2016 AStatistical Compendium Adelaide University of Adelaide Press (forthcoming) Also tobe available as an e-book at wwwadelaideeduaupress

Anderson K and Pinilla V (with the assistance of A J Holmes) (2017)Annual Database ofGlobal Wine Markets 1835 to 2016 Wine Economics Research Centre University ofAdelaide to be posted in October at wwwadelaideeduauwine-econdatabases

Anderson K and Strutt A (2016) Impacts of Asiarsquos rise on African and Latin Americantrade Projections to 2030 World Economy 39(2) 172ndash194

Anderson K andWittwer G (2013) Modeling global wine markets to 2018 Exchange ratestaste changes and Chinarsquos import growth Journal of Wine Economics 8(2) 131ndash158

Anderson K and Wittwer G (2015) Asiarsquos evolving role in global wine markets ChinaEconomic Review 35 1ndash14

250 UK and Global Wine Markets by 2025 and Implications of Brexit

Armington P A (1969) A theory of demand for products distinguished by place of produc-tion IMF Staff Papers 16 159ndash178

Baldwin R E (ed) (2016) Brexit Beckons Thinking Ahead by Leading Economists LondonCentre for Economic Policy Research

Campos N F (2016) Lousy experts Looking back at the ex ante estimates of the costs ofBrexit In R E Baldwin (ed) Brexit Beckons Thinking Ahead by Leading Economists(ch 3) London Centre for Economic Policy Research

Dixon P Parmenter B Sutton J and Vincent D (1982) ORANI A Multisectoral Modelof the Australian Economy Contributions to Economic Analysis 142 Amsterdam North-Holland

European Commission (2015) Trade for All Towards a More Responsible Trade andInvestment Policy Brussels European Commission

Harrison J Horridge M Jerie M and Pearson K (2014) GEMPACK ManualMelbourne GEMPACK Software

HM Treasury (2017) Spring Budget 2017 London HM TreasuryHolmes A J and Anderson K (2017a)Annual Database of National Beverage Consumption

Volumes and Expenditures 1950 to 2015 Wine Economics Research Centre University ofAdelaide posted at wwwadelaideeduauwine-econdatabases

Holmes A J and Anderson K (2017b) Convergence in National Alcohol ConsumptionPatterns New Global Indicators Journal of Wine Economics 12(2) 117ndash148

Ludington C C (2013) The Politics of Wine in Britain A New Cultural History BasingstokeUK Palgrave Macmillan

Office of National Statistics (2017) Analysis of real earnings London Office of NationalStatistics Available at httpswwwonsgovukemploymentandlabourmarketpeopleinworkearningsandworkinghoursarticlessupplementaryanalysisofaverageweeklyearningslatest

Rollo J Borchert I Dawar K Holmes P and Winters L A (2016) The World TradeOrganisation A safety net for a post-Brexit UK trade policy Briefing Paper 1UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-1-final-1pdf

Smith A (2017) Brexit Hard truths and hard choices UKTPO blog 19 June Available athttpsblogssussexacukuktpo20170619brexit-hard-truths-and-hard-choices

Swinbank A (2017) World trade rules and the policy options for British agriculture post-Brexit Briefing Paper 7 UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-7pdf

Viner J (1950) The Customs Union Issue New York Carnegie Endowment for InternationalPeace

Wine Australia (2015) Export Market Guide European Union Adelaide Wine AustraliaWittwer G Berger N and Anderson K (2003) A model of the worldrsquos wine markets

Economic Modelling 20(3) 487ndash506

Kym Anderson and Glyn Wittwer 251

Page 6: U.K. and Global Wine Markets by 2025, and Implications of ... · U.K. and Global Wine Markets by 2025, and Implications of Brexit* Kym Andersona and Glyn Wittwerb Abstract The United

leaves a union the reverse happens because the leaving countryrsquos tariffs now apply toits imports from union countries as well as from the rest of the world Hence importsfrom union countries fall because other things equal these countriesrsquo preferentialaccess to the leaving country no longer applies That is also the case for countriesthat enjoyed an FTA with the union Of significance to wine markets are theEuropean Unionrsquos FTAs with Chile and South Africa

The size of that trade-diverting impact of leaving the union on wine depends onthe external tariff imposed on wine imports not only by the union but also by theleaving country Some have suggested that the United Kingdom should becomethe Hong Kong of Europe and move immediately to free trade on all productsOthers have suggested that this move would impose huge structural changes onthe UK economy that society would not tolerate ndash at least not without majorcompensation packages But all agree that a new trade policy that sets most-favored-nation (MFN) tariff rates is needed before the United Kingdom can beginto negotiate new preferential trading arrangements with the EU27 its FTA partnerssuch as Chile and South Africa or other countries Rollo Borchert Dawar Holmesand Winters (2016) suggest that the most practical trade policy for the UnitedKingdom to adopt at the outset is the European Unionrsquos tariff schedules previouslyagreed to at the WTO In all but one of the Brexit scenarios examined below we

Figure 3

UK Price Relative to World Price of Wine Imports and UK Wine-Import Intensitya 1950 to2015 ()

Fig3-Colou

ron

line

BW

inprint

a Import intensity is defined as the United Kingdomrsquos share of the value of global wine imports divided by the United Kingdomrsquos share of theworldrsquos total merchandise importsSource Compiled from data in Anderson and Pinilla (2017)

226 UK and Global Wine Markets by 2025 and Implications of Brexit

Table1

Shares

ofUKW

ineIm

portsfrom

TodayrsquosKey

Wine-Exp

ortin

gCountries1

675to

2014

andProjected

to2025

with

outandwith

Brexit()

Volum

eFrance

Spain

Portugal

Italy

Germany

South

Africa

Australia

United

States

Chile

New

Zealand

Argentin

aOther

countries

Total

1675ndash1696

2542

231

90

00

00

00

100

1697ndash1862

526

491

32

00

00

014

100

1863ndash1919

2626

220

00

20

00

025

100

1920ndash1940

1218

320

05

140

00

020

100

1995

volume

3210

318

143

73

21

07

100

value

439

413

102

83

21

05

100

2010ndash2014

volume

159

117

48

2110

84

12

100

value

358

215

44

115

66

12

100

Exp

ortersrsquoshareof

world

wineexports2010ndash2014

volume

1520

322

44

74

72

39

100

value

309

319

42

64

53

312

100

2025

projected

noBrexit

100

volume

188

188

26

207

45

45

65

43

64

17

20

92

100

value

323

100

30

218

44

23

58

44

54

41

20

45

100

2025

projected

with

Brexit

volume

190

188

27

208

43

44

66

42

63

17

19

93

100

value

311

96

29

210

41

24

59

44

55

41

20

70

100

SourcesCom

piledfrom

data

inAnd

ersonan

dPinilla

(201

7)to

1940

UnitedNations

COMTRADEh

ttpscomtrad

eun

orgdata

for19

95ndash2

014

andau

thorsrsquomod

elresults

for20

25

Kym Anderson and Glyn Wittwer 227

assume that Rollo et alrsquos suggestion will be the new UK trade policy commitmentto WTO members that subsequent negotiations for preferential arrangements willtake years and that any consequent agreements with the European Union andothers will be gradually implemented after that

The impact of leaving a customs union on wine markets does not only come fromtariff changes Also relevant are any effects of leaving on real UK incomes and thevalue of the British pound If the United Kingdom were to move immediately to freetrade on all products (the Hong Kong option) its per-capita income could eventuallyrise but only after considerable adjustment Should the United Kingdom insteadcommit to the current EU tariff schedule at the WTO in the first instance as weassume below then its per-capita income growth rate and the British poundrsquosexchange rates almost certainly will be lower for some time ndash at least until newtrade agreement negotiations with the EU27 and others are sufficiently advanced

Table 2Taxes on British Wine Imports by Source 1660ndash1862 (British Pounds per Kiloliter)

France Germany Spain Portugal South Africa

1660ndash1665 7 9 8 81666ndash1684 7 9 8 81685ndash1691 14 20 19 181692ndash1695 22 20 19 181696 47 20 19 181697ndash1702 51 25 23 221703 52 27 24 231704ndash1744 55 31 26 251745ndash1762 63 35 30 291763ndash1777 71 39 34 331778 79 43 38 371779 84 41 40 391780ndash1781 92 49 44 431782ndash1785 96 51 47 46 441786 65 51 37 37 371787ndash1794 47 51 32 32 371795 78 64 51 51 571796ndash1797 108 92 71 71 771798 111 96 73 73 791799ndash1801 107 92 71 71 771802 112 97 74 74 801803 131 109 87 87 871804 142 117 95 95 951805ndash1824 144 119 96 96 961825ndash1830 78 50 50 50 251831ndash1859 58 58 58 58 291860 32 32 32 32 321861 16 21 21 21 211862 11 26 26 26 26

Source Summarized from Ludington (2013 Table A1)

228 UK and Global Wine Markets by 2025 and Implications of Brexit

Figure 4

UK Shares in Wine Exports of Key Wine-Exporting Countries 2010ndash2014 ()

Fig4-BW

onlin

eBW

inprint

Source Compiled from data in Anderson and Pinilla (2017)

Table 3Import Duties Excise Duties and VAT Affecting Consumer Prices of Wine and Other Alcohol

in the United Kingdom April 1 2017

VAT 20

MFN import duties on wine pound per liter-bottled still wine lt13 alc 0114-bottled still wine 13ndash15 alc 0134-bottled still wine 15ndash18 alc 0162-bulk still wine lt13 alc 0086-bulk still wine 13ndash15 alc 0105-bulk still wine 15ndash18 alc 0134-sparkling wine 0278Weighted averagea 0130

Excise duties on alcohol pound per liter-still wine lt15 alc 2887-still wine 15ndash22 alc 3848-sparkling wine 55ndash85 alc 2795-sparkling wine 85ndash15 alc 3697-spirits (assumed 40 alc) 11551-beer (assumed 5 alc) 0954

VAT on alcohol 20

a Assumes that one-third of UK wine imports arrives in bulk that half has less than 13 alcohol and that one-tenth of imports is sparkling

Sources HM Revenue httpswwwgovukgovernmentpublicationsalcohol-duty-rate-changes for excise duties and VAT accessed April 92017 and Wine Australia (2015) for import duties converted at the 2016 average exchange rate of pound0740634 per euro

Kym Anderson and Glyn Wittwer 229

as to restore investor and consumer confidence in the United Kingdomrsquos economyThe size of those impacts is uncertain of course (see Baldwin 2016 and especiallyCampos 2016) so we consider a range of possibilities below

Those assumed adverse macroeconomic effects will add to the initial impact ofaltered wine tariffs on aggregate wine consumption in the United Kingdom andhence on its bilateral trades in wine They will make the loss of sales to theUnited Kingdom by EU (and Chilean and South African) suppliers greater thanwould otherwise be the case And they will reduce the likelihood that other countriesrsquosales of wine in the United Kingdom will be higher than in the baseline Indeed themacroeconomic effects could outweigh the trade-diverting effects so that even coun-tries that are currently discriminated against by the EU28rsquos wine-trade policy may beworse off because of Brexit

IV Global Wine Markets Model and Database

Our model of the worldrsquos wine markets first published by Wittwer Berger andAnderson (2003) and revised by Anderson and Wittwer (2013) is summarized inthe Appendix It disaggregates wine markets into four types namely nonpremiumcommercial-premium and superpremium still wines and sparkling wines2 There aretwo types of grapes premium and nonpremium Nonpremium wine uses nonpre-mium grapes exclusively superpremium wines use premium grapes exclusivelyand commercial-premium and sparkling wines use both types of grapes to varyingextents across countries The world is divided into 44 individual nations and 7 com-posite geographic regions that capture all other countries

The modelrsquos database is calibrated to 2014 based on the comprehensive wine-market volume and value data and trade and excise-tax data provided inAnderson and Pinilla (2017) and in Anderson Nelgen and Pinilla (2017) It is pro-jected assuming that aggregate national consumption population and realexchange rates change between 2014 and 2025 to the extent shown in AppendixTable 13 The Brexit alternatives to that baseline also are projected to 2025

Concerning preferences we assume a continued considerable swing toward allwine types in China and a swing away from nonpremium wines in all other countriesuntil 2025

In our baseline scenario grape- and wine-industry total-factor productivity areassumed to grow at 1 per year everywhere while grape- and wine-industry

2Commercial-premium still wines are defined byAnderson Nelgen and Pinilla (2017) to cost between US$250 and US$750 per liter pretax at a countryrsquos border or wholesale3The real exchange rate changes over the projection period are the changes expected in the nominal valueof country irsquos currency relative to the US dollar times the expected ratio of the GDP deflator for theUnited States versus that for country i

230 UK and Global Wine Markets by 2025 and Implications of Brexit

capital is assumed to grow net of depreciation at 15 per year in China but zero else-where (consistent with the almost-zero growth in global wine production and con-sumption over the past two decades)

Two alternative scenarios are considered (ldquolargerdquo and ldquosmallrdquo) for the initialimpact of Brexit to provide a range of results We assume that because of theUnited Kingdomrsquos decision to exit the European Union the rate of UK realgross domestic product (GDP) growth is only one-third or two-thirds as fast overthe projection period (09 or 18 per year instead of 26) and the Britishpound will be 20 or 10 lower in real terms than in our modelrsquos core baselineprojection4 In both alternative scenarios we assume that the United Kingdomapplies the European Unionrsquos external tariffs on wine from March 30 2019 atthe end of the two-year period following the United Kingdomrsquos formal triggeringof Article 50 (see Table 3)

We then consider a subsequent impact of Brexit presumed to result from negoti-ating signing implementing and responding to an FTA with the EU27 by 2025(Chile and South Africa currently have preferential access to EU wine marketsbut we continue to assume in this scenario that the United Kingdom does not imple-ment new bilateral FTAs with them or others in our time frame)5 This subsequentscenario assumes the British pound returns to what it would have been in 2025 in theabsence of Brexit and that real incomes regain two-thirds of the value they wouldhave had without Brexit as compared with our ldquosmallrdquo initial Brexit scenario

This global model has supply-and-demand equations and hence quantitiesprices and price elasticities for each of the grape and wine products and for asingle composite of all other products in each country Income elasticities ofdemand also exist for each final product Grapes are assumed to not be traded inter-nationally but other products are exported and imported Each market is assumed tohave been in equilibrium before any shock and to find a new market-clearingoutcome following any exogenously introduced shock

4The nominal price of the British pound in US dollars in the fortnight following the Brexit vote on June23 2016 dropped 13 to US$130 and a year later the pound sat at the same rate having dropped to anaverage of $124 (an 18 devaluation) between October 2016 and April 2017 Our choice of a low of 10and a high of 20 aims to capture future possible rates while uncertainties remain The average real wagein the United Kingdom fell in the first half of 2017 and projected real GDP growth during 2018ndash2020 hasbeen revised down to 17 in the United Kingdomrsquos latest budget (HM Treasury 2017)5South Africa currently has duty-free access to the EU28 for just 50 megaliters (ML) of wine per yearbeyond which the MFN tariff rate applies South Africa currently exports around 320 ML to theEuropean Union one-third of which initially goes to the United Kingdom How that quota of 50 MLis divided between the United Kingdom and the EU27 is subject to future negotiation (Rollo et al2016 Swinbank 2017) In the following discussion we assume none of it is accepted by the UnitedKingdom so that all South African wine imports pay the United Kingdomrsquos MFN tariff under Brexitbut altering that assumption makes very little difference to our results for any countryrsquos trade exceptSouth Africarsquos

Kym Anderson and Glyn Wittwer 231

V Projecting Global Wine Markets to 2025

Global wine production and exports are projected in the baseline from 2014 to 2025consistent with past trends The modelrsquos global volume of production (and consump-tion) rises little over that 11-year period (9) made up of a 6 decline in nonpre-mium wine and a one-sixth rise in commercial-premium and superpremium wineIn real (2014 US$) value though global wine output and consumption increaseby about 50 in total and 60 in the two premium categories The internationaltrade projections are similar although a little larger with the share of global wineproduction exported (equivalent to the share of global consumption imported)rising 2 percentage points between 2014 and 2025

The baseline projection does not greatly alter the 2014 shares of various countriesin global wine production apart from China because we assume vineyard expansionthere is faster than elsewhere6 In value terms this assumption means that Chinamoves from fifth to fourth by 2025 behind France the United States and ItalySpain remains barely ahead of Australia and they and Germany take the nextthree places (Figure 5(a)) In total wine-production volume terms China movesfrom sixth to fifth place and Argentina drops from fifth to eighth (and fromeighth to ninth in value terms)

When their products are subdivided into fine wines (superpremium still plus spark-ing) commercial-premium wines and nonpremium wines France and the UnitedStates retain the highest two places on the global ladder for fine-wine productionand Spain and Italy retain the top two places for nonpremium wine As for commer-cial-premium wine production (defined to cost between US$250 and US$750 perliter pretax at a countryrsquos wholesale level or national border) Italy retains the topranking over our projections period but at least in terms of value China challengesFrance for second place

The country rankings by projected value of total wine consumption change some-what more than those for production by 2025 with China taking second place afterthe United States ahead of France and Germany and then the United Kingdomslightly overtaking Italy to slip into fifth place (Figure 5(b)) The United StatesFrance and Germany retain the top three rankings for consuming fine wine butCanada slightly overtakes Italy for fourth place in terms of value at least In thecase of commercial-wine consumption China strengthens its number-one positionahead of the United States and the United Kingdom does likewise vis-agrave-visGermany for third place

As for the projected changes in consumption volumes China is projected to dom-inate the increase in aggregate although the United States is projected to lead the

6 In fact Chinarsquos wine production fell steadily between 2012 and 2016 by a total of one-sixth so Chinarsquoswine imports may grow faster in practice than in this baseline projection

232 UK and Global Wine Markets by 2025 and Implications of Brexit

Figure 5

Value of Wine Production and Consumption in Key Countries 2014 and Projected Baseline for2025 (2014 US$ Million at WineryWholesale Pretax Prices)

Fig5-BW

onlin

eBW

inprint

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 233

increase in consumption of fine wine In Western Europe and in the SouthernHemispherersquos New World countries fine wines are projected to substitute for com-mercial wines (defined as the sum of commercial-premium and nonpremiumwines) with almost no change in total wine consumption Sub-Saharan Africa isthe next region that is projected to take off with its growth accounting for morethan one-third of the rest of the worldrsquos increase in volume consumed

Those differences in production versus consumption rankings are reflected ininternational trade Figure 6 shows that France Italy and Spain remain the threedominant exporters of wine in aggregate value but the rankings of the next fewchange with Australia being slightly ahead of Chile and the United StatesGermany and New Zealand being nearly tied for sixth place in value termsFrance and Italy are even more dominant in fine-wine exports and remain so by2025 while Italy outranks France in the commercial-premium export categoryand Spain outranks Italy Australia and Chile in the nonpremium export class

Among the importers the United States and the United Kingdom are projected tocontinue to hold the first two places in 2025 in value terms but China moves intothird place slightly ahead of Germany followed well behind by Canada HongKong Belgium-Luxembourg the Netherlands and Japan (Figure 6(b)) OtherAfrica (excluding South Africa) is projected to experience the largest increase inimports among all the other regions followed by Other Asia which becomes asbig as Germany in value terms (Figure 7(a)) In terms of total volume of wineimports Germany and the United Kingdom held the top two shares in 2014 butby 2025 the United Kingdom is projected to be well ahead of Germany (Figure 7(b)) However this projection ignores the effects of Brexit to which we now turn

VI How Might Wine Markets Be Affected by the United Kingdomrsquos Exitfrom the European Union

As mentioned earlier for our two alternative scenarios to capture the initial effects ofBrexit (ldquolargerdquo and ldquosmallrdquo) we assume that following the United Kingdomrsquos exitfrom the European Union the United Kingdomrsquos rate of economic growth wouldbe only one-third or two-thirds as fast for the period to 2025 the British poundwould be 20 or 10 lower in real terms than in our modelrsquos baseline projectionand the United Kingdom would apply the European Unionrsquos external tariff onwine to imports from EU member countries (as part of establishing MFN ratesvia the WTO to then start new bilateral FTA negotiations) In these initial scenarioswe assume the United Kingdom does not implement any new FTAs particularlywith the EU27 Chile and South Africa Following the discussion of those initialresults we present the results of a subsequent scenario that includes the implemen-tation of an FTA between the United Kingdom and the EU27 We assume these aretwo discrete steps with no agreed arrangements to smooth the transition betweenthem

234 UK and Global Wine Markets by 2025 and Implications of Brexit

Figure 6

Value of Wine Exports and Imports Key Wine-Trading Countries 2014 and Projected Baselinefor 2025 (2014 US$ million)

Fig6-BW

onlin

eBW

inprint

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 235

Figure 7

National Shares of Global Wine Import Value and Volume 2014 and Projected Baselinefor 2025 ()

Fig7-BW

onlin

eBW

inprint

Source Authorsrsquo model results

236 UK and Global Wine Markets by 2025 and Implications of Brexit

A Initial Impact of Brexit

Because the British pound dropped by one-sixth against the US dollar in the fourmonths following the Brexit vote in June 2016 and because the UnitedKingdomrsquos average real wage fell 1 between November 2016 and April 2017(according to the latest available data see Office of National Statistics 2017) weviewed our ldquolargerdquo scenario as more likely soon after the Brexit vote But wecompare those results with ones from our ldquosmallrdquo initial-impact scenario becauseit has become more likely following the June 2017 general election that ldquosofterrdquooptions will be considered by the new government Generally the results areabout half the size in the ldquosmallrdquo scenario with the exception of the bilateraltrade effects To show the sensitivity of results to our assumptions we point outthe differences when the ldquosmallrdquo results are not close to half the results shown forthe ldquolargerdquo scenario

Before turning to those results we first present a ldquofree-traderdquo scenario whichassumes the United Kingdom chooses the radical option of becoming the HongKong of Europe Unlikely as this scenario is the results provide assurance thatour global wine model generates the standard types of effects of exiting a customsunion In this ldquofree-traderdquo scenario the United Kingdom is assumed to move tozero tariffs on all wine imports Thus the current preferences on EU Chileanand South African wine imports disappear as all other wine exporters enjoy thesame free access to UK wine markets We assume in this scenario that Brexit hasno impact on the United Kingdomrsquos real income or consumption growth rates by2025 and that by then the British pound has returned to its 2014 value relative toother currencies (similar to the real exchange rate assumed in our baseline scenarioto 2025 see Appendix Table A1)7 Under these assumptions EU Chilean andSouth African wine exports to the United Kingdom in 2025 are 003 lower thanin the no-Brexit baseline and UK imports from other wine exporters are 025higher but the United Kingdomrsquos total wine imports are hardly any different(because local wine prices fall by only 05 and incomes are unchanged) In this sce-nario UK consumers expand their volume of consumption by just 05 while theEU27 exports more and other countries export less to the rest of the world such thatworld wine exports are almost unchanged In short all these effects are in the direc-tion that comparative static-customs-union theory would predict but they are smallbecause we assume the United Kingdomrsquos real income and currency are the same asin our baseline for 2025 as described in the previous section

In the ldquolargerdquo alternate initial scenario involving Brexit as compared with thebaseline scenario to 2025 the consumer price of wine in 2025 is 22 higher in the

7If this scenario were seen as a credible long-run trade policy it would encourage more investment thatwould eventually raise UK incomes but not without major structural changes that would initiallydisrupt the economy Implicitly we assume in our free-trade scenario that these two opposite impactson real UK incomes and the British pound exactly offset each other by 2025

Kym Anderson and Glyn Wittwer 237

United Kingdom in local currency terms (20 because of real depreciation of theBritish pound 4 because of new tariffs on EU Chilean and South Africanwines and ndash2 because of slower UK income growth) The volume of UKwine consumption is 28 lower 16 because of slower UK economic growth7 because of real depreciation of the British pound and 5 because of newtariffs Superpremium still-wine sales are the most affected dropping by two-fifths while sparkling and commercial-premium wines drop a bit less than one-quarter Because the average price rises by more than the decline in volume soldthe aggregate value of UK sales even in local currency terms falls under thisldquolargerdquo Brexit scenario Under the ldquosmallrdquo Brexit scenario the consumer price ofwine in 2025 is 11 higher in the United Kingdom and its volume of wine consump-tion is 17 lower

The volume of projected UK imports in 2025 is 427 ML or nearly one-quarterless in the ldquolargerdquo scenario than in the baseline scenario comprising 58ML less spar-kling wine 31 ML less superpremium still wine and 339 ML less commercial-premium wine World imports are lower by just 239 ML because imports byother countries are 189 ML higher in response to the lower international wineprices in this scenario In value terms UK imports are $175 billion (or 27)lower in 2025 because of ldquolargerdquo Brexit $113 billion because of lower incomes$038 billion because of the fall in the British pound and $014 billion because ofthe rise in wine-import tariffs (Table 4) These aggregate trade impacts are a littlemore than half this size under the ldquosmallrdquo Brexit scenario

Despite the import levels falling because of raised import tariffs domestic con-sumption of all three quality categories of UK-produced wine is lower with thanwithout Brexit because of shrunken demand for all wines resulting from loweredUK incomes and raised local prices because of devaluation of the British poundThe British poundrsquos devaluation does make it easier for the United Kingdom tosell wines abroad though The countryrsquos exports are 7 ML or nearly 5 higherin 2025 in the ldquolargerdquo Brexit scenario and UK production is 3 higher ThoseUK exports (or re-exports of imported bulk wine after it is bottled in the UnitedKingdom) that go to EU27 countries are reduced though because of the tariffnow imposed at the new EU border

Without Brexit the United Kingdomrsquos shares of global wine imports are slightlyhigher in volume terms in 2025 than in 2010ndash2015 but 2 percentage points lower invalue terms thanks to East Asiarsquos expanding demand for imports of premium winesWith ldquolargerdquo Brexit however that value share is an additional 2 percentage pointslower and the volume share is almost 5 points lower (Figure 8) The net effects ofthese impacts on global trade are shown in Figure 9 Most of the initial tradeeffect of Brexit is a large decline in net imports of wine by the United Kingdomwith very little offsetting positive effect on trade in the rest of the world Theldquosmallrdquo Brexit numbers are a bit more than half of those for ldquolargerdquo Brexit

238 UK and Global Wine Markets by 2025 and Implications of Brexit

Table4

Difference

in2025

Projected

VolumeandValue

ofWineIm

portsb

ytheUnitedKingdom

andtheResto

fthe

World

asaConsequence

oftheInitial

BrexitS

hock

(MLand2014US$

Million

ldquoLargerdquoSc

enario)

Volum

e(M

L)

Value

(US$

million)

NP+CPa

SuperPrb

Sparkling

TOTAL

NP+CPa

SuperPrb

Sparkling

TOTAL

ΔUK

impo

rtsdu

eto

Low

erincomes

minus198

minus20

minus29

minus247

58minus644

minus253

minus234

minus1131

65Low

erpo

und

minus70

minus10

minus14

minus93

22minus248

minus127

minus102

minus476

27Highertariffs

minus71

minus1minus16

minus87

20minus110

minus8minus2

4minus143

8TOTA

Lminus339

minus31

minus58

minus427

100

minus1001

minus388

minus360

minus1750

100

difffrom

base

2332

3325

2432

3227

of

totalc

uts

797

14100

5722

21100

ΔROW

netim

ports

143

2125

189

230

143

minus181

192

ΔWORLD

TRADE

minus195

minus10

minus34

minus239

minus763

minus246

minus543

minus1552

aNon

prem

ium

plus

commercial-premium

still

wines

bSu

perpremium

still

wines

SourceA

utho

rsrsquomod

elresults

Kym Anderson and Glyn Wittwer 239

The aggregate effect of ldquolargerdquo Brexit on the market shares of various wine-export-ing countries in the United Kingdom is almost indiscernible even with one decimalpoint The projected 2025 shares are quite different from the actual 2014 shares forseveral countries They are much smaller in 2025 for South Africa Australia andNew Zealand (and the United States in volume terms) and are much larger involume for Spain and in value for Italy These results occur because wine-exportingcountries benefit differentially from the varying rates of growth in net importdemand for wine in non-UK countries over this projection period The most impor-tant projected changes are the increase in the real value of annual wine importsbetween 2014 and 2025 by China (200 or $3 billion) Other Asia (110 or $22billion) and Africa (270 or $16 billion) More than half of Australiarsquos increasein annual exports from 2014 to 2025 go to Asia and more than half of SouthAfricarsquos increase in exports go to Other Africa

Table 5 reveals that ldquolargerdquo Brexit lowers EU Chilean and South African wineexports by 150 ML or US$12 billion in the case of the European Union with someof their exports diverted from the United Kingdom to the EU27 and other marketsin competition with New World exporters Although the United States Australiaand Argentina sell only a little less to the United Kingdom they also sell less toother countries For Chile and South Africa which lose their preferential access toUK (but not to EU27) markets in this Brexit scenario some exports are redirected

Figure 8

UK Shares of World Wine Imports 2010ndash2015 and Projected to 2025 without and with theldquoLargerdquo Initial Brexit Shock ()

Fig8-BW

onlin

eBW

inprint

Sources Anderson and Pinilla (2017) and authorsrsquo model results

240 UK and Global Wine Markets by 2025 and Implications of Brexit

from the United Kingdom to EU27 countries ndash but again they export less overallGlobal wine trade in 2025 is 240 ML (19) or $18 billion (35) less under thisldquolargerdquo Brexit scenario The percentage by which wine exportersrsquo trade shrinks isgreater for values than for volumes because of changes in relative prices of differ-ent-quality wines Those differences are shown in the numbers in parentheses inTable 5

Three other points are worth making about Table 5 One is that Australia sellsslightly more to the United Kingdom in the ldquosmallrdquo Brexit scenario rather thanslightly less as in the ldquolargerdquo Brexit case Evidently the negative income and price(devaluation) effects do not more than offset the positive trade-diverting effect onAustralian exports to the United Kingdom of removing preferences in the ldquosmallrdquoscenario Second New Zealand sells slightly more to non-UK countries underBrexit despite greater competition from the EU27 Chile and South Africa Thisanomaly is due to changes in the relative prices of different qualities of wine inglobal wine markets bearing in mind that New Zealand has the worldrsquos highestaverage price for still-wine exports And third the value (but not the volume) ofexports of ldquoOtherrdquo countries to markets other than the United Kingdom arehigher under Brexit This result too is due to changes in the relative prices of differ-ent qualities of wine in global wine markets

Figure 9

Difference in 2025 Wine Import Volumes and Values as a Result of the ldquoLargerdquo Initial BrexitShock (ML and US$ Million in 2014 US dollars)

Fig9-BW

onlin

eBW

inprint

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 241

B Subsequent Impact of Brexit from a UK-EU27 FTA

The next-most-likely step in the Brexit process is for the United Kingdom to nego-tiate a new trade arrangement with the EU27 We therefore assume that a UK-EU27FTAwith free bilateral wine trade is implemented and adjusted to by 2025 and thatprogress toward that end occurs soon enough that the adverse macroeconomicshocks from the initial impact of uncertainty over the Brexit process are confinedto those assumed in the ldquosmallrdquo scenario outlined above In this subsequent scenariowe assume that the British pound returns to the value it would have reached in 2025in the absence of Brexit (ie reversing the 10 devaluation assumed in the ldquosmallrdquoinitial Brexit scenario) and that real incomes in the United Kingdom are 8 higher

Table 5Difference in 2025 BilateralWine Import Volumes and Values fromKey Exporters by the United

Kingdom and the Rest of the World (RoW) as a Result of Initial Brexit Shock(ML and 2014US$ Million)a

(a) ldquolargerdquo scenario

Volume (ML) Value (2014US$ million)

UK RoW WORLD () UK RoW WORLD ()

EU27 minus287 136 minus150 (ndash17) minus1187 minus5 minus1192 (ndash31)Chile minus59 35 minus25 (ndash30) minus169 31 minus138 (ndash48)South Africa minus53 35 minus18 (ndash32) minus105 20 minus85 (ndash67)USA minus7 minus6 minus13 (ndash24) minus75 minus40 minus115 (ndash50)Australia minus4 minus3 minus7 (ndash09) minus25 minus65 minus90 (ndash30)Argentina minus3 minus9 minus12 (ndash48) minus16 minus39 minus55 (ndash52)New Zealand minus11 9 minus2 (ndash09) minus162 71 minus91 (ndash43)Others minus2 minus10 minus12 (ndash02) minus11 minus52 minus63 (ndash44)WORLD minus427 187 minus240 (ndash19) minus1750 minus79 minus1829 (ndash35)

(a) ldquosmallrdquo scenario

Volume (ML) Value (2014US$ million)

UK RoW WORLD () UK RoW WORLD ()

EU27 minus178 82 minus96 (ndash12) minus692 minus43 minus736 (ndash19)Chile minus46 28 minus18 (ndash24) minus128 36 minus91 (ndash32)South Africa minus43 29 minus14 (ndash42) minus82 23 minus59 (ndash47)USA 1 minus6 minus5 (ndash11) minus23 minus28 minus51 (ndash22)Australia 5 minus10 minus5 (ndash06) 19 minus56 minus38 (ndash13)Argentina 0 minus6 minus6 (ndash26) minus3 minus25 minus29 (ndash27)New Zealand minus5 4 minus1 (ndash06) minus80 34 minus46 (ndash22)Others 0 minus9 minus9 (ndash01) minus1 minus33 minus34 (ndash24)WORLD minus266 112 minus154 (ndash13) minus991 minus92 minus1083 (ndash21)

a Numbers in parentheses are the percentage difference between the Brexit and baseline scenarios for 2025 projected wine-import volumes orvalues by source

Source Authorsrsquo model results

242 UK and Global Wine Markets by 2025 and Implications of Brexit

in 2025 than in the ldquosmallrdquo scenario ndash that is they regain two-thirds of the value theywould have reached without Brexit-related uncertainty

This subsequent development in the Brexit process reverses most of but not allthe initial effects of Brexit by 2025 because of our assumption that the lost growthin the initial years of uncertainty following the Brexit vote are only partly recoveredby 2025 following the implementation of a UK-EU27 FTA Moreover the longer ittakes before this FTA is finalized and implemented the longer the estimated initialadverse effects persist and the larger the cumulative cost of Brexit to UK wine con-sumers and to grape and wine producers in wine-exporting countries

Table 6 summarizes the subsequent trade effects for 2025 It suggests that all butone-tenth of the loss in value of world trade in wine from the initial ldquosmallrdquo impact isrestored most of it because of a smaller reduction in the United Kingdomrsquos wineimports following the FTA Most of that improved outcome is because of recoveredimports from the EU27 commensurate with the latterrsquos high share of UK importsEven though Chile and South Africa are assumed in this scenario to have not yetsigned an FTA with the United Kingdom they export slightly more to the UnitedKingdom (and even more to all other countries) than in the initial ldquosmallrdquo scenario

Within the United Kingdom this FTA brings down the local currency consumerprice of wine by 9 largely offsetting the 11 rise in the initial ldquosmallrdquo Brexit sce-nario and it raises the volume consumed in the United Kingdom by 18 fully off-setting the 17 fall in the initial ldquosmallrdquo Brexit scenario

Table 6Difference in 2025 Bilateral Wine-Import Volumes and Values from Key Exporters by the

United Kingdom and the Rest of the World (RoW) as a Result of Implementing a UK-EU27FTA Following Initial ldquoSmallrdquo Brexit Shock (ML and 2014US$ Million Difference Relative to

Initial ldquoSmallrdquo Brexit Scenario)a

Volume (ML) Value (2014US$ million)

UK RoW WORLD (ldquosmallrdquo)a UK RoW WORLD (ldquosmallrdquo)a

EU27 212 minus112 100 (minus96) 750 0 750 (minus806)Chile 1 3 4 (minus18) 3 40 43 (minus92)South Africa 3 2 5 (minus14) 3 18 21 (minus60)USA 3 3 6 (minus5) 26 21 47 (minus50)Australia 1 1 2 (minus5) 8 38 46 (minus33)Argentina 1 4 5 (minus6) 4 20 24 (minus28)New Zealand 4 minus3 1 (minus1) 61 minus25 36 (minus46)Others 1 6 7 (minus9) 5 minus46 minus41 (92)WORLD 226 minus96 130 (minus154) 860 66 926 (minus1022)

a Numbers in parentheses are the world trade differences between the ldquosmallrdquo initial Brexit scenario and the baseline scenario copied fromcolumns 3 and 7 of Table 5(b)

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 243

In short this subsequent step in the Brexit process can be expected to restore muchof the initial adverse effects in the United Kingdom of the Brexit vote by 2025 Thewine exporters in the EU27 Chile and South Africa also lose less in this scenarioas compared with the initial ldquosmallrdquo scenario while other key exporters sell nearlyas much in 2025 in this scenario as they are projected to in the baseline ndash that isthey are beneficiaries of the reduced discrimination in global wine markets in thiscase where Chile and South Africa no longer enjoy preferential access to the UKmarket (and even though EU27 producers are again allowed to do so) Keep inmind though that although the estimated losses in the initial scenarios are reportedjust for 2025 they are expected to be felt throughout the preceding years of uncertaintyTherefore the longer it is before uncertainty abates and the United Kingdomrsquos currentpreferential trading arrangements are replaced by newones such as a UK-EU27 FTAthe higher the cumulative cost of the Brexit vote to the wine trade

VII Caveats and Conclusions

The above Brexit simulations are just a few of many scenarios that could be modeledObvious additional ones could also assume that FTAs are reached between theUnited Kingdom and other trading partners including Chile and South Africaand such countries as Australia and New Zealand Some of these other countriesare already exploring the ECrsquos invitation (see European Commission 2015) to nego-tiate an FTA with what will be the EU27 The sequence in which FTAs are signedand the speed with which they are implemented will matter (as was also the casewith the sequential signing over the past decade of bilateral FTAs with NortheastAsian countries by Chile Australia and New Zealand see Anderson andWittwer 2015) Even if the United Kingdom were able to sign additional bilateralFTAs and begin implementing them before 2025 it would have little effect on theabove results (because wine tariffs are such a minor contributor to them) unlessthose FTAs were to accelerate the United Kingdomrsquos economic growth and therise in the British poundrsquos value before 2025

We assume above that no changes are made to alcohol excise duties in the UnitedKingdom following Brexit even though they are scheduled to be progressively raisedwith inflation and may be raised even more for wine relative to spirits to offset theopposite effects of Brexit on those two domestic industries Nor do we make anyallowance for UK increases in consumer wine prices that may be needed to coverthe higher cost of clearing customs on imports from EU27 countries (which maybe more or less offset by increased smuggling and duty-free purchasing)Importers of fine wines including individuals who buy Bordeaux en primeur maywell reduce their demand because of concerns that delays on docks will affect thequality of their wines or because of greater currency uncertainty

To summarize Brexit is costly initially to UK consumers of wine (and of manyother tradable products) because the domestic retail price in local currency tax-

244 UK and Global Wine Markets by 2025 and Implications of Brexit

inclusive terms is 22 higher than otherwise in 2025 and the volume of wine con-sumed domestically is 28 lower in the ldquolargerdquo Brexit scenario (or 11 higher pricesand 17 lower quantities in the ldquosmallrdquo case) by 2025 unless a UK-EU27 FTA isdeveloped before then Even if such an FTA does get signed ratified by all 28 par-liaments and implemented by 2025 the slower income growth in the interim meansa smaller UK wine market in 2025 than would otherwise have been the case Thevolume reduction is a blow to many participants in UK wine-bottling transportingstoring wholesaling and retailing businesses in addition to restaurants and pubsVery little of that initial impact occurs because of higher import tariffs most impor-tant is the assumed fall in UK real incomes

The small but growing number of local UK vignerons (who supply less than 05of all domestic-wine sales and less than 4 of sparkling-wine sales) is initially pro-jected to sell less wine domestically because of reduced demand and to export only alittle extra abroad while the British pound is devalued Overall their production isonly 3 higher in 2025 even in the ldquolargerdquo Brexit scenario (not taking intoaccount the fact that their casual labor is likely to be more expensive under Brexitbecause of tighter restrictions on immigration) and their production diminishes ifand when a UK-EU27 FTA is in place

Even with a UK-EU27 FTA in place EU27 wine exporters export US$56 millionless wine in 2025 thanks to Brexit Chile and South Africa export $88 million lesswine and wine exports from the rest of the world are little different in aggregatealthough slightly larger from Australia

Clearly there will be great uncertainly for some time over the possible policy out-comes to flow from Brexit and their consequent sequential impacts on UK house-hold disposable incomes foreign exchange rates and bilateral wine tariffsMeanwhile the above projections under explicit assumptions provide some idea ofhow wine markets might be affected by the most-likely first two stages of theBrexit process (agreeing on a new tariff schedule at the WTO and agreeing to andimplementing a UK-EU27 FTA) In particular they make clear that nontrivialinitial impacts could affect the domestic wine market impacts that are likely to belarger than just the direct impact of changes in bilateral tariffs If the UnitedKingdom succeeds in getting countries to agree to transition arrangements thatdelay the changes in tariffs until new FTAs are signed and ratified the initialeffects will be less dramatic than in our first scenarios but it remains to be seenwhether any such agreements can be reached In any event the net effect of Brexiton the welfare of the worldrsquos consumers and producers of wine as a whole will benegative not just initially but permanently unless new trade policy commitmentsby the United Kingdom with major wine-exporting countries are sufficiently moreliberal than current arrangements

Kym Anderson and Glyn Wittwer 245

Appendix Key Equations in the Global Wine Markets Model

Amodel of the worldrsquos wine markets was first published byWittwer et al (2003) andmuch improved by Anderson and Wittwer (2013) Several significant enhancementshave been made to that original model Wine types have been disaggregated intomore types ndash namely nonpremium (including generic bulk) commercial-premiumand superpremium still wines and sparkling wines8 As in the original modelthere are two types of grapes premium and nonpremium Nonpremium wines usenonpremium grapes exclusively superpremium wines use mainly premium grapesand commercial-premium and sparkling wines use both types of grapes In theregional dimension the number of countries and country groups has expandedfrom 10 in the original model to 51 44 individual nations and 7 compositeregions The modelrsquos database is calibrated to 2014 for this paper based on datain Anderson et al (2017) trade data for which are downloaded from httpscomtradeunorg The modelrsquos supply-and-demand equations are based on the orig-inal ORANI modelrsquos theory (see Dixon Parmenter Sutton and Vincent 1982) andthe model is implemented using GEMPACK software (Harrison Horridge Jerieand Pearson 2014)

An enhancement of importance to the present study is the inclusion of exchange-rate variables in the model which allow a distinction between price impacts asobserved in local currency units and those observed in 2014 US dollars

In the model the grape and wine sectors minimize costs of intermediate inputssubject to weak constant elasticity of substitution (CES) substitutability betweeninputs By assumption no intermediate inputs are imported from other countries9Hence

Xcid frac14 f ethX1id CESfrac12P

cid=P1id $THORN eth1THORN

P1id X1id frac14X

c

Xcid P

cid eth2THORN

whereXcid is the quantity demanded of commodity c by grape or wine industry i in

region d Pcid is the corresponding price and X1id and P1id are the respective inter-

mediate composite quantities and prices

Two primary factors are employed in the sector labor (the quantity of which isendogenous with perfectly elastic supply) and capital Capital is usually treated asexogenous in quantity with rates of return bearing all the adjustment in thevarious scenarios This reflects the fact that grapes (a perennial crop) and wine-

8Commercial-premium still wines are defined by Anderson and Nelgen (2011) to be those priced betweenUS$250 and $750 per litre pre-tax at a countryrsquos border or wholesale9An exception concerns bulk wine imports used mainly in the bottling of commercial premium wine in theUK USA Canada Japan and Germany

246 UK and Global Wine Markets by 2025 and Implications of Brexit

plant capacity adjust slowly to market signals

Lid frac14 f ethFid CESfrac12Wid=PFid $THORN eth3THORN

Kid frac14 f ethFid CESethRid=PFid $THORN eth4THORN

PFid Fid frac14 Lid Wid thorn Kid Rid eth5THORN

Grape and wine producers are assumed to minimize costs subject to CES substitu-tion between capital and labor Equations (3) to (5) show primary factor demandsfor the labor composite Lid and capital Kid subject to a composite factor demandFid by industry i in region d The factor prices are Wid for labor Rid for capitalrentals and PFid for composite factor prices

Appendix Table A1Cumulative Consumption and Population Growth Rates and Changes in the Real Exchange Rate

(RER) Relative to the US Dollar 2014 to 2025 without Brexit ()

Aggregateconsumption Poprsquon RER

Aggregateconsumption Poprsquon RER

France 18 4 minus11 Australia 35 11 minus17Italy 11 2 minus9 New Zealand 32 9 minus26Portugal 14 0 minus9 Canada 27 8 minus18Spain 26 8 minus9 United States 31 8 0Austria 19 4 minus7 Argentina 7 10 109Belgium 20 7 minus9 Brazil 16 8 minus29Denmark 22 2 minus9 Chile 55 8 minus2Finland 21 3 minus7 Mexico 42 12 minus8Germany 14 minus2 minus11 Uruguay 45 3 1Greece 22 minus1 minus14 Other Latin America 60 10 minus5Ireland 42 12 minus9 South Africa 36 12 minus1Netherlands 21 4 minus9 Turkey 50 8 20Sweden 24 9 minus13 North Africa 53 11 0Switzerland 18 8 minus6 Other Africa 109 18 84United Kingdom 32 6 1 Middle East 52 18 minus12Other West Europe 21 10 minus1 China 79 3 5Bulgaria 41 minus7 7 Hong Kong 42 3 2Croatia 20 minus2 minus1 India 134 13 17Georgia 35 0 23 Japan 11 minus3 minus24Hungary 25 minus3 minus11 Korea 38 1 minus9Moldova 49 minus11 13 Malaysia 62 15 minus16Romania 45 minus4 22 Philippines 75 18 7Russia 18 minus2 minus8 Singapore 44 21 minus22Ukraine 22 minus5 14 Taiwan 29 1 minus13Other East Europe 40 minus5 48 Thailand 47 3 minus9

Other Asia 99 10 10

Source Authorsrsquo compilation from projections by various international agencies and from global economy-wide modeling by Anderson andStrutt (2016)

Kym Anderson and Glyn Wittwer 247

The composite factor demand Fid is proportional to total output Qid subject to aprimary factor using technology Aid Hence

Fid frac14 Qid Aid eth6THORN

The perfectly competitive zero-pure-profit condition is that total revenue valued atthe output price P0s

i multiplied by Qid equals the total production cost

P0si Qid frac14

X

cPcid X1cid thorn

X

oWo

id Loid thorn Rid Kid eth7THORN

Household demands follow a linear expenditure system in each region We reducethe optimizing problem for household consumption of each commodity subjectto a budget constraint to equations describing subsistence and discretionarydemands Aggregate subsistence expenditure WSUBd depends only on consumerprices P3cd for each commodity and the number of households N as per-capita sub-sistence quantities XSUBcd subject to given preferences are constant

WSUBd frac14X

c

P3cd XSUBcd Nd eth8THORN

Discretionary expenditures for each commodity (the left-hand side of equation (9))are equal to the marginal budget share (βcd) of aggregate discretionary expenditureThis aggregate is the bracketed term on the right-hand side of equation (9) whereW3TOTd is aggregate nominal expenditure

P3cdethX3cd XSUBcd NdTHORN frac14 βcdethW3TOTd WSUBdTHORN eth9THORN

Because real aggregate consumption is usually exogenous in our partial equilibriumsimulations the linear expenditure system determines the consumption shares ofindividual final commodities (ie the five wine types plus a composite of all otherconsumption items) driven by changes in relative prices as faced by domestic con-sumers The income elasticity of demand for each commodity is equal to the mar-ginal budget share divided by the expenditure share This number varies from 05for nonpremium wine to 17 for superpremium still wine The income elasticity ofdemand for other consumption is very close to 10 because wine accounts for anaverage of only 03 of aggregate expenditures globally and no more than 11in any country (Anderson and Nelgen 2011 Table 166)

A feature of our revised model of world wine markets is the inclusion of nominalexchange rates These appear directly in the equation linking retail prices (P3scd ) toproducer prices by country of origin (P0s

c ) where c denotes the wine type

P3 scd frac14 P0s

cfd

fsTtarcd T

taxcd thorn Pm

cd eth10THORN

248 UK and Global Wine Markets by 2025 and Implications of Brexit

The exchange rates in the consuming and producing regions are fd and fs respec-tively expressed as local currency units per $US Ttar

cd is the power of the tariff in theconsuming region and Ttax

cd is the power of the domestic consumption (or excise) taxprior to any generic value-added or goods-and-services tax Pm

cd is the price of marginm assumed to be locally supplied nontradable and therefore unaffected by theexchange rate

A given level of consumption for wine type c (X3cd) is satisfied using theArmington (1969) assumption in which wines from different countries of originare imperfectly substitutable First domestic wine is imperfectly substitutable witha composite of imports

X3sscd frac14 f ethX3cd CESethP3sscd=P3cdTHORNTHORN ss frac14 domestic imports eth11THORN

and then imports by origin (X3scd ) are determined in a second CES equation

X3scd frac14 f ethX3ssfrac14importscd CESethP3scd=P3

ssfrac14importscd THORNTHORN eth12THORN

The model enables us to show how changes in international competitiveness affectthe worldrsquos wine markets A crucial part of this exercise is explaining how pricesdetermined outside the grape and wine markets influence these markets Becausethe model is in partial equilibrium to depict the impacts of changes in internationalcompetitiveness outside price changes need to be imposed as shocks on the modelThe price of intermediate inputs other than grapes and wine shown in equations (1)and (2) is set equal to the price of GDP (Pg

d ) multiplied by a shifter Fcd

Pcid frac14 Fc

dPgd eth13THORN

If no specific price observations are available the shifter Fcd remains exogenous and

unshocked with the change in price being determined by a shock to the price ofGDP If observations are available for specific input price movements the shifterFc

dbecomes endogenous with Pc

id now exogenous and shocked

Wid frac14 Fwd P

gd eth14THORN

Wage rates are treated similarly In equation (14) if the wage shifter Fwd is exogenous

changes in wage rates Wid are determined by changes in the price of GDP If wagerate data are available Fw

d becomes endogenous andwage rates are shocked directly

Pmcd frac14 Fm

d Pgd eth15THORN

The prices of trade and transport margins are also determined by the price of GDP ifthe shifter Fm

d in equation (15) is exogenous

Changes in international competitiveness depend on changes in relative pricelevels and changes in nominal exchange rates In equation (16) fR

s denotes real

Kym Anderson and Glyn Wittwer 249

exchange-rate movements relative to the US dollar

fRs frac14 Pg

s =frac12PgUSA

fs$ eth16THORN

In equation (16) the nominal exchange rate for the United States is alwaysunchanged because nominal and real exchange rates are expressed relative to theUS currency

Changes in international market conditions may have impacts in one direction onproducer prices as expressed in US dollars P0s

i and potentially in the opposite direc-tion in local currency units Hence we calculate real producer pricesP0s

iloc in localcurrency terms (ie the price most relevant to domestic producers)

P0siloc frac14 P0s

ifs=P

gs eth17THORN

To obtain real price changes in local currency terms we convert US dollar prices(P3 s

cd for the source-specific price and P3cd for the source-composite price) to reallocal currency prices (P3 s

cdloc and P3cdloc) using the consumer price index (CPI)(Pc

d ) as the deflator

P3 scdloc frac14 P3 s

cdfd=P

cd eth18THORN

and

P3cdloc frac14 P3cdfd=Pcd eth19THORN

References

Anderson K and Nelgen S (2011) Global Wine Markets 1961 to 2009 A StatisticalCompendium Adelaide University of Adelaide Press Also available as an e-book atwwwadelaideeduaupresstitlesglobal-wine

Anderson K Nelgen S and Pinilla V (2017) Global Wine Markets 1860 to 2016 AStatistical Compendium Adelaide University of Adelaide Press (forthcoming) Also tobe available as an e-book at wwwadelaideeduaupress

Anderson K and Pinilla V (with the assistance of A J Holmes) (2017)Annual Database ofGlobal Wine Markets 1835 to 2016 Wine Economics Research Centre University ofAdelaide to be posted in October at wwwadelaideeduauwine-econdatabases

Anderson K and Strutt A (2016) Impacts of Asiarsquos rise on African and Latin Americantrade Projections to 2030 World Economy 39(2) 172ndash194

Anderson K andWittwer G (2013) Modeling global wine markets to 2018 Exchange ratestaste changes and Chinarsquos import growth Journal of Wine Economics 8(2) 131ndash158

Anderson K and Wittwer G (2015) Asiarsquos evolving role in global wine markets ChinaEconomic Review 35 1ndash14

250 UK and Global Wine Markets by 2025 and Implications of Brexit

Armington P A (1969) A theory of demand for products distinguished by place of produc-tion IMF Staff Papers 16 159ndash178

Baldwin R E (ed) (2016) Brexit Beckons Thinking Ahead by Leading Economists LondonCentre for Economic Policy Research

Campos N F (2016) Lousy experts Looking back at the ex ante estimates of the costs ofBrexit In R E Baldwin (ed) Brexit Beckons Thinking Ahead by Leading Economists(ch 3) London Centre for Economic Policy Research

Dixon P Parmenter B Sutton J and Vincent D (1982) ORANI A Multisectoral Modelof the Australian Economy Contributions to Economic Analysis 142 Amsterdam North-Holland

European Commission (2015) Trade for All Towards a More Responsible Trade andInvestment Policy Brussels European Commission

Harrison J Horridge M Jerie M and Pearson K (2014) GEMPACK ManualMelbourne GEMPACK Software

HM Treasury (2017) Spring Budget 2017 London HM TreasuryHolmes A J and Anderson K (2017a)Annual Database of National Beverage Consumption

Volumes and Expenditures 1950 to 2015 Wine Economics Research Centre University ofAdelaide posted at wwwadelaideeduauwine-econdatabases

Holmes A J and Anderson K (2017b) Convergence in National Alcohol ConsumptionPatterns New Global Indicators Journal of Wine Economics 12(2) 117ndash148

Ludington C C (2013) The Politics of Wine in Britain A New Cultural History BasingstokeUK Palgrave Macmillan

Office of National Statistics (2017) Analysis of real earnings London Office of NationalStatistics Available at httpswwwonsgovukemploymentandlabourmarketpeopleinworkearningsandworkinghoursarticlessupplementaryanalysisofaverageweeklyearningslatest

Rollo J Borchert I Dawar K Holmes P and Winters L A (2016) The World TradeOrganisation A safety net for a post-Brexit UK trade policy Briefing Paper 1UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-1-final-1pdf

Smith A (2017) Brexit Hard truths and hard choices UKTPO blog 19 June Available athttpsblogssussexacukuktpo20170619brexit-hard-truths-and-hard-choices

Swinbank A (2017) World trade rules and the policy options for British agriculture post-Brexit Briefing Paper 7 UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-7pdf

Viner J (1950) The Customs Union Issue New York Carnegie Endowment for InternationalPeace

Wine Australia (2015) Export Market Guide European Union Adelaide Wine AustraliaWittwer G Berger N and Anderson K (2003) A model of the worldrsquos wine markets

Economic Modelling 20(3) 487ndash506

Kym Anderson and Glyn Wittwer 251

Page 7: U.K. and Global Wine Markets by 2025, and Implications of ... · U.K. and Global Wine Markets by 2025, and Implications of Brexit* Kym Andersona and Glyn Wittwerb Abstract The United

Table1

Shares

ofUKW

ineIm

portsfrom

TodayrsquosKey

Wine-Exp

ortin

gCountries1

675to

2014

andProjected

to2025

with

outandwith

Brexit()

Volum

eFrance

Spain

Portugal

Italy

Germany

South

Africa

Australia

United

States

Chile

New

Zealand

Argentin

aOther

countries

Total

1675ndash1696

2542

231

90

00

00

00

100

1697ndash1862

526

491

32

00

00

014

100

1863ndash1919

2626

220

00

20

00

025

100

1920ndash1940

1218

320

05

140

00

020

100

1995

volume

3210

318

143

73

21

07

100

value

439

413

102

83

21

05

100

2010ndash2014

volume

159

117

48

2110

84

12

100

value

358

215

44

115

66

12

100

Exp

ortersrsquoshareof

world

wineexports2010ndash2014

volume

1520

322

44

74

72

39

100

value

309

319

42

64

53

312

100

2025

projected

noBrexit

100

volume

188

188

26

207

45

45

65

43

64

17

20

92

100

value

323

100

30

218

44

23

58

44

54

41

20

45

100

2025

projected

with

Brexit

volume

190

188

27

208

43

44

66

42

63

17

19

93

100

value

311

96

29

210

41

24

59

44

55

41

20

70

100

SourcesCom

piledfrom

data

inAnd

ersonan

dPinilla

(201

7)to

1940

UnitedNations

COMTRADEh

ttpscomtrad

eun

orgdata

for19

95ndash2

014

andau

thorsrsquomod

elresults

for20

25

Kym Anderson and Glyn Wittwer 227

assume that Rollo et alrsquos suggestion will be the new UK trade policy commitmentto WTO members that subsequent negotiations for preferential arrangements willtake years and that any consequent agreements with the European Union andothers will be gradually implemented after that

The impact of leaving a customs union on wine markets does not only come fromtariff changes Also relevant are any effects of leaving on real UK incomes and thevalue of the British pound If the United Kingdom were to move immediately to freetrade on all products (the Hong Kong option) its per-capita income could eventuallyrise but only after considerable adjustment Should the United Kingdom insteadcommit to the current EU tariff schedule at the WTO in the first instance as weassume below then its per-capita income growth rate and the British poundrsquosexchange rates almost certainly will be lower for some time ndash at least until newtrade agreement negotiations with the EU27 and others are sufficiently advanced

Table 2Taxes on British Wine Imports by Source 1660ndash1862 (British Pounds per Kiloliter)

France Germany Spain Portugal South Africa

1660ndash1665 7 9 8 81666ndash1684 7 9 8 81685ndash1691 14 20 19 181692ndash1695 22 20 19 181696 47 20 19 181697ndash1702 51 25 23 221703 52 27 24 231704ndash1744 55 31 26 251745ndash1762 63 35 30 291763ndash1777 71 39 34 331778 79 43 38 371779 84 41 40 391780ndash1781 92 49 44 431782ndash1785 96 51 47 46 441786 65 51 37 37 371787ndash1794 47 51 32 32 371795 78 64 51 51 571796ndash1797 108 92 71 71 771798 111 96 73 73 791799ndash1801 107 92 71 71 771802 112 97 74 74 801803 131 109 87 87 871804 142 117 95 95 951805ndash1824 144 119 96 96 961825ndash1830 78 50 50 50 251831ndash1859 58 58 58 58 291860 32 32 32 32 321861 16 21 21 21 211862 11 26 26 26 26

Source Summarized from Ludington (2013 Table A1)

228 UK and Global Wine Markets by 2025 and Implications of Brexit

Figure 4

UK Shares in Wine Exports of Key Wine-Exporting Countries 2010ndash2014 ()

Fig4-BW

onlin

eBW

inprint

Source Compiled from data in Anderson and Pinilla (2017)

Table 3Import Duties Excise Duties and VAT Affecting Consumer Prices of Wine and Other Alcohol

in the United Kingdom April 1 2017

VAT 20

MFN import duties on wine pound per liter-bottled still wine lt13 alc 0114-bottled still wine 13ndash15 alc 0134-bottled still wine 15ndash18 alc 0162-bulk still wine lt13 alc 0086-bulk still wine 13ndash15 alc 0105-bulk still wine 15ndash18 alc 0134-sparkling wine 0278Weighted averagea 0130

Excise duties on alcohol pound per liter-still wine lt15 alc 2887-still wine 15ndash22 alc 3848-sparkling wine 55ndash85 alc 2795-sparkling wine 85ndash15 alc 3697-spirits (assumed 40 alc) 11551-beer (assumed 5 alc) 0954

VAT on alcohol 20

a Assumes that one-third of UK wine imports arrives in bulk that half has less than 13 alcohol and that one-tenth of imports is sparkling

Sources HM Revenue httpswwwgovukgovernmentpublicationsalcohol-duty-rate-changes for excise duties and VAT accessed April 92017 and Wine Australia (2015) for import duties converted at the 2016 average exchange rate of pound0740634 per euro

Kym Anderson and Glyn Wittwer 229

as to restore investor and consumer confidence in the United Kingdomrsquos economyThe size of those impacts is uncertain of course (see Baldwin 2016 and especiallyCampos 2016) so we consider a range of possibilities below

Those assumed adverse macroeconomic effects will add to the initial impact ofaltered wine tariffs on aggregate wine consumption in the United Kingdom andhence on its bilateral trades in wine They will make the loss of sales to theUnited Kingdom by EU (and Chilean and South African) suppliers greater thanwould otherwise be the case And they will reduce the likelihood that other countriesrsquosales of wine in the United Kingdom will be higher than in the baseline Indeed themacroeconomic effects could outweigh the trade-diverting effects so that even coun-tries that are currently discriminated against by the EU28rsquos wine-trade policy may beworse off because of Brexit

IV Global Wine Markets Model and Database

Our model of the worldrsquos wine markets first published by Wittwer Berger andAnderson (2003) and revised by Anderson and Wittwer (2013) is summarized inthe Appendix It disaggregates wine markets into four types namely nonpremiumcommercial-premium and superpremium still wines and sparkling wines2 There aretwo types of grapes premium and nonpremium Nonpremium wine uses nonpre-mium grapes exclusively superpremium wines use premium grapes exclusivelyand commercial-premium and sparkling wines use both types of grapes to varyingextents across countries The world is divided into 44 individual nations and 7 com-posite geographic regions that capture all other countries

The modelrsquos database is calibrated to 2014 based on the comprehensive wine-market volume and value data and trade and excise-tax data provided inAnderson and Pinilla (2017) and in Anderson Nelgen and Pinilla (2017) It is pro-jected assuming that aggregate national consumption population and realexchange rates change between 2014 and 2025 to the extent shown in AppendixTable 13 The Brexit alternatives to that baseline also are projected to 2025

Concerning preferences we assume a continued considerable swing toward allwine types in China and a swing away from nonpremium wines in all other countriesuntil 2025

In our baseline scenario grape- and wine-industry total-factor productivity areassumed to grow at 1 per year everywhere while grape- and wine-industry

2Commercial-premium still wines are defined byAnderson Nelgen and Pinilla (2017) to cost between US$250 and US$750 per liter pretax at a countryrsquos border or wholesale3The real exchange rate changes over the projection period are the changes expected in the nominal valueof country irsquos currency relative to the US dollar times the expected ratio of the GDP deflator for theUnited States versus that for country i

230 UK and Global Wine Markets by 2025 and Implications of Brexit

capital is assumed to grow net of depreciation at 15 per year in China but zero else-where (consistent with the almost-zero growth in global wine production and con-sumption over the past two decades)

Two alternative scenarios are considered (ldquolargerdquo and ldquosmallrdquo) for the initialimpact of Brexit to provide a range of results We assume that because of theUnited Kingdomrsquos decision to exit the European Union the rate of UK realgross domestic product (GDP) growth is only one-third or two-thirds as fast overthe projection period (09 or 18 per year instead of 26) and the Britishpound will be 20 or 10 lower in real terms than in our modelrsquos core baselineprojection4 In both alternative scenarios we assume that the United Kingdomapplies the European Unionrsquos external tariffs on wine from March 30 2019 atthe end of the two-year period following the United Kingdomrsquos formal triggeringof Article 50 (see Table 3)

We then consider a subsequent impact of Brexit presumed to result from negoti-ating signing implementing and responding to an FTA with the EU27 by 2025(Chile and South Africa currently have preferential access to EU wine marketsbut we continue to assume in this scenario that the United Kingdom does not imple-ment new bilateral FTAs with them or others in our time frame)5 This subsequentscenario assumes the British pound returns to what it would have been in 2025 in theabsence of Brexit and that real incomes regain two-thirds of the value they wouldhave had without Brexit as compared with our ldquosmallrdquo initial Brexit scenario

This global model has supply-and-demand equations and hence quantitiesprices and price elasticities for each of the grape and wine products and for asingle composite of all other products in each country Income elasticities ofdemand also exist for each final product Grapes are assumed to not be traded inter-nationally but other products are exported and imported Each market is assumed tohave been in equilibrium before any shock and to find a new market-clearingoutcome following any exogenously introduced shock

4The nominal price of the British pound in US dollars in the fortnight following the Brexit vote on June23 2016 dropped 13 to US$130 and a year later the pound sat at the same rate having dropped to anaverage of $124 (an 18 devaluation) between October 2016 and April 2017 Our choice of a low of 10and a high of 20 aims to capture future possible rates while uncertainties remain The average real wagein the United Kingdom fell in the first half of 2017 and projected real GDP growth during 2018ndash2020 hasbeen revised down to 17 in the United Kingdomrsquos latest budget (HM Treasury 2017)5South Africa currently has duty-free access to the EU28 for just 50 megaliters (ML) of wine per yearbeyond which the MFN tariff rate applies South Africa currently exports around 320 ML to theEuropean Union one-third of which initially goes to the United Kingdom How that quota of 50 MLis divided between the United Kingdom and the EU27 is subject to future negotiation (Rollo et al2016 Swinbank 2017) In the following discussion we assume none of it is accepted by the UnitedKingdom so that all South African wine imports pay the United Kingdomrsquos MFN tariff under Brexitbut altering that assumption makes very little difference to our results for any countryrsquos trade exceptSouth Africarsquos

Kym Anderson and Glyn Wittwer 231

V Projecting Global Wine Markets to 2025

Global wine production and exports are projected in the baseline from 2014 to 2025consistent with past trends The modelrsquos global volume of production (and consump-tion) rises little over that 11-year period (9) made up of a 6 decline in nonpre-mium wine and a one-sixth rise in commercial-premium and superpremium wineIn real (2014 US$) value though global wine output and consumption increaseby about 50 in total and 60 in the two premium categories The internationaltrade projections are similar although a little larger with the share of global wineproduction exported (equivalent to the share of global consumption imported)rising 2 percentage points between 2014 and 2025

The baseline projection does not greatly alter the 2014 shares of various countriesin global wine production apart from China because we assume vineyard expansionthere is faster than elsewhere6 In value terms this assumption means that Chinamoves from fifth to fourth by 2025 behind France the United States and ItalySpain remains barely ahead of Australia and they and Germany take the nextthree places (Figure 5(a)) In total wine-production volume terms China movesfrom sixth to fifth place and Argentina drops from fifth to eighth (and fromeighth to ninth in value terms)

When their products are subdivided into fine wines (superpremium still plus spark-ing) commercial-premium wines and nonpremium wines France and the UnitedStates retain the highest two places on the global ladder for fine-wine productionand Spain and Italy retain the top two places for nonpremium wine As for commer-cial-premium wine production (defined to cost between US$250 and US$750 perliter pretax at a countryrsquos wholesale level or national border) Italy retains the topranking over our projections period but at least in terms of value China challengesFrance for second place

The country rankings by projected value of total wine consumption change some-what more than those for production by 2025 with China taking second place afterthe United States ahead of France and Germany and then the United Kingdomslightly overtaking Italy to slip into fifth place (Figure 5(b)) The United StatesFrance and Germany retain the top three rankings for consuming fine wine butCanada slightly overtakes Italy for fourth place in terms of value at least In thecase of commercial-wine consumption China strengthens its number-one positionahead of the United States and the United Kingdom does likewise vis-agrave-visGermany for third place

As for the projected changes in consumption volumes China is projected to dom-inate the increase in aggregate although the United States is projected to lead the

6 In fact Chinarsquos wine production fell steadily between 2012 and 2016 by a total of one-sixth so Chinarsquoswine imports may grow faster in practice than in this baseline projection

232 UK and Global Wine Markets by 2025 and Implications of Brexit

Figure 5

Value of Wine Production and Consumption in Key Countries 2014 and Projected Baseline for2025 (2014 US$ Million at WineryWholesale Pretax Prices)

Fig5-BW

onlin

eBW

inprint

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 233

increase in consumption of fine wine In Western Europe and in the SouthernHemispherersquos New World countries fine wines are projected to substitute for com-mercial wines (defined as the sum of commercial-premium and nonpremiumwines) with almost no change in total wine consumption Sub-Saharan Africa isthe next region that is projected to take off with its growth accounting for morethan one-third of the rest of the worldrsquos increase in volume consumed

Those differences in production versus consumption rankings are reflected ininternational trade Figure 6 shows that France Italy and Spain remain the threedominant exporters of wine in aggregate value but the rankings of the next fewchange with Australia being slightly ahead of Chile and the United StatesGermany and New Zealand being nearly tied for sixth place in value termsFrance and Italy are even more dominant in fine-wine exports and remain so by2025 while Italy outranks France in the commercial-premium export categoryand Spain outranks Italy Australia and Chile in the nonpremium export class

Among the importers the United States and the United Kingdom are projected tocontinue to hold the first two places in 2025 in value terms but China moves intothird place slightly ahead of Germany followed well behind by Canada HongKong Belgium-Luxembourg the Netherlands and Japan (Figure 6(b)) OtherAfrica (excluding South Africa) is projected to experience the largest increase inimports among all the other regions followed by Other Asia which becomes asbig as Germany in value terms (Figure 7(a)) In terms of total volume of wineimports Germany and the United Kingdom held the top two shares in 2014 butby 2025 the United Kingdom is projected to be well ahead of Germany (Figure 7(b)) However this projection ignores the effects of Brexit to which we now turn

VI How Might Wine Markets Be Affected by the United Kingdomrsquos Exitfrom the European Union

As mentioned earlier for our two alternative scenarios to capture the initial effects ofBrexit (ldquolargerdquo and ldquosmallrdquo) we assume that following the United Kingdomrsquos exitfrom the European Union the United Kingdomrsquos rate of economic growth wouldbe only one-third or two-thirds as fast for the period to 2025 the British poundwould be 20 or 10 lower in real terms than in our modelrsquos baseline projectionand the United Kingdom would apply the European Unionrsquos external tariff onwine to imports from EU member countries (as part of establishing MFN ratesvia the WTO to then start new bilateral FTA negotiations) In these initial scenarioswe assume the United Kingdom does not implement any new FTAs particularlywith the EU27 Chile and South Africa Following the discussion of those initialresults we present the results of a subsequent scenario that includes the implemen-tation of an FTA between the United Kingdom and the EU27 We assume these aretwo discrete steps with no agreed arrangements to smooth the transition betweenthem

234 UK and Global Wine Markets by 2025 and Implications of Brexit

Figure 6

Value of Wine Exports and Imports Key Wine-Trading Countries 2014 and Projected Baselinefor 2025 (2014 US$ million)

Fig6-BW

onlin

eBW

inprint

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 235

Figure 7

National Shares of Global Wine Import Value and Volume 2014 and Projected Baselinefor 2025 ()

Fig7-BW

onlin

eBW

inprint

Source Authorsrsquo model results

236 UK and Global Wine Markets by 2025 and Implications of Brexit

A Initial Impact of Brexit

Because the British pound dropped by one-sixth against the US dollar in the fourmonths following the Brexit vote in June 2016 and because the UnitedKingdomrsquos average real wage fell 1 between November 2016 and April 2017(according to the latest available data see Office of National Statistics 2017) weviewed our ldquolargerdquo scenario as more likely soon after the Brexit vote But wecompare those results with ones from our ldquosmallrdquo initial-impact scenario becauseit has become more likely following the June 2017 general election that ldquosofterrdquooptions will be considered by the new government Generally the results areabout half the size in the ldquosmallrdquo scenario with the exception of the bilateraltrade effects To show the sensitivity of results to our assumptions we point outthe differences when the ldquosmallrdquo results are not close to half the results shown forthe ldquolargerdquo scenario

Before turning to those results we first present a ldquofree-traderdquo scenario whichassumes the United Kingdom chooses the radical option of becoming the HongKong of Europe Unlikely as this scenario is the results provide assurance thatour global wine model generates the standard types of effects of exiting a customsunion In this ldquofree-traderdquo scenario the United Kingdom is assumed to move tozero tariffs on all wine imports Thus the current preferences on EU Chileanand South African wine imports disappear as all other wine exporters enjoy thesame free access to UK wine markets We assume in this scenario that Brexit hasno impact on the United Kingdomrsquos real income or consumption growth rates by2025 and that by then the British pound has returned to its 2014 value relative toother currencies (similar to the real exchange rate assumed in our baseline scenarioto 2025 see Appendix Table A1)7 Under these assumptions EU Chilean andSouth African wine exports to the United Kingdom in 2025 are 003 lower thanin the no-Brexit baseline and UK imports from other wine exporters are 025higher but the United Kingdomrsquos total wine imports are hardly any different(because local wine prices fall by only 05 and incomes are unchanged) In this sce-nario UK consumers expand their volume of consumption by just 05 while theEU27 exports more and other countries export less to the rest of the world such thatworld wine exports are almost unchanged In short all these effects are in the direc-tion that comparative static-customs-union theory would predict but they are smallbecause we assume the United Kingdomrsquos real income and currency are the same asin our baseline for 2025 as described in the previous section

In the ldquolargerdquo alternate initial scenario involving Brexit as compared with thebaseline scenario to 2025 the consumer price of wine in 2025 is 22 higher in the

7If this scenario were seen as a credible long-run trade policy it would encourage more investment thatwould eventually raise UK incomes but not without major structural changes that would initiallydisrupt the economy Implicitly we assume in our free-trade scenario that these two opposite impactson real UK incomes and the British pound exactly offset each other by 2025

Kym Anderson and Glyn Wittwer 237

United Kingdom in local currency terms (20 because of real depreciation of theBritish pound 4 because of new tariffs on EU Chilean and South Africanwines and ndash2 because of slower UK income growth) The volume of UKwine consumption is 28 lower 16 because of slower UK economic growth7 because of real depreciation of the British pound and 5 because of newtariffs Superpremium still-wine sales are the most affected dropping by two-fifths while sparkling and commercial-premium wines drop a bit less than one-quarter Because the average price rises by more than the decline in volume soldthe aggregate value of UK sales even in local currency terms falls under thisldquolargerdquo Brexit scenario Under the ldquosmallrdquo Brexit scenario the consumer price ofwine in 2025 is 11 higher in the United Kingdom and its volume of wine consump-tion is 17 lower

The volume of projected UK imports in 2025 is 427 ML or nearly one-quarterless in the ldquolargerdquo scenario than in the baseline scenario comprising 58ML less spar-kling wine 31 ML less superpremium still wine and 339 ML less commercial-premium wine World imports are lower by just 239 ML because imports byother countries are 189 ML higher in response to the lower international wineprices in this scenario In value terms UK imports are $175 billion (or 27)lower in 2025 because of ldquolargerdquo Brexit $113 billion because of lower incomes$038 billion because of the fall in the British pound and $014 billion because ofthe rise in wine-import tariffs (Table 4) These aggregate trade impacts are a littlemore than half this size under the ldquosmallrdquo Brexit scenario

Despite the import levels falling because of raised import tariffs domestic con-sumption of all three quality categories of UK-produced wine is lower with thanwithout Brexit because of shrunken demand for all wines resulting from loweredUK incomes and raised local prices because of devaluation of the British poundThe British poundrsquos devaluation does make it easier for the United Kingdom tosell wines abroad though The countryrsquos exports are 7 ML or nearly 5 higherin 2025 in the ldquolargerdquo Brexit scenario and UK production is 3 higher ThoseUK exports (or re-exports of imported bulk wine after it is bottled in the UnitedKingdom) that go to EU27 countries are reduced though because of the tariffnow imposed at the new EU border

Without Brexit the United Kingdomrsquos shares of global wine imports are slightlyhigher in volume terms in 2025 than in 2010ndash2015 but 2 percentage points lower invalue terms thanks to East Asiarsquos expanding demand for imports of premium winesWith ldquolargerdquo Brexit however that value share is an additional 2 percentage pointslower and the volume share is almost 5 points lower (Figure 8) The net effects ofthese impacts on global trade are shown in Figure 9 Most of the initial tradeeffect of Brexit is a large decline in net imports of wine by the United Kingdomwith very little offsetting positive effect on trade in the rest of the world Theldquosmallrdquo Brexit numbers are a bit more than half of those for ldquolargerdquo Brexit

238 UK and Global Wine Markets by 2025 and Implications of Brexit

Table4

Difference

in2025

Projected

VolumeandValue

ofWineIm

portsb

ytheUnitedKingdom

andtheResto

fthe

World

asaConsequence

oftheInitial

BrexitS

hock

(MLand2014US$

Million

ldquoLargerdquoSc

enario)

Volum

e(M

L)

Value

(US$

million)

NP+CPa

SuperPrb

Sparkling

TOTAL

NP+CPa

SuperPrb

Sparkling

TOTAL

ΔUK

impo

rtsdu

eto

Low

erincomes

minus198

minus20

minus29

minus247

58minus644

minus253

minus234

minus1131

65Low

erpo

und

minus70

minus10

minus14

minus93

22minus248

minus127

minus102

minus476

27Highertariffs

minus71

minus1minus16

minus87

20minus110

minus8minus2

4minus143

8TOTA

Lminus339

minus31

minus58

minus427

100

minus1001

minus388

minus360

minus1750

100

difffrom

base

2332

3325

2432

3227

of

totalc

uts

797

14100

5722

21100

ΔROW

netim

ports

143

2125

189

230

143

minus181

192

ΔWORLD

TRADE

minus195

minus10

minus34

minus239

minus763

minus246

minus543

minus1552

aNon

prem

ium

plus

commercial-premium

still

wines

bSu

perpremium

still

wines

SourceA

utho

rsrsquomod

elresults

Kym Anderson and Glyn Wittwer 239

The aggregate effect of ldquolargerdquo Brexit on the market shares of various wine-export-ing countries in the United Kingdom is almost indiscernible even with one decimalpoint The projected 2025 shares are quite different from the actual 2014 shares forseveral countries They are much smaller in 2025 for South Africa Australia andNew Zealand (and the United States in volume terms) and are much larger involume for Spain and in value for Italy These results occur because wine-exportingcountries benefit differentially from the varying rates of growth in net importdemand for wine in non-UK countries over this projection period The most impor-tant projected changes are the increase in the real value of annual wine importsbetween 2014 and 2025 by China (200 or $3 billion) Other Asia (110 or $22billion) and Africa (270 or $16 billion) More than half of Australiarsquos increasein annual exports from 2014 to 2025 go to Asia and more than half of SouthAfricarsquos increase in exports go to Other Africa

Table 5 reveals that ldquolargerdquo Brexit lowers EU Chilean and South African wineexports by 150 ML or US$12 billion in the case of the European Union with someof their exports diverted from the United Kingdom to the EU27 and other marketsin competition with New World exporters Although the United States Australiaand Argentina sell only a little less to the United Kingdom they also sell less toother countries For Chile and South Africa which lose their preferential access toUK (but not to EU27) markets in this Brexit scenario some exports are redirected

Figure 8

UK Shares of World Wine Imports 2010ndash2015 and Projected to 2025 without and with theldquoLargerdquo Initial Brexit Shock ()

Fig8-BW

onlin

eBW

inprint

Sources Anderson and Pinilla (2017) and authorsrsquo model results

240 UK and Global Wine Markets by 2025 and Implications of Brexit

from the United Kingdom to EU27 countries ndash but again they export less overallGlobal wine trade in 2025 is 240 ML (19) or $18 billion (35) less under thisldquolargerdquo Brexit scenario The percentage by which wine exportersrsquo trade shrinks isgreater for values than for volumes because of changes in relative prices of differ-ent-quality wines Those differences are shown in the numbers in parentheses inTable 5

Three other points are worth making about Table 5 One is that Australia sellsslightly more to the United Kingdom in the ldquosmallrdquo Brexit scenario rather thanslightly less as in the ldquolargerdquo Brexit case Evidently the negative income and price(devaluation) effects do not more than offset the positive trade-diverting effect onAustralian exports to the United Kingdom of removing preferences in the ldquosmallrdquoscenario Second New Zealand sells slightly more to non-UK countries underBrexit despite greater competition from the EU27 Chile and South Africa Thisanomaly is due to changes in the relative prices of different qualities of wine inglobal wine markets bearing in mind that New Zealand has the worldrsquos highestaverage price for still-wine exports And third the value (but not the volume) ofexports of ldquoOtherrdquo countries to markets other than the United Kingdom arehigher under Brexit This result too is due to changes in the relative prices of differ-ent qualities of wine in global wine markets

Figure 9

Difference in 2025 Wine Import Volumes and Values as a Result of the ldquoLargerdquo Initial BrexitShock (ML and US$ Million in 2014 US dollars)

Fig9-BW

onlin

eBW

inprint

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 241

B Subsequent Impact of Brexit from a UK-EU27 FTA

The next-most-likely step in the Brexit process is for the United Kingdom to nego-tiate a new trade arrangement with the EU27 We therefore assume that a UK-EU27FTAwith free bilateral wine trade is implemented and adjusted to by 2025 and thatprogress toward that end occurs soon enough that the adverse macroeconomicshocks from the initial impact of uncertainty over the Brexit process are confinedto those assumed in the ldquosmallrdquo scenario outlined above In this subsequent scenariowe assume that the British pound returns to the value it would have reached in 2025in the absence of Brexit (ie reversing the 10 devaluation assumed in the ldquosmallrdquoinitial Brexit scenario) and that real incomes in the United Kingdom are 8 higher

Table 5Difference in 2025 BilateralWine Import Volumes and Values fromKey Exporters by the United

Kingdom and the Rest of the World (RoW) as a Result of Initial Brexit Shock(ML and 2014US$ Million)a

(a) ldquolargerdquo scenario

Volume (ML) Value (2014US$ million)

UK RoW WORLD () UK RoW WORLD ()

EU27 minus287 136 minus150 (ndash17) minus1187 minus5 minus1192 (ndash31)Chile minus59 35 minus25 (ndash30) minus169 31 minus138 (ndash48)South Africa minus53 35 minus18 (ndash32) minus105 20 minus85 (ndash67)USA minus7 minus6 minus13 (ndash24) minus75 minus40 minus115 (ndash50)Australia minus4 minus3 minus7 (ndash09) minus25 minus65 minus90 (ndash30)Argentina minus3 minus9 minus12 (ndash48) minus16 minus39 minus55 (ndash52)New Zealand minus11 9 minus2 (ndash09) minus162 71 minus91 (ndash43)Others minus2 minus10 minus12 (ndash02) minus11 minus52 minus63 (ndash44)WORLD minus427 187 minus240 (ndash19) minus1750 minus79 minus1829 (ndash35)

(a) ldquosmallrdquo scenario

Volume (ML) Value (2014US$ million)

UK RoW WORLD () UK RoW WORLD ()

EU27 minus178 82 minus96 (ndash12) minus692 minus43 minus736 (ndash19)Chile minus46 28 minus18 (ndash24) minus128 36 minus91 (ndash32)South Africa minus43 29 minus14 (ndash42) minus82 23 minus59 (ndash47)USA 1 minus6 minus5 (ndash11) minus23 minus28 minus51 (ndash22)Australia 5 minus10 minus5 (ndash06) 19 minus56 minus38 (ndash13)Argentina 0 minus6 minus6 (ndash26) minus3 minus25 minus29 (ndash27)New Zealand minus5 4 minus1 (ndash06) minus80 34 minus46 (ndash22)Others 0 minus9 minus9 (ndash01) minus1 minus33 minus34 (ndash24)WORLD minus266 112 minus154 (ndash13) minus991 minus92 minus1083 (ndash21)

a Numbers in parentheses are the percentage difference between the Brexit and baseline scenarios for 2025 projected wine-import volumes orvalues by source

Source Authorsrsquo model results

242 UK and Global Wine Markets by 2025 and Implications of Brexit

in 2025 than in the ldquosmallrdquo scenario ndash that is they regain two-thirds of the value theywould have reached without Brexit-related uncertainty

This subsequent development in the Brexit process reverses most of but not allthe initial effects of Brexit by 2025 because of our assumption that the lost growthin the initial years of uncertainty following the Brexit vote are only partly recoveredby 2025 following the implementation of a UK-EU27 FTA Moreover the longer ittakes before this FTA is finalized and implemented the longer the estimated initialadverse effects persist and the larger the cumulative cost of Brexit to UK wine con-sumers and to grape and wine producers in wine-exporting countries

Table 6 summarizes the subsequent trade effects for 2025 It suggests that all butone-tenth of the loss in value of world trade in wine from the initial ldquosmallrdquo impact isrestored most of it because of a smaller reduction in the United Kingdomrsquos wineimports following the FTA Most of that improved outcome is because of recoveredimports from the EU27 commensurate with the latterrsquos high share of UK importsEven though Chile and South Africa are assumed in this scenario to have not yetsigned an FTA with the United Kingdom they export slightly more to the UnitedKingdom (and even more to all other countries) than in the initial ldquosmallrdquo scenario

Within the United Kingdom this FTA brings down the local currency consumerprice of wine by 9 largely offsetting the 11 rise in the initial ldquosmallrdquo Brexit sce-nario and it raises the volume consumed in the United Kingdom by 18 fully off-setting the 17 fall in the initial ldquosmallrdquo Brexit scenario

Table 6Difference in 2025 Bilateral Wine-Import Volumes and Values from Key Exporters by the

United Kingdom and the Rest of the World (RoW) as a Result of Implementing a UK-EU27FTA Following Initial ldquoSmallrdquo Brexit Shock (ML and 2014US$ Million Difference Relative to

Initial ldquoSmallrdquo Brexit Scenario)a

Volume (ML) Value (2014US$ million)

UK RoW WORLD (ldquosmallrdquo)a UK RoW WORLD (ldquosmallrdquo)a

EU27 212 minus112 100 (minus96) 750 0 750 (minus806)Chile 1 3 4 (minus18) 3 40 43 (minus92)South Africa 3 2 5 (minus14) 3 18 21 (minus60)USA 3 3 6 (minus5) 26 21 47 (minus50)Australia 1 1 2 (minus5) 8 38 46 (minus33)Argentina 1 4 5 (minus6) 4 20 24 (minus28)New Zealand 4 minus3 1 (minus1) 61 minus25 36 (minus46)Others 1 6 7 (minus9) 5 minus46 minus41 (92)WORLD 226 minus96 130 (minus154) 860 66 926 (minus1022)

a Numbers in parentheses are the world trade differences between the ldquosmallrdquo initial Brexit scenario and the baseline scenario copied fromcolumns 3 and 7 of Table 5(b)

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 243

In short this subsequent step in the Brexit process can be expected to restore muchof the initial adverse effects in the United Kingdom of the Brexit vote by 2025 Thewine exporters in the EU27 Chile and South Africa also lose less in this scenarioas compared with the initial ldquosmallrdquo scenario while other key exporters sell nearlyas much in 2025 in this scenario as they are projected to in the baseline ndash that isthey are beneficiaries of the reduced discrimination in global wine markets in thiscase where Chile and South Africa no longer enjoy preferential access to the UKmarket (and even though EU27 producers are again allowed to do so) Keep inmind though that although the estimated losses in the initial scenarios are reportedjust for 2025 they are expected to be felt throughout the preceding years of uncertaintyTherefore the longer it is before uncertainty abates and the United Kingdomrsquos currentpreferential trading arrangements are replaced by newones such as a UK-EU27 FTAthe higher the cumulative cost of the Brexit vote to the wine trade

VII Caveats and Conclusions

The above Brexit simulations are just a few of many scenarios that could be modeledObvious additional ones could also assume that FTAs are reached between theUnited Kingdom and other trading partners including Chile and South Africaand such countries as Australia and New Zealand Some of these other countriesare already exploring the ECrsquos invitation (see European Commission 2015) to nego-tiate an FTA with what will be the EU27 The sequence in which FTAs are signedand the speed with which they are implemented will matter (as was also the casewith the sequential signing over the past decade of bilateral FTAs with NortheastAsian countries by Chile Australia and New Zealand see Anderson andWittwer 2015) Even if the United Kingdom were able to sign additional bilateralFTAs and begin implementing them before 2025 it would have little effect on theabove results (because wine tariffs are such a minor contributor to them) unlessthose FTAs were to accelerate the United Kingdomrsquos economic growth and therise in the British poundrsquos value before 2025

We assume above that no changes are made to alcohol excise duties in the UnitedKingdom following Brexit even though they are scheduled to be progressively raisedwith inflation and may be raised even more for wine relative to spirits to offset theopposite effects of Brexit on those two domestic industries Nor do we make anyallowance for UK increases in consumer wine prices that may be needed to coverthe higher cost of clearing customs on imports from EU27 countries (which maybe more or less offset by increased smuggling and duty-free purchasing)Importers of fine wines including individuals who buy Bordeaux en primeur maywell reduce their demand because of concerns that delays on docks will affect thequality of their wines or because of greater currency uncertainty

To summarize Brexit is costly initially to UK consumers of wine (and of manyother tradable products) because the domestic retail price in local currency tax-

244 UK and Global Wine Markets by 2025 and Implications of Brexit

inclusive terms is 22 higher than otherwise in 2025 and the volume of wine con-sumed domestically is 28 lower in the ldquolargerdquo Brexit scenario (or 11 higher pricesand 17 lower quantities in the ldquosmallrdquo case) by 2025 unless a UK-EU27 FTA isdeveloped before then Even if such an FTA does get signed ratified by all 28 par-liaments and implemented by 2025 the slower income growth in the interim meansa smaller UK wine market in 2025 than would otherwise have been the case Thevolume reduction is a blow to many participants in UK wine-bottling transportingstoring wholesaling and retailing businesses in addition to restaurants and pubsVery little of that initial impact occurs because of higher import tariffs most impor-tant is the assumed fall in UK real incomes

The small but growing number of local UK vignerons (who supply less than 05of all domestic-wine sales and less than 4 of sparkling-wine sales) is initially pro-jected to sell less wine domestically because of reduced demand and to export only alittle extra abroad while the British pound is devalued Overall their production isonly 3 higher in 2025 even in the ldquolargerdquo Brexit scenario (not taking intoaccount the fact that their casual labor is likely to be more expensive under Brexitbecause of tighter restrictions on immigration) and their production diminishes ifand when a UK-EU27 FTA is in place

Even with a UK-EU27 FTA in place EU27 wine exporters export US$56 millionless wine in 2025 thanks to Brexit Chile and South Africa export $88 million lesswine and wine exports from the rest of the world are little different in aggregatealthough slightly larger from Australia

Clearly there will be great uncertainly for some time over the possible policy out-comes to flow from Brexit and their consequent sequential impacts on UK house-hold disposable incomes foreign exchange rates and bilateral wine tariffsMeanwhile the above projections under explicit assumptions provide some idea ofhow wine markets might be affected by the most-likely first two stages of theBrexit process (agreeing on a new tariff schedule at the WTO and agreeing to andimplementing a UK-EU27 FTA) In particular they make clear that nontrivialinitial impacts could affect the domestic wine market impacts that are likely to belarger than just the direct impact of changes in bilateral tariffs If the UnitedKingdom succeeds in getting countries to agree to transition arrangements thatdelay the changes in tariffs until new FTAs are signed and ratified the initialeffects will be less dramatic than in our first scenarios but it remains to be seenwhether any such agreements can be reached In any event the net effect of Brexiton the welfare of the worldrsquos consumers and producers of wine as a whole will benegative not just initially but permanently unless new trade policy commitmentsby the United Kingdom with major wine-exporting countries are sufficiently moreliberal than current arrangements

Kym Anderson and Glyn Wittwer 245

Appendix Key Equations in the Global Wine Markets Model

Amodel of the worldrsquos wine markets was first published byWittwer et al (2003) andmuch improved by Anderson and Wittwer (2013) Several significant enhancementshave been made to that original model Wine types have been disaggregated intomore types ndash namely nonpremium (including generic bulk) commercial-premiumand superpremium still wines and sparkling wines8 As in the original modelthere are two types of grapes premium and nonpremium Nonpremium wines usenonpremium grapes exclusively superpremium wines use mainly premium grapesand commercial-premium and sparkling wines use both types of grapes In theregional dimension the number of countries and country groups has expandedfrom 10 in the original model to 51 44 individual nations and 7 compositeregions The modelrsquos database is calibrated to 2014 for this paper based on datain Anderson et al (2017) trade data for which are downloaded from httpscomtradeunorg The modelrsquos supply-and-demand equations are based on the orig-inal ORANI modelrsquos theory (see Dixon Parmenter Sutton and Vincent 1982) andthe model is implemented using GEMPACK software (Harrison Horridge Jerieand Pearson 2014)

An enhancement of importance to the present study is the inclusion of exchange-rate variables in the model which allow a distinction between price impacts asobserved in local currency units and those observed in 2014 US dollars

In the model the grape and wine sectors minimize costs of intermediate inputssubject to weak constant elasticity of substitution (CES) substitutability betweeninputs By assumption no intermediate inputs are imported from other countries9Hence

Xcid frac14 f ethX1id CESfrac12P

cid=P1id $THORN eth1THORN

P1id X1id frac14X

c

Xcid P

cid eth2THORN

whereXcid is the quantity demanded of commodity c by grape or wine industry i in

region d Pcid is the corresponding price and X1id and P1id are the respective inter-

mediate composite quantities and prices

Two primary factors are employed in the sector labor (the quantity of which isendogenous with perfectly elastic supply) and capital Capital is usually treated asexogenous in quantity with rates of return bearing all the adjustment in thevarious scenarios This reflects the fact that grapes (a perennial crop) and wine-

8Commercial-premium still wines are defined by Anderson and Nelgen (2011) to be those priced betweenUS$250 and $750 per litre pre-tax at a countryrsquos border or wholesale9An exception concerns bulk wine imports used mainly in the bottling of commercial premium wine in theUK USA Canada Japan and Germany

246 UK and Global Wine Markets by 2025 and Implications of Brexit

plant capacity adjust slowly to market signals

Lid frac14 f ethFid CESfrac12Wid=PFid $THORN eth3THORN

Kid frac14 f ethFid CESethRid=PFid $THORN eth4THORN

PFid Fid frac14 Lid Wid thorn Kid Rid eth5THORN

Grape and wine producers are assumed to minimize costs subject to CES substitu-tion between capital and labor Equations (3) to (5) show primary factor demandsfor the labor composite Lid and capital Kid subject to a composite factor demandFid by industry i in region d The factor prices are Wid for labor Rid for capitalrentals and PFid for composite factor prices

Appendix Table A1Cumulative Consumption and Population Growth Rates and Changes in the Real Exchange Rate

(RER) Relative to the US Dollar 2014 to 2025 without Brexit ()

Aggregateconsumption Poprsquon RER

Aggregateconsumption Poprsquon RER

France 18 4 minus11 Australia 35 11 minus17Italy 11 2 minus9 New Zealand 32 9 minus26Portugal 14 0 minus9 Canada 27 8 minus18Spain 26 8 minus9 United States 31 8 0Austria 19 4 minus7 Argentina 7 10 109Belgium 20 7 minus9 Brazil 16 8 minus29Denmark 22 2 minus9 Chile 55 8 minus2Finland 21 3 minus7 Mexico 42 12 minus8Germany 14 minus2 minus11 Uruguay 45 3 1Greece 22 minus1 minus14 Other Latin America 60 10 minus5Ireland 42 12 minus9 South Africa 36 12 minus1Netherlands 21 4 minus9 Turkey 50 8 20Sweden 24 9 minus13 North Africa 53 11 0Switzerland 18 8 minus6 Other Africa 109 18 84United Kingdom 32 6 1 Middle East 52 18 minus12Other West Europe 21 10 minus1 China 79 3 5Bulgaria 41 minus7 7 Hong Kong 42 3 2Croatia 20 minus2 minus1 India 134 13 17Georgia 35 0 23 Japan 11 minus3 minus24Hungary 25 minus3 minus11 Korea 38 1 minus9Moldova 49 minus11 13 Malaysia 62 15 minus16Romania 45 minus4 22 Philippines 75 18 7Russia 18 minus2 minus8 Singapore 44 21 minus22Ukraine 22 minus5 14 Taiwan 29 1 minus13Other East Europe 40 minus5 48 Thailand 47 3 minus9

Other Asia 99 10 10

Source Authorsrsquo compilation from projections by various international agencies and from global economy-wide modeling by Anderson andStrutt (2016)

Kym Anderson and Glyn Wittwer 247

The composite factor demand Fid is proportional to total output Qid subject to aprimary factor using technology Aid Hence

Fid frac14 Qid Aid eth6THORN

The perfectly competitive zero-pure-profit condition is that total revenue valued atthe output price P0s

i multiplied by Qid equals the total production cost

P0si Qid frac14

X

cPcid X1cid thorn

X

oWo

id Loid thorn Rid Kid eth7THORN

Household demands follow a linear expenditure system in each region We reducethe optimizing problem for household consumption of each commodity subjectto a budget constraint to equations describing subsistence and discretionarydemands Aggregate subsistence expenditure WSUBd depends only on consumerprices P3cd for each commodity and the number of households N as per-capita sub-sistence quantities XSUBcd subject to given preferences are constant

WSUBd frac14X

c

P3cd XSUBcd Nd eth8THORN

Discretionary expenditures for each commodity (the left-hand side of equation (9))are equal to the marginal budget share (βcd) of aggregate discretionary expenditureThis aggregate is the bracketed term on the right-hand side of equation (9) whereW3TOTd is aggregate nominal expenditure

P3cdethX3cd XSUBcd NdTHORN frac14 βcdethW3TOTd WSUBdTHORN eth9THORN

Because real aggregate consumption is usually exogenous in our partial equilibriumsimulations the linear expenditure system determines the consumption shares ofindividual final commodities (ie the five wine types plus a composite of all otherconsumption items) driven by changes in relative prices as faced by domestic con-sumers The income elasticity of demand for each commodity is equal to the mar-ginal budget share divided by the expenditure share This number varies from 05for nonpremium wine to 17 for superpremium still wine The income elasticity ofdemand for other consumption is very close to 10 because wine accounts for anaverage of only 03 of aggregate expenditures globally and no more than 11in any country (Anderson and Nelgen 2011 Table 166)

A feature of our revised model of world wine markets is the inclusion of nominalexchange rates These appear directly in the equation linking retail prices (P3scd ) toproducer prices by country of origin (P0s

c ) where c denotes the wine type

P3 scd frac14 P0s

cfd

fsTtarcd T

taxcd thorn Pm

cd eth10THORN

248 UK and Global Wine Markets by 2025 and Implications of Brexit

The exchange rates in the consuming and producing regions are fd and fs respec-tively expressed as local currency units per $US Ttar

cd is the power of the tariff in theconsuming region and Ttax

cd is the power of the domestic consumption (or excise) taxprior to any generic value-added or goods-and-services tax Pm

cd is the price of marginm assumed to be locally supplied nontradable and therefore unaffected by theexchange rate

A given level of consumption for wine type c (X3cd) is satisfied using theArmington (1969) assumption in which wines from different countries of originare imperfectly substitutable First domestic wine is imperfectly substitutable witha composite of imports

X3sscd frac14 f ethX3cd CESethP3sscd=P3cdTHORNTHORN ss frac14 domestic imports eth11THORN

and then imports by origin (X3scd ) are determined in a second CES equation

X3scd frac14 f ethX3ssfrac14importscd CESethP3scd=P3

ssfrac14importscd THORNTHORN eth12THORN

The model enables us to show how changes in international competitiveness affectthe worldrsquos wine markets A crucial part of this exercise is explaining how pricesdetermined outside the grape and wine markets influence these markets Becausethe model is in partial equilibrium to depict the impacts of changes in internationalcompetitiveness outside price changes need to be imposed as shocks on the modelThe price of intermediate inputs other than grapes and wine shown in equations (1)and (2) is set equal to the price of GDP (Pg

d ) multiplied by a shifter Fcd

Pcid frac14 Fc

dPgd eth13THORN

If no specific price observations are available the shifter Fcd remains exogenous and

unshocked with the change in price being determined by a shock to the price ofGDP If observations are available for specific input price movements the shifterFc

dbecomes endogenous with Pc

id now exogenous and shocked

Wid frac14 Fwd P

gd eth14THORN

Wage rates are treated similarly In equation (14) if the wage shifter Fwd is exogenous

changes in wage rates Wid are determined by changes in the price of GDP If wagerate data are available Fw

d becomes endogenous andwage rates are shocked directly

Pmcd frac14 Fm

d Pgd eth15THORN

The prices of trade and transport margins are also determined by the price of GDP ifthe shifter Fm

d in equation (15) is exogenous

Changes in international competitiveness depend on changes in relative pricelevels and changes in nominal exchange rates In equation (16) fR

s denotes real

Kym Anderson and Glyn Wittwer 249

exchange-rate movements relative to the US dollar

fRs frac14 Pg

s =frac12PgUSA

fs$ eth16THORN

In equation (16) the nominal exchange rate for the United States is alwaysunchanged because nominal and real exchange rates are expressed relative to theUS currency

Changes in international market conditions may have impacts in one direction onproducer prices as expressed in US dollars P0s

i and potentially in the opposite direc-tion in local currency units Hence we calculate real producer pricesP0s

iloc in localcurrency terms (ie the price most relevant to domestic producers)

P0siloc frac14 P0s

ifs=P

gs eth17THORN

To obtain real price changes in local currency terms we convert US dollar prices(P3 s

cd for the source-specific price and P3cd for the source-composite price) to reallocal currency prices (P3 s

cdloc and P3cdloc) using the consumer price index (CPI)(Pc

d ) as the deflator

P3 scdloc frac14 P3 s

cdfd=P

cd eth18THORN

and

P3cdloc frac14 P3cdfd=Pcd eth19THORN

References

Anderson K and Nelgen S (2011) Global Wine Markets 1961 to 2009 A StatisticalCompendium Adelaide University of Adelaide Press Also available as an e-book atwwwadelaideeduaupresstitlesglobal-wine

Anderson K Nelgen S and Pinilla V (2017) Global Wine Markets 1860 to 2016 AStatistical Compendium Adelaide University of Adelaide Press (forthcoming) Also tobe available as an e-book at wwwadelaideeduaupress

Anderson K and Pinilla V (with the assistance of A J Holmes) (2017)Annual Database ofGlobal Wine Markets 1835 to 2016 Wine Economics Research Centre University ofAdelaide to be posted in October at wwwadelaideeduauwine-econdatabases

Anderson K and Strutt A (2016) Impacts of Asiarsquos rise on African and Latin Americantrade Projections to 2030 World Economy 39(2) 172ndash194

Anderson K andWittwer G (2013) Modeling global wine markets to 2018 Exchange ratestaste changes and Chinarsquos import growth Journal of Wine Economics 8(2) 131ndash158

Anderson K and Wittwer G (2015) Asiarsquos evolving role in global wine markets ChinaEconomic Review 35 1ndash14

250 UK and Global Wine Markets by 2025 and Implications of Brexit

Armington P A (1969) A theory of demand for products distinguished by place of produc-tion IMF Staff Papers 16 159ndash178

Baldwin R E (ed) (2016) Brexit Beckons Thinking Ahead by Leading Economists LondonCentre for Economic Policy Research

Campos N F (2016) Lousy experts Looking back at the ex ante estimates of the costs ofBrexit In R E Baldwin (ed) Brexit Beckons Thinking Ahead by Leading Economists(ch 3) London Centre for Economic Policy Research

Dixon P Parmenter B Sutton J and Vincent D (1982) ORANI A Multisectoral Modelof the Australian Economy Contributions to Economic Analysis 142 Amsterdam North-Holland

European Commission (2015) Trade for All Towards a More Responsible Trade andInvestment Policy Brussels European Commission

Harrison J Horridge M Jerie M and Pearson K (2014) GEMPACK ManualMelbourne GEMPACK Software

HM Treasury (2017) Spring Budget 2017 London HM TreasuryHolmes A J and Anderson K (2017a)Annual Database of National Beverage Consumption

Volumes and Expenditures 1950 to 2015 Wine Economics Research Centre University ofAdelaide posted at wwwadelaideeduauwine-econdatabases

Holmes A J and Anderson K (2017b) Convergence in National Alcohol ConsumptionPatterns New Global Indicators Journal of Wine Economics 12(2) 117ndash148

Ludington C C (2013) The Politics of Wine in Britain A New Cultural History BasingstokeUK Palgrave Macmillan

Office of National Statistics (2017) Analysis of real earnings London Office of NationalStatistics Available at httpswwwonsgovukemploymentandlabourmarketpeopleinworkearningsandworkinghoursarticlessupplementaryanalysisofaverageweeklyearningslatest

Rollo J Borchert I Dawar K Holmes P and Winters L A (2016) The World TradeOrganisation A safety net for a post-Brexit UK trade policy Briefing Paper 1UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-1-final-1pdf

Smith A (2017) Brexit Hard truths and hard choices UKTPO blog 19 June Available athttpsblogssussexacukuktpo20170619brexit-hard-truths-and-hard-choices

Swinbank A (2017) World trade rules and the policy options for British agriculture post-Brexit Briefing Paper 7 UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-7pdf

Viner J (1950) The Customs Union Issue New York Carnegie Endowment for InternationalPeace

Wine Australia (2015) Export Market Guide European Union Adelaide Wine AustraliaWittwer G Berger N and Anderson K (2003) A model of the worldrsquos wine markets

Economic Modelling 20(3) 487ndash506

Kym Anderson and Glyn Wittwer 251

Page 8: U.K. and Global Wine Markets by 2025, and Implications of ... · U.K. and Global Wine Markets by 2025, and Implications of Brexit* Kym Andersona and Glyn Wittwerb Abstract The United

assume that Rollo et alrsquos suggestion will be the new UK trade policy commitmentto WTO members that subsequent negotiations for preferential arrangements willtake years and that any consequent agreements with the European Union andothers will be gradually implemented after that

The impact of leaving a customs union on wine markets does not only come fromtariff changes Also relevant are any effects of leaving on real UK incomes and thevalue of the British pound If the United Kingdom were to move immediately to freetrade on all products (the Hong Kong option) its per-capita income could eventuallyrise but only after considerable adjustment Should the United Kingdom insteadcommit to the current EU tariff schedule at the WTO in the first instance as weassume below then its per-capita income growth rate and the British poundrsquosexchange rates almost certainly will be lower for some time ndash at least until newtrade agreement negotiations with the EU27 and others are sufficiently advanced

Table 2Taxes on British Wine Imports by Source 1660ndash1862 (British Pounds per Kiloliter)

France Germany Spain Portugal South Africa

1660ndash1665 7 9 8 81666ndash1684 7 9 8 81685ndash1691 14 20 19 181692ndash1695 22 20 19 181696 47 20 19 181697ndash1702 51 25 23 221703 52 27 24 231704ndash1744 55 31 26 251745ndash1762 63 35 30 291763ndash1777 71 39 34 331778 79 43 38 371779 84 41 40 391780ndash1781 92 49 44 431782ndash1785 96 51 47 46 441786 65 51 37 37 371787ndash1794 47 51 32 32 371795 78 64 51 51 571796ndash1797 108 92 71 71 771798 111 96 73 73 791799ndash1801 107 92 71 71 771802 112 97 74 74 801803 131 109 87 87 871804 142 117 95 95 951805ndash1824 144 119 96 96 961825ndash1830 78 50 50 50 251831ndash1859 58 58 58 58 291860 32 32 32 32 321861 16 21 21 21 211862 11 26 26 26 26

Source Summarized from Ludington (2013 Table A1)

228 UK and Global Wine Markets by 2025 and Implications of Brexit

Figure 4

UK Shares in Wine Exports of Key Wine-Exporting Countries 2010ndash2014 ()

Fig4-BW

onlin

eBW

inprint

Source Compiled from data in Anderson and Pinilla (2017)

Table 3Import Duties Excise Duties and VAT Affecting Consumer Prices of Wine and Other Alcohol

in the United Kingdom April 1 2017

VAT 20

MFN import duties on wine pound per liter-bottled still wine lt13 alc 0114-bottled still wine 13ndash15 alc 0134-bottled still wine 15ndash18 alc 0162-bulk still wine lt13 alc 0086-bulk still wine 13ndash15 alc 0105-bulk still wine 15ndash18 alc 0134-sparkling wine 0278Weighted averagea 0130

Excise duties on alcohol pound per liter-still wine lt15 alc 2887-still wine 15ndash22 alc 3848-sparkling wine 55ndash85 alc 2795-sparkling wine 85ndash15 alc 3697-spirits (assumed 40 alc) 11551-beer (assumed 5 alc) 0954

VAT on alcohol 20

a Assumes that one-third of UK wine imports arrives in bulk that half has less than 13 alcohol and that one-tenth of imports is sparkling

Sources HM Revenue httpswwwgovukgovernmentpublicationsalcohol-duty-rate-changes for excise duties and VAT accessed April 92017 and Wine Australia (2015) for import duties converted at the 2016 average exchange rate of pound0740634 per euro

Kym Anderson and Glyn Wittwer 229

as to restore investor and consumer confidence in the United Kingdomrsquos economyThe size of those impacts is uncertain of course (see Baldwin 2016 and especiallyCampos 2016) so we consider a range of possibilities below

Those assumed adverse macroeconomic effects will add to the initial impact ofaltered wine tariffs on aggregate wine consumption in the United Kingdom andhence on its bilateral trades in wine They will make the loss of sales to theUnited Kingdom by EU (and Chilean and South African) suppliers greater thanwould otherwise be the case And they will reduce the likelihood that other countriesrsquosales of wine in the United Kingdom will be higher than in the baseline Indeed themacroeconomic effects could outweigh the trade-diverting effects so that even coun-tries that are currently discriminated against by the EU28rsquos wine-trade policy may beworse off because of Brexit

IV Global Wine Markets Model and Database

Our model of the worldrsquos wine markets first published by Wittwer Berger andAnderson (2003) and revised by Anderson and Wittwer (2013) is summarized inthe Appendix It disaggregates wine markets into four types namely nonpremiumcommercial-premium and superpremium still wines and sparkling wines2 There aretwo types of grapes premium and nonpremium Nonpremium wine uses nonpre-mium grapes exclusively superpremium wines use premium grapes exclusivelyand commercial-premium and sparkling wines use both types of grapes to varyingextents across countries The world is divided into 44 individual nations and 7 com-posite geographic regions that capture all other countries

The modelrsquos database is calibrated to 2014 based on the comprehensive wine-market volume and value data and trade and excise-tax data provided inAnderson and Pinilla (2017) and in Anderson Nelgen and Pinilla (2017) It is pro-jected assuming that aggregate national consumption population and realexchange rates change between 2014 and 2025 to the extent shown in AppendixTable 13 The Brexit alternatives to that baseline also are projected to 2025

Concerning preferences we assume a continued considerable swing toward allwine types in China and a swing away from nonpremium wines in all other countriesuntil 2025

In our baseline scenario grape- and wine-industry total-factor productivity areassumed to grow at 1 per year everywhere while grape- and wine-industry

2Commercial-premium still wines are defined byAnderson Nelgen and Pinilla (2017) to cost between US$250 and US$750 per liter pretax at a countryrsquos border or wholesale3The real exchange rate changes over the projection period are the changes expected in the nominal valueof country irsquos currency relative to the US dollar times the expected ratio of the GDP deflator for theUnited States versus that for country i

230 UK and Global Wine Markets by 2025 and Implications of Brexit

capital is assumed to grow net of depreciation at 15 per year in China but zero else-where (consistent with the almost-zero growth in global wine production and con-sumption over the past two decades)

Two alternative scenarios are considered (ldquolargerdquo and ldquosmallrdquo) for the initialimpact of Brexit to provide a range of results We assume that because of theUnited Kingdomrsquos decision to exit the European Union the rate of UK realgross domestic product (GDP) growth is only one-third or two-thirds as fast overthe projection period (09 or 18 per year instead of 26) and the Britishpound will be 20 or 10 lower in real terms than in our modelrsquos core baselineprojection4 In both alternative scenarios we assume that the United Kingdomapplies the European Unionrsquos external tariffs on wine from March 30 2019 atthe end of the two-year period following the United Kingdomrsquos formal triggeringof Article 50 (see Table 3)

We then consider a subsequent impact of Brexit presumed to result from negoti-ating signing implementing and responding to an FTA with the EU27 by 2025(Chile and South Africa currently have preferential access to EU wine marketsbut we continue to assume in this scenario that the United Kingdom does not imple-ment new bilateral FTAs with them or others in our time frame)5 This subsequentscenario assumes the British pound returns to what it would have been in 2025 in theabsence of Brexit and that real incomes regain two-thirds of the value they wouldhave had without Brexit as compared with our ldquosmallrdquo initial Brexit scenario

This global model has supply-and-demand equations and hence quantitiesprices and price elasticities for each of the grape and wine products and for asingle composite of all other products in each country Income elasticities ofdemand also exist for each final product Grapes are assumed to not be traded inter-nationally but other products are exported and imported Each market is assumed tohave been in equilibrium before any shock and to find a new market-clearingoutcome following any exogenously introduced shock

4The nominal price of the British pound in US dollars in the fortnight following the Brexit vote on June23 2016 dropped 13 to US$130 and a year later the pound sat at the same rate having dropped to anaverage of $124 (an 18 devaluation) between October 2016 and April 2017 Our choice of a low of 10and a high of 20 aims to capture future possible rates while uncertainties remain The average real wagein the United Kingdom fell in the first half of 2017 and projected real GDP growth during 2018ndash2020 hasbeen revised down to 17 in the United Kingdomrsquos latest budget (HM Treasury 2017)5South Africa currently has duty-free access to the EU28 for just 50 megaliters (ML) of wine per yearbeyond which the MFN tariff rate applies South Africa currently exports around 320 ML to theEuropean Union one-third of which initially goes to the United Kingdom How that quota of 50 MLis divided between the United Kingdom and the EU27 is subject to future negotiation (Rollo et al2016 Swinbank 2017) In the following discussion we assume none of it is accepted by the UnitedKingdom so that all South African wine imports pay the United Kingdomrsquos MFN tariff under Brexitbut altering that assumption makes very little difference to our results for any countryrsquos trade exceptSouth Africarsquos

Kym Anderson and Glyn Wittwer 231

V Projecting Global Wine Markets to 2025

Global wine production and exports are projected in the baseline from 2014 to 2025consistent with past trends The modelrsquos global volume of production (and consump-tion) rises little over that 11-year period (9) made up of a 6 decline in nonpre-mium wine and a one-sixth rise in commercial-premium and superpremium wineIn real (2014 US$) value though global wine output and consumption increaseby about 50 in total and 60 in the two premium categories The internationaltrade projections are similar although a little larger with the share of global wineproduction exported (equivalent to the share of global consumption imported)rising 2 percentage points between 2014 and 2025

The baseline projection does not greatly alter the 2014 shares of various countriesin global wine production apart from China because we assume vineyard expansionthere is faster than elsewhere6 In value terms this assumption means that Chinamoves from fifth to fourth by 2025 behind France the United States and ItalySpain remains barely ahead of Australia and they and Germany take the nextthree places (Figure 5(a)) In total wine-production volume terms China movesfrom sixth to fifth place and Argentina drops from fifth to eighth (and fromeighth to ninth in value terms)

When their products are subdivided into fine wines (superpremium still plus spark-ing) commercial-premium wines and nonpremium wines France and the UnitedStates retain the highest two places on the global ladder for fine-wine productionand Spain and Italy retain the top two places for nonpremium wine As for commer-cial-premium wine production (defined to cost between US$250 and US$750 perliter pretax at a countryrsquos wholesale level or national border) Italy retains the topranking over our projections period but at least in terms of value China challengesFrance for second place

The country rankings by projected value of total wine consumption change some-what more than those for production by 2025 with China taking second place afterthe United States ahead of France and Germany and then the United Kingdomslightly overtaking Italy to slip into fifth place (Figure 5(b)) The United StatesFrance and Germany retain the top three rankings for consuming fine wine butCanada slightly overtakes Italy for fourth place in terms of value at least In thecase of commercial-wine consumption China strengthens its number-one positionahead of the United States and the United Kingdom does likewise vis-agrave-visGermany for third place

As for the projected changes in consumption volumes China is projected to dom-inate the increase in aggregate although the United States is projected to lead the

6 In fact Chinarsquos wine production fell steadily between 2012 and 2016 by a total of one-sixth so Chinarsquoswine imports may grow faster in practice than in this baseline projection

232 UK and Global Wine Markets by 2025 and Implications of Brexit

Figure 5

Value of Wine Production and Consumption in Key Countries 2014 and Projected Baseline for2025 (2014 US$ Million at WineryWholesale Pretax Prices)

Fig5-BW

onlin

eBW

inprint

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 233

increase in consumption of fine wine In Western Europe and in the SouthernHemispherersquos New World countries fine wines are projected to substitute for com-mercial wines (defined as the sum of commercial-premium and nonpremiumwines) with almost no change in total wine consumption Sub-Saharan Africa isthe next region that is projected to take off with its growth accounting for morethan one-third of the rest of the worldrsquos increase in volume consumed

Those differences in production versus consumption rankings are reflected ininternational trade Figure 6 shows that France Italy and Spain remain the threedominant exporters of wine in aggregate value but the rankings of the next fewchange with Australia being slightly ahead of Chile and the United StatesGermany and New Zealand being nearly tied for sixth place in value termsFrance and Italy are even more dominant in fine-wine exports and remain so by2025 while Italy outranks France in the commercial-premium export categoryand Spain outranks Italy Australia and Chile in the nonpremium export class

Among the importers the United States and the United Kingdom are projected tocontinue to hold the first two places in 2025 in value terms but China moves intothird place slightly ahead of Germany followed well behind by Canada HongKong Belgium-Luxembourg the Netherlands and Japan (Figure 6(b)) OtherAfrica (excluding South Africa) is projected to experience the largest increase inimports among all the other regions followed by Other Asia which becomes asbig as Germany in value terms (Figure 7(a)) In terms of total volume of wineimports Germany and the United Kingdom held the top two shares in 2014 butby 2025 the United Kingdom is projected to be well ahead of Germany (Figure 7(b)) However this projection ignores the effects of Brexit to which we now turn

VI How Might Wine Markets Be Affected by the United Kingdomrsquos Exitfrom the European Union

As mentioned earlier for our two alternative scenarios to capture the initial effects ofBrexit (ldquolargerdquo and ldquosmallrdquo) we assume that following the United Kingdomrsquos exitfrom the European Union the United Kingdomrsquos rate of economic growth wouldbe only one-third or two-thirds as fast for the period to 2025 the British poundwould be 20 or 10 lower in real terms than in our modelrsquos baseline projectionand the United Kingdom would apply the European Unionrsquos external tariff onwine to imports from EU member countries (as part of establishing MFN ratesvia the WTO to then start new bilateral FTA negotiations) In these initial scenarioswe assume the United Kingdom does not implement any new FTAs particularlywith the EU27 Chile and South Africa Following the discussion of those initialresults we present the results of a subsequent scenario that includes the implemen-tation of an FTA between the United Kingdom and the EU27 We assume these aretwo discrete steps with no agreed arrangements to smooth the transition betweenthem

234 UK and Global Wine Markets by 2025 and Implications of Brexit

Figure 6

Value of Wine Exports and Imports Key Wine-Trading Countries 2014 and Projected Baselinefor 2025 (2014 US$ million)

Fig6-BW

onlin

eBW

inprint

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 235

Figure 7

National Shares of Global Wine Import Value and Volume 2014 and Projected Baselinefor 2025 ()

Fig7-BW

onlin

eBW

inprint

Source Authorsrsquo model results

236 UK and Global Wine Markets by 2025 and Implications of Brexit

A Initial Impact of Brexit

Because the British pound dropped by one-sixth against the US dollar in the fourmonths following the Brexit vote in June 2016 and because the UnitedKingdomrsquos average real wage fell 1 between November 2016 and April 2017(according to the latest available data see Office of National Statistics 2017) weviewed our ldquolargerdquo scenario as more likely soon after the Brexit vote But wecompare those results with ones from our ldquosmallrdquo initial-impact scenario becauseit has become more likely following the June 2017 general election that ldquosofterrdquooptions will be considered by the new government Generally the results areabout half the size in the ldquosmallrdquo scenario with the exception of the bilateraltrade effects To show the sensitivity of results to our assumptions we point outthe differences when the ldquosmallrdquo results are not close to half the results shown forthe ldquolargerdquo scenario

Before turning to those results we first present a ldquofree-traderdquo scenario whichassumes the United Kingdom chooses the radical option of becoming the HongKong of Europe Unlikely as this scenario is the results provide assurance thatour global wine model generates the standard types of effects of exiting a customsunion In this ldquofree-traderdquo scenario the United Kingdom is assumed to move tozero tariffs on all wine imports Thus the current preferences on EU Chileanand South African wine imports disappear as all other wine exporters enjoy thesame free access to UK wine markets We assume in this scenario that Brexit hasno impact on the United Kingdomrsquos real income or consumption growth rates by2025 and that by then the British pound has returned to its 2014 value relative toother currencies (similar to the real exchange rate assumed in our baseline scenarioto 2025 see Appendix Table A1)7 Under these assumptions EU Chilean andSouth African wine exports to the United Kingdom in 2025 are 003 lower thanin the no-Brexit baseline and UK imports from other wine exporters are 025higher but the United Kingdomrsquos total wine imports are hardly any different(because local wine prices fall by only 05 and incomes are unchanged) In this sce-nario UK consumers expand their volume of consumption by just 05 while theEU27 exports more and other countries export less to the rest of the world such thatworld wine exports are almost unchanged In short all these effects are in the direc-tion that comparative static-customs-union theory would predict but they are smallbecause we assume the United Kingdomrsquos real income and currency are the same asin our baseline for 2025 as described in the previous section

In the ldquolargerdquo alternate initial scenario involving Brexit as compared with thebaseline scenario to 2025 the consumer price of wine in 2025 is 22 higher in the

7If this scenario were seen as a credible long-run trade policy it would encourage more investment thatwould eventually raise UK incomes but not without major structural changes that would initiallydisrupt the economy Implicitly we assume in our free-trade scenario that these two opposite impactson real UK incomes and the British pound exactly offset each other by 2025

Kym Anderson and Glyn Wittwer 237

United Kingdom in local currency terms (20 because of real depreciation of theBritish pound 4 because of new tariffs on EU Chilean and South Africanwines and ndash2 because of slower UK income growth) The volume of UKwine consumption is 28 lower 16 because of slower UK economic growth7 because of real depreciation of the British pound and 5 because of newtariffs Superpremium still-wine sales are the most affected dropping by two-fifths while sparkling and commercial-premium wines drop a bit less than one-quarter Because the average price rises by more than the decline in volume soldthe aggregate value of UK sales even in local currency terms falls under thisldquolargerdquo Brexit scenario Under the ldquosmallrdquo Brexit scenario the consumer price ofwine in 2025 is 11 higher in the United Kingdom and its volume of wine consump-tion is 17 lower

The volume of projected UK imports in 2025 is 427 ML or nearly one-quarterless in the ldquolargerdquo scenario than in the baseline scenario comprising 58ML less spar-kling wine 31 ML less superpremium still wine and 339 ML less commercial-premium wine World imports are lower by just 239 ML because imports byother countries are 189 ML higher in response to the lower international wineprices in this scenario In value terms UK imports are $175 billion (or 27)lower in 2025 because of ldquolargerdquo Brexit $113 billion because of lower incomes$038 billion because of the fall in the British pound and $014 billion because ofthe rise in wine-import tariffs (Table 4) These aggregate trade impacts are a littlemore than half this size under the ldquosmallrdquo Brexit scenario

Despite the import levels falling because of raised import tariffs domestic con-sumption of all three quality categories of UK-produced wine is lower with thanwithout Brexit because of shrunken demand for all wines resulting from loweredUK incomes and raised local prices because of devaluation of the British poundThe British poundrsquos devaluation does make it easier for the United Kingdom tosell wines abroad though The countryrsquos exports are 7 ML or nearly 5 higherin 2025 in the ldquolargerdquo Brexit scenario and UK production is 3 higher ThoseUK exports (or re-exports of imported bulk wine after it is bottled in the UnitedKingdom) that go to EU27 countries are reduced though because of the tariffnow imposed at the new EU border

Without Brexit the United Kingdomrsquos shares of global wine imports are slightlyhigher in volume terms in 2025 than in 2010ndash2015 but 2 percentage points lower invalue terms thanks to East Asiarsquos expanding demand for imports of premium winesWith ldquolargerdquo Brexit however that value share is an additional 2 percentage pointslower and the volume share is almost 5 points lower (Figure 8) The net effects ofthese impacts on global trade are shown in Figure 9 Most of the initial tradeeffect of Brexit is a large decline in net imports of wine by the United Kingdomwith very little offsetting positive effect on trade in the rest of the world Theldquosmallrdquo Brexit numbers are a bit more than half of those for ldquolargerdquo Brexit

238 UK and Global Wine Markets by 2025 and Implications of Brexit

Table4

Difference

in2025

Projected

VolumeandValue

ofWineIm

portsb

ytheUnitedKingdom

andtheResto

fthe

World

asaConsequence

oftheInitial

BrexitS

hock

(MLand2014US$

Million

ldquoLargerdquoSc

enario)

Volum

e(M

L)

Value

(US$

million)

NP+CPa

SuperPrb

Sparkling

TOTAL

NP+CPa

SuperPrb

Sparkling

TOTAL

ΔUK

impo

rtsdu

eto

Low

erincomes

minus198

minus20

minus29

minus247

58minus644

minus253

minus234

minus1131

65Low

erpo

und

minus70

minus10

minus14

minus93

22minus248

minus127

minus102

minus476

27Highertariffs

minus71

minus1minus16

minus87

20minus110

minus8minus2

4minus143

8TOTA

Lminus339

minus31

minus58

minus427

100

minus1001

minus388

minus360

minus1750

100

difffrom

base

2332

3325

2432

3227

of

totalc

uts

797

14100

5722

21100

ΔROW

netim

ports

143

2125

189

230

143

minus181

192

ΔWORLD

TRADE

minus195

minus10

minus34

minus239

minus763

minus246

minus543

minus1552

aNon

prem

ium

plus

commercial-premium

still

wines

bSu

perpremium

still

wines

SourceA

utho

rsrsquomod

elresults

Kym Anderson and Glyn Wittwer 239

The aggregate effect of ldquolargerdquo Brexit on the market shares of various wine-export-ing countries in the United Kingdom is almost indiscernible even with one decimalpoint The projected 2025 shares are quite different from the actual 2014 shares forseveral countries They are much smaller in 2025 for South Africa Australia andNew Zealand (and the United States in volume terms) and are much larger involume for Spain and in value for Italy These results occur because wine-exportingcountries benefit differentially from the varying rates of growth in net importdemand for wine in non-UK countries over this projection period The most impor-tant projected changes are the increase in the real value of annual wine importsbetween 2014 and 2025 by China (200 or $3 billion) Other Asia (110 or $22billion) and Africa (270 or $16 billion) More than half of Australiarsquos increasein annual exports from 2014 to 2025 go to Asia and more than half of SouthAfricarsquos increase in exports go to Other Africa

Table 5 reveals that ldquolargerdquo Brexit lowers EU Chilean and South African wineexports by 150 ML or US$12 billion in the case of the European Union with someof their exports diverted from the United Kingdom to the EU27 and other marketsin competition with New World exporters Although the United States Australiaand Argentina sell only a little less to the United Kingdom they also sell less toother countries For Chile and South Africa which lose their preferential access toUK (but not to EU27) markets in this Brexit scenario some exports are redirected

Figure 8

UK Shares of World Wine Imports 2010ndash2015 and Projected to 2025 without and with theldquoLargerdquo Initial Brexit Shock ()

Fig8-BW

onlin

eBW

inprint

Sources Anderson and Pinilla (2017) and authorsrsquo model results

240 UK and Global Wine Markets by 2025 and Implications of Brexit

from the United Kingdom to EU27 countries ndash but again they export less overallGlobal wine trade in 2025 is 240 ML (19) or $18 billion (35) less under thisldquolargerdquo Brexit scenario The percentage by which wine exportersrsquo trade shrinks isgreater for values than for volumes because of changes in relative prices of differ-ent-quality wines Those differences are shown in the numbers in parentheses inTable 5

Three other points are worth making about Table 5 One is that Australia sellsslightly more to the United Kingdom in the ldquosmallrdquo Brexit scenario rather thanslightly less as in the ldquolargerdquo Brexit case Evidently the negative income and price(devaluation) effects do not more than offset the positive trade-diverting effect onAustralian exports to the United Kingdom of removing preferences in the ldquosmallrdquoscenario Second New Zealand sells slightly more to non-UK countries underBrexit despite greater competition from the EU27 Chile and South Africa Thisanomaly is due to changes in the relative prices of different qualities of wine inglobal wine markets bearing in mind that New Zealand has the worldrsquos highestaverage price for still-wine exports And third the value (but not the volume) ofexports of ldquoOtherrdquo countries to markets other than the United Kingdom arehigher under Brexit This result too is due to changes in the relative prices of differ-ent qualities of wine in global wine markets

Figure 9

Difference in 2025 Wine Import Volumes and Values as a Result of the ldquoLargerdquo Initial BrexitShock (ML and US$ Million in 2014 US dollars)

Fig9-BW

onlin

eBW

inprint

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 241

B Subsequent Impact of Brexit from a UK-EU27 FTA

The next-most-likely step in the Brexit process is for the United Kingdom to nego-tiate a new trade arrangement with the EU27 We therefore assume that a UK-EU27FTAwith free bilateral wine trade is implemented and adjusted to by 2025 and thatprogress toward that end occurs soon enough that the adverse macroeconomicshocks from the initial impact of uncertainty over the Brexit process are confinedto those assumed in the ldquosmallrdquo scenario outlined above In this subsequent scenariowe assume that the British pound returns to the value it would have reached in 2025in the absence of Brexit (ie reversing the 10 devaluation assumed in the ldquosmallrdquoinitial Brexit scenario) and that real incomes in the United Kingdom are 8 higher

Table 5Difference in 2025 BilateralWine Import Volumes and Values fromKey Exporters by the United

Kingdom and the Rest of the World (RoW) as a Result of Initial Brexit Shock(ML and 2014US$ Million)a

(a) ldquolargerdquo scenario

Volume (ML) Value (2014US$ million)

UK RoW WORLD () UK RoW WORLD ()

EU27 minus287 136 minus150 (ndash17) minus1187 minus5 minus1192 (ndash31)Chile minus59 35 minus25 (ndash30) minus169 31 minus138 (ndash48)South Africa minus53 35 minus18 (ndash32) minus105 20 minus85 (ndash67)USA minus7 minus6 minus13 (ndash24) minus75 minus40 minus115 (ndash50)Australia minus4 minus3 minus7 (ndash09) minus25 minus65 minus90 (ndash30)Argentina minus3 minus9 minus12 (ndash48) minus16 minus39 minus55 (ndash52)New Zealand minus11 9 minus2 (ndash09) minus162 71 minus91 (ndash43)Others minus2 minus10 minus12 (ndash02) minus11 minus52 minus63 (ndash44)WORLD minus427 187 minus240 (ndash19) minus1750 minus79 minus1829 (ndash35)

(a) ldquosmallrdquo scenario

Volume (ML) Value (2014US$ million)

UK RoW WORLD () UK RoW WORLD ()

EU27 minus178 82 minus96 (ndash12) minus692 minus43 minus736 (ndash19)Chile minus46 28 minus18 (ndash24) minus128 36 minus91 (ndash32)South Africa minus43 29 minus14 (ndash42) minus82 23 minus59 (ndash47)USA 1 minus6 minus5 (ndash11) minus23 minus28 minus51 (ndash22)Australia 5 minus10 minus5 (ndash06) 19 minus56 minus38 (ndash13)Argentina 0 minus6 minus6 (ndash26) minus3 minus25 minus29 (ndash27)New Zealand minus5 4 minus1 (ndash06) minus80 34 minus46 (ndash22)Others 0 minus9 minus9 (ndash01) minus1 minus33 minus34 (ndash24)WORLD minus266 112 minus154 (ndash13) minus991 minus92 minus1083 (ndash21)

a Numbers in parentheses are the percentage difference between the Brexit and baseline scenarios for 2025 projected wine-import volumes orvalues by source

Source Authorsrsquo model results

242 UK and Global Wine Markets by 2025 and Implications of Brexit

in 2025 than in the ldquosmallrdquo scenario ndash that is they regain two-thirds of the value theywould have reached without Brexit-related uncertainty

This subsequent development in the Brexit process reverses most of but not allthe initial effects of Brexit by 2025 because of our assumption that the lost growthin the initial years of uncertainty following the Brexit vote are only partly recoveredby 2025 following the implementation of a UK-EU27 FTA Moreover the longer ittakes before this FTA is finalized and implemented the longer the estimated initialadverse effects persist and the larger the cumulative cost of Brexit to UK wine con-sumers and to grape and wine producers in wine-exporting countries

Table 6 summarizes the subsequent trade effects for 2025 It suggests that all butone-tenth of the loss in value of world trade in wine from the initial ldquosmallrdquo impact isrestored most of it because of a smaller reduction in the United Kingdomrsquos wineimports following the FTA Most of that improved outcome is because of recoveredimports from the EU27 commensurate with the latterrsquos high share of UK importsEven though Chile and South Africa are assumed in this scenario to have not yetsigned an FTA with the United Kingdom they export slightly more to the UnitedKingdom (and even more to all other countries) than in the initial ldquosmallrdquo scenario

Within the United Kingdom this FTA brings down the local currency consumerprice of wine by 9 largely offsetting the 11 rise in the initial ldquosmallrdquo Brexit sce-nario and it raises the volume consumed in the United Kingdom by 18 fully off-setting the 17 fall in the initial ldquosmallrdquo Brexit scenario

Table 6Difference in 2025 Bilateral Wine-Import Volumes and Values from Key Exporters by the

United Kingdom and the Rest of the World (RoW) as a Result of Implementing a UK-EU27FTA Following Initial ldquoSmallrdquo Brexit Shock (ML and 2014US$ Million Difference Relative to

Initial ldquoSmallrdquo Brexit Scenario)a

Volume (ML) Value (2014US$ million)

UK RoW WORLD (ldquosmallrdquo)a UK RoW WORLD (ldquosmallrdquo)a

EU27 212 minus112 100 (minus96) 750 0 750 (minus806)Chile 1 3 4 (minus18) 3 40 43 (minus92)South Africa 3 2 5 (minus14) 3 18 21 (minus60)USA 3 3 6 (minus5) 26 21 47 (minus50)Australia 1 1 2 (minus5) 8 38 46 (minus33)Argentina 1 4 5 (minus6) 4 20 24 (minus28)New Zealand 4 minus3 1 (minus1) 61 minus25 36 (minus46)Others 1 6 7 (minus9) 5 minus46 minus41 (92)WORLD 226 minus96 130 (minus154) 860 66 926 (minus1022)

a Numbers in parentheses are the world trade differences between the ldquosmallrdquo initial Brexit scenario and the baseline scenario copied fromcolumns 3 and 7 of Table 5(b)

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 243

In short this subsequent step in the Brexit process can be expected to restore muchof the initial adverse effects in the United Kingdom of the Brexit vote by 2025 Thewine exporters in the EU27 Chile and South Africa also lose less in this scenarioas compared with the initial ldquosmallrdquo scenario while other key exporters sell nearlyas much in 2025 in this scenario as they are projected to in the baseline ndash that isthey are beneficiaries of the reduced discrimination in global wine markets in thiscase where Chile and South Africa no longer enjoy preferential access to the UKmarket (and even though EU27 producers are again allowed to do so) Keep inmind though that although the estimated losses in the initial scenarios are reportedjust for 2025 they are expected to be felt throughout the preceding years of uncertaintyTherefore the longer it is before uncertainty abates and the United Kingdomrsquos currentpreferential trading arrangements are replaced by newones such as a UK-EU27 FTAthe higher the cumulative cost of the Brexit vote to the wine trade

VII Caveats and Conclusions

The above Brexit simulations are just a few of many scenarios that could be modeledObvious additional ones could also assume that FTAs are reached between theUnited Kingdom and other trading partners including Chile and South Africaand such countries as Australia and New Zealand Some of these other countriesare already exploring the ECrsquos invitation (see European Commission 2015) to nego-tiate an FTA with what will be the EU27 The sequence in which FTAs are signedand the speed with which they are implemented will matter (as was also the casewith the sequential signing over the past decade of bilateral FTAs with NortheastAsian countries by Chile Australia and New Zealand see Anderson andWittwer 2015) Even if the United Kingdom were able to sign additional bilateralFTAs and begin implementing them before 2025 it would have little effect on theabove results (because wine tariffs are such a minor contributor to them) unlessthose FTAs were to accelerate the United Kingdomrsquos economic growth and therise in the British poundrsquos value before 2025

We assume above that no changes are made to alcohol excise duties in the UnitedKingdom following Brexit even though they are scheduled to be progressively raisedwith inflation and may be raised even more for wine relative to spirits to offset theopposite effects of Brexit on those two domestic industries Nor do we make anyallowance for UK increases in consumer wine prices that may be needed to coverthe higher cost of clearing customs on imports from EU27 countries (which maybe more or less offset by increased smuggling and duty-free purchasing)Importers of fine wines including individuals who buy Bordeaux en primeur maywell reduce their demand because of concerns that delays on docks will affect thequality of their wines or because of greater currency uncertainty

To summarize Brexit is costly initially to UK consumers of wine (and of manyother tradable products) because the domestic retail price in local currency tax-

244 UK and Global Wine Markets by 2025 and Implications of Brexit

inclusive terms is 22 higher than otherwise in 2025 and the volume of wine con-sumed domestically is 28 lower in the ldquolargerdquo Brexit scenario (or 11 higher pricesand 17 lower quantities in the ldquosmallrdquo case) by 2025 unless a UK-EU27 FTA isdeveloped before then Even if such an FTA does get signed ratified by all 28 par-liaments and implemented by 2025 the slower income growth in the interim meansa smaller UK wine market in 2025 than would otherwise have been the case Thevolume reduction is a blow to many participants in UK wine-bottling transportingstoring wholesaling and retailing businesses in addition to restaurants and pubsVery little of that initial impact occurs because of higher import tariffs most impor-tant is the assumed fall in UK real incomes

The small but growing number of local UK vignerons (who supply less than 05of all domestic-wine sales and less than 4 of sparkling-wine sales) is initially pro-jected to sell less wine domestically because of reduced demand and to export only alittle extra abroad while the British pound is devalued Overall their production isonly 3 higher in 2025 even in the ldquolargerdquo Brexit scenario (not taking intoaccount the fact that their casual labor is likely to be more expensive under Brexitbecause of tighter restrictions on immigration) and their production diminishes ifand when a UK-EU27 FTA is in place

Even with a UK-EU27 FTA in place EU27 wine exporters export US$56 millionless wine in 2025 thanks to Brexit Chile and South Africa export $88 million lesswine and wine exports from the rest of the world are little different in aggregatealthough slightly larger from Australia

Clearly there will be great uncertainly for some time over the possible policy out-comes to flow from Brexit and their consequent sequential impacts on UK house-hold disposable incomes foreign exchange rates and bilateral wine tariffsMeanwhile the above projections under explicit assumptions provide some idea ofhow wine markets might be affected by the most-likely first two stages of theBrexit process (agreeing on a new tariff schedule at the WTO and agreeing to andimplementing a UK-EU27 FTA) In particular they make clear that nontrivialinitial impacts could affect the domestic wine market impacts that are likely to belarger than just the direct impact of changes in bilateral tariffs If the UnitedKingdom succeeds in getting countries to agree to transition arrangements thatdelay the changes in tariffs until new FTAs are signed and ratified the initialeffects will be less dramatic than in our first scenarios but it remains to be seenwhether any such agreements can be reached In any event the net effect of Brexiton the welfare of the worldrsquos consumers and producers of wine as a whole will benegative not just initially but permanently unless new trade policy commitmentsby the United Kingdom with major wine-exporting countries are sufficiently moreliberal than current arrangements

Kym Anderson and Glyn Wittwer 245

Appendix Key Equations in the Global Wine Markets Model

Amodel of the worldrsquos wine markets was first published byWittwer et al (2003) andmuch improved by Anderson and Wittwer (2013) Several significant enhancementshave been made to that original model Wine types have been disaggregated intomore types ndash namely nonpremium (including generic bulk) commercial-premiumand superpremium still wines and sparkling wines8 As in the original modelthere are two types of grapes premium and nonpremium Nonpremium wines usenonpremium grapes exclusively superpremium wines use mainly premium grapesand commercial-premium and sparkling wines use both types of grapes In theregional dimension the number of countries and country groups has expandedfrom 10 in the original model to 51 44 individual nations and 7 compositeregions The modelrsquos database is calibrated to 2014 for this paper based on datain Anderson et al (2017) trade data for which are downloaded from httpscomtradeunorg The modelrsquos supply-and-demand equations are based on the orig-inal ORANI modelrsquos theory (see Dixon Parmenter Sutton and Vincent 1982) andthe model is implemented using GEMPACK software (Harrison Horridge Jerieand Pearson 2014)

An enhancement of importance to the present study is the inclusion of exchange-rate variables in the model which allow a distinction between price impacts asobserved in local currency units and those observed in 2014 US dollars

In the model the grape and wine sectors minimize costs of intermediate inputssubject to weak constant elasticity of substitution (CES) substitutability betweeninputs By assumption no intermediate inputs are imported from other countries9Hence

Xcid frac14 f ethX1id CESfrac12P

cid=P1id $THORN eth1THORN

P1id X1id frac14X

c

Xcid P

cid eth2THORN

whereXcid is the quantity demanded of commodity c by grape or wine industry i in

region d Pcid is the corresponding price and X1id and P1id are the respective inter-

mediate composite quantities and prices

Two primary factors are employed in the sector labor (the quantity of which isendogenous with perfectly elastic supply) and capital Capital is usually treated asexogenous in quantity with rates of return bearing all the adjustment in thevarious scenarios This reflects the fact that grapes (a perennial crop) and wine-

8Commercial-premium still wines are defined by Anderson and Nelgen (2011) to be those priced betweenUS$250 and $750 per litre pre-tax at a countryrsquos border or wholesale9An exception concerns bulk wine imports used mainly in the bottling of commercial premium wine in theUK USA Canada Japan and Germany

246 UK and Global Wine Markets by 2025 and Implications of Brexit

plant capacity adjust slowly to market signals

Lid frac14 f ethFid CESfrac12Wid=PFid $THORN eth3THORN

Kid frac14 f ethFid CESethRid=PFid $THORN eth4THORN

PFid Fid frac14 Lid Wid thorn Kid Rid eth5THORN

Grape and wine producers are assumed to minimize costs subject to CES substitu-tion between capital and labor Equations (3) to (5) show primary factor demandsfor the labor composite Lid and capital Kid subject to a composite factor demandFid by industry i in region d The factor prices are Wid for labor Rid for capitalrentals and PFid for composite factor prices

Appendix Table A1Cumulative Consumption and Population Growth Rates and Changes in the Real Exchange Rate

(RER) Relative to the US Dollar 2014 to 2025 without Brexit ()

Aggregateconsumption Poprsquon RER

Aggregateconsumption Poprsquon RER

France 18 4 minus11 Australia 35 11 minus17Italy 11 2 minus9 New Zealand 32 9 minus26Portugal 14 0 minus9 Canada 27 8 minus18Spain 26 8 minus9 United States 31 8 0Austria 19 4 minus7 Argentina 7 10 109Belgium 20 7 minus9 Brazil 16 8 minus29Denmark 22 2 minus9 Chile 55 8 minus2Finland 21 3 minus7 Mexico 42 12 minus8Germany 14 minus2 minus11 Uruguay 45 3 1Greece 22 minus1 minus14 Other Latin America 60 10 minus5Ireland 42 12 minus9 South Africa 36 12 minus1Netherlands 21 4 minus9 Turkey 50 8 20Sweden 24 9 minus13 North Africa 53 11 0Switzerland 18 8 minus6 Other Africa 109 18 84United Kingdom 32 6 1 Middle East 52 18 minus12Other West Europe 21 10 minus1 China 79 3 5Bulgaria 41 minus7 7 Hong Kong 42 3 2Croatia 20 minus2 minus1 India 134 13 17Georgia 35 0 23 Japan 11 minus3 minus24Hungary 25 minus3 minus11 Korea 38 1 minus9Moldova 49 minus11 13 Malaysia 62 15 minus16Romania 45 minus4 22 Philippines 75 18 7Russia 18 minus2 minus8 Singapore 44 21 minus22Ukraine 22 minus5 14 Taiwan 29 1 minus13Other East Europe 40 minus5 48 Thailand 47 3 minus9

Other Asia 99 10 10

Source Authorsrsquo compilation from projections by various international agencies and from global economy-wide modeling by Anderson andStrutt (2016)

Kym Anderson and Glyn Wittwer 247

The composite factor demand Fid is proportional to total output Qid subject to aprimary factor using technology Aid Hence

Fid frac14 Qid Aid eth6THORN

The perfectly competitive zero-pure-profit condition is that total revenue valued atthe output price P0s

i multiplied by Qid equals the total production cost

P0si Qid frac14

X

cPcid X1cid thorn

X

oWo

id Loid thorn Rid Kid eth7THORN

Household demands follow a linear expenditure system in each region We reducethe optimizing problem for household consumption of each commodity subjectto a budget constraint to equations describing subsistence and discretionarydemands Aggregate subsistence expenditure WSUBd depends only on consumerprices P3cd for each commodity and the number of households N as per-capita sub-sistence quantities XSUBcd subject to given preferences are constant

WSUBd frac14X

c

P3cd XSUBcd Nd eth8THORN

Discretionary expenditures for each commodity (the left-hand side of equation (9))are equal to the marginal budget share (βcd) of aggregate discretionary expenditureThis aggregate is the bracketed term on the right-hand side of equation (9) whereW3TOTd is aggregate nominal expenditure

P3cdethX3cd XSUBcd NdTHORN frac14 βcdethW3TOTd WSUBdTHORN eth9THORN

Because real aggregate consumption is usually exogenous in our partial equilibriumsimulations the linear expenditure system determines the consumption shares ofindividual final commodities (ie the five wine types plus a composite of all otherconsumption items) driven by changes in relative prices as faced by domestic con-sumers The income elasticity of demand for each commodity is equal to the mar-ginal budget share divided by the expenditure share This number varies from 05for nonpremium wine to 17 for superpremium still wine The income elasticity ofdemand for other consumption is very close to 10 because wine accounts for anaverage of only 03 of aggregate expenditures globally and no more than 11in any country (Anderson and Nelgen 2011 Table 166)

A feature of our revised model of world wine markets is the inclusion of nominalexchange rates These appear directly in the equation linking retail prices (P3scd ) toproducer prices by country of origin (P0s

c ) where c denotes the wine type

P3 scd frac14 P0s

cfd

fsTtarcd T

taxcd thorn Pm

cd eth10THORN

248 UK and Global Wine Markets by 2025 and Implications of Brexit

The exchange rates in the consuming and producing regions are fd and fs respec-tively expressed as local currency units per $US Ttar

cd is the power of the tariff in theconsuming region and Ttax

cd is the power of the domestic consumption (or excise) taxprior to any generic value-added or goods-and-services tax Pm

cd is the price of marginm assumed to be locally supplied nontradable and therefore unaffected by theexchange rate

A given level of consumption for wine type c (X3cd) is satisfied using theArmington (1969) assumption in which wines from different countries of originare imperfectly substitutable First domestic wine is imperfectly substitutable witha composite of imports

X3sscd frac14 f ethX3cd CESethP3sscd=P3cdTHORNTHORN ss frac14 domestic imports eth11THORN

and then imports by origin (X3scd ) are determined in a second CES equation

X3scd frac14 f ethX3ssfrac14importscd CESethP3scd=P3

ssfrac14importscd THORNTHORN eth12THORN

The model enables us to show how changes in international competitiveness affectthe worldrsquos wine markets A crucial part of this exercise is explaining how pricesdetermined outside the grape and wine markets influence these markets Becausethe model is in partial equilibrium to depict the impacts of changes in internationalcompetitiveness outside price changes need to be imposed as shocks on the modelThe price of intermediate inputs other than grapes and wine shown in equations (1)and (2) is set equal to the price of GDP (Pg

d ) multiplied by a shifter Fcd

Pcid frac14 Fc

dPgd eth13THORN

If no specific price observations are available the shifter Fcd remains exogenous and

unshocked with the change in price being determined by a shock to the price ofGDP If observations are available for specific input price movements the shifterFc

dbecomes endogenous with Pc

id now exogenous and shocked

Wid frac14 Fwd P

gd eth14THORN

Wage rates are treated similarly In equation (14) if the wage shifter Fwd is exogenous

changes in wage rates Wid are determined by changes in the price of GDP If wagerate data are available Fw

d becomes endogenous andwage rates are shocked directly

Pmcd frac14 Fm

d Pgd eth15THORN

The prices of trade and transport margins are also determined by the price of GDP ifthe shifter Fm

d in equation (15) is exogenous

Changes in international competitiveness depend on changes in relative pricelevels and changes in nominal exchange rates In equation (16) fR

s denotes real

Kym Anderson and Glyn Wittwer 249

exchange-rate movements relative to the US dollar

fRs frac14 Pg

s =frac12PgUSA

fs$ eth16THORN

In equation (16) the nominal exchange rate for the United States is alwaysunchanged because nominal and real exchange rates are expressed relative to theUS currency

Changes in international market conditions may have impacts in one direction onproducer prices as expressed in US dollars P0s

i and potentially in the opposite direc-tion in local currency units Hence we calculate real producer pricesP0s

iloc in localcurrency terms (ie the price most relevant to domestic producers)

P0siloc frac14 P0s

ifs=P

gs eth17THORN

To obtain real price changes in local currency terms we convert US dollar prices(P3 s

cd for the source-specific price and P3cd for the source-composite price) to reallocal currency prices (P3 s

cdloc and P3cdloc) using the consumer price index (CPI)(Pc

d ) as the deflator

P3 scdloc frac14 P3 s

cdfd=P

cd eth18THORN

and

P3cdloc frac14 P3cdfd=Pcd eth19THORN

References

Anderson K and Nelgen S (2011) Global Wine Markets 1961 to 2009 A StatisticalCompendium Adelaide University of Adelaide Press Also available as an e-book atwwwadelaideeduaupresstitlesglobal-wine

Anderson K Nelgen S and Pinilla V (2017) Global Wine Markets 1860 to 2016 AStatistical Compendium Adelaide University of Adelaide Press (forthcoming) Also tobe available as an e-book at wwwadelaideeduaupress

Anderson K and Pinilla V (with the assistance of A J Holmes) (2017)Annual Database ofGlobal Wine Markets 1835 to 2016 Wine Economics Research Centre University ofAdelaide to be posted in October at wwwadelaideeduauwine-econdatabases

Anderson K and Strutt A (2016) Impacts of Asiarsquos rise on African and Latin Americantrade Projections to 2030 World Economy 39(2) 172ndash194

Anderson K andWittwer G (2013) Modeling global wine markets to 2018 Exchange ratestaste changes and Chinarsquos import growth Journal of Wine Economics 8(2) 131ndash158

Anderson K and Wittwer G (2015) Asiarsquos evolving role in global wine markets ChinaEconomic Review 35 1ndash14

250 UK and Global Wine Markets by 2025 and Implications of Brexit

Armington P A (1969) A theory of demand for products distinguished by place of produc-tion IMF Staff Papers 16 159ndash178

Baldwin R E (ed) (2016) Brexit Beckons Thinking Ahead by Leading Economists LondonCentre for Economic Policy Research

Campos N F (2016) Lousy experts Looking back at the ex ante estimates of the costs ofBrexit In R E Baldwin (ed) Brexit Beckons Thinking Ahead by Leading Economists(ch 3) London Centre for Economic Policy Research

Dixon P Parmenter B Sutton J and Vincent D (1982) ORANI A Multisectoral Modelof the Australian Economy Contributions to Economic Analysis 142 Amsterdam North-Holland

European Commission (2015) Trade for All Towards a More Responsible Trade andInvestment Policy Brussels European Commission

Harrison J Horridge M Jerie M and Pearson K (2014) GEMPACK ManualMelbourne GEMPACK Software

HM Treasury (2017) Spring Budget 2017 London HM TreasuryHolmes A J and Anderson K (2017a)Annual Database of National Beverage Consumption

Volumes and Expenditures 1950 to 2015 Wine Economics Research Centre University ofAdelaide posted at wwwadelaideeduauwine-econdatabases

Holmes A J and Anderson K (2017b) Convergence in National Alcohol ConsumptionPatterns New Global Indicators Journal of Wine Economics 12(2) 117ndash148

Ludington C C (2013) The Politics of Wine in Britain A New Cultural History BasingstokeUK Palgrave Macmillan

Office of National Statistics (2017) Analysis of real earnings London Office of NationalStatistics Available at httpswwwonsgovukemploymentandlabourmarketpeopleinworkearningsandworkinghoursarticlessupplementaryanalysisofaverageweeklyearningslatest

Rollo J Borchert I Dawar K Holmes P and Winters L A (2016) The World TradeOrganisation A safety net for a post-Brexit UK trade policy Briefing Paper 1UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-1-final-1pdf

Smith A (2017) Brexit Hard truths and hard choices UKTPO blog 19 June Available athttpsblogssussexacukuktpo20170619brexit-hard-truths-and-hard-choices

Swinbank A (2017) World trade rules and the policy options for British agriculture post-Brexit Briefing Paper 7 UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-7pdf

Viner J (1950) The Customs Union Issue New York Carnegie Endowment for InternationalPeace

Wine Australia (2015) Export Market Guide European Union Adelaide Wine AustraliaWittwer G Berger N and Anderson K (2003) A model of the worldrsquos wine markets

Economic Modelling 20(3) 487ndash506

Kym Anderson and Glyn Wittwer 251

Page 9: U.K. and Global Wine Markets by 2025, and Implications of ... · U.K. and Global Wine Markets by 2025, and Implications of Brexit* Kym Andersona and Glyn Wittwerb Abstract The United

Figure 4

UK Shares in Wine Exports of Key Wine-Exporting Countries 2010ndash2014 ()

Fig4-BW

onlin

eBW

inprint

Source Compiled from data in Anderson and Pinilla (2017)

Table 3Import Duties Excise Duties and VAT Affecting Consumer Prices of Wine and Other Alcohol

in the United Kingdom April 1 2017

VAT 20

MFN import duties on wine pound per liter-bottled still wine lt13 alc 0114-bottled still wine 13ndash15 alc 0134-bottled still wine 15ndash18 alc 0162-bulk still wine lt13 alc 0086-bulk still wine 13ndash15 alc 0105-bulk still wine 15ndash18 alc 0134-sparkling wine 0278Weighted averagea 0130

Excise duties on alcohol pound per liter-still wine lt15 alc 2887-still wine 15ndash22 alc 3848-sparkling wine 55ndash85 alc 2795-sparkling wine 85ndash15 alc 3697-spirits (assumed 40 alc) 11551-beer (assumed 5 alc) 0954

VAT on alcohol 20

a Assumes that one-third of UK wine imports arrives in bulk that half has less than 13 alcohol and that one-tenth of imports is sparkling

Sources HM Revenue httpswwwgovukgovernmentpublicationsalcohol-duty-rate-changes for excise duties and VAT accessed April 92017 and Wine Australia (2015) for import duties converted at the 2016 average exchange rate of pound0740634 per euro

Kym Anderson and Glyn Wittwer 229

as to restore investor and consumer confidence in the United Kingdomrsquos economyThe size of those impacts is uncertain of course (see Baldwin 2016 and especiallyCampos 2016) so we consider a range of possibilities below

Those assumed adverse macroeconomic effects will add to the initial impact ofaltered wine tariffs on aggregate wine consumption in the United Kingdom andhence on its bilateral trades in wine They will make the loss of sales to theUnited Kingdom by EU (and Chilean and South African) suppliers greater thanwould otherwise be the case And they will reduce the likelihood that other countriesrsquosales of wine in the United Kingdom will be higher than in the baseline Indeed themacroeconomic effects could outweigh the trade-diverting effects so that even coun-tries that are currently discriminated against by the EU28rsquos wine-trade policy may beworse off because of Brexit

IV Global Wine Markets Model and Database

Our model of the worldrsquos wine markets first published by Wittwer Berger andAnderson (2003) and revised by Anderson and Wittwer (2013) is summarized inthe Appendix It disaggregates wine markets into four types namely nonpremiumcommercial-premium and superpremium still wines and sparkling wines2 There aretwo types of grapes premium and nonpremium Nonpremium wine uses nonpre-mium grapes exclusively superpremium wines use premium grapes exclusivelyand commercial-premium and sparkling wines use both types of grapes to varyingextents across countries The world is divided into 44 individual nations and 7 com-posite geographic regions that capture all other countries

The modelrsquos database is calibrated to 2014 based on the comprehensive wine-market volume and value data and trade and excise-tax data provided inAnderson and Pinilla (2017) and in Anderson Nelgen and Pinilla (2017) It is pro-jected assuming that aggregate national consumption population and realexchange rates change between 2014 and 2025 to the extent shown in AppendixTable 13 The Brexit alternatives to that baseline also are projected to 2025

Concerning preferences we assume a continued considerable swing toward allwine types in China and a swing away from nonpremium wines in all other countriesuntil 2025

In our baseline scenario grape- and wine-industry total-factor productivity areassumed to grow at 1 per year everywhere while grape- and wine-industry

2Commercial-premium still wines are defined byAnderson Nelgen and Pinilla (2017) to cost between US$250 and US$750 per liter pretax at a countryrsquos border or wholesale3The real exchange rate changes over the projection period are the changes expected in the nominal valueof country irsquos currency relative to the US dollar times the expected ratio of the GDP deflator for theUnited States versus that for country i

230 UK and Global Wine Markets by 2025 and Implications of Brexit

capital is assumed to grow net of depreciation at 15 per year in China but zero else-where (consistent with the almost-zero growth in global wine production and con-sumption over the past two decades)

Two alternative scenarios are considered (ldquolargerdquo and ldquosmallrdquo) for the initialimpact of Brexit to provide a range of results We assume that because of theUnited Kingdomrsquos decision to exit the European Union the rate of UK realgross domestic product (GDP) growth is only one-third or two-thirds as fast overthe projection period (09 or 18 per year instead of 26) and the Britishpound will be 20 or 10 lower in real terms than in our modelrsquos core baselineprojection4 In both alternative scenarios we assume that the United Kingdomapplies the European Unionrsquos external tariffs on wine from March 30 2019 atthe end of the two-year period following the United Kingdomrsquos formal triggeringof Article 50 (see Table 3)

We then consider a subsequent impact of Brexit presumed to result from negoti-ating signing implementing and responding to an FTA with the EU27 by 2025(Chile and South Africa currently have preferential access to EU wine marketsbut we continue to assume in this scenario that the United Kingdom does not imple-ment new bilateral FTAs with them or others in our time frame)5 This subsequentscenario assumes the British pound returns to what it would have been in 2025 in theabsence of Brexit and that real incomes regain two-thirds of the value they wouldhave had without Brexit as compared with our ldquosmallrdquo initial Brexit scenario

This global model has supply-and-demand equations and hence quantitiesprices and price elasticities for each of the grape and wine products and for asingle composite of all other products in each country Income elasticities ofdemand also exist for each final product Grapes are assumed to not be traded inter-nationally but other products are exported and imported Each market is assumed tohave been in equilibrium before any shock and to find a new market-clearingoutcome following any exogenously introduced shock

4The nominal price of the British pound in US dollars in the fortnight following the Brexit vote on June23 2016 dropped 13 to US$130 and a year later the pound sat at the same rate having dropped to anaverage of $124 (an 18 devaluation) between October 2016 and April 2017 Our choice of a low of 10and a high of 20 aims to capture future possible rates while uncertainties remain The average real wagein the United Kingdom fell in the first half of 2017 and projected real GDP growth during 2018ndash2020 hasbeen revised down to 17 in the United Kingdomrsquos latest budget (HM Treasury 2017)5South Africa currently has duty-free access to the EU28 for just 50 megaliters (ML) of wine per yearbeyond which the MFN tariff rate applies South Africa currently exports around 320 ML to theEuropean Union one-third of which initially goes to the United Kingdom How that quota of 50 MLis divided between the United Kingdom and the EU27 is subject to future negotiation (Rollo et al2016 Swinbank 2017) In the following discussion we assume none of it is accepted by the UnitedKingdom so that all South African wine imports pay the United Kingdomrsquos MFN tariff under Brexitbut altering that assumption makes very little difference to our results for any countryrsquos trade exceptSouth Africarsquos

Kym Anderson and Glyn Wittwer 231

V Projecting Global Wine Markets to 2025

Global wine production and exports are projected in the baseline from 2014 to 2025consistent with past trends The modelrsquos global volume of production (and consump-tion) rises little over that 11-year period (9) made up of a 6 decline in nonpre-mium wine and a one-sixth rise in commercial-premium and superpremium wineIn real (2014 US$) value though global wine output and consumption increaseby about 50 in total and 60 in the two premium categories The internationaltrade projections are similar although a little larger with the share of global wineproduction exported (equivalent to the share of global consumption imported)rising 2 percentage points between 2014 and 2025

The baseline projection does not greatly alter the 2014 shares of various countriesin global wine production apart from China because we assume vineyard expansionthere is faster than elsewhere6 In value terms this assumption means that Chinamoves from fifth to fourth by 2025 behind France the United States and ItalySpain remains barely ahead of Australia and they and Germany take the nextthree places (Figure 5(a)) In total wine-production volume terms China movesfrom sixth to fifth place and Argentina drops from fifth to eighth (and fromeighth to ninth in value terms)

When their products are subdivided into fine wines (superpremium still plus spark-ing) commercial-premium wines and nonpremium wines France and the UnitedStates retain the highest two places on the global ladder for fine-wine productionand Spain and Italy retain the top two places for nonpremium wine As for commer-cial-premium wine production (defined to cost between US$250 and US$750 perliter pretax at a countryrsquos wholesale level or national border) Italy retains the topranking over our projections period but at least in terms of value China challengesFrance for second place

The country rankings by projected value of total wine consumption change some-what more than those for production by 2025 with China taking second place afterthe United States ahead of France and Germany and then the United Kingdomslightly overtaking Italy to slip into fifth place (Figure 5(b)) The United StatesFrance and Germany retain the top three rankings for consuming fine wine butCanada slightly overtakes Italy for fourth place in terms of value at least In thecase of commercial-wine consumption China strengthens its number-one positionahead of the United States and the United Kingdom does likewise vis-agrave-visGermany for third place

As for the projected changes in consumption volumes China is projected to dom-inate the increase in aggregate although the United States is projected to lead the

6 In fact Chinarsquos wine production fell steadily between 2012 and 2016 by a total of one-sixth so Chinarsquoswine imports may grow faster in practice than in this baseline projection

232 UK and Global Wine Markets by 2025 and Implications of Brexit

Figure 5

Value of Wine Production and Consumption in Key Countries 2014 and Projected Baseline for2025 (2014 US$ Million at WineryWholesale Pretax Prices)

Fig5-BW

onlin

eBW

inprint

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 233

increase in consumption of fine wine In Western Europe and in the SouthernHemispherersquos New World countries fine wines are projected to substitute for com-mercial wines (defined as the sum of commercial-premium and nonpremiumwines) with almost no change in total wine consumption Sub-Saharan Africa isthe next region that is projected to take off with its growth accounting for morethan one-third of the rest of the worldrsquos increase in volume consumed

Those differences in production versus consumption rankings are reflected ininternational trade Figure 6 shows that France Italy and Spain remain the threedominant exporters of wine in aggregate value but the rankings of the next fewchange with Australia being slightly ahead of Chile and the United StatesGermany and New Zealand being nearly tied for sixth place in value termsFrance and Italy are even more dominant in fine-wine exports and remain so by2025 while Italy outranks France in the commercial-premium export categoryand Spain outranks Italy Australia and Chile in the nonpremium export class

Among the importers the United States and the United Kingdom are projected tocontinue to hold the first two places in 2025 in value terms but China moves intothird place slightly ahead of Germany followed well behind by Canada HongKong Belgium-Luxembourg the Netherlands and Japan (Figure 6(b)) OtherAfrica (excluding South Africa) is projected to experience the largest increase inimports among all the other regions followed by Other Asia which becomes asbig as Germany in value terms (Figure 7(a)) In terms of total volume of wineimports Germany and the United Kingdom held the top two shares in 2014 butby 2025 the United Kingdom is projected to be well ahead of Germany (Figure 7(b)) However this projection ignores the effects of Brexit to which we now turn

VI How Might Wine Markets Be Affected by the United Kingdomrsquos Exitfrom the European Union

As mentioned earlier for our two alternative scenarios to capture the initial effects ofBrexit (ldquolargerdquo and ldquosmallrdquo) we assume that following the United Kingdomrsquos exitfrom the European Union the United Kingdomrsquos rate of economic growth wouldbe only one-third or two-thirds as fast for the period to 2025 the British poundwould be 20 or 10 lower in real terms than in our modelrsquos baseline projectionand the United Kingdom would apply the European Unionrsquos external tariff onwine to imports from EU member countries (as part of establishing MFN ratesvia the WTO to then start new bilateral FTA negotiations) In these initial scenarioswe assume the United Kingdom does not implement any new FTAs particularlywith the EU27 Chile and South Africa Following the discussion of those initialresults we present the results of a subsequent scenario that includes the implemen-tation of an FTA between the United Kingdom and the EU27 We assume these aretwo discrete steps with no agreed arrangements to smooth the transition betweenthem

234 UK and Global Wine Markets by 2025 and Implications of Brexit

Figure 6

Value of Wine Exports and Imports Key Wine-Trading Countries 2014 and Projected Baselinefor 2025 (2014 US$ million)

Fig6-BW

onlin

eBW

inprint

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 235

Figure 7

National Shares of Global Wine Import Value and Volume 2014 and Projected Baselinefor 2025 ()

Fig7-BW

onlin

eBW

inprint

Source Authorsrsquo model results

236 UK and Global Wine Markets by 2025 and Implications of Brexit

A Initial Impact of Brexit

Because the British pound dropped by one-sixth against the US dollar in the fourmonths following the Brexit vote in June 2016 and because the UnitedKingdomrsquos average real wage fell 1 between November 2016 and April 2017(according to the latest available data see Office of National Statistics 2017) weviewed our ldquolargerdquo scenario as more likely soon after the Brexit vote But wecompare those results with ones from our ldquosmallrdquo initial-impact scenario becauseit has become more likely following the June 2017 general election that ldquosofterrdquooptions will be considered by the new government Generally the results areabout half the size in the ldquosmallrdquo scenario with the exception of the bilateraltrade effects To show the sensitivity of results to our assumptions we point outthe differences when the ldquosmallrdquo results are not close to half the results shown forthe ldquolargerdquo scenario

Before turning to those results we first present a ldquofree-traderdquo scenario whichassumes the United Kingdom chooses the radical option of becoming the HongKong of Europe Unlikely as this scenario is the results provide assurance thatour global wine model generates the standard types of effects of exiting a customsunion In this ldquofree-traderdquo scenario the United Kingdom is assumed to move tozero tariffs on all wine imports Thus the current preferences on EU Chileanand South African wine imports disappear as all other wine exporters enjoy thesame free access to UK wine markets We assume in this scenario that Brexit hasno impact on the United Kingdomrsquos real income or consumption growth rates by2025 and that by then the British pound has returned to its 2014 value relative toother currencies (similar to the real exchange rate assumed in our baseline scenarioto 2025 see Appendix Table A1)7 Under these assumptions EU Chilean andSouth African wine exports to the United Kingdom in 2025 are 003 lower thanin the no-Brexit baseline and UK imports from other wine exporters are 025higher but the United Kingdomrsquos total wine imports are hardly any different(because local wine prices fall by only 05 and incomes are unchanged) In this sce-nario UK consumers expand their volume of consumption by just 05 while theEU27 exports more and other countries export less to the rest of the world such thatworld wine exports are almost unchanged In short all these effects are in the direc-tion that comparative static-customs-union theory would predict but they are smallbecause we assume the United Kingdomrsquos real income and currency are the same asin our baseline for 2025 as described in the previous section

In the ldquolargerdquo alternate initial scenario involving Brexit as compared with thebaseline scenario to 2025 the consumer price of wine in 2025 is 22 higher in the

7If this scenario were seen as a credible long-run trade policy it would encourage more investment thatwould eventually raise UK incomes but not without major structural changes that would initiallydisrupt the economy Implicitly we assume in our free-trade scenario that these two opposite impactson real UK incomes and the British pound exactly offset each other by 2025

Kym Anderson and Glyn Wittwer 237

United Kingdom in local currency terms (20 because of real depreciation of theBritish pound 4 because of new tariffs on EU Chilean and South Africanwines and ndash2 because of slower UK income growth) The volume of UKwine consumption is 28 lower 16 because of slower UK economic growth7 because of real depreciation of the British pound and 5 because of newtariffs Superpremium still-wine sales are the most affected dropping by two-fifths while sparkling and commercial-premium wines drop a bit less than one-quarter Because the average price rises by more than the decline in volume soldthe aggregate value of UK sales even in local currency terms falls under thisldquolargerdquo Brexit scenario Under the ldquosmallrdquo Brexit scenario the consumer price ofwine in 2025 is 11 higher in the United Kingdom and its volume of wine consump-tion is 17 lower

The volume of projected UK imports in 2025 is 427 ML or nearly one-quarterless in the ldquolargerdquo scenario than in the baseline scenario comprising 58ML less spar-kling wine 31 ML less superpremium still wine and 339 ML less commercial-premium wine World imports are lower by just 239 ML because imports byother countries are 189 ML higher in response to the lower international wineprices in this scenario In value terms UK imports are $175 billion (or 27)lower in 2025 because of ldquolargerdquo Brexit $113 billion because of lower incomes$038 billion because of the fall in the British pound and $014 billion because ofthe rise in wine-import tariffs (Table 4) These aggregate trade impacts are a littlemore than half this size under the ldquosmallrdquo Brexit scenario

Despite the import levels falling because of raised import tariffs domestic con-sumption of all three quality categories of UK-produced wine is lower with thanwithout Brexit because of shrunken demand for all wines resulting from loweredUK incomes and raised local prices because of devaluation of the British poundThe British poundrsquos devaluation does make it easier for the United Kingdom tosell wines abroad though The countryrsquos exports are 7 ML or nearly 5 higherin 2025 in the ldquolargerdquo Brexit scenario and UK production is 3 higher ThoseUK exports (or re-exports of imported bulk wine after it is bottled in the UnitedKingdom) that go to EU27 countries are reduced though because of the tariffnow imposed at the new EU border

Without Brexit the United Kingdomrsquos shares of global wine imports are slightlyhigher in volume terms in 2025 than in 2010ndash2015 but 2 percentage points lower invalue terms thanks to East Asiarsquos expanding demand for imports of premium winesWith ldquolargerdquo Brexit however that value share is an additional 2 percentage pointslower and the volume share is almost 5 points lower (Figure 8) The net effects ofthese impacts on global trade are shown in Figure 9 Most of the initial tradeeffect of Brexit is a large decline in net imports of wine by the United Kingdomwith very little offsetting positive effect on trade in the rest of the world Theldquosmallrdquo Brexit numbers are a bit more than half of those for ldquolargerdquo Brexit

238 UK and Global Wine Markets by 2025 and Implications of Brexit

Table4

Difference

in2025

Projected

VolumeandValue

ofWineIm

portsb

ytheUnitedKingdom

andtheResto

fthe

World

asaConsequence

oftheInitial

BrexitS

hock

(MLand2014US$

Million

ldquoLargerdquoSc

enario)

Volum

e(M

L)

Value

(US$

million)

NP+CPa

SuperPrb

Sparkling

TOTAL

NP+CPa

SuperPrb

Sparkling

TOTAL

ΔUK

impo

rtsdu

eto

Low

erincomes

minus198

minus20

minus29

minus247

58minus644

minus253

minus234

minus1131

65Low

erpo

und

minus70

minus10

minus14

minus93

22minus248

minus127

minus102

minus476

27Highertariffs

minus71

minus1minus16

minus87

20minus110

minus8minus2

4minus143

8TOTA

Lminus339

minus31

minus58

minus427

100

minus1001

minus388

minus360

minus1750

100

difffrom

base

2332

3325

2432

3227

of

totalc

uts

797

14100

5722

21100

ΔROW

netim

ports

143

2125

189

230

143

minus181

192

ΔWORLD

TRADE

minus195

minus10

minus34

minus239

minus763

minus246

minus543

minus1552

aNon

prem

ium

plus

commercial-premium

still

wines

bSu

perpremium

still

wines

SourceA

utho

rsrsquomod

elresults

Kym Anderson and Glyn Wittwer 239

The aggregate effect of ldquolargerdquo Brexit on the market shares of various wine-export-ing countries in the United Kingdom is almost indiscernible even with one decimalpoint The projected 2025 shares are quite different from the actual 2014 shares forseveral countries They are much smaller in 2025 for South Africa Australia andNew Zealand (and the United States in volume terms) and are much larger involume for Spain and in value for Italy These results occur because wine-exportingcountries benefit differentially from the varying rates of growth in net importdemand for wine in non-UK countries over this projection period The most impor-tant projected changes are the increase in the real value of annual wine importsbetween 2014 and 2025 by China (200 or $3 billion) Other Asia (110 or $22billion) and Africa (270 or $16 billion) More than half of Australiarsquos increasein annual exports from 2014 to 2025 go to Asia and more than half of SouthAfricarsquos increase in exports go to Other Africa

Table 5 reveals that ldquolargerdquo Brexit lowers EU Chilean and South African wineexports by 150 ML or US$12 billion in the case of the European Union with someof their exports diverted from the United Kingdom to the EU27 and other marketsin competition with New World exporters Although the United States Australiaand Argentina sell only a little less to the United Kingdom they also sell less toother countries For Chile and South Africa which lose their preferential access toUK (but not to EU27) markets in this Brexit scenario some exports are redirected

Figure 8

UK Shares of World Wine Imports 2010ndash2015 and Projected to 2025 without and with theldquoLargerdquo Initial Brexit Shock ()

Fig8-BW

onlin

eBW

inprint

Sources Anderson and Pinilla (2017) and authorsrsquo model results

240 UK and Global Wine Markets by 2025 and Implications of Brexit

from the United Kingdom to EU27 countries ndash but again they export less overallGlobal wine trade in 2025 is 240 ML (19) or $18 billion (35) less under thisldquolargerdquo Brexit scenario The percentage by which wine exportersrsquo trade shrinks isgreater for values than for volumes because of changes in relative prices of differ-ent-quality wines Those differences are shown in the numbers in parentheses inTable 5

Three other points are worth making about Table 5 One is that Australia sellsslightly more to the United Kingdom in the ldquosmallrdquo Brexit scenario rather thanslightly less as in the ldquolargerdquo Brexit case Evidently the negative income and price(devaluation) effects do not more than offset the positive trade-diverting effect onAustralian exports to the United Kingdom of removing preferences in the ldquosmallrdquoscenario Second New Zealand sells slightly more to non-UK countries underBrexit despite greater competition from the EU27 Chile and South Africa Thisanomaly is due to changes in the relative prices of different qualities of wine inglobal wine markets bearing in mind that New Zealand has the worldrsquos highestaverage price for still-wine exports And third the value (but not the volume) ofexports of ldquoOtherrdquo countries to markets other than the United Kingdom arehigher under Brexit This result too is due to changes in the relative prices of differ-ent qualities of wine in global wine markets

Figure 9

Difference in 2025 Wine Import Volumes and Values as a Result of the ldquoLargerdquo Initial BrexitShock (ML and US$ Million in 2014 US dollars)

Fig9-BW

onlin

eBW

inprint

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 241

B Subsequent Impact of Brexit from a UK-EU27 FTA

The next-most-likely step in the Brexit process is for the United Kingdom to nego-tiate a new trade arrangement with the EU27 We therefore assume that a UK-EU27FTAwith free bilateral wine trade is implemented and adjusted to by 2025 and thatprogress toward that end occurs soon enough that the adverse macroeconomicshocks from the initial impact of uncertainty over the Brexit process are confinedto those assumed in the ldquosmallrdquo scenario outlined above In this subsequent scenariowe assume that the British pound returns to the value it would have reached in 2025in the absence of Brexit (ie reversing the 10 devaluation assumed in the ldquosmallrdquoinitial Brexit scenario) and that real incomes in the United Kingdom are 8 higher

Table 5Difference in 2025 BilateralWine Import Volumes and Values fromKey Exporters by the United

Kingdom and the Rest of the World (RoW) as a Result of Initial Brexit Shock(ML and 2014US$ Million)a

(a) ldquolargerdquo scenario

Volume (ML) Value (2014US$ million)

UK RoW WORLD () UK RoW WORLD ()

EU27 minus287 136 minus150 (ndash17) minus1187 minus5 minus1192 (ndash31)Chile minus59 35 minus25 (ndash30) minus169 31 minus138 (ndash48)South Africa minus53 35 minus18 (ndash32) minus105 20 minus85 (ndash67)USA minus7 minus6 minus13 (ndash24) minus75 minus40 minus115 (ndash50)Australia minus4 minus3 minus7 (ndash09) minus25 minus65 minus90 (ndash30)Argentina minus3 minus9 minus12 (ndash48) minus16 minus39 minus55 (ndash52)New Zealand minus11 9 minus2 (ndash09) minus162 71 minus91 (ndash43)Others minus2 minus10 minus12 (ndash02) minus11 minus52 minus63 (ndash44)WORLD minus427 187 minus240 (ndash19) minus1750 minus79 minus1829 (ndash35)

(a) ldquosmallrdquo scenario

Volume (ML) Value (2014US$ million)

UK RoW WORLD () UK RoW WORLD ()

EU27 minus178 82 minus96 (ndash12) minus692 minus43 minus736 (ndash19)Chile minus46 28 minus18 (ndash24) minus128 36 minus91 (ndash32)South Africa minus43 29 minus14 (ndash42) minus82 23 minus59 (ndash47)USA 1 minus6 minus5 (ndash11) minus23 minus28 minus51 (ndash22)Australia 5 minus10 minus5 (ndash06) 19 minus56 minus38 (ndash13)Argentina 0 minus6 minus6 (ndash26) minus3 minus25 minus29 (ndash27)New Zealand minus5 4 minus1 (ndash06) minus80 34 minus46 (ndash22)Others 0 minus9 minus9 (ndash01) minus1 minus33 minus34 (ndash24)WORLD minus266 112 minus154 (ndash13) minus991 minus92 minus1083 (ndash21)

a Numbers in parentheses are the percentage difference between the Brexit and baseline scenarios for 2025 projected wine-import volumes orvalues by source

Source Authorsrsquo model results

242 UK and Global Wine Markets by 2025 and Implications of Brexit

in 2025 than in the ldquosmallrdquo scenario ndash that is they regain two-thirds of the value theywould have reached without Brexit-related uncertainty

This subsequent development in the Brexit process reverses most of but not allthe initial effects of Brexit by 2025 because of our assumption that the lost growthin the initial years of uncertainty following the Brexit vote are only partly recoveredby 2025 following the implementation of a UK-EU27 FTA Moreover the longer ittakes before this FTA is finalized and implemented the longer the estimated initialadverse effects persist and the larger the cumulative cost of Brexit to UK wine con-sumers and to grape and wine producers in wine-exporting countries

Table 6 summarizes the subsequent trade effects for 2025 It suggests that all butone-tenth of the loss in value of world trade in wine from the initial ldquosmallrdquo impact isrestored most of it because of a smaller reduction in the United Kingdomrsquos wineimports following the FTA Most of that improved outcome is because of recoveredimports from the EU27 commensurate with the latterrsquos high share of UK importsEven though Chile and South Africa are assumed in this scenario to have not yetsigned an FTA with the United Kingdom they export slightly more to the UnitedKingdom (and even more to all other countries) than in the initial ldquosmallrdquo scenario

Within the United Kingdom this FTA brings down the local currency consumerprice of wine by 9 largely offsetting the 11 rise in the initial ldquosmallrdquo Brexit sce-nario and it raises the volume consumed in the United Kingdom by 18 fully off-setting the 17 fall in the initial ldquosmallrdquo Brexit scenario

Table 6Difference in 2025 Bilateral Wine-Import Volumes and Values from Key Exporters by the

United Kingdom and the Rest of the World (RoW) as a Result of Implementing a UK-EU27FTA Following Initial ldquoSmallrdquo Brexit Shock (ML and 2014US$ Million Difference Relative to

Initial ldquoSmallrdquo Brexit Scenario)a

Volume (ML) Value (2014US$ million)

UK RoW WORLD (ldquosmallrdquo)a UK RoW WORLD (ldquosmallrdquo)a

EU27 212 minus112 100 (minus96) 750 0 750 (minus806)Chile 1 3 4 (minus18) 3 40 43 (minus92)South Africa 3 2 5 (minus14) 3 18 21 (minus60)USA 3 3 6 (minus5) 26 21 47 (minus50)Australia 1 1 2 (minus5) 8 38 46 (minus33)Argentina 1 4 5 (minus6) 4 20 24 (minus28)New Zealand 4 minus3 1 (minus1) 61 minus25 36 (minus46)Others 1 6 7 (minus9) 5 minus46 minus41 (92)WORLD 226 minus96 130 (minus154) 860 66 926 (minus1022)

a Numbers in parentheses are the world trade differences between the ldquosmallrdquo initial Brexit scenario and the baseline scenario copied fromcolumns 3 and 7 of Table 5(b)

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 243

In short this subsequent step in the Brexit process can be expected to restore muchof the initial adverse effects in the United Kingdom of the Brexit vote by 2025 Thewine exporters in the EU27 Chile and South Africa also lose less in this scenarioas compared with the initial ldquosmallrdquo scenario while other key exporters sell nearlyas much in 2025 in this scenario as they are projected to in the baseline ndash that isthey are beneficiaries of the reduced discrimination in global wine markets in thiscase where Chile and South Africa no longer enjoy preferential access to the UKmarket (and even though EU27 producers are again allowed to do so) Keep inmind though that although the estimated losses in the initial scenarios are reportedjust for 2025 they are expected to be felt throughout the preceding years of uncertaintyTherefore the longer it is before uncertainty abates and the United Kingdomrsquos currentpreferential trading arrangements are replaced by newones such as a UK-EU27 FTAthe higher the cumulative cost of the Brexit vote to the wine trade

VII Caveats and Conclusions

The above Brexit simulations are just a few of many scenarios that could be modeledObvious additional ones could also assume that FTAs are reached between theUnited Kingdom and other trading partners including Chile and South Africaand such countries as Australia and New Zealand Some of these other countriesare already exploring the ECrsquos invitation (see European Commission 2015) to nego-tiate an FTA with what will be the EU27 The sequence in which FTAs are signedand the speed with which they are implemented will matter (as was also the casewith the sequential signing over the past decade of bilateral FTAs with NortheastAsian countries by Chile Australia and New Zealand see Anderson andWittwer 2015) Even if the United Kingdom were able to sign additional bilateralFTAs and begin implementing them before 2025 it would have little effect on theabove results (because wine tariffs are such a minor contributor to them) unlessthose FTAs were to accelerate the United Kingdomrsquos economic growth and therise in the British poundrsquos value before 2025

We assume above that no changes are made to alcohol excise duties in the UnitedKingdom following Brexit even though they are scheduled to be progressively raisedwith inflation and may be raised even more for wine relative to spirits to offset theopposite effects of Brexit on those two domestic industries Nor do we make anyallowance for UK increases in consumer wine prices that may be needed to coverthe higher cost of clearing customs on imports from EU27 countries (which maybe more or less offset by increased smuggling and duty-free purchasing)Importers of fine wines including individuals who buy Bordeaux en primeur maywell reduce their demand because of concerns that delays on docks will affect thequality of their wines or because of greater currency uncertainty

To summarize Brexit is costly initially to UK consumers of wine (and of manyother tradable products) because the domestic retail price in local currency tax-

244 UK and Global Wine Markets by 2025 and Implications of Brexit

inclusive terms is 22 higher than otherwise in 2025 and the volume of wine con-sumed domestically is 28 lower in the ldquolargerdquo Brexit scenario (or 11 higher pricesand 17 lower quantities in the ldquosmallrdquo case) by 2025 unless a UK-EU27 FTA isdeveloped before then Even if such an FTA does get signed ratified by all 28 par-liaments and implemented by 2025 the slower income growth in the interim meansa smaller UK wine market in 2025 than would otherwise have been the case Thevolume reduction is a blow to many participants in UK wine-bottling transportingstoring wholesaling and retailing businesses in addition to restaurants and pubsVery little of that initial impact occurs because of higher import tariffs most impor-tant is the assumed fall in UK real incomes

The small but growing number of local UK vignerons (who supply less than 05of all domestic-wine sales and less than 4 of sparkling-wine sales) is initially pro-jected to sell less wine domestically because of reduced demand and to export only alittle extra abroad while the British pound is devalued Overall their production isonly 3 higher in 2025 even in the ldquolargerdquo Brexit scenario (not taking intoaccount the fact that their casual labor is likely to be more expensive under Brexitbecause of tighter restrictions on immigration) and their production diminishes ifand when a UK-EU27 FTA is in place

Even with a UK-EU27 FTA in place EU27 wine exporters export US$56 millionless wine in 2025 thanks to Brexit Chile and South Africa export $88 million lesswine and wine exports from the rest of the world are little different in aggregatealthough slightly larger from Australia

Clearly there will be great uncertainly for some time over the possible policy out-comes to flow from Brexit and their consequent sequential impacts on UK house-hold disposable incomes foreign exchange rates and bilateral wine tariffsMeanwhile the above projections under explicit assumptions provide some idea ofhow wine markets might be affected by the most-likely first two stages of theBrexit process (agreeing on a new tariff schedule at the WTO and agreeing to andimplementing a UK-EU27 FTA) In particular they make clear that nontrivialinitial impacts could affect the domestic wine market impacts that are likely to belarger than just the direct impact of changes in bilateral tariffs If the UnitedKingdom succeeds in getting countries to agree to transition arrangements thatdelay the changes in tariffs until new FTAs are signed and ratified the initialeffects will be less dramatic than in our first scenarios but it remains to be seenwhether any such agreements can be reached In any event the net effect of Brexiton the welfare of the worldrsquos consumers and producers of wine as a whole will benegative not just initially but permanently unless new trade policy commitmentsby the United Kingdom with major wine-exporting countries are sufficiently moreliberal than current arrangements

Kym Anderson and Glyn Wittwer 245

Appendix Key Equations in the Global Wine Markets Model

Amodel of the worldrsquos wine markets was first published byWittwer et al (2003) andmuch improved by Anderson and Wittwer (2013) Several significant enhancementshave been made to that original model Wine types have been disaggregated intomore types ndash namely nonpremium (including generic bulk) commercial-premiumand superpremium still wines and sparkling wines8 As in the original modelthere are two types of grapes premium and nonpremium Nonpremium wines usenonpremium grapes exclusively superpremium wines use mainly premium grapesand commercial-premium and sparkling wines use both types of grapes In theregional dimension the number of countries and country groups has expandedfrom 10 in the original model to 51 44 individual nations and 7 compositeregions The modelrsquos database is calibrated to 2014 for this paper based on datain Anderson et al (2017) trade data for which are downloaded from httpscomtradeunorg The modelrsquos supply-and-demand equations are based on the orig-inal ORANI modelrsquos theory (see Dixon Parmenter Sutton and Vincent 1982) andthe model is implemented using GEMPACK software (Harrison Horridge Jerieand Pearson 2014)

An enhancement of importance to the present study is the inclusion of exchange-rate variables in the model which allow a distinction between price impacts asobserved in local currency units and those observed in 2014 US dollars

In the model the grape and wine sectors minimize costs of intermediate inputssubject to weak constant elasticity of substitution (CES) substitutability betweeninputs By assumption no intermediate inputs are imported from other countries9Hence

Xcid frac14 f ethX1id CESfrac12P

cid=P1id $THORN eth1THORN

P1id X1id frac14X

c

Xcid P

cid eth2THORN

whereXcid is the quantity demanded of commodity c by grape or wine industry i in

region d Pcid is the corresponding price and X1id and P1id are the respective inter-

mediate composite quantities and prices

Two primary factors are employed in the sector labor (the quantity of which isendogenous with perfectly elastic supply) and capital Capital is usually treated asexogenous in quantity with rates of return bearing all the adjustment in thevarious scenarios This reflects the fact that grapes (a perennial crop) and wine-

8Commercial-premium still wines are defined by Anderson and Nelgen (2011) to be those priced betweenUS$250 and $750 per litre pre-tax at a countryrsquos border or wholesale9An exception concerns bulk wine imports used mainly in the bottling of commercial premium wine in theUK USA Canada Japan and Germany

246 UK and Global Wine Markets by 2025 and Implications of Brexit

plant capacity adjust slowly to market signals

Lid frac14 f ethFid CESfrac12Wid=PFid $THORN eth3THORN

Kid frac14 f ethFid CESethRid=PFid $THORN eth4THORN

PFid Fid frac14 Lid Wid thorn Kid Rid eth5THORN

Grape and wine producers are assumed to minimize costs subject to CES substitu-tion between capital and labor Equations (3) to (5) show primary factor demandsfor the labor composite Lid and capital Kid subject to a composite factor demandFid by industry i in region d The factor prices are Wid for labor Rid for capitalrentals and PFid for composite factor prices

Appendix Table A1Cumulative Consumption and Population Growth Rates and Changes in the Real Exchange Rate

(RER) Relative to the US Dollar 2014 to 2025 without Brexit ()

Aggregateconsumption Poprsquon RER

Aggregateconsumption Poprsquon RER

France 18 4 minus11 Australia 35 11 minus17Italy 11 2 minus9 New Zealand 32 9 minus26Portugal 14 0 minus9 Canada 27 8 minus18Spain 26 8 minus9 United States 31 8 0Austria 19 4 minus7 Argentina 7 10 109Belgium 20 7 minus9 Brazil 16 8 minus29Denmark 22 2 minus9 Chile 55 8 minus2Finland 21 3 minus7 Mexico 42 12 minus8Germany 14 minus2 minus11 Uruguay 45 3 1Greece 22 minus1 minus14 Other Latin America 60 10 minus5Ireland 42 12 minus9 South Africa 36 12 minus1Netherlands 21 4 minus9 Turkey 50 8 20Sweden 24 9 minus13 North Africa 53 11 0Switzerland 18 8 minus6 Other Africa 109 18 84United Kingdom 32 6 1 Middle East 52 18 minus12Other West Europe 21 10 minus1 China 79 3 5Bulgaria 41 minus7 7 Hong Kong 42 3 2Croatia 20 minus2 minus1 India 134 13 17Georgia 35 0 23 Japan 11 minus3 minus24Hungary 25 minus3 minus11 Korea 38 1 minus9Moldova 49 minus11 13 Malaysia 62 15 minus16Romania 45 minus4 22 Philippines 75 18 7Russia 18 minus2 minus8 Singapore 44 21 minus22Ukraine 22 minus5 14 Taiwan 29 1 minus13Other East Europe 40 minus5 48 Thailand 47 3 minus9

Other Asia 99 10 10

Source Authorsrsquo compilation from projections by various international agencies and from global economy-wide modeling by Anderson andStrutt (2016)

Kym Anderson and Glyn Wittwer 247

The composite factor demand Fid is proportional to total output Qid subject to aprimary factor using technology Aid Hence

Fid frac14 Qid Aid eth6THORN

The perfectly competitive zero-pure-profit condition is that total revenue valued atthe output price P0s

i multiplied by Qid equals the total production cost

P0si Qid frac14

X

cPcid X1cid thorn

X

oWo

id Loid thorn Rid Kid eth7THORN

Household demands follow a linear expenditure system in each region We reducethe optimizing problem for household consumption of each commodity subjectto a budget constraint to equations describing subsistence and discretionarydemands Aggregate subsistence expenditure WSUBd depends only on consumerprices P3cd for each commodity and the number of households N as per-capita sub-sistence quantities XSUBcd subject to given preferences are constant

WSUBd frac14X

c

P3cd XSUBcd Nd eth8THORN

Discretionary expenditures for each commodity (the left-hand side of equation (9))are equal to the marginal budget share (βcd) of aggregate discretionary expenditureThis aggregate is the bracketed term on the right-hand side of equation (9) whereW3TOTd is aggregate nominal expenditure

P3cdethX3cd XSUBcd NdTHORN frac14 βcdethW3TOTd WSUBdTHORN eth9THORN

Because real aggregate consumption is usually exogenous in our partial equilibriumsimulations the linear expenditure system determines the consumption shares ofindividual final commodities (ie the five wine types plus a composite of all otherconsumption items) driven by changes in relative prices as faced by domestic con-sumers The income elasticity of demand for each commodity is equal to the mar-ginal budget share divided by the expenditure share This number varies from 05for nonpremium wine to 17 for superpremium still wine The income elasticity ofdemand for other consumption is very close to 10 because wine accounts for anaverage of only 03 of aggregate expenditures globally and no more than 11in any country (Anderson and Nelgen 2011 Table 166)

A feature of our revised model of world wine markets is the inclusion of nominalexchange rates These appear directly in the equation linking retail prices (P3scd ) toproducer prices by country of origin (P0s

c ) where c denotes the wine type

P3 scd frac14 P0s

cfd

fsTtarcd T

taxcd thorn Pm

cd eth10THORN

248 UK and Global Wine Markets by 2025 and Implications of Brexit

The exchange rates in the consuming and producing regions are fd and fs respec-tively expressed as local currency units per $US Ttar

cd is the power of the tariff in theconsuming region and Ttax

cd is the power of the domestic consumption (or excise) taxprior to any generic value-added or goods-and-services tax Pm

cd is the price of marginm assumed to be locally supplied nontradable and therefore unaffected by theexchange rate

A given level of consumption for wine type c (X3cd) is satisfied using theArmington (1969) assumption in which wines from different countries of originare imperfectly substitutable First domestic wine is imperfectly substitutable witha composite of imports

X3sscd frac14 f ethX3cd CESethP3sscd=P3cdTHORNTHORN ss frac14 domestic imports eth11THORN

and then imports by origin (X3scd ) are determined in a second CES equation

X3scd frac14 f ethX3ssfrac14importscd CESethP3scd=P3

ssfrac14importscd THORNTHORN eth12THORN

The model enables us to show how changes in international competitiveness affectthe worldrsquos wine markets A crucial part of this exercise is explaining how pricesdetermined outside the grape and wine markets influence these markets Becausethe model is in partial equilibrium to depict the impacts of changes in internationalcompetitiveness outside price changes need to be imposed as shocks on the modelThe price of intermediate inputs other than grapes and wine shown in equations (1)and (2) is set equal to the price of GDP (Pg

d ) multiplied by a shifter Fcd

Pcid frac14 Fc

dPgd eth13THORN

If no specific price observations are available the shifter Fcd remains exogenous and

unshocked with the change in price being determined by a shock to the price ofGDP If observations are available for specific input price movements the shifterFc

dbecomes endogenous with Pc

id now exogenous and shocked

Wid frac14 Fwd P

gd eth14THORN

Wage rates are treated similarly In equation (14) if the wage shifter Fwd is exogenous

changes in wage rates Wid are determined by changes in the price of GDP If wagerate data are available Fw

d becomes endogenous andwage rates are shocked directly

Pmcd frac14 Fm

d Pgd eth15THORN

The prices of trade and transport margins are also determined by the price of GDP ifthe shifter Fm

d in equation (15) is exogenous

Changes in international competitiveness depend on changes in relative pricelevels and changes in nominal exchange rates In equation (16) fR

s denotes real

Kym Anderson and Glyn Wittwer 249

exchange-rate movements relative to the US dollar

fRs frac14 Pg

s =frac12PgUSA

fs$ eth16THORN

In equation (16) the nominal exchange rate for the United States is alwaysunchanged because nominal and real exchange rates are expressed relative to theUS currency

Changes in international market conditions may have impacts in one direction onproducer prices as expressed in US dollars P0s

i and potentially in the opposite direc-tion in local currency units Hence we calculate real producer pricesP0s

iloc in localcurrency terms (ie the price most relevant to domestic producers)

P0siloc frac14 P0s

ifs=P

gs eth17THORN

To obtain real price changes in local currency terms we convert US dollar prices(P3 s

cd for the source-specific price and P3cd for the source-composite price) to reallocal currency prices (P3 s

cdloc and P3cdloc) using the consumer price index (CPI)(Pc

d ) as the deflator

P3 scdloc frac14 P3 s

cdfd=P

cd eth18THORN

and

P3cdloc frac14 P3cdfd=Pcd eth19THORN

References

Anderson K and Nelgen S (2011) Global Wine Markets 1961 to 2009 A StatisticalCompendium Adelaide University of Adelaide Press Also available as an e-book atwwwadelaideeduaupresstitlesglobal-wine

Anderson K Nelgen S and Pinilla V (2017) Global Wine Markets 1860 to 2016 AStatistical Compendium Adelaide University of Adelaide Press (forthcoming) Also tobe available as an e-book at wwwadelaideeduaupress

Anderson K and Pinilla V (with the assistance of A J Holmes) (2017)Annual Database ofGlobal Wine Markets 1835 to 2016 Wine Economics Research Centre University ofAdelaide to be posted in October at wwwadelaideeduauwine-econdatabases

Anderson K and Strutt A (2016) Impacts of Asiarsquos rise on African and Latin Americantrade Projections to 2030 World Economy 39(2) 172ndash194

Anderson K andWittwer G (2013) Modeling global wine markets to 2018 Exchange ratestaste changes and Chinarsquos import growth Journal of Wine Economics 8(2) 131ndash158

Anderson K and Wittwer G (2015) Asiarsquos evolving role in global wine markets ChinaEconomic Review 35 1ndash14

250 UK and Global Wine Markets by 2025 and Implications of Brexit

Armington P A (1969) A theory of demand for products distinguished by place of produc-tion IMF Staff Papers 16 159ndash178

Baldwin R E (ed) (2016) Brexit Beckons Thinking Ahead by Leading Economists LondonCentre for Economic Policy Research

Campos N F (2016) Lousy experts Looking back at the ex ante estimates of the costs ofBrexit In R E Baldwin (ed) Brexit Beckons Thinking Ahead by Leading Economists(ch 3) London Centre for Economic Policy Research

Dixon P Parmenter B Sutton J and Vincent D (1982) ORANI A Multisectoral Modelof the Australian Economy Contributions to Economic Analysis 142 Amsterdam North-Holland

European Commission (2015) Trade for All Towards a More Responsible Trade andInvestment Policy Brussels European Commission

Harrison J Horridge M Jerie M and Pearson K (2014) GEMPACK ManualMelbourne GEMPACK Software

HM Treasury (2017) Spring Budget 2017 London HM TreasuryHolmes A J and Anderson K (2017a)Annual Database of National Beverage Consumption

Volumes and Expenditures 1950 to 2015 Wine Economics Research Centre University ofAdelaide posted at wwwadelaideeduauwine-econdatabases

Holmes A J and Anderson K (2017b) Convergence in National Alcohol ConsumptionPatterns New Global Indicators Journal of Wine Economics 12(2) 117ndash148

Ludington C C (2013) The Politics of Wine in Britain A New Cultural History BasingstokeUK Palgrave Macmillan

Office of National Statistics (2017) Analysis of real earnings London Office of NationalStatistics Available at httpswwwonsgovukemploymentandlabourmarketpeopleinworkearningsandworkinghoursarticlessupplementaryanalysisofaverageweeklyearningslatest

Rollo J Borchert I Dawar K Holmes P and Winters L A (2016) The World TradeOrganisation A safety net for a post-Brexit UK trade policy Briefing Paper 1UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-1-final-1pdf

Smith A (2017) Brexit Hard truths and hard choices UKTPO blog 19 June Available athttpsblogssussexacukuktpo20170619brexit-hard-truths-and-hard-choices

Swinbank A (2017) World trade rules and the policy options for British agriculture post-Brexit Briefing Paper 7 UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-7pdf

Viner J (1950) The Customs Union Issue New York Carnegie Endowment for InternationalPeace

Wine Australia (2015) Export Market Guide European Union Adelaide Wine AustraliaWittwer G Berger N and Anderson K (2003) A model of the worldrsquos wine markets

Economic Modelling 20(3) 487ndash506

Kym Anderson and Glyn Wittwer 251

Page 10: U.K. and Global Wine Markets by 2025, and Implications of ... · U.K. and Global Wine Markets by 2025, and Implications of Brexit* Kym Andersona and Glyn Wittwerb Abstract The United

as to restore investor and consumer confidence in the United Kingdomrsquos economyThe size of those impacts is uncertain of course (see Baldwin 2016 and especiallyCampos 2016) so we consider a range of possibilities below

Those assumed adverse macroeconomic effects will add to the initial impact ofaltered wine tariffs on aggregate wine consumption in the United Kingdom andhence on its bilateral trades in wine They will make the loss of sales to theUnited Kingdom by EU (and Chilean and South African) suppliers greater thanwould otherwise be the case And they will reduce the likelihood that other countriesrsquosales of wine in the United Kingdom will be higher than in the baseline Indeed themacroeconomic effects could outweigh the trade-diverting effects so that even coun-tries that are currently discriminated against by the EU28rsquos wine-trade policy may beworse off because of Brexit

IV Global Wine Markets Model and Database

Our model of the worldrsquos wine markets first published by Wittwer Berger andAnderson (2003) and revised by Anderson and Wittwer (2013) is summarized inthe Appendix It disaggregates wine markets into four types namely nonpremiumcommercial-premium and superpremium still wines and sparkling wines2 There aretwo types of grapes premium and nonpremium Nonpremium wine uses nonpre-mium grapes exclusively superpremium wines use premium grapes exclusivelyand commercial-premium and sparkling wines use both types of grapes to varyingextents across countries The world is divided into 44 individual nations and 7 com-posite geographic regions that capture all other countries

The modelrsquos database is calibrated to 2014 based on the comprehensive wine-market volume and value data and trade and excise-tax data provided inAnderson and Pinilla (2017) and in Anderson Nelgen and Pinilla (2017) It is pro-jected assuming that aggregate national consumption population and realexchange rates change between 2014 and 2025 to the extent shown in AppendixTable 13 The Brexit alternatives to that baseline also are projected to 2025

Concerning preferences we assume a continued considerable swing toward allwine types in China and a swing away from nonpremium wines in all other countriesuntil 2025

In our baseline scenario grape- and wine-industry total-factor productivity areassumed to grow at 1 per year everywhere while grape- and wine-industry

2Commercial-premium still wines are defined byAnderson Nelgen and Pinilla (2017) to cost between US$250 and US$750 per liter pretax at a countryrsquos border or wholesale3The real exchange rate changes over the projection period are the changes expected in the nominal valueof country irsquos currency relative to the US dollar times the expected ratio of the GDP deflator for theUnited States versus that for country i

230 UK and Global Wine Markets by 2025 and Implications of Brexit

capital is assumed to grow net of depreciation at 15 per year in China but zero else-where (consistent with the almost-zero growth in global wine production and con-sumption over the past two decades)

Two alternative scenarios are considered (ldquolargerdquo and ldquosmallrdquo) for the initialimpact of Brexit to provide a range of results We assume that because of theUnited Kingdomrsquos decision to exit the European Union the rate of UK realgross domestic product (GDP) growth is only one-third or two-thirds as fast overthe projection period (09 or 18 per year instead of 26) and the Britishpound will be 20 or 10 lower in real terms than in our modelrsquos core baselineprojection4 In both alternative scenarios we assume that the United Kingdomapplies the European Unionrsquos external tariffs on wine from March 30 2019 atthe end of the two-year period following the United Kingdomrsquos formal triggeringof Article 50 (see Table 3)

We then consider a subsequent impact of Brexit presumed to result from negoti-ating signing implementing and responding to an FTA with the EU27 by 2025(Chile and South Africa currently have preferential access to EU wine marketsbut we continue to assume in this scenario that the United Kingdom does not imple-ment new bilateral FTAs with them or others in our time frame)5 This subsequentscenario assumes the British pound returns to what it would have been in 2025 in theabsence of Brexit and that real incomes regain two-thirds of the value they wouldhave had without Brexit as compared with our ldquosmallrdquo initial Brexit scenario

This global model has supply-and-demand equations and hence quantitiesprices and price elasticities for each of the grape and wine products and for asingle composite of all other products in each country Income elasticities ofdemand also exist for each final product Grapes are assumed to not be traded inter-nationally but other products are exported and imported Each market is assumed tohave been in equilibrium before any shock and to find a new market-clearingoutcome following any exogenously introduced shock

4The nominal price of the British pound in US dollars in the fortnight following the Brexit vote on June23 2016 dropped 13 to US$130 and a year later the pound sat at the same rate having dropped to anaverage of $124 (an 18 devaluation) between October 2016 and April 2017 Our choice of a low of 10and a high of 20 aims to capture future possible rates while uncertainties remain The average real wagein the United Kingdom fell in the first half of 2017 and projected real GDP growth during 2018ndash2020 hasbeen revised down to 17 in the United Kingdomrsquos latest budget (HM Treasury 2017)5South Africa currently has duty-free access to the EU28 for just 50 megaliters (ML) of wine per yearbeyond which the MFN tariff rate applies South Africa currently exports around 320 ML to theEuropean Union one-third of which initially goes to the United Kingdom How that quota of 50 MLis divided between the United Kingdom and the EU27 is subject to future negotiation (Rollo et al2016 Swinbank 2017) In the following discussion we assume none of it is accepted by the UnitedKingdom so that all South African wine imports pay the United Kingdomrsquos MFN tariff under Brexitbut altering that assumption makes very little difference to our results for any countryrsquos trade exceptSouth Africarsquos

Kym Anderson and Glyn Wittwer 231

V Projecting Global Wine Markets to 2025

Global wine production and exports are projected in the baseline from 2014 to 2025consistent with past trends The modelrsquos global volume of production (and consump-tion) rises little over that 11-year period (9) made up of a 6 decline in nonpre-mium wine and a one-sixth rise in commercial-premium and superpremium wineIn real (2014 US$) value though global wine output and consumption increaseby about 50 in total and 60 in the two premium categories The internationaltrade projections are similar although a little larger with the share of global wineproduction exported (equivalent to the share of global consumption imported)rising 2 percentage points between 2014 and 2025

The baseline projection does not greatly alter the 2014 shares of various countriesin global wine production apart from China because we assume vineyard expansionthere is faster than elsewhere6 In value terms this assumption means that Chinamoves from fifth to fourth by 2025 behind France the United States and ItalySpain remains barely ahead of Australia and they and Germany take the nextthree places (Figure 5(a)) In total wine-production volume terms China movesfrom sixth to fifth place and Argentina drops from fifth to eighth (and fromeighth to ninth in value terms)

When their products are subdivided into fine wines (superpremium still plus spark-ing) commercial-premium wines and nonpremium wines France and the UnitedStates retain the highest two places on the global ladder for fine-wine productionand Spain and Italy retain the top two places for nonpremium wine As for commer-cial-premium wine production (defined to cost between US$250 and US$750 perliter pretax at a countryrsquos wholesale level or national border) Italy retains the topranking over our projections period but at least in terms of value China challengesFrance for second place

The country rankings by projected value of total wine consumption change some-what more than those for production by 2025 with China taking second place afterthe United States ahead of France and Germany and then the United Kingdomslightly overtaking Italy to slip into fifth place (Figure 5(b)) The United StatesFrance and Germany retain the top three rankings for consuming fine wine butCanada slightly overtakes Italy for fourth place in terms of value at least In thecase of commercial-wine consumption China strengthens its number-one positionahead of the United States and the United Kingdom does likewise vis-agrave-visGermany for third place

As for the projected changes in consumption volumes China is projected to dom-inate the increase in aggregate although the United States is projected to lead the

6 In fact Chinarsquos wine production fell steadily between 2012 and 2016 by a total of one-sixth so Chinarsquoswine imports may grow faster in practice than in this baseline projection

232 UK and Global Wine Markets by 2025 and Implications of Brexit

Figure 5

Value of Wine Production and Consumption in Key Countries 2014 and Projected Baseline for2025 (2014 US$ Million at WineryWholesale Pretax Prices)

Fig5-BW

onlin

eBW

inprint

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 233

increase in consumption of fine wine In Western Europe and in the SouthernHemispherersquos New World countries fine wines are projected to substitute for com-mercial wines (defined as the sum of commercial-premium and nonpremiumwines) with almost no change in total wine consumption Sub-Saharan Africa isthe next region that is projected to take off with its growth accounting for morethan one-third of the rest of the worldrsquos increase in volume consumed

Those differences in production versus consumption rankings are reflected ininternational trade Figure 6 shows that France Italy and Spain remain the threedominant exporters of wine in aggregate value but the rankings of the next fewchange with Australia being slightly ahead of Chile and the United StatesGermany and New Zealand being nearly tied for sixth place in value termsFrance and Italy are even more dominant in fine-wine exports and remain so by2025 while Italy outranks France in the commercial-premium export categoryand Spain outranks Italy Australia and Chile in the nonpremium export class

Among the importers the United States and the United Kingdom are projected tocontinue to hold the first two places in 2025 in value terms but China moves intothird place slightly ahead of Germany followed well behind by Canada HongKong Belgium-Luxembourg the Netherlands and Japan (Figure 6(b)) OtherAfrica (excluding South Africa) is projected to experience the largest increase inimports among all the other regions followed by Other Asia which becomes asbig as Germany in value terms (Figure 7(a)) In terms of total volume of wineimports Germany and the United Kingdom held the top two shares in 2014 butby 2025 the United Kingdom is projected to be well ahead of Germany (Figure 7(b)) However this projection ignores the effects of Brexit to which we now turn

VI How Might Wine Markets Be Affected by the United Kingdomrsquos Exitfrom the European Union

As mentioned earlier for our two alternative scenarios to capture the initial effects ofBrexit (ldquolargerdquo and ldquosmallrdquo) we assume that following the United Kingdomrsquos exitfrom the European Union the United Kingdomrsquos rate of economic growth wouldbe only one-third or two-thirds as fast for the period to 2025 the British poundwould be 20 or 10 lower in real terms than in our modelrsquos baseline projectionand the United Kingdom would apply the European Unionrsquos external tariff onwine to imports from EU member countries (as part of establishing MFN ratesvia the WTO to then start new bilateral FTA negotiations) In these initial scenarioswe assume the United Kingdom does not implement any new FTAs particularlywith the EU27 Chile and South Africa Following the discussion of those initialresults we present the results of a subsequent scenario that includes the implemen-tation of an FTA between the United Kingdom and the EU27 We assume these aretwo discrete steps with no agreed arrangements to smooth the transition betweenthem

234 UK and Global Wine Markets by 2025 and Implications of Brexit

Figure 6

Value of Wine Exports and Imports Key Wine-Trading Countries 2014 and Projected Baselinefor 2025 (2014 US$ million)

Fig6-BW

onlin

eBW

inprint

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 235

Figure 7

National Shares of Global Wine Import Value and Volume 2014 and Projected Baselinefor 2025 ()

Fig7-BW

onlin

eBW

inprint

Source Authorsrsquo model results

236 UK and Global Wine Markets by 2025 and Implications of Brexit

A Initial Impact of Brexit

Because the British pound dropped by one-sixth against the US dollar in the fourmonths following the Brexit vote in June 2016 and because the UnitedKingdomrsquos average real wage fell 1 between November 2016 and April 2017(according to the latest available data see Office of National Statistics 2017) weviewed our ldquolargerdquo scenario as more likely soon after the Brexit vote But wecompare those results with ones from our ldquosmallrdquo initial-impact scenario becauseit has become more likely following the June 2017 general election that ldquosofterrdquooptions will be considered by the new government Generally the results areabout half the size in the ldquosmallrdquo scenario with the exception of the bilateraltrade effects To show the sensitivity of results to our assumptions we point outthe differences when the ldquosmallrdquo results are not close to half the results shown forthe ldquolargerdquo scenario

Before turning to those results we first present a ldquofree-traderdquo scenario whichassumes the United Kingdom chooses the radical option of becoming the HongKong of Europe Unlikely as this scenario is the results provide assurance thatour global wine model generates the standard types of effects of exiting a customsunion In this ldquofree-traderdquo scenario the United Kingdom is assumed to move tozero tariffs on all wine imports Thus the current preferences on EU Chileanand South African wine imports disappear as all other wine exporters enjoy thesame free access to UK wine markets We assume in this scenario that Brexit hasno impact on the United Kingdomrsquos real income or consumption growth rates by2025 and that by then the British pound has returned to its 2014 value relative toother currencies (similar to the real exchange rate assumed in our baseline scenarioto 2025 see Appendix Table A1)7 Under these assumptions EU Chilean andSouth African wine exports to the United Kingdom in 2025 are 003 lower thanin the no-Brexit baseline and UK imports from other wine exporters are 025higher but the United Kingdomrsquos total wine imports are hardly any different(because local wine prices fall by only 05 and incomes are unchanged) In this sce-nario UK consumers expand their volume of consumption by just 05 while theEU27 exports more and other countries export less to the rest of the world such thatworld wine exports are almost unchanged In short all these effects are in the direc-tion that comparative static-customs-union theory would predict but they are smallbecause we assume the United Kingdomrsquos real income and currency are the same asin our baseline for 2025 as described in the previous section

In the ldquolargerdquo alternate initial scenario involving Brexit as compared with thebaseline scenario to 2025 the consumer price of wine in 2025 is 22 higher in the

7If this scenario were seen as a credible long-run trade policy it would encourage more investment thatwould eventually raise UK incomes but not without major structural changes that would initiallydisrupt the economy Implicitly we assume in our free-trade scenario that these two opposite impactson real UK incomes and the British pound exactly offset each other by 2025

Kym Anderson and Glyn Wittwer 237

United Kingdom in local currency terms (20 because of real depreciation of theBritish pound 4 because of new tariffs on EU Chilean and South Africanwines and ndash2 because of slower UK income growth) The volume of UKwine consumption is 28 lower 16 because of slower UK economic growth7 because of real depreciation of the British pound and 5 because of newtariffs Superpremium still-wine sales are the most affected dropping by two-fifths while sparkling and commercial-premium wines drop a bit less than one-quarter Because the average price rises by more than the decline in volume soldthe aggregate value of UK sales even in local currency terms falls under thisldquolargerdquo Brexit scenario Under the ldquosmallrdquo Brexit scenario the consumer price ofwine in 2025 is 11 higher in the United Kingdom and its volume of wine consump-tion is 17 lower

The volume of projected UK imports in 2025 is 427 ML or nearly one-quarterless in the ldquolargerdquo scenario than in the baseline scenario comprising 58ML less spar-kling wine 31 ML less superpremium still wine and 339 ML less commercial-premium wine World imports are lower by just 239 ML because imports byother countries are 189 ML higher in response to the lower international wineprices in this scenario In value terms UK imports are $175 billion (or 27)lower in 2025 because of ldquolargerdquo Brexit $113 billion because of lower incomes$038 billion because of the fall in the British pound and $014 billion because ofthe rise in wine-import tariffs (Table 4) These aggregate trade impacts are a littlemore than half this size under the ldquosmallrdquo Brexit scenario

Despite the import levels falling because of raised import tariffs domestic con-sumption of all three quality categories of UK-produced wine is lower with thanwithout Brexit because of shrunken demand for all wines resulting from loweredUK incomes and raised local prices because of devaluation of the British poundThe British poundrsquos devaluation does make it easier for the United Kingdom tosell wines abroad though The countryrsquos exports are 7 ML or nearly 5 higherin 2025 in the ldquolargerdquo Brexit scenario and UK production is 3 higher ThoseUK exports (or re-exports of imported bulk wine after it is bottled in the UnitedKingdom) that go to EU27 countries are reduced though because of the tariffnow imposed at the new EU border

Without Brexit the United Kingdomrsquos shares of global wine imports are slightlyhigher in volume terms in 2025 than in 2010ndash2015 but 2 percentage points lower invalue terms thanks to East Asiarsquos expanding demand for imports of premium winesWith ldquolargerdquo Brexit however that value share is an additional 2 percentage pointslower and the volume share is almost 5 points lower (Figure 8) The net effects ofthese impacts on global trade are shown in Figure 9 Most of the initial tradeeffect of Brexit is a large decline in net imports of wine by the United Kingdomwith very little offsetting positive effect on trade in the rest of the world Theldquosmallrdquo Brexit numbers are a bit more than half of those for ldquolargerdquo Brexit

238 UK and Global Wine Markets by 2025 and Implications of Brexit

Table4

Difference

in2025

Projected

VolumeandValue

ofWineIm

portsb

ytheUnitedKingdom

andtheResto

fthe

World

asaConsequence

oftheInitial

BrexitS

hock

(MLand2014US$

Million

ldquoLargerdquoSc

enario)

Volum

e(M

L)

Value

(US$

million)

NP+CPa

SuperPrb

Sparkling

TOTAL

NP+CPa

SuperPrb

Sparkling

TOTAL

ΔUK

impo

rtsdu

eto

Low

erincomes

minus198

minus20

minus29

minus247

58minus644

minus253

minus234

minus1131

65Low

erpo

und

minus70

minus10

minus14

minus93

22minus248

minus127

minus102

minus476

27Highertariffs

minus71

minus1minus16

minus87

20minus110

minus8minus2

4minus143

8TOTA

Lminus339

minus31

minus58

minus427

100

minus1001

minus388

minus360

minus1750

100

difffrom

base

2332

3325

2432

3227

of

totalc

uts

797

14100

5722

21100

ΔROW

netim

ports

143

2125

189

230

143

minus181

192

ΔWORLD

TRADE

minus195

minus10

minus34

minus239

minus763

minus246

minus543

minus1552

aNon

prem

ium

plus

commercial-premium

still

wines

bSu

perpremium

still

wines

SourceA

utho

rsrsquomod

elresults

Kym Anderson and Glyn Wittwer 239

The aggregate effect of ldquolargerdquo Brexit on the market shares of various wine-export-ing countries in the United Kingdom is almost indiscernible even with one decimalpoint The projected 2025 shares are quite different from the actual 2014 shares forseveral countries They are much smaller in 2025 for South Africa Australia andNew Zealand (and the United States in volume terms) and are much larger involume for Spain and in value for Italy These results occur because wine-exportingcountries benefit differentially from the varying rates of growth in net importdemand for wine in non-UK countries over this projection period The most impor-tant projected changes are the increase in the real value of annual wine importsbetween 2014 and 2025 by China (200 or $3 billion) Other Asia (110 or $22billion) and Africa (270 or $16 billion) More than half of Australiarsquos increasein annual exports from 2014 to 2025 go to Asia and more than half of SouthAfricarsquos increase in exports go to Other Africa

Table 5 reveals that ldquolargerdquo Brexit lowers EU Chilean and South African wineexports by 150 ML or US$12 billion in the case of the European Union with someof their exports diverted from the United Kingdom to the EU27 and other marketsin competition with New World exporters Although the United States Australiaand Argentina sell only a little less to the United Kingdom they also sell less toother countries For Chile and South Africa which lose their preferential access toUK (but not to EU27) markets in this Brexit scenario some exports are redirected

Figure 8

UK Shares of World Wine Imports 2010ndash2015 and Projected to 2025 without and with theldquoLargerdquo Initial Brexit Shock ()

Fig8-BW

onlin

eBW

inprint

Sources Anderson and Pinilla (2017) and authorsrsquo model results

240 UK and Global Wine Markets by 2025 and Implications of Brexit

from the United Kingdom to EU27 countries ndash but again they export less overallGlobal wine trade in 2025 is 240 ML (19) or $18 billion (35) less under thisldquolargerdquo Brexit scenario The percentage by which wine exportersrsquo trade shrinks isgreater for values than for volumes because of changes in relative prices of differ-ent-quality wines Those differences are shown in the numbers in parentheses inTable 5

Three other points are worth making about Table 5 One is that Australia sellsslightly more to the United Kingdom in the ldquosmallrdquo Brexit scenario rather thanslightly less as in the ldquolargerdquo Brexit case Evidently the negative income and price(devaluation) effects do not more than offset the positive trade-diverting effect onAustralian exports to the United Kingdom of removing preferences in the ldquosmallrdquoscenario Second New Zealand sells slightly more to non-UK countries underBrexit despite greater competition from the EU27 Chile and South Africa Thisanomaly is due to changes in the relative prices of different qualities of wine inglobal wine markets bearing in mind that New Zealand has the worldrsquos highestaverage price for still-wine exports And third the value (but not the volume) ofexports of ldquoOtherrdquo countries to markets other than the United Kingdom arehigher under Brexit This result too is due to changes in the relative prices of differ-ent qualities of wine in global wine markets

Figure 9

Difference in 2025 Wine Import Volumes and Values as a Result of the ldquoLargerdquo Initial BrexitShock (ML and US$ Million in 2014 US dollars)

Fig9-BW

onlin

eBW

inprint

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 241

B Subsequent Impact of Brexit from a UK-EU27 FTA

The next-most-likely step in the Brexit process is for the United Kingdom to nego-tiate a new trade arrangement with the EU27 We therefore assume that a UK-EU27FTAwith free bilateral wine trade is implemented and adjusted to by 2025 and thatprogress toward that end occurs soon enough that the adverse macroeconomicshocks from the initial impact of uncertainty over the Brexit process are confinedto those assumed in the ldquosmallrdquo scenario outlined above In this subsequent scenariowe assume that the British pound returns to the value it would have reached in 2025in the absence of Brexit (ie reversing the 10 devaluation assumed in the ldquosmallrdquoinitial Brexit scenario) and that real incomes in the United Kingdom are 8 higher

Table 5Difference in 2025 BilateralWine Import Volumes and Values fromKey Exporters by the United

Kingdom and the Rest of the World (RoW) as a Result of Initial Brexit Shock(ML and 2014US$ Million)a

(a) ldquolargerdquo scenario

Volume (ML) Value (2014US$ million)

UK RoW WORLD () UK RoW WORLD ()

EU27 minus287 136 minus150 (ndash17) minus1187 minus5 minus1192 (ndash31)Chile minus59 35 minus25 (ndash30) minus169 31 minus138 (ndash48)South Africa minus53 35 minus18 (ndash32) minus105 20 minus85 (ndash67)USA minus7 minus6 minus13 (ndash24) minus75 minus40 minus115 (ndash50)Australia minus4 minus3 minus7 (ndash09) minus25 minus65 minus90 (ndash30)Argentina minus3 minus9 minus12 (ndash48) minus16 minus39 minus55 (ndash52)New Zealand minus11 9 minus2 (ndash09) minus162 71 minus91 (ndash43)Others minus2 minus10 minus12 (ndash02) minus11 minus52 minus63 (ndash44)WORLD minus427 187 minus240 (ndash19) minus1750 minus79 minus1829 (ndash35)

(a) ldquosmallrdquo scenario

Volume (ML) Value (2014US$ million)

UK RoW WORLD () UK RoW WORLD ()

EU27 minus178 82 minus96 (ndash12) minus692 minus43 minus736 (ndash19)Chile minus46 28 minus18 (ndash24) minus128 36 minus91 (ndash32)South Africa minus43 29 minus14 (ndash42) minus82 23 minus59 (ndash47)USA 1 minus6 minus5 (ndash11) minus23 minus28 minus51 (ndash22)Australia 5 minus10 minus5 (ndash06) 19 minus56 minus38 (ndash13)Argentina 0 minus6 minus6 (ndash26) minus3 minus25 minus29 (ndash27)New Zealand minus5 4 minus1 (ndash06) minus80 34 minus46 (ndash22)Others 0 minus9 minus9 (ndash01) minus1 minus33 minus34 (ndash24)WORLD minus266 112 minus154 (ndash13) minus991 minus92 minus1083 (ndash21)

a Numbers in parentheses are the percentage difference between the Brexit and baseline scenarios for 2025 projected wine-import volumes orvalues by source

Source Authorsrsquo model results

242 UK and Global Wine Markets by 2025 and Implications of Brexit

in 2025 than in the ldquosmallrdquo scenario ndash that is they regain two-thirds of the value theywould have reached without Brexit-related uncertainty

This subsequent development in the Brexit process reverses most of but not allthe initial effects of Brexit by 2025 because of our assumption that the lost growthin the initial years of uncertainty following the Brexit vote are only partly recoveredby 2025 following the implementation of a UK-EU27 FTA Moreover the longer ittakes before this FTA is finalized and implemented the longer the estimated initialadverse effects persist and the larger the cumulative cost of Brexit to UK wine con-sumers and to grape and wine producers in wine-exporting countries

Table 6 summarizes the subsequent trade effects for 2025 It suggests that all butone-tenth of the loss in value of world trade in wine from the initial ldquosmallrdquo impact isrestored most of it because of a smaller reduction in the United Kingdomrsquos wineimports following the FTA Most of that improved outcome is because of recoveredimports from the EU27 commensurate with the latterrsquos high share of UK importsEven though Chile and South Africa are assumed in this scenario to have not yetsigned an FTA with the United Kingdom they export slightly more to the UnitedKingdom (and even more to all other countries) than in the initial ldquosmallrdquo scenario

Within the United Kingdom this FTA brings down the local currency consumerprice of wine by 9 largely offsetting the 11 rise in the initial ldquosmallrdquo Brexit sce-nario and it raises the volume consumed in the United Kingdom by 18 fully off-setting the 17 fall in the initial ldquosmallrdquo Brexit scenario

Table 6Difference in 2025 Bilateral Wine-Import Volumes and Values from Key Exporters by the

United Kingdom and the Rest of the World (RoW) as a Result of Implementing a UK-EU27FTA Following Initial ldquoSmallrdquo Brexit Shock (ML and 2014US$ Million Difference Relative to

Initial ldquoSmallrdquo Brexit Scenario)a

Volume (ML) Value (2014US$ million)

UK RoW WORLD (ldquosmallrdquo)a UK RoW WORLD (ldquosmallrdquo)a

EU27 212 minus112 100 (minus96) 750 0 750 (minus806)Chile 1 3 4 (minus18) 3 40 43 (minus92)South Africa 3 2 5 (minus14) 3 18 21 (minus60)USA 3 3 6 (minus5) 26 21 47 (minus50)Australia 1 1 2 (minus5) 8 38 46 (minus33)Argentina 1 4 5 (minus6) 4 20 24 (minus28)New Zealand 4 minus3 1 (minus1) 61 minus25 36 (minus46)Others 1 6 7 (minus9) 5 minus46 minus41 (92)WORLD 226 minus96 130 (minus154) 860 66 926 (minus1022)

a Numbers in parentheses are the world trade differences between the ldquosmallrdquo initial Brexit scenario and the baseline scenario copied fromcolumns 3 and 7 of Table 5(b)

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 243

In short this subsequent step in the Brexit process can be expected to restore muchof the initial adverse effects in the United Kingdom of the Brexit vote by 2025 Thewine exporters in the EU27 Chile and South Africa also lose less in this scenarioas compared with the initial ldquosmallrdquo scenario while other key exporters sell nearlyas much in 2025 in this scenario as they are projected to in the baseline ndash that isthey are beneficiaries of the reduced discrimination in global wine markets in thiscase where Chile and South Africa no longer enjoy preferential access to the UKmarket (and even though EU27 producers are again allowed to do so) Keep inmind though that although the estimated losses in the initial scenarios are reportedjust for 2025 they are expected to be felt throughout the preceding years of uncertaintyTherefore the longer it is before uncertainty abates and the United Kingdomrsquos currentpreferential trading arrangements are replaced by newones such as a UK-EU27 FTAthe higher the cumulative cost of the Brexit vote to the wine trade

VII Caveats and Conclusions

The above Brexit simulations are just a few of many scenarios that could be modeledObvious additional ones could also assume that FTAs are reached between theUnited Kingdom and other trading partners including Chile and South Africaand such countries as Australia and New Zealand Some of these other countriesare already exploring the ECrsquos invitation (see European Commission 2015) to nego-tiate an FTA with what will be the EU27 The sequence in which FTAs are signedand the speed with which they are implemented will matter (as was also the casewith the sequential signing over the past decade of bilateral FTAs with NortheastAsian countries by Chile Australia and New Zealand see Anderson andWittwer 2015) Even if the United Kingdom were able to sign additional bilateralFTAs and begin implementing them before 2025 it would have little effect on theabove results (because wine tariffs are such a minor contributor to them) unlessthose FTAs were to accelerate the United Kingdomrsquos economic growth and therise in the British poundrsquos value before 2025

We assume above that no changes are made to alcohol excise duties in the UnitedKingdom following Brexit even though they are scheduled to be progressively raisedwith inflation and may be raised even more for wine relative to spirits to offset theopposite effects of Brexit on those two domestic industries Nor do we make anyallowance for UK increases in consumer wine prices that may be needed to coverthe higher cost of clearing customs on imports from EU27 countries (which maybe more or less offset by increased smuggling and duty-free purchasing)Importers of fine wines including individuals who buy Bordeaux en primeur maywell reduce their demand because of concerns that delays on docks will affect thequality of their wines or because of greater currency uncertainty

To summarize Brexit is costly initially to UK consumers of wine (and of manyother tradable products) because the domestic retail price in local currency tax-

244 UK and Global Wine Markets by 2025 and Implications of Brexit

inclusive terms is 22 higher than otherwise in 2025 and the volume of wine con-sumed domestically is 28 lower in the ldquolargerdquo Brexit scenario (or 11 higher pricesand 17 lower quantities in the ldquosmallrdquo case) by 2025 unless a UK-EU27 FTA isdeveloped before then Even if such an FTA does get signed ratified by all 28 par-liaments and implemented by 2025 the slower income growth in the interim meansa smaller UK wine market in 2025 than would otherwise have been the case Thevolume reduction is a blow to many participants in UK wine-bottling transportingstoring wholesaling and retailing businesses in addition to restaurants and pubsVery little of that initial impact occurs because of higher import tariffs most impor-tant is the assumed fall in UK real incomes

The small but growing number of local UK vignerons (who supply less than 05of all domestic-wine sales and less than 4 of sparkling-wine sales) is initially pro-jected to sell less wine domestically because of reduced demand and to export only alittle extra abroad while the British pound is devalued Overall their production isonly 3 higher in 2025 even in the ldquolargerdquo Brexit scenario (not taking intoaccount the fact that their casual labor is likely to be more expensive under Brexitbecause of tighter restrictions on immigration) and their production diminishes ifand when a UK-EU27 FTA is in place

Even with a UK-EU27 FTA in place EU27 wine exporters export US$56 millionless wine in 2025 thanks to Brexit Chile and South Africa export $88 million lesswine and wine exports from the rest of the world are little different in aggregatealthough slightly larger from Australia

Clearly there will be great uncertainly for some time over the possible policy out-comes to flow from Brexit and their consequent sequential impacts on UK house-hold disposable incomes foreign exchange rates and bilateral wine tariffsMeanwhile the above projections under explicit assumptions provide some idea ofhow wine markets might be affected by the most-likely first two stages of theBrexit process (agreeing on a new tariff schedule at the WTO and agreeing to andimplementing a UK-EU27 FTA) In particular they make clear that nontrivialinitial impacts could affect the domestic wine market impacts that are likely to belarger than just the direct impact of changes in bilateral tariffs If the UnitedKingdom succeeds in getting countries to agree to transition arrangements thatdelay the changes in tariffs until new FTAs are signed and ratified the initialeffects will be less dramatic than in our first scenarios but it remains to be seenwhether any such agreements can be reached In any event the net effect of Brexiton the welfare of the worldrsquos consumers and producers of wine as a whole will benegative not just initially but permanently unless new trade policy commitmentsby the United Kingdom with major wine-exporting countries are sufficiently moreliberal than current arrangements

Kym Anderson and Glyn Wittwer 245

Appendix Key Equations in the Global Wine Markets Model

Amodel of the worldrsquos wine markets was first published byWittwer et al (2003) andmuch improved by Anderson and Wittwer (2013) Several significant enhancementshave been made to that original model Wine types have been disaggregated intomore types ndash namely nonpremium (including generic bulk) commercial-premiumand superpremium still wines and sparkling wines8 As in the original modelthere are two types of grapes premium and nonpremium Nonpremium wines usenonpremium grapes exclusively superpremium wines use mainly premium grapesand commercial-premium and sparkling wines use both types of grapes In theregional dimension the number of countries and country groups has expandedfrom 10 in the original model to 51 44 individual nations and 7 compositeregions The modelrsquos database is calibrated to 2014 for this paper based on datain Anderson et al (2017) trade data for which are downloaded from httpscomtradeunorg The modelrsquos supply-and-demand equations are based on the orig-inal ORANI modelrsquos theory (see Dixon Parmenter Sutton and Vincent 1982) andthe model is implemented using GEMPACK software (Harrison Horridge Jerieand Pearson 2014)

An enhancement of importance to the present study is the inclusion of exchange-rate variables in the model which allow a distinction between price impacts asobserved in local currency units and those observed in 2014 US dollars

In the model the grape and wine sectors minimize costs of intermediate inputssubject to weak constant elasticity of substitution (CES) substitutability betweeninputs By assumption no intermediate inputs are imported from other countries9Hence

Xcid frac14 f ethX1id CESfrac12P

cid=P1id $THORN eth1THORN

P1id X1id frac14X

c

Xcid P

cid eth2THORN

whereXcid is the quantity demanded of commodity c by grape or wine industry i in

region d Pcid is the corresponding price and X1id and P1id are the respective inter-

mediate composite quantities and prices

Two primary factors are employed in the sector labor (the quantity of which isendogenous with perfectly elastic supply) and capital Capital is usually treated asexogenous in quantity with rates of return bearing all the adjustment in thevarious scenarios This reflects the fact that grapes (a perennial crop) and wine-

8Commercial-premium still wines are defined by Anderson and Nelgen (2011) to be those priced betweenUS$250 and $750 per litre pre-tax at a countryrsquos border or wholesale9An exception concerns bulk wine imports used mainly in the bottling of commercial premium wine in theUK USA Canada Japan and Germany

246 UK and Global Wine Markets by 2025 and Implications of Brexit

plant capacity adjust slowly to market signals

Lid frac14 f ethFid CESfrac12Wid=PFid $THORN eth3THORN

Kid frac14 f ethFid CESethRid=PFid $THORN eth4THORN

PFid Fid frac14 Lid Wid thorn Kid Rid eth5THORN

Grape and wine producers are assumed to minimize costs subject to CES substitu-tion between capital and labor Equations (3) to (5) show primary factor demandsfor the labor composite Lid and capital Kid subject to a composite factor demandFid by industry i in region d The factor prices are Wid for labor Rid for capitalrentals and PFid for composite factor prices

Appendix Table A1Cumulative Consumption and Population Growth Rates and Changes in the Real Exchange Rate

(RER) Relative to the US Dollar 2014 to 2025 without Brexit ()

Aggregateconsumption Poprsquon RER

Aggregateconsumption Poprsquon RER

France 18 4 minus11 Australia 35 11 minus17Italy 11 2 minus9 New Zealand 32 9 minus26Portugal 14 0 minus9 Canada 27 8 minus18Spain 26 8 minus9 United States 31 8 0Austria 19 4 minus7 Argentina 7 10 109Belgium 20 7 minus9 Brazil 16 8 minus29Denmark 22 2 minus9 Chile 55 8 minus2Finland 21 3 minus7 Mexico 42 12 minus8Germany 14 minus2 minus11 Uruguay 45 3 1Greece 22 minus1 minus14 Other Latin America 60 10 minus5Ireland 42 12 minus9 South Africa 36 12 minus1Netherlands 21 4 minus9 Turkey 50 8 20Sweden 24 9 minus13 North Africa 53 11 0Switzerland 18 8 minus6 Other Africa 109 18 84United Kingdom 32 6 1 Middle East 52 18 minus12Other West Europe 21 10 minus1 China 79 3 5Bulgaria 41 minus7 7 Hong Kong 42 3 2Croatia 20 minus2 minus1 India 134 13 17Georgia 35 0 23 Japan 11 minus3 minus24Hungary 25 minus3 minus11 Korea 38 1 minus9Moldova 49 minus11 13 Malaysia 62 15 minus16Romania 45 minus4 22 Philippines 75 18 7Russia 18 minus2 minus8 Singapore 44 21 minus22Ukraine 22 minus5 14 Taiwan 29 1 minus13Other East Europe 40 minus5 48 Thailand 47 3 minus9

Other Asia 99 10 10

Source Authorsrsquo compilation from projections by various international agencies and from global economy-wide modeling by Anderson andStrutt (2016)

Kym Anderson and Glyn Wittwer 247

The composite factor demand Fid is proportional to total output Qid subject to aprimary factor using technology Aid Hence

Fid frac14 Qid Aid eth6THORN

The perfectly competitive zero-pure-profit condition is that total revenue valued atthe output price P0s

i multiplied by Qid equals the total production cost

P0si Qid frac14

X

cPcid X1cid thorn

X

oWo

id Loid thorn Rid Kid eth7THORN

Household demands follow a linear expenditure system in each region We reducethe optimizing problem for household consumption of each commodity subjectto a budget constraint to equations describing subsistence and discretionarydemands Aggregate subsistence expenditure WSUBd depends only on consumerprices P3cd for each commodity and the number of households N as per-capita sub-sistence quantities XSUBcd subject to given preferences are constant

WSUBd frac14X

c

P3cd XSUBcd Nd eth8THORN

Discretionary expenditures for each commodity (the left-hand side of equation (9))are equal to the marginal budget share (βcd) of aggregate discretionary expenditureThis aggregate is the bracketed term on the right-hand side of equation (9) whereW3TOTd is aggregate nominal expenditure

P3cdethX3cd XSUBcd NdTHORN frac14 βcdethW3TOTd WSUBdTHORN eth9THORN

Because real aggregate consumption is usually exogenous in our partial equilibriumsimulations the linear expenditure system determines the consumption shares ofindividual final commodities (ie the five wine types plus a composite of all otherconsumption items) driven by changes in relative prices as faced by domestic con-sumers The income elasticity of demand for each commodity is equal to the mar-ginal budget share divided by the expenditure share This number varies from 05for nonpremium wine to 17 for superpremium still wine The income elasticity ofdemand for other consumption is very close to 10 because wine accounts for anaverage of only 03 of aggregate expenditures globally and no more than 11in any country (Anderson and Nelgen 2011 Table 166)

A feature of our revised model of world wine markets is the inclusion of nominalexchange rates These appear directly in the equation linking retail prices (P3scd ) toproducer prices by country of origin (P0s

c ) where c denotes the wine type

P3 scd frac14 P0s

cfd

fsTtarcd T

taxcd thorn Pm

cd eth10THORN

248 UK and Global Wine Markets by 2025 and Implications of Brexit

The exchange rates in the consuming and producing regions are fd and fs respec-tively expressed as local currency units per $US Ttar

cd is the power of the tariff in theconsuming region and Ttax

cd is the power of the domestic consumption (or excise) taxprior to any generic value-added or goods-and-services tax Pm

cd is the price of marginm assumed to be locally supplied nontradable and therefore unaffected by theexchange rate

A given level of consumption for wine type c (X3cd) is satisfied using theArmington (1969) assumption in which wines from different countries of originare imperfectly substitutable First domestic wine is imperfectly substitutable witha composite of imports

X3sscd frac14 f ethX3cd CESethP3sscd=P3cdTHORNTHORN ss frac14 domestic imports eth11THORN

and then imports by origin (X3scd ) are determined in a second CES equation

X3scd frac14 f ethX3ssfrac14importscd CESethP3scd=P3

ssfrac14importscd THORNTHORN eth12THORN

The model enables us to show how changes in international competitiveness affectthe worldrsquos wine markets A crucial part of this exercise is explaining how pricesdetermined outside the grape and wine markets influence these markets Becausethe model is in partial equilibrium to depict the impacts of changes in internationalcompetitiveness outside price changes need to be imposed as shocks on the modelThe price of intermediate inputs other than grapes and wine shown in equations (1)and (2) is set equal to the price of GDP (Pg

d ) multiplied by a shifter Fcd

Pcid frac14 Fc

dPgd eth13THORN

If no specific price observations are available the shifter Fcd remains exogenous and

unshocked with the change in price being determined by a shock to the price ofGDP If observations are available for specific input price movements the shifterFc

dbecomes endogenous with Pc

id now exogenous and shocked

Wid frac14 Fwd P

gd eth14THORN

Wage rates are treated similarly In equation (14) if the wage shifter Fwd is exogenous

changes in wage rates Wid are determined by changes in the price of GDP If wagerate data are available Fw

d becomes endogenous andwage rates are shocked directly

Pmcd frac14 Fm

d Pgd eth15THORN

The prices of trade and transport margins are also determined by the price of GDP ifthe shifter Fm

d in equation (15) is exogenous

Changes in international competitiveness depend on changes in relative pricelevels and changes in nominal exchange rates In equation (16) fR

s denotes real

Kym Anderson and Glyn Wittwer 249

exchange-rate movements relative to the US dollar

fRs frac14 Pg

s =frac12PgUSA

fs$ eth16THORN

In equation (16) the nominal exchange rate for the United States is alwaysunchanged because nominal and real exchange rates are expressed relative to theUS currency

Changes in international market conditions may have impacts in one direction onproducer prices as expressed in US dollars P0s

i and potentially in the opposite direc-tion in local currency units Hence we calculate real producer pricesP0s

iloc in localcurrency terms (ie the price most relevant to domestic producers)

P0siloc frac14 P0s

ifs=P

gs eth17THORN

To obtain real price changes in local currency terms we convert US dollar prices(P3 s

cd for the source-specific price and P3cd for the source-composite price) to reallocal currency prices (P3 s

cdloc and P3cdloc) using the consumer price index (CPI)(Pc

d ) as the deflator

P3 scdloc frac14 P3 s

cdfd=P

cd eth18THORN

and

P3cdloc frac14 P3cdfd=Pcd eth19THORN

References

Anderson K and Nelgen S (2011) Global Wine Markets 1961 to 2009 A StatisticalCompendium Adelaide University of Adelaide Press Also available as an e-book atwwwadelaideeduaupresstitlesglobal-wine

Anderson K Nelgen S and Pinilla V (2017) Global Wine Markets 1860 to 2016 AStatistical Compendium Adelaide University of Adelaide Press (forthcoming) Also tobe available as an e-book at wwwadelaideeduaupress

Anderson K and Pinilla V (with the assistance of A J Holmes) (2017)Annual Database ofGlobal Wine Markets 1835 to 2016 Wine Economics Research Centre University ofAdelaide to be posted in October at wwwadelaideeduauwine-econdatabases

Anderson K and Strutt A (2016) Impacts of Asiarsquos rise on African and Latin Americantrade Projections to 2030 World Economy 39(2) 172ndash194

Anderson K andWittwer G (2013) Modeling global wine markets to 2018 Exchange ratestaste changes and Chinarsquos import growth Journal of Wine Economics 8(2) 131ndash158

Anderson K and Wittwer G (2015) Asiarsquos evolving role in global wine markets ChinaEconomic Review 35 1ndash14

250 UK and Global Wine Markets by 2025 and Implications of Brexit

Armington P A (1969) A theory of demand for products distinguished by place of produc-tion IMF Staff Papers 16 159ndash178

Baldwin R E (ed) (2016) Brexit Beckons Thinking Ahead by Leading Economists LondonCentre for Economic Policy Research

Campos N F (2016) Lousy experts Looking back at the ex ante estimates of the costs ofBrexit In R E Baldwin (ed) Brexit Beckons Thinking Ahead by Leading Economists(ch 3) London Centre for Economic Policy Research

Dixon P Parmenter B Sutton J and Vincent D (1982) ORANI A Multisectoral Modelof the Australian Economy Contributions to Economic Analysis 142 Amsterdam North-Holland

European Commission (2015) Trade for All Towards a More Responsible Trade andInvestment Policy Brussels European Commission

Harrison J Horridge M Jerie M and Pearson K (2014) GEMPACK ManualMelbourne GEMPACK Software

HM Treasury (2017) Spring Budget 2017 London HM TreasuryHolmes A J and Anderson K (2017a)Annual Database of National Beverage Consumption

Volumes and Expenditures 1950 to 2015 Wine Economics Research Centre University ofAdelaide posted at wwwadelaideeduauwine-econdatabases

Holmes A J and Anderson K (2017b) Convergence in National Alcohol ConsumptionPatterns New Global Indicators Journal of Wine Economics 12(2) 117ndash148

Ludington C C (2013) The Politics of Wine in Britain A New Cultural History BasingstokeUK Palgrave Macmillan

Office of National Statistics (2017) Analysis of real earnings London Office of NationalStatistics Available at httpswwwonsgovukemploymentandlabourmarketpeopleinworkearningsandworkinghoursarticlessupplementaryanalysisofaverageweeklyearningslatest

Rollo J Borchert I Dawar K Holmes P and Winters L A (2016) The World TradeOrganisation A safety net for a post-Brexit UK trade policy Briefing Paper 1UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-1-final-1pdf

Smith A (2017) Brexit Hard truths and hard choices UKTPO blog 19 June Available athttpsblogssussexacukuktpo20170619brexit-hard-truths-and-hard-choices

Swinbank A (2017) World trade rules and the policy options for British agriculture post-Brexit Briefing Paper 7 UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-7pdf

Viner J (1950) The Customs Union Issue New York Carnegie Endowment for InternationalPeace

Wine Australia (2015) Export Market Guide European Union Adelaide Wine AustraliaWittwer G Berger N and Anderson K (2003) A model of the worldrsquos wine markets

Economic Modelling 20(3) 487ndash506

Kym Anderson and Glyn Wittwer 251

Page 11: U.K. and Global Wine Markets by 2025, and Implications of ... · U.K. and Global Wine Markets by 2025, and Implications of Brexit* Kym Andersona and Glyn Wittwerb Abstract The United

capital is assumed to grow net of depreciation at 15 per year in China but zero else-where (consistent with the almost-zero growth in global wine production and con-sumption over the past two decades)

Two alternative scenarios are considered (ldquolargerdquo and ldquosmallrdquo) for the initialimpact of Brexit to provide a range of results We assume that because of theUnited Kingdomrsquos decision to exit the European Union the rate of UK realgross domestic product (GDP) growth is only one-third or two-thirds as fast overthe projection period (09 or 18 per year instead of 26) and the Britishpound will be 20 or 10 lower in real terms than in our modelrsquos core baselineprojection4 In both alternative scenarios we assume that the United Kingdomapplies the European Unionrsquos external tariffs on wine from March 30 2019 atthe end of the two-year period following the United Kingdomrsquos formal triggeringof Article 50 (see Table 3)

We then consider a subsequent impact of Brexit presumed to result from negoti-ating signing implementing and responding to an FTA with the EU27 by 2025(Chile and South Africa currently have preferential access to EU wine marketsbut we continue to assume in this scenario that the United Kingdom does not imple-ment new bilateral FTAs with them or others in our time frame)5 This subsequentscenario assumes the British pound returns to what it would have been in 2025 in theabsence of Brexit and that real incomes regain two-thirds of the value they wouldhave had without Brexit as compared with our ldquosmallrdquo initial Brexit scenario

This global model has supply-and-demand equations and hence quantitiesprices and price elasticities for each of the grape and wine products and for asingle composite of all other products in each country Income elasticities ofdemand also exist for each final product Grapes are assumed to not be traded inter-nationally but other products are exported and imported Each market is assumed tohave been in equilibrium before any shock and to find a new market-clearingoutcome following any exogenously introduced shock

4The nominal price of the British pound in US dollars in the fortnight following the Brexit vote on June23 2016 dropped 13 to US$130 and a year later the pound sat at the same rate having dropped to anaverage of $124 (an 18 devaluation) between October 2016 and April 2017 Our choice of a low of 10and a high of 20 aims to capture future possible rates while uncertainties remain The average real wagein the United Kingdom fell in the first half of 2017 and projected real GDP growth during 2018ndash2020 hasbeen revised down to 17 in the United Kingdomrsquos latest budget (HM Treasury 2017)5South Africa currently has duty-free access to the EU28 for just 50 megaliters (ML) of wine per yearbeyond which the MFN tariff rate applies South Africa currently exports around 320 ML to theEuropean Union one-third of which initially goes to the United Kingdom How that quota of 50 MLis divided between the United Kingdom and the EU27 is subject to future negotiation (Rollo et al2016 Swinbank 2017) In the following discussion we assume none of it is accepted by the UnitedKingdom so that all South African wine imports pay the United Kingdomrsquos MFN tariff under Brexitbut altering that assumption makes very little difference to our results for any countryrsquos trade exceptSouth Africarsquos

Kym Anderson and Glyn Wittwer 231

V Projecting Global Wine Markets to 2025

Global wine production and exports are projected in the baseline from 2014 to 2025consistent with past trends The modelrsquos global volume of production (and consump-tion) rises little over that 11-year period (9) made up of a 6 decline in nonpre-mium wine and a one-sixth rise in commercial-premium and superpremium wineIn real (2014 US$) value though global wine output and consumption increaseby about 50 in total and 60 in the two premium categories The internationaltrade projections are similar although a little larger with the share of global wineproduction exported (equivalent to the share of global consumption imported)rising 2 percentage points between 2014 and 2025

The baseline projection does not greatly alter the 2014 shares of various countriesin global wine production apart from China because we assume vineyard expansionthere is faster than elsewhere6 In value terms this assumption means that Chinamoves from fifth to fourth by 2025 behind France the United States and ItalySpain remains barely ahead of Australia and they and Germany take the nextthree places (Figure 5(a)) In total wine-production volume terms China movesfrom sixth to fifth place and Argentina drops from fifth to eighth (and fromeighth to ninth in value terms)

When their products are subdivided into fine wines (superpremium still plus spark-ing) commercial-premium wines and nonpremium wines France and the UnitedStates retain the highest two places on the global ladder for fine-wine productionand Spain and Italy retain the top two places for nonpremium wine As for commer-cial-premium wine production (defined to cost between US$250 and US$750 perliter pretax at a countryrsquos wholesale level or national border) Italy retains the topranking over our projections period but at least in terms of value China challengesFrance for second place

The country rankings by projected value of total wine consumption change some-what more than those for production by 2025 with China taking second place afterthe United States ahead of France and Germany and then the United Kingdomslightly overtaking Italy to slip into fifth place (Figure 5(b)) The United StatesFrance and Germany retain the top three rankings for consuming fine wine butCanada slightly overtakes Italy for fourth place in terms of value at least In thecase of commercial-wine consumption China strengthens its number-one positionahead of the United States and the United Kingdom does likewise vis-agrave-visGermany for third place

As for the projected changes in consumption volumes China is projected to dom-inate the increase in aggregate although the United States is projected to lead the

6 In fact Chinarsquos wine production fell steadily between 2012 and 2016 by a total of one-sixth so Chinarsquoswine imports may grow faster in practice than in this baseline projection

232 UK and Global Wine Markets by 2025 and Implications of Brexit

Figure 5

Value of Wine Production and Consumption in Key Countries 2014 and Projected Baseline for2025 (2014 US$ Million at WineryWholesale Pretax Prices)

Fig5-BW

onlin

eBW

inprint

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 233

increase in consumption of fine wine In Western Europe and in the SouthernHemispherersquos New World countries fine wines are projected to substitute for com-mercial wines (defined as the sum of commercial-premium and nonpremiumwines) with almost no change in total wine consumption Sub-Saharan Africa isthe next region that is projected to take off with its growth accounting for morethan one-third of the rest of the worldrsquos increase in volume consumed

Those differences in production versus consumption rankings are reflected ininternational trade Figure 6 shows that France Italy and Spain remain the threedominant exporters of wine in aggregate value but the rankings of the next fewchange with Australia being slightly ahead of Chile and the United StatesGermany and New Zealand being nearly tied for sixth place in value termsFrance and Italy are even more dominant in fine-wine exports and remain so by2025 while Italy outranks France in the commercial-premium export categoryand Spain outranks Italy Australia and Chile in the nonpremium export class

Among the importers the United States and the United Kingdom are projected tocontinue to hold the first two places in 2025 in value terms but China moves intothird place slightly ahead of Germany followed well behind by Canada HongKong Belgium-Luxembourg the Netherlands and Japan (Figure 6(b)) OtherAfrica (excluding South Africa) is projected to experience the largest increase inimports among all the other regions followed by Other Asia which becomes asbig as Germany in value terms (Figure 7(a)) In terms of total volume of wineimports Germany and the United Kingdom held the top two shares in 2014 butby 2025 the United Kingdom is projected to be well ahead of Germany (Figure 7(b)) However this projection ignores the effects of Brexit to which we now turn

VI How Might Wine Markets Be Affected by the United Kingdomrsquos Exitfrom the European Union

As mentioned earlier for our two alternative scenarios to capture the initial effects ofBrexit (ldquolargerdquo and ldquosmallrdquo) we assume that following the United Kingdomrsquos exitfrom the European Union the United Kingdomrsquos rate of economic growth wouldbe only one-third or two-thirds as fast for the period to 2025 the British poundwould be 20 or 10 lower in real terms than in our modelrsquos baseline projectionand the United Kingdom would apply the European Unionrsquos external tariff onwine to imports from EU member countries (as part of establishing MFN ratesvia the WTO to then start new bilateral FTA negotiations) In these initial scenarioswe assume the United Kingdom does not implement any new FTAs particularlywith the EU27 Chile and South Africa Following the discussion of those initialresults we present the results of a subsequent scenario that includes the implemen-tation of an FTA between the United Kingdom and the EU27 We assume these aretwo discrete steps with no agreed arrangements to smooth the transition betweenthem

234 UK and Global Wine Markets by 2025 and Implications of Brexit

Figure 6

Value of Wine Exports and Imports Key Wine-Trading Countries 2014 and Projected Baselinefor 2025 (2014 US$ million)

Fig6-BW

onlin

eBW

inprint

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 235

Figure 7

National Shares of Global Wine Import Value and Volume 2014 and Projected Baselinefor 2025 ()

Fig7-BW

onlin

eBW

inprint

Source Authorsrsquo model results

236 UK and Global Wine Markets by 2025 and Implications of Brexit

A Initial Impact of Brexit

Because the British pound dropped by one-sixth against the US dollar in the fourmonths following the Brexit vote in June 2016 and because the UnitedKingdomrsquos average real wage fell 1 between November 2016 and April 2017(according to the latest available data see Office of National Statistics 2017) weviewed our ldquolargerdquo scenario as more likely soon after the Brexit vote But wecompare those results with ones from our ldquosmallrdquo initial-impact scenario becauseit has become more likely following the June 2017 general election that ldquosofterrdquooptions will be considered by the new government Generally the results areabout half the size in the ldquosmallrdquo scenario with the exception of the bilateraltrade effects To show the sensitivity of results to our assumptions we point outthe differences when the ldquosmallrdquo results are not close to half the results shown forthe ldquolargerdquo scenario

Before turning to those results we first present a ldquofree-traderdquo scenario whichassumes the United Kingdom chooses the radical option of becoming the HongKong of Europe Unlikely as this scenario is the results provide assurance thatour global wine model generates the standard types of effects of exiting a customsunion In this ldquofree-traderdquo scenario the United Kingdom is assumed to move tozero tariffs on all wine imports Thus the current preferences on EU Chileanand South African wine imports disappear as all other wine exporters enjoy thesame free access to UK wine markets We assume in this scenario that Brexit hasno impact on the United Kingdomrsquos real income or consumption growth rates by2025 and that by then the British pound has returned to its 2014 value relative toother currencies (similar to the real exchange rate assumed in our baseline scenarioto 2025 see Appendix Table A1)7 Under these assumptions EU Chilean andSouth African wine exports to the United Kingdom in 2025 are 003 lower thanin the no-Brexit baseline and UK imports from other wine exporters are 025higher but the United Kingdomrsquos total wine imports are hardly any different(because local wine prices fall by only 05 and incomes are unchanged) In this sce-nario UK consumers expand their volume of consumption by just 05 while theEU27 exports more and other countries export less to the rest of the world such thatworld wine exports are almost unchanged In short all these effects are in the direc-tion that comparative static-customs-union theory would predict but they are smallbecause we assume the United Kingdomrsquos real income and currency are the same asin our baseline for 2025 as described in the previous section

In the ldquolargerdquo alternate initial scenario involving Brexit as compared with thebaseline scenario to 2025 the consumer price of wine in 2025 is 22 higher in the

7If this scenario were seen as a credible long-run trade policy it would encourage more investment thatwould eventually raise UK incomes but not without major structural changes that would initiallydisrupt the economy Implicitly we assume in our free-trade scenario that these two opposite impactson real UK incomes and the British pound exactly offset each other by 2025

Kym Anderson and Glyn Wittwer 237

United Kingdom in local currency terms (20 because of real depreciation of theBritish pound 4 because of new tariffs on EU Chilean and South Africanwines and ndash2 because of slower UK income growth) The volume of UKwine consumption is 28 lower 16 because of slower UK economic growth7 because of real depreciation of the British pound and 5 because of newtariffs Superpremium still-wine sales are the most affected dropping by two-fifths while sparkling and commercial-premium wines drop a bit less than one-quarter Because the average price rises by more than the decline in volume soldthe aggregate value of UK sales even in local currency terms falls under thisldquolargerdquo Brexit scenario Under the ldquosmallrdquo Brexit scenario the consumer price ofwine in 2025 is 11 higher in the United Kingdom and its volume of wine consump-tion is 17 lower

The volume of projected UK imports in 2025 is 427 ML or nearly one-quarterless in the ldquolargerdquo scenario than in the baseline scenario comprising 58ML less spar-kling wine 31 ML less superpremium still wine and 339 ML less commercial-premium wine World imports are lower by just 239 ML because imports byother countries are 189 ML higher in response to the lower international wineprices in this scenario In value terms UK imports are $175 billion (or 27)lower in 2025 because of ldquolargerdquo Brexit $113 billion because of lower incomes$038 billion because of the fall in the British pound and $014 billion because ofthe rise in wine-import tariffs (Table 4) These aggregate trade impacts are a littlemore than half this size under the ldquosmallrdquo Brexit scenario

Despite the import levels falling because of raised import tariffs domestic con-sumption of all three quality categories of UK-produced wine is lower with thanwithout Brexit because of shrunken demand for all wines resulting from loweredUK incomes and raised local prices because of devaluation of the British poundThe British poundrsquos devaluation does make it easier for the United Kingdom tosell wines abroad though The countryrsquos exports are 7 ML or nearly 5 higherin 2025 in the ldquolargerdquo Brexit scenario and UK production is 3 higher ThoseUK exports (or re-exports of imported bulk wine after it is bottled in the UnitedKingdom) that go to EU27 countries are reduced though because of the tariffnow imposed at the new EU border

Without Brexit the United Kingdomrsquos shares of global wine imports are slightlyhigher in volume terms in 2025 than in 2010ndash2015 but 2 percentage points lower invalue terms thanks to East Asiarsquos expanding demand for imports of premium winesWith ldquolargerdquo Brexit however that value share is an additional 2 percentage pointslower and the volume share is almost 5 points lower (Figure 8) The net effects ofthese impacts on global trade are shown in Figure 9 Most of the initial tradeeffect of Brexit is a large decline in net imports of wine by the United Kingdomwith very little offsetting positive effect on trade in the rest of the world Theldquosmallrdquo Brexit numbers are a bit more than half of those for ldquolargerdquo Brexit

238 UK and Global Wine Markets by 2025 and Implications of Brexit

Table4

Difference

in2025

Projected

VolumeandValue

ofWineIm

portsb

ytheUnitedKingdom

andtheResto

fthe

World

asaConsequence

oftheInitial

BrexitS

hock

(MLand2014US$

Million

ldquoLargerdquoSc

enario)

Volum

e(M

L)

Value

(US$

million)

NP+CPa

SuperPrb

Sparkling

TOTAL

NP+CPa

SuperPrb

Sparkling

TOTAL

ΔUK

impo

rtsdu

eto

Low

erincomes

minus198

minus20

minus29

minus247

58minus644

minus253

minus234

minus1131

65Low

erpo

und

minus70

minus10

minus14

minus93

22minus248

minus127

minus102

minus476

27Highertariffs

minus71

minus1minus16

minus87

20minus110

minus8minus2

4minus143

8TOTA

Lminus339

minus31

minus58

minus427

100

minus1001

minus388

minus360

minus1750

100

difffrom

base

2332

3325

2432

3227

of

totalc

uts

797

14100

5722

21100

ΔROW

netim

ports

143

2125

189

230

143

minus181

192

ΔWORLD

TRADE

minus195

minus10

minus34

minus239

minus763

minus246

minus543

minus1552

aNon

prem

ium

plus

commercial-premium

still

wines

bSu

perpremium

still

wines

SourceA

utho

rsrsquomod

elresults

Kym Anderson and Glyn Wittwer 239

The aggregate effect of ldquolargerdquo Brexit on the market shares of various wine-export-ing countries in the United Kingdom is almost indiscernible even with one decimalpoint The projected 2025 shares are quite different from the actual 2014 shares forseveral countries They are much smaller in 2025 for South Africa Australia andNew Zealand (and the United States in volume terms) and are much larger involume for Spain and in value for Italy These results occur because wine-exportingcountries benefit differentially from the varying rates of growth in net importdemand for wine in non-UK countries over this projection period The most impor-tant projected changes are the increase in the real value of annual wine importsbetween 2014 and 2025 by China (200 or $3 billion) Other Asia (110 or $22billion) and Africa (270 or $16 billion) More than half of Australiarsquos increasein annual exports from 2014 to 2025 go to Asia and more than half of SouthAfricarsquos increase in exports go to Other Africa

Table 5 reveals that ldquolargerdquo Brexit lowers EU Chilean and South African wineexports by 150 ML or US$12 billion in the case of the European Union with someof their exports diverted from the United Kingdom to the EU27 and other marketsin competition with New World exporters Although the United States Australiaand Argentina sell only a little less to the United Kingdom they also sell less toother countries For Chile and South Africa which lose their preferential access toUK (but not to EU27) markets in this Brexit scenario some exports are redirected

Figure 8

UK Shares of World Wine Imports 2010ndash2015 and Projected to 2025 without and with theldquoLargerdquo Initial Brexit Shock ()

Fig8-BW

onlin

eBW

inprint

Sources Anderson and Pinilla (2017) and authorsrsquo model results

240 UK and Global Wine Markets by 2025 and Implications of Brexit

from the United Kingdom to EU27 countries ndash but again they export less overallGlobal wine trade in 2025 is 240 ML (19) or $18 billion (35) less under thisldquolargerdquo Brexit scenario The percentage by which wine exportersrsquo trade shrinks isgreater for values than for volumes because of changes in relative prices of differ-ent-quality wines Those differences are shown in the numbers in parentheses inTable 5

Three other points are worth making about Table 5 One is that Australia sellsslightly more to the United Kingdom in the ldquosmallrdquo Brexit scenario rather thanslightly less as in the ldquolargerdquo Brexit case Evidently the negative income and price(devaluation) effects do not more than offset the positive trade-diverting effect onAustralian exports to the United Kingdom of removing preferences in the ldquosmallrdquoscenario Second New Zealand sells slightly more to non-UK countries underBrexit despite greater competition from the EU27 Chile and South Africa Thisanomaly is due to changes in the relative prices of different qualities of wine inglobal wine markets bearing in mind that New Zealand has the worldrsquos highestaverage price for still-wine exports And third the value (but not the volume) ofexports of ldquoOtherrdquo countries to markets other than the United Kingdom arehigher under Brexit This result too is due to changes in the relative prices of differ-ent qualities of wine in global wine markets

Figure 9

Difference in 2025 Wine Import Volumes and Values as a Result of the ldquoLargerdquo Initial BrexitShock (ML and US$ Million in 2014 US dollars)

Fig9-BW

onlin

eBW

inprint

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 241

B Subsequent Impact of Brexit from a UK-EU27 FTA

The next-most-likely step in the Brexit process is for the United Kingdom to nego-tiate a new trade arrangement with the EU27 We therefore assume that a UK-EU27FTAwith free bilateral wine trade is implemented and adjusted to by 2025 and thatprogress toward that end occurs soon enough that the adverse macroeconomicshocks from the initial impact of uncertainty over the Brexit process are confinedto those assumed in the ldquosmallrdquo scenario outlined above In this subsequent scenariowe assume that the British pound returns to the value it would have reached in 2025in the absence of Brexit (ie reversing the 10 devaluation assumed in the ldquosmallrdquoinitial Brexit scenario) and that real incomes in the United Kingdom are 8 higher

Table 5Difference in 2025 BilateralWine Import Volumes and Values fromKey Exporters by the United

Kingdom and the Rest of the World (RoW) as a Result of Initial Brexit Shock(ML and 2014US$ Million)a

(a) ldquolargerdquo scenario

Volume (ML) Value (2014US$ million)

UK RoW WORLD () UK RoW WORLD ()

EU27 minus287 136 minus150 (ndash17) minus1187 minus5 minus1192 (ndash31)Chile minus59 35 minus25 (ndash30) minus169 31 minus138 (ndash48)South Africa minus53 35 minus18 (ndash32) minus105 20 minus85 (ndash67)USA minus7 minus6 minus13 (ndash24) minus75 minus40 minus115 (ndash50)Australia minus4 minus3 minus7 (ndash09) minus25 minus65 minus90 (ndash30)Argentina minus3 minus9 minus12 (ndash48) minus16 minus39 minus55 (ndash52)New Zealand minus11 9 minus2 (ndash09) minus162 71 minus91 (ndash43)Others minus2 minus10 minus12 (ndash02) minus11 minus52 minus63 (ndash44)WORLD minus427 187 minus240 (ndash19) minus1750 minus79 minus1829 (ndash35)

(a) ldquosmallrdquo scenario

Volume (ML) Value (2014US$ million)

UK RoW WORLD () UK RoW WORLD ()

EU27 minus178 82 minus96 (ndash12) minus692 minus43 minus736 (ndash19)Chile minus46 28 minus18 (ndash24) minus128 36 minus91 (ndash32)South Africa minus43 29 minus14 (ndash42) minus82 23 minus59 (ndash47)USA 1 minus6 minus5 (ndash11) minus23 minus28 minus51 (ndash22)Australia 5 minus10 minus5 (ndash06) 19 minus56 minus38 (ndash13)Argentina 0 minus6 minus6 (ndash26) minus3 minus25 minus29 (ndash27)New Zealand minus5 4 minus1 (ndash06) minus80 34 minus46 (ndash22)Others 0 minus9 minus9 (ndash01) minus1 minus33 minus34 (ndash24)WORLD minus266 112 minus154 (ndash13) minus991 minus92 minus1083 (ndash21)

a Numbers in parentheses are the percentage difference between the Brexit and baseline scenarios for 2025 projected wine-import volumes orvalues by source

Source Authorsrsquo model results

242 UK and Global Wine Markets by 2025 and Implications of Brexit

in 2025 than in the ldquosmallrdquo scenario ndash that is they regain two-thirds of the value theywould have reached without Brexit-related uncertainty

This subsequent development in the Brexit process reverses most of but not allthe initial effects of Brexit by 2025 because of our assumption that the lost growthin the initial years of uncertainty following the Brexit vote are only partly recoveredby 2025 following the implementation of a UK-EU27 FTA Moreover the longer ittakes before this FTA is finalized and implemented the longer the estimated initialadverse effects persist and the larger the cumulative cost of Brexit to UK wine con-sumers and to grape and wine producers in wine-exporting countries

Table 6 summarizes the subsequent trade effects for 2025 It suggests that all butone-tenth of the loss in value of world trade in wine from the initial ldquosmallrdquo impact isrestored most of it because of a smaller reduction in the United Kingdomrsquos wineimports following the FTA Most of that improved outcome is because of recoveredimports from the EU27 commensurate with the latterrsquos high share of UK importsEven though Chile and South Africa are assumed in this scenario to have not yetsigned an FTA with the United Kingdom they export slightly more to the UnitedKingdom (and even more to all other countries) than in the initial ldquosmallrdquo scenario

Within the United Kingdom this FTA brings down the local currency consumerprice of wine by 9 largely offsetting the 11 rise in the initial ldquosmallrdquo Brexit sce-nario and it raises the volume consumed in the United Kingdom by 18 fully off-setting the 17 fall in the initial ldquosmallrdquo Brexit scenario

Table 6Difference in 2025 Bilateral Wine-Import Volumes and Values from Key Exporters by the

United Kingdom and the Rest of the World (RoW) as a Result of Implementing a UK-EU27FTA Following Initial ldquoSmallrdquo Brexit Shock (ML and 2014US$ Million Difference Relative to

Initial ldquoSmallrdquo Brexit Scenario)a

Volume (ML) Value (2014US$ million)

UK RoW WORLD (ldquosmallrdquo)a UK RoW WORLD (ldquosmallrdquo)a

EU27 212 minus112 100 (minus96) 750 0 750 (minus806)Chile 1 3 4 (minus18) 3 40 43 (minus92)South Africa 3 2 5 (minus14) 3 18 21 (minus60)USA 3 3 6 (minus5) 26 21 47 (minus50)Australia 1 1 2 (minus5) 8 38 46 (minus33)Argentina 1 4 5 (minus6) 4 20 24 (minus28)New Zealand 4 minus3 1 (minus1) 61 minus25 36 (minus46)Others 1 6 7 (minus9) 5 minus46 minus41 (92)WORLD 226 minus96 130 (minus154) 860 66 926 (minus1022)

a Numbers in parentheses are the world trade differences between the ldquosmallrdquo initial Brexit scenario and the baseline scenario copied fromcolumns 3 and 7 of Table 5(b)

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 243

In short this subsequent step in the Brexit process can be expected to restore muchof the initial adverse effects in the United Kingdom of the Brexit vote by 2025 Thewine exporters in the EU27 Chile and South Africa also lose less in this scenarioas compared with the initial ldquosmallrdquo scenario while other key exporters sell nearlyas much in 2025 in this scenario as they are projected to in the baseline ndash that isthey are beneficiaries of the reduced discrimination in global wine markets in thiscase where Chile and South Africa no longer enjoy preferential access to the UKmarket (and even though EU27 producers are again allowed to do so) Keep inmind though that although the estimated losses in the initial scenarios are reportedjust for 2025 they are expected to be felt throughout the preceding years of uncertaintyTherefore the longer it is before uncertainty abates and the United Kingdomrsquos currentpreferential trading arrangements are replaced by newones such as a UK-EU27 FTAthe higher the cumulative cost of the Brexit vote to the wine trade

VII Caveats and Conclusions

The above Brexit simulations are just a few of many scenarios that could be modeledObvious additional ones could also assume that FTAs are reached between theUnited Kingdom and other trading partners including Chile and South Africaand such countries as Australia and New Zealand Some of these other countriesare already exploring the ECrsquos invitation (see European Commission 2015) to nego-tiate an FTA with what will be the EU27 The sequence in which FTAs are signedand the speed with which they are implemented will matter (as was also the casewith the sequential signing over the past decade of bilateral FTAs with NortheastAsian countries by Chile Australia and New Zealand see Anderson andWittwer 2015) Even if the United Kingdom were able to sign additional bilateralFTAs and begin implementing them before 2025 it would have little effect on theabove results (because wine tariffs are such a minor contributor to them) unlessthose FTAs were to accelerate the United Kingdomrsquos economic growth and therise in the British poundrsquos value before 2025

We assume above that no changes are made to alcohol excise duties in the UnitedKingdom following Brexit even though they are scheduled to be progressively raisedwith inflation and may be raised even more for wine relative to spirits to offset theopposite effects of Brexit on those two domestic industries Nor do we make anyallowance for UK increases in consumer wine prices that may be needed to coverthe higher cost of clearing customs on imports from EU27 countries (which maybe more or less offset by increased smuggling and duty-free purchasing)Importers of fine wines including individuals who buy Bordeaux en primeur maywell reduce their demand because of concerns that delays on docks will affect thequality of their wines or because of greater currency uncertainty

To summarize Brexit is costly initially to UK consumers of wine (and of manyother tradable products) because the domestic retail price in local currency tax-

244 UK and Global Wine Markets by 2025 and Implications of Brexit

inclusive terms is 22 higher than otherwise in 2025 and the volume of wine con-sumed domestically is 28 lower in the ldquolargerdquo Brexit scenario (or 11 higher pricesand 17 lower quantities in the ldquosmallrdquo case) by 2025 unless a UK-EU27 FTA isdeveloped before then Even if such an FTA does get signed ratified by all 28 par-liaments and implemented by 2025 the slower income growth in the interim meansa smaller UK wine market in 2025 than would otherwise have been the case Thevolume reduction is a blow to many participants in UK wine-bottling transportingstoring wholesaling and retailing businesses in addition to restaurants and pubsVery little of that initial impact occurs because of higher import tariffs most impor-tant is the assumed fall in UK real incomes

The small but growing number of local UK vignerons (who supply less than 05of all domestic-wine sales and less than 4 of sparkling-wine sales) is initially pro-jected to sell less wine domestically because of reduced demand and to export only alittle extra abroad while the British pound is devalued Overall their production isonly 3 higher in 2025 even in the ldquolargerdquo Brexit scenario (not taking intoaccount the fact that their casual labor is likely to be more expensive under Brexitbecause of tighter restrictions on immigration) and their production diminishes ifand when a UK-EU27 FTA is in place

Even with a UK-EU27 FTA in place EU27 wine exporters export US$56 millionless wine in 2025 thanks to Brexit Chile and South Africa export $88 million lesswine and wine exports from the rest of the world are little different in aggregatealthough slightly larger from Australia

Clearly there will be great uncertainly for some time over the possible policy out-comes to flow from Brexit and their consequent sequential impacts on UK house-hold disposable incomes foreign exchange rates and bilateral wine tariffsMeanwhile the above projections under explicit assumptions provide some idea ofhow wine markets might be affected by the most-likely first two stages of theBrexit process (agreeing on a new tariff schedule at the WTO and agreeing to andimplementing a UK-EU27 FTA) In particular they make clear that nontrivialinitial impacts could affect the domestic wine market impacts that are likely to belarger than just the direct impact of changes in bilateral tariffs If the UnitedKingdom succeeds in getting countries to agree to transition arrangements thatdelay the changes in tariffs until new FTAs are signed and ratified the initialeffects will be less dramatic than in our first scenarios but it remains to be seenwhether any such agreements can be reached In any event the net effect of Brexiton the welfare of the worldrsquos consumers and producers of wine as a whole will benegative not just initially but permanently unless new trade policy commitmentsby the United Kingdom with major wine-exporting countries are sufficiently moreliberal than current arrangements

Kym Anderson and Glyn Wittwer 245

Appendix Key Equations in the Global Wine Markets Model

Amodel of the worldrsquos wine markets was first published byWittwer et al (2003) andmuch improved by Anderson and Wittwer (2013) Several significant enhancementshave been made to that original model Wine types have been disaggregated intomore types ndash namely nonpremium (including generic bulk) commercial-premiumand superpremium still wines and sparkling wines8 As in the original modelthere are two types of grapes premium and nonpremium Nonpremium wines usenonpremium grapes exclusively superpremium wines use mainly premium grapesand commercial-premium and sparkling wines use both types of grapes In theregional dimension the number of countries and country groups has expandedfrom 10 in the original model to 51 44 individual nations and 7 compositeregions The modelrsquos database is calibrated to 2014 for this paper based on datain Anderson et al (2017) trade data for which are downloaded from httpscomtradeunorg The modelrsquos supply-and-demand equations are based on the orig-inal ORANI modelrsquos theory (see Dixon Parmenter Sutton and Vincent 1982) andthe model is implemented using GEMPACK software (Harrison Horridge Jerieand Pearson 2014)

An enhancement of importance to the present study is the inclusion of exchange-rate variables in the model which allow a distinction between price impacts asobserved in local currency units and those observed in 2014 US dollars

In the model the grape and wine sectors minimize costs of intermediate inputssubject to weak constant elasticity of substitution (CES) substitutability betweeninputs By assumption no intermediate inputs are imported from other countries9Hence

Xcid frac14 f ethX1id CESfrac12P

cid=P1id $THORN eth1THORN

P1id X1id frac14X

c

Xcid P

cid eth2THORN

whereXcid is the quantity demanded of commodity c by grape or wine industry i in

region d Pcid is the corresponding price and X1id and P1id are the respective inter-

mediate composite quantities and prices

Two primary factors are employed in the sector labor (the quantity of which isendogenous with perfectly elastic supply) and capital Capital is usually treated asexogenous in quantity with rates of return bearing all the adjustment in thevarious scenarios This reflects the fact that grapes (a perennial crop) and wine-

8Commercial-premium still wines are defined by Anderson and Nelgen (2011) to be those priced betweenUS$250 and $750 per litre pre-tax at a countryrsquos border or wholesale9An exception concerns bulk wine imports used mainly in the bottling of commercial premium wine in theUK USA Canada Japan and Germany

246 UK and Global Wine Markets by 2025 and Implications of Brexit

plant capacity adjust slowly to market signals

Lid frac14 f ethFid CESfrac12Wid=PFid $THORN eth3THORN

Kid frac14 f ethFid CESethRid=PFid $THORN eth4THORN

PFid Fid frac14 Lid Wid thorn Kid Rid eth5THORN

Grape and wine producers are assumed to minimize costs subject to CES substitu-tion between capital and labor Equations (3) to (5) show primary factor demandsfor the labor composite Lid and capital Kid subject to a composite factor demandFid by industry i in region d The factor prices are Wid for labor Rid for capitalrentals and PFid for composite factor prices

Appendix Table A1Cumulative Consumption and Population Growth Rates and Changes in the Real Exchange Rate

(RER) Relative to the US Dollar 2014 to 2025 without Brexit ()

Aggregateconsumption Poprsquon RER

Aggregateconsumption Poprsquon RER

France 18 4 minus11 Australia 35 11 minus17Italy 11 2 minus9 New Zealand 32 9 minus26Portugal 14 0 minus9 Canada 27 8 minus18Spain 26 8 minus9 United States 31 8 0Austria 19 4 minus7 Argentina 7 10 109Belgium 20 7 minus9 Brazil 16 8 minus29Denmark 22 2 minus9 Chile 55 8 minus2Finland 21 3 minus7 Mexico 42 12 minus8Germany 14 minus2 minus11 Uruguay 45 3 1Greece 22 minus1 minus14 Other Latin America 60 10 minus5Ireland 42 12 minus9 South Africa 36 12 minus1Netherlands 21 4 minus9 Turkey 50 8 20Sweden 24 9 minus13 North Africa 53 11 0Switzerland 18 8 minus6 Other Africa 109 18 84United Kingdom 32 6 1 Middle East 52 18 minus12Other West Europe 21 10 minus1 China 79 3 5Bulgaria 41 minus7 7 Hong Kong 42 3 2Croatia 20 minus2 minus1 India 134 13 17Georgia 35 0 23 Japan 11 minus3 minus24Hungary 25 minus3 minus11 Korea 38 1 minus9Moldova 49 minus11 13 Malaysia 62 15 minus16Romania 45 minus4 22 Philippines 75 18 7Russia 18 minus2 minus8 Singapore 44 21 minus22Ukraine 22 minus5 14 Taiwan 29 1 minus13Other East Europe 40 minus5 48 Thailand 47 3 minus9

Other Asia 99 10 10

Source Authorsrsquo compilation from projections by various international agencies and from global economy-wide modeling by Anderson andStrutt (2016)

Kym Anderson and Glyn Wittwer 247

The composite factor demand Fid is proportional to total output Qid subject to aprimary factor using technology Aid Hence

Fid frac14 Qid Aid eth6THORN

The perfectly competitive zero-pure-profit condition is that total revenue valued atthe output price P0s

i multiplied by Qid equals the total production cost

P0si Qid frac14

X

cPcid X1cid thorn

X

oWo

id Loid thorn Rid Kid eth7THORN

Household demands follow a linear expenditure system in each region We reducethe optimizing problem for household consumption of each commodity subjectto a budget constraint to equations describing subsistence and discretionarydemands Aggregate subsistence expenditure WSUBd depends only on consumerprices P3cd for each commodity and the number of households N as per-capita sub-sistence quantities XSUBcd subject to given preferences are constant

WSUBd frac14X

c

P3cd XSUBcd Nd eth8THORN

Discretionary expenditures for each commodity (the left-hand side of equation (9))are equal to the marginal budget share (βcd) of aggregate discretionary expenditureThis aggregate is the bracketed term on the right-hand side of equation (9) whereW3TOTd is aggregate nominal expenditure

P3cdethX3cd XSUBcd NdTHORN frac14 βcdethW3TOTd WSUBdTHORN eth9THORN

Because real aggregate consumption is usually exogenous in our partial equilibriumsimulations the linear expenditure system determines the consumption shares ofindividual final commodities (ie the five wine types plus a composite of all otherconsumption items) driven by changes in relative prices as faced by domestic con-sumers The income elasticity of demand for each commodity is equal to the mar-ginal budget share divided by the expenditure share This number varies from 05for nonpremium wine to 17 for superpremium still wine The income elasticity ofdemand for other consumption is very close to 10 because wine accounts for anaverage of only 03 of aggregate expenditures globally and no more than 11in any country (Anderson and Nelgen 2011 Table 166)

A feature of our revised model of world wine markets is the inclusion of nominalexchange rates These appear directly in the equation linking retail prices (P3scd ) toproducer prices by country of origin (P0s

c ) where c denotes the wine type

P3 scd frac14 P0s

cfd

fsTtarcd T

taxcd thorn Pm

cd eth10THORN

248 UK and Global Wine Markets by 2025 and Implications of Brexit

The exchange rates in the consuming and producing regions are fd and fs respec-tively expressed as local currency units per $US Ttar

cd is the power of the tariff in theconsuming region and Ttax

cd is the power of the domestic consumption (or excise) taxprior to any generic value-added or goods-and-services tax Pm

cd is the price of marginm assumed to be locally supplied nontradable and therefore unaffected by theexchange rate

A given level of consumption for wine type c (X3cd) is satisfied using theArmington (1969) assumption in which wines from different countries of originare imperfectly substitutable First domestic wine is imperfectly substitutable witha composite of imports

X3sscd frac14 f ethX3cd CESethP3sscd=P3cdTHORNTHORN ss frac14 domestic imports eth11THORN

and then imports by origin (X3scd ) are determined in a second CES equation

X3scd frac14 f ethX3ssfrac14importscd CESethP3scd=P3

ssfrac14importscd THORNTHORN eth12THORN

The model enables us to show how changes in international competitiveness affectthe worldrsquos wine markets A crucial part of this exercise is explaining how pricesdetermined outside the grape and wine markets influence these markets Becausethe model is in partial equilibrium to depict the impacts of changes in internationalcompetitiveness outside price changes need to be imposed as shocks on the modelThe price of intermediate inputs other than grapes and wine shown in equations (1)and (2) is set equal to the price of GDP (Pg

d ) multiplied by a shifter Fcd

Pcid frac14 Fc

dPgd eth13THORN

If no specific price observations are available the shifter Fcd remains exogenous and

unshocked with the change in price being determined by a shock to the price ofGDP If observations are available for specific input price movements the shifterFc

dbecomes endogenous with Pc

id now exogenous and shocked

Wid frac14 Fwd P

gd eth14THORN

Wage rates are treated similarly In equation (14) if the wage shifter Fwd is exogenous

changes in wage rates Wid are determined by changes in the price of GDP If wagerate data are available Fw

d becomes endogenous andwage rates are shocked directly

Pmcd frac14 Fm

d Pgd eth15THORN

The prices of trade and transport margins are also determined by the price of GDP ifthe shifter Fm

d in equation (15) is exogenous

Changes in international competitiveness depend on changes in relative pricelevels and changes in nominal exchange rates In equation (16) fR

s denotes real

Kym Anderson and Glyn Wittwer 249

exchange-rate movements relative to the US dollar

fRs frac14 Pg

s =frac12PgUSA

fs$ eth16THORN

In equation (16) the nominal exchange rate for the United States is alwaysunchanged because nominal and real exchange rates are expressed relative to theUS currency

Changes in international market conditions may have impacts in one direction onproducer prices as expressed in US dollars P0s

i and potentially in the opposite direc-tion in local currency units Hence we calculate real producer pricesP0s

iloc in localcurrency terms (ie the price most relevant to domestic producers)

P0siloc frac14 P0s

ifs=P

gs eth17THORN

To obtain real price changes in local currency terms we convert US dollar prices(P3 s

cd for the source-specific price and P3cd for the source-composite price) to reallocal currency prices (P3 s

cdloc and P3cdloc) using the consumer price index (CPI)(Pc

d ) as the deflator

P3 scdloc frac14 P3 s

cdfd=P

cd eth18THORN

and

P3cdloc frac14 P3cdfd=Pcd eth19THORN

References

Anderson K and Nelgen S (2011) Global Wine Markets 1961 to 2009 A StatisticalCompendium Adelaide University of Adelaide Press Also available as an e-book atwwwadelaideeduaupresstitlesglobal-wine

Anderson K Nelgen S and Pinilla V (2017) Global Wine Markets 1860 to 2016 AStatistical Compendium Adelaide University of Adelaide Press (forthcoming) Also tobe available as an e-book at wwwadelaideeduaupress

Anderson K and Pinilla V (with the assistance of A J Holmes) (2017)Annual Database ofGlobal Wine Markets 1835 to 2016 Wine Economics Research Centre University ofAdelaide to be posted in October at wwwadelaideeduauwine-econdatabases

Anderson K and Strutt A (2016) Impacts of Asiarsquos rise on African and Latin Americantrade Projections to 2030 World Economy 39(2) 172ndash194

Anderson K andWittwer G (2013) Modeling global wine markets to 2018 Exchange ratestaste changes and Chinarsquos import growth Journal of Wine Economics 8(2) 131ndash158

Anderson K and Wittwer G (2015) Asiarsquos evolving role in global wine markets ChinaEconomic Review 35 1ndash14

250 UK and Global Wine Markets by 2025 and Implications of Brexit

Armington P A (1969) A theory of demand for products distinguished by place of produc-tion IMF Staff Papers 16 159ndash178

Baldwin R E (ed) (2016) Brexit Beckons Thinking Ahead by Leading Economists LondonCentre for Economic Policy Research

Campos N F (2016) Lousy experts Looking back at the ex ante estimates of the costs ofBrexit In R E Baldwin (ed) Brexit Beckons Thinking Ahead by Leading Economists(ch 3) London Centre for Economic Policy Research

Dixon P Parmenter B Sutton J and Vincent D (1982) ORANI A Multisectoral Modelof the Australian Economy Contributions to Economic Analysis 142 Amsterdam North-Holland

European Commission (2015) Trade for All Towards a More Responsible Trade andInvestment Policy Brussels European Commission

Harrison J Horridge M Jerie M and Pearson K (2014) GEMPACK ManualMelbourne GEMPACK Software

HM Treasury (2017) Spring Budget 2017 London HM TreasuryHolmes A J and Anderson K (2017a)Annual Database of National Beverage Consumption

Volumes and Expenditures 1950 to 2015 Wine Economics Research Centre University ofAdelaide posted at wwwadelaideeduauwine-econdatabases

Holmes A J and Anderson K (2017b) Convergence in National Alcohol ConsumptionPatterns New Global Indicators Journal of Wine Economics 12(2) 117ndash148

Ludington C C (2013) The Politics of Wine in Britain A New Cultural History BasingstokeUK Palgrave Macmillan

Office of National Statistics (2017) Analysis of real earnings London Office of NationalStatistics Available at httpswwwonsgovukemploymentandlabourmarketpeopleinworkearningsandworkinghoursarticlessupplementaryanalysisofaverageweeklyearningslatest

Rollo J Borchert I Dawar K Holmes P and Winters L A (2016) The World TradeOrganisation A safety net for a post-Brexit UK trade policy Briefing Paper 1UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-1-final-1pdf

Smith A (2017) Brexit Hard truths and hard choices UKTPO blog 19 June Available athttpsblogssussexacukuktpo20170619brexit-hard-truths-and-hard-choices

Swinbank A (2017) World trade rules and the policy options for British agriculture post-Brexit Briefing Paper 7 UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-7pdf

Viner J (1950) The Customs Union Issue New York Carnegie Endowment for InternationalPeace

Wine Australia (2015) Export Market Guide European Union Adelaide Wine AustraliaWittwer G Berger N and Anderson K (2003) A model of the worldrsquos wine markets

Economic Modelling 20(3) 487ndash506

Kym Anderson and Glyn Wittwer 251

Page 12: U.K. and Global Wine Markets by 2025, and Implications of ... · U.K. and Global Wine Markets by 2025, and Implications of Brexit* Kym Andersona and Glyn Wittwerb Abstract The United

V Projecting Global Wine Markets to 2025

Global wine production and exports are projected in the baseline from 2014 to 2025consistent with past trends The modelrsquos global volume of production (and consump-tion) rises little over that 11-year period (9) made up of a 6 decline in nonpre-mium wine and a one-sixth rise in commercial-premium and superpremium wineIn real (2014 US$) value though global wine output and consumption increaseby about 50 in total and 60 in the two premium categories The internationaltrade projections are similar although a little larger with the share of global wineproduction exported (equivalent to the share of global consumption imported)rising 2 percentage points between 2014 and 2025

The baseline projection does not greatly alter the 2014 shares of various countriesin global wine production apart from China because we assume vineyard expansionthere is faster than elsewhere6 In value terms this assumption means that Chinamoves from fifth to fourth by 2025 behind France the United States and ItalySpain remains barely ahead of Australia and they and Germany take the nextthree places (Figure 5(a)) In total wine-production volume terms China movesfrom sixth to fifth place and Argentina drops from fifth to eighth (and fromeighth to ninth in value terms)

When their products are subdivided into fine wines (superpremium still plus spark-ing) commercial-premium wines and nonpremium wines France and the UnitedStates retain the highest two places on the global ladder for fine-wine productionand Spain and Italy retain the top two places for nonpremium wine As for commer-cial-premium wine production (defined to cost between US$250 and US$750 perliter pretax at a countryrsquos wholesale level or national border) Italy retains the topranking over our projections period but at least in terms of value China challengesFrance for second place

The country rankings by projected value of total wine consumption change some-what more than those for production by 2025 with China taking second place afterthe United States ahead of France and Germany and then the United Kingdomslightly overtaking Italy to slip into fifth place (Figure 5(b)) The United StatesFrance and Germany retain the top three rankings for consuming fine wine butCanada slightly overtakes Italy for fourth place in terms of value at least In thecase of commercial-wine consumption China strengthens its number-one positionahead of the United States and the United Kingdom does likewise vis-agrave-visGermany for third place

As for the projected changes in consumption volumes China is projected to dom-inate the increase in aggregate although the United States is projected to lead the

6 In fact Chinarsquos wine production fell steadily between 2012 and 2016 by a total of one-sixth so Chinarsquoswine imports may grow faster in practice than in this baseline projection

232 UK and Global Wine Markets by 2025 and Implications of Brexit

Figure 5

Value of Wine Production and Consumption in Key Countries 2014 and Projected Baseline for2025 (2014 US$ Million at WineryWholesale Pretax Prices)

Fig5-BW

onlin

eBW

inprint

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 233

increase in consumption of fine wine In Western Europe and in the SouthernHemispherersquos New World countries fine wines are projected to substitute for com-mercial wines (defined as the sum of commercial-premium and nonpremiumwines) with almost no change in total wine consumption Sub-Saharan Africa isthe next region that is projected to take off with its growth accounting for morethan one-third of the rest of the worldrsquos increase in volume consumed

Those differences in production versus consumption rankings are reflected ininternational trade Figure 6 shows that France Italy and Spain remain the threedominant exporters of wine in aggregate value but the rankings of the next fewchange with Australia being slightly ahead of Chile and the United StatesGermany and New Zealand being nearly tied for sixth place in value termsFrance and Italy are even more dominant in fine-wine exports and remain so by2025 while Italy outranks France in the commercial-premium export categoryand Spain outranks Italy Australia and Chile in the nonpremium export class

Among the importers the United States and the United Kingdom are projected tocontinue to hold the first two places in 2025 in value terms but China moves intothird place slightly ahead of Germany followed well behind by Canada HongKong Belgium-Luxembourg the Netherlands and Japan (Figure 6(b)) OtherAfrica (excluding South Africa) is projected to experience the largest increase inimports among all the other regions followed by Other Asia which becomes asbig as Germany in value terms (Figure 7(a)) In terms of total volume of wineimports Germany and the United Kingdom held the top two shares in 2014 butby 2025 the United Kingdom is projected to be well ahead of Germany (Figure 7(b)) However this projection ignores the effects of Brexit to which we now turn

VI How Might Wine Markets Be Affected by the United Kingdomrsquos Exitfrom the European Union

As mentioned earlier for our two alternative scenarios to capture the initial effects ofBrexit (ldquolargerdquo and ldquosmallrdquo) we assume that following the United Kingdomrsquos exitfrom the European Union the United Kingdomrsquos rate of economic growth wouldbe only one-third or two-thirds as fast for the period to 2025 the British poundwould be 20 or 10 lower in real terms than in our modelrsquos baseline projectionand the United Kingdom would apply the European Unionrsquos external tariff onwine to imports from EU member countries (as part of establishing MFN ratesvia the WTO to then start new bilateral FTA negotiations) In these initial scenarioswe assume the United Kingdom does not implement any new FTAs particularlywith the EU27 Chile and South Africa Following the discussion of those initialresults we present the results of a subsequent scenario that includes the implemen-tation of an FTA between the United Kingdom and the EU27 We assume these aretwo discrete steps with no agreed arrangements to smooth the transition betweenthem

234 UK and Global Wine Markets by 2025 and Implications of Brexit

Figure 6

Value of Wine Exports and Imports Key Wine-Trading Countries 2014 and Projected Baselinefor 2025 (2014 US$ million)

Fig6-BW

onlin

eBW

inprint

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 235

Figure 7

National Shares of Global Wine Import Value and Volume 2014 and Projected Baselinefor 2025 ()

Fig7-BW

onlin

eBW

inprint

Source Authorsrsquo model results

236 UK and Global Wine Markets by 2025 and Implications of Brexit

A Initial Impact of Brexit

Because the British pound dropped by one-sixth against the US dollar in the fourmonths following the Brexit vote in June 2016 and because the UnitedKingdomrsquos average real wage fell 1 between November 2016 and April 2017(according to the latest available data see Office of National Statistics 2017) weviewed our ldquolargerdquo scenario as more likely soon after the Brexit vote But wecompare those results with ones from our ldquosmallrdquo initial-impact scenario becauseit has become more likely following the June 2017 general election that ldquosofterrdquooptions will be considered by the new government Generally the results areabout half the size in the ldquosmallrdquo scenario with the exception of the bilateraltrade effects To show the sensitivity of results to our assumptions we point outthe differences when the ldquosmallrdquo results are not close to half the results shown forthe ldquolargerdquo scenario

Before turning to those results we first present a ldquofree-traderdquo scenario whichassumes the United Kingdom chooses the radical option of becoming the HongKong of Europe Unlikely as this scenario is the results provide assurance thatour global wine model generates the standard types of effects of exiting a customsunion In this ldquofree-traderdquo scenario the United Kingdom is assumed to move tozero tariffs on all wine imports Thus the current preferences on EU Chileanand South African wine imports disappear as all other wine exporters enjoy thesame free access to UK wine markets We assume in this scenario that Brexit hasno impact on the United Kingdomrsquos real income or consumption growth rates by2025 and that by then the British pound has returned to its 2014 value relative toother currencies (similar to the real exchange rate assumed in our baseline scenarioto 2025 see Appendix Table A1)7 Under these assumptions EU Chilean andSouth African wine exports to the United Kingdom in 2025 are 003 lower thanin the no-Brexit baseline and UK imports from other wine exporters are 025higher but the United Kingdomrsquos total wine imports are hardly any different(because local wine prices fall by only 05 and incomes are unchanged) In this sce-nario UK consumers expand their volume of consumption by just 05 while theEU27 exports more and other countries export less to the rest of the world such thatworld wine exports are almost unchanged In short all these effects are in the direc-tion that comparative static-customs-union theory would predict but they are smallbecause we assume the United Kingdomrsquos real income and currency are the same asin our baseline for 2025 as described in the previous section

In the ldquolargerdquo alternate initial scenario involving Brexit as compared with thebaseline scenario to 2025 the consumer price of wine in 2025 is 22 higher in the

7If this scenario were seen as a credible long-run trade policy it would encourage more investment thatwould eventually raise UK incomes but not without major structural changes that would initiallydisrupt the economy Implicitly we assume in our free-trade scenario that these two opposite impactson real UK incomes and the British pound exactly offset each other by 2025

Kym Anderson and Glyn Wittwer 237

United Kingdom in local currency terms (20 because of real depreciation of theBritish pound 4 because of new tariffs on EU Chilean and South Africanwines and ndash2 because of slower UK income growth) The volume of UKwine consumption is 28 lower 16 because of slower UK economic growth7 because of real depreciation of the British pound and 5 because of newtariffs Superpremium still-wine sales are the most affected dropping by two-fifths while sparkling and commercial-premium wines drop a bit less than one-quarter Because the average price rises by more than the decline in volume soldthe aggregate value of UK sales even in local currency terms falls under thisldquolargerdquo Brexit scenario Under the ldquosmallrdquo Brexit scenario the consumer price ofwine in 2025 is 11 higher in the United Kingdom and its volume of wine consump-tion is 17 lower

The volume of projected UK imports in 2025 is 427 ML or nearly one-quarterless in the ldquolargerdquo scenario than in the baseline scenario comprising 58ML less spar-kling wine 31 ML less superpremium still wine and 339 ML less commercial-premium wine World imports are lower by just 239 ML because imports byother countries are 189 ML higher in response to the lower international wineprices in this scenario In value terms UK imports are $175 billion (or 27)lower in 2025 because of ldquolargerdquo Brexit $113 billion because of lower incomes$038 billion because of the fall in the British pound and $014 billion because ofthe rise in wine-import tariffs (Table 4) These aggregate trade impacts are a littlemore than half this size under the ldquosmallrdquo Brexit scenario

Despite the import levels falling because of raised import tariffs domestic con-sumption of all three quality categories of UK-produced wine is lower with thanwithout Brexit because of shrunken demand for all wines resulting from loweredUK incomes and raised local prices because of devaluation of the British poundThe British poundrsquos devaluation does make it easier for the United Kingdom tosell wines abroad though The countryrsquos exports are 7 ML or nearly 5 higherin 2025 in the ldquolargerdquo Brexit scenario and UK production is 3 higher ThoseUK exports (or re-exports of imported bulk wine after it is bottled in the UnitedKingdom) that go to EU27 countries are reduced though because of the tariffnow imposed at the new EU border

Without Brexit the United Kingdomrsquos shares of global wine imports are slightlyhigher in volume terms in 2025 than in 2010ndash2015 but 2 percentage points lower invalue terms thanks to East Asiarsquos expanding demand for imports of premium winesWith ldquolargerdquo Brexit however that value share is an additional 2 percentage pointslower and the volume share is almost 5 points lower (Figure 8) The net effects ofthese impacts on global trade are shown in Figure 9 Most of the initial tradeeffect of Brexit is a large decline in net imports of wine by the United Kingdomwith very little offsetting positive effect on trade in the rest of the world Theldquosmallrdquo Brexit numbers are a bit more than half of those for ldquolargerdquo Brexit

238 UK and Global Wine Markets by 2025 and Implications of Brexit

Table4

Difference

in2025

Projected

VolumeandValue

ofWineIm

portsb

ytheUnitedKingdom

andtheResto

fthe

World

asaConsequence

oftheInitial

BrexitS

hock

(MLand2014US$

Million

ldquoLargerdquoSc

enario)

Volum

e(M

L)

Value

(US$

million)

NP+CPa

SuperPrb

Sparkling

TOTAL

NP+CPa

SuperPrb

Sparkling

TOTAL

ΔUK

impo

rtsdu

eto

Low

erincomes

minus198

minus20

minus29

minus247

58minus644

minus253

minus234

minus1131

65Low

erpo

und

minus70

minus10

minus14

minus93

22minus248

minus127

minus102

minus476

27Highertariffs

minus71

minus1minus16

minus87

20minus110

minus8minus2

4minus143

8TOTA

Lminus339

minus31

minus58

minus427

100

minus1001

minus388

minus360

minus1750

100

difffrom

base

2332

3325

2432

3227

of

totalc

uts

797

14100

5722

21100

ΔROW

netim

ports

143

2125

189

230

143

minus181

192

ΔWORLD

TRADE

minus195

minus10

minus34

minus239

minus763

minus246

minus543

minus1552

aNon

prem

ium

plus

commercial-premium

still

wines

bSu

perpremium

still

wines

SourceA

utho

rsrsquomod

elresults

Kym Anderson and Glyn Wittwer 239

The aggregate effect of ldquolargerdquo Brexit on the market shares of various wine-export-ing countries in the United Kingdom is almost indiscernible even with one decimalpoint The projected 2025 shares are quite different from the actual 2014 shares forseveral countries They are much smaller in 2025 for South Africa Australia andNew Zealand (and the United States in volume terms) and are much larger involume for Spain and in value for Italy These results occur because wine-exportingcountries benefit differentially from the varying rates of growth in net importdemand for wine in non-UK countries over this projection period The most impor-tant projected changes are the increase in the real value of annual wine importsbetween 2014 and 2025 by China (200 or $3 billion) Other Asia (110 or $22billion) and Africa (270 or $16 billion) More than half of Australiarsquos increasein annual exports from 2014 to 2025 go to Asia and more than half of SouthAfricarsquos increase in exports go to Other Africa

Table 5 reveals that ldquolargerdquo Brexit lowers EU Chilean and South African wineexports by 150 ML or US$12 billion in the case of the European Union with someof their exports diverted from the United Kingdom to the EU27 and other marketsin competition with New World exporters Although the United States Australiaand Argentina sell only a little less to the United Kingdom they also sell less toother countries For Chile and South Africa which lose their preferential access toUK (but not to EU27) markets in this Brexit scenario some exports are redirected

Figure 8

UK Shares of World Wine Imports 2010ndash2015 and Projected to 2025 without and with theldquoLargerdquo Initial Brexit Shock ()

Fig8-BW

onlin

eBW

inprint

Sources Anderson and Pinilla (2017) and authorsrsquo model results

240 UK and Global Wine Markets by 2025 and Implications of Brexit

from the United Kingdom to EU27 countries ndash but again they export less overallGlobal wine trade in 2025 is 240 ML (19) or $18 billion (35) less under thisldquolargerdquo Brexit scenario The percentage by which wine exportersrsquo trade shrinks isgreater for values than for volumes because of changes in relative prices of differ-ent-quality wines Those differences are shown in the numbers in parentheses inTable 5

Three other points are worth making about Table 5 One is that Australia sellsslightly more to the United Kingdom in the ldquosmallrdquo Brexit scenario rather thanslightly less as in the ldquolargerdquo Brexit case Evidently the negative income and price(devaluation) effects do not more than offset the positive trade-diverting effect onAustralian exports to the United Kingdom of removing preferences in the ldquosmallrdquoscenario Second New Zealand sells slightly more to non-UK countries underBrexit despite greater competition from the EU27 Chile and South Africa Thisanomaly is due to changes in the relative prices of different qualities of wine inglobal wine markets bearing in mind that New Zealand has the worldrsquos highestaverage price for still-wine exports And third the value (but not the volume) ofexports of ldquoOtherrdquo countries to markets other than the United Kingdom arehigher under Brexit This result too is due to changes in the relative prices of differ-ent qualities of wine in global wine markets

Figure 9

Difference in 2025 Wine Import Volumes and Values as a Result of the ldquoLargerdquo Initial BrexitShock (ML and US$ Million in 2014 US dollars)

Fig9-BW

onlin

eBW

inprint

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 241

B Subsequent Impact of Brexit from a UK-EU27 FTA

The next-most-likely step in the Brexit process is for the United Kingdom to nego-tiate a new trade arrangement with the EU27 We therefore assume that a UK-EU27FTAwith free bilateral wine trade is implemented and adjusted to by 2025 and thatprogress toward that end occurs soon enough that the adverse macroeconomicshocks from the initial impact of uncertainty over the Brexit process are confinedto those assumed in the ldquosmallrdquo scenario outlined above In this subsequent scenariowe assume that the British pound returns to the value it would have reached in 2025in the absence of Brexit (ie reversing the 10 devaluation assumed in the ldquosmallrdquoinitial Brexit scenario) and that real incomes in the United Kingdom are 8 higher

Table 5Difference in 2025 BilateralWine Import Volumes and Values fromKey Exporters by the United

Kingdom and the Rest of the World (RoW) as a Result of Initial Brexit Shock(ML and 2014US$ Million)a

(a) ldquolargerdquo scenario

Volume (ML) Value (2014US$ million)

UK RoW WORLD () UK RoW WORLD ()

EU27 minus287 136 minus150 (ndash17) minus1187 minus5 minus1192 (ndash31)Chile minus59 35 minus25 (ndash30) minus169 31 minus138 (ndash48)South Africa minus53 35 minus18 (ndash32) minus105 20 minus85 (ndash67)USA minus7 minus6 minus13 (ndash24) minus75 minus40 minus115 (ndash50)Australia minus4 minus3 minus7 (ndash09) minus25 minus65 minus90 (ndash30)Argentina minus3 minus9 minus12 (ndash48) minus16 minus39 minus55 (ndash52)New Zealand minus11 9 minus2 (ndash09) minus162 71 minus91 (ndash43)Others minus2 minus10 minus12 (ndash02) minus11 minus52 minus63 (ndash44)WORLD minus427 187 minus240 (ndash19) minus1750 minus79 minus1829 (ndash35)

(a) ldquosmallrdquo scenario

Volume (ML) Value (2014US$ million)

UK RoW WORLD () UK RoW WORLD ()

EU27 minus178 82 minus96 (ndash12) minus692 minus43 minus736 (ndash19)Chile minus46 28 minus18 (ndash24) minus128 36 minus91 (ndash32)South Africa minus43 29 minus14 (ndash42) minus82 23 minus59 (ndash47)USA 1 minus6 minus5 (ndash11) minus23 minus28 minus51 (ndash22)Australia 5 minus10 minus5 (ndash06) 19 minus56 minus38 (ndash13)Argentina 0 minus6 minus6 (ndash26) minus3 minus25 minus29 (ndash27)New Zealand minus5 4 minus1 (ndash06) minus80 34 minus46 (ndash22)Others 0 minus9 minus9 (ndash01) minus1 minus33 minus34 (ndash24)WORLD minus266 112 minus154 (ndash13) minus991 minus92 minus1083 (ndash21)

a Numbers in parentheses are the percentage difference between the Brexit and baseline scenarios for 2025 projected wine-import volumes orvalues by source

Source Authorsrsquo model results

242 UK and Global Wine Markets by 2025 and Implications of Brexit

in 2025 than in the ldquosmallrdquo scenario ndash that is they regain two-thirds of the value theywould have reached without Brexit-related uncertainty

This subsequent development in the Brexit process reverses most of but not allthe initial effects of Brexit by 2025 because of our assumption that the lost growthin the initial years of uncertainty following the Brexit vote are only partly recoveredby 2025 following the implementation of a UK-EU27 FTA Moreover the longer ittakes before this FTA is finalized and implemented the longer the estimated initialadverse effects persist and the larger the cumulative cost of Brexit to UK wine con-sumers and to grape and wine producers in wine-exporting countries

Table 6 summarizes the subsequent trade effects for 2025 It suggests that all butone-tenth of the loss in value of world trade in wine from the initial ldquosmallrdquo impact isrestored most of it because of a smaller reduction in the United Kingdomrsquos wineimports following the FTA Most of that improved outcome is because of recoveredimports from the EU27 commensurate with the latterrsquos high share of UK importsEven though Chile and South Africa are assumed in this scenario to have not yetsigned an FTA with the United Kingdom they export slightly more to the UnitedKingdom (and even more to all other countries) than in the initial ldquosmallrdquo scenario

Within the United Kingdom this FTA brings down the local currency consumerprice of wine by 9 largely offsetting the 11 rise in the initial ldquosmallrdquo Brexit sce-nario and it raises the volume consumed in the United Kingdom by 18 fully off-setting the 17 fall in the initial ldquosmallrdquo Brexit scenario

Table 6Difference in 2025 Bilateral Wine-Import Volumes and Values from Key Exporters by the

United Kingdom and the Rest of the World (RoW) as a Result of Implementing a UK-EU27FTA Following Initial ldquoSmallrdquo Brexit Shock (ML and 2014US$ Million Difference Relative to

Initial ldquoSmallrdquo Brexit Scenario)a

Volume (ML) Value (2014US$ million)

UK RoW WORLD (ldquosmallrdquo)a UK RoW WORLD (ldquosmallrdquo)a

EU27 212 minus112 100 (minus96) 750 0 750 (minus806)Chile 1 3 4 (minus18) 3 40 43 (minus92)South Africa 3 2 5 (minus14) 3 18 21 (minus60)USA 3 3 6 (minus5) 26 21 47 (minus50)Australia 1 1 2 (minus5) 8 38 46 (minus33)Argentina 1 4 5 (minus6) 4 20 24 (minus28)New Zealand 4 minus3 1 (minus1) 61 minus25 36 (minus46)Others 1 6 7 (minus9) 5 minus46 minus41 (92)WORLD 226 minus96 130 (minus154) 860 66 926 (minus1022)

a Numbers in parentheses are the world trade differences between the ldquosmallrdquo initial Brexit scenario and the baseline scenario copied fromcolumns 3 and 7 of Table 5(b)

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 243

In short this subsequent step in the Brexit process can be expected to restore muchof the initial adverse effects in the United Kingdom of the Brexit vote by 2025 Thewine exporters in the EU27 Chile and South Africa also lose less in this scenarioas compared with the initial ldquosmallrdquo scenario while other key exporters sell nearlyas much in 2025 in this scenario as they are projected to in the baseline ndash that isthey are beneficiaries of the reduced discrimination in global wine markets in thiscase where Chile and South Africa no longer enjoy preferential access to the UKmarket (and even though EU27 producers are again allowed to do so) Keep inmind though that although the estimated losses in the initial scenarios are reportedjust for 2025 they are expected to be felt throughout the preceding years of uncertaintyTherefore the longer it is before uncertainty abates and the United Kingdomrsquos currentpreferential trading arrangements are replaced by newones such as a UK-EU27 FTAthe higher the cumulative cost of the Brexit vote to the wine trade

VII Caveats and Conclusions

The above Brexit simulations are just a few of many scenarios that could be modeledObvious additional ones could also assume that FTAs are reached between theUnited Kingdom and other trading partners including Chile and South Africaand such countries as Australia and New Zealand Some of these other countriesare already exploring the ECrsquos invitation (see European Commission 2015) to nego-tiate an FTA with what will be the EU27 The sequence in which FTAs are signedand the speed with which they are implemented will matter (as was also the casewith the sequential signing over the past decade of bilateral FTAs with NortheastAsian countries by Chile Australia and New Zealand see Anderson andWittwer 2015) Even if the United Kingdom were able to sign additional bilateralFTAs and begin implementing them before 2025 it would have little effect on theabove results (because wine tariffs are such a minor contributor to them) unlessthose FTAs were to accelerate the United Kingdomrsquos economic growth and therise in the British poundrsquos value before 2025

We assume above that no changes are made to alcohol excise duties in the UnitedKingdom following Brexit even though they are scheduled to be progressively raisedwith inflation and may be raised even more for wine relative to spirits to offset theopposite effects of Brexit on those two domestic industries Nor do we make anyallowance for UK increases in consumer wine prices that may be needed to coverthe higher cost of clearing customs on imports from EU27 countries (which maybe more or less offset by increased smuggling and duty-free purchasing)Importers of fine wines including individuals who buy Bordeaux en primeur maywell reduce their demand because of concerns that delays on docks will affect thequality of their wines or because of greater currency uncertainty

To summarize Brexit is costly initially to UK consumers of wine (and of manyother tradable products) because the domestic retail price in local currency tax-

244 UK and Global Wine Markets by 2025 and Implications of Brexit

inclusive terms is 22 higher than otherwise in 2025 and the volume of wine con-sumed domestically is 28 lower in the ldquolargerdquo Brexit scenario (or 11 higher pricesand 17 lower quantities in the ldquosmallrdquo case) by 2025 unless a UK-EU27 FTA isdeveloped before then Even if such an FTA does get signed ratified by all 28 par-liaments and implemented by 2025 the slower income growth in the interim meansa smaller UK wine market in 2025 than would otherwise have been the case Thevolume reduction is a blow to many participants in UK wine-bottling transportingstoring wholesaling and retailing businesses in addition to restaurants and pubsVery little of that initial impact occurs because of higher import tariffs most impor-tant is the assumed fall in UK real incomes

The small but growing number of local UK vignerons (who supply less than 05of all domestic-wine sales and less than 4 of sparkling-wine sales) is initially pro-jected to sell less wine domestically because of reduced demand and to export only alittle extra abroad while the British pound is devalued Overall their production isonly 3 higher in 2025 even in the ldquolargerdquo Brexit scenario (not taking intoaccount the fact that their casual labor is likely to be more expensive under Brexitbecause of tighter restrictions on immigration) and their production diminishes ifand when a UK-EU27 FTA is in place

Even with a UK-EU27 FTA in place EU27 wine exporters export US$56 millionless wine in 2025 thanks to Brexit Chile and South Africa export $88 million lesswine and wine exports from the rest of the world are little different in aggregatealthough slightly larger from Australia

Clearly there will be great uncertainly for some time over the possible policy out-comes to flow from Brexit and their consequent sequential impacts on UK house-hold disposable incomes foreign exchange rates and bilateral wine tariffsMeanwhile the above projections under explicit assumptions provide some idea ofhow wine markets might be affected by the most-likely first two stages of theBrexit process (agreeing on a new tariff schedule at the WTO and agreeing to andimplementing a UK-EU27 FTA) In particular they make clear that nontrivialinitial impacts could affect the domestic wine market impacts that are likely to belarger than just the direct impact of changes in bilateral tariffs If the UnitedKingdom succeeds in getting countries to agree to transition arrangements thatdelay the changes in tariffs until new FTAs are signed and ratified the initialeffects will be less dramatic than in our first scenarios but it remains to be seenwhether any such agreements can be reached In any event the net effect of Brexiton the welfare of the worldrsquos consumers and producers of wine as a whole will benegative not just initially but permanently unless new trade policy commitmentsby the United Kingdom with major wine-exporting countries are sufficiently moreliberal than current arrangements

Kym Anderson and Glyn Wittwer 245

Appendix Key Equations in the Global Wine Markets Model

Amodel of the worldrsquos wine markets was first published byWittwer et al (2003) andmuch improved by Anderson and Wittwer (2013) Several significant enhancementshave been made to that original model Wine types have been disaggregated intomore types ndash namely nonpremium (including generic bulk) commercial-premiumand superpremium still wines and sparkling wines8 As in the original modelthere are two types of grapes premium and nonpremium Nonpremium wines usenonpremium grapes exclusively superpremium wines use mainly premium grapesand commercial-premium and sparkling wines use both types of grapes In theregional dimension the number of countries and country groups has expandedfrom 10 in the original model to 51 44 individual nations and 7 compositeregions The modelrsquos database is calibrated to 2014 for this paper based on datain Anderson et al (2017) trade data for which are downloaded from httpscomtradeunorg The modelrsquos supply-and-demand equations are based on the orig-inal ORANI modelrsquos theory (see Dixon Parmenter Sutton and Vincent 1982) andthe model is implemented using GEMPACK software (Harrison Horridge Jerieand Pearson 2014)

An enhancement of importance to the present study is the inclusion of exchange-rate variables in the model which allow a distinction between price impacts asobserved in local currency units and those observed in 2014 US dollars

In the model the grape and wine sectors minimize costs of intermediate inputssubject to weak constant elasticity of substitution (CES) substitutability betweeninputs By assumption no intermediate inputs are imported from other countries9Hence

Xcid frac14 f ethX1id CESfrac12P

cid=P1id $THORN eth1THORN

P1id X1id frac14X

c

Xcid P

cid eth2THORN

whereXcid is the quantity demanded of commodity c by grape or wine industry i in

region d Pcid is the corresponding price and X1id and P1id are the respective inter-

mediate composite quantities and prices

Two primary factors are employed in the sector labor (the quantity of which isendogenous with perfectly elastic supply) and capital Capital is usually treated asexogenous in quantity with rates of return bearing all the adjustment in thevarious scenarios This reflects the fact that grapes (a perennial crop) and wine-

8Commercial-premium still wines are defined by Anderson and Nelgen (2011) to be those priced betweenUS$250 and $750 per litre pre-tax at a countryrsquos border or wholesale9An exception concerns bulk wine imports used mainly in the bottling of commercial premium wine in theUK USA Canada Japan and Germany

246 UK and Global Wine Markets by 2025 and Implications of Brexit

plant capacity adjust slowly to market signals

Lid frac14 f ethFid CESfrac12Wid=PFid $THORN eth3THORN

Kid frac14 f ethFid CESethRid=PFid $THORN eth4THORN

PFid Fid frac14 Lid Wid thorn Kid Rid eth5THORN

Grape and wine producers are assumed to minimize costs subject to CES substitu-tion between capital and labor Equations (3) to (5) show primary factor demandsfor the labor composite Lid and capital Kid subject to a composite factor demandFid by industry i in region d The factor prices are Wid for labor Rid for capitalrentals and PFid for composite factor prices

Appendix Table A1Cumulative Consumption and Population Growth Rates and Changes in the Real Exchange Rate

(RER) Relative to the US Dollar 2014 to 2025 without Brexit ()

Aggregateconsumption Poprsquon RER

Aggregateconsumption Poprsquon RER

France 18 4 minus11 Australia 35 11 minus17Italy 11 2 minus9 New Zealand 32 9 minus26Portugal 14 0 minus9 Canada 27 8 minus18Spain 26 8 minus9 United States 31 8 0Austria 19 4 minus7 Argentina 7 10 109Belgium 20 7 minus9 Brazil 16 8 minus29Denmark 22 2 minus9 Chile 55 8 minus2Finland 21 3 minus7 Mexico 42 12 minus8Germany 14 minus2 minus11 Uruguay 45 3 1Greece 22 minus1 minus14 Other Latin America 60 10 minus5Ireland 42 12 minus9 South Africa 36 12 minus1Netherlands 21 4 minus9 Turkey 50 8 20Sweden 24 9 minus13 North Africa 53 11 0Switzerland 18 8 minus6 Other Africa 109 18 84United Kingdom 32 6 1 Middle East 52 18 minus12Other West Europe 21 10 minus1 China 79 3 5Bulgaria 41 minus7 7 Hong Kong 42 3 2Croatia 20 minus2 minus1 India 134 13 17Georgia 35 0 23 Japan 11 minus3 minus24Hungary 25 minus3 minus11 Korea 38 1 minus9Moldova 49 minus11 13 Malaysia 62 15 minus16Romania 45 minus4 22 Philippines 75 18 7Russia 18 minus2 minus8 Singapore 44 21 minus22Ukraine 22 minus5 14 Taiwan 29 1 minus13Other East Europe 40 minus5 48 Thailand 47 3 minus9

Other Asia 99 10 10

Source Authorsrsquo compilation from projections by various international agencies and from global economy-wide modeling by Anderson andStrutt (2016)

Kym Anderson and Glyn Wittwer 247

The composite factor demand Fid is proportional to total output Qid subject to aprimary factor using technology Aid Hence

Fid frac14 Qid Aid eth6THORN

The perfectly competitive zero-pure-profit condition is that total revenue valued atthe output price P0s

i multiplied by Qid equals the total production cost

P0si Qid frac14

X

cPcid X1cid thorn

X

oWo

id Loid thorn Rid Kid eth7THORN

Household demands follow a linear expenditure system in each region We reducethe optimizing problem for household consumption of each commodity subjectto a budget constraint to equations describing subsistence and discretionarydemands Aggregate subsistence expenditure WSUBd depends only on consumerprices P3cd for each commodity and the number of households N as per-capita sub-sistence quantities XSUBcd subject to given preferences are constant

WSUBd frac14X

c

P3cd XSUBcd Nd eth8THORN

Discretionary expenditures for each commodity (the left-hand side of equation (9))are equal to the marginal budget share (βcd) of aggregate discretionary expenditureThis aggregate is the bracketed term on the right-hand side of equation (9) whereW3TOTd is aggregate nominal expenditure

P3cdethX3cd XSUBcd NdTHORN frac14 βcdethW3TOTd WSUBdTHORN eth9THORN

Because real aggregate consumption is usually exogenous in our partial equilibriumsimulations the linear expenditure system determines the consumption shares ofindividual final commodities (ie the five wine types plus a composite of all otherconsumption items) driven by changes in relative prices as faced by domestic con-sumers The income elasticity of demand for each commodity is equal to the mar-ginal budget share divided by the expenditure share This number varies from 05for nonpremium wine to 17 for superpremium still wine The income elasticity ofdemand for other consumption is very close to 10 because wine accounts for anaverage of only 03 of aggregate expenditures globally and no more than 11in any country (Anderson and Nelgen 2011 Table 166)

A feature of our revised model of world wine markets is the inclusion of nominalexchange rates These appear directly in the equation linking retail prices (P3scd ) toproducer prices by country of origin (P0s

c ) where c denotes the wine type

P3 scd frac14 P0s

cfd

fsTtarcd T

taxcd thorn Pm

cd eth10THORN

248 UK and Global Wine Markets by 2025 and Implications of Brexit

The exchange rates in the consuming and producing regions are fd and fs respec-tively expressed as local currency units per $US Ttar

cd is the power of the tariff in theconsuming region and Ttax

cd is the power of the domestic consumption (or excise) taxprior to any generic value-added or goods-and-services tax Pm

cd is the price of marginm assumed to be locally supplied nontradable and therefore unaffected by theexchange rate

A given level of consumption for wine type c (X3cd) is satisfied using theArmington (1969) assumption in which wines from different countries of originare imperfectly substitutable First domestic wine is imperfectly substitutable witha composite of imports

X3sscd frac14 f ethX3cd CESethP3sscd=P3cdTHORNTHORN ss frac14 domestic imports eth11THORN

and then imports by origin (X3scd ) are determined in a second CES equation

X3scd frac14 f ethX3ssfrac14importscd CESethP3scd=P3

ssfrac14importscd THORNTHORN eth12THORN

The model enables us to show how changes in international competitiveness affectthe worldrsquos wine markets A crucial part of this exercise is explaining how pricesdetermined outside the grape and wine markets influence these markets Becausethe model is in partial equilibrium to depict the impacts of changes in internationalcompetitiveness outside price changes need to be imposed as shocks on the modelThe price of intermediate inputs other than grapes and wine shown in equations (1)and (2) is set equal to the price of GDP (Pg

d ) multiplied by a shifter Fcd

Pcid frac14 Fc

dPgd eth13THORN

If no specific price observations are available the shifter Fcd remains exogenous and

unshocked with the change in price being determined by a shock to the price ofGDP If observations are available for specific input price movements the shifterFc

dbecomes endogenous with Pc

id now exogenous and shocked

Wid frac14 Fwd P

gd eth14THORN

Wage rates are treated similarly In equation (14) if the wage shifter Fwd is exogenous

changes in wage rates Wid are determined by changes in the price of GDP If wagerate data are available Fw

d becomes endogenous andwage rates are shocked directly

Pmcd frac14 Fm

d Pgd eth15THORN

The prices of trade and transport margins are also determined by the price of GDP ifthe shifter Fm

d in equation (15) is exogenous

Changes in international competitiveness depend on changes in relative pricelevels and changes in nominal exchange rates In equation (16) fR

s denotes real

Kym Anderson and Glyn Wittwer 249

exchange-rate movements relative to the US dollar

fRs frac14 Pg

s =frac12PgUSA

fs$ eth16THORN

In equation (16) the nominal exchange rate for the United States is alwaysunchanged because nominal and real exchange rates are expressed relative to theUS currency

Changes in international market conditions may have impacts in one direction onproducer prices as expressed in US dollars P0s

i and potentially in the opposite direc-tion in local currency units Hence we calculate real producer pricesP0s

iloc in localcurrency terms (ie the price most relevant to domestic producers)

P0siloc frac14 P0s

ifs=P

gs eth17THORN

To obtain real price changes in local currency terms we convert US dollar prices(P3 s

cd for the source-specific price and P3cd for the source-composite price) to reallocal currency prices (P3 s

cdloc and P3cdloc) using the consumer price index (CPI)(Pc

d ) as the deflator

P3 scdloc frac14 P3 s

cdfd=P

cd eth18THORN

and

P3cdloc frac14 P3cdfd=Pcd eth19THORN

References

Anderson K and Nelgen S (2011) Global Wine Markets 1961 to 2009 A StatisticalCompendium Adelaide University of Adelaide Press Also available as an e-book atwwwadelaideeduaupresstitlesglobal-wine

Anderson K Nelgen S and Pinilla V (2017) Global Wine Markets 1860 to 2016 AStatistical Compendium Adelaide University of Adelaide Press (forthcoming) Also tobe available as an e-book at wwwadelaideeduaupress

Anderson K and Pinilla V (with the assistance of A J Holmes) (2017)Annual Database ofGlobal Wine Markets 1835 to 2016 Wine Economics Research Centre University ofAdelaide to be posted in October at wwwadelaideeduauwine-econdatabases

Anderson K and Strutt A (2016) Impacts of Asiarsquos rise on African and Latin Americantrade Projections to 2030 World Economy 39(2) 172ndash194

Anderson K andWittwer G (2013) Modeling global wine markets to 2018 Exchange ratestaste changes and Chinarsquos import growth Journal of Wine Economics 8(2) 131ndash158

Anderson K and Wittwer G (2015) Asiarsquos evolving role in global wine markets ChinaEconomic Review 35 1ndash14

250 UK and Global Wine Markets by 2025 and Implications of Brexit

Armington P A (1969) A theory of demand for products distinguished by place of produc-tion IMF Staff Papers 16 159ndash178

Baldwin R E (ed) (2016) Brexit Beckons Thinking Ahead by Leading Economists LondonCentre for Economic Policy Research

Campos N F (2016) Lousy experts Looking back at the ex ante estimates of the costs ofBrexit In R E Baldwin (ed) Brexit Beckons Thinking Ahead by Leading Economists(ch 3) London Centre for Economic Policy Research

Dixon P Parmenter B Sutton J and Vincent D (1982) ORANI A Multisectoral Modelof the Australian Economy Contributions to Economic Analysis 142 Amsterdam North-Holland

European Commission (2015) Trade for All Towards a More Responsible Trade andInvestment Policy Brussels European Commission

Harrison J Horridge M Jerie M and Pearson K (2014) GEMPACK ManualMelbourne GEMPACK Software

HM Treasury (2017) Spring Budget 2017 London HM TreasuryHolmes A J and Anderson K (2017a)Annual Database of National Beverage Consumption

Volumes and Expenditures 1950 to 2015 Wine Economics Research Centre University ofAdelaide posted at wwwadelaideeduauwine-econdatabases

Holmes A J and Anderson K (2017b) Convergence in National Alcohol ConsumptionPatterns New Global Indicators Journal of Wine Economics 12(2) 117ndash148

Ludington C C (2013) The Politics of Wine in Britain A New Cultural History BasingstokeUK Palgrave Macmillan

Office of National Statistics (2017) Analysis of real earnings London Office of NationalStatistics Available at httpswwwonsgovukemploymentandlabourmarketpeopleinworkearningsandworkinghoursarticlessupplementaryanalysisofaverageweeklyearningslatest

Rollo J Borchert I Dawar K Holmes P and Winters L A (2016) The World TradeOrganisation A safety net for a post-Brexit UK trade policy Briefing Paper 1UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-1-final-1pdf

Smith A (2017) Brexit Hard truths and hard choices UKTPO blog 19 June Available athttpsblogssussexacukuktpo20170619brexit-hard-truths-and-hard-choices

Swinbank A (2017) World trade rules and the policy options for British agriculture post-Brexit Briefing Paper 7 UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-7pdf

Viner J (1950) The Customs Union Issue New York Carnegie Endowment for InternationalPeace

Wine Australia (2015) Export Market Guide European Union Adelaide Wine AustraliaWittwer G Berger N and Anderson K (2003) A model of the worldrsquos wine markets

Economic Modelling 20(3) 487ndash506

Kym Anderson and Glyn Wittwer 251

Page 13: U.K. and Global Wine Markets by 2025, and Implications of ... · U.K. and Global Wine Markets by 2025, and Implications of Brexit* Kym Andersona and Glyn Wittwerb Abstract The United

Figure 5

Value of Wine Production and Consumption in Key Countries 2014 and Projected Baseline for2025 (2014 US$ Million at WineryWholesale Pretax Prices)

Fig5-BW

onlin

eBW

inprint

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 233

increase in consumption of fine wine In Western Europe and in the SouthernHemispherersquos New World countries fine wines are projected to substitute for com-mercial wines (defined as the sum of commercial-premium and nonpremiumwines) with almost no change in total wine consumption Sub-Saharan Africa isthe next region that is projected to take off with its growth accounting for morethan one-third of the rest of the worldrsquos increase in volume consumed

Those differences in production versus consumption rankings are reflected ininternational trade Figure 6 shows that France Italy and Spain remain the threedominant exporters of wine in aggregate value but the rankings of the next fewchange with Australia being slightly ahead of Chile and the United StatesGermany and New Zealand being nearly tied for sixth place in value termsFrance and Italy are even more dominant in fine-wine exports and remain so by2025 while Italy outranks France in the commercial-premium export categoryand Spain outranks Italy Australia and Chile in the nonpremium export class

Among the importers the United States and the United Kingdom are projected tocontinue to hold the first two places in 2025 in value terms but China moves intothird place slightly ahead of Germany followed well behind by Canada HongKong Belgium-Luxembourg the Netherlands and Japan (Figure 6(b)) OtherAfrica (excluding South Africa) is projected to experience the largest increase inimports among all the other regions followed by Other Asia which becomes asbig as Germany in value terms (Figure 7(a)) In terms of total volume of wineimports Germany and the United Kingdom held the top two shares in 2014 butby 2025 the United Kingdom is projected to be well ahead of Germany (Figure 7(b)) However this projection ignores the effects of Brexit to which we now turn

VI How Might Wine Markets Be Affected by the United Kingdomrsquos Exitfrom the European Union

As mentioned earlier for our two alternative scenarios to capture the initial effects ofBrexit (ldquolargerdquo and ldquosmallrdquo) we assume that following the United Kingdomrsquos exitfrom the European Union the United Kingdomrsquos rate of economic growth wouldbe only one-third or two-thirds as fast for the period to 2025 the British poundwould be 20 or 10 lower in real terms than in our modelrsquos baseline projectionand the United Kingdom would apply the European Unionrsquos external tariff onwine to imports from EU member countries (as part of establishing MFN ratesvia the WTO to then start new bilateral FTA negotiations) In these initial scenarioswe assume the United Kingdom does not implement any new FTAs particularlywith the EU27 Chile and South Africa Following the discussion of those initialresults we present the results of a subsequent scenario that includes the implemen-tation of an FTA between the United Kingdom and the EU27 We assume these aretwo discrete steps with no agreed arrangements to smooth the transition betweenthem

234 UK and Global Wine Markets by 2025 and Implications of Brexit

Figure 6

Value of Wine Exports and Imports Key Wine-Trading Countries 2014 and Projected Baselinefor 2025 (2014 US$ million)

Fig6-BW

onlin

eBW

inprint

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 235

Figure 7

National Shares of Global Wine Import Value and Volume 2014 and Projected Baselinefor 2025 ()

Fig7-BW

onlin

eBW

inprint

Source Authorsrsquo model results

236 UK and Global Wine Markets by 2025 and Implications of Brexit

A Initial Impact of Brexit

Because the British pound dropped by one-sixth against the US dollar in the fourmonths following the Brexit vote in June 2016 and because the UnitedKingdomrsquos average real wage fell 1 between November 2016 and April 2017(according to the latest available data see Office of National Statistics 2017) weviewed our ldquolargerdquo scenario as more likely soon after the Brexit vote But wecompare those results with ones from our ldquosmallrdquo initial-impact scenario becauseit has become more likely following the June 2017 general election that ldquosofterrdquooptions will be considered by the new government Generally the results areabout half the size in the ldquosmallrdquo scenario with the exception of the bilateraltrade effects To show the sensitivity of results to our assumptions we point outthe differences when the ldquosmallrdquo results are not close to half the results shown forthe ldquolargerdquo scenario

Before turning to those results we first present a ldquofree-traderdquo scenario whichassumes the United Kingdom chooses the radical option of becoming the HongKong of Europe Unlikely as this scenario is the results provide assurance thatour global wine model generates the standard types of effects of exiting a customsunion In this ldquofree-traderdquo scenario the United Kingdom is assumed to move tozero tariffs on all wine imports Thus the current preferences on EU Chileanand South African wine imports disappear as all other wine exporters enjoy thesame free access to UK wine markets We assume in this scenario that Brexit hasno impact on the United Kingdomrsquos real income or consumption growth rates by2025 and that by then the British pound has returned to its 2014 value relative toother currencies (similar to the real exchange rate assumed in our baseline scenarioto 2025 see Appendix Table A1)7 Under these assumptions EU Chilean andSouth African wine exports to the United Kingdom in 2025 are 003 lower thanin the no-Brexit baseline and UK imports from other wine exporters are 025higher but the United Kingdomrsquos total wine imports are hardly any different(because local wine prices fall by only 05 and incomes are unchanged) In this sce-nario UK consumers expand their volume of consumption by just 05 while theEU27 exports more and other countries export less to the rest of the world such thatworld wine exports are almost unchanged In short all these effects are in the direc-tion that comparative static-customs-union theory would predict but they are smallbecause we assume the United Kingdomrsquos real income and currency are the same asin our baseline for 2025 as described in the previous section

In the ldquolargerdquo alternate initial scenario involving Brexit as compared with thebaseline scenario to 2025 the consumer price of wine in 2025 is 22 higher in the

7If this scenario were seen as a credible long-run trade policy it would encourage more investment thatwould eventually raise UK incomes but not without major structural changes that would initiallydisrupt the economy Implicitly we assume in our free-trade scenario that these two opposite impactson real UK incomes and the British pound exactly offset each other by 2025

Kym Anderson and Glyn Wittwer 237

United Kingdom in local currency terms (20 because of real depreciation of theBritish pound 4 because of new tariffs on EU Chilean and South Africanwines and ndash2 because of slower UK income growth) The volume of UKwine consumption is 28 lower 16 because of slower UK economic growth7 because of real depreciation of the British pound and 5 because of newtariffs Superpremium still-wine sales are the most affected dropping by two-fifths while sparkling and commercial-premium wines drop a bit less than one-quarter Because the average price rises by more than the decline in volume soldthe aggregate value of UK sales even in local currency terms falls under thisldquolargerdquo Brexit scenario Under the ldquosmallrdquo Brexit scenario the consumer price ofwine in 2025 is 11 higher in the United Kingdom and its volume of wine consump-tion is 17 lower

The volume of projected UK imports in 2025 is 427 ML or nearly one-quarterless in the ldquolargerdquo scenario than in the baseline scenario comprising 58ML less spar-kling wine 31 ML less superpremium still wine and 339 ML less commercial-premium wine World imports are lower by just 239 ML because imports byother countries are 189 ML higher in response to the lower international wineprices in this scenario In value terms UK imports are $175 billion (or 27)lower in 2025 because of ldquolargerdquo Brexit $113 billion because of lower incomes$038 billion because of the fall in the British pound and $014 billion because ofthe rise in wine-import tariffs (Table 4) These aggregate trade impacts are a littlemore than half this size under the ldquosmallrdquo Brexit scenario

Despite the import levels falling because of raised import tariffs domestic con-sumption of all three quality categories of UK-produced wine is lower with thanwithout Brexit because of shrunken demand for all wines resulting from loweredUK incomes and raised local prices because of devaluation of the British poundThe British poundrsquos devaluation does make it easier for the United Kingdom tosell wines abroad though The countryrsquos exports are 7 ML or nearly 5 higherin 2025 in the ldquolargerdquo Brexit scenario and UK production is 3 higher ThoseUK exports (or re-exports of imported bulk wine after it is bottled in the UnitedKingdom) that go to EU27 countries are reduced though because of the tariffnow imposed at the new EU border

Without Brexit the United Kingdomrsquos shares of global wine imports are slightlyhigher in volume terms in 2025 than in 2010ndash2015 but 2 percentage points lower invalue terms thanks to East Asiarsquos expanding demand for imports of premium winesWith ldquolargerdquo Brexit however that value share is an additional 2 percentage pointslower and the volume share is almost 5 points lower (Figure 8) The net effects ofthese impacts on global trade are shown in Figure 9 Most of the initial tradeeffect of Brexit is a large decline in net imports of wine by the United Kingdomwith very little offsetting positive effect on trade in the rest of the world Theldquosmallrdquo Brexit numbers are a bit more than half of those for ldquolargerdquo Brexit

238 UK and Global Wine Markets by 2025 and Implications of Brexit

Table4

Difference

in2025

Projected

VolumeandValue

ofWineIm

portsb

ytheUnitedKingdom

andtheResto

fthe

World

asaConsequence

oftheInitial

BrexitS

hock

(MLand2014US$

Million

ldquoLargerdquoSc

enario)

Volum

e(M

L)

Value

(US$

million)

NP+CPa

SuperPrb

Sparkling

TOTAL

NP+CPa

SuperPrb

Sparkling

TOTAL

ΔUK

impo

rtsdu

eto

Low

erincomes

minus198

minus20

minus29

minus247

58minus644

minus253

minus234

minus1131

65Low

erpo

und

minus70

minus10

minus14

minus93

22minus248

minus127

minus102

minus476

27Highertariffs

minus71

minus1minus16

minus87

20minus110

minus8minus2

4minus143

8TOTA

Lminus339

minus31

minus58

minus427

100

minus1001

minus388

minus360

minus1750

100

difffrom

base

2332

3325

2432

3227

of

totalc

uts

797

14100

5722

21100

ΔROW

netim

ports

143

2125

189

230

143

minus181

192

ΔWORLD

TRADE

minus195

minus10

minus34

minus239

minus763

minus246

minus543

minus1552

aNon

prem

ium

plus

commercial-premium

still

wines

bSu

perpremium

still

wines

SourceA

utho

rsrsquomod

elresults

Kym Anderson and Glyn Wittwer 239

The aggregate effect of ldquolargerdquo Brexit on the market shares of various wine-export-ing countries in the United Kingdom is almost indiscernible even with one decimalpoint The projected 2025 shares are quite different from the actual 2014 shares forseveral countries They are much smaller in 2025 for South Africa Australia andNew Zealand (and the United States in volume terms) and are much larger involume for Spain and in value for Italy These results occur because wine-exportingcountries benefit differentially from the varying rates of growth in net importdemand for wine in non-UK countries over this projection period The most impor-tant projected changes are the increase in the real value of annual wine importsbetween 2014 and 2025 by China (200 or $3 billion) Other Asia (110 or $22billion) and Africa (270 or $16 billion) More than half of Australiarsquos increasein annual exports from 2014 to 2025 go to Asia and more than half of SouthAfricarsquos increase in exports go to Other Africa

Table 5 reveals that ldquolargerdquo Brexit lowers EU Chilean and South African wineexports by 150 ML or US$12 billion in the case of the European Union with someof their exports diverted from the United Kingdom to the EU27 and other marketsin competition with New World exporters Although the United States Australiaand Argentina sell only a little less to the United Kingdom they also sell less toother countries For Chile and South Africa which lose their preferential access toUK (but not to EU27) markets in this Brexit scenario some exports are redirected

Figure 8

UK Shares of World Wine Imports 2010ndash2015 and Projected to 2025 without and with theldquoLargerdquo Initial Brexit Shock ()

Fig8-BW

onlin

eBW

inprint

Sources Anderson and Pinilla (2017) and authorsrsquo model results

240 UK and Global Wine Markets by 2025 and Implications of Brexit

from the United Kingdom to EU27 countries ndash but again they export less overallGlobal wine trade in 2025 is 240 ML (19) or $18 billion (35) less under thisldquolargerdquo Brexit scenario The percentage by which wine exportersrsquo trade shrinks isgreater for values than for volumes because of changes in relative prices of differ-ent-quality wines Those differences are shown in the numbers in parentheses inTable 5

Three other points are worth making about Table 5 One is that Australia sellsslightly more to the United Kingdom in the ldquosmallrdquo Brexit scenario rather thanslightly less as in the ldquolargerdquo Brexit case Evidently the negative income and price(devaluation) effects do not more than offset the positive trade-diverting effect onAustralian exports to the United Kingdom of removing preferences in the ldquosmallrdquoscenario Second New Zealand sells slightly more to non-UK countries underBrexit despite greater competition from the EU27 Chile and South Africa Thisanomaly is due to changes in the relative prices of different qualities of wine inglobal wine markets bearing in mind that New Zealand has the worldrsquos highestaverage price for still-wine exports And third the value (but not the volume) ofexports of ldquoOtherrdquo countries to markets other than the United Kingdom arehigher under Brexit This result too is due to changes in the relative prices of differ-ent qualities of wine in global wine markets

Figure 9

Difference in 2025 Wine Import Volumes and Values as a Result of the ldquoLargerdquo Initial BrexitShock (ML and US$ Million in 2014 US dollars)

Fig9-BW

onlin

eBW

inprint

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 241

B Subsequent Impact of Brexit from a UK-EU27 FTA

The next-most-likely step in the Brexit process is for the United Kingdom to nego-tiate a new trade arrangement with the EU27 We therefore assume that a UK-EU27FTAwith free bilateral wine trade is implemented and adjusted to by 2025 and thatprogress toward that end occurs soon enough that the adverse macroeconomicshocks from the initial impact of uncertainty over the Brexit process are confinedto those assumed in the ldquosmallrdquo scenario outlined above In this subsequent scenariowe assume that the British pound returns to the value it would have reached in 2025in the absence of Brexit (ie reversing the 10 devaluation assumed in the ldquosmallrdquoinitial Brexit scenario) and that real incomes in the United Kingdom are 8 higher

Table 5Difference in 2025 BilateralWine Import Volumes and Values fromKey Exporters by the United

Kingdom and the Rest of the World (RoW) as a Result of Initial Brexit Shock(ML and 2014US$ Million)a

(a) ldquolargerdquo scenario

Volume (ML) Value (2014US$ million)

UK RoW WORLD () UK RoW WORLD ()

EU27 minus287 136 minus150 (ndash17) minus1187 minus5 minus1192 (ndash31)Chile minus59 35 minus25 (ndash30) minus169 31 minus138 (ndash48)South Africa minus53 35 minus18 (ndash32) minus105 20 minus85 (ndash67)USA minus7 minus6 minus13 (ndash24) minus75 minus40 minus115 (ndash50)Australia minus4 minus3 minus7 (ndash09) minus25 minus65 minus90 (ndash30)Argentina minus3 minus9 minus12 (ndash48) minus16 minus39 minus55 (ndash52)New Zealand minus11 9 minus2 (ndash09) minus162 71 minus91 (ndash43)Others minus2 minus10 minus12 (ndash02) minus11 minus52 minus63 (ndash44)WORLD minus427 187 minus240 (ndash19) minus1750 minus79 minus1829 (ndash35)

(a) ldquosmallrdquo scenario

Volume (ML) Value (2014US$ million)

UK RoW WORLD () UK RoW WORLD ()

EU27 minus178 82 minus96 (ndash12) minus692 minus43 minus736 (ndash19)Chile minus46 28 minus18 (ndash24) minus128 36 minus91 (ndash32)South Africa minus43 29 minus14 (ndash42) minus82 23 minus59 (ndash47)USA 1 minus6 minus5 (ndash11) minus23 minus28 minus51 (ndash22)Australia 5 minus10 minus5 (ndash06) 19 minus56 minus38 (ndash13)Argentina 0 minus6 minus6 (ndash26) minus3 minus25 minus29 (ndash27)New Zealand minus5 4 minus1 (ndash06) minus80 34 minus46 (ndash22)Others 0 minus9 minus9 (ndash01) minus1 minus33 minus34 (ndash24)WORLD minus266 112 minus154 (ndash13) minus991 minus92 minus1083 (ndash21)

a Numbers in parentheses are the percentage difference between the Brexit and baseline scenarios for 2025 projected wine-import volumes orvalues by source

Source Authorsrsquo model results

242 UK and Global Wine Markets by 2025 and Implications of Brexit

in 2025 than in the ldquosmallrdquo scenario ndash that is they regain two-thirds of the value theywould have reached without Brexit-related uncertainty

This subsequent development in the Brexit process reverses most of but not allthe initial effects of Brexit by 2025 because of our assumption that the lost growthin the initial years of uncertainty following the Brexit vote are only partly recoveredby 2025 following the implementation of a UK-EU27 FTA Moreover the longer ittakes before this FTA is finalized and implemented the longer the estimated initialadverse effects persist and the larger the cumulative cost of Brexit to UK wine con-sumers and to grape and wine producers in wine-exporting countries

Table 6 summarizes the subsequent trade effects for 2025 It suggests that all butone-tenth of the loss in value of world trade in wine from the initial ldquosmallrdquo impact isrestored most of it because of a smaller reduction in the United Kingdomrsquos wineimports following the FTA Most of that improved outcome is because of recoveredimports from the EU27 commensurate with the latterrsquos high share of UK importsEven though Chile and South Africa are assumed in this scenario to have not yetsigned an FTA with the United Kingdom they export slightly more to the UnitedKingdom (and even more to all other countries) than in the initial ldquosmallrdquo scenario

Within the United Kingdom this FTA brings down the local currency consumerprice of wine by 9 largely offsetting the 11 rise in the initial ldquosmallrdquo Brexit sce-nario and it raises the volume consumed in the United Kingdom by 18 fully off-setting the 17 fall in the initial ldquosmallrdquo Brexit scenario

Table 6Difference in 2025 Bilateral Wine-Import Volumes and Values from Key Exporters by the

United Kingdom and the Rest of the World (RoW) as a Result of Implementing a UK-EU27FTA Following Initial ldquoSmallrdquo Brexit Shock (ML and 2014US$ Million Difference Relative to

Initial ldquoSmallrdquo Brexit Scenario)a

Volume (ML) Value (2014US$ million)

UK RoW WORLD (ldquosmallrdquo)a UK RoW WORLD (ldquosmallrdquo)a

EU27 212 minus112 100 (minus96) 750 0 750 (minus806)Chile 1 3 4 (minus18) 3 40 43 (minus92)South Africa 3 2 5 (minus14) 3 18 21 (minus60)USA 3 3 6 (minus5) 26 21 47 (minus50)Australia 1 1 2 (minus5) 8 38 46 (minus33)Argentina 1 4 5 (minus6) 4 20 24 (minus28)New Zealand 4 minus3 1 (minus1) 61 minus25 36 (minus46)Others 1 6 7 (minus9) 5 minus46 minus41 (92)WORLD 226 minus96 130 (minus154) 860 66 926 (minus1022)

a Numbers in parentheses are the world trade differences between the ldquosmallrdquo initial Brexit scenario and the baseline scenario copied fromcolumns 3 and 7 of Table 5(b)

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 243

In short this subsequent step in the Brexit process can be expected to restore muchof the initial adverse effects in the United Kingdom of the Brexit vote by 2025 Thewine exporters in the EU27 Chile and South Africa also lose less in this scenarioas compared with the initial ldquosmallrdquo scenario while other key exporters sell nearlyas much in 2025 in this scenario as they are projected to in the baseline ndash that isthey are beneficiaries of the reduced discrimination in global wine markets in thiscase where Chile and South Africa no longer enjoy preferential access to the UKmarket (and even though EU27 producers are again allowed to do so) Keep inmind though that although the estimated losses in the initial scenarios are reportedjust for 2025 they are expected to be felt throughout the preceding years of uncertaintyTherefore the longer it is before uncertainty abates and the United Kingdomrsquos currentpreferential trading arrangements are replaced by newones such as a UK-EU27 FTAthe higher the cumulative cost of the Brexit vote to the wine trade

VII Caveats and Conclusions

The above Brexit simulations are just a few of many scenarios that could be modeledObvious additional ones could also assume that FTAs are reached between theUnited Kingdom and other trading partners including Chile and South Africaand such countries as Australia and New Zealand Some of these other countriesare already exploring the ECrsquos invitation (see European Commission 2015) to nego-tiate an FTA with what will be the EU27 The sequence in which FTAs are signedand the speed with which they are implemented will matter (as was also the casewith the sequential signing over the past decade of bilateral FTAs with NortheastAsian countries by Chile Australia and New Zealand see Anderson andWittwer 2015) Even if the United Kingdom were able to sign additional bilateralFTAs and begin implementing them before 2025 it would have little effect on theabove results (because wine tariffs are such a minor contributor to them) unlessthose FTAs were to accelerate the United Kingdomrsquos economic growth and therise in the British poundrsquos value before 2025

We assume above that no changes are made to alcohol excise duties in the UnitedKingdom following Brexit even though they are scheduled to be progressively raisedwith inflation and may be raised even more for wine relative to spirits to offset theopposite effects of Brexit on those two domestic industries Nor do we make anyallowance for UK increases in consumer wine prices that may be needed to coverthe higher cost of clearing customs on imports from EU27 countries (which maybe more or less offset by increased smuggling and duty-free purchasing)Importers of fine wines including individuals who buy Bordeaux en primeur maywell reduce their demand because of concerns that delays on docks will affect thequality of their wines or because of greater currency uncertainty

To summarize Brexit is costly initially to UK consumers of wine (and of manyother tradable products) because the domestic retail price in local currency tax-

244 UK and Global Wine Markets by 2025 and Implications of Brexit

inclusive terms is 22 higher than otherwise in 2025 and the volume of wine con-sumed domestically is 28 lower in the ldquolargerdquo Brexit scenario (or 11 higher pricesand 17 lower quantities in the ldquosmallrdquo case) by 2025 unless a UK-EU27 FTA isdeveloped before then Even if such an FTA does get signed ratified by all 28 par-liaments and implemented by 2025 the slower income growth in the interim meansa smaller UK wine market in 2025 than would otherwise have been the case Thevolume reduction is a blow to many participants in UK wine-bottling transportingstoring wholesaling and retailing businesses in addition to restaurants and pubsVery little of that initial impact occurs because of higher import tariffs most impor-tant is the assumed fall in UK real incomes

The small but growing number of local UK vignerons (who supply less than 05of all domestic-wine sales and less than 4 of sparkling-wine sales) is initially pro-jected to sell less wine domestically because of reduced demand and to export only alittle extra abroad while the British pound is devalued Overall their production isonly 3 higher in 2025 even in the ldquolargerdquo Brexit scenario (not taking intoaccount the fact that their casual labor is likely to be more expensive under Brexitbecause of tighter restrictions on immigration) and their production diminishes ifand when a UK-EU27 FTA is in place

Even with a UK-EU27 FTA in place EU27 wine exporters export US$56 millionless wine in 2025 thanks to Brexit Chile and South Africa export $88 million lesswine and wine exports from the rest of the world are little different in aggregatealthough slightly larger from Australia

Clearly there will be great uncertainly for some time over the possible policy out-comes to flow from Brexit and their consequent sequential impacts on UK house-hold disposable incomes foreign exchange rates and bilateral wine tariffsMeanwhile the above projections under explicit assumptions provide some idea ofhow wine markets might be affected by the most-likely first two stages of theBrexit process (agreeing on a new tariff schedule at the WTO and agreeing to andimplementing a UK-EU27 FTA) In particular they make clear that nontrivialinitial impacts could affect the domestic wine market impacts that are likely to belarger than just the direct impact of changes in bilateral tariffs If the UnitedKingdom succeeds in getting countries to agree to transition arrangements thatdelay the changes in tariffs until new FTAs are signed and ratified the initialeffects will be less dramatic than in our first scenarios but it remains to be seenwhether any such agreements can be reached In any event the net effect of Brexiton the welfare of the worldrsquos consumers and producers of wine as a whole will benegative not just initially but permanently unless new trade policy commitmentsby the United Kingdom with major wine-exporting countries are sufficiently moreliberal than current arrangements

Kym Anderson and Glyn Wittwer 245

Appendix Key Equations in the Global Wine Markets Model

Amodel of the worldrsquos wine markets was first published byWittwer et al (2003) andmuch improved by Anderson and Wittwer (2013) Several significant enhancementshave been made to that original model Wine types have been disaggregated intomore types ndash namely nonpremium (including generic bulk) commercial-premiumand superpremium still wines and sparkling wines8 As in the original modelthere are two types of grapes premium and nonpremium Nonpremium wines usenonpremium grapes exclusively superpremium wines use mainly premium grapesand commercial-premium and sparkling wines use both types of grapes In theregional dimension the number of countries and country groups has expandedfrom 10 in the original model to 51 44 individual nations and 7 compositeregions The modelrsquos database is calibrated to 2014 for this paper based on datain Anderson et al (2017) trade data for which are downloaded from httpscomtradeunorg The modelrsquos supply-and-demand equations are based on the orig-inal ORANI modelrsquos theory (see Dixon Parmenter Sutton and Vincent 1982) andthe model is implemented using GEMPACK software (Harrison Horridge Jerieand Pearson 2014)

An enhancement of importance to the present study is the inclusion of exchange-rate variables in the model which allow a distinction between price impacts asobserved in local currency units and those observed in 2014 US dollars

In the model the grape and wine sectors minimize costs of intermediate inputssubject to weak constant elasticity of substitution (CES) substitutability betweeninputs By assumption no intermediate inputs are imported from other countries9Hence

Xcid frac14 f ethX1id CESfrac12P

cid=P1id $THORN eth1THORN

P1id X1id frac14X

c

Xcid P

cid eth2THORN

whereXcid is the quantity demanded of commodity c by grape or wine industry i in

region d Pcid is the corresponding price and X1id and P1id are the respective inter-

mediate composite quantities and prices

Two primary factors are employed in the sector labor (the quantity of which isendogenous with perfectly elastic supply) and capital Capital is usually treated asexogenous in quantity with rates of return bearing all the adjustment in thevarious scenarios This reflects the fact that grapes (a perennial crop) and wine-

8Commercial-premium still wines are defined by Anderson and Nelgen (2011) to be those priced betweenUS$250 and $750 per litre pre-tax at a countryrsquos border or wholesale9An exception concerns bulk wine imports used mainly in the bottling of commercial premium wine in theUK USA Canada Japan and Germany

246 UK and Global Wine Markets by 2025 and Implications of Brexit

plant capacity adjust slowly to market signals

Lid frac14 f ethFid CESfrac12Wid=PFid $THORN eth3THORN

Kid frac14 f ethFid CESethRid=PFid $THORN eth4THORN

PFid Fid frac14 Lid Wid thorn Kid Rid eth5THORN

Grape and wine producers are assumed to minimize costs subject to CES substitu-tion between capital and labor Equations (3) to (5) show primary factor demandsfor the labor composite Lid and capital Kid subject to a composite factor demandFid by industry i in region d The factor prices are Wid for labor Rid for capitalrentals and PFid for composite factor prices

Appendix Table A1Cumulative Consumption and Population Growth Rates and Changes in the Real Exchange Rate

(RER) Relative to the US Dollar 2014 to 2025 without Brexit ()

Aggregateconsumption Poprsquon RER

Aggregateconsumption Poprsquon RER

France 18 4 minus11 Australia 35 11 minus17Italy 11 2 minus9 New Zealand 32 9 minus26Portugal 14 0 minus9 Canada 27 8 minus18Spain 26 8 minus9 United States 31 8 0Austria 19 4 minus7 Argentina 7 10 109Belgium 20 7 minus9 Brazil 16 8 minus29Denmark 22 2 minus9 Chile 55 8 minus2Finland 21 3 minus7 Mexico 42 12 minus8Germany 14 minus2 minus11 Uruguay 45 3 1Greece 22 minus1 minus14 Other Latin America 60 10 minus5Ireland 42 12 minus9 South Africa 36 12 minus1Netherlands 21 4 minus9 Turkey 50 8 20Sweden 24 9 minus13 North Africa 53 11 0Switzerland 18 8 minus6 Other Africa 109 18 84United Kingdom 32 6 1 Middle East 52 18 minus12Other West Europe 21 10 minus1 China 79 3 5Bulgaria 41 minus7 7 Hong Kong 42 3 2Croatia 20 minus2 minus1 India 134 13 17Georgia 35 0 23 Japan 11 minus3 minus24Hungary 25 minus3 minus11 Korea 38 1 minus9Moldova 49 minus11 13 Malaysia 62 15 minus16Romania 45 minus4 22 Philippines 75 18 7Russia 18 minus2 minus8 Singapore 44 21 minus22Ukraine 22 minus5 14 Taiwan 29 1 minus13Other East Europe 40 minus5 48 Thailand 47 3 minus9

Other Asia 99 10 10

Source Authorsrsquo compilation from projections by various international agencies and from global economy-wide modeling by Anderson andStrutt (2016)

Kym Anderson and Glyn Wittwer 247

The composite factor demand Fid is proportional to total output Qid subject to aprimary factor using technology Aid Hence

Fid frac14 Qid Aid eth6THORN

The perfectly competitive zero-pure-profit condition is that total revenue valued atthe output price P0s

i multiplied by Qid equals the total production cost

P0si Qid frac14

X

cPcid X1cid thorn

X

oWo

id Loid thorn Rid Kid eth7THORN

Household demands follow a linear expenditure system in each region We reducethe optimizing problem for household consumption of each commodity subjectto a budget constraint to equations describing subsistence and discretionarydemands Aggregate subsistence expenditure WSUBd depends only on consumerprices P3cd for each commodity and the number of households N as per-capita sub-sistence quantities XSUBcd subject to given preferences are constant

WSUBd frac14X

c

P3cd XSUBcd Nd eth8THORN

Discretionary expenditures for each commodity (the left-hand side of equation (9))are equal to the marginal budget share (βcd) of aggregate discretionary expenditureThis aggregate is the bracketed term on the right-hand side of equation (9) whereW3TOTd is aggregate nominal expenditure

P3cdethX3cd XSUBcd NdTHORN frac14 βcdethW3TOTd WSUBdTHORN eth9THORN

Because real aggregate consumption is usually exogenous in our partial equilibriumsimulations the linear expenditure system determines the consumption shares ofindividual final commodities (ie the five wine types plus a composite of all otherconsumption items) driven by changes in relative prices as faced by domestic con-sumers The income elasticity of demand for each commodity is equal to the mar-ginal budget share divided by the expenditure share This number varies from 05for nonpremium wine to 17 for superpremium still wine The income elasticity ofdemand for other consumption is very close to 10 because wine accounts for anaverage of only 03 of aggregate expenditures globally and no more than 11in any country (Anderson and Nelgen 2011 Table 166)

A feature of our revised model of world wine markets is the inclusion of nominalexchange rates These appear directly in the equation linking retail prices (P3scd ) toproducer prices by country of origin (P0s

c ) where c denotes the wine type

P3 scd frac14 P0s

cfd

fsTtarcd T

taxcd thorn Pm

cd eth10THORN

248 UK and Global Wine Markets by 2025 and Implications of Brexit

The exchange rates in the consuming and producing regions are fd and fs respec-tively expressed as local currency units per $US Ttar

cd is the power of the tariff in theconsuming region and Ttax

cd is the power of the domestic consumption (or excise) taxprior to any generic value-added or goods-and-services tax Pm

cd is the price of marginm assumed to be locally supplied nontradable and therefore unaffected by theexchange rate

A given level of consumption for wine type c (X3cd) is satisfied using theArmington (1969) assumption in which wines from different countries of originare imperfectly substitutable First domestic wine is imperfectly substitutable witha composite of imports

X3sscd frac14 f ethX3cd CESethP3sscd=P3cdTHORNTHORN ss frac14 domestic imports eth11THORN

and then imports by origin (X3scd ) are determined in a second CES equation

X3scd frac14 f ethX3ssfrac14importscd CESethP3scd=P3

ssfrac14importscd THORNTHORN eth12THORN

The model enables us to show how changes in international competitiveness affectthe worldrsquos wine markets A crucial part of this exercise is explaining how pricesdetermined outside the grape and wine markets influence these markets Becausethe model is in partial equilibrium to depict the impacts of changes in internationalcompetitiveness outside price changes need to be imposed as shocks on the modelThe price of intermediate inputs other than grapes and wine shown in equations (1)and (2) is set equal to the price of GDP (Pg

d ) multiplied by a shifter Fcd

Pcid frac14 Fc

dPgd eth13THORN

If no specific price observations are available the shifter Fcd remains exogenous and

unshocked with the change in price being determined by a shock to the price ofGDP If observations are available for specific input price movements the shifterFc

dbecomes endogenous with Pc

id now exogenous and shocked

Wid frac14 Fwd P

gd eth14THORN

Wage rates are treated similarly In equation (14) if the wage shifter Fwd is exogenous

changes in wage rates Wid are determined by changes in the price of GDP If wagerate data are available Fw

d becomes endogenous andwage rates are shocked directly

Pmcd frac14 Fm

d Pgd eth15THORN

The prices of trade and transport margins are also determined by the price of GDP ifthe shifter Fm

d in equation (15) is exogenous

Changes in international competitiveness depend on changes in relative pricelevels and changes in nominal exchange rates In equation (16) fR

s denotes real

Kym Anderson and Glyn Wittwer 249

exchange-rate movements relative to the US dollar

fRs frac14 Pg

s =frac12PgUSA

fs$ eth16THORN

In equation (16) the nominal exchange rate for the United States is alwaysunchanged because nominal and real exchange rates are expressed relative to theUS currency

Changes in international market conditions may have impacts in one direction onproducer prices as expressed in US dollars P0s

i and potentially in the opposite direc-tion in local currency units Hence we calculate real producer pricesP0s

iloc in localcurrency terms (ie the price most relevant to domestic producers)

P0siloc frac14 P0s

ifs=P

gs eth17THORN

To obtain real price changes in local currency terms we convert US dollar prices(P3 s

cd for the source-specific price and P3cd for the source-composite price) to reallocal currency prices (P3 s

cdloc and P3cdloc) using the consumer price index (CPI)(Pc

d ) as the deflator

P3 scdloc frac14 P3 s

cdfd=P

cd eth18THORN

and

P3cdloc frac14 P3cdfd=Pcd eth19THORN

References

Anderson K and Nelgen S (2011) Global Wine Markets 1961 to 2009 A StatisticalCompendium Adelaide University of Adelaide Press Also available as an e-book atwwwadelaideeduaupresstitlesglobal-wine

Anderson K Nelgen S and Pinilla V (2017) Global Wine Markets 1860 to 2016 AStatistical Compendium Adelaide University of Adelaide Press (forthcoming) Also tobe available as an e-book at wwwadelaideeduaupress

Anderson K and Pinilla V (with the assistance of A J Holmes) (2017)Annual Database ofGlobal Wine Markets 1835 to 2016 Wine Economics Research Centre University ofAdelaide to be posted in October at wwwadelaideeduauwine-econdatabases

Anderson K and Strutt A (2016) Impacts of Asiarsquos rise on African and Latin Americantrade Projections to 2030 World Economy 39(2) 172ndash194

Anderson K andWittwer G (2013) Modeling global wine markets to 2018 Exchange ratestaste changes and Chinarsquos import growth Journal of Wine Economics 8(2) 131ndash158

Anderson K and Wittwer G (2015) Asiarsquos evolving role in global wine markets ChinaEconomic Review 35 1ndash14

250 UK and Global Wine Markets by 2025 and Implications of Brexit

Armington P A (1969) A theory of demand for products distinguished by place of produc-tion IMF Staff Papers 16 159ndash178

Baldwin R E (ed) (2016) Brexit Beckons Thinking Ahead by Leading Economists LondonCentre for Economic Policy Research

Campos N F (2016) Lousy experts Looking back at the ex ante estimates of the costs ofBrexit In R E Baldwin (ed) Brexit Beckons Thinking Ahead by Leading Economists(ch 3) London Centre for Economic Policy Research

Dixon P Parmenter B Sutton J and Vincent D (1982) ORANI A Multisectoral Modelof the Australian Economy Contributions to Economic Analysis 142 Amsterdam North-Holland

European Commission (2015) Trade for All Towards a More Responsible Trade andInvestment Policy Brussels European Commission

Harrison J Horridge M Jerie M and Pearson K (2014) GEMPACK ManualMelbourne GEMPACK Software

HM Treasury (2017) Spring Budget 2017 London HM TreasuryHolmes A J and Anderson K (2017a)Annual Database of National Beverage Consumption

Volumes and Expenditures 1950 to 2015 Wine Economics Research Centre University ofAdelaide posted at wwwadelaideeduauwine-econdatabases

Holmes A J and Anderson K (2017b) Convergence in National Alcohol ConsumptionPatterns New Global Indicators Journal of Wine Economics 12(2) 117ndash148

Ludington C C (2013) The Politics of Wine in Britain A New Cultural History BasingstokeUK Palgrave Macmillan

Office of National Statistics (2017) Analysis of real earnings London Office of NationalStatistics Available at httpswwwonsgovukemploymentandlabourmarketpeopleinworkearningsandworkinghoursarticlessupplementaryanalysisofaverageweeklyearningslatest

Rollo J Borchert I Dawar K Holmes P and Winters L A (2016) The World TradeOrganisation A safety net for a post-Brexit UK trade policy Briefing Paper 1UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-1-final-1pdf

Smith A (2017) Brexit Hard truths and hard choices UKTPO blog 19 June Available athttpsblogssussexacukuktpo20170619brexit-hard-truths-and-hard-choices

Swinbank A (2017) World trade rules and the policy options for British agriculture post-Brexit Briefing Paper 7 UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-7pdf

Viner J (1950) The Customs Union Issue New York Carnegie Endowment for InternationalPeace

Wine Australia (2015) Export Market Guide European Union Adelaide Wine AustraliaWittwer G Berger N and Anderson K (2003) A model of the worldrsquos wine markets

Economic Modelling 20(3) 487ndash506

Kym Anderson and Glyn Wittwer 251

Page 14: U.K. and Global Wine Markets by 2025, and Implications of ... · U.K. and Global Wine Markets by 2025, and Implications of Brexit* Kym Andersona and Glyn Wittwerb Abstract The United

increase in consumption of fine wine In Western Europe and in the SouthernHemispherersquos New World countries fine wines are projected to substitute for com-mercial wines (defined as the sum of commercial-premium and nonpremiumwines) with almost no change in total wine consumption Sub-Saharan Africa isthe next region that is projected to take off with its growth accounting for morethan one-third of the rest of the worldrsquos increase in volume consumed

Those differences in production versus consumption rankings are reflected ininternational trade Figure 6 shows that France Italy and Spain remain the threedominant exporters of wine in aggregate value but the rankings of the next fewchange with Australia being slightly ahead of Chile and the United StatesGermany and New Zealand being nearly tied for sixth place in value termsFrance and Italy are even more dominant in fine-wine exports and remain so by2025 while Italy outranks France in the commercial-premium export categoryand Spain outranks Italy Australia and Chile in the nonpremium export class

Among the importers the United States and the United Kingdom are projected tocontinue to hold the first two places in 2025 in value terms but China moves intothird place slightly ahead of Germany followed well behind by Canada HongKong Belgium-Luxembourg the Netherlands and Japan (Figure 6(b)) OtherAfrica (excluding South Africa) is projected to experience the largest increase inimports among all the other regions followed by Other Asia which becomes asbig as Germany in value terms (Figure 7(a)) In terms of total volume of wineimports Germany and the United Kingdom held the top two shares in 2014 butby 2025 the United Kingdom is projected to be well ahead of Germany (Figure 7(b)) However this projection ignores the effects of Brexit to which we now turn

VI How Might Wine Markets Be Affected by the United Kingdomrsquos Exitfrom the European Union

As mentioned earlier for our two alternative scenarios to capture the initial effects ofBrexit (ldquolargerdquo and ldquosmallrdquo) we assume that following the United Kingdomrsquos exitfrom the European Union the United Kingdomrsquos rate of economic growth wouldbe only one-third or two-thirds as fast for the period to 2025 the British poundwould be 20 or 10 lower in real terms than in our modelrsquos baseline projectionand the United Kingdom would apply the European Unionrsquos external tariff onwine to imports from EU member countries (as part of establishing MFN ratesvia the WTO to then start new bilateral FTA negotiations) In these initial scenarioswe assume the United Kingdom does not implement any new FTAs particularlywith the EU27 Chile and South Africa Following the discussion of those initialresults we present the results of a subsequent scenario that includes the implemen-tation of an FTA between the United Kingdom and the EU27 We assume these aretwo discrete steps with no agreed arrangements to smooth the transition betweenthem

234 UK and Global Wine Markets by 2025 and Implications of Brexit

Figure 6

Value of Wine Exports and Imports Key Wine-Trading Countries 2014 and Projected Baselinefor 2025 (2014 US$ million)

Fig6-BW

onlin

eBW

inprint

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 235

Figure 7

National Shares of Global Wine Import Value and Volume 2014 and Projected Baselinefor 2025 ()

Fig7-BW

onlin

eBW

inprint

Source Authorsrsquo model results

236 UK and Global Wine Markets by 2025 and Implications of Brexit

A Initial Impact of Brexit

Because the British pound dropped by one-sixth against the US dollar in the fourmonths following the Brexit vote in June 2016 and because the UnitedKingdomrsquos average real wage fell 1 between November 2016 and April 2017(according to the latest available data see Office of National Statistics 2017) weviewed our ldquolargerdquo scenario as more likely soon after the Brexit vote But wecompare those results with ones from our ldquosmallrdquo initial-impact scenario becauseit has become more likely following the June 2017 general election that ldquosofterrdquooptions will be considered by the new government Generally the results areabout half the size in the ldquosmallrdquo scenario with the exception of the bilateraltrade effects To show the sensitivity of results to our assumptions we point outthe differences when the ldquosmallrdquo results are not close to half the results shown forthe ldquolargerdquo scenario

Before turning to those results we first present a ldquofree-traderdquo scenario whichassumes the United Kingdom chooses the radical option of becoming the HongKong of Europe Unlikely as this scenario is the results provide assurance thatour global wine model generates the standard types of effects of exiting a customsunion In this ldquofree-traderdquo scenario the United Kingdom is assumed to move tozero tariffs on all wine imports Thus the current preferences on EU Chileanand South African wine imports disappear as all other wine exporters enjoy thesame free access to UK wine markets We assume in this scenario that Brexit hasno impact on the United Kingdomrsquos real income or consumption growth rates by2025 and that by then the British pound has returned to its 2014 value relative toother currencies (similar to the real exchange rate assumed in our baseline scenarioto 2025 see Appendix Table A1)7 Under these assumptions EU Chilean andSouth African wine exports to the United Kingdom in 2025 are 003 lower thanin the no-Brexit baseline and UK imports from other wine exporters are 025higher but the United Kingdomrsquos total wine imports are hardly any different(because local wine prices fall by only 05 and incomes are unchanged) In this sce-nario UK consumers expand their volume of consumption by just 05 while theEU27 exports more and other countries export less to the rest of the world such thatworld wine exports are almost unchanged In short all these effects are in the direc-tion that comparative static-customs-union theory would predict but they are smallbecause we assume the United Kingdomrsquos real income and currency are the same asin our baseline for 2025 as described in the previous section

In the ldquolargerdquo alternate initial scenario involving Brexit as compared with thebaseline scenario to 2025 the consumer price of wine in 2025 is 22 higher in the

7If this scenario were seen as a credible long-run trade policy it would encourage more investment thatwould eventually raise UK incomes but not without major structural changes that would initiallydisrupt the economy Implicitly we assume in our free-trade scenario that these two opposite impactson real UK incomes and the British pound exactly offset each other by 2025

Kym Anderson and Glyn Wittwer 237

United Kingdom in local currency terms (20 because of real depreciation of theBritish pound 4 because of new tariffs on EU Chilean and South Africanwines and ndash2 because of slower UK income growth) The volume of UKwine consumption is 28 lower 16 because of slower UK economic growth7 because of real depreciation of the British pound and 5 because of newtariffs Superpremium still-wine sales are the most affected dropping by two-fifths while sparkling and commercial-premium wines drop a bit less than one-quarter Because the average price rises by more than the decline in volume soldthe aggregate value of UK sales even in local currency terms falls under thisldquolargerdquo Brexit scenario Under the ldquosmallrdquo Brexit scenario the consumer price ofwine in 2025 is 11 higher in the United Kingdom and its volume of wine consump-tion is 17 lower

The volume of projected UK imports in 2025 is 427 ML or nearly one-quarterless in the ldquolargerdquo scenario than in the baseline scenario comprising 58ML less spar-kling wine 31 ML less superpremium still wine and 339 ML less commercial-premium wine World imports are lower by just 239 ML because imports byother countries are 189 ML higher in response to the lower international wineprices in this scenario In value terms UK imports are $175 billion (or 27)lower in 2025 because of ldquolargerdquo Brexit $113 billion because of lower incomes$038 billion because of the fall in the British pound and $014 billion because ofthe rise in wine-import tariffs (Table 4) These aggregate trade impacts are a littlemore than half this size under the ldquosmallrdquo Brexit scenario

Despite the import levels falling because of raised import tariffs domestic con-sumption of all three quality categories of UK-produced wine is lower with thanwithout Brexit because of shrunken demand for all wines resulting from loweredUK incomes and raised local prices because of devaluation of the British poundThe British poundrsquos devaluation does make it easier for the United Kingdom tosell wines abroad though The countryrsquos exports are 7 ML or nearly 5 higherin 2025 in the ldquolargerdquo Brexit scenario and UK production is 3 higher ThoseUK exports (or re-exports of imported bulk wine after it is bottled in the UnitedKingdom) that go to EU27 countries are reduced though because of the tariffnow imposed at the new EU border

Without Brexit the United Kingdomrsquos shares of global wine imports are slightlyhigher in volume terms in 2025 than in 2010ndash2015 but 2 percentage points lower invalue terms thanks to East Asiarsquos expanding demand for imports of premium winesWith ldquolargerdquo Brexit however that value share is an additional 2 percentage pointslower and the volume share is almost 5 points lower (Figure 8) The net effects ofthese impacts on global trade are shown in Figure 9 Most of the initial tradeeffect of Brexit is a large decline in net imports of wine by the United Kingdomwith very little offsetting positive effect on trade in the rest of the world Theldquosmallrdquo Brexit numbers are a bit more than half of those for ldquolargerdquo Brexit

238 UK and Global Wine Markets by 2025 and Implications of Brexit

Table4

Difference

in2025

Projected

VolumeandValue

ofWineIm

portsb

ytheUnitedKingdom

andtheResto

fthe

World

asaConsequence

oftheInitial

BrexitS

hock

(MLand2014US$

Million

ldquoLargerdquoSc

enario)

Volum

e(M

L)

Value

(US$

million)

NP+CPa

SuperPrb

Sparkling

TOTAL

NP+CPa

SuperPrb

Sparkling

TOTAL

ΔUK

impo

rtsdu

eto

Low

erincomes

minus198

minus20

minus29

minus247

58minus644

minus253

minus234

minus1131

65Low

erpo

und

minus70

minus10

minus14

minus93

22minus248

minus127

minus102

minus476

27Highertariffs

minus71

minus1minus16

minus87

20minus110

minus8minus2

4minus143

8TOTA

Lminus339

minus31

minus58

minus427

100

minus1001

minus388

minus360

minus1750

100

difffrom

base

2332

3325

2432

3227

of

totalc

uts

797

14100

5722

21100

ΔROW

netim

ports

143

2125

189

230

143

minus181

192

ΔWORLD

TRADE

minus195

minus10

minus34

minus239

minus763

minus246

minus543

minus1552

aNon

prem

ium

plus

commercial-premium

still

wines

bSu

perpremium

still

wines

SourceA

utho

rsrsquomod

elresults

Kym Anderson and Glyn Wittwer 239

The aggregate effect of ldquolargerdquo Brexit on the market shares of various wine-export-ing countries in the United Kingdom is almost indiscernible even with one decimalpoint The projected 2025 shares are quite different from the actual 2014 shares forseveral countries They are much smaller in 2025 for South Africa Australia andNew Zealand (and the United States in volume terms) and are much larger involume for Spain and in value for Italy These results occur because wine-exportingcountries benefit differentially from the varying rates of growth in net importdemand for wine in non-UK countries over this projection period The most impor-tant projected changes are the increase in the real value of annual wine importsbetween 2014 and 2025 by China (200 or $3 billion) Other Asia (110 or $22billion) and Africa (270 or $16 billion) More than half of Australiarsquos increasein annual exports from 2014 to 2025 go to Asia and more than half of SouthAfricarsquos increase in exports go to Other Africa

Table 5 reveals that ldquolargerdquo Brexit lowers EU Chilean and South African wineexports by 150 ML or US$12 billion in the case of the European Union with someof their exports diverted from the United Kingdom to the EU27 and other marketsin competition with New World exporters Although the United States Australiaand Argentina sell only a little less to the United Kingdom they also sell less toother countries For Chile and South Africa which lose their preferential access toUK (but not to EU27) markets in this Brexit scenario some exports are redirected

Figure 8

UK Shares of World Wine Imports 2010ndash2015 and Projected to 2025 without and with theldquoLargerdquo Initial Brexit Shock ()

Fig8-BW

onlin

eBW

inprint

Sources Anderson and Pinilla (2017) and authorsrsquo model results

240 UK and Global Wine Markets by 2025 and Implications of Brexit

from the United Kingdom to EU27 countries ndash but again they export less overallGlobal wine trade in 2025 is 240 ML (19) or $18 billion (35) less under thisldquolargerdquo Brexit scenario The percentage by which wine exportersrsquo trade shrinks isgreater for values than for volumes because of changes in relative prices of differ-ent-quality wines Those differences are shown in the numbers in parentheses inTable 5

Three other points are worth making about Table 5 One is that Australia sellsslightly more to the United Kingdom in the ldquosmallrdquo Brexit scenario rather thanslightly less as in the ldquolargerdquo Brexit case Evidently the negative income and price(devaluation) effects do not more than offset the positive trade-diverting effect onAustralian exports to the United Kingdom of removing preferences in the ldquosmallrdquoscenario Second New Zealand sells slightly more to non-UK countries underBrexit despite greater competition from the EU27 Chile and South Africa Thisanomaly is due to changes in the relative prices of different qualities of wine inglobal wine markets bearing in mind that New Zealand has the worldrsquos highestaverage price for still-wine exports And third the value (but not the volume) ofexports of ldquoOtherrdquo countries to markets other than the United Kingdom arehigher under Brexit This result too is due to changes in the relative prices of differ-ent qualities of wine in global wine markets

Figure 9

Difference in 2025 Wine Import Volumes and Values as a Result of the ldquoLargerdquo Initial BrexitShock (ML and US$ Million in 2014 US dollars)

Fig9-BW

onlin

eBW

inprint

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 241

B Subsequent Impact of Brexit from a UK-EU27 FTA

The next-most-likely step in the Brexit process is for the United Kingdom to nego-tiate a new trade arrangement with the EU27 We therefore assume that a UK-EU27FTAwith free bilateral wine trade is implemented and adjusted to by 2025 and thatprogress toward that end occurs soon enough that the adverse macroeconomicshocks from the initial impact of uncertainty over the Brexit process are confinedto those assumed in the ldquosmallrdquo scenario outlined above In this subsequent scenariowe assume that the British pound returns to the value it would have reached in 2025in the absence of Brexit (ie reversing the 10 devaluation assumed in the ldquosmallrdquoinitial Brexit scenario) and that real incomes in the United Kingdom are 8 higher

Table 5Difference in 2025 BilateralWine Import Volumes and Values fromKey Exporters by the United

Kingdom and the Rest of the World (RoW) as a Result of Initial Brexit Shock(ML and 2014US$ Million)a

(a) ldquolargerdquo scenario

Volume (ML) Value (2014US$ million)

UK RoW WORLD () UK RoW WORLD ()

EU27 minus287 136 minus150 (ndash17) minus1187 minus5 minus1192 (ndash31)Chile minus59 35 minus25 (ndash30) minus169 31 minus138 (ndash48)South Africa minus53 35 minus18 (ndash32) minus105 20 minus85 (ndash67)USA minus7 minus6 minus13 (ndash24) minus75 minus40 minus115 (ndash50)Australia minus4 minus3 minus7 (ndash09) minus25 minus65 minus90 (ndash30)Argentina minus3 minus9 minus12 (ndash48) minus16 minus39 minus55 (ndash52)New Zealand minus11 9 minus2 (ndash09) minus162 71 minus91 (ndash43)Others minus2 minus10 minus12 (ndash02) minus11 minus52 minus63 (ndash44)WORLD minus427 187 minus240 (ndash19) minus1750 minus79 minus1829 (ndash35)

(a) ldquosmallrdquo scenario

Volume (ML) Value (2014US$ million)

UK RoW WORLD () UK RoW WORLD ()

EU27 minus178 82 minus96 (ndash12) minus692 minus43 minus736 (ndash19)Chile minus46 28 minus18 (ndash24) minus128 36 minus91 (ndash32)South Africa minus43 29 minus14 (ndash42) minus82 23 minus59 (ndash47)USA 1 minus6 minus5 (ndash11) minus23 minus28 minus51 (ndash22)Australia 5 minus10 minus5 (ndash06) 19 minus56 minus38 (ndash13)Argentina 0 minus6 minus6 (ndash26) minus3 minus25 minus29 (ndash27)New Zealand minus5 4 minus1 (ndash06) minus80 34 minus46 (ndash22)Others 0 minus9 minus9 (ndash01) minus1 minus33 minus34 (ndash24)WORLD minus266 112 minus154 (ndash13) minus991 minus92 minus1083 (ndash21)

a Numbers in parentheses are the percentage difference between the Brexit and baseline scenarios for 2025 projected wine-import volumes orvalues by source

Source Authorsrsquo model results

242 UK and Global Wine Markets by 2025 and Implications of Brexit

in 2025 than in the ldquosmallrdquo scenario ndash that is they regain two-thirds of the value theywould have reached without Brexit-related uncertainty

This subsequent development in the Brexit process reverses most of but not allthe initial effects of Brexit by 2025 because of our assumption that the lost growthin the initial years of uncertainty following the Brexit vote are only partly recoveredby 2025 following the implementation of a UK-EU27 FTA Moreover the longer ittakes before this FTA is finalized and implemented the longer the estimated initialadverse effects persist and the larger the cumulative cost of Brexit to UK wine con-sumers and to grape and wine producers in wine-exporting countries

Table 6 summarizes the subsequent trade effects for 2025 It suggests that all butone-tenth of the loss in value of world trade in wine from the initial ldquosmallrdquo impact isrestored most of it because of a smaller reduction in the United Kingdomrsquos wineimports following the FTA Most of that improved outcome is because of recoveredimports from the EU27 commensurate with the latterrsquos high share of UK importsEven though Chile and South Africa are assumed in this scenario to have not yetsigned an FTA with the United Kingdom they export slightly more to the UnitedKingdom (and even more to all other countries) than in the initial ldquosmallrdquo scenario

Within the United Kingdom this FTA brings down the local currency consumerprice of wine by 9 largely offsetting the 11 rise in the initial ldquosmallrdquo Brexit sce-nario and it raises the volume consumed in the United Kingdom by 18 fully off-setting the 17 fall in the initial ldquosmallrdquo Brexit scenario

Table 6Difference in 2025 Bilateral Wine-Import Volumes and Values from Key Exporters by the

United Kingdom and the Rest of the World (RoW) as a Result of Implementing a UK-EU27FTA Following Initial ldquoSmallrdquo Brexit Shock (ML and 2014US$ Million Difference Relative to

Initial ldquoSmallrdquo Brexit Scenario)a

Volume (ML) Value (2014US$ million)

UK RoW WORLD (ldquosmallrdquo)a UK RoW WORLD (ldquosmallrdquo)a

EU27 212 minus112 100 (minus96) 750 0 750 (minus806)Chile 1 3 4 (minus18) 3 40 43 (minus92)South Africa 3 2 5 (minus14) 3 18 21 (minus60)USA 3 3 6 (minus5) 26 21 47 (minus50)Australia 1 1 2 (minus5) 8 38 46 (minus33)Argentina 1 4 5 (minus6) 4 20 24 (minus28)New Zealand 4 minus3 1 (minus1) 61 minus25 36 (minus46)Others 1 6 7 (minus9) 5 minus46 minus41 (92)WORLD 226 minus96 130 (minus154) 860 66 926 (minus1022)

a Numbers in parentheses are the world trade differences between the ldquosmallrdquo initial Brexit scenario and the baseline scenario copied fromcolumns 3 and 7 of Table 5(b)

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 243

In short this subsequent step in the Brexit process can be expected to restore muchof the initial adverse effects in the United Kingdom of the Brexit vote by 2025 Thewine exporters in the EU27 Chile and South Africa also lose less in this scenarioas compared with the initial ldquosmallrdquo scenario while other key exporters sell nearlyas much in 2025 in this scenario as they are projected to in the baseline ndash that isthey are beneficiaries of the reduced discrimination in global wine markets in thiscase where Chile and South Africa no longer enjoy preferential access to the UKmarket (and even though EU27 producers are again allowed to do so) Keep inmind though that although the estimated losses in the initial scenarios are reportedjust for 2025 they are expected to be felt throughout the preceding years of uncertaintyTherefore the longer it is before uncertainty abates and the United Kingdomrsquos currentpreferential trading arrangements are replaced by newones such as a UK-EU27 FTAthe higher the cumulative cost of the Brexit vote to the wine trade

VII Caveats and Conclusions

The above Brexit simulations are just a few of many scenarios that could be modeledObvious additional ones could also assume that FTAs are reached between theUnited Kingdom and other trading partners including Chile and South Africaand such countries as Australia and New Zealand Some of these other countriesare already exploring the ECrsquos invitation (see European Commission 2015) to nego-tiate an FTA with what will be the EU27 The sequence in which FTAs are signedand the speed with which they are implemented will matter (as was also the casewith the sequential signing over the past decade of bilateral FTAs with NortheastAsian countries by Chile Australia and New Zealand see Anderson andWittwer 2015) Even if the United Kingdom were able to sign additional bilateralFTAs and begin implementing them before 2025 it would have little effect on theabove results (because wine tariffs are such a minor contributor to them) unlessthose FTAs were to accelerate the United Kingdomrsquos economic growth and therise in the British poundrsquos value before 2025

We assume above that no changes are made to alcohol excise duties in the UnitedKingdom following Brexit even though they are scheduled to be progressively raisedwith inflation and may be raised even more for wine relative to spirits to offset theopposite effects of Brexit on those two domestic industries Nor do we make anyallowance for UK increases in consumer wine prices that may be needed to coverthe higher cost of clearing customs on imports from EU27 countries (which maybe more or less offset by increased smuggling and duty-free purchasing)Importers of fine wines including individuals who buy Bordeaux en primeur maywell reduce their demand because of concerns that delays on docks will affect thequality of their wines or because of greater currency uncertainty

To summarize Brexit is costly initially to UK consumers of wine (and of manyother tradable products) because the domestic retail price in local currency tax-

244 UK and Global Wine Markets by 2025 and Implications of Brexit

inclusive terms is 22 higher than otherwise in 2025 and the volume of wine con-sumed domestically is 28 lower in the ldquolargerdquo Brexit scenario (or 11 higher pricesand 17 lower quantities in the ldquosmallrdquo case) by 2025 unless a UK-EU27 FTA isdeveloped before then Even if such an FTA does get signed ratified by all 28 par-liaments and implemented by 2025 the slower income growth in the interim meansa smaller UK wine market in 2025 than would otherwise have been the case Thevolume reduction is a blow to many participants in UK wine-bottling transportingstoring wholesaling and retailing businesses in addition to restaurants and pubsVery little of that initial impact occurs because of higher import tariffs most impor-tant is the assumed fall in UK real incomes

The small but growing number of local UK vignerons (who supply less than 05of all domestic-wine sales and less than 4 of sparkling-wine sales) is initially pro-jected to sell less wine domestically because of reduced demand and to export only alittle extra abroad while the British pound is devalued Overall their production isonly 3 higher in 2025 even in the ldquolargerdquo Brexit scenario (not taking intoaccount the fact that their casual labor is likely to be more expensive under Brexitbecause of tighter restrictions on immigration) and their production diminishes ifand when a UK-EU27 FTA is in place

Even with a UK-EU27 FTA in place EU27 wine exporters export US$56 millionless wine in 2025 thanks to Brexit Chile and South Africa export $88 million lesswine and wine exports from the rest of the world are little different in aggregatealthough slightly larger from Australia

Clearly there will be great uncertainly for some time over the possible policy out-comes to flow from Brexit and their consequent sequential impacts on UK house-hold disposable incomes foreign exchange rates and bilateral wine tariffsMeanwhile the above projections under explicit assumptions provide some idea ofhow wine markets might be affected by the most-likely first two stages of theBrexit process (agreeing on a new tariff schedule at the WTO and agreeing to andimplementing a UK-EU27 FTA) In particular they make clear that nontrivialinitial impacts could affect the domestic wine market impacts that are likely to belarger than just the direct impact of changes in bilateral tariffs If the UnitedKingdom succeeds in getting countries to agree to transition arrangements thatdelay the changes in tariffs until new FTAs are signed and ratified the initialeffects will be less dramatic than in our first scenarios but it remains to be seenwhether any such agreements can be reached In any event the net effect of Brexiton the welfare of the worldrsquos consumers and producers of wine as a whole will benegative not just initially but permanently unless new trade policy commitmentsby the United Kingdom with major wine-exporting countries are sufficiently moreliberal than current arrangements

Kym Anderson and Glyn Wittwer 245

Appendix Key Equations in the Global Wine Markets Model

Amodel of the worldrsquos wine markets was first published byWittwer et al (2003) andmuch improved by Anderson and Wittwer (2013) Several significant enhancementshave been made to that original model Wine types have been disaggregated intomore types ndash namely nonpremium (including generic bulk) commercial-premiumand superpremium still wines and sparkling wines8 As in the original modelthere are two types of grapes premium and nonpremium Nonpremium wines usenonpremium grapes exclusively superpremium wines use mainly premium grapesand commercial-premium and sparkling wines use both types of grapes In theregional dimension the number of countries and country groups has expandedfrom 10 in the original model to 51 44 individual nations and 7 compositeregions The modelrsquos database is calibrated to 2014 for this paper based on datain Anderson et al (2017) trade data for which are downloaded from httpscomtradeunorg The modelrsquos supply-and-demand equations are based on the orig-inal ORANI modelrsquos theory (see Dixon Parmenter Sutton and Vincent 1982) andthe model is implemented using GEMPACK software (Harrison Horridge Jerieand Pearson 2014)

An enhancement of importance to the present study is the inclusion of exchange-rate variables in the model which allow a distinction between price impacts asobserved in local currency units and those observed in 2014 US dollars

In the model the grape and wine sectors minimize costs of intermediate inputssubject to weak constant elasticity of substitution (CES) substitutability betweeninputs By assumption no intermediate inputs are imported from other countries9Hence

Xcid frac14 f ethX1id CESfrac12P

cid=P1id $THORN eth1THORN

P1id X1id frac14X

c

Xcid P

cid eth2THORN

whereXcid is the quantity demanded of commodity c by grape or wine industry i in

region d Pcid is the corresponding price and X1id and P1id are the respective inter-

mediate composite quantities and prices

Two primary factors are employed in the sector labor (the quantity of which isendogenous with perfectly elastic supply) and capital Capital is usually treated asexogenous in quantity with rates of return bearing all the adjustment in thevarious scenarios This reflects the fact that grapes (a perennial crop) and wine-

8Commercial-premium still wines are defined by Anderson and Nelgen (2011) to be those priced betweenUS$250 and $750 per litre pre-tax at a countryrsquos border or wholesale9An exception concerns bulk wine imports used mainly in the bottling of commercial premium wine in theUK USA Canada Japan and Germany

246 UK and Global Wine Markets by 2025 and Implications of Brexit

plant capacity adjust slowly to market signals

Lid frac14 f ethFid CESfrac12Wid=PFid $THORN eth3THORN

Kid frac14 f ethFid CESethRid=PFid $THORN eth4THORN

PFid Fid frac14 Lid Wid thorn Kid Rid eth5THORN

Grape and wine producers are assumed to minimize costs subject to CES substitu-tion between capital and labor Equations (3) to (5) show primary factor demandsfor the labor composite Lid and capital Kid subject to a composite factor demandFid by industry i in region d The factor prices are Wid for labor Rid for capitalrentals and PFid for composite factor prices

Appendix Table A1Cumulative Consumption and Population Growth Rates and Changes in the Real Exchange Rate

(RER) Relative to the US Dollar 2014 to 2025 without Brexit ()

Aggregateconsumption Poprsquon RER

Aggregateconsumption Poprsquon RER

France 18 4 minus11 Australia 35 11 minus17Italy 11 2 minus9 New Zealand 32 9 minus26Portugal 14 0 minus9 Canada 27 8 minus18Spain 26 8 minus9 United States 31 8 0Austria 19 4 minus7 Argentina 7 10 109Belgium 20 7 minus9 Brazil 16 8 minus29Denmark 22 2 minus9 Chile 55 8 minus2Finland 21 3 minus7 Mexico 42 12 minus8Germany 14 minus2 minus11 Uruguay 45 3 1Greece 22 minus1 minus14 Other Latin America 60 10 minus5Ireland 42 12 minus9 South Africa 36 12 minus1Netherlands 21 4 minus9 Turkey 50 8 20Sweden 24 9 minus13 North Africa 53 11 0Switzerland 18 8 minus6 Other Africa 109 18 84United Kingdom 32 6 1 Middle East 52 18 minus12Other West Europe 21 10 minus1 China 79 3 5Bulgaria 41 minus7 7 Hong Kong 42 3 2Croatia 20 minus2 minus1 India 134 13 17Georgia 35 0 23 Japan 11 minus3 minus24Hungary 25 minus3 minus11 Korea 38 1 minus9Moldova 49 minus11 13 Malaysia 62 15 minus16Romania 45 minus4 22 Philippines 75 18 7Russia 18 minus2 minus8 Singapore 44 21 minus22Ukraine 22 minus5 14 Taiwan 29 1 minus13Other East Europe 40 minus5 48 Thailand 47 3 minus9

Other Asia 99 10 10

Source Authorsrsquo compilation from projections by various international agencies and from global economy-wide modeling by Anderson andStrutt (2016)

Kym Anderson and Glyn Wittwer 247

The composite factor demand Fid is proportional to total output Qid subject to aprimary factor using technology Aid Hence

Fid frac14 Qid Aid eth6THORN

The perfectly competitive zero-pure-profit condition is that total revenue valued atthe output price P0s

i multiplied by Qid equals the total production cost

P0si Qid frac14

X

cPcid X1cid thorn

X

oWo

id Loid thorn Rid Kid eth7THORN

Household demands follow a linear expenditure system in each region We reducethe optimizing problem for household consumption of each commodity subjectto a budget constraint to equations describing subsistence and discretionarydemands Aggregate subsistence expenditure WSUBd depends only on consumerprices P3cd for each commodity and the number of households N as per-capita sub-sistence quantities XSUBcd subject to given preferences are constant

WSUBd frac14X

c

P3cd XSUBcd Nd eth8THORN

Discretionary expenditures for each commodity (the left-hand side of equation (9))are equal to the marginal budget share (βcd) of aggregate discretionary expenditureThis aggregate is the bracketed term on the right-hand side of equation (9) whereW3TOTd is aggregate nominal expenditure

P3cdethX3cd XSUBcd NdTHORN frac14 βcdethW3TOTd WSUBdTHORN eth9THORN

Because real aggregate consumption is usually exogenous in our partial equilibriumsimulations the linear expenditure system determines the consumption shares ofindividual final commodities (ie the five wine types plus a composite of all otherconsumption items) driven by changes in relative prices as faced by domestic con-sumers The income elasticity of demand for each commodity is equal to the mar-ginal budget share divided by the expenditure share This number varies from 05for nonpremium wine to 17 for superpremium still wine The income elasticity ofdemand for other consumption is very close to 10 because wine accounts for anaverage of only 03 of aggregate expenditures globally and no more than 11in any country (Anderson and Nelgen 2011 Table 166)

A feature of our revised model of world wine markets is the inclusion of nominalexchange rates These appear directly in the equation linking retail prices (P3scd ) toproducer prices by country of origin (P0s

c ) where c denotes the wine type

P3 scd frac14 P0s

cfd

fsTtarcd T

taxcd thorn Pm

cd eth10THORN

248 UK and Global Wine Markets by 2025 and Implications of Brexit

The exchange rates in the consuming and producing regions are fd and fs respec-tively expressed as local currency units per $US Ttar

cd is the power of the tariff in theconsuming region and Ttax

cd is the power of the domestic consumption (or excise) taxprior to any generic value-added or goods-and-services tax Pm

cd is the price of marginm assumed to be locally supplied nontradable and therefore unaffected by theexchange rate

A given level of consumption for wine type c (X3cd) is satisfied using theArmington (1969) assumption in which wines from different countries of originare imperfectly substitutable First domestic wine is imperfectly substitutable witha composite of imports

X3sscd frac14 f ethX3cd CESethP3sscd=P3cdTHORNTHORN ss frac14 domestic imports eth11THORN

and then imports by origin (X3scd ) are determined in a second CES equation

X3scd frac14 f ethX3ssfrac14importscd CESethP3scd=P3

ssfrac14importscd THORNTHORN eth12THORN

The model enables us to show how changes in international competitiveness affectthe worldrsquos wine markets A crucial part of this exercise is explaining how pricesdetermined outside the grape and wine markets influence these markets Becausethe model is in partial equilibrium to depict the impacts of changes in internationalcompetitiveness outside price changes need to be imposed as shocks on the modelThe price of intermediate inputs other than grapes and wine shown in equations (1)and (2) is set equal to the price of GDP (Pg

d ) multiplied by a shifter Fcd

Pcid frac14 Fc

dPgd eth13THORN

If no specific price observations are available the shifter Fcd remains exogenous and

unshocked with the change in price being determined by a shock to the price ofGDP If observations are available for specific input price movements the shifterFc

dbecomes endogenous with Pc

id now exogenous and shocked

Wid frac14 Fwd P

gd eth14THORN

Wage rates are treated similarly In equation (14) if the wage shifter Fwd is exogenous

changes in wage rates Wid are determined by changes in the price of GDP If wagerate data are available Fw

d becomes endogenous andwage rates are shocked directly

Pmcd frac14 Fm

d Pgd eth15THORN

The prices of trade and transport margins are also determined by the price of GDP ifthe shifter Fm

d in equation (15) is exogenous

Changes in international competitiveness depend on changes in relative pricelevels and changes in nominal exchange rates In equation (16) fR

s denotes real

Kym Anderson and Glyn Wittwer 249

exchange-rate movements relative to the US dollar

fRs frac14 Pg

s =frac12PgUSA

fs$ eth16THORN

In equation (16) the nominal exchange rate for the United States is alwaysunchanged because nominal and real exchange rates are expressed relative to theUS currency

Changes in international market conditions may have impacts in one direction onproducer prices as expressed in US dollars P0s

i and potentially in the opposite direc-tion in local currency units Hence we calculate real producer pricesP0s

iloc in localcurrency terms (ie the price most relevant to domestic producers)

P0siloc frac14 P0s

ifs=P

gs eth17THORN

To obtain real price changes in local currency terms we convert US dollar prices(P3 s

cd for the source-specific price and P3cd for the source-composite price) to reallocal currency prices (P3 s

cdloc and P3cdloc) using the consumer price index (CPI)(Pc

d ) as the deflator

P3 scdloc frac14 P3 s

cdfd=P

cd eth18THORN

and

P3cdloc frac14 P3cdfd=Pcd eth19THORN

References

Anderson K and Nelgen S (2011) Global Wine Markets 1961 to 2009 A StatisticalCompendium Adelaide University of Adelaide Press Also available as an e-book atwwwadelaideeduaupresstitlesglobal-wine

Anderson K Nelgen S and Pinilla V (2017) Global Wine Markets 1860 to 2016 AStatistical Compendium Adelaide University of Adelaide Press (forthcoming) Also tobe available as an e-book at wwwadelaideeduaupress

Anderson K and Pinilla V (with the assistance of A J Holmes) (2017)Annual Database ofGlobal Wine Markets 1835 to 2016 Wine Economics Research Centre University ofAdelaide to be posted in October at wwwadelaideeduauwine-econdatabases

Anderson K and Strutt A (2016) Impacts of Asiarsquos rise on African and Latin Americantrade Projections to 2030 World Economy 39(2) 172ndash194

Anderson K andWittwer G (2013) Modeling global wine markets to 2018 Exchange ratestaste changes and Chinarsquos import growth Journal of Wine Economics 8(2) 131ndash158

Anderson K and Wittwer G (2015) Asiarsquos evolving role in global wine markets ChinaEconomic Review 35 1ndash14

250 UK and Global Wine Markets by 2025 and Implications of Brexit

Armington P A (1969) A theory of demand for products distinguished by place of produc-tion IMF Staff Papers 16 159ndash178

Baldwin R E (ed) (2016) Brexit Beckons Thinking Ahead by Leading Economists LondonCentre for Economic Policy Research

Campos N F (2016) Lousy experts Looking back at the ex ante estimates of the costs ofBrexit In R E Baldwin (ed) Brexit Beckons Thinking Ahead by Leading Economists(ch 3) London Centre for Economic Policy Research

Dixon P Parmenter B Sutton J and Vincent D (1982) ORANI A Multisectoral Modelof the Australian Economy Contributions to Economic Analysis 142 Amsterdam North-Holland

European Commission (2015) Trade for All Towards a More Responsible Trade andInvestment Policy Brussels European Commission

Harrison J Horridge M Jerie M and Pearson K (2014) GEMPACK ManualMelbourne GEMPACK Software

HM Treasury (2017) Spring Budget 2017 London HM TreasuryHolmes A J and Anderson K (2017a)Annual Database of National Beverage Consumption

Volumes and Expenditures 1950 to 2015 Wine Economics Research Centre University ofAdelaide posted at wwwadelaideeduauwine-econdatabases

Holmes A J and Anderson K (2017b) Convergence in National Alcohol ConsumptionPatterns New Global Indicators Journal of Wine Economics 12(2) 117ndash148

Ludington C C (2013) The Politics of Wine in Britain A New Cultural History BasingstokeUK Palgrave Macmillan

Office of National Statistics (2017) Analysis of real earnings London Office of NationalStatistics Available at httpswwwonsgovukemploymentandlabourmarketpeopleinworkearningsandworkinghoursarticlessupplementaryanalysisofaverageweeklyearningslatest

Rollo J Borchert I Dawar K Holmes P and Winters L A (2016) The World TradeOrganisation A safety net for a post-Brexit UK trade policy Briefing Paper 1UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-1-final-1pdf

Smith A (2017) Brexit Hard truths and hard choices UKTPO blog 19 June Available athttpsblogssussexacukuktpo20170619brexit-hard-truths-and-hard-choices

Swinbank A (2017) World trade rules and the policy options for British agriculture post-Brexit Briefing Paper 7 UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-7pdf

Viner J (1950) The Customs Union Issue New York Carnegie Endowment for InternationalPeace

Wine Australia (2015) Export Market Guide European Union Adelaide Wine AustraliaWittwer G Berger N and Anderson K (2003) A model of the worldrsquos wine markets

Economic Modelling 20(3) 487ndash506

Kym Anderson and Glyn Wittwer 251

Page 15: U.K. and Global Wine Markets by 2025, and Implications of ... · U.K. and Global Wine Markets by 2025, and Implications of Brexit* Kym Andersona and Glyn Wittwerb Abstract The United

Figure 6

Value of Wine Exports and Imports Key Wine-Trading Countries 2014 and Projected Baselinefor 2025 (2014 US$ million)

Fig6-BW

onlin

eBW

inprint

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 235

Figure 7

National Shares of Global Wine Import Value and Volume 2014 and Projected Baselinefor 2025 ()

Fig7-BW

onlin

eBW

inprint

Source Authorsrsquo model results

236 UK and Global Wine Markets by 2025 and Implications of Brexit

A Initial Impact of Brexit

Because the British pound dropped by one-sixth against the US dollar in the fourmonths following the Brexit vote in June 2016 and because the UnitedKingdomrsquos average real wage fell 1 between November 2016 and April 2017(according to the latest available data see Office of National Statistics 2017) weviewed our ldquolargerdquo scenario as more likely soon after the Brexit vote But wecompare those results with ones from our ldquosmallrdquo initial-impact scenario becauseit has become more likely following the June 2017 general election that ldquosofterrdquooptions will be considered by the new government Generally the results areabout half the size in the ldquosmallrdquo scenario with the exception of the bilateraltrade effects To show the sensitivity of results to our assumptions we point outthe differences when the ldquosmallrdquo results are not close to half the results shown forthe ldquolargerdquo scenario

Before turning to those results we first present a ldquofree-traderdquo scenario whichassumes the United Kingdom chooses the radical option of becoming the HongKong of Europe Unlikely as this scenario is the results provide assurance thatour global wine model generates the standard types of effects of exiting a customsunion In this ldquofree-traderdquo scenario the United Kingdom is assumed to move tozero tariffs on all wine imports Thus the current preferences on EU Chileanand South African wine imports disappear as all other wine exporters enjoy thesame free access to UK wine markets We assume in this scenario that Brexit hasno impact on the United Kingdomrsquos real income or consumption growth rates by2025 and that by then the British pound has returned to its 2014 value relative toother currencies (similar to the real exchange rate assumed in our baseline scenarioto 2025 see Appendix Table A1)7 Under these assumptions EU Chilean andSouth African wine exports to the United Kingdom in 2025 are 003 lower thanin the no-Brexit baseline and UK imports from other wine exporters are 025higher but the United Kingdomrsquos total wine imports are hardly any different(because local wine prices fall by only 05 and incomes are unchanged) In this sce-nario UK consumers expand their volume of consumption by just 05 while theEU27 exports more and other countries export less to the rest of the world such thatworld wine exports are almost unchanged In short all these effects are in the direc-tion that comparative static-customs-union theory would predict but they are smallbecause we assume the United Kingdomrsquos real income and currency are the same asin our baseline for 2025 as described in the previous section

In the ldquolargerdquo alternate initial scenario involving Brexit as compared with thebaseline scenario to 2025 the consumer price of wine in 2025 is 22 higher in the

7If this scenario were seen as a credible long-run trade policy it would encourage more investment thatwould eventually raise UK incomes but not without major structural changes that would initiallydisrupt the economy Implicitly we assume in our free-trade scenario that these two opposite impactson real UK incomes and the British pound exactly offset each other by 2025

Kym Anderson and Glyn Wittwer 237

United Kingdom in local currency terms (20 because of real depreciation of theBritish pound 4 because of new tariffs on EU Chilean and South Africanwines and ndash2 because of slower UK income growth) The volume of UKwine consumption is 28 lower 16 because of slower UK economic growth7 because of real depreciation of the British pound and 5 because of newtariffs Superpremium still-wine sales are the most affected dropping by two-fifths while sparkling and commercial-premium wines drop a bit less than one-quarter Because the average price rises by more than the decline in volume soldthe aggregate value of UK sales even in local currency terms falls under thisldquolargerdquo Brexit scenario Under the ldquosmallrdquo Brexit scenario the consumer price ofwine in 2025 is 11 higher in the United Kingdom and its volume of wine consump-tion is 17 lower

The volume of projected UK imports in 2025 is 427 ML or nearly one-quarterless in the ldquolargerdquo scenario than in the baseline scenario comprising 58ML less spar-kling wine 31 ML less superpremium still wine and 339 ML less commercial-premium wine World imports are lower by just 239 ML because imports byother countries are 189 ML higher in response to the lower international wineprices in this scenario In value terms UK imports are $175 billion (or 27)lower in 2025 because of ldquolargerdquo Brexit $113 billion because of lower incomes$038 billion because of the fall in the British pound and $014 billion because ofthe rise in wine-import tariffs (Table 4) These aggregate trade impacts are a littlemore than half this size under the ldquosmallrdquo Brexit scenario

Despite the import levels falling because of raised import tariffs domestic con-sumption of all three quality categories of UK-produced wine is lower with thanwithout Brexit because of shrunken demand for all wines resulting from loweredUK incomes and raised local prices because of devaluation of the British poundThe British poundrsquos devaluation does make it easier for the United Kingdom tosell wines abroad though The countryrsquos exports are 7 ML or nearly 5 higherin 2025 in the ldquolargerdquo Brexit scenario and UK production is 3 higher ThoseUK exports (or re-exports of imported bulk wine after it is bottled in the UnitedKingdom) that go to EU27 countries are reduced though because of the tariffnow imposed at the new EU border

Without Brexit the United Kingdomrsquos shares of global wine imports are slightlyhigher in volume terms in 2025 than in 2010ndash2015 but 2 percentage points lower invalue terms thanks to East Asiarsquos expanding demand for imports of premium winesWith ldquolargerdquo Brexit however that value share is an additional 2 percentage pointslower and the volume share is almost 5 points lower (Figure 8) The net effects ofthese impacts on global trade are shown in Figure 9 Most of the initial tradeeffect of Brexit is a large decline in net imports of wine by the United Kingdomwith very little offsetting positive effect on trade in the rest of the world Theldquosmallrdquo Brexit numbers are a bit more than half of those for ldquolargerdquo Brexit

238 UK and Global Wine Markets by 2025 and Implications of Brexit

Table4

Difference

in2025

Projected

VolumeandValue

ofWineIm

portsb

ytheUnitedKingdom

andtheResto

fthe

World

asaConsequence

oftheInitial

BrexitS

hock

(MLand2014US$

Million

ldquoLargerdquoSc

enario)

Volum

e(M

L)

Value

(US$

million)

NP+CPa

SuperPrb

Sparkling

TOTAL

NP+CPa

SuperPrb

Sparkling

TOTAL

ΔUK

impo

rtsdu

eto

Low

erincomes

minus198

minus20

minus29

minus247

58minus644

minus253

minus234

minus1131

65Low

erpo

und

minus70

minus10

minus14

minus93

22minus248

minus127

minus102

minus476

27Highertariffs

minus71

minus1minus16

minus87

20minus110

minus8minus2

4minus143

8TOTA

Lminus339

minus31

minus58

minus427

100

minus1001

minus388

minus360

minus1750

100

difffrom

base

2332

3325

2432

3227

of

totalc

uts

797

14100

5722

21100

ΔROW

netim

ports

143

2125

189

230

143

minus181

192

ΔWORLD

TRADE

minus195

minus10

minus34

minus239

minus763

minus246

minus543

minus1552

aNon

prem

ium

plus

commercial-premium

still

wines

bSu

perpremium

still

wines

SourceA

utho

rsrsquomod

elresults

Kym Anderson and Glyn Wittwer 239

The aggregate effect of ldquolargerdquo Brexit on the market shares of various wine-export-ing countries in the United Kingdom is almost indiscernible even with one decimalpoint The projected 2025 shares are quite different from the actual 2014 shares forseveral countries They are much smaller in 2025 for South Africa Australia andNew Zealand (and the United States in volume terms) and are much larger involume for Spain and in value for Italy These results occur because wine-exportingcountries benefit differentially from the varying rates of growth in net importdemand for wine in non-UK countries over this projection period The most impor-tant projected changes are the increase in the real value of annual wine importsbetween 2014 and 2025 by China (200 or $3 billion) Other Asia (110 or $22billion) and Africa (270 or $16 billion) More than half of Australiarsquos increasein annual exports from 2014 to 2025 go to Asia and more than half of SouthAfricarsquos increase in exports go to Other Africa

Table 5 reveals that ldquolargerdquo Brexit lowers EU Chilean and South African wineexports by 150 ML or US$12 billion in the case of the European Union with someof their exports diverted from the United Kingdom to the EU27 and other marketsin competition with New World exporters Although the United States Australiaand Argentina sell only a little less to the United Kingdom they also sell less toother countries For Chile and South Africa which lose their preferential access toUK (but not to EU27) markets in this Brexit scenario some exports are redirected

Figure 8

UK Shares of World Wine Imports 2010ndash2015 and Projected to 2025 without and with theldquoLargerdquo Initial Brexit Shock ()

Fig8-BW

onlin

eBW

inprint

Sources Anderson and Pinilla (2017) and authorsrsquo model results

240 UK and Global Wine Markets by 2025 and Implications of Brexit

from the United Kingdom to EU27 countries ndash but again they export less overallGlobal wine trade in 2025 is 240 ML (19) or $18 billion (35) less under thisldquolargerdquo Brexit scenario The percentage by which wine exportersrsquo trade shrinks isgreater for values than for volumes because of changes in relative prices of differ-ent-quality wines Those differences are shown in the numbers in parentheses inTable 5

Three other points are worth making about Table 5 One is that Australia sellsslightly more to the United Kingdom in the ldquosmallrdquo Brexit scenario rather thanslightly less as in the ldquolargerdquo Brexit case Evidently the negative income and price(devaluation) effects do not more than offset the positive trade-diverting effect onAustralian exports to the United Kingdom of removing preferences in the ldquosmallrdquoscenario Second New Zealand sells slightly more to non-UK countries underBrexit despite greater competition from the EU27 Chile and South Africa Thisanomaly is due to changes in the relative prices of different qualities of wine inglobal wine markets bearing in mind that New Zealand has the worldrsquos highestaverage price for still-wine exports And third the value (but not the volume) ofexports of ldquoOtherrdquo countries to markets other than the United Kingdom arehigher under Brexit This result too is due to changes in the relative prices of differ-ent qualities of wine in global wine markets

Figure 9

Difference in 2025 Wine Import Volumes and Values as a Result of the ldquoLargerdquo Initial BrexitShock (ML and US$ Million in 2014 US dollars)

Fig9-BW

onlin

eBW

inprint

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 241

B Subsequent Impact of Brexit from a UK-EU27 FTA

The next-most-likely step in the Brexit process is for the United Kingdom to nego-tiate a new trade arrangement with the EU27 We therefore assume that a UK-EU27FTAwith free bilateral wine trade is implemented and adjusted to by 2025 and thatprogress toward that end occurs soon enough that the adverse macroeconomicshocks from the initial impact of uncertainty over the Brexit process are confinedto those assumed in the ldquosmallrdquo scenario outlined above In this subsequent scenariowe assume that the British pound returns to the value it would have reached in 2025in the absence of Brexit (ie reversing the 10 devaluation assumed in the ldquosmallrdquoinitial Brexit scenario) and that real incomes in the United Kingdom are 8 higher

Table 5Difference in 2025 BilateralWine Import Volumes and Values fromKey Exporters by the United

Kingdom and the Rest of the World (RoW) as a Result of Initial Brexit Shock(ML and 2014US$ Million)a

(a) ldquolargerdquo scenario

Volume (ML) Value (2014US$ million)

UK RoW WORLD () UK RoW WORLD ()

EU27 minus287 136 minus150 (ndash17) minus1187 minus5 minus1192 (ndash31)Chile minus59 35 minus25 (ndash30) minus169 31 minus138 (ndash48)South Africa minus53 35 minus18 (ndash32) minus105 20 minus85 (ndash67)USA minus7 minus6 minus13 (ndash24) minus75 minus40 minus115 (ndash50)Australia minus4 minus3 minus7 (ndash09) minus25 minus65 minus90 (ndash30)Argentina minus3 minus9 minus12 (ndash48) minus16 minus39 minus55 (ndash52)New Zealand minus11 9 minus2 (ndash09) minus162 71 minus91 (ndash43)Others minus2 minus10 minus12 (ndash02) minus11 minus52 minus63 (ndash44)WORLD minus427 187 minus240 (ndash19) minus1750 minus79 minus1829 (ndash35)

(a) ldquosmallrdquo scenario

Volume (ML) Value (2014US$ million)

UK RoW WORLD () UK RoW WORLD ()

EU27 minus178 82 minus96 (ndash12) minus692 minus43 minus736 (ndash19)Chile minus46 28 minus18 (ndash24) minus128 36 minus91 (ndash32)South Africa minus43 29 minus14 (ndash42) minus82 23 minus59 (ndash47)USA 1 minus6 minus5 (ndash11) minus23 minus28 minus51 (ndash22)Australia 5 minus10 minus5 (ndash06) 19 minus56 minus38 (ndash13)Argentina 0 minus6 minus6 (ndash26) minus3 minus25 minus29 (ndash27)New Zealand minus5 4 minus1 (ndash06) minus80 34 minus46 (ndash22)Others 0 minus9 minus9 (ndash01) minus1 minus33 minus34 (ndash24)WORLD minus266 112 minus154 (ndash13) minus991 minus92 minus1083 (ndash21)

a Numbers in parentheses are the percentage difference between the Brexit and baseline scenarios for 2025 projected wine-import volumes orvalues by source

Source Authorsrsquo model results

242 UK and Global Wine Markets by 2025 and Implications of Brexit

in 2025 than in the ldquosmallrdquo scenario ndash that is they regain two-thirds of the value theywould have reached without Brexit-related uncertainty

This subsequent development in the Brexit process reverses most of but not allthe initial effects of Brexit by 2025 because of our assumption that the lost growthin the initial years of uncertainty following the Brexit vote are only partly recoveredby 2025 following the implementation of a UK-EU27 FTA Moreover the longer ittakes before this FTA is finalized and implemented the longer the estimated initialadverse effects persist and the larger the cumulative cost of Brexit to UK wine con-sumers and to grape and wine producers in wine-exporting countries

Table 6 summarizes the subsequent trade effects for 2025 It suggests that all butone-tenth of the loss in value of world trade in wine from the initial ldquosmallrdquo impact isrestored most of it because of a smaller reduction in the United Kingdomrsquos wineimports following the FTA Most of that improved outcome is because of recoveredimports from the EU27 commensurate with the latterrsquos high share of UK importsEven though Chile and South Africa are assumed in this scenario to have not yetsigned an FTA with the United Kingdom they export slightly more to the UnitedKingdom (and even more to all other countries) than in the initial ldquosmallrdquo scenario

Within the United Kingdom this FTA brings down the local currency consumerprice of wine by 9 largely offsetting the 11 rise in the initial ldquosmallrdquo Brexit sce-nario and it raises the volume consumed in the United Kingdom by 18 fully off-setting the 17 fall in the initial ldquosmallrdquo Brexit scenario

Table 6Difference in 2025 Bilateral Wine-Import Volumes and Values from Key Exporters by the

United Kingdom and the Rest of the World (RoW) as a Result of Implementing a UK-EU27FTA Following Initial ldquoSmallrdquo Brexit Shock (ML and 2014US$ Million Difference Relative to

Initial ldquoSmallrdquo Brexit Scenario)a

Volume (ML) Value (2014US$ million)

UK RoW WORLD (ldquosmallrdquo)a UK RoW WORLD (ldquosmallrdquo)a

EU27 212 minus112 100 (minus96) 750 0 750 (minus806)Chile 1 3 4 (minus18) 3 40 43 (minus92)South Africa 3 2 5 (minus14) 3 18 21 (minus60)USA 3 3 6 (minus5) 26 21 47 (minus50)Australia 1 1 2 (minus5) 8 38 46 (minus33)Argentina 1 4 5 (minus6) 4 20 24 (minus28)New Zealand 4 minus3 1 (minus1) 61 minus25 36 (minus46)Others 1 6 7 (minus9) 5 minus46 minus41 (92)WORLD 226 minus96 130 (minus154) 860 66 926 (minus1022)

a Numbers in parentheses are the world trade differences between the ldquosmallrdquo initial Brexit scenario and the baseline scenario copied fromcolumns 3 and 7 of Table 5(b)

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 243

In short this subsequent step in the Brexit process can be expected to restore muchof the initial adverse effects in the United Kingdom of the Brexit vote by 2025 Thewine exporters in the EU27 Chile and South Africa also lose less in this scenarioas compared with the initial ldquosmallrdquo scenario while other key exporters sell nearlyas much in 2025 in this scenario as they are projected to in the baseline ndash that isthey are beneficiaries of the reduced discrimination in global wine markets in thiscase where Chile and South Africa no longer enjoy preferential access to the UKmarket (and even though EU27 producers are again allowed to do so) Keep inmind though that although the estimated losses in the initial scenarios are reportedjust for 2025 they are expected to be felt throughout the preceding years of uncertaintyTherefore the longer it is before uncertainty abates and the United Kingdomrsquos currentpreferential trading arrangements are replaced by newones such as a UK-EU27 FTAthe higher the cumulative cost of the Brexit vote to the wine trade

VII Caveats and Conclusions

The above Brexit simulations are just a few of many scenarios that could be modeledObvious additional ones could also assume that FTAs are reached between theUnited Kingdom and other trading partners including Chile and South Africaand such countries as Australia and New Zealand Some of these other countriesare already exploring the ECrsquos invitation (see European Commission 2015) to nego-tiate an FTA with what will be the EU27 The sequence in which FTAs are signedand the speed with which they are implemented will matter (as was also the casewith the sequential signing over the past decade of bilateral FTAs with NortheastAsian countries by Chile Australia and New Zealand see Anderson andWittwer 2015) Even if the United Kingdom were able to sign additional bilateralFTAs and begin implementing them before 2025 it would have little effect on theabove results (because wine tariffs are such a minor contributor to them) unlessthose FTAs were to accelerate the United Kingdomrsquos economic growth and therise in the British poundrsquos value before 2025

We assume above that no changes are made to alcohol excise duties in the UnitedKingdom following Brexit even though they are scheduled to be progressively raisedwith inflation and may be raised even more for wine relative to spirits to offset theopposite effects of Brexit on those two domestic industries Nor do we make anyallowance for UK increases in consumer wine prices that may be needed to coverthe higher cost of clearing customs on imports from EU27 countries (which maybe more or less offset by increased smuggling and duty-free purchasing)Importers of fine wines including individuals who buy Bordeaux en primeur maywell reduce their demand because of concerns that delays on docks will affect thequality of their wines or because of greater currency uncertainty

To summarize Brexit is costly initially to UK consumers of wine (and of manyother tradable products) because the domestic retail price in local currency tax-

244 UK and Global Wine Markets by 2025 and Implications of Brexit

inclusive terms is 22 higher than otherwise in 2025 and the volume of wine con-sumed domestically is 28 lower in the ldquolargerdquo Brexit scenario (or 11 higher pricesand 17 lower quantities in the ldquosmallrdquo case) by 2025 unless a UK-EU27 FTA isdeveloped before then Even if such an FTA does get signed ratified by all 28 par-liaments and implemented by 2025 the slower income growth in the interim meansa smaller UK wine market in 2025 than would otherwise have been the case Thevolume reduction is a blow to many participants in UK wine-bottling transportingstoring wholesaling and retailing businesses in addition to restaurants and pubsVery little of that initial impact occurs because of higher import tariffs most impor-tant is the assumed fall in UK real incomes

The small but growing number of local UK vignerons (who supply less than 05of all domestic-wine sales and less than 4 of sparkling-wine sales) is initially pro-jected to sell less wine domestically because of reduced demand and to export only alittle extra abroad while the British pound is devalued Overall their production isonly 3 higher in 2025 even in the ldquolargerdquo Brexit scenario (not taking intoaccount the fact that their casual labor is likely to be more expensive under Brexitbecause of tighter restrictions on immigration) and their production diminishes ifand when a UK-EU27 FTA is in place

Even with a UK-EU27 FTA in place EU27 wine exporters export US$56 millionless wine in 2025 thanks to Brexit Chile and South Africa export $88 million lesswine and wine exports from the rest of the world are little different in aggregatealthough slightly larger from Australia

Clearly there will be great uncertainly for some time over the possible policy out-comes to flow from Brexit and their consequent sequential impacts on UK house-hold disposable incomes foreign exchange rates and bilateral wine tariffsMeanwhile the above projections under explicit assumptions provide some idea ofhow wine markets might be affected by the most-likely first two stages of theBrexit process (agreeing on a new tariff schedule at the WTO and agreeing to andimplementing a UK-EU27 FTA) In particular they make clear that nontrivialinitial impacts could affect the domestic wine market impacts that are likely to belarger than just the direct impact of changes in bilateral tariffs If the UnitedKingdom succeeds in getting countries to agree to transition arrangements thatdelay the changes in tariffs until new FTAs are signed and ratified the initialeffects will be less dramatic than in our first scenarios but it remains to be seenwhether any such agreements can be reached In any event the net effect of Brexiton the welfare of the worldrsquos consumers and producers of wine as a whole will benegative not just initially but permanently unless new trade policy commitmentsby the United Kingdom with major wine-exporting countries are sufficiently moreliberal than current arrangements

Kym Anderson and Glyn Wittwer 245

Appendix Key Equations in the Global Wine Markets Model

Amodel of the worldrsquos wine markets was first published byWittwer et al (2003) andmuch improved by Anderson and Wittwer (2013) Several significant enhancementshave been made to that original model Wine types have been disaggregated intomore types ndash namely nonpremium (including generic bulk) commercial-premiumand superpremium still wines and sparkling wines8 As in the original modelthere are two types of grapes premium and nonpremium Nonpremium wines usenonpremium grapes exclusively superpremium wines use mainly premium grapesand commercial-premium and sparkling wines use both types of grapes In theregional dimension the number of countries and country groups has expandedfrom 10 in the original model to 51 44 individual nations and 7 compositeregions The modelrsquos database is calibrated to 2014 for this paper based on datain Anderson et al (2017) trade data for which are downloaded from httpscomtradeunorg The modelrsquos supply-and-demand equations are based on the orig-inal ORANI modelrsquos theory (see Dixon Parmenter Sutton and Vincent 1982) andthe model is implemented using GEMPACK software (Harrison Horridge Jerieand Pearson 2014)

An enhancement of importance to the present study is the inclusion of exchange-rate variables in the model which allow a distinction between price impacts asobserved in local currency units and those observed in 2014 US dollars

In the model the grape and wine sectors minimize costs of intermediate inputssubject to weak constant elasticity of substitution (CES) substitutability betweeninputs By assumption no intermediate inputs are imported from other countries9Hence

Xcid frac14 f ethX1id CESfrac12P

cid=P1id $THORN eth1THORN

P1id X1id frac14X

c

Xcid P

cid eth2THORN

whereXcid is the quantity demanded of commodity c by grape or wine industry i in

region d Pcid is the corresponding price and X1id and P1id are the respective inter-

mediate composite quantities and prices

Two primary factors are employed in the sector labor (the quantity of which isendogenous with perfectly elastic supply) and capital Capital is usually treated asexogenous in quantity with rates of return bearing all the adjustment in thevarious scenarios This reflects the fact that grapes (a perennial crop) and wine-

8Commercial-premium still wines are defined by Anderson and Nelgen (2011) to be those priced betweenUS$250 and $750 per litre pre-tax at a countryrsquos border or wholesale9An exception concerns bulk wine imports used mainly in the bottling of commercial premium wine in theUK USA Canada Japan and Germany

246 UK and Global Wine Markets by 2025 and Implications of Brexit

plant capacity adjust slowly to market signals

Lid frac14 f ethFid CESfrac12Wid=PFid $THORN eth3THORN

Kid frac14 f ethFid CESethRid=PFid $THORN eth4THORN

PFid Fid frac14 Lid Wid thorn Kid Rid eth5THORN

Grape and wine producers are assumed to minimize costs subject to CES substitu-tion between capital and labor Equations (3) to (5) show primary factor demandsfor the labor composite Lid and capital Kid subject to a composite factor demandFid by industry i in region d The factor prices are Wid for labor Rid for capitalrentals and PFid for composite factor prices

Appendix Table A1Cumulative Consumption and Population Growth Rates and Changes in the Real Exchange Rate

(RER) Relative to the US Dollar 2014 to 2025 without Brexit ()

Aggregateconsumption Poprsquon RER

Aggregateconsumption Poprsquon RER

France 18 4 minus11 Australia 35 11 minus17Italy 11 2 minus9 New Zealand 32 9 minus26Portugal 14 0 minus9 Canada 27 8 minus18Spain 26 8 minus9 United States 31 8 0Austria 19 4 minus7 Argentina 7 10 109Belgium 20 7 minus9 Brazil 16 8 minus29Denmark 22 2 minus9 Chile 55 8 minus2Finland 21 3 minus7 Mexico 42 12 minus8Germany 14 minus2 minus11 Uruguay 45 3 1Greece 22 minus1 minus14 Other Latin America 60 10 minus5Ireland 42 12 minus9 South Africa 36 12 minus1Netherlands 21 4 minus9 Turkey 50 8 20Sweden 24 9 minus13 North Africa 53 11 0Switzerland 18 8 minus6 Other Africa 109 18 84United Kingdom 32 6 1 Middle East 52 18 minus12Other West Europe 21 10 minus1 China 79 3 5Bulgaria 41 minus7 7 Hong Kong 42 3 2Croatia 20 minus2 minus1 India 134 13 17Georgia 35 0 23 Japan 11 minus3 minus24Hungary 25 minus3 minus11 Korea 38 1 minus9Moldova 49 minus11 13 Malaysia 62 15 minus16Romania 45 minus4 22 Philippines 75 18 7Russia 18 minus2 minus8 Singapore 44 21 minus22Ukraine 22 minus5 14 Taiwan 29 1 minus13Other East Europe 40 minus5 48 Thailand 47 3 minus9

Other Asia 99 10 10

Source Authorsrsquo compilation from projections by various international agencies and from global economy-wide modeling by Anderson andStrutt (2016)

Kym Anderson and Glyn Wittwer 247

The composite factor demand Fid is proportional to total output Qid subject to aprimary factor using technology Aid Hence

Fid frac14 Qid Aid eth6THORN

The perfectly competitive zero-pure-profit condition is that total revenue valued atthe output price P0s

i multiplied by Qid equals the total production cost

P0si Qid frac14

X

cPcid X1cid thorn

X

oWo

id Loid thorn Rid Kid eth7THORN

Household demands follow a linear expenditure system in each region We reducethe optimizing problem for household consumption of each commodity subjectto a budget constraint to equations describing subsistence and discretionarydemands Aggregate subsistence expenditure WSUBd depends only on consumerprices P3cd for each commodity and the number of households N as per-capita sub-sistence quantities XSUBcd subject to given preferences are constant

WSUBd frac14X

c

P3cd XSUBcd Nd eth8THORN

Discretionary expenditures for each commodity (the left-hand side of equation (9))are equal to the marginal budget share (βcd) of aggregate discretionary expenditureThis aggregate is the bracketed term on the right-hand side of equation (9) whereW3TOTd is aggregate nominal expenditure

P3cdethX3cd XSUBcd NdTHORN frac14 βcdethW3TOTd WSUBdTHORN eth9THORN

Because real aggregate consumption is usually exogenous in our partial equilibriumsimulations the linear expenditure system determines the consumption shares ofindividual final commodities (ie the five wine types plus a composite of all otherconsumption items) driven by changes in relative prices as faced by domestic con-sumers The income elasticity of demand for each commodity is equal to the mar-ginal budget share divided by the expenditure share This number varies from 05for nonpremium wine to 17 for superpremium still wine The income elasticity ofdemand for other consumption is very close to 10 because wine accounts for anaverage of only 03 of aggregate expenditures globally and no more than 11in any country (Anderson and Nelgen 2011 Table 166)

A feature of our revised model of world wine markets is the inclusion of nominalexchange rates These appear directly in the equation linking retail prices (P3scd ) toproducer prices by country of origin (P0s

c ) where c denotes the wine type

P3 scd frac14 P0s

cfd

fsTtarcd T

taxcd thorn Pm

cd eth10THORN

248 UK and Global Wine Markets by 2025 and Implications of Brexit

The exchange rates in the consuming and producing regions are fd and fs respec-tively expressed as local currency units per $US Ttar

cd is the power of the tariff in theconsuming region and Ttax

cd is the power of the domestic consumption (or excise) taxprior to any generic value-added or goods-and-services tax Pm

cd is the price of marginm assumed to be locally supplied nontradable and therefore unaffected by theexchange rate

A given level of consumption for wine type c (X3cd) is satisfied using theArmington (1969) assumption in which wines from different countries of originare imperfectly substitutable First domestic wine is imperfectly substitutable witha composite of imports

X3sscd frac14 f ethX3cd CESethP3sscd=P3cdTHORNTHORN ss frac14 domestic imports eth11THORN

and then imports by origin (X3scd ) are determined in a second CES equation

X3scd frac14 f ethX3ssfrac14importscd CESethP3scd=P3

ssfrac14importscd THORNTHORN eth12THORN

The model enables us to show how changes in international competitiveness affectthe worldrsquos wine markets A crucial part of this exercise is explaining how pricesdetermined outside the grape and wine markets influence these markets Becausethe model is in partial equilibrium to depict the impacts of changes in internationalcompetitiveness outside price changes need to be imposed as shocks on the modelThe price of intermediate inputs other than grapes and wine shown in equations (1)and (2) is set equal to the price of GDP (Pg

d ) multiplied by a shifter Fcd

Pcid frac14 Fc

dPgd eth13THORN

If no specific price observations are available the shifter Fcd remains exogenous and

unshocked with the change in price being determined by a shock to the price ofGDP If observations are available for specific input price movements the shifterFc

dbecomes endogenous with Pc

id now exogenous and shocked

Wid frac14 Fwd P

gd eth14THORN

Wage rates are treated similarly In equation (14) if the wage shifter Fwd is exogenous

changes in wage rates Wid are determined by changes in the price of GDP If wagerate data are available Fw

d becomes endogenous andwage rates are shocked directly

Pmcd frac14 Fm

d Pgd eth15THORN

The prices of trade and transport margins are also determined by the price of GDP ifthe shifter Fm

d in equation (15) is exogenous

Changes in international competitiveness depend on changes in relative pricelevels and changes in nominal exchange rates In equation (16) fR

s denotes real

Kym Anderson and Glyn Wittwer 249

exchange-rate movements relative to the US dollar

fRs frac14 Pg

s =frac12PgUSA

fs$ eth16THORN

In equation (16) the nominal exchange rate for the United States is alwaysunchanged because nominal and real exchange rates are expressed relative to theUS currency

Changes in international market conditions may have impacts in one direction onproducer prices as expressed in US dollars P0s

i and potentially in the opposite direc-tion in local currency units Hence we calculate real producer pricesP0s

iloc in localcurrency terms (ie the price most relevant to domestic producers)

P0siloc frac14 P0s

ifs=P

gs eth17THORN

To obtain real price changes in local currency terms we convert US dollar prices(P3 s

cd for the source-specific price and P3cd for the source-composite price) to reallocal currency prices (P3 s

cdloc and P3cdloc) using the consumer price index (CPI)(Pc

d ) as the deflator

P3 scdloc frac14 P3 s

cdfd=P

cd eth18THORN

and

P3cdloc frac14 P3cdfd=Pcd eth19THORN

References

Anderson K and Nelgen S (2011) Global Wine Markets 1961 to 2009 A StatisticalCompendium Adelaide University of Adelaide Press Also available as an e-book atwwwadelaideeduaupresstitlesglobal-wine

Anderson K Nelgen S and Pinilla V (2017) Global Wine Markets 1860 to 2016 AStatistical Compendium Adelaide University of Adelaide Press (forthcoming) Also tobe available as an e-book at wwwadelaideeduaupress

Anderson K and Pinilla V (with the assistance of A J Holmes) (2017)Annual Database ofGlobal Wine Markets 1835 to 2016 Wine Economics Research Centre University ofAdelaide to be posted in October at wwwadelaideeduauwine-econdatabases

Anderson K and Strutt A (2016) Impacts of Asiarsquos rise on African and Latin Americantrade Projections to 2030 World Economy 39(2) 172ndash194

Anderson K andWittwer G (2013) Modeling global wine markets to 2018 Exchange ratestaste changes and Chinarsquos import growth Journal of Wine Economics 8(2) 131ndash158

Anderson K and Wittwer G (2015) Asiarsquos evolving role in global wine markets ChinaEconomic Review 35 1ndash14

250 UK and Global Wine Markets by 2025 and Implications of Brexit

Armington P A (1969) A theory of demand for products distinguished by place of produc-tion IMF Staff Papers 16 159ndash178

Baldwin R E (ed) (2016) Brexit Beckons Thinking Ahead by Leading Economists LondonCentre for Economic Policy Research

Campos N F (2016) Lousy experts Looking back at the ex ante estimates of the costs ofBrexit In R E Baldwin (ed) Brexit Beckons Thinking Ahead by Leading Economists(ch 3) London Centre for Economic Policy Research

Dixon P Parmenter B Sutton J and Vincent D (1982) ORANI A Multisectoral Modelof the Australian Economy Contributions to Economic Analysis 142 Amsterdam North-Holland

European Commission (2015) Trade for All Towards a More Responsible Trade andInvestment Policy Brussels European Commission

Harrison J Horridge M Jerie M and Pearson K (2014) GEMPACK ManualMelbourne GEMPACK Software

HM Treasury (2017) Spring Budget 2017 London HM TreasuryHolmes A J and Anderson K (2017a)Annual Database of National Beverage Consumption

Volumes and Expenditures 1950 to 2015 Wine Economics Research Centre University ofAdelaide posted at wwwadelaideeduauwine-econdatabases

Holmes A J and Anderson K (2017b) Convergence in National Alcohol ConsumptionPatterns New Global Indicators Journal of Wine Economics 12(2) 117ndash148

Ludington C C (2013) The Politics of Wine in Britain A New Cultural History BasingstokeUK Palgrave Macmillan

Office of National Statistics (2017) Analysis of real earnings London Office of NationalStatistics Available at httpswwwonsgovukemploymentandlabourmarketpeopleinworkearningsandworkinghoursarticlessupplementaryanalysisofaverageweeklyearningslatest

Rollo J Borchert I Dawar K Holmes P and Winters L A (2016) The World TradeOrganisation A safety net for a post-Brexit UK trade policy Briefing Paper 1UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-1-final-1pdf

Smith A (2017) Brexit Hard truths and hard choices UKTPO blog 19 June Available athttpsblogssussexacukuktpo20170619brexit-hard-truths-and-hard-choices

Swinbank A (2017) World trade rules and the policy options for British agriculture post-Brexit Briefing Paper 7 UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-7pdf

Viner J (1950) The Customs Union Issue New York Carnegie Endowment for InternationalPeace

Wine Australia (2015) Export Market Guide European Union Adelaide Wine AustraliaWittwer G Berger N and Anderson K (2003) A model of the worldrsquos wine markets

Economic Modelling 20(3) 487ndash506

Kym Anderson and Glyn Wittwer 251

Page 16: U.K. and Global Wine Markets by 2025, and Implications of ... · U.K. and Global Wine Markets by 2025, and Implications of Brexit* Kym Andersona and Glyn Wittwerb Abstract The United

Figure 7

National Shares of Global Wine Import Value and Volume 2014 and Projected Baselinefor 2025 ()

Fig7-BW

onlin

eBW

inprint

Source Authorsrsquo model results

236 UK and Global Wine Markets by 2025 and Implications of Brexit

A Initial Impact of Brexit

Because the British pound dropped by one-sixth against the US dollar in the fourmonths following the Brexit vote in June 2016 and because the UnitedKingdomrsquos average real wage fell 1 between November 2016 and April 2017(according to the latest available data see Office of National Statistics 2017) weviewed our ldquolargerdquo scenario as more likely soon after the Brexit vote But wecompare those results with ones from our ldquosmallrdquo initial-impact scenario becauseit has become more likely following the June 2017 general election that ldquosofterrdquooptions will be considered by the new government Generally the results areabout half the size in the ldquosmallrdquo scenario with the exception of the bilateraltrade effects To show the sensitivity of results to our assumptions we point outthe differences when the ldquosmallrdquo results are not close to half the results shown forthe ldquolargerdquo scenario

Before turning to those results we first present a ldquofree-traderdquo scenario whichassumes the United Kingdom chooses the radical option of becoming the HongKong of Europe Unlikely as this scenario is the results provide assurance thatour global wine model generates the standard types of effects of exiting a customsunion In this ldquofree-traderdquo scenario the United Kingdom is assumed to move tozero tariffs on all wine imports Thus the current preferences on EU Chileanand South African wine imports disappear as all other wine exporters enjoy thesame free access to UK wine markets We assume in this scenario that Brexit hasno impact on the United Kingdomrsquos real income or consumption growth rates by2025 and that by then the British pound has returned to its 2014 value relative toother currencies (similar to the real exchange rate assumed in our baseline scenarioto 2025 see Appendix Table A1)7 Under these assumptions EU Chilean andSouth African wine exports to the United Kingdom in 2025 are 003 lower thanin the no-Brexit baseline and UK imports from other wine exporters are 025higher but the United Kingdomrsquos total wine imports are hardly any different(because local wine prices fall by only 05 and incomes are unchanged) In this sce-nario UK consumers expand their volume of consumption by just 05 while theEU27 exports more and other countries export less to the rest of the world such thatworld wine exports are almost unchanged In short all these effects are in the direc-tion that comparative static-customs-union theory would predict but they are smallbecause we assume the United Kingdomrsquos real income and currency are the same asin our baseline for 2025 as described in the previous section

In the ldquolargerdquo alternate initial scenario involving Brexit as compared with thebaseline scenario to 2025 the consumer price of wine in 2025 is 22 higher in the

7If this scenario were seen as a credible long-run trade policy it would encourage more investment thatwould eventually raise UK incomes but not without major structural changes that would initiallydisrupt the economy Implicitly we assume in our free-trade scenario that these two opposite impactson real UK incomes and the British pound exactly offset each other by 2025

Kym Anderson and Glyn Wittwer 237

United Kingdom in local currency terms (20 because of real depreciation of theBritish pound 4 because of new tariffs on EU Chilean and South Africanwines and ndash2 because of slower UK income growth) The volume of UKwine consumption is 28 lower 16 because of slower UK economic growth7 because of real depreciation of the British pound and 5 because of newtariffs Superpremium still-wine sales are the most affected dropping by two-fifths while sparkling and commercial-premium wines drop a bit less than one-quarter Because the average price rises by more than the decline in volume soldthe aggregate value of UK sales even in local currency terms falls under thisldquolargerdquo Brexit scenario Under the ldquosmallrdquo Brexit scenario the consumer price ofwine in 2025 is 11 higher in the United Kingdom and its volume of wine consump-tion is 17 lower

The volume of projected UK imports in 2025 is 427 ML or nearly one-quarterless in the ldquolargerdquo scenario than in the baseline scenario comprising 58ML less spar-kling wine 31 ML less superpremium still wine and 339 ML less commercial-premium wine World imports are lower by just 239 ML because imports byother countries are 189 ML higher in response to the lower international wineprices in this scenario In value terms UK imports are $175 billion (or 27)lower in 2025 because of ldquolargerdquo Brexit $113 billion because of lower incomes$038 billion because of the fall in the British pound and $014 billion because ofthe rise in wine-import tariffs (Table 4) These aggregate trade impacts are a littlemore than half this size under the ldquosmallrdquo Brexit scenario

Despite the import levels falling because of raised import tariffs domestic con-sumption of all three quality categories of UK-produced wine is lower with thanwithout Brexit because of shrunken demand for all wines resulting from loweredUK incomes and raised local prices because of devaluation of the British poundThe British poundrsquos devaluation does make it easier for the United Kingdom tosell wines abroad though The countryrsquos exports are 7 ML or nearly 5 higherin 2025 in the ldquolargerdquo Brexit scenario and UK production is 3 higher ThoseUK exports (or re-exports of imported bulk wine after it is bottled in the UnitedKingdom) that go to EU27 countries are reduced though because of the tariffnow imposed at the new EU border

Without Brexit the United Kingdomrsquos shares of global wine imports are slightlyhigher in volume terms in 2025 than in 2010ndash2015 but 2 percentage points lower invalue terms thanks to East Asiarsquos expanding demand for imports of premium winesWith ldquolargerdquo Brexit however that value share is an additional 2 percentage pointslower and the volume share is almost 5 points lower (Figure 8) The net effects ofthese impacts on global trade are shown in Figure 9 Most of the initial tradeeffect of Brexit is a large decline in net imports of wine by the United Kingdomwith very little offsetting positive effect on trade in the rest of the world Theldquosmallrdquo Brexit numbers are a bit more than half of those for ldquolargerdquo Brexit

238 UK and Global Wine Markets by 2025 and Implications of Brexit

Table4

Difference

in2025

Projected

VolumeandValue

ofWineIm

portsb

ytheUnitedKingdom

andtheResto

fthe

World

asaConsequence

oftheInitial

BrexitS

hock

(MLand2014US$

Million

ldquoLargerdquoSc

enario)

Volum

e(M

L)

Value

(US$

million)

NP+CPa

SuperPrb

Sparkling

TOTAL

NP+CPa

SuperPrb

Sparkling

TOTAL

ΔUK

impo

rtsdu

eto

Low

erincomes

minus198

minus20

minus29

minus247

58minus644

minus253

minus234

minus1131

65Low

erpo

und

minus70

minus10

minus14

minus93

22minus248

minus127

minus102

minus476

27Highertariffs

minus71

minus1minus16

minus87

20minus110

minus8minus2

4minus143

8TOTA

Lminus339

minus31

minus58

minus427

100

minus1001

minus388

minus360

minus1750

100

difffrom

base

2332

3325

2432

3227

of

totalc

uts

797

14100

5722

21100

ΔROW

netim

ports

143

2125

189

230

143

minus181

192

ΔWORLD

TRADE

minus195

minus10

minus34

minus239

minus763

minus246

minus543

minus1552

aNon

prem

ium

plus

commercial-premium

still

wines

bSu

perpremium

still

wines

SourceA

utho

rsrsquomod

elresults

Kym Anderson and Glyn Wittwer 239

The aggregate effect of ldquolargerdquo Brexit on the market shares of various wine-export-ing countries in the United Kingdom is almost indiscernible even with one decimalpoint The projected 2025 shares are quite different from the actual 2014 shares forseveral countries They are much smaller in 2025 for South Africa Australia andNew Zealand (and the United States in volume terms) and are much larger involume for Spain and in value for Italy These results occur because wine-exportingcountries benefit differentially from the varying rates of growth in net importdemand for wine in non-UK countries over this projection period The most impor-tant projected changes are the increase in the real value of annual wine importsbetween 2014 and 2025 by China (200 or $3 billion) Other Asia (110 or $22billion) and Africa (270 or $16 billion) More than half of Australiarsquos increasein annual exports from 2014 to 2025 go to Asia and more than half of SouthAfricarsquos increase in exports go to Other Africa

Table 5 reveals that ldquolargerdquo Brexit lowers EU Chilean and South African wineexports by 150 ML or US$12 billion in the case of the European Union with someof their exports diverted from the United Kingdom to the EU27 and other marketsin competition with New World exporters Although the United States Australiaand Argentina sell only a little less to the United Kingdom they also sell less toother countries For Chile and South Africa which lose their preferential access toUK (but not to EU27) markets in this Brexit scenario some exports are redirected

Figure 8

UK Shares of World Wine Imports 2010ndash2015 and Projected to 2025 without and with theldquoLargerdquo Initial Brexit Shock ()

Fig8-BW

onlin

eBW

inprint

Sources Anderson and Pinilla (2017) and authorsrsquo model results

240 UK and Global Wine Markets by 2025 and Implications of Brexit

from the United Kingdom to EU27 countries ndash but again they export less overallGlobal wine trade in 2025 is 240 ML (19) or $18 billion (35) less under thisldquolargerdquo Brexit scenario The percentage by which wine exportersrsquo trade shrinks isgreater for values than for volumes because of changes in relative prices of differ-ent-quality wines Those differences are shown in the numbers in parentheses inTable 5

Three other points are worth making about Table 5 One is that Australia sellsslightly more to the United Kingdom in the ldquosmallrdquo Brexit scenario rather thanslightly less as in the ldquolargerdquo Brexit case Evidently the negative income and price(devaluation) effects do not more than offset the positive trade-diverting effect onAustralian exports to the United Kingdom of removing preferences in the ldquosmallrdquoscenario Second New Zealand sells slightly more to non-UK countries underBrexit despite greater competition from the EU27 Chile and South Africa Thisanomaly is due to changes in the relative prices of different qualities of wine inglobal wine markets bearing in mind that New Zealand has the worldrsquos highestaverage price for still-wine exports And third the value (but not the volume) ofexports of ldquoOtherrdquo countries to markets other than the United Kingdom arehigher under Brexit This result too is due to changes in the relative prices of differ-ent qualities of wine in global wine markets

Figure 9

Difference in 2025 Wine Import Volumes and Values as a Result of the ldquoLargerdquo Initial BrexitShock (ML and US$ Million in 2014 US dollars)

Fig9-BW

onlin

eBW

inprint

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 241

B Subsequent Impact of Brexit from a UK-EU27 FTA

The next-most-likely step in the Brexit process is for the United Kingdom to nego-tiate a new trade arrangement with the EU27 We therefore assume that a UK-EU27FTAwith free bilateral wine trade is implemented and adjusted to by 2025 and thatprogress toward that end occurs soon enough that the adverse macroeconomicshocks from the initial impact of uncertainty over the Brexit process are confinedto those assumed in the ldquosmallrdquo scenario outlined above In this subsequent scenariowe assume that the British pound returns to the value it would have reached in 2025in the absence of Brexit (ie reversing the 10 devaluation assumed in the ldquosmallrdquoinitial Brexit scenario) and that real incomes in the United Kingdom are 8 higher

Table 5Difference in 2025 BilateralWine Import Volumes and Values fromKey Exporters by the United

Kingdom and the Rest of the World (RoW) as a Result of Initial Brexit Shock(ML and 2014US$ Million)a

(a) ldquolargerdquo scenario

Volume (ML) Value (2014US$ million)

UK RoW WORLD () UK RoW WORLD ()

EU27 minus287 136 minus150 (ndash17) minus1187 minus5 minus1192 (ndash31)Chile minus59 35 minus25 (ndash30) minus169 31 minus138 (ndash48)South Africa minus53 35 minus18 (ndash32) minus105 20 minus85 (ndash67)USA minus7 minus6 minus13 (ndash24) minus75 minus40 minus115 (ndash50)Australia minus4 minus3 minus7 (ndash09) minus25 minus65 minus90 (ndash30)Argentina minus3 minus9 minus12 (ndash48) minus16 minus39 minus55 (ndash52)New Zealand minus11 9 minus2 (ndash09) minus162 71 minus91 (ndash43)Others minus2 minus10 minus12 (ndash02) minus11 minus52 minus63 (ndash44)WORLD minus427 187 minus240 (ndash19) minus1750 minus79 minus1829 (ndash35)

(a) ldquosmallrdquo scenario

Volume (ML) Value (2014US$ million)

UK RoW WORLD () UK RoW WORLD ()

EU27 minus178 82 minus96 (ndash12) minus692 minus43 minus736 (ndash19)Chile minus46 28 minus18 (ndash24) minus128 36 minus91 (ndash32)South Africa minus43 29 minus14 (ndash42) minus82 23 minus59 (ndash47)USA 1 minus6 minus5 (ndash11) minus23 minus28 minus51 (ndash22)Australia 5 minus10 minus5 (ndash06) 19 minus56 minus38 (ndash13)Argentina 0 minus6 minus6 (ndash26) minus3 minus25 minus29 (ndash27)New Zealand minus5 4 minus1 (ndash06) minus80 34 minus46 (ndash22)Others 0 minus9 minus9 (ndash01) minus1 minus33 minus34 (ndash24)WORLD minus266 112 minus154 (ndash13) minus991 minus92 minus1083 (ndash21)

a Numbers in parentheses are the percentage difference between the Brexit and baseline scenarios for 2025 projected wine-import volumes orvalues by source

Source Authorsrsquo model results

242 UK and Global Wine Markets by 2025 and Implications of Brexit

in 2025 than in the ldquosmallrdquo scenario ndash that is they regain two-thirds of the value theywould have reached without Brexit-related uncertainty

This subsequent development in the Brexit process reverses most of but not allthe initial effects of Brexit by 2025 because of our assumption that the lost growthin the initial years of uncertainty following the Brexit vote are only partly recoveredby 2025 following the implementation of a UK-EU27 FTA Moreover the longer ittakes before this FTA is finalized and implemented the longer the estimated initialadverse effects persist and the larger the cumulative cost of Brexit to UK wine con-sumers and to grape and wine producers in wine-exporting countries

Table 6 summarizes the subsequent trade effects for 2025 It suggests that all butone-tenth of the loss in value of world trade in wine from the initial ldquosmallrdquo impact isrestored most of it because of a smaller reduction in the United Kingdomrsquos wineimports following the FTA Most of that improved outcome is because of recoveredimports from the EU27 commensurate with the latterrsquos high share of UK importsEven though Chile and South Africa are assumed in this scenario to have not yetsigned an FTA with the United Kingdom they export slightly more to the UnitedKingdom (and even more to all other countries) than in the initial ldquosmallrdquo scenario

Within the United Kingdom this FTA brings down the local currency consumerprice of wine by 9 largely offsetting the 11 rise in the initial ldquosmallrdquo Brexit sce-nario and it raises the volume consumed in the United Kingdom by 18 fully off-setting the 17 fall in the initial ldquosmallrdquo Brexit scenario

Table 6Difference in 2025 Bilateral Wine-Import Volumes and Values from Key Exporters by the

United Kingdom and the Rest of the World (RoW) as a Result of Implementing a UK-EU27FTA Following Initial ldquoSmallrdquo Brexit Shock (ML and 2014US$ Million Difference Relative to

Initial ldquoSmallrdquo Brexit Scenario)a

Volume (ML) Value (2014US$ million)

UK RoW WORLD (ldquosmallrdquo)a UK RoW WORLD (ldquosmallrdquo)a

EU27 212 minus112 100 (minus96) 750 0 750 (minus806)Chile 1 3 4 (minus18) 3 40 43 (minus92)South Africa 3 2 5 (minus14) 3 18 21 (minus60)USA 3 3 6 (minus5) 26 21 47 (minus50)Australia 1 1 2 (minus5) 8 38 46 (minus33)Argentina 1 4 5 (minus6) 4 20 24 (minus28)New Zealand 4 minus3 1 (minus1) 61 minus25 36 (minus46)Others 1 6 7 (minus9) 5 minus46 minus41 (92)WORLD 226 minus96 130 (minus154) 860 66 926 (minus1022)

a Numbers in parentheses are the world trade differences between the ldquosmallrdquo initial Brexit scenario and the baseline scenario copied fromcolumns 3 and 7 of Table 5(b)

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 243

In short this subsequent step in the Brexit process can be expected to restore muchof the initial adverse effects in the United Kingdom of the Brexit vote by 2025 Thewine exporters in the EU27 Chile and South Africa also lose less in this scenarioas compared with the initial ldquosmallrdquo scenario while other key exporters sell nearlyas much in 2025 in this scenario as they are projected to in the baseline ndash that isthey are beneficiaries of the reduced discrimination in global wine markets in thiscase where Chile and South Africa no longer enjoy preferential access to the UKmarket (and even though EU27 producers are again allowed to do so) Keep inmind though that although the estimated losses in the initial scenarios are reportedjust for 2025 they are expected to be felt throughout the preceding years of uncertaintyTherefore the longer it is before uncertainty abates and the United Kingdomrsquos currentpreferential trading arrangements are replaced by newones such as a UK-EU27 FTAthe higher the cumulative cost of the Brexit vote to the wine trade

VII Caveats and Conclusions

The above Brexit simulations are just a few of many scenarios that could be modeledObvious additional ones could also assume that FTAs are reached between theUnited Kingdom and other trading partners including Chile and South Africaand such countries as Australia and New Zealand Some of these other countriesare already exploring the ECrsquos invitation (see European Commission 2015) to nego-tiate an FTA with what will be the EU27 The sequence in which FTAs are signedand the speed with which they are implemented will matter (as was also the casewith the sequential signing over the past decade of bilateral FTAs with NortheastAsian countries by Chile Australia and New Zealand see Anderson andWittwer 2015) Even if the United Kingdom were able to sign additional bilateralFTAs and begin implementing them before 2025 it would have little effect on theabove results (because wine tariffs are such a minor contributor to them) unlessthose FTAs were to accelerate the United Kingdomrsquos economic growth and therise in the British poundrsquos value before 2025

We assume above that no changes are made to alcohol excise duties in the UnitedKingdom following Brexit even though they are scheduled to be progressively raisedwith inflation and may be raised even more for wine relative to spirits to offset theopposite effects of Brexit on those two domestic industries Nor do we make anyallowance for UK increases in consumer wine prices that may be needed to coverthe higher cost of clearing customs on imports from EU27 countries (which maybe more or less offset by increased smuggling and duty-free purchasing)Importers of fine wines including individuals who buy Bordeaux en primeur maywell reduce their demand because of concerns that delays on docks will affect thequality of their wines or because of greater currency uncertainty

To summarize Brexit is costly initially to UK consumers of wine (and of manyother tradable products) because the domestic retail price in local currency tax-

244 UK and Global Wine Markets by 2025 and Implications of Brexit

inclusive terms is 22 higher than otherwise in 2025 and the volume of wine con-sumed domestically is 28 lower in the ldquolargerdquo Brexit scenario (or 11 higher pricesand 17 lower quantities in the ldquosmallrdquo case) by 2025 unless a UK-EU27 FTA isdeveloped before then Even if such an FTA does get signed ratified by all 28 par-liaments and implemented by 2025 the slower income growth in the interim meansa smaller UK wine market in 2025 than would otherwise have been the case Thevolume reduction is a blow to many participants in UK wine-bottling transportingstoring wholesaling and retailing businesses in addition to restaurants and pubsVery little of that initial impact occurs because of higher import tariffs most impor-tant is the assumed fall in UK real incomes

The small but growing number of local UK vignerons (who supply less than 05of all domestic-wine sales and less than 4 of sparkling-wine sales) is initially pro-jected to sell less wine domestically because of reduced demand and to export only alittle extra abroad while the British pound is devalued Overall their production isonly 3 higher in 2025 even in the ldquolargerdquo Brexit scenario (not taking intoaccount the fact that their casual labor is likely to be more expensive under Brexitbecause of tighter restrictions on immigration) and their production diminishes ifand when a UK-EU27 FTA is in place

Even with a UK-EU27 FTA in place EU27 wine exporters export US$56 millionless wine in 2025 thanks to Brexit Chile and South Africa export $88 million lesswine and wine exports from the rest of the world are little different in aggregatealthough slightly larger from Australia

Clearly there will be great uncertainly for some time over the possible policy out-comes to flow from Brexit and their consequent sequential impacts on UK house-hold disposable incomes foreign exchange rates and bilateral wine tariffsMeanwhile the above projections under explicit assumptions provide some idea ofhow wine markets might be affected by the most-likely first two stages of theBrexit process (agreeing on a new tariff schedule at the WTO and agreeing to andimplementing a UK-EU27 FTA) In particular they make clear that nontrivialinitial impacts could affect the domestic wine market impacts that are likely to belarger than just the direct impact of changes in bilateral tariffs If the UnitedKingdom succeeds in getting countries to agree to transition arrangements thatdelay the changes in tariffs until new FTAs are signed and ratified the initialeffects will be less dramatic than in our first scenarios but it remains to be seenwhether any such agreements can be reached In any event the net effect of Brexiton the welfare of the worldrsquos consumers and producers of wine as a whole will benegative not just initially but permanently unless new trade policy commitmentsby the United Kingdom with major wine-exporting countries are sufficiently moreliberal than current arrangements

Kym Anderson and Glyn Wittwer 245

Appendix Key Equations in the Global Wine Markets Model

Amodel of the worldrsquos wine markets was first published byWittwer et al (2003) andmuch improved by Anderson and Wittwer (2013) Several significant enhancementshave been made to that original model Wine types have been disaggregated intomore types ndash namely nonpremium (including generic bulk) commercial-premiumand superpremium still wines and sparkling wines8 As in the original modelthere are two types of grapes premium and nonpremium Nonpremium wines usenonpremium grapes exclusively superpremium wines use mainly premium grapesand commercial-premium and sparkling wines use both types of grapes In theregional dimension the number of countries and country groups has expandedfrom 10 in the original model to 51 44 individual nations and 7 compositeregions The modelrsquos database is calibrated to 2014 for this paper based on datain Anderson et al (2017) trade data for which are downloaded from httpscomtradeunorg The modelrsquos supply-and-demand equations are based on the orig-inal ORANI modelrsquos theory (see Dixon Parmenter Sutton and Vincent 1982) andthe model is implemented using GEMPACK software (Harrison Horridge Jerieand Pearson 2014)

An enhancement of importance to the present study is the inclusion of exchange-rate variables in the model which allow a distinction between price impacts asobserved in local currency units and those observed in 2014 US dollars

In the model the grape and wine sectors minimize costs of intermediate inputssubject to weak constant elasticity of substitution (CES) substitutability betweeninputs By assumption no intermediate inputs are imported from other countries9Hence

Xcid frac14 f ethX1id CESfrac12P

cid=P1id $THORN eth1THORN

P1id X1id frac14X

c

Xcid P

cid eth2THORN

whereXcid is the quantity demanded of commodity c by grape or wine industry i in

region d Pcid is the corresponding price and X1id and P1id are the respective inter-

mediate composite quantities and prices

Two primary factors are employed in the sector labor (the quantity of which isendogenous with perfectly elastic supply) and capital Capital is usually treated asexogenous in quantity with rates of return bearing all the adjustment in thevarious scenarios This reflects the fact that grapes (a perennial crop) and wine-

8Commercial-premium still wines are defined by Anderson and Nelgen (2011) to be those priced betweenUS$250 and $750 per litre pre-tax at a countryrsquos border or wholesale9An exception concerns bulk wine imports used mainly in the bottling of commercial premium wine in theUK USA Canada Japan and Germany

246 UK and Global Wine Markets by 2025 and Implications of Brexit

plant capacity adjust slowly to market signals

Lid frac14 f ethFid CESfrac12Wid=PFid $THORN eth3THORN

Kid frac14 f ethFid CESethRid=PFid $THORN eth4THORN

PFid Fid frac14 Lid Wid thorn Kid Rid eth5THORN

Grape and wine producers are assumed to minimize costs subject to CES substitu-tion between capital and labor Equations (3) to (5) show primary factor demandsfor the labor composite Lid and capital Kid subject to a composite factor demandFid by industry i in region d The factor prices are Wid for labor Rid for capitalrentals and PFid for composite factor prices

Appendix Table A1Cumulative Consumption and Population Growth Rates and Changes in the Real Exchange Rate

(RER) Relative to the US Dollar 2014 to 2025 without Brexit ()

Aggregateconsumption Poprsquon RER

Aggregateconsumption Poprsquon RER

France 18 4 minus11 Australia 35 11 minus17Italy 11 2 minus9 New Zealand 32 9 minus26Portugal 14 0 minus9 Canada 27 8 minus18Spain 26 8 minus9 United States 31 8 0Austria 19 4 minus7 Argentina 7 10 109Belgium 20 7 minus9 Brazil 16 8 minus29Denmark 22 2 minus9 Chile 55 8 minus2Finland 21 3 minus7 Mexico 42 12 minus8Germany 14 minus2 minus11 Uruguay 45 3 1Greece 22 minus1 minus14 Other Latin America 60 10 minus5Ireland 42 12 minus9 South Africa 36 12 minus1Netherlands 21 4 minus9 Turkey 50 8 20Sweden 24 9 minus13 North Africa 53 11 0Switzerland 18 8 minus6 Other Africa 109 18 84United Kingdom 32 6 1 Middle East 52 18 minus12Other West Europe 21 10 minus1 China 79 3 5Bulgaria 41 minus7 7 Hong Kong 42 3 2Croatia 20 minus2 minus1 India 134 13 17Georgia 35 0 23 Japan 11 minus3 minus24Hungary 25 minus3 minus11 Korea 38 1 minus9Moldova 49 minus11 13 Malaysia 62 15 minus16Romania 45 minus4 22 Philippines 75 18 7Russia 18 minus2 minus8 Singapore 44 21 minus22Ukraine 22 minus5 14 Taiwan 29 1 minus13Other East Europe 40 minus5 48 Thailand 47 3 minus9

Other Asia 99 10 10

Source Authorsrsquo compilation from projections by various international agencies and from global economy-wide modeling by Anderson andStrutt (2016)

Kym Anderson and Glyn Wittwer 247

The composite factor demand Fid is proportional to total output Qid subject to aprimary factor using technology Aid Hence

Fid frac14 Qid Aid eth6THORN

The perfectly competitive zero-pure-profit condition is that total revenue valued atthe output price P0s

i multiplied by Qid equals the total production cost

P0si Qid frac14

X

cPcid X1cid thorn

X

oWo

id Loid thorn Rid Kid eth7THORN

Household demands follow a linear expenditure system in each region We reducethe optimizing problem for household consumption of each commodity subjectto a budget constraint to equations describing subsistence and discretionarydemands Aggregate subsistence expenditure WSUBd depends only on consumerprices P3cd for each commodity and the number of households N as per-capita sub-sistence quantities XSUBcd subject to given preferences are constant

WSUBd frac14X

c

P3cd XSUBcd Nd eth8THORN

Discretionary expenditures for each commodity (the left-hand side of equation (9))are equal to the marginal budget share (βcd) of aggregate discretionary expenditureThis aggregate is the bracketed term on the right-hand side of equation (9) whereW3TOTd is aggregate nominal expenditure

P3cdethX3cd XSUBcd NdTHORN frac14 βcdethW3TOTd WSUBdTHORN eth9THORN

Because real aggregate consumption is usually exogenous in our partial equilibriumsimulations the linear expenditure system determines the consumption shares ofindividual final commodities (ie the five wine types plus a composite of all otherconsumption items) driven by changes in relative prices as faced by domestic con-sumers The income elasticity of demand for each commodity is equal to the mar-ginal budget share divided by the expenditure share This number varies from 05for nonpremium wine to 17 for superpremium still wine The income elasticity ofdemand for other consumption is very close to 10 because wine accounts for anaverage of only 03 of aggregate expenditures globally and no more than 11in any country (Anderson and Nelgen 2011 Table 166)

A feature of our revised model of world wine markets is the inclusion of nominalexchange rates These appear directly in the equation linking retail prices (P3scd ) toproducer prices by country of origin (P0s

c ) where c denotes the wine type

P3 scd frac14 P0s

cfd

fsTtarcd T

taxcd thorn Pm

cd eth10THORN

248 UK and Global Wine Markets by 2025 and Implications of Brexit

The exchange rates in the consuming and producing regions are fd and fs respec-tively expressed as local currency units per $US Ttar

cd is the power of the tariff in theconsuming region and Ttax

cd is the power of the domestic consumption (or excise) taxprior to any generic value-added or goods-and-services tax Pm

cd is the price of marginm assumed to be locally supplied nontradable and therefore unaffected by theexchange rate

A given level of consumption for wine type c (X3cd) is satisfied using theArmington (1969) assumption in which wines from different countries of originare imperfectly substitutable First domestic wine is imperfectly substitutable witha composite of imports

X3sscd frac14 f ethX3cd CESethP3sscd=P3cdTHORNTHORN ss frac14 domestic imports eth11THORN

and then imports by origin (X3scd ) are determined in a second CES equation

X3scd frac14 f ethX3ssfrac14importscd CESethP3scd=P3

ssfrac14importscd THORNTHORN eth12THORN

The model enables us to show how changes in international competitiveness affectthe worldrsquos wine markets A crucial part of this exercise is explaining how pricesdetermined outside the grape and wine markets influence these markets Becausethe model is in partial equilibrium to depict the impacts of changes in internationalcompetitiveness outside price changes need to be imposed as shocks on the modelThe price of intermediate inputs other than grapes and wine shown in equations (1)and (2) is set equal to the price of GDP (Pg

d ) multiplied by a shifter Fcd

Pcid frac14 Fc

dPgd eth13THORN

If no specific price observations are available the shifter Fcd remains exogenous and

unshocked with the change in price being determined by a shock to the price ofGDP If observations are available for specific input price movements the shifterFc

dbecomes endogenous with Pc

id now exogenous and shocked

Wid frac14 Fwd P

gd eth14THORN

Wage rates are treated similarly In equation (14) if the wage shifter Fwd is exogenous

changes in wage rates Wid are determined by changes in the price of GDP If wagerate data are available Fw

d becomes endogenous andwage rates are shocked directly

Pmcd frac14 Fm

d Pgd eth15THORN

The prices of trade and transport margins are also determined by the price of GDP ifthe shifter Fm

d in equation (15) is exogenous

Changes in international competitiveness depend on changes in relative pricelevels and changes in nominal exchange rates In equation (16) fR

s denotes real

Kym Anderson and Glyn Wittwer 249

exchange-rate movements relative to the US dollar

fRs frac14 Pg

s =frac12PgUSA

fs$ eth16THORN

In equation (16) the nominal exchange rate for the United States is alwaysunchanged because nominal and real exchange rates are expressed relative to theUS currency

Changes in international market conditions may have impacts in one direction onproducer prices as expressed in US dollars P0s

i and potentially in the opposite direc-tion in local currency units Hence we calculate real producer pricesP0s

iloc in localcurrency terms (ie the price most relevant to domestic producers)

P0siloc frac14 P0s

ifs=P

gs eth17THORN

To obtain real price changes in local currency terms we convert US dollar prices(P3 s

cd for the source-specific price and P3cd for the source-composite price) to reallocal currency prices (P3 s

cdloc and P3cdloc) using the consumer price index (CPI)(Pc

d ) as the deflator

P3 scdloc frac14 P3 s

cdfd=P

cd eth18THORN

and

P3cdloc frac14 P3cdfd=Pcd eth19THORN

References

Anderson K and Nelgen S (2011) Global Wine Markets 1961 to 2009 A StatisticalCompendium Adelaide University of Adelaide Press Also available as an e-book atwwwadelaideeduaupresstitlesglobal-wine

Anderson K Nelgen S and Pinilla V (2017) Global Wine Markets 1860 to 2016 AStatistical Compendium Adelaide University of Adelaide Press (forthcoming) Also tobe available as an e-book at wwwadelaideeduaupress

Anderson K and Pinilla V (with the assistance of A J Holmes) (2017)Annual Database ofGlobal Wine Markets 1835 to 2016 Wine Economics Research Centre University ofAdelaide to be posted in October at wwwadelaideeduauwine-econdatabases

Anderson K and Strutt A (2016) Impacts of Asiarsquos rise on African and Latin Americantrade Projections to 2030 World Economy 39(2) 172ndash194

Anderson K andWittwer G (2013) Modeling global wine markets to 2018 Exchange ratestaste changes and Chinarsquos import growth Journal of Wine Economics 8(2) 131ndash158

Anderson K and Wittwer G (2015) Asiarsquos evolving role in global wine markets ChinaEconomic Review 35 1ndash14

250 UK and Global Wine Markets by 2025 and Implications of Brexit

Armington P A (1969) A theory of demand for products distinguished by place of produc-tion IMF Staff Papers 16 159ndash178

Baldwin R E (ed) (2016) Brexit Beckons Thinking Ahead by Leading Economists LondonCentre for Economic Policy Research

Campos N F (2016) Lousy experts Looking back at the ex ante estimates of the costs ofBrexit In R E Baldwin (ed) Brexit Beckons Thinking Ahead by Leading Economists(ch 3) London Centre for Economic Policy Research

Dixon P Parmenter B Sutton J and Vincent D (1982) ORANI A Multisectoral Modelof the Australian Economy Contributions to Economic Analysis 142 Amsterdam North-Holland

European Commission (2015) Trade for All Towards a More Responsible Trade andInvestment Policy Brussels European Commission

Harrison J Horridge M Jerie M and Pearson K (2014) GEMPACK ManualMelbourne GEMPACK Software

HM Treasury (2017) Spring Budget 2017 London HM TreasuryHolmes A J and Anderson K (2017a)Annual Database of National Beverage Consumption

Volumes and Expenditures 1950 to 2015 Wine Economics Research Centre University ofAdelaide posted at wwwadelaideeduauwine-econdatabases

Holmes A J and Anderson K (2017b) Convergence in National Alcohol ConsumptionPatterns New Global Indicators Journal of Wine Economics 12(2) 117ndash148

Ludington C C (2013) The Politics of Wine in Britain A New Cultural History BasingstokeUK Palgrave Macmillan

Office of National Statistics (2017) Analysis of real earnings London Office of NationalStatistics Available at httpswwwonsgovukemploymentandlabourmarketpeopleinworkearningsandworkinghoursarticlessupplementaryanalysisofaverageweeklyearningslatest

Rollo J Borchert I Dawar K Holmes P and Winters L A (2016) The World TradeOrganisation A safety net for a post-Brexit UK trade policy Briefing Paper 1UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-1-final-1pdf

Smith A (2017) Brexit Hard truths and hard choices UKTPO blog 19 June Available athttpsblogssussexacukuktpo20170619brexit-hard-truths-and-hard-choices

Swinbank A (2017) World trade rules and the policy options for British agriculture post-Brexit Briefing Paper 7 UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-7pdf

Viner J (1950) The Customs Union Issue New York Carnegie Endowment for InternationalPeace

Wine Australia (2015) Export Market Guide European Union Adelaide Wine AustraliaWittwer G Berger N and Anderson K (2003) A model of the worldrsquos wine markets

Economic Modelling 20(3) 487ndash506

Kym Anderson and Glyn Wittwer 251

Page 17: U.K. and Global Wine Markets by 2025, and Implications of ... · U.K. and Global Wine Markets by 2025, and Implications of Brexit* Kym Andersona and Glyn Wittwerb Abstract The United

A Initial Impact of Brexit

Because the British pound dropped by one-sixth against the US dollar in the fourmonths following the Brexit vote in June 2016 and because the UnitedKingdomrsquos average real wage fell 1 between November 2016 and April 2017(according to the latest available data see Office of National Statistics 2017) weviewed our ldquolargerdquo scenario as more likely soon after the Brexit vote But wecompare those results with ones from our ldquosmallrdquo initial-impact scenario becauseit has become more likely following the June 2017 general election that ldquosofterrdquooptions will be considered by the new government Generally the results areabout half the size in the ldquosmallrdquo scenario with the exception of the bilateraltrade effects To show the sensitivity of results to our assumptions we point outthe differences when the ldquosmallrdquo results are not close to half the results shown forthe ldquolargerdquo scenario

Before turning to those results we first present a ldquofree-traderdquo scenario whichassumes the United Kingdom chooses the radical option of becoming the HongKong of Europe Unlikely as this scenario is the results provide assurance thatour global wine model generates the standard types of effects of exiting a customsunion In this ldquofree-traderdquo scenario the United Kingdom is assumed to move tozero tariffs on all wine imports Thus the current preferences on EU Chileanand South African wine imports disappear as all other wine exporters enjoy thesame free access to UK wine markets We assume in this scenario that Brexit hasno impact on the United Kingdomrsquos real income or consumption growth rates by2025 and that by then the British pound has returned to its 2014 value relative toother currencies (similar to the real exchange rate assumed in our baseline scenarioto 2025 see Appendix Table A1)7 Under these assumptions EU Chilean andSouth African wine exports to the United Kingdom in 2025 are 003 lower thanin the no-Brexit baseline and UK imports from other wine exporters are 025higher but the United Kingdomrsquos total wine imports are hardly any different(because local wine prices fall by only 05 and incomes are unchanged) In this sce-nario UK consumers expand their volume of consumption by just 05 while theEU27 exports more and other countries export less to the rest of the world such thatworld wine exports are almost unchanged In short all these effects are in the direc-tion that comparative static-customs-union theory would predict but they are smallbecause we assume the United Kingdomrsquos real income and currency are the same asin our baseline for 2025 as described in the previous section

In the ldquolargerdquo alternate initial scenario involving Brexit as compared with thebaseline scenario to 2025 the consumer price of wine in 2025 is 22 higher in the

7If this scenario were seen as a credible long-run trade policy it would encourage more investment thatwould eventually raise UK incomes but not without major structural changes that would initiallydisrupt the economy Implicitly we assume in our free-trade scenario that these two opposite impactson real UK incomes and the British pound exactly offset each other by 2025

Kym Anderson and Glyn Wittwer 237

United Kingdom in local currency terms (20 because of real depreciation of theBritish pound 4 because of new tariffs on EU Chilean and South Africanwines and ndash2 because of slower UK income growth) The volume of UKwine consumption is 28 lower 16 because of slower UK economic growth7 because of real depreciation of the British pound and 5 because of newtariffs Superpremium still-wine sales are the most affected dropping by two-fifths while sparkling and commercial-premium wines drop a bit less than one-quarter Because the average price rises by more than the decline in volume soldthe aggregate value of UK sales even in local currency terms falls under thisldquolargerdquo Brexit scenario Under the ldquosmallrdquo Brexit scenario the consumer price ofwine in 2025 is 11 higher in the United Kingdom and its volume of wine consump-tion is 17 lower

The volume of projected UK imports in 2025 is 427 ML or nearly one-quarterless in the ldquolargerdquo scenario than in the baseline scenario comprising 58ML less spar-kling wine 31 ML less superpremium still wine and 339 ML less commercial-premium wine World imports are lower by just 239 ML because imports byother countries are 189 ML higher in response to the lower international wineprices in this scenario In value terms UK imports are $175 billion (or 27)lower in 2025 because of ldquolargerdquo Brexit $113 billion because of lower incomes$038 billion because of the fall in the British pound and $014 billion because ofthe rise in wine-import tariffs (Table 4) These aggregate trade impacts are a littlemore than half this size under the ldquosmallrdquo Brexit scenario

Despite the import levels falling because of raised import tariffs domestic con-sumption of all three quality categories of UK-produced wine is lower with thanwithout Brexit because of shrunken demand for all wines resulting from loweredUK incomes and raised local prices because of devaluation of the British poundThe British poundrsquos devaluation does make it easier for the United Kingdom tosell wines abroad though The countryrsquos exports are 7 ML or nearly 5 higherin 2025 in the ldquolargerdquo Brexit scenario and UK production is 3 higher ThoseUK exports (or re-exports of imported bulk wine after it is bottled in the UnitedKingdom) that go to EU27 countries are reduced though because of the tariffnow imposed at the new EU border

Without Brexit the United Kingdomrsquos shares of global wine imports are slightlyhigher in volume terms in 2025 than in 2010ndash2015 but 2 percentage points lower invalue terms thanks to East Asiarsquos expanding demand for imports of premium winesWith ldquolargerdquo Brexit however that value share is an additional 2 percentage pointslower and the volume share is almost 5 points lower (Figure 8) The net effects ofthese impacts on global trade are shown in Figure 9 Most of the initial tradeeffect of Brexit is a large decline in net imports of wine by the United Kingdomwith very little offsetting positive effect on trade in the rest of the world Theldquosmallrdquo Brexit numbers are a bit more than half of those for ldquolargerdquo Brexit

238 UK and Global Wine Markets by 2025 and Implications of Brexit

Table4

Difference

in2025

Projected

VolumeandValue

ofWineIm

portsb

ytheUnitedKingdom

andtheResto

fthe

World

asaConsequence

oftheInitial

BrexitS

hock

(MLand2014US$

Million

ldquoLargerdquoSc

enario)

Volum

e(M

L)

Value

(US$

million)

NP+CPa

SuperPrb

Sparkling

TOTAL

NP+CPa

SuperPrb

Sparkling

TOTAL

ΔUK

impo

rtsdu

eto

Low

erincomes

minus198

minus20

minus29

minus247

58minus644

minus253

minus234

minus1131

65Low

erpo

und

minus70

minus10

minus14

minus93

22minus248

minus127

minus102

minus476

27Highertariffs

minus71

minus1minus16

minus87

20minus110

minus8minus2

4minus143

8TOTA

Lminus339

minus31

minus58

minus427

100

minus1001

minus388

minus360

minus1750

100

difffrom

base

2332

3325

2432

3227

of

totalc

uts

797

14100

5722

21100

ΔROW

netim

ports

143

2125

189

230

143

minus181

192

ΔWORLD

TRADE

minus195

minus10

minus34

minus239

minus763

minus246

minus543

minus1552

aNon

prem

ium

plus

commercial-premium

still

wines

bSu

perpremium

still

wines

SourceA

utho

rsrsquomod

elresults

Kym Anderson and Glyn Wittwer 239

The aggregate effect of ldquolargerdquo Brexit on the market shares of various wine-export-ing countries in the United Kingdom is almost indiscernible even with one decimalpoint The projected 2025 shares are quite different from the actual 2014 shares forseveral countries They are much smaller in 2025 for South Africa Australia andNew Zealand (and the United States in volume terms) and are much larger involume for Spain and in value for Italy These results occur because wine-exportingcountries benefit differentially from the varying rates of growth in net importdemand for wine in non-UK countries over this projection period The most impor-tant projected changes are the increase in the real value of annual wine importsbetween 2014 and 2025 by China (200 or $3 billion) Other Asia (110 or $22billion) and Africa (270 or $16 billion) More than half of Australiarsquos increasein annual exports from 2014 to 2025 go to Asia and more than half of SouthAfricarsquos increase in exports go to Other Africa

Table 5 reveals that ldquolargerdquo Brexit lowers EU Chilean and South African wineexports by 150 ML or US$12 billion in the case of the European Union with someof their exports diverted from the United Kingdom to the EU27 and other marketsin competition with New World exporters Although the United States Australiaand Argentina sell only a little less to the United Kingdom they also sell less toother countries For Chile and South Africa which lose their preferential access toUK (but not to EU27) markets in this Brexit scenario some exports are redirected

Figure 8

UK Shares of World Wine Imports 2010ndash2015 and Projected to 2025 without and with theldquoLargerdquo Initial Brexit Shock ()

Fig8-BW

onlin

eBW

inprint

Sources Anderson and Pinilla (2017) and authorsrsquo model results

240 UK and Global Wine Markets by 2025 and Implications of Brexit

from the United Kingdom to EU27 countries ndash but again they export less overallGlobal wine trade in 2025 is 240 ML (19) or $18 billion (35) less under thisldquolargerdquo Brexit scenario The percentage by which wine exportersrsquo trade shrinks isgreater for values than for volumes because of changes in relative prices of differ-ent-quality wines Those differences are shown in the numbers in parentheses inTable 5

Three other points are worth making about Table 5 One is that Australia sellsslightly more to the United Kingdom in the ldquosmallrdquo Brexit scenario rather thanslightly less as in the ldquolargerdquo Brexit case Evidently the negative income and price(devaluation) effects do not more than offset the positive trade-diverting effect onAustralian exports to the United Kingdom of removing preferences in the ldquosmallrdquoscenario Second New Zealand sells slightly more to non-UK countries underBrexit despite greater competition from the EU27 Chile and South Africa Thisanomaly is due to changes in the relative prices of different qualities of wine inglobal wine markets bearing in mind that New Zealand has the worldrsquos highestaverage price for still-wine exports And third the value (but not the volume) ofexports of ldquoOtherrdquo countries to markets other than the United Kingdom arehigher under Brexit This result too is due to changes in the relative prices of differ-ent qualities of wine in global wine markets

Figure 9

Difference in 2025 Wine Import Volumes and Values as a Result of the ldquoLargerdquo Initial BrexitShock (ML and US$ Million in 2014 US dollars)

Fig9-BW

onlin

eBW

inprint

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 241

B Subsequent Impact of Brexit from a UK-EU27 FTA

The next-most-likely step in the Brexit process is for the United Kingdom to nego-tiate a new trade arrangement with the EU27 We therefore assume that a UK-EU27FTAwith free bilateral wine trade is implemented and adjusted to by 2025 and thatprogress toward that end occurs soon enough that the adverse macroeconomicshocks from the initial impact of uncertainty over the Brexit process are confinedto those assumed in the ldquosmallrdquo scenario outlined above In this subsequent scenariowe assume that the British pound returns to the value it would have reached in 2025in the absence of Brexit (ie reversing the 10 devaluation assumed in the ldquosmallrdquoinitial Brexit scenario) and that real incomes in the United Kingdom are 8 higher

Table 5Difference in 2025 BilateralWine Import Volumes and Values fromKey Exporters by the United

Kingdom and the Rest of the World (RoW) as a Result of Initial Brexit Shock(ML and 2014US$ Million)a

(a) ldquolargerdquo scenario

Volume (ML) Value (2014US$ million)

UK RoW WORLD () UK RoW WORLD ()

EU27 minus287 136 minus150 (ndash17) minus1187 minus5 minus1192 (ndash31)Chile minus59 35 minus25 (ndash30) minus169 31 minus138 (ndash48)South Africa minus53 35 minus18 (ndash32) minus105 20 minus85 (ndash67)USA minus7 minus6 minus13 (ndash24) minus75 minus40 minus115 (ndash50)Australia minus4 minus3 minus7 (ndash09) minus25 minus65 minus90 (ndash30)Argentina minus3 minus9 minus12 (ndash48) minus16 minus39 minus55 (ndash52)New Zealand minus11 9 minus2 (ndash09) minus162 71 minus91 (ndash43)Others minus2 minus10 minus12 (ndash02) minus11 minus52 minus63 (ndash44)WORLD minus427 187 minus240 (ndash19) minus1750 minus79 minus1829 (ndash35)

(a) ldquosmallrdquo scenario

Volume (ML) Value (2014US$ million)

UK RoW WORLD () UK RoW WORLD ()

EU27 minus178 82 minus96 (ndash12) minus692 minus43 minus736 (ndash19)Chile minus46 28 minus18 (ndash24) minus128 36 minus91 (ndash32)South Africa minus43 29 minus14 (ndash42) minus82 23 minus59 (ndash47)USA 1 minus6 minus5 (ndash11) minus23 minus28 minus51 (ndash22)Australia 5 minus10 minus5 (ndash06) 19 minus56 minus38 (ndash13)Argentina 0 minus6 minus6 (ndash26) minus3 minus25 minus29 (ndash27)New Zealand minus5 4 minus1 (ndash06) minus80 34 minus46 (ndash22)Others 0 minus9 minus9 (ndash01) minus1 minus33 minus34 (ndash24)WORLD minus266 112 minus154 (ndash13) minus991 minus92 minus1083 (ndash21)

a Numbers in parentheses are the percentage difference between the Brexit and baseline scenarios for 2025 projected wine-import volumes orvalues by source

Source Authorsrsquo model results

242 UK and Global Wine Markets by 2025 and Implications of Brexit

in 2025 than in the ldquosmallrdquo scenario ndash that is they regain two-thirds of the value theywould have reached without Brexit-related uncertainty

This subsequent development in the Brexit process reverses most of but not allthe initial effects of Brexit by 2025 because of our assumption that the lost growthin the initial years of uncertainty following the Brexit vote are only partly recoveredby 2025 following the implementation of a UK-EU27 FTA Moreover the longer ittakes before this FTA is finalized and implemented the longer the estimated initialadverse effects persist and the larger the cumulative cost of Brexit to UK wine con-sumers and to grape and wine producers in wine-exporting countries

Table 6 summarizes the subsequent trade effects for 2025 It suggests that all butone-tenth of the loss in value of world trade in wine from the initial ldquosmallrdquo impact isrestored most of it because of a smaller reduction in the United Kingdomrsquos wineimports following the FTA Most of that improved outcome is because of recoveredimports from the EU27 commensurate with the latterrsquos high share of UK importsEven though Chile and South Africa are assumed in this scenario to have not yetsigned an FTA with the United Kingdom they export slightly more to the UnitedKingdom (and even more to all other countries) than in the initial ldquosmallrdquo scenario

Within the United Kingdom this FTA brings down the local currency consumerprice of wine by 9 largely offsetting the 11 rise in the initial ldquosmallrdquo Brexit sce-nario and it raises the volume consumed in the United Kingdom by 18 fully off-setting the 17 fall in the initial ldquosmallrdquo Brexit scenario

Table 6Difference in 2025 Bilateral Wine-Import Volumes and Values from Key Exporters by the

United Kingdom and the Rest of the World (RoW) as a Result of Implementing a UK-EU27FTA Following Initial ldquoSmallrdquo Brexit Shock (ML and 2014US$ Million Difference Relative to

Initial ldquoSmallrdquo Brexit Scenario)a

Volume (ML) Value (2014US$ million)

UK RoW WORLD (ldquosmallrdquo)a UK RoW WORLD (ldquosmallrdquo)a

EU27 212 minus112 100 (minus96) 750 0 750 (minus806)Chile 1 3 4 (minus18) 3 40 43 (minus92)South Africa 3 2 5 (minus14) 3 18 21 (minus60)USA 3 3 6 (minus5) 26 21 47 (minus50)Australia 1 1 2 (minus5) 8 38 46 (minus33)Argentina 1 4 5 (minus6) 4 20 24 (minus28)New Zealand 4 minus3 1 (minus1) 61 minus25 36 (minus46)Others 1 6 7 (minus9) 5 minus46 minus41 (92)WORLD 226 minus96 130 (minus154) 860 66 926 (minus1022)

a Numbers in parentheses are the world trade differences between the ldquosmallrdquo initial Brexit scenario and the baseline scenario copied fromcolumns 3 and 7 of Table 5(b)

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 243

In short this subsequent step in the Brexit process can be expected to restore muchof the initial adverse effects in the United Kingdom of the Brexit vote by 2025 Thewine exporters in the EU27 Chile and South Africa also lose less in this scenarioas compared with the initial ldquosmallrdquo scenario while other key exporters sell nearlyas much in 2025 in this scenario as they are projected to in the baseline ndash that isthey are beneficiaries of the reduced discrimination in global wine markets in thiscase where Chile and South Africa no longer enjoy preferential access to the UKmarket (and even though EU27 producers are again allowed to do so) Keep inmind though that although the estimated losses in the initial scenarios are reportedjust for 2025 they are expected to be felt throughout the preceding years of uncertaintyTherefore the longer it is before uncertainty abates and the United Kingdomrsquos currentpreferential trading arrangements are replaced by newones such as a UK-EU27 FTAthe higher the cumulative cost of the Brexit vote to the wine trade

VII Caveats and Conclusions

The above Brexit simulations are just a few of many scenarios that could be modeledObvious additional ones could also assume that FTAs are reached between theUnited Kingdom and other trading partners including Chile and South Africaand such countries as Australia and New Zealand Some of these other countriesare already exploring the ECrsquos invitation (see European Commission 2015) to nego-tiate an FTA with what will be the EU27 The sequence in which FTAs are signedand the speed with which they are implemented will matter (as was also the casewith the sequential signing over the past decade of bilateral FTAs with NortheastAsian countries by Chile Australia and New Zealand see Anderson andWittwer 2015) Even if the United Kingdom were able to sign additional bilateralFTAs and begin implementing them before 2025 it would have little effect on theabove results (because wine tariffs are such a minor contributor to them) unlessthose FTAs were to accelerate the United Kingdomrsquos economic growth and therise in the British poundrsquos value before 2025

We assume above that no changes are made to alcohol excise duties in the UnitedKingdom following Brexit even though they are scheduled to be progressively raisedwith inflation and may be raised even more for wine relative to spirits to offset theopposite effects of Brexit on those two domestic industries Nor do we make anyallowance for UK increases in consumer wine prices that may be needed to coverthe higher cost of clearing customs on imports from EU27 countries (which maybe more or less offset by increased smuggling and duty-free purchasing)Importers of fine wines including individuals who buy Bordeaux en primeur maywell reduce their demand because of concerns that delays on docks will affect thequality of their wines or because of greater currency uncertainty

To summarize Brexit is costly initially to UK consumers of wine (and of manyother tradable products) because the domestic retail price in local currency tax-

244 UK and Global Wine Markets by 2025 and Implications of Brexit

inclusive terms is 22 higher than otherwise in 2025 and the volume of wine con-sumed domestically is 28 lower in the ldquolargerdquo Brexit scenario (or 11 higher pricesand 17 lower quantities in the ldquosmallrdquo case) by 2025 unless a UK-EU27 FTA isdeveloped before then Even if such an FTA does get signed ratified by all 28 par-liaments and implemented by 2025 the slower income growth in the interim meansa smaller UK wine market in 2025 than would otherwise have been the case Thevolume reduction is a blow to many participants in UK wine-bottling transportingstoring wholesaling and retailing businesses in addition to restaurants and pubsVery little of that initial impact occurs because of higher import tariffs most impor-tant is the assumed fall in UK real incomes

The small but growing number of local UK vignerons (who supply less than 05of all domestic-wine sales and less than 4 of sparkling-wine sales) is initially pro-jected to sell less wine domestically because of reduced demand and to export only alittle extra abroad while the British pound is devalued Overall their production isonly 3 higher in 2025 even in the ldquolargerdquo Brexit scenario (not taking intoaccount the fact that their casual labor is likely to be more expensive under Brexitbecause of tighter restrictions on immigration) and their production diminishes ifand when a UK-EU27 FTA is in place

Even with a UK-EU27 FTA in place EU27 wine exporters export US$56 millionless wine in 2025 thanks to Brexit Chile and South Africa export $88 million lesswine and wine exports from the rest of the world are little different in aggregatealthough slightly larger from Australia

Clearly there will be great uncertainly for some time over the possible policy out-comes to flow from Brexit and their consequent sequential impacts on UK house-hold disposable incomes foreign exchange rates and bilateral wine tariffsMeanwhile the above projections under explicit assumptions provide some idea ofhow wine markets might be affected by the most-likely first two stages of theBrexit process (agreeing on a new tariff schedule at the WTO and agreeing to andimplementing a UK-EU27 FTA) In particular they make clear that nontrivialinitial impacts could affect the domestic wine market impacts that are likely to belarger than just the direct impact of changes in bilateral tariffs If the UnitedKingdom succeeds in getting countries to agree to transition arrangements thatdelay the changes in tariffs until new FTAs are signed and ratified the initialeffects will be less dramatic than in our first scenarios but it remains to be seenwhether any such agreements can be reached In any event the net effect of Brexiton the welfare of the worldrsquos consumers and producers of wine as a whole will benegative not just initially but permanently unless new trade policy commitmentsby the United Kingdom with major wine-exporting countries are sufficiently moreliberal than current arrangements

Kym Anderson and Glyn Wittwer 245

Appendix Key Equations in the Global Wine Markets Model

Amodel of the worldrsquos wine markets was first published byWittwer et al (2003) andmuch improved by Anderson and Wittwer (2013) Several significant enhancementshave been made to that original model Wine types have been disaggregated intomore types ndash namely nonpremium (including generic bulk) commercial-premiumand superpremium still wines and sparkling wines8 As in the original modelthere are two types of grapes premium and nonpremium Nonpremium wines usenonpremium grapes exclusively superpremium wines use mainly premium grapesand commercial-premium and sparkling wines use both types of grapes In theregional dimension the number of countries and country groups has expandedfrom 10 in the original model to 51 44 individual nations and 7 compositeregions The modelrsquos database is calibrated to 2014 for this paper based on datain Anderson et al (2017) trade data for which are downloaded from httpscomtradeunorg The modelrsquos supply-and-demand equations are based on the orig-inal ORANI modelrsquos theory (see Dixon Parmenter Sutton and Vincent 1982) andthe model is implemented using GEMPACK software (Harrison Horridge Jerieand Pearson 2014)

An enhancement of importance to the present study is the inclusion of exchange-rate variables in the model which allow a distinction between price impacts asobserved in local currency units and those observed in 2014 US dollars

In the model the grape and wine sectors minimize costs of intermediate inputssubject to weak constant elasticity of substitution (CES) substitutability betweeninputs By assumption no intermediate inputs are imported from other countries9Hence

Xcid frac14 f ethX1id CESfrac12P

cid=P1id $THORN eth1THORN

P1id X1id frac14X

c

Xcid P

cid eth2THORN

whereXcid is the quantity demanded of commodity c by grape or wine industry i in

region d Pcid is the corresponding price and X1id and P1id are the respective inter-

mediate composite quantities and prices

Two primary factors are employed in the sector labor (the quantity of which isendogenous with perfectly elastic supply) and capital Capital is usually treated asexogenous in quantity with rates of return bearing all the adjustment in thevarious scenarios This reflects the fact that grapes (a perennial crop) and wine-

8Commercial-premium still wines are defined by Anderson and Nelgen (2011) to be those priced betweenUS$250 and $750 per litre pre-tax at a countryrsquos border or wholesale9An exception concerns bulk wine imports used mainly in the bottling of commercial premium wine in theUK USA Canada Japan and Germany

246 UK and Global Wine Markets by 2025 and Implications of Brexit

plant capacity adjust slowly to market signals

Lid frac14 f ethFid CESfrac12Wid=PFid $THORN eth3THORN

Kid frac14 f ethFid CESethRid=PFid $THORN eth4THORN

PFid Fid frac14 Lid Wid thorn Kid Rid eth5THORN

Grape and wine producers are assumed to minimize costs subject to CES substitu-tion between capital and labor Equations (3) to (5) show primary factor demandsfor the labor composite Lid and capital Kid subject to a composite factor demandFid by industry i in region d The factor prices are Wid for labor Rid for capitalrentals and PFid for composite factor prices

Appendix Table A1Cumulative Consumption and Population Growth Rates and Changes in the Real Exchange Rate

(RER) Relative to the US Dollar 2014 to 2025 without Brexit ()

Aggregateconsumption Poprsquon RER

Aggregateconsumption Poprsquon RER

France 18 4 minus11 Australia 35 11 minus17Italy 11 2 minus9 New Zealand 32 9 minus26Portugal 14 0 minus9 Canada 27 8 minus18Spain 26 8 minus9 United States 31 8 0Austria 19 4 minus7 Argentina 7 10 109Belgium 20 7 minus9 Brazil 16 8 minus29Denmark 22 2 minus9 Chile 55 8 minus2Finland 21 3 minus7 Mexico 42 12 minus8Germany 14 minus2 minus11 Uruguay 45 3 1Greece 22 minus1 minus14 Other Latin America 60 10 minus5Ireland 42 12 minus9 South Africa 36 12 minus1Netherlands 21 4 minus9 Turkey 50 8 20Sweden 24 9 minus13 North Africa 53 11 0Switzerland 18 8 minus6 Other Africa 109 18 84United Kingdom 32 6 1 Middle East 52 18 minus12Other West Europe 21 10 minus1 China 79 3 5Bulgaria 41 minus7 7 Hong Kong 42 3 2Croatia 20 minus2 minus1 India 134 13 17Georgia 35 0 23 Japan 11 minus3 minus24Hungary 25 minus3 minus11 Korea 38 1 minus9Moldova 49 minus11 13 Malaysia 62 15 minus16Romania 45 minus4 22 Philippines 75 18 7Russia 18 minus2 minus8 Singapore 44 21 minus22Ukraine 22 minus5 14 Taiwan 29 1 minus13Other East Europe 40 minus5 48 Thailand 47 3 minus9

Other Asia 99 10 10

Source Authorsrsquo compilation from projections by various international agencies and from global economy-wide modeling by Anderson andStrutt (2016)

Kym Anderson and Glyn Wittwer 247

The composite factor demand Fid is proportional to total output Qid subject to aprimary factor using technology Aid Hence

Fid frac14 Qid Aid eth6THORN

The perfectly competitive zero-pure-profit condition is that total revenue valued atthe output price P0s

i multiplied by Qid equals the total production cost

P0si Qid frac14

X

cPcid X1cid thorn

X

oWo

id Loid thorn Rid Kid eth7THORN

Household demands follow a linear expenditure system in each region We reducethe optimizing problem for household consumption of each commodity subjectto a budget constraint to equations describing subsistence and discretionarydemands Aggregate subsistence expenditure WSUBd depends only on consumerprices P3cd for each commodity and the number of households N as per-capita sub-sistence quantities XSUBcd subject to given preferences are constant

WSUBd frac14X

c

P3cd XSUBcd Nd eth8THORN

Discretionary expenditures for each commodity (the left-hand side of equation (9))are equal to the marginal budget share (βcd) of aggregate discretionary expenditureThis aggregate is the bracketed term on the right-hand side of equation (9) whereW3TOTd is aggregate nominal expenditure

P3cdethX3cd XSUBcd NdTHORN frac14 βcdethW3TOTd WSUBdTHORN eth9THORN

Because real aggregate consumption is usually exogenous in our partial equilibriumsimulations the linear expenditure system determines the consumption shares ofindividual final commodities (ie the five wine types plus a composite of all otherconsumption items) driven by changes in relative prices as faced by domestic con-sumers The income elasticity of demand for each commodity is equal to the mar-ginal budget share divided by the expenditure share This number varies from 05for nonpremium wine to 17 for superpremium still wine The income elasticity ofdemand for other consumption is very close to 10 because wine accounts for anaverage of only 03 of aggregate expenditures globally and no more than 11in any country (Anderson and Nelgen 2011 Table 166)

A feature of our revised model of world wine markets is the inclusion of nominalexchange rates These appear directly in the equation linking retail prices (P3scd ) toproducer prices by country of origin (P0s

c ) where c denotes the wine type

P3 scd frac14 P0s

cfd

fsTtarcd T

taxcd thorn Pm

cd eth10THORN

248 UK and Global Wine Markets by 2025 and Implications of Brexit

The exchange rates in the consuming and producing regions are fd and fs respec-tively expressed as local currency units per $US Ttar

cd is the power of the tariff in theconsuming region and Ttax

cd is the power of the domestic consumption (or excise) taxprior to any generic value-added or goods-and-services tax Pm

cd is the price of marginm assumed to be locally supplied nontradable and therefore unaffected by theexchange rate

A given level of consumption for wine type c (X3cd) is satisfied using theArmington (1969) assumption in which wines from different countries of originare imperfectly substitutable First domestic wine is imperfectly substitutable witha composite of imports

X3sscd frac14 f ethX3cd CESethP3sscd=P3cdTHORNTHORN ss frac14 domestic imports eth11THORN

and then imports by origin (X3scd ) are determined in a second CES equation

X3scd frac14 f ethX3ssfrac14importscd CESethP3scd=P3

ssfrac14importscd THORNTHORN eth12THORN

The model enables us to show how changes in international competitiveness affectthe worldrsquos wine markets A crucial part of this exercise is explaining how pricesdetermined outside the grape and wine markets influence these markets Becausethe model is in partial equilibrium to depict the impacts of changes in internationalcompetitiveness outside price changes need to be imposed as shocks on the modelThe price of intermediate inputs other than grapes and wine shown in equations (1)and (2) is set equal to the price of GDP (Pg

d ) multiplied by a shifter Fcd

Pcid frac14 Fc

dPgd eth13THORN

If no specific price observations are available the shifter Fcd remains exogenous and

unshocked with the change in price being determined by a shock to the price ofGDP If observations are available for specific input price movements the shifterFc

dbecomes endogenous with Pc

id now exogenous and shocked

Wid frac14 Fwd P

gd eth14THORN

Wage rates are treated similarly In equation (14) if the wage shifter Fwd is exogenous

changes in wage rates Wid are determined by changes in the price of GDP If wagerate data are available Fw

d becomes endogenous andwage rates are shocked directly

Pmcd frac14 Fm

d Pgd eth15THORN

The prices of trade and transport margins are also determined by the price of GDP ifthe shifter Fm

d in equation (15) is exogenous

Changes in international competitiveness depend on changes in relative pricelevels and changes in nominal exchange rates In equation (16) fR

s denotes real

Kym Anderson and Glyn Wittwer 249

exchange-rate movements relative to the US dollar

fRs frac14 Pg

s =frac12PgUSA

fs$ eth16THORN

In equation (16) the nominal exchange rate for the United States is alwaysunchanged because nominal and real exchange rates are expressed relative to theUS currency

Changes in international market conditions may have impacts in one direction onproducer prices as expressed in US dollars P0s

i and potentially in the opposite direc-tion in local currency units Hence we calculate real producer pricesP0s

iloc in localcurrency terms (ie the price most relevant to domestic producers)

P0siloc frac14 P0s

ifs=P

gs eth17THORN

To obtain real price changes in local currency terms we convert US dollar prices(P3 s

cd for the source-specific price and P3cd for the source-composite price) to reallocal currency prices (P3 s

cdloc and P3cdloc) using the consumer price index (CPI)(Pc

d ) as the deflator

P3 scdloc frac14 P3 s

cdfd=P

cd eth18THORN

and

P3cdloc frac14 P3cdfd=Pcd eth19THORN

References

Anderson K and Nelgen S (2011) Global Wine Markets 1961 to 2009 A StatisticalCompendium Adelaide University of Adelaide Press Also available as an e-book atwwwadelaideeduaupresstitlesglobal-wine

Anderson K Nelgen S and Pinilla V (2017) Global Wine Markets 1860 to 2016 AStatistical Compendium Adelaide University of Adelaide Press (forthcoming) Also tobe available as an e-book at wwwadelaideeduaupress

Anderson K and Pinilla V (with the assistance of A J Holmes) (2017)Annual Database ofGlobal Wine Markets 1835 to 2016 Wine Economics Research Centre University ofAdelaide to be posted in October at wwwadelaideeduauwine-econdatabases

Anderson K and Strutt A (2016) Impacts of Asiarsquos rise on African and Latin Americantrade Projections to 2030 World Economy 39(2) 172ndash194

Anderson K andWittwer G (2013) Modeling global wine markets to 2018 Exchange ratestaste changes and Chinarsquos import growth Journal of Wine Economics 8(2) 131ndash158

Anderson K and Wittwer G (2015) Asiarsquos evolving role in global wine markets ChinaEconomic Review 35 1ndash14

250 UK and Global Wine Markets by 2025 and Implications of Brexit

Armington P A (1969) A theory of demand for products distinguished by place of produc-tion IMF Staff Papers 16 159ndash178

Baldwin R E (ed) (2016) Brexit Beckons Thinking Ahead by Leading Economists LondonCentre for Economic Policy Research

Campos N F (2016) Lousy experts Looking back at the ex ante estimates of the costs ofBrexit In R E Baldwin (ed) Brexit Beckons Thinking Ahead by Leading Economists(ch 3) London Centre for Economic Policy Research

Dixon P Parmenter B Sutton J and Vincent D (1982) ORANI A Multisectoral Modelof the Australian Economy Contributions to Economic Analysis 142 Amsterdam North-Holland

European Commission (2015) Trade for All Towards a More Responsible Trade andInvestment Policy Brussels European Commission

Harrison J Horridge M Jerie M and Pearson K (2014) GEMPACK ManualMelbourne GEMPACK Software

HM Treasury (2017) Spring Budget 2017 London HM TreasuryHolmes A J and Anderson K (2017a)Annual Database of National Beverage Consumption

Volumes and Expenditures 1950 to 2015 Wine Economics Research Centre University ofAdelaide posted at wwwadelaideeduauwine-econdatabases

Holmes A J and Anderson K (2017b) Convergence in National Alcohol ConsumptionPatterns New Global Indicators Journal of Wine Economics 12(2) 117ndash148

Ludington C C (2013) The Politics of Wine in Britain A New Cultural History BasingstokeUK Palgrave Macmillan

Office of National Statistics (2017) Analysis of real earnings London Office of NationalStatistics Available at httpswwwonsgovukemploymentandlabourmarketpeopleinworkearningsandworkinghoursarticlessupplementaryanalysisofaverageweeklyearningslatest

Rollo J Borchert I Dawar K Holmes P and Winters L A (2016) The World TradeOrganisation A safety net for a post-Brexit UK trade policy Briefing Paper 1UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-1-final-1pdf

Smith A (2017) Brexit Hard truths and hard choices UKTPO blog 19 June Available athttpsblogssussexacukuktpo20170619brexit-hard-truths-and-hard-choices

Swinbank A (2017) World trade rules and the policy options for British agriculture post-Brexit Briefing Paper 7 UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-7pdf

Viner J (1950) The Customs Union Issue New York Carnegie Endowment for InternationalPeace

Wine Australia (2015) Export Market Guide European Union Adelaide Wine AustraliaWittwer G Berger N and Anderson K (2003) A model of the worldrsquos wine markets

Economic Modelling 20(3) 487ndash506

Kym Anderson and Glyn Wittwer 251

Page 18: U.K. and Global Wine Markets by 2025, and Implications of ... · U.K. and Global Wine Markets by 2025, and Implications of Brexit* Kym Andersona and Glyn Wittwerb Abstract The United

United Kingdom in local currency terms (20 because of real depreciation of theBritish pound 4 because of new tariffs on EU Chilean and South Africanwines and ndash2 because of slower UK income growth) The volume of UKwine consumption is 28 lower 16 because of slower UK economic growth7 because of real depreciation of the British pound and 5 because of newtariffs Superpremium still-wine sales are the most affected dropping by two-fifths while sparkling and commercial-premium wines drop a bit less than one-quarter Because the average price rises by more than the decline in volume soldthe aggregate value of UK sales even in local currency terms falls under thisldquolargerdquo Brexit scenario Under the ldquosmallrdquo Brexit scenario the consumer price ofwine in 2025 is 11 higher in the United Kingdom and its volume of wine consump-tion is 17 lower

The volume of projected UK imports in 2025 is 427 ML or nearly one-quarterless in the ldquolargerdquo scenario than in the baseline scenario comprising 58ML less spar-kling wine 31 ML less superpremium still wine and 339 ML less commercial-premium wine World imports are lower by just 239 ML because imports byother countries are 189 ML higher in response to the lower international wineprices in this scenario In value terms UK imports are $175 billion (or 27)lower in 2025 because of ldquolargerdquo Brexit $113 billion because of lower incomes$038 billion because of the fall in the British pound and $014 billion because ofthe rise in wine-import tariffs (Table 4) These aggregate trade impacts are a littlemore than half this size under the ldquosmallrdquo Brexit scenario

Despite the import levels falling because of raised import tariffs domestic con-sumption of all three quality categories of UK-produced wine is lower with thanwithout Brexit because of shrunken demand for all wines resulting from loweredUK incomes and raised local prices because of devaluation of the British poundThe British poundrsquos devaluation does make it easier for the United Kingdom tosell wines abroad though The countryrsquos exports are 7 ML or nearly 5 higherin 2025 in the ldquolargerdquo Brexit scenario and UK production is 3 higher ThoseUK exports (or re-exports of imported bulk wine after it is bottled in the UnitedKingdom) that go to EU27 countries are reduced though because of the tariffnow imposed at the new EU border

Without Brexit the United Kingdomrsquos shares of global wine imports are slightlyhigher in volume terms in 2025 than in 2010ndash2015 but 2 percentage points lower invalue terms thanks to East Asiarsquos expanding demand for imports of premium winesWith ldquolargerdquo Brexit however that value share is an additional 2 percentage pointslower and the volume share is almost 5 points lower (Figure 8) The net effects ofthese impacts on global trade are shown in Figure 9 Most of the initial tradeeffect of Brexit is a large decline in net imports of wine by the United Kingdomwith very little offsetting positive effect on trade in the rest of the world Theldquosmallrdquo Brexit numbers are a bit more than half of those for ldquolargerdquo Brexit

238 UK and Global Wine Markets by 2025 and Implications of Brexit

Table4

Difference

in2025

Projected

VolumeandValue

ofWineIm

portsb

ytheUnitedKingdom

andtheResto

fthe

World

asaConsequence

oftheInitial

BrexitS

hock

(MLand2014US$

Million

ldquoLargerdquoSc

enario)

Volum

e(M

L)

Value

(US$

million)

NP+CPa

SuperPrb

Sparkling

TOTAL

NP+CPa

SuperPrb

Sparkling

TOTAL

ΔUK

impo

rtsdu

eto

Low

erincomes

minus198

minus20

minus29

minus247

58minus644

minus253

minus234

minus1131

65Low

erpo

und

minus70

minus10

minus14

minus93

22minus248

minus127

minus102

minus476

27Highertariffs

minus71

minus1minus16

minus87

20minus110

minus8minus2

4minus143

8TOTA

Lminus339

minus31

minus58

minus427

100

minus1001

minus388

minus360

minus1750

100

difffrom

base

2332

3325

2432

3227

of

totalc

uts

797

14100

5722

21100

ΔROW

netim

ports

143

2125

189

230

143

minus181

192

ΔWORLD

TRADE

minus195

minus10

minus34

minus239

minus763

minus246

minus543

minus1552

aNon

prem

ium

plus

commercial-premium

still

wines

bSu

perpremium

still

wines

SourceA

utho

rsrsquomod

elresults

Kym Anderson and Glyn Wittwer 239

The aggregate effect of ldquolargerdquo Brexit on the market shares of various wine-export-ing countries in the United Kingdom is almost indiscernible even with one decimalpoint The projected 2025 shares are quite different from the actual 2014 shares forseveral countries They are much smaller in 2025 for South Africa Australia andNew Zealand (and the United States in volume terms) and are much larger involume for Spain and in value for Italy These results occur because wine-exportingcountries benefit differentially from the varying rates of growth in net importdemand for wine in non-UK countries over this projection period The most impor-tant projected changes are the increase in the real value of annual wine importsbetween 2014 and 2025 by China (200 or $3 billion) Other Asia (110 or $22billion) and Africa (270 or $16 billion) More than half of Australiarsquos increasein annual exports from 2014 to 2025 go to Asia and more than half of SouthAfricarsquos increase in exports go to Other Africa

Table 5 reveals that ldquolargerdquo Brexit lowers EU Chilean and South African wineexports by 150 ML or US$12 billion in the case of the European Union with someof their exports diverted from the United Kingdom to the EU27 and other marketsin competition with New World exporters Although the United States Australiaand Argentina sell only a little less to the United Kingdom they also sell less toother countries For Chile and South Africa which lose their preferential access toUK (but not to EU27) markets in this Brexit scenario some exports are redirected

Figure 8

UK Shares of World Wine Imports 2010ndash2015 and Projected to 2025 without and with theldquoLargerdquo Initial Brexit Shock ()

Fig8-BW

onlin

eBW

inprint

Sources Anderson and Pinilla (2017) and authorsrsquo model results

240 UK and Global Wine Markets by 2025 and Implications of Brexit

from the United Kingdom to EU27 countries ndash but again they export less overallGlobal wine trade in 2025 is 240 ML (19) or $18 billion (35) less under thisldquolargerdquo Brexit scenario The percentage by which wine exportersrsquo trade shrinks isgreater for values than for volumes because of changes in relative prices of differ-ent-quality wines Those differences are shown in the numbers in parentheses inTable 5

Three other points are worth making about Table 5 One is that Australia sellsslightly more to the United Kingdom in the ldquosmallrdquo Brexit scenario rather thanslightly less as in the ldquolargerdquo Brexit case Evidently the negative income and price(devaluation) effects do not more than offset the positive trade-diverting effect onAustralian exports to the United Kingdom of removing preferences in the ldquosmallrdquoscenario Second New Zealand sells slightly more to non-UK countries underBrexit despite greater competition from the EU27 Chile and South Africa Thisanomaly is due to changes in the relative prices of different qualities of wine inglobal wine markets bearing in mind that New Zealand has the worldrsquos highestaverage price for still-wine exports And third the value (but not the volume) ofexports of ldquoOtherrdquo countries to markets other than the United Kingdom arehigher under Brexit This result too is due to changes in the relative prices of differ-ent qualities of wine in global wine markets

Figure 9

Difference in 2025 Wine Import Volumes and Values as a Result of the ldquoLargerdquo Initial BrexitShock (ML and US$ Million in 2014 US dollars)

Fig9-BW

onlin

eBW

inprint

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 241

B Subsequent Impact of Brexit from a UK-EU27 FTA

The next-most-likely step in the Brexit process is for the United Kingdom to nego-tiate a new trade arrangement with the EU27 We therefore assume that a UK-EU27FTAwith free bilateral wine trade is implemented and adjusted to by 2025 and thatprogress toward that end occurs soon enough that the adverse macroeconomicshocks from the initial impact of uncertainty over the Brexit process are confinedto those assumed in the ldquosmallrdquo scenario outlined above In this subsequent scenariowe assume that the British pound returns to the value it would have reached in 2025in the absence of Brexit (ie reversing the 10 devaluation assumed in the ldquosmallrdquoinitial Brexit scenario) and that real incomes in the United Kingdom are 8 higher

Table 5Difference in 2025 BilateralWine Import Volumes and Values fromKey Exporters by the United

Kingdom and the Rest of the World (RoW) as a Result of Initial Brexit Shock(ML and 2014US$ Million)a

(a) ldquolargerdquo scenario

Volume (ML) Value (2014US$ million)

UK RoW WORLD () UK RoW WORLD ()

EU27 minus287 136 minus150 (ndash17) minus1187 minus5 minus1192 (ndash31)Chile minus59 35 minus25 (ndash30) minus169 31 minus138 (ndash48)South Africa minus53 35 minus18 (ndash32) minus105 20 minus85 (ndash67)USA minus7 minus6 minus13 (ndash24) minus75 minus40 minus115 (ndash50)Australia minus4 minus3 minus7 (ndash09) minus25 minus65 minus90 (ndash30)Argentina minus3 minus9 minus12 (ndash48) minus16 minus39 minus55 (ndash52)New Zealand minus11 9 minus2 (ndash09) minus162 71 minus91 (ndash43)Others minus2 minus10 minus12 (ndash02) minus11 minus52 minus63 (ndash44)WORLD minus427 187 minus240 (ndash19) minus1750 minus79 minus1829 (ndash35)

(a) ldquosmallrdquo scenario

Volume (ML) Value (2014US$ million)

UK RoW WORLD () UK RoW WORLD ()

EU27 minus178 82 minus96 (ndash12) minus692 minus43 minus736 (ndash19)Chile minus46 28 minus18 (ndash24) minus128 36 minus91 (ndash32)South Africa minus43 29 minus14 (ndash42) minus82 23 minus59 (ndash47)USA 1 minus6 minus5 (ndash11) minus23 minus28 minus51 (ndash22)Australia 5 minus10 minus5 (ndash06) 19 minus56 minus38 (ndash13)Argentina 0 minus6 minus6 (ndash26) minus3 minus25 minus29 (ndash27)New Zealand minus5 4 minus1 (ndash06) minus80 34 minus46 (ndash22)Others 0 minus9 minus9 (ndash01) minus1 minus33 minus34 (ndash24)WORLD minus266 112 minus154 (ndash13) minus991 minus92 minus1083 (ndash21)

a Numbers in parentheses are the percentage difference between the Brexit and baseline scenarios for 2025 projected wine-import volumes orvalues by source

Source Authorsrsquo model results

242 UK and Global Wine Markets by 2025 and Implications of Brexit

in 2025 than in the ldquosmallrdquo scenario ndash that is they regain two-thirds of the value theywould have reached without Brexit-related uncertainty

This subsequent development in the Brexit process reverses most of but not allthe initial effects of Brexit by 2025 because of our assumption that the lost growthin the initial years of uncertainty following the Brexit vote are only partly recoveredby 2025 following the implementation of a UK-EU27 FTA Moreover the longer ittakes before this FTA is finalized and implemented the longer the estimated initialadverse effects persist and the larger the cumulative cost of Brexit to UK wine con-sumers and to grape and wine producers in wine-exporting countries

Table 6 summarizes the subsequent trade effects for 2025 It suggests that all butone-tenth of the loss in value of world trade in wine from the initial ldquosmallrdquo impact isrestored most of it because of a smaller reduction in the United Kingdomrsquos wineimports following the FTA Most of that improved outcome is because of recoveredimports from the EU27 commensurate with the latterrsquos high share of UK importsEven though Chile and South Africa are assumed in this scenario to have not yetsigned an FTA with the United Kingdom they export slightly more to the UnitedKingdom (and even more to all other countries) than in the initial ldquosmallrdquo scenario

Within the United Kingdom this FTA brings down the local currency consumerprice of wine by 9 largely offsetting the 11 rise in the initial ldquosmallrdquo Brexit sce-nario and it raises the volume consumed in the United Kingdom by 18 fully off-setting the 17 fall in the initial ldquosmallrdquo Brexit scenario

Table 6Difference in 2025 Bilateral Wine-Import Volumes and Values from Key Exporters by the

United Kingdom and the Rest of the World (RoW) as a Result of Implementing a UK-EU27FTA Following Initial ldquoSmallrdquo Brexit Shock (ML and 2014US$ Million Difference Relative to

Initial ldquoSmallrdquo Brexit Scenario)a

Volume (ML) Value (2014US$ million)

UK RoW WORLD (ldquosmallrdquo)a UK RoW WORLD (ldquosmallrdquo)a

EU27 212 minus112 100 (minus96) 750 0 750 (minus806)Chile 1 3 4 (minus18) 3 40 43 (minus92)South Africa 3 2 5 (minus14) 3 18 21 (minus60)USA 3 3 6 (minus5) 26 21 47 (minus50)Australia 1 1 2 (minus5) 8 38 46 (minus33)Argentina 1 4 5 (minus6) 4 20 24 (minus28)New Zealand 4 minus3 1 (minus1) 61 minus25 36 (minus46)Others 1 6 7 (minus9) 5 minus46 minus41 (92)WORLD 226 minus96 130 (minus154) 860 66 926 (minus1022)

a Numbers in parentheses are the world trade differences between the ldquosmallrdquo initial Brexit scenario and the baseline scenario copied fromcolumns 3 and 7 of Table 5(b)

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 243

In short this subsequent step in the Brexit process can be expected to restore muchof the initial adverse effects in the United Kingdom of the Brexit vote by 2025 Thewine exporters in the EU27 Chile and South Africa also lose less in this scenarioas compared with the initial ldquosmallrdquo scenario while other key exporters sell nearlyas much in 2025 in this scenario as they are projected to in the baseline ndash that isthey are beneficiaries of the reduced discrimination in global wine markets in thiscase where Chile and South Africa no longer enjoy preferential access to the UKmarket (and even though EU27 producers are again allowed to do so) Keep inmind though that although the estimated losses in the initial scenarios are reportedjust for 2025 they are expected to be felt throughout the preceding years of uncertaintyTherefore the longer it is before uncertainty abates and the United Kingdomrsquos currentpreferential trading arrangements are replaced by newones such as a UK-EU27 FTAthe higher the cumulative cost of the Brexit vote to the wine trade

VII Caveats and Conclusions

The above Brexit simulations are just a few of many scenarios that could be modeledObvious additional ones could also assume that FTAs are reached between theUnited Kingdom and other trading partners including Chile and South Africaand such countries as Australia and New Zealand Some of these other countriesare already exploring the ECrsquos invitation (see European Commission 2015) to nego-tiate an FTA with what will be the EU27 The sequence in which FTAs are signedand the speed with which they are implemented will matter (as was also the casewith the sequential signing over the past decade of bilateral FTAs with NortheastAsian countries by Chile Australia and New Zealand see Anderson andWittwer 2015) Even if the United Kingdom were able to sign additional bilateralFTAs and begin implementing them before 2025 it would have little effect on theabove results (because wine tariffs are such a minor contributor to them) unlessthose FTAs were to accelerate the United Kingdomrsquos economic growth and therise in the British poundrsquos value before 2025

We assume above that no changes are made to alcohol excise duties in the UnitedKingdom following Brexit even though they are scheduled to be progressively raisedwith inflation and may be raised even more for wine relative to spirits to offset theopposite effects of Brexit on those two domestic industries Nor do we make anyallowance for UK increases in consumer wine prices that may be needed to coverthe higher cost of clearing customs on imports from EU27 countries (which maybe more or less offset by increased smuggling and duty-free purchasing)Importers of fine wines including individuals who buy Bordeaux en primeur maywell reduce their demand because of concerns that delays on docks will affect thequality of their wines or because of greater currency uncertainty

To summarize Brexit is costly initially to UK consumers of wine (and of manyother tradable products) because the domestic retail price in local currency tax-

244 UK and Global Wine Markets by 2025 and Implications of Brexit

inclusive terms is 22 higher than otherwise in 2025 and the volume of wine con-sumed domestically is 28 lower in the ldquolargerdquo Brexit scenario (or 11 higher pricesand 17 lower quantities in the ldquosmallrdquo case) by 2025 unless a UK-EU27 FTA isdeveloped before then Even if such an FTA does get signed ratified by all 28 par-liaments and implemented by 2025 the slower income growth in the interim meansa smaller UK wine market in 2025 than would otherwise have been the case Thevolume reduction is a blow to many participants in UK wine-bottling transportingstoring wholesaling and retailing businesses in addition to restaurants and pubsVery little of that initial impact occurs because of higher import tariffs most impor-tant is the assumed fall in UK real incomes

The small but growing number of local UK vignerons (who supply less than 05of all domestic-wine sales and less than 4 of sparkling-wine sales) is initially pro-jected to sell less wine domestically because of reduced demand and to export only alittle extra abroad while the British pound is devalued Overall their production isonly 3 higher in 2025 even in the ldquolargerdquo Brexit scenario (not taking intoaccount the fact that their casual labor is likely to be more expensive under Brexitbecause of tighter restrictions on immigration) and their production diminishes ifand when a UK-EU27 FTA is in place

Even with a UK-EU27 FTA in place EU27 wine exporters export US$56 millionless wine in 2025 thanks to Brexit Chile and South Africa export $88 million lesswine and wine exports from the rest of the world are little different in aggregatealthough slightly larger from Australia

Clearly there will be great uncertainly for some time over the possible policy out-comes to flow from Brexit and their consequent sequential impacts on UK house-hold disposable incomes foreign exchange rates and bilateral wine tariffsMeanwhile the above projections under explicit assumptions provide some idea ofhow wine markets might be affected by the most-likely first two stages of theBrexit process (agreeing on a new tariff schedule at the WTO and agreeing to andimplementing a UK-EU27 FTA) In particular they make clear that nontrivialinitial impacts could affect the domestic wine market impacts that are likely to belarger than just the direct impact of changes in bilateral tariffs If the UnitedKingdom succeeds in getting countries to agree to transition arrangements thatdelay the changes in tariffs until new FTAs are signed and ratified the initialeffects will be less dramatic than in our first scenarios but it remains to be seenwhether any such agreements can be reached In any event the net effect of Brexiton the welfare of the worldrsquos consumers and producers of wine as a whole will benegative not just initially but permanently unless new trade policy commitmentsby the United Kingdom with major wine-exporting countries are sufficiently moreliberal than current arrangements

Kym Anderson and Glyn Wittwer 245

Appendix Key Equations in the Global Wine Markets Model

Amodel of the worldrsquos wine markets was first published byWittwer et al (2003) andmuch improved by Anderson and Wittwer (2013) Several significant enhancementshave been made to that original model Wine types have been disaggregated intomore types ndash namely nonpremium (including generic bulk) commercial-premiumand superpremium still wines and sparkling wines8 As in the original modelthere are two types of grapes premium and nonpremium Nonpremium wines usenonpremium grapes exclusively superpremium wines use mainly premium grapesand commercial-premium and sparkling wines use both types of grapes In theregional dimension the number of countries and country groups has expandedfrom 10 in the original model to 51 44 individual nations and 7 compositeregions The modelrsquos database is calibrated to 2014 for this paper based on datain Anderson et al (2017) trade data for which are downloaded from httpscomtradeunorg The modelrsquos supply-and-demand equations are based on the orig-inal ORANI modelrsquos theory (see Dixon Parmenter Sutton and Vincent 1982) andthe model is implemented using GEMPACK software (Harrison Horridge Jerieand Pearson 2014)

An enhancement of importance to the present study is the inclusion of exchange-rate variables in the model which allow a distinction between price impacts asobserved in local currency units and those observed in 2014 US dollars

In the model the grape and wine sectors minimize costs of intermediate inputssubject to weak constant elasticity of substitution (CES) substitutability betweeninputs By assumption no intermediate inputs are imported from other countries9Hence

Xcid frac14 f ethX1id CESfrac12P

cid=P1id $THORN eth1THORN

P1id X1id frac14X

c

Xcid P

cid eth2THORN

whereXcid is the quantity demanded of commodity c by grape or wine industry i in

region d Pcid is the corresponding price and X1id and P1id are the respective inter-

mediate composite quantities and prices

Two primary factors are employed in the sector labor (the quantity of which isendogenous with perfectly elastic supply) and capital Capital is usually treated asexogenous in quantity with rates of return bearing all the adjustment in thevarious scenarios This reflects the fact that grapes (a perennial crop) and wine-

8Commercial-premium still wines are defined by Anderson and Nelgen (2011) to be those priced betweenUS$250 and $750 per litre pre-tax at a countryrsquos border or wholesale9An exception concerns bulk wine imports used mainly in the bottling of commercial premium wine in theUK USA Canada Japan and Germany

246 UK and Global Wine Markets by 2025 and Implications of Brexit

plant capacity adjust slowly to market signals

Lid frac14 f ethFid CESfrac12Wid=PFid $THORN eth3THORN

Kid frac14 f ethFid CESethRid=PFid $THORN eth4THORN

PFid Fid frac14 Lid Wid thorn Kid Rid eth5THORN

Grape and wine producers are assumed to minimize costs subject to CES substitu-tion between capital and labor Equations (3) to (5) show primary factor demandsfor the labor composite Lid and capital Kid subject to a composite factor demandFid by industry i in region d The factor prices are Wid for labor Rid for capitalrentals and PFid for composite factor prices

Appendix Table A1Cumulative Consumption and Population Growth Rates and Changes in the Real Exchange Rate

(RER) Relative to the US Dollar 2014 to 2025 without Brexit ()

Aggregateconsumption Poprsquon RER

Aggregateconsumption Poprsquon RER

France 18 4 minus11 Australia 35 11 minus17Italy 11 2 minus9 New Zealand 32 9 minus26Portugal 14 0 minus9 Canada 27 8 minus18Spain 26 8 minus9 United States 31 8 0Austria 19 4 minus7 Argentina 7 10 109Belgium 20 7 minus9 Brazil 16 8 minus29Denmark 22 2 minus9 Chile 55 8 minus2Finland 21 3 minus7 Mexico 42 12 minus8Germany 14 minus2 minus11 Uruguay 45 3 1Greece 22 minus1 minus14 Other Latin America 60 10 minus5Ireland 42 12 minus9 South Africa 36 12 minus1Netherlands 21 4 minus9 Turkey 50 8 20Sweden 24 9 minus13 North Africa 53 11 0Switzerland 18 8 minus6 Other Africa 109 18 84United Kingdom 32 6 1 Middle East 52 18 minus12Other West Europe 21 10 minus1 China 79 3 5Bulgaria 41 minus7 7 Hong Kong 42 3 2Croatia 20 minus2 minus1 India 134 13 17Georgia 35 0 23 Japan 11 minus3 minus24Hungary 25 minus3 minus11 Korea 38 1 minus9Moldova 49 minus11 13 Malaysia 62 15 minus16Romania 45 minus4 22 Philippines 75 18 7Russia 18 minus2 minus8 Singapore 44 21 minus22Ukraine 22 minus5 14 Taiwan 29 1 minus13Other East Europe 40 minus5 48 Thailand 47 3 minus9

Other Asia 99 10 10

Source Authorsrsquo compilation from projections by various international agencies and from global economy-wide modeling by Anderson andStrutt (2016)

Kym Anderson and Glyn Wittwer 247

The composite factor demand Fid is proportional to total output Qid subject to aprimary factor using technology Aid Hence

Fid frac14 Qid Aid eth6THORN

The perfectly competitive zero-pure-profit condition is that total revenue valued atthe output price P0s

i multiplied by Qid equals the total production cost

P0si Qid frac14

X

cPcid X1cid thorn

X

oWo

id Loid thorn Rid Kid eth7THORN

Household demands follow a linear expenditure system in each region We reducethe optimizing problem for household consumption of each commodity subjectto a budget constraint to equations describing subsistence and discretionarydemands Aggregate subsistence expenditure WSUBd depends only on consumerprices P3cd for each commodity and the number of households N as per-capita sub-sistence quantities XSUBcd subject to given preferences are constant

WSUBd frac14X

c

P3cd XSUBcd Nd eth8THORN

Discretionary expenditures for each commodity (the left-hand side of equation (9))are equal to the marginal budget share (βcd) of aggregate discretionary expenditureThis aggregate is the bracketed term on the right-hand side of equation (9) whereW3TOTd is aggregate nominal expenditure

P3cdethX3cd XSUBcd NdTHORN frac14 βcdethW3TOTd WSUBdTHORN eth9THORN

Because real aggregate consumption is usually exogenous in our partial equilibriumsimulations the linear expenditure system determines the consumption shares ofindividual final commodities (ie the five wine types plus a composite of all otherconsumption items) driven by changes in relative prices as faced by domestic con-sumers The income elasticity of demand for each commodity is equal to the mar-ginal budget share divided by the expenditure share This number varies from 05for nonpremium wine to 17 for superpremium still wine The income elasticity ofdemand for other consumption is very close to 10 because wine accounts for anaverage of only 03 of aggregate expenditures globally and no more than 11in any country (Anderson and Nelgen 2011 Table 166)

A feature of our revised model of world wine markets is the inclusion of nominalexchange rates These appear directly in the equation linking retail prices (P3scd ) toproducer prices by country of origin (P0s

c ) where c denotes the wine type

P3 scd frac14 P0s

cfd

fsTtarcd T

taxcd thorn Pm

cd eth10THORN

248 UK and Global Wine Markets by 2025 and Implications of Brexit

The exchange rates in the consuming and producing regions are fd and fs respec-tively expressed as local currency units per $US Ttar

cd is the power of the tariff in theconsuming region and Ttax

cd is the power of the domestic consumption (or excise) taxprior to any generic value-added or goods-and-services tax Pm

cd is the price of marginm assumed to be locally supplied nontradable and therefore unaffected by theexchange rate

A given level of consumption for wine type c (X3cd) is satisfied using theArmington (1969) assumption in which wines from different countries of originare imperfectly substitutable First domestic wine is imperfectly substitutable witha composite of imports

X3sscd frac14 f ethX3cd CESethP3sscd=P3cdTHORNTHORN ss frac14 domestic imports eth11THORN

and then imports by origin (X3scd ) are determined in a second CES equation

X3scd frac14 f ethX3ssfrac14importscd CESethP3scd=P3

ssfrac14importscd THORNTHORN eth12THORN

The model enables us to show how changes in international competitiveness affectthe worldrsquos wine markets A crucial part of this exercise is explaining how pricesdetermined outside the grape and wine markets influence these markets Becausethe model is in partial equilibrium to depict the impacts of changes in internationalcompetitiveness outside price changes need to be imposed as shocks on the modelThe price of intermediate inputs other than grapes and wine shown in equations (1)and (2) is set equal to the price of GDP (Pg

d ) multiplied by a shifter Fcd

Pcid frac14 Fc

dPgd eth13THORN

If no specific price observations are available the shifter Fcd remains exogenous and

unshocked with the change in price being determined by a shock to the price ofGDP If observations are available for specific input price movements the shifterFc

dbecomes endogenous with Pc

id now exogenous and shocked

Wid frac14 Fwd P

gd eth14THORN

Wage rates are treated similarly In equation (14) if the wage shifter Fwd is exogenous

changes in wage rates Wid are determined by changes in the price of GDP If wagerate data are available Fw

d becomes endogenous andwage rates are shocked directly

Pmcd frac14 Fm

d Pgd eth15THORN

The prices of trade and transport margins are also determined by the price of GDP ifthe shifter Fm

d in equation (15) is exogenous

Changes in international competitiveness depend on changes in relative pricelevels and changes in nominal exchange rates In equation (16) fR

s denotes real

Kym Anderson and Glyn Wittwer 249

exchange-rate movements relative to the US dollar

fRs frac14 Pg

s =frac12PgUSA

fs$ eth16THORN

In equation (16) the nominal exchange rate for the United States is alwaysunchanged because nominal and real exchange rates are expressed relative to theUS currency

Changes in international market conditions may have impacts in one direction onproducer prices as expressed in US dollars P0s

i and potentially in the opposite direc-tion in local currency units Hence we calculate real producer pricesP0s

iloc in localcurrency terms (ie the price most relevant to domestic producers)

P0siloc frac14 P0s

ifs=P

gs eth17THORN

To obtain real price changes in local currency terms we convert US dollar prices(P3 s

cd for the source-specific price and P3cd for the source-composite price) to reallocal currency prices (P3 s

cdloc and P3cdloc) using the consumer price index (CPI)(Pc

d ) as the deflator

P3 scdloc frac14 P3 s

cdfd=P

cd eth18THORN

and

P3cdloc frac14 P3cdfd=Pcd eth19THORN

References

Anderson K and Nelgen S (2011) Global Wine Markets 1961 to 2009 A StatisticalCompendium Adelaide University of Adelaide Press Also available as an e-book atwwwadelaideeduaupresstitlesglobal-wine

Anderson K Nelgen S and Pinilla V (2017) Global Wine Markets 1860 to 2016 AStatistical Compendium Adelaide University of Adelaide Press (forthcoming) Also tobe available as an e-book at wwwadelaideeduaupress

Anderson K and Pinilla V (with the assistance of A J Holmes) (2017)Annual Database ofGlobal Wine Markets 1835 to 2016 Wine Economics Research Centre University ofAdelaide to be posted in October at wwwadelaideeduauwine-econdatabases

Anderson K and Strutt A (2016) Impacts of Asiarsquos rise on African and Latin Americantrade Projections to 2030 World Economy 39(2) 172ndash194

Anderson K andWittwer G (2013) Modeling global wine markets to 2018 Exchange ratestaste changes and Chinarsquos import growth Journal of Wine Economics 8(2) 131ndash158

Anderson K and Wittwer G (2015) Asiarsquos evolving role in global wine markets ChinaEconomic Review 35 1ndash14

250 UK and Global Wine Markets by 2025 and Implications of Brexit

Armington P A (1969) A theory of demand for products distinguished by place of produc-tion IMF Staff Papers 16 159ndash178

Baldwin R E (ed) (2016) Brexit Beckons Thinking Ahead by Leading Economists LondonCentre for Economic Policy Research

Campos N F (2016) Lousy experts Looking back at the ex ante estimates of the costs ofBrexit In R E Baldwin (ed) Brexit Beckons Thinking Ahead by Leading Economists(ch 3) London Centre for Economic Policy Research

Dixon P Parmenter B Sutton J and Vincent D (1982) ORANI A Multisectoral Modelof the Australian Economy Contributions to Economic Analysis 142 Amsterdam North-Holland

European Commission (2015) Trade for All Towards a More Responsible Trade andInvestment Policy Brussels European Commission

Harrison J Horridge M Jerie M and Pearson K (2014) GEMPACK ManualMelbourne GEMPACK Software

HM Treasury (2017) Spring Budget 2017 London HM TreasuryHolmes A J and Anderson K (2017a)Annual Database of National Beverage Consumption

Volumes and Expenditures 1950 to 2015 Wine Economics Research Centre University ofAdelaide posted at wwwadelaideeduauwine-econdatabases

Holmes A J and Anderson K (2017b) Convergence in National Alcohol ConsumptionPatterns New Global Indicators Journal of Wine Economics 12(2) 117ndash148

Ludington C C (2013) The Politics of Wine in Britain A New Cultural History BasingstokeUK Palgrave Macmillan

Office of National Statistics (2017) Analysis of real earnings London Office of NationalStatistics Available at httpswwwonsgovukemploymentandlabourmarketpeopleinworkearningsandworkinghoursarticlessupplementaryanalysisofaverageweeklyearningslatest

Rollo J Borchert I Dawar K Holmes P and Winters L A (2016) The World TradeOrganisation A safety net for a post-Brexit UK trade policy Briefing Paper 1UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-1-final-1pdf

Smith A (2017) Brexit Hard truths and hard choices UKTPO blog 19 June Available athttpsblogssussexacukuktpo20170619brexit-hard-truths-and-hard-choices

Swinbank A (2017) World trade rules and the policy options for British agriculture post-Brexit Briefing Paper 7 UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-7pdf

Viner J (1950) The Customs Union Issue New York Carnegie Endowment for InternationalPeace

Wine Australia (2015) Export Market Guide European Union Adelaide Wine AustraliaWittwer G Berger N and Anderson K (2003) A model of the worldrsquos wine markets

Economic Modelling 20(3) 487ndash506

Kym Anderson and Glyn Wittwer 251

Page 19: U.K. and Global Wine Markets by 2025, and Implications of ... · U.K. and Global Wine Markets by 2025, and Implications of Brexit* Kym Andersona and Glyn Wittwerb Abstract The United

Table4

Difference

in2025

Projected

VolumeandValue

ofWineIm

portsb

ytheUnitedKingdom

andtheResto

fthe

World

asaConsequence

oftheInitial

BrexitS

hock

(MLand2014US$

Million

ldquoLargerdquoSc

enario)

Volum

e(M

L)

Value

(US$

million)

NP+CPa

SuperPrb

Sparkling

TOTAL

NP+CPa

SuperPrb

Sparkling

TOTAL

ΔUK

impo

rtsdu

eto

Low

erincomes

minus198

minus20

minus29

minus247

58minus644

minus253

minus234

minus1131

65Low

erpo

und

minus70

minus10

minus14

minus93

22minus248

minus127

minus102

minus476

27Highertariffs

minus71

minus1minus16

minus87

20minus110

minus8minus2

4minus143

8TOTA

Lminus339

minus31

minus58

minus427

100

minus1001

minus388

minus360

minus1750

100

difffrom

base

2332

3325

2432

3227

of

totalc

uts

797

14100

5722

21100

ΔROW

netim

ports

143

2125

189

230

143

minus181

192

ΔWORLD

TRADE

minus195

minus10

minus34

minus239

minus763

minus246

minus543

minus1552

aNon

prem

ium

plus

commercial-premium

still

wines

bSu

perpremium

still

wines

SourceA

utho

rsrsquomod

elresults

Kym Anderson and Glyn Wittwer 239

The aggregate effect of ldquolargerdquo Brexit on the market shares of various wine-export-ing countries in the United Kingdom is almost indiscernible even with one decimalpoint The projected 2025 shares are quite different from the actual 2014 shares forseveral countries They are much smaller in 2025 for South Africa Australia andNew Zealand (and the United States in volume terms) and are much larger involume for Spain and in value for Italy These results occur because wine-exportingcountries benefit differentially from the varying rates of growth in net importdemand for wine in non-UK countries over this projection period The most impor-tant projected changes are the increase in the real value of annual wine importsbetween 2014 and 2025 by China (200 or $3 billion) Other Asia (110 or $22billion) and Africa (270 or $16 billion) More than half of Australiarsquos increasein annual exports from 2014 to 2025 go to Asia and more than half of SouthAfricarsquos increase in exports go to Other Africa

Table 5 reveals that ldquolargerdquo Brexit lowers EU Chilean and South African wineexports by 150 ML or US$12 billion in the case of the European Union with someof their exports diverted from the United Kingdom to the EU27 and other marketsin competition with New World exporters Although the United States Australiaand Argentina sell only a little less to the United Kingdom they also sell less toother countries For Chile and South Africa which lose their preferential access toUK (but not to EU27) markets in this Brexit scenario some exports are redirected

Figure 8

UK Shares of World Wine Imports 2010ndash2015 and Projected to 2025 without and with theldquoLargerdquo Initial Brexit Shock ()

Fig8-BW

onlin

eBW

inprint

Sources Anderson and Pinilla (2017) and authorsrsquo model results

240 UK and Global Wine Markets by 2025 and Implications of Brexit

from the United Kingdom to EU27 countries ndash but again they export less overallGlobal wine trade in 2025 is 240 ML (19) or $18 billion (35) less under thisldquolargerdquo Brexit scenario The percentage by which wine exportersrsquo trade shrinks isgreater for values than for volumes because of changes in relative prices of differ-ent-quality wines Those differences are shown in the numbers in parentheses inTable 5

Three other points are worth making about Table 5 One is that Australia sellsslightly more to the United Kingdom in the ldquosmallrdquo Brexit scenario rather thanslightly less as in the ldquolargerdquo Brexit case Evidently the negative income and price(devaluation) effects do not more than offset the positive trade-diverting effect onAustralian exports to the United Kingdom of removing preferences in the ldquosmallrdquoscenario Second New Zealand sells slightly more to non-UK countries underBrexit despite greater competition from the EU27 Chile and South Africa Thisanomaly is due to changes in the relative prices of different qualities of wine inglobal wine markets bearing in mind that New Zealand has the worldrsquos highestaverage price for still-wine exports And third the value (but not the volume) ofexports of ldquoOtherrdquo countries to markets other than the United Kingdom arehigher under Brexit This result too is due to changes in the relative prices of differ-ent qualities of wine in global wine markets

Figure 9

Difference in 2025 Wine Import Volumes and Values as a Result of the ldquoLargerdquo Initial BrexitShock (ML and US$ Million in 2014 US dollars)

Fig9-BW

onlin

eBW

inprint

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 241

B Subsequent Impact of Brexit from a UK-EU27 FTA

The next-most-likely step in the Brexit process is for the United Kingdom to nego-tiate a new trade arrangement with the EU27 We therefore assume that a UK-EU27FTAwith free bilateral wine trade is implemented and adjusted to by 2025 and thatprogress toward that end occurs soon enough that the adverse macroeconomicshocks from the initial impact of uncertainty over the Brexit process are confinedto those assumed in the ldquosmallrdquo scenario outlined above In this subsequent scenariowe assume that the British pound returns to the value it would have reached in 2025in the absence of Brexit (ie reversing the 10 devaluation assumed in the ldquosmallrdquoinitial Brexit scenario) and that real incomes in the United Kingdom are 8 higher

Table 5Difference in 2025 BilateralWine Import Volumes and Values fromKey Exporters by the United

Kingdom and the Rest of the World (RoW) as a Result of Initial Brexit Shock(ML and 2014US$ Million)a

(a) ldquolargerdquo scenario

Volume (ML) Value (2014US$ million)

UK RoW WORLD () UK RoW WORLD ()

EU27 minus287 136 minus150 (ndash17) minus1187 minus5 minus1192 (ndash31)Chile minus59 35 minus25 (ndash30) minus169 31 minus138 (ndash48)South Africa minus53 35 minus18 (ndash32) minus105 20 minus85 (ndash67)USA minus7 minus6 minus13 (ndash24) minus75 minus40 minus115 (ndash50)Australia minus4 minus3 minus7 (ndash09) minus25 minus65 minus90 (ndash30)Argentina minus3 minus9 minus12 (ndash48) minus16 minus39 minus55 (ndash52)New Zealand minus11 9 minus2 (ndash09) minus162 71 minus91 (ndash43)Others minus2 minus10 minus12 (ndash02) minus11 minus52 minus63 (ndash44)WORLD minus427 187 minus240 (ndash19) minus1750 minus79 minus1829 (ndash35)

(a) ldquosmallrdquo scenario

Volume (ML) Value (2014US$ million)

UK RoW WORLD () UK RoW WORLD ()

EU27 minus178 82 minus96 (ndash12) minus692 minus43 minus736 (ndash19)Chile minus46 28 minus18 (ndash24) minus128 36 minus91 (ndash32)South Africa minus43 29 minus14 (ndash42) minus82 23 minus59 (ndash47)USA 1 minus6 minus5 (ndash11) minus23 minus28 minus51 (ndash22)Australia 5 minus10 minus5 (ndash06) 19 minus56 minus38 (ndash13)Argentina 0 minus6 minus6 (ndash26) minus3 minus25 minus29 (ndash27)New Zealand minus5 4 minus1 (ndash06) minus80 34 minus46 (ndash22)Others 0 minus9 minus9 (ndash01) minus1 minus33 minus34 (ndash24)WORLD minus266 112 minus154 (ndash13) minus991 minus92 minus1083 (ndash21)

a Numbers in parentheses are the percentage difference between the Brexit and baseline scenarios for 2025 projected wine-import volumes orvalues by source

Source Authorsrsquo model results

242 UK and Global Wine Markets by 2025 and Implications of Brexit

in 2025 than in the ldquosmallrdquo scenario ndash that is they regain two-thirds of the value theywould have reached without Brexit-related uncertainty

This subsequent development in the Brexit process reverses most of but not allthe initial effects of Brexit by 2025 because of our assumption that the lost growthin the initial years of uncertainty following the Brexit vote are only partly recoveredby 2025 following the implementation of a UK-EU27 FTA Moreover the longer ittakes before this FTA is finalized and implemented the longer the estimated initialadverse effects persist and the larger the cumulative cost of Brexit to UK wine con-sumers and to grape and wine producers in wine-exporting countries

Table 6 summarizes the subsequent trade effects for 2025 It suggests that all butone-tenth of the loss in value of world trade in wine from the initial ldquosmallrdquo impact isrestored most of it because of a smaller reduction in the United Kingdomrsquos wineimports following the FTA Most of that improved outcome is because of recoveredimports from the EU27 commensurate with the latterrsquos high share of UK importsEven though Chile and South Africa are assumed in this scenario to have not yetsigned an FTA with the United Kingdom they export slightly more to the UnitedKingdom (and even more to all other countries) than in the initial ldquosmallrdquo scenario

Within the United Kingdom this FTA brings down the local currency consumerprice of wine by 9 largely offsetting the 11 rise in the initial ldquosmallrdquo Brexit sce-nario and it raises the volume consumed in the United Kingdom by 18 fully off-setting the 17 fall in the initial ldquosmallrdquo Brexit scenario

Table 6Difference in 2025 Bilateral Wine-Import Volumes and Values from Key Exporters by the

United Kingdom and the Rest of the World (RoW) as a Result of Implementing a UK-EU27FTA Following Initial ldquoSmallrdquo Brexit Shock (ML and 2014US$ Million Difference Relative to

Initial ldquoSmallrdquo Brexit Scenario)a

Volume (ML) Value (2014US$ million)

UK RoW WORLD (ldquosmallrdquo)a UK RoW WORLD (ldquosmallrdquo)a

EU27 212 minus112 100 (minus96) 750 0 750 (minus806)Chile 1 3 4 (minus18) 3 40 43 (minus92)South Africa 3 2 5 (minus14) 3 18 21 (minus60)USA 3 3 6 (minus5) 26 21 47 (minus50)Australia 1 1 2 (minus5) 8 38 46 (minus33)Argentina 1 4 5 (minus6) 4 20 24 (minus28)New Zealand 4 minus3 1 (minus1) 61 minus25 36 (minus46)Others 1 6 7 (minus9) 5 minus46 minus41 (92)WORLD 226 minus96 130 (minus154) 860 66 926 (minus1022)

a Numbers in parentheses are the world trade differences between the ldquosmallrdquo initial Brexit scenario and the baseline scenario copied fromcolumns 3 and 7 of Table 5(b)

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 243

In short this subsequent step in the Brexit process can be expected to restore muchof the initial adverse effects in the United Kingdom of the Brexit vote by 2025 Thewine exporters in the EU27 Chile and South Africa also lose less in this scenarioas compared with the initial ldquosmallrdquo scenario while other key exporters sell nearlyas much in 2025 in this scenario as they are projected to in the baseline ndash that isthey are beneficiaries of the reduced discrimination in global wine markets in thiscase where Chile and South Africa no longer enjoy preferential access to the UKmarket (and even though EU27 producers are again allowed to do so) Keep inmind though that although the estimated losses in the initial scenarios are reportedjust for 2025 they are expected to be felt throughout the preceding years of uncertaintyTherefore the longer it is before uncertainty abates and the United Kingdomrsquos currentpreferential trading arrangements are replaced by newones such as a UK-EU27 FTAthe higher the cumulative cost of the Brexit vote to the wine trade

VII Caveats and Conclusions

The above Brexit simulations are just a few of many scenarios that could be modeledObvious additional ones could also assume that FTAs are reached between theUnited Kingdom and other trading partners including Chile and South Africaand such countries as Australia and New Zealand Some of these other countriesare already exploring the ECrsquos invitation (see European Commission 2015) to nego-tiate an FTA with what will be the EU27 The sequence in which FTAs are signedand the speed with which they are implemented will matter (as was also the casewith the sequential signing over the past decade of bilateral FTAs with NortheastAsian countries by Chile Australia and New Zealand see Anderson andWittwer 2015) Even if the United Kingdom were able to sign additional bilateralFTAs and begin implementing them before 2025 it would have little effect on theabove results (because wine tariffs are such a minor contributor to them) unlessthose FTAs were to accelerate the United Kingdomrsquos economic growth and therise in the British poundrsquos value before 2025

We assume above that no changes are made to alcohol excise duties in the UnitedKingdom following Brexit even though they are scheduled to be progressively raisedwith inflation and may be raised even more for wine relative to spirits to offset theopposite effects of Brexit on those two domestic industries Nor do we make anyallowance for UK increases in consumer wine prices that may be needed to coverthe higher cost of clearing customs on imports from EU27 countries (which maybe more or less offset by increased smuggling and duty-free purchasing)Importers of fine wines including individuals who buy Bordeaux en primeur maywell reduce their demand because of concerns that delays on docks will affect thequality of their wines or because of greater currency uncertainty

To summarize Brexit is costly initially to UK consumers of wine (and of manyother tradable products) because the domestic retail price in local currency tax-

244 UK and Global Wine Markets by 2025 and Implications of Brexit

inclusive terms is 22 higher than otherwise in 2025 and the volume of wine con-sumed domestically is 28 lower in the ldquolargerdquo Brexit scenario (or 11 higher pricesand 17 lower quantities in the ldquosmallrdquo case) by 2025 unless a UK-EU27 FTA isdeveloped before then Even if such an FTA does get signed ratified by all 28 par-liaments and implemented by 2025 the slower income growth in the interim meansa smaller UK wine market in 2025 than would otherwise have been the case Thevolume reduction is a blow to many participants in UK wine-bottling transportingstoring wholesaling and retailing businesses in addition to restaurants and pubsVery little of that initial impact occurs because of higher import tariffs most impor-tant is the assumed fall in UK real incomes

The small but growing number of local UK vignerons (who supply less than 05of all domestic-wine sales and less than 4 of sparkling-wine sales) is initially pro-jected to sell less wine domestically because of reduced demand and to export only alittle extra abroad while the British pound is devalued Overall their production isonly 3 higher in 2025 even in the ldquolargerdquo Brexit scenario (not taking intoaccount the fact that their casual labor is likely to be more expensive under Brexitbecause of tighter restrictions on immigration) and their production diminishes ifand when a UK-EU27 FTA is in place

Even with a UK-EU27 FTA in place EU27 wine exporters export US$56 millionless wine in 2025 thanks to Brexit Chile and South Africa export $88 million lesswine and wine exports from the rest of the world are little different in aggregatealthough slightly larger from Australia

Clearly there will be great uncertainly for some time over the possible policy out-comes to flow from Brexit and their consequent sequential impacts on UK house-hold disposable incomes foreign exchange rates and bilateral wine tariffsMeanwhile the above projections under explicit assumptions provide some idea ofhow wine markets might be affected by the most-likely first two stages of theBrexit process (agreeing on a new tariff schedule at the WTO and agreeing to andimplementing a UK-EU27 FTA) In particular they make clear that nontrivialinitial impacts could affect the domestic wine market impacts that are likely to belarger than just the direct impact of changes in bilateral tariffs If the UnitedKingdom succeeds in getting countries to agree to transition arrangements thatdelay the changes in tariffs until new FTAs are signed and ratified the initialeffects will be less dramatic than in our first scenarios but it remains to be seenwhether any such agreements can be reached In any event the net effect of Brexiton the welfare of the worldrsquos consumers and producers of wine as a whole will benegative not just initially but permanently unless new trade policy commitmentsby the United Kingdom with major wine-exporting countries are sufficiently moreliberal than current arrangements

Kym Anderson and Glyn Wittwer 245

Appendix Key Equations in the Global Wine Markets Model

Amodel of the worldrsquos wine markets was first published byWittwer et al (2003) andmuch improved by Anderson and Wittwer (2013) Several significant enhancementshave been made to that original model Wine types have been disaggregated intomore types ndash namely nonpremium (including generic bulk) commercial-premiumand superpremium still wines and sparkling wines8 As in the original modelthere are two types of grapes premium and nonpremium Nonpremium wines usenonpremium grapes exclusively superpremium wines use mainly premium grapesand commercial-premium and sparkling wines use both types of grapes In theregional dimension the number of countries and country groups has expandedfrom 10 in the original model to 51 44 individual nations and 7 compositeregions The modelrsquos database is calibrated to 2014 for this paper based on datain Anderson et al (2017) trade data for which are downloaded from httpscomtradeunorg The modelrsquos supply-and-demand equations are based on the orig-inal ORANI modelrsquos theory (see Dixon Parmenter Sutton and Vincent 1982) andthe model is implemented using GEMPACK software (Harrison Horridge Jerieand Pearson 2014)

An enhancement of importance to the present study is the inclusion of exchange-rate variables in the model which allow a distinction between price impacts asobserved in local currency units and those observed in 2014 US dollars

In the model the grape and wine sectors minimize costs of intermediate inputssubject to weak constant elasticity of substitution (CES) substitutability betweeninputs By assumption no intermediate inputs are imported from other countries9Hence

Xcid frac14 f ethX1id CESfrac12P

cid=P1id $THORN eth1THORN

P1id X1id frac14X

c

Xcid P

cid eth2THORN

whereXcid is the quantity demanded of commodity c by grape or wine industry i in

region d Pcid is the corresponding price and X1id and P1id are the respective inter-

mediate composite quantities and prices

Two primary factors are employed in the sector labor (the quantity of which isendogenous with perfectly elastic supply) and capital Capital is usually treated asexogenous in quantity with rates of return bearing all the adjustment in thevarious scenarios This reflects the fact that grapes (a perennial crop) and wine-

8Commercial-premium still wines are defined by Anderson and Nelgen (2011) to be those priced betweenUS$250 and $750 per litre pre-tax at a countryrsquos border or wholesale9An exception concerns bulk wine imports used mainly in the bottling of commercial premium wine in theUK USA Canada Japan and Germany

246 UK and Global Wine Markets by 2025 and Implications of Brexit

plant capacity adjust slowly to market signals

Lid frac14 f ethFid CESfrac12Wid=PFid $THORN eth3THORN

Kid frac14 f ethFid CESethRid=PFid $THORN eth4THORN

PFid Fid frac14 Lid Wid thorn Kid Rid eth5THORN

Grape and wine producers are assumed to minimize costs subject to CES substitu-tion between capital and labor Equations (3) to (5) show primary factor demandsfor the labor composite Lid and capital Kid subject to a composite factor demandFid by industry i in region d The factor prices are Wid for labor Rid for capitalrentals and PFid for composite factor prices

Appendix Table A1Cumulative Consumption and Population Growth Rates and Changes in the Real Exchange Rate

(RER) Relative to the US Dollar 2014 to 2025 without Brexit ()

Aggregateconsumption Poprsquon RER

Aggregateconsumption Poprsquon RER

France 18 4 minus11 Australia 35 11 minus17Italy 11 2 minus9 New Zealand 32 9 minus26Portugal 14 0 minus9 Canada 27 8 minus18Spain 26 8 minus9 United States 31 8 0Austria 19 4 minus7 Argentina 7 10 109Belgium 20 7 minus9 Brazil 16 8 minus29Denmark 22 2 minus9 Chile 55 8 minus2Finland 21 3 minus7 Mexico 42 12 minus8Germany 14 minus2 minus11 Uruguay 45 3 1Greece 22 minus1 minus14 Other Latin America 60 10 minus5Ireland 42 12 minus9 South Africa 36 12 minus1Netherlands 21 4 minus9 Turkey 50 8 20Sweden 24 9 minus13 North Africa 53 11 0Switzerland 18 8 minus6 Other Africa 109 18 84United Kingdom 32 6 1 Middle East 52 18 minus12Other West Europe 21 10 minus1 China 79 3 5Bulgaria 41 minus7 7 Hong Kong 42 3 2Croatia 20 minus2 minus1 India 134 13 17Georgia 35 0 23 Japan 11 minus3 minus24Hungary 25 minus3 minus11 Korea 38 1 minus9Moldova 49 minus11 13 Malaysia 62 15 minus16Romania 45 minus4 22 Philippines 75 18 7Russia 18 minus2 minus8 Singapore 44 21 minus22Ukraine 22 minus5 14 Taiwan 29 1 minus13Other East Europe 40 minus5 48 Thailand 47 3 minus9

Other Asia 99 10 10

Source Authorsrsquo compilation from projections by various international agencies and from global economy-wide modeling by Anderson andStrutt (2016)

Kym Anderson and Glyn Wittwer 247

The composite factor demand Fid is proportional to total output Qid subject to aprimary factor using technology Aid Hence

Fid frac14 Qid Aid eth6THORN

The perfectly competitive zero-pure-profit condition is that total revenue valued atthe output price P0s

i multiplied by Qid equals the total production cost

P0si Qid frac14

X

cPcid X1cid thorn

X

oWo

id Loid thorn Rid Kid eth7THORN

Household demands follow a linear expenditure system in each region We reducethe optimizing problem for household consumption of each commodity subjectto a budget constraint to equations describing subsistence and discretionarydemands Aggregate subsistence expenditure WSUBd depends only on consumerprices P3cd for each commodity and the number of households N as per-capita sub-sistence quantities XSUBcd subject to given preferences are constant

WSUBd frac14X

c

P3cd XSUBcd Nd eth8THORN

Discretionary expenditures for each commodity (the left-hand side of equation (9))are equal to the marginal budget share (βcd) of aggregate discretionary expenditureThis aggregate is the bracketed term on the right-hand side of equation (9) whereW3TOTd is aggregate nominal expenditure

P3cdethX3cd XSUBcd NdTHORN frac14 βcdethW3TOTd WSUBdTHORN eth9THORN

Because real aggregate consumption is usually exogenous in our partial equilibriumsimulations the linear expenditure system determines the consumption shares ofindividual final commodities (ie the five wine types plus a composite of all otherconsumption items) driven by changes in relative prices as faced by domestic con-sumers The income elasticity of demand for each commodity is equal to the mar-ginal budget share divided by the expenditure share This number varies from 05for nonpremium wine to 17 for superpremium still wine The income elasticity ofdemand for other consumption is very close to 10 because wine accounts for anaverage of only 03 of aggregate expenditures globally and no more than 11in any country (Anderson and Nelgen 2011 Table 166)

A feature of our revised model of world wine markets is the inclusion of nominalexchange rates These appear directly in the equation linking retail prices (P3scd ) toproducer prices by country of origin (P0s

c ) where c denotes the wine type

P3 scd frac14 P0s

cfd

fsTtarcd T

taxcd thorn Pm

cd eth10THORN

248 UK and Global Wine Markets by 2025 and Implications of Brexit

The exchange rates in the consuming and producing regions are fd and fs respec-tively expressed as local currency units per $US Ttar

cd is the power of the tariff in theconsuming region and Ttax

cd is the power of the domestic consumption (or excise) taxprior to any generic value-added or goods-and-services tax Pm

cd is the price of marginm assumed to be locally supplied nontradable and therefore unaffected by theexchange rate

A given level of consumption for wine type c (X3cd) is satisfied using theArmington (1969) assumption in which wines from different countries of originare imperfectly substitutable First domestic wine is imperfectly substitutable witha composite of imports

X3sscd frac14 f ethX3cd CESethP3sscd=P3cdTHORNTHORN ss frac14 domestic imports eth11THORN

and then imports by origin (X3scd ) are determined in a second CES equation

X3scd frac14 f ethX3ssfrac14importscd CESethP3scd=P3

ssfrac14importscd THORNTHORN eth12THORN

The model enables us to show how changes in international competitiveness affectthe worldrsquos wine markets A crucial part of this exercise is explaining how pricesdetermined outside the grape and wine markets influence these markets Becausethe model is in partial equilibrium to depict the impacts of changes in internationalcompetitiveness outside price changes need to be imposed as shocks on the modelThe price of intermediate inputs other than grapes and wine shown in equations (1)and (2) is set equal to the price of GDP (Pg

d ) multiplied by a shifter Fcd

Pcid frac14 Fc

dPgd eth13THORN

If no specific price observations are available the shifter Fcd remains exogenous and

unshocked with the change in price being determined by a shock to the price ofGDP If observations are available for specific input price movements the shifterFc

dbecomes endogenous with Pc

id now exogenous and shocked

Wid frac14 Fwd P

gd eth14THORN

Wage rates are treated similarly In equation (14) if the wage shifter Fwd is exogenous

changes in wage rates Wid are determined by changes in the price of GDP If wagerate data are available Fw

d becomes endogenous andwage rates are shocked directly

Pmcd frac14 Fm

d Pgd eth15THORN

The prices of trade and transport margins are also determined by the price of GDP ifthe shifter Fm

d in equation (15) is exogenous

Changes in international competitiveness depend on changes in relative pricelevels and changes in nominal exchange rates In equation (16) fR

s denotes real

Kym Anderson and Glyn Wittwer 249

exchange-rate movements relative to the US dollar

fRs frac14 Pg

s =frac12PgUSA

fs$ eth16THORN

In equation (16) the nominal exchange rate for the United States is alwaysunchanged because nominal and real exchange rates are expressed relative to theUS currency

Changes in international market conditions may have impacts in one direction onproducer prices as expressed in US dollars P0s

i and potentially in the opposite direc-tion in local currency units Hence we calculate real producer pricesP0s

iloc in localcurrency terms (ie the price most relevant to domestic producers)

P0siloc frac14 P0s

ifs=P

gs eth17THORN

To obtain real price changes in local currency terms we convert US dollar prices(P3 s

cd for the source-specific price and P3cd for the source-composite price) to reallocal currency prices (P3 s

cdloc and P3cdloc) using the consumer price index (CPI)(Pc

d ) as the deflator

P3 scdloc frac14 P3 s

cdfd=P

cd eth18THORN

and

P3cdloc frac14 P3cdfd=Pcd eth19THORN

References

Anderson K and Nelgen S (2011) Global Wine Markets 1961 to 2009 A StatisticalCompendium Adelaide University of Adelaide Press Also available as an e-book atwwwadelaideeduaupresstitlesglobal-wine

Anderson K Nelgen S and Pinilla V (2017) Global Wine Markets 1860 to 2016 AStatistical Compendium Adelaide University of Adelaide Press (forthcoming) Also tobe available as an e-book at wwwadelaideeduaupress

Anderson K and Pinilla V (with the assistance of A J Holmes) (2017)Annual Database ofGlobal Wine Markets 1835 to 2016 Wine Economics Research Centre University ofAdelaide to be posted in October at wwwadelaideeduauwine-econdatabases

Anderson K and Strutt A (2016) Impacts of Asiarsquos rise on African and Latin Americantrade Projections to 2030 World Economy 39(2) 172ndash194

Anderson K andWittwer G (2013) Modeling global wine markets to 2018 Exchange ratestaste changes and Chinarsquos import growth Journal of Wine Economics 8(2) 131ndash158

Anderson K and Wittwer G (2015) Asiarsquos evolving role in global wine markets ChinaEconomic Review 35 1ndash14

250 UK and Global Wine Markets by 2025 and Implications of Brexit

Armington P A (1969) A theory of demand for products distinguished by place of produc-tion IMF Staff Papers 16 159ndash178

Baldwin R E (ed) (2016) Brexit Beckons Thinking Ahead by Leading Economists LondonCentre for Economic Policy Research

Campos N F (2016) Lousy experts Looking back at the ex ante estimates of the costs ofBrexit In R E Baldwin (ed) Brexit Beckons Thinking Ahead by Leading Economists(ch 3) London Centre for Economic Policy Research

Dixon P Parmenter B Sutton J and Vincent D (1982) ORANI A Multisectoral Modelof the Australian Economy Contributions to Economic Analysis 142 Amsterdam North-Holland

European Commission (2015) Trade for All Towards a More Responsible Trade andInvestment Policy Brussels European Commission

Harrison J Horridge M Jerie M and Pearson K (2014) GEMPACK ManualMelbourne GEMPACK Software

HM Treasury (2017) Spring Budget 2017 London HM TreasuryHolmes A J and Anderson K (2017a)Annual Database of National Beverage Consumption

Volumes and Expenditures 1950 to 2015 Wine Economics Research Centre University ofAdelaide posted at wwwadelaideeduauwine-econdatabases

Holmes A J and Anderson K (2017b) Convergence in National Alcohol ConsumptionPatterns New Global Indicators Journal of Wine Economics 12(2) 117ndash148

Ludington C C (2013) The Politics of Wine in Britain A New Cultural History BasingstokeUK Palgrave Macmillan

Office of National Statistics (2017) Analysis of real earnings London Office of NationalStatistics Available at httpswwwonsgovukemploymentandlabourmarketpeopleinworkearningsandworkinghoursarticlessupplementaryanalysisofaverageweeklyearningslatest

Rollo J Borchert I Dawar K Holmes P and Winters L A (2016) The World TradeOrganisation A safety net for a post-Brexit UK trade policy Briefing Paper 1UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-1-final-1pdf

Smith A (2017) Brexit Hard truths and hard choices UKTPO blog 19 June Available athttpsblogssussexacukuktpo20170619brexit-hard-truths-and-hard-choices

Swinbank A (2017) World trade rules and the policy options for British agriculture post-Brexit Briefing Paper 7 UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-7pdf

Viner J (1950) The Customs Union Issue New York Carnegie Endowment for InternationalPeace

Wine Australia (2015) Export Market Guide European Union Adelaide Wine AustraliaWittwer G Berger N and Anderson K (2003) A model of the worldrsquos wine markets

Economic Modelling 20(3) 487ndash506

Kym Anderson and Glyn Wittwer 251

Page 20: U.K. and Global Wine Markets by 2025, and Implications of ... · U.K. and Global Wine Markets by 2025, and Implications of Brexit* Kym Andersona and Glyn Wittwerb Abstract The United

The aggregate effect of ldquolargerdquo Brexit on the market shares of various wine-export-ing countries in the United Kingdom is almost indiscernible even with one decimalpoint The projected 2025 shares are quite different from the actual 2014 shares forseveral countries They are much smaller in 2025 for South Africa Australia andNew Zealand (and the United States in volume terms) and are much larger involume for Spain and in value for Italy These results occur because wine-exportingcountries benefit differentially from the varying rates of growth in net importdemand for wine in non-UK countries over this projection period The most impor-tant projected changes are the increase in the real value of annual wine importsbetween 2014 and 2025 by China (200 or $3 billion) Other Asia (110 or $22billion) and Africa (270 or $16 billion) More than half of Australiarsquos increasein annual exports from 2014 to 2025 go to Asia and more than half of SouthAfricarsquos increase in exports go to Other Africa

Table 5 reveals that ldquolargerdquo Brexit lowers EU Chilean and South African wineexports by 150 ML or US$12 billion in the case of the European Union with someof their exports diverted from the United Kingdom to the EU27 and other marketsin competition with New World exporters Although the United States Australiaand Argentina sell only a little less to the United Kingdom they also sell less toother countries For Chile and South Africa which lose their preferential access toUK (but not to EU27) markets in this Brexit scenario some exports are redirected

Figure 8

UK Shares of World Wine Imports 2010ndash2015 and Projected to 2025 without and with theldquoLargerdquo Initial Brexit Shock ()

Fig8-BW

onlin

eBW

inprint

Sources Anderson and Pinilla (2017) and authorsrsquo model results

240 UK and Global Wine Markets by 2025 and Implications of Brexit

from the United Kingdom to EU27 countries ndash but again they export less overallGlobal wine trade in 2025 is 240 ML (19) or $18 billion (35) less under thisldquolargerdquo Brexit scenario The percentage by which wine exportersrsquo trade shrinks isgreater for values than for volumes because of changes in relative prices of differ-ent-quality wines Those differences are shown in the numbers in parentheses inTable 5

Three other points are worth making about Table 5 One is that Australia sellsslightly more to the United Kingdom in the ldquosmallrdquo Brexit scenario rather thanslightly less as in the ldquolargerdquo Brexit case Evidently the negative income and price(devaluation) effects do not more than offset the positive trade-diverting effect onAustralian exports to the United Kingdom of removing preferences in the ldquosmallrdquoscenario Second New Zealand sells slightly more to non-UK countries underBrexit despite greater competition from the EU27 Chile and South Africa Thisanomaly is due to changes in the relative prices of different qualities of wine inglobal wine markets bearing in mind that New Zealand has the worldrsquos highestaverage price for still-wine exports And third the value (but not the volume) ofexports of ldquoOtherrdquo countries to markets other than the United Kingdom arehigher under Brexit This result too is due to changes in the relative prices of differ-ent qualities of wine in global wine markets

Figure 9

Difference in 2025 Wine Import Volumes and Values as a Result of the ldquoLargerdquo Initial BrexitShock (ML and US$ Million in 2014 US dollars)

Fig9-BW

onlin

eBW

inprint

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 241

B Subsequent Impact of Brexit from a UK-EU27 FTA

The next-most-likely step in the Brexit process is for the United Kingdom to nego-tiate a new trade arrangement with the EU27 We therefore assume that a UK-EU27FTAwith free bilateral wine trade is implemented and adjusted to by 2025 and thatprogress toward that end occurs soon enough that the adverse macroeconomicshocks from the initial impact of uncertainty over the Brexit process are confinedto those assumed in the ldquosmallrdquo scenario outlined above In this subsequent scenariowe assume that the British pound returns to the value it would have reached in 2025in the absence of Brexit (ie reversing the 10 devaluation assumed in the ldquosmallrdquoinitial Brexit scenario) and that real incomes in the United Kingdom are 8 higher

Table 5Difference in 2025 BilateralWine Import Volumes and Values fromKey Exporters by the United

Kingdom and the Rest of the World (RoW) as a Result of Initial Brexit Shock(ML and 2014US$ Million)a

(a) ldquolargerdquo scenario

Volume (ML) Value (2014US$ million)

UK RoW WORLD () UK RoW WORLD ()

EU27 minus287 136 minus150 (ndash17) minus1187 minus5 minus1192 (ndash31)Chile minus59 35 minus25 (ndash30) minus169 31 minus138 (ndash48)South Africa minus53 35 minus18 (ndash32) minus105 20 minus85 (ndash67)USA minus7 minus6 minus13 (ndash24) minus75 minus40 minus115 (ndash50)Australia minus4 minus3 minus7 (ndash09) minus25 minus65 minus90 (ndash30)Argentina minus3 minus9 minus12 (ndash48) minus16 minus39 minus55 (ndash52)New Zealand minus11 9 minus2 (ndash09) minus162 71 minus91 (ndash43)Others minus2 minus10 minus12 (ndash02) minus11 minus52 minus63 (ndash44)WORLD minus427 187 minus240 (ndash19) minus1750 minus79 minus1829 (ndash35)

(a) ldquosmallrdquo scenario

Volume (ML) Value (2014US$ million)

UK RoW WORLD () UK RoW WORLD ()

EU27 minus178 82 minus96 (ndash12) minus692 minus43 minus736 (ndash19)Chile minus46 28 minus18 (ndash24) minus128 36 minus91 (ndash32)South Africa minus43 29 minus14 (ndash42) minus82 23 minus59 (ndash47)USA 1 minus6 minus5 (ndash11) minus23 minus28 minus51 (ndash22)Australia 5 minus10 minus5 (ndash06) 19 minus56 minus38 (ndash13)Argentina 0 minus6 minus6 (ndash26) minus3 minus25 minus29 (ndash27)New Zealand minus5 4 minus1 (ndash06) minus80 34 minus46 (ndash22)Others 0 minus9 minus9 (ndash01) minus1 minus33 minus34 (ndash24)WORLD minus266 112 minus154 (ndash13) minus991 minus92 minus1083 (ndash21)

a Numbers in parentheses are the percentage difference between the Brexit and baseline scenarios for 2025 projected wine-import volumes orvalues by source

Source Authorsrsquo model results

242 UK and Global Wine Markets by 2025 and Implications of Brexit

in 2025 than in the ldquosmallrdquo scenario ndash that is they regain two-thirds of the value theywould have reached without Brexit-related uncertainty

This subsequent development in the Brexit process reverses most of but not allthe initial effects of Brexit by 2025 because of our assumption that the lost growthin the initial years of uncertainty following the Brexit vote are only partly recoveredby 2025 following the implementation of a UK-EU27 FTA Moreover the longer ittakes before this FTA is finalized and implemented the longer the estimated initialadverse effects persist and the larger the cumulative cost of Brexit to UK wine con-sumers and to grape and wine producers in wine-exporting countries

Table 6 summarizes the subsequent trade effects for 2025 It suggests that all butone-tenth of the loss in value of world trade in wine from the initial ldquosmallrdquo impact isrestored most of it because of a smaller reduction in the United Kingdomrsquos wineimports following the FTA Most of that improved outcome is because of recoveredimports from the EU27 commensurate with the latterrsquos high share of UK importsEven though Chile and South Africa are assumed in this scenario to have not yetsigned an FTA with the United Kingdom they export slightly more to the UnitedKingdom (and even more to all other countries) than in the initial ldquosmallrdquo scenario

Within the United Kingdom this FTA brings down the local currency consumerprice of wine by 9 largely offsetting the 11 rise in the initial ldquosmallrdquo Brexit sce-nario and it raises the volume consumed in the United Kingdom by 18 fully off-setting the 17 fall in the initial ldquosmallrdquo Brexit scenario

Table 6Difference in 2025 Bilateral Wine-Import Volumes and Values from Key Exporters by the

United Kingdom and the Rest of the World (RoW) as a Result of Implementing a UK-EU27FTA Following Initial ldquoSmallrdquo Brexit Shock (ML and 2014US$ Million Difference Relative to

Initial ldquoSmallrdquo Brexit Scenario)a

Volume (ML) Value (2014US$ million)

UK RoW WORLD (ldquosmallrdquo)a UK RoW WORLD (ldquosmallrdquo)a

EU27 212 minus112 100 (minus96) 750 0 750 (minus806)Chile 1 3 4 (minus18) 3 40 43 (minus92)South Africa 3 2 5 (minus14) 3 18 21 (minus60)USA 3 3 6 (minus5) 26 21 47 (minus50)Australia 1 1 2 (minus5) 8 38 46 (minus33)Argentina 1 4 5 (minus6) 4 20 24 (minus28)New Zealand 4 minus3 1 (minus1) 61 minus25 36 (minus46)Others 1 6 7 (minus9) 5 minus46 minus41 (92)WORLD 226 minus96 130 (minus154) 860 66 926 (minus1022)

a Numbers in parentheses are the world trade differences between the ldquosmallrdquo initial Brexit scenario and the baseline scenario copied fromcolumns 3 and 7 of Table 5(b)

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 243

In short this subsequent step in the Brexit process can be expected to restore muchof the initial adverse effects in the United Kingdom of the Brexit vote by 2025 Thewine exporters in the EU27 Chile and South Africa also lose less in this scenarioas compared with the initial ldquosmallrdquo scenario while other key exporters sell nearlyas much in 2025 in this scenario as they are projected to in the baseline ndash that isthey are beneficiaries of the reduced discrimination in global wine markets in thiscase where Chile and South Africa no longer enjoy preferential access to the UKmarket (and even though EU27 producers are again allowed to do so) Keep inmind though that although the estimated losses in the initial scenarios are reportedjust for 2025 they are expected to be felt throughout the preceding years of uncertaintyTherefore the longer it is before uncertainty abates and the United Kingdomrsquos currentpreferential trading arrangements are replaced by newones such as a UK-EU27 FTAthe higher the cumulative cost of the Brexit vote to the wine trade

VII Caveats and Conclusions

The above Brexit simulations are just a few of many scenarios that could be modeledObvious additional ones could also assume that FTAs are reached between theUnited Kingdom and other trading partners including Chile and South Africaand such countries as Australia and New Zealand Some of these other countriesare already exploring the ECrsquos invitation (see European Commission 2015) to nego-tiate an FTA with what will be the EU27 The sequence in which FTAs are signedand the speed with which they are implemented will matter (as was also the casewith the sequential signing over the past decade of bilateral FTAs with NortheastAsian countries by Chile Australia and New Zealand see Anderson andWittwer 2015) Even if the United Kingdom were able to sign additional bilateralFTAs and begin implementing them before 2025 it would have little effect on theabove results (because wine tariffs are such a minor contributor to them) unlessthose FTAs were to accelerate the United Kingdomrsquos economic growth and therise in the British poundrsquos value before 2025

We assume above that no changes are made to alcohol excise duties in the UnitedKingdom following Brexit even though they are scheduled to be progressively raisedwith inflation and may be raised even more for wine relative to spirits to offset theopposite effects of Brexit on those two domestic industries Nor do we make anyallowance for UK increases in consumer wine prices that may be needed to coverthe higher cost of clearing customs on imports from EU27 countries (which maybe more or less offset by increased smuggling and duty-free purchasing)Importers of fine wines including individuals who buy Bordeaux en primeur maywell reduce their demand because of concerns that delays on docks will affect thequality of their wines or because of greater currency uncertainty

To summarize Brexit is costly initially to UK consumers of wine (and of manyother tradable products) because the domestic retail price in local currency tax-

244 UK and Global Wine Markets by 2025 and Implications of Brexit

inclusive terms is 22 higher than otherwise in 2025 and the volume of wine con-sumed domestically is 28 lower in the ldquolargerdquo Brexit scenario (or 11 higher pricesand 17 lower quantities in the ldquosmallrdquo case) by 2025 unless a UK-EU27 FTA isdeveloped before then Even if such an FTA does get signed ratified by all 28 par-liaments and implemented by 2025 the slower income growth in the interim meansa smaller UK wine market in 2025 than would otherwise have been the case Thevolume reduction is a blow to many participants in UK wine-bottling transportingstoring wholesaling and retailing businesses in addition to restaurants and pubsVery little of that initial impact occurs because of higher import tariffs most impor-tant is the assumed fall in UK real incomes

The small but growing number of local UK vignerons (who supply less than 05of all domestic-wine sales and less than 4 of sparkling-wine sales) is initially pro-jected to sell less wine domestically because of reduced demand and to export only alittle extra abroad while the British pound is devalued Overall their production isonly 3 higher in 2025 even in the ldquolargerdquo Brexit scenario (not taking intoaccount the fact that their casual labor is likely to be more expensive under Brexitbecause of tighter restrictions on immigration) and their production diminishes ifand when a UK-EU27 FTA is in place

Even with a UK-EU27 FTA in place EU27 wine exporters export US$56 millionless wine in 2025 thanks to Brexit Chile and South Africa export $88 million lesswine and wine exports from the rest of the world are little different in aggregatealthough slightly larger from Australia

Clearly there will be great uncertainly for some time over the possible policy out-comes to flow from Brexit and their consequent sequential impacts on UK house-hold disposable incomes foreign exchange rates and bilateral wine tariffsMeanwhile the above projections under explicit assumptions provide some idea ofhow wine markets might be affected by the most-likely first two stages of theBrexit process (agreeing on a new tariff schedule at the WTO and agreeing to andimplementing a UK-EU27 FTA) In particular they make clear that nontrivialinitial impacts could affect the domestic wine market impacts that are likely to belarger than just the direct impact of changes in bilateral tariffs If the UnitedKingdom succeeds in getting countries to agree to transition arrangements thatdelay the changes in tariffs until new FTAs are signed and ratified the initialeffects will be less dramatic than in our first scenarios but it remains to be seenwhether any such agreements can be reached In any event the net effect of Brexiton the welfare of the worldrsquos consumers and producers of wine as a whole will benegative not just initially but permanently unless new trade policy commitmentsby the United Kingdom with major wine-exporting countries are sufficiently moreliberal than current arrangements

Kym Anderson and Glyn Wittwer 245

Appendix Key Equations in the Global Wine Markets Model

Amodel of the worldrsquos wine markets was first published byWittwer et al (2003) andmuch improved by Anderson and Wittwer (2013) Several significant enhancementshave been made to that original model Wine types have been disaggregated intomore types ndash namely nonpremium (including generic bulk) commercial-premiumand superpremium still wines and sparkling wines8 As in the original modelthere are two types of grapes premium and nonpremium Nonpremium wines usenonpremium grapes exclusively superpremium wines use mainly premium grapesand commercial-premium and sparkling wines use both types of grapes In theregional dimension the number of countries and country groups has expandedfrom 10 in the original model to 51 44 individual nations and 7 compositeregions The modelrsquos database is calibrated to 2014 for this paper based on datain Anderson et al (2017) trade data for which are downloaded from httpscomtradeunorg The modelrsquos supply-and-demand equations are based on the orig-inal ORANI modelrsquos theory (see Dixon Parmenter Sutton and Vincent 1982) andthe model is implemented using GEMPACK software (Harrison Horridge Jerieand Pearson 2014)

An enhancement of importance to the present study is the inclusion of exchange-rate variables in the model which allow a distinction between price impacts asobserved in local currency units and those observed in 2014 US dollars

In the model the grape and wine sectors minimize costs of intermediate inputssubject to weak constant elasticity of substitution (CES) substitutability betweeninputs By assumption no intermediate inputs are imported from other countries9Hence

Xcid frac14 f ethX1id CESfrac12P

cid=P1id $THORN eth1THORN

P1id X1id frac14X

c

Xcid P

cid eth2THORN

whereXcid is the quantity demanded of commodity c by grape or wine industry i in

region d Pcid is the corresponding price and X1id and P1id are the respective inter-

mediate composite quantities and prices

Two primary factors are employed in the sector labor (the quantity of which isendogenous with perfectly elastic supply) and capital Capital is usually treated asexogenous in quantity with rates of return bearing all the adjustment in thevarious scenarios This reflects the fact that grapes (a perennial crop) and wine-

8Commercial-premium still wines are defined by Anderson and Nelgen (2011) to be those priced betweenUS$250 and $750 per litre pre-tax at a countryrsquos border or wholesale9An exception concerns bulk wine imports used mainly in the bottling of commercial premium wine in theUK USA Canada Japan and Germany

246 UK and Global Wine Markets by 2025 and Implications of Brexit

plant capacity adjust slowly to market signals

Lid frac14 f ethFid CESfrac12Wid=PFid $THORN eth3THORN

Kid frac14 f ethFid CESethRid=PFid $THORN eth4THORN

PFid Fid frac14 Lid Wid thorn Kid Rid eth5THORN

Grape and wine producers are assumed to minimize costs subject to CES substitu-tion between capital and labor Equations (3) to (5) show primary factor demandsfor the labor composite Lid and capital Kid subject to a composite factor demandFid by industry i in region d The factor prices are Wid for labor Rid for capitalrentals and PFid for composite factor prices

Appendix Table A1Cumulative Consumption and Population Growth Rates and Changes in the Real Exchange Rate

(RER) Relative to the US Dollar 2014 to 2025 without Brexit ()

Aggregateconsumption Poprsquon RER

Aggregateconsumption Poprsquon RER

France 18 4 minus11 Australia 35 11 minus17Italy 11 2 minus9 New Zealand 32 9 minus26Portugal 14 0 minus9 Canada 27 8 minus18Spain 26 8 minus9 United States 31 8 0Austria 19 4 minus7 Argentina 7 10 109Belgium 20 7 minus9 Brazil 16 8 minus29Denmark 22 2 minus9 Chile 55 8 minus2Finland 21 3 minus7 Mexico 42 12 minus8Germany 14 minus2 minus11 Uruguay 45 3 1Greece 22 minus1 minus14 Other Latin America 60 10 minus5Ireland 42 12 minus9 South Africa 36 12 minus1Netherlands 21 4 minus9 Turkey 50 8 20Sweden 24 9 minus13 North Africa 53 11 0Switzerland 18 8 minus6 Other Africa 109 18 84United Kingdom 32 6 1 Middle East 52 18 minus12Other West Europe 21 10 minus1 China 79 3 5Bulgaria 41 minus7 7 Hong Kong 42 3 2Croatia 20 minus2 minus1 India 134 13 17Georgia 35 0 23 Japan 11 minus3 minus24Hungary 25 minus3 minus11 Korea 38 1 minus9Moldova 49 minus11 13 Malaysia 62 15 minus16Romania 45 minus4 22 Philippines 75 18 7Russia 18 minus2 minus8 Singapore 44 21 minus22Ukraine 22 minus5 14 Taiwan 29 1 minus13Other East Europe 40 minus5 48 Thailand 47 3 minus9

Other Asia 99 10 10

Source Authorsrsquo compilation from projections by various international agencies and from global economy-wide modeling by Anderson andStrutt (2016)

Kym Anderson and Glyn Wittwer 247

The composite factor demand Fid is proportional to total output Qid subject to aprimary factor using technology Aid Hence

Fid frac14 Qid Aid eth6THORN

The perfectly competitive zero-pure-profit condition is that total revenue valued atthe output price P0s

i multiplied by Qid equals the total production cost

P0si Qid frac14

X

cPcid X1cid thorn

X

oWo

id Loid thorn Rid Kid eth7THORN

Household demands follow a linear expenditure system in each region We reducethe optimizing problem for household consumption of each commodity subjectto a budget constraint to equations describing subsistence and discretionarydemands Aggregate subsistence expenditure WSUBd depends only on consumerprices P3cd for each commodity and the number of households N as per-capita sub-sistence quantities XSUBcd subject to given preferences are constant

WSUBd frac14X

c

P3cd XSUBcd Nd eth8THORN

Discretionary expenditures for each commodity (the left-hand side of equation (9))are equal to the marginal budget share (βcd) of aggregate discretionary expenditureThis aggregate is the bracketed term on the right-hand side of equation (9) whereW3TOTd is aggregate nominal expenditure

P3cdethX3cd XSUBcd NdTHORN frac14 βcdethW3TOTd WSUBdTHORN eth9THORN

Because real aggregate consumption is usually exogenous in our partial equilibriumsimulations the linear expenditure system determines the consumption shares ofindividual final commodities (ie the five wine types plus a composite of all otherconsumption items) driven by changes in relative prices as faced by domestic con-sumers The income elasticity of demand for each commodity is equal to the mar-ginal budget share divided by the expenditure share This number varies from 05for nonpremium wine to 17 for superpremium still wine The income elasticity ofdemand for other consumption is very close to 10 because wine accounts for anaverage of only 03 of aggregate expenditures globally and no more than 11in any country (Anderson and Nelgen 2011 Table 166)

A feature of our revised model of world wine markets is the inclusion of nominalexchange rates These appear directly in the equation linking retail prices (P3scd ) toproducer prices by country of origin (P0s

c ) where c denotes the wine type

P3 scd frac14 P0s

cfd

fsTtarcd T

taxcd thorn Pm

cd eth10THORN

248 UK and Global Wine Markets by 2025 and Implications of Brexit

The exchange rates in the consuming and producing regions are fd and fs respec-tively expressed as local currency units per $US Ttar

cd is the power of the tariff in theconsuming region and Ttax

cd is the power of the domestic consumption (or excise) taxprior to any generic value-added or goods-and-services tax Pm

cd is the price of marginm assumed to be locally supplied nontradable and therefore unaffected by theexchange rate

A given level of consumption for wine type c (X3cd) is satisfied using theArmington (1969) assumption in which wines from different countries of originare imperfectly substitutable First domestic wine is imperfectly substitutable witha composite of imports

X3sscd frac14 f ethX3cd CESethP3sscd=P3cdTHORNTHORN ss frac14 domestic imports eth11THORN

and then imports by origin (X3scd ) are determined in a second CES equation

X3scd frac14 f ethX3ssfrac14importscd CESethP3scd=P3

ssfrac14importscd THORNTHORN eth12THORN

The model enables us to show how changes in international competitiveness affectthe worldrsquos wine markets A crucial part of this exercise is explaining how pricesdetermined outside the grape and wine markets influence these markets Becausethe model is in partial equilibrium to depict the impacts of changes in internationalcompetitiveness outside price changes need to be imposed as shocks on the modelThe price of intermediate inputs other than grapes and wine shown in equations (1)and (2) is set equal to the price of GDP (Pg

d ) multiplied by a shifter Fcd

Pcid frac14 Fc

dPgd eth13THORN

If no specific price observations are available the shifter Fcd remains exogenous and

unshocked with the change in price being determined by a shock to the price ofGDP If observations are available for specific input price movements the shifterFc

dbecomes endogenous with Pc

id now exogenous and shocked

Wid frac14 Fwd P

gd eth14THORN

Wage rates are treated similarly In equation (14) if the wage shifter Fwd is exogenous

changes in wage rates Wid are determined by changes in the price of GDP If wagerate data are available Fw

d becomes endogenous andwage rates are shocked directly

Pmcd frac14 Fm

d Pgd eth15THORN

The prices of trade and transport margins are also determined by the price of GDP ifthe shifter Fm

d in equation (15) is exogenous

Changes in international competitiveness depend on changes in relative pricelevels and changes in nominal exchange rates In equation (16) fR

s denotes real

Kym Anderson and Glyn Wittwer 249

exchange-rate movements relative to the US dollar

fRs frac14 Pg

s =frac12PgUSA

fs$ eth16THORN

In equation (16) the nominal exchange rate for the United States is alwaysunchanged because nominal and real exchange rates are expressed relative to theUS currency

Changes in international market conditions may have impacts in one direction onproducer prices as expressed in US dollars P0s

i and potentially in the opposite direc-tion in local currency units Hence we calculate real producer pricesP0s

iloc in localcurrency terms (ie the price most relevant to domestic producers)

P0siloc frac14 P0s

ifs=P

gs eth17THORN

To obtain real price changes in local currency terms we convert US dollar prices(P3 s

cd for the source-specific price and P3cd for the source-composite price) to reallocal currency prices (P3 s

cdloc and P3cdloc) using the consumer price index (CPI)(Pc

d ) as the deflator

P3 scdloc frac14 P3 s

cdfd=P

cd eth18THORN

and

P3cdloc frac14 P3cdfd=Pcd eth19THORN

References

Anderson K and Nelgen S (2011) Global Wine Markets 1961 to 2009 A StatisticalCompendium Adelaide University of Adelaide Press Also available as an e-book atwwwadelaideeduaupresstitlesglobal-wine

Anderson K Nelgen S and Pinilla V (2017) Global Wine Markets 1860 to 2016 AStatistical Compendium Adelaide University of Adelaide Press (forthcoming) Also tobe available as an e-book at wwwadelaideeduaupress

Anderson K and Pinilla V (with the assistance of A J Holmes) (2017)Annual Database ofGlobal Wine Markets 1835 to 2016 Wine Economics Research Centre University ofAdelaide to be posted in October at wwwadelaideeduauwine-econdatabases

Anderson K and Strutt A (2016) Impacts of Asiarsquos rise on African and Latin Americantrade Projections to 2030 World Economy 39(2) 172ndash194

Anderson K andWittwer G (2013) Modeling global wine markets to 2018 Exchange ratestaste changes and Chinarsquos import growth Journal of Wine Economics 8(2) 131ndash158

Anderson K and Wittwer G (2015) Asiarsquos evolving role in global wine markets ChinaEconomic Review 35 1ndash14

250 UK and Global Wine Markets by 2025 and Implications of Brexit

Armington P A (1969) A theory of demand for products distinguished by place of produc-tion IMF Staff Papers 16 159ndash178

Baldwin R E (ed) (2016) Brexit Beckons Thinking Ahead by Leading Economists LondonCentre for Economic Policy Research

Campos N F (2016) Lousy experts Looking back at the ex ante estimates of the costs ofBrexit In R E Baldwin (ed) Brexit Beckons Thinking Ahead by Leading Economists(ch 3) London Centre for Economic Policy Research

Dixon P Parmenter B Sutton J and Vincent D (1982) ORANI A Multisectoral Modelof the Australian Economy Contributions to Economic Analysis 142 Amsterdam North-Holland

European Commission (2015) Trade for All Towards a More Responsible Trade andInvestment Policy Brussels European Commission

Harrison J Horridge M Jerie M and Pearson K (2014) GEMPACK ManualMelbourne GEMPACK Software

HM Treasury (2017) Spring Budget 2017 London HM TreasuryHolmes A J and Anderson K (2017a)Annual Database of National Beverage Consumption

Volumes and Expenditures 1950 to 2015 Wine Economics Research Centre University ofAdelaide posted at wwwadelaideeduauwine-econdatabases

Holmes A J and Anderson K (2017b) Convergence in National Alcohol ConsumptionPatterns New Global Indicators Journal of Wine Economics 12(2) 117ndash148

Ludington C C (2013) The Politics of Wine in Britain A New Cultural History BasingstokeUK Palgrave Macmillan

Office of National Statistics (2017) Analysis of real earnings London Office of NationalStatistics Available at httpswwwonsgovukemploymentandlabourmarketpeopleinworkearningsandworkinghoursarticlessupplementaryanalysisofaverageweeklyearningslatest

Rollo J Borchert I Dawar K Holmes P and Winters L A (2016) The World TradeOrganisation A safety net for a post-Brexit UK trade policy Briefing Paper 1UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-1-final-1pdf

Smith A (2017) Brexit Hard truths and hard choices UKTPO blog 19 June Available athttpsblogssussexacukuktpo20170619brexit-hard-truths-and-hard-choices

Swinbank A (2017) World trade rules and the policy options for British agriculture post-Brexit Briefing Paper 7 UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-7pdf

Viner J (1950) The Customs Union Issue New York Carnegie Endowment for InternationalPeace

Wine Australia (2015) Export Market Guide European Union Adelaide Wine AustraliaWittwer G Berger N and Anderson K (2003) A model of the worldrsquos wine markets

Economic Modelling 20(3) 487ndash506

Kym Anderson and Glyn Wittwer 251

Page 21: U.K. and Global Wine Markets by 2025, and Implications of ... · U.K. and Global Wine Markets by 2025, and Implications of Brexit* Kym Andersona and Glyn Wittwerb Abstract The United

from the United Kingdom to EU27 countries ndash but again they export less overallGlobal wine trade in 2025 is 240 ML (19) or $18 billion (35) less under thisldquolargerdquo Brexit scenario The percentage by which wine exportersrsquo trade shrinks isgreater for values than for volumes because of changes in relative prices of differ-ent-quality wines Those differences are shown in the numbers in parentheses inTable 5

Three other points are worth making about Table 5 One is that Australia sellsslightly more to the United Kingdom in the ldquosmallrdquo Brexit scenario rather thanslightly less as in the ldquolargerdquo Brexit case Evidently the negative income and price(devaluation) effects do not more than offset the positive trade-diverting effect onAustralian exports to the United Kingdom of removing preferences in the ldquosmallrdquoscenario Second New Zealand sells slightly more to non-UK countries underBrexit despite greater competition from the EU27 Chile and South Africa Thisanomaly is due to changes in the relative prices of different qualities of wine inglobal wine markets bearing in mind that New Zealand has the worldrsquos highestaverage price for still-wine exports And third the value (but not the volume) ofexports of ldquoOtherrdquo countries to markets other than the United Kingdom arehigher under Brexit This result too is due to changes in the relative prices of differ-ent qualities of wine in global wine markets

Figure 9

Difference in 2025 Wine Import Volumes and Values as a Result of the ldquoLargerdquo Initial BrexitShock (ML and US$ Million in 2014 US dollars)

Fig9-BW

onlin

eBW

inprint

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 241

B Subsequent Impact of Brexit from a UK-EU27 FTA

The next-most-likely step in the Brexit process is for the United Kingdom to nego-tiate a new trade arrangement with the EU27 We therefore assume that a UK-EU27FTAwith free bilateral wine trade is implemented and adjusted to by 2025 and thatprogress toward that end occurs soon enough that the adverse macroeconomicshocks from the initial impact of uncertainty over the Brexit process are confinedto those assumed in the ldquosmallrdquo scenario outlined above In this subsequent scenariowe assume that the British pound returns to the value it would have reached in 2025in the absence of Brexit (ie reversing the 10 devaluation assumed in the ldquosmallrdquoinitial Brexit scenario) and that real incomes in the United Kingdom are 8 higher

Table 5Difference in 2025 BilateralWine Import Volumes and Values fromKey Exporters by the United

Kingdom and the Rest of the World (RoW) as a Result of Initial Brexit Shock(ML and 2014US$ Million)a

(a) ldquolargerdquo scenario

Volume (ML) Value (2014US$ million)

UK RoW WORLD () UK RoW WORLD ()

EU27 minus287 136 minus150 (ndash17) minus1187 minus5 minus1192 (ndash31)Chile minus59 35 minus25 (ndash30) minus169 31 minus138 (ndash48)South Africa minus53 35 minus18 (ndash32) minus105 20 minus85 (ndash67)USA minus7 minus6 minus13 (ndash24) minus75 minus40 minus115 (ndash50)Australia minus4 minus3 minus7 (ndash09) minus25 minus65 minus90 (ndash30)Argentina minus3 minus9 minus12 (ndash48) minus16 minus39 minus55 (ndash52)New Zealand minus11 9 minus2 (ndash09) minus162 71 minus91 (ndash43)Others minus2 minus10 minus12 (ndash02) minus11 minus52 minus63 (ndash44)WORLD minus427 187 minus240 (ndash19) minus1750 minus79 minus1829 (ndash35)

(a) ldquosmallrdquo scenario

Volume (ML) Value (2014US$ million)

UK RoW WORLD () UK RoW WORLD ()

EU27 minus178 82 minus96 (ndash12) minus692 minus43 minus736 (ndash19)Chile minus46 28 minus18 (ndash24) minus128 36 minus91 (ndash32)South Africa minus43 29 minus14 (ndash42) minus82 23 minus59 (ndash47)USA 1 minus6 minus5 (ndash11) minus23 minus28 minus51 (ndash22)Australia 5 minus10 minus5 (ndash06) 19 minus56 minus38 (ndash13)Argentina 0 minus6 minus6 (ndash26) minus3 minus25 minus29 (ndash27)New Zealand minus5 4 minus1 (ndash06) minus80 34 minus46 (ndash22)Others 0 minus9 minus9 (ndash01) minus1 minus33 minus34 (ndash24)WORLD minus266 112 minus154 (ndash13) minus991 minus92 minus1083 (ndash21)

a Numbers in parentheses are the percentage difference between the Brexit and baseline scenarios for 2025 projected wine-import volumes orvalues by source

Source Authorsrsquo model results

242 UK and Global Wine Markets by 2025 and Implications of Brexit

in 2025 than in the ldquosmallrdquo scenario ndash that is they regain two-thirds of the value theywould have reached without Brexit-related uncertainty

This subsequent development in the Brexit process reverses most of but not allthe initial effects of Brexit by 2025 because of our assumption that the lost growthin the initial years of uncertainty following the Brexit vote are only partly recoveredby 2025 following the implementation of a UK-EU27 FTA Moreover the longer ittakes before this FTA is finalized and implemented the longer the estimated initialadverse effects persist and the larger the cumulative cost of Brexit to UK wine con-sumers and to grape and wine producers in wine-exporting countries

Table 6 summarizes the subsequent trade effects for 2025 It suggests that all butone-tenth of the loss in value of world trade in wine from the initial ldquosmallrdquo impact isrestored most of it because of a smaller reduction in the United Kingdomrsquos wineimports following the FTA Most of that improved outcome is because of recoveredimports from the EU27 commensurate with the latterrsquos high share of UK importsEven though Chile and South Africa are assumed in this scenario to have not yetsigned an FTA with the United Kingdom they export slightly more to the UnitedKingdom (and even more to all other countries) than in the initial ldquosmallrdquo scenario

Within the United Kingdom this FTA brings down the local currency consumerprice of wine by 9 largely offsetting the 11 rise in the initial ldquosmallrdquo Brexit sce-nario and it raises the volume consumed in the United Kingdom by 18 fully off-setting the 17 fall in the initial ldquosmallrdquo Brexit scenario

Table 6Difference in 2025 Bilateral Wine-Import Volumes and Values from Key Exporters by the

United Kingdom and the Rest of the World (RoW) as a Result of Implementing a UK-EU27FTA Following Initial ldquoSmallrdquo Brexit Shock (ML and 2014US$ Million Difference Relative to

Initial ldquoSmallrdquo Brexit Scenario)a

Volume (ML) Value (2014US$ million)

UK RoW WORLD (ldquosmallrdquo)a UK RoW WORLD (ldquosmallrdquo)a

EU27 212 minus112 100 (minus96) 750 0 750 (minus806)Chile 1 3 4 (minus18) 3 40 43 (minus92)South Africa 3 2 5 (minus14) 3 18 21 (minus60)USA 3 3 6 (minus5) 26 21 47 (minus50)Australia 1 1 2 (minus5) 8 38 46 (minus33)Argentina 1 4 5 (minus6) 4 20 24 (minus28)New Zealand 4 minus3 1 (minus1) 61 minus25 36 (minus46)Others 1 6 7 (minus9) 5 minus46 minus41 (92)WORLD 226 minus96 130 (minus154) 860 66 926 (minus1022)

a Numbers in parentheses are the world trade differences between the ldquosmallrdquo initial Brexit scenario and the baseline scenario copied fromcolumns 3 and 7 of Table 5(b)

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 243

In short this subsequent step in the Brexit process can be expected to restore muchof the initial adverse effects in the United Kingdom of the Brexit vote by 2025 Thewine exporters in the EU27 Chile and South Africa also lose less in this scenarioas compared with the initial ldquosmallrdquo scenario while other key exporters sell nearlyas much in 2025 in this scenario as they are projected to in the baseline ndash that isthey are beneficiaries of the reduced discrimination in global wine markets in thiscase where Chile and South Africa no longer enjoy preferential access to the UKmarket (and even though EU27 producers are again allowed to do so) Keep inmind though that although the estimated losses in the initial scenarios are reportedjust for 2025 they are expected to be felt throughout the preceding years of uncertaintyTherefore the longer it is before uncertainty abates and the United Kingdomrsquos currentpreferential trading arrangements are replaced by newones such as a UK-EU27 FTAthe higher the cumulative cost of the Brexit vote to the wine trade

VII Caveats and Conclusions

The above Brexit simulations are just a few of many scenarios that could be modeledObvious additional ones could also assume that FTAs are reached between theUnited Kingdom and other trading partners including Chile and South Africaand such countries as Australia and New Zealand Some of these other countriesare already exploring the ECrsquos invitation (see European Commission 2015) to nego-tiate an FTA with what will be the EU27 The sequence in which FTAs are signedand the speed with which they are implemented will matter (as was also the casewith the sequential signing over the past decade of bilateral FTAs with NortheastAsian countries by Chile Australia and New Zealand see Anderson andWittwer 2015) Even if the United Kingdom were able to sign additional bilateralFTAs and begin implementing them before 2025 it would have little effect on theabove results (because wine tariffs are such a minor contributor to them) unlessthose FTAs were to accelerate the United Kingdomrsquos economic growth and therise in the British poundrsquos value before 2025

We assume above that no changes are made to alcohol excise duties in the UnitedKingdom following Brexit even though they are scheduled to be progressively raisedwith inflation and may be raised even more for wine relative to spirits to offset theopposite effects of Brexit on those two domestic industries Nor do we make anyallowance for UK increases in consumer wine prices that may be needed to coverthe higher cost of clearing customs on imports from EU27 countries (which maybe more or less offset by increased smuggling and duty-free purchasing)Importers of fine wines including individuals who buy Bordeaux en primeur maywell reduce their demand because of concerns that delays on docks will affect thequality of their wines or because of greater currency uncertainty

To summarize Brexit is costly initially to UK consumers of wine (and of manyother tradable products) because the domestic retail price in local currency tax-

244 UK and Global Wine Markets by 2025 and Implications of Brexit

inclusive terms is 22 higher than otherwise in 2025 and the volume of wine con-sumed domestically is 28 lower in the ldquolargerdquo Brexit scenario (or 11 higher pricesand 17 lower quantities in the ldquosmallrdquo case) by 2025 unless a UK-EU27 FTA isdeveloped before then Even if such an FTA does get signed ratified by all 28 par-liaments and implemented by 2025 the slower income growth in the interim meansa smaller UK wine market in 2025 than would otherwise have been the case Thevolume reduction is a blow to many participants in UK wine-bottling transportingstoring wholesaling and retailing businesses in addition to restaurants and pubsVery little of that initial impact occurs because of higher import tariffs most impor-tant is the assumed fall in UK real incomes

The small but growing number of local UK vignerons (who supply less than 05of all domestic-wine sales and less than 4 of sparkling-wine sales) is initially pro-jected to sell less wine domestically because of reduced demand and to export only alittle extra abroad while the British pound is devalued Overall their production isonly 3 higher in 2025 even in the ldquolargerdquo Brexit scenario (not taking intoaccount the fact that their casual labor is likely to be more expensive under Brexitbecause of tighter restrictions on immigration) and their production diminishes ifand when a UK-EU27 FTA is in place

Even with a UK-EU27 FTA in place EU27 wine exporters export US$56 millionless wine in 2025 thanks to Brexit Chile and South Africa export $88 million lesswine and wine exports from the rest of the world are little different in aggregatealthough slightly larger from Australia

Clearly there will be great uncertainly for some time over the possible policy out-comes to flow from Brexit and their consequent sequential impacts on UK house-hold disposable incomes foreign exchange rates and bilateral wine tariffsMeanwhile the above projections under explicit assumptions provide some idea ofhow wine markets might be affected by the most-likely first two stages of theBrexit process (agreeing on a new tariff schedule at the WTO and agreeing to andimplementing a UK-EU27 FTA) In particular they make clear that nontrivialinitial impacts could affect the domestic wine market impacts that are likely to belarger than just the direct impact of changes in bilateral tariffs If the UnitedKingdom succeeds in getting countries to agree to transition arrangements thatdelay the changes in tariffs until new FTAs are signed and ratified the initialeffects will be less dramatic than in our first scenarios but it remains to be seenwhether any such agreements can be reached In any event the net effect of Brexiton the welfare of the worldrsquos consumers and producers of wine as a whole will benegative not just initially but permanently unless new trade policy commitmentsby the United Kingdom with major wine-exporting countries are sufficiently moreliberal than current arrangements

Kym Anderson and Glyn Wittwer 245

Appendix Key Equations in the Global Wine Markets Model

Amodel of the worldrsquos wine markets was first published byWittwer et al (2003) andmuch improved by Anderson and Wittwer (2013) Several significant enhancementshave been made to that original model Wine types have been disaggregated intomore types ndash namely nonpremium (including generic bulk) commercial-premiumand superpremium still wines and sparkling wines8 As in the original modelthere are two types of grapes premium and nonpremium Nonpremium wines usenonpremium grapes exclusively superpremium wines use mainly premium grapesand commercial-premium and sparkling wines use both types of grapes In theregional dimension the number of countries and country groups has expandedfrom 10 in the original model to 51 44 individual nations and 7 compositeregions The modelrsquos database is calibrated to 2014 for this paper based on datain Anderson et al (2017) trade data for which are downloaded from httpscomtradeunorg The modelrsquos supply-and-demand equations are based on the orig-inal ORANI modelrsquos theory (see Dixon Parmenter Sutton and Vincent 1982) andthe model is implemented using GEMPACK software (Harrison Horridge Jerieand Pearson 2014)

An enhancement of importance to the present study is the inclusion of exchange-rate variables in the model which allow a distinction between price impacts asobserved in local currency units and those observed in 2014 US dollars

In the model the grape and wine sectors minimize costs of intermediate inputssubject to weak constant elasticity of substitution (CES) substitutability betweeninputs By assumption no intermediate inputs are imported from other countries9Hence

Xcid frac14 f ethX1id CESfrac12P

cid=P1id $THORN eth1THORN

P1id X1id frac14X

c

Xcid P

cid eth2THORN

whereXcid is the quantity demanded of commodity c by grape or wine industry i in

region d Pcid is the corresponding price and X1id and P1id are the respective inter-

mediate composite quantities and prices

Two primary factors are employed in the sector labor (the quantity of which isendogenous with perfectly elastic supply) and capital Capital is usually treated asexogenous in quantity with rates of return bearing all the adjustment in thevarious scenarios This reflects the fact that grapes (a perennial crop) and wine-

8Commercial-premium still wines are defined by Anderson and Nelgen (2011) to be those priced betweenUS$250 and $750 per litre pre-tax at a countryrsquos border or wholesale9An exception concerns bulk wine imports used mainly in the bottling of commercial premium wine in theUK USA Canada Japan and Germany

246 UK and Global Wine Markets by 2025 and Implications of Brexit

plant capacity adjust slowly to market signals

Lid frac14 f ethFid CESfrac12Wid=PFid $THORN eth3THORN

Kid frac14 f ethFid CESethRid=PFid $THORN eth4THORN

PFid Fid frac14 Lid Wid thorn Kid Rid eth5THORN

Grape and wine producers are assumed to minimize costs subject to CES substitu-tion between capital and labor Equations (3) to (5) show primary factor demandsfor the labor composite Lid and capital Kid subject to a composite factor demandFid by industry i in region d The factor prices are Wid for labor Rid for capitalrentals and PFid for composite factor prices

Appendix Table A1Cumulative Consumption and Population Growth Rates and Changes in the Real Exchange Rate

(RER) Relative to the US Dollar 2014 to 2025 without Brexit ()

Aggregateconsumption Poprsquon RER

Aggregateconsumption Poprsquon RER

France 18 4 minus11 Australia 35 11 minus17Italy 11 2 minus9 New Zealand 32 9 minus26Portugal 14 0 minus9 Canada 27 8 minus18Spain 26 8 minus9 United States 31 8 0Austria 19 4 minus7 Argentina 7 10 109Belgium 20 7 minus9 Brazil 16 8 minus29Denmark 22 2 minus9 Chile 55 8 minus2Finland 21 3 minus7 Mexico 42 12 minus8Germany 14 minus2 minus11 Uruguay 45 3 1Greece 22 minus1 minus14 Other Latin America 60 10 minus5Ireland 42 12 minus9 South Africa 36 12 minus1Netherlands 21 4 minus9 Turkey 50 8 20Sweden 24 9 minus13 North Africa 53 11 0Switzerland 18 8 minus6 Other Africa 109 18 84United Kingdom 32 6 1 Middle East 52 18 minus12Other West Europe 21 10 minus1 China 79 3 5Bulgaria 41 minus7 7 Hong Kong 42 3 2Croatia 20 minus2 minus1 India 134 13 17Georgia 35 0 23 Japan 11 minus3 minus24Hungary 25 minus3 minus11 Korea 38 1 minus9Moldova 49 minus11 13 Malaysia 62 15 minus16Romania 45 minus4 22 Philippines 75 18 7Russia 18 minus2 minus8 Singapore 44 21 minus22Ukraine 22 minus5 14 Taiwan 29 1 minus13Other East Europe 40 minus5 48 Thailand 47 3 minus9

Other Asia 99 10 10

Source Authorsrsquo compilation from projections by various international agencies and from global economy-wide modeling by Anderson andStrutt (2016)

Kym Anderson and Glyn Wittwer 247

The composite factor demand Fid is proportional to total output Qid subject to aprimary factor using technology Aid Hence

Fid frac14 Qid Aid eth6THORN

The perfectly competitive zero-pure-profit condition is that total revenue valued atthe output price P0s

i multiplied by Qid equals the total production cost

P0si Qid frac14

X

cPcid X1cid thorn

X

oWo

id Loid thorn Rid Kid eth7THORN

Household demands follow a linear expenditure system in each region We reducethe optimizing problem for household consumption of each commodity subjectto a budget constraint to equations describing subsistence and discretionarydemands Aggregate subsistence expenditure WSUBd depends only on consumerprices P3cd for each commodity and the number of households N as per-capita sub-sistence quantities XSUBcd subject to given preferences are constant

WSUBd frac14X

c

P3cd XSUBcd Nd eth8THORN

Discretionary expenditures for each commodity (the left-hand side of equation (9))are equal to the marginal budget share (βcd) of aggregate discretionary expenditureThis aggregate is the bracketed term on the right-hand side of equation (9) whereW3TOTd is aggregate nominal expenditure

P3cdethX3cd XSUBcd NdTHORN frac14 βcdethW3TOTd WSUBdTHORN eth9THORN

Because real aggregate consumption is usually exogenous in our partial equilibriumsimulations the linear expenditure system determines the consumption shares ofindividual final commodities (ie the five wine types plus a composite of all otherconsumption items) driven by changes in relative prices as faced by domestic con-sumers The income elasticity of demand for each commodity is equal to the mar-ginal budget share divided by the expenditure share This number varies from 05for nonpremium wine to 17 for superpremium still wine The income elasticity ofdemand for other consumption is very close to 10 because wine accounts for anaverage of only 03 of aggregate expenditures globally and no more than 11in any country (Anderson and Nelgen 2011 Table 166)

A feature of our revised model of world wine markets is the inclusion of nominalexchange rates These appear directly in the equation linking retail prices (P3scd ) toproducer prices by country of origin (P0s

c ) where c denotes the wine type

P3 scd frac14 P0s

cfd

fsTtarcd T

taxcd thorn Pm

cd eth10THORN

248 UK and Global Wine Markets by 2025 and Implications of Brexit

The exchange rates in the consuming and producing regions are fd and fs respec-tively expressed as local currency units per $US Ttar

cd is the power of the tariff in theconsuming region and Ttax

cd is the power of the domestic consumption (or excise) taxprior to any generic value-added or goods-and-services tax Pm

cd is the price of marginm assumed to be locally supplied nontradable and therefore unaffected by theexchange rate

A given level of consumption for wine type c (X3cd) is satisfied using theArmington (1969) assumption in which wines from different countries of originare imperfectly substitutable First domestic wine is imperfectly substitutable witha composite of imports

X3sscd frac14 f ethX3cd CESethP3sscd=P3cdTHORNTHORN ss frac14 domestic imports eth11THORN

and then imports by origin (X3scd ) are determined in a second CES equation

X3scd frac14 f ethX3ssfrac14importscd CESethP3scd=P3

ssfrac14importscd THORNTHORN eth12THORN

The model enables us to show how changes in international competitiveness affectthe worldrsquos wine markets A crucial part of this exercise is explaining how pricesdetermined outside the grape and wine markets influence these markets Becausethe model is in partial equilibrium to depict the impacts of changes in internationalcompetitiveness outside price changes need to be imposed as shocks on the modelThe price of intermediate inputs other than grapes and wine shown in equations (1)and (2) is set equal to the price of GDP (Pg

d ) multiplied by a shifter Fcd

Pcid frac14 Fc

dPgd eth13THORN

If no specific price observations are available the shifter Fcd remains exogenous and

unshocked with the change in price being determined by a shock to the price ofGDP If observations are available for specific input price movements the shifterFc

dbecomes endogenous with Pc

id now exogenous and shocked

Wid frac14 Fwd P

gd eth14THORN

Wage rates are treated similarly In equation (14) if the wage shifter Fwd is exogenous

changes in wage rates Wid are determined by changes in the price of GDP If wagerate data are available Fw

d becomes endogenous andwage rates are shocked directly

Pmcd frac14 Fm

d Pgd eth15THORN

The prices of trade and transport margins are also determined by the price of GDP ifthe shifter Fm

d in equation (15) is exogenous

Changes in international competitiveness depend on changes in relative pricelevels and changes in nominal exchange rates In equation (16) fR

s denotes real

Kym Anderson and Glyn Wittwer 249

exchange-rate movements relative to the US dollar

fRs frac14 Pg

s =frac12PgUSA

fs$ eth16THORN

In equation (16) the nominal exchange rate for the United States is alwaysunchanged because nominal and real exchange rates are expressed relative to theUS currency

Changes in international market conditions may have impacts in one direction onproducer prices as expressed in US dollars P0s

i and potentially in the opposite direc-tion in local currency units Hence we calculate real producer pricesP0s

iloc in localcurrency terms (ie the price most relevant to domestic producers)

P0siloc frac14 P0s

ifs=P

gs eth17THORN

To obtain real price changes in local currency terms we convert US dollar prices(P3 s

cd for the source-specific price and P3cd for the source-composite price) to reallocal currency prices (P3 s

cdloc and P3cdloc) using the consumer price index (CPI)(Pc

d ) as the deflator

P3 scdloc frac14 P3 s

cdfd=P

cd eth18THORN

and

P3cdloc frac14 P3cdfd=Pcd eth19THORN

References

Anderson K and Nelgen S (2011) Global Wine Markets 1961 to 2009 A StatisticalCompendium Adelaide University of Adelaide Press Also available as an e-book atwwwadelaideeduaupresstitlesglobal-wine

Anderson K Nelgen S and Pinilla V (2017) Global Wine Markets 1860 to 2016 AStatistical Compendium Adelaide University of Adelaide Press (forthcoming) Also tobe available as an e-book at wwwadelaideeduaupress

Anderson K and Pinilla V (with the assistance of A J Holmes) (2017)Annual Database ofGlobal Wine Markets 1835 to 2016 Wine Economics Research Centre University ofAdelaide to be posted in October at wwwadelaideeduauwine-econdatabases

Anderson K and Strutt A (2016) Impacts of Asiarsquos rise on African and Latin Americantrade Projections to 2030 World Economy 39(2) 172ndash194

Anderson K andWittwer G (2013) Modeling global wine markets to 2018 Exchange ratestaste changes and Chinarsquos import growth Journal of Wine Economics 8(2) 131ndash158

Anderson K and Wittwer G (2015) Asiarsquos evolving role in global wine markets ChinaEconomic Review 35 1ndash14

250 UK and Global Wine Markets by 2025 and Implications of Brexit

Armington P A (1969) A theory of demand for products distinguished by place of produc-tion IMF Staff Papers 16 159ndash178

Baldwin R E (ed) (2016) Brexit Beckons Thinking Ahead by Leading Economists LondonCentre for Economic Policy Research

Campos N F (2016) Lousy experts Looking back at the ex ante estimates of the costs ofBrexit In R E Baldwin (ed) Brexit Beckons Thinking Ahead by Leading Economists(ch 3) London Centre for Economic Policy Research

Dixon P Parmenter B Sutton J and Vincent D (1982) ORANI A Multisectoral Modelof the Australian Economy Contributions to Economic Analysis 142 Amsterdam North-Holland

European Commission (2015) Trade for All Towards a More Responsible Trade andInvestment Policy Brussels European Commission

Harrison J Horridge M Jerie M and Pearson K (2014) GEMPACK ManualMelbourne GEMPACK Software

HM Treasury (2017) Spring Budget 2017 London HM TreasuryHolmes A J and Anderson K (2017a)Annual Database of National Beverage Consumption

Volumes and Expenditures 1950 to 2015 Wine Economics Research Centre University ofAdelaide posted at wwwadelaideeduauwine-econdatabases

Holmes A J and Anderson K (2017b) Convergence in National Alcohol ConsumptionPatterns New Global Indicators Journal of Wine Economics 12(2) 117ndash148

Ludington C C (2013) The Politics of Wine in Britain A New Cultural History BasingstokeUK Palgrave Macmillan

Office of National Statistics (2017) Analysis of real earnings London Office of NationalStatistics Available at httpswwwonsgovukemploymentandlabourmarketpeopleinworkearningsandworkinghoursarticlessupplementaryanalysisofaverageweeklyearningslatest

Rollo J Borchert I Dawar K Holmes P and Winters L A (2016) The World TradeOrganisation A safety net for a post-Brexit UK trade policy Briefing Paper 1UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-1-final-1pdf

Smith A (2017) Brexit Hard truths and hard choices UKTPO blog 19 June Available athttpsblogssussexacukuktpo20170619brexit-hard-truths-and-hard-choices

Swinbank A (2017) World trade rules and the policy options for British agriculture post-Brexit Briefing Paper 7 UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-7pdf

Viner J (1950) The Customs Union Issue New York Carnegie Endowment for InternationalPeace

Wine Australia (2015) Export Market Guide European Union Adelaide Wine AustraliaWittwer G Berger N and Anderson K (2003) A model of the worldrsquos wine markets

Economic Modelling 20(3) 487ndash506

Kym Anderson and Glyn Wittwer 251

Page 22: U.K. and Global Wine Markets by 2025, and Implications of ... · U.K. and Global Wine Markets by 2025, and Implications of Brexit* Kym Andersona and Glyn Wittwerb Abstract The United

B Subsequent Impact of Brexit from a UK-EU27 FTA

The next-most-likely step in the Brexit process is for the United Kingdom to nego-tiate a new trade arrangement with the EU27 We therefore assume that a UK-EU27FTAwith free bilateral wine trade is implemented and adjusted to by 2025 and thatprogress toward that end occurs soon enough that the adverse macroeconomicshocks from the initial impact of uncertainty over the Brexit process are confinedto those assumed in the ldquosmallrdquo scenario outlined above In this subsequent scenariowe assume that the British pound returns to the value it would have reached in 2025in the absence of Brexit (ie reversing the 10 devaluation assumed in the ldquosmallrdquoinitial Brexit scenario) and that real incomes in the United Kingdom are 8 higher

Table 5Difference in 2025 BilateralWine Import Volumes and Values fromKey Exporters by the United

Kingdom and the Rest of the World (RoW) as a Result of Initial Brexit Shock(ML and 2014US$ Million)a

(a) ldquolargerdquo scenario

Volume (ML) Value (2014US$ million)

UK RoW WORLD () UK RoW WORLD ()

EU27 minus287 136 minus150 (ndash17) minus1187 minus5 minus1192 (ndash31)Chile minus59 35 minus25 (ndash30) minus169 31 minus138 (ndash48)South Africa minus53 35 minus18 (ndash32) minus105 20 minus85 (ndash67)USA minus7 minus6 minus13 (ndash24) minus75 minus40 minus115 (ndash50)Australia minus4 minus3 minus7 (ndash09) minus25 minus65 minus90 (ndash30)Argentina minus3 minus9 minus12 (ndash48) minus16 minus39 minus55 (ndash52)New Zealand minus11 9 minus2 (ndash09) minus162 71 minus91 (ndash43)Others minus2 minus10 minus12 (ndash02) minus11 minus52 minus63 (ndash44)WORLD minus427 187 minus240 (ndash19) minus1750 minus79 minus1829 (ndash35)

(a) ldquosmallrdquo scenario

Volume (ML) Value (2014US$ million)

UK RoW WORLD () UK RoW WORLD ()

EU27 minus178 82 minus96 (ndash12) minus692 minus43 minus736 (ndash19)Chile minus46 28 minus18 (ndash24) minus128 36 minus91 (ndash32)South Africa minus43 29 minus14 (ndash42) minus82 23 minus59 (ndash47)USA 1 minus6 minus5 (ndash11) minus23 minus28 minus51 (ndash22)Australia 5 minus10 minus5 (ndash06) 19 minus56 minus38 (ndash13)Argentina 0 minus6 minus6 (ndash26) minus3 minus25 minus29 (ndash27)New Zealand minus5 4 minus1 (ndash06) minus80 34 minus46 (ndash22)Others 0 minus9 minus9 (ndash01) minus1 minus33 minus34 (ndash24)WORLD minus266 112 minus154 (ndash13) minus991 minus92 minus1083 (ndash21)

a Numbers in parentheses are the percentage difference between the Brexit and baseline scenarios for 2025 projected wine-import volumes orvalues by source

Source Authorsrsquo model results

242 UK and Global Wine Markets by 2025 and Implications of Brexit

in 2025 than in the ldquosmallrdquo scenario ndash that is they regain two-thirds of the value theywould have reached without Brexit-related uncertainty

This subsequent development in the Brexit process reverses most of but not allthe initial effects of Brexit by 2025 because of our assumption that the lost growthin the initial years of uncertainty following the Brexit vote are only partly recoveredby 2025 following the implementation of a UK-EU27 FTA Moreover the longer ittakes before this FTA is finalized and implemented the longer the estimated initialadverse effects persist and the larger the cumulative cost of Brexit to UK wine con-sumers and to grape and wine producers in wine-exporting countries

Table 6 summarizes the subsequent trade effects for 2025 It suggests that all butone-tenth of the loss in value of world trade in wine from the initial ldquosmallrdquo impact isrestored most of it because of a smaller reduction in the United Kingdomrsquos wineimports following the FTA Most of that improved outcome is because of recoveredimports from the EU27 commensurate with the latterrsquos high share of UK importsEven though Chile and South Africa are assumed in this scenario to have not yetsigned an FTA with the United Kingdom they export slightly more to the UnitedKingdom (and even more to all other countries) than in the initial ldquosmallrdquo scenario

Within the United Kingdom this FTA brings down the local currency consumerprice of wine by 9 largely offsetting the 11 rise in the initial ldquosmallrdquo Brexit sce-nario and it raises the volume consumed in the United Kingdom by 18 fully off-setting the 17 fall in the initial ldquosmallrdquo Brexit scenario

Table 6Difference in 2025 Bilateral Wine-Import Volumes and Values from Key Exporters by the

United Kingdom and the Rest of the World (RoW) as a Result of Implementing a UK-EU27FTA Following Initial ldquoSmallrdquo Brexit Shock (ML and 2014US$ Million Difference Relative to

Initial ldquoSmallrdquo Brexit Scenario)a

Volume (ML) Value (2014US$ million)

UK RoW WORLD (ldquosmallrdquo)a UK RoW WORLD (ldquosmallrdquo)a

EU27 212 minus112 100 (minus96) 750 0 750 (minus806)Chile 1 3 4 (minus18) 3 40 43 (minus92)South Africa 3 2 5 (minus14) 3 18 21 (minus60)USA 3 3 6 (minus5) 26 21 47 (minus50)Australia 1 1 2 (minus5) 8 38 46 (minus33)Argentina 1 4 5 (minus6) 4 20 24 (minus28)New Zealand 4 minus3 1 (minus1) 61 minus25 36 (minus46)Others 1 6 7 (minus9) 5 minus46 minus41 (92)WORLD 226 minus96 130 (minus154) 860 66 926 (minus1022)

a Numbers in parentheses are the world trade differences between the ldquosmallrdquo initial Brexit scenario and the baseline scenario copied fromcolumns 3 and 7 of Table 5(b)

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 243

In short this subsequent step in the Brexit process can be expected to restore muchof the initial adverse effects in the United Kingdom of the Brexit vote by 2025 Thewine exporters in the EU27 Chile and South Africa also lose less in this scenarioas compared with the initial ldquosmallrdquo scenario while other key exporters sell nearlyas much in 2025 in this scenario as they are projected to in the baseline ndash that isthey are beneficiaries of the reduced discrimination in global wine markets in thiscase where Chile and South Africa no longer enjoy preferential access to the UKmarket (and even though EU27 producers are again allowed to do so) Keep inmind though that although the estimated losses in the initial scenarios are reportedjust for 2025 they are expected to be felt throughout the preceding years of uncertaintyTherefore the longer it is before uncertainty abates and the United Kingdomrsquos currentpreferential trading arrangements are replaced by newones such as a UK-EU27 FTAthe higher the cumulative cost of the Brexit vote to the wine trade

VII Caveats and Conclusions

The above Brexit simulations are just a few of many scenarios that could be modeledObvious additional ones could also assume that FTAs are reached between theUnited Kingdom and other trading partners including Chile and South Africaand such countries as Australia and New Zealand Some of these other countriesare already exploring the ECrsquos invitation (see European Commission 2015) to nego-tiate an FTA with what will be the EU27 The sequence in which FTAs are signedand the speed with which they are implemented will matter (as was also the casewith the sequential signing over the past decade of bilateral FTAs with NortheastAsian countries by Chile Australia and New Zealand see Anderson andWittwer 2015) Even if the United Kingdom were able to sign additional bilateralFTAs and begin implementing them before 2025 it would have little effect on theabove results (because wine tariffs are such a minor contributor to them) unlessthose FTAs were to accelerate the United Kingdomrsquos economic growth and therise in the British poundrsquos value before 2025

We assume above that no changes are made to alcohol excise duties in the UnitedKingdom following Brexit even though they are scheduled to be progressively raisedwith inflation and may be raised even more for wine relative to spirits to offset theopposite effects of Brexit on those two domestic industries Nor do we make anyallowance for UK increases in consumer wine prices that may be needed to coverthe higher cost of clearing customs on imports from EU27 countries (which maybe more or less offset by increased smuggling and duty-free purchasing)Importers of fine wines including individuals who buy Bordeaux en primeur maywell reduce their demand because of concerns that delays on docks will affect thequality of their wines or because of greater currency uncertainty

To summarize Brexit is costly initially to UK consumers of wine (and of manyother tradable products) because the domestic retail price in local currency tax-

244 UK and Global Wine Markets by 2025 and Implications of Brexit

inclusive terms is 22 higher than otherwise in 2025 and the volume of wine con-sumed domestically is 28 lower in the ldquolargerdquo Brexit scenario (or 11 higher pricesand 17 lower quantities in the ldquosmallrdquo case) by 2025 unless a UK-EU27 FTA isdeveloped before then Even if such an FTA does get signed ratified by all 28 par-liaments and implemented by 2025 the slower income growth in the interim meansa smaller UK wine market in 2025 than would otherwise have been the case Thevolume reduction is a blow to many participants in UK wine-bottling transportingstoring wholesaling and retailing businesses in addition to restaurants and pubsVery little of that initial impact occurs because of higher import tariffs most impor-tant is the assumed fall in UK real incomes

The small but growing number of local UK vignerons (who supply less than 05of all domestic-wine sales and less than 4 of sparkling-wine sales) is initially pro-jected to sell less wine domestically because of reduced demand and to export only alittle extra abroad while the British pound is devalued Overall their production isonly 3 higher in 2025 even in the ldquolargerdquo Brexit scenario (not taking intoaccount the fact that their casual labor is likely to be more expensive under Brexitbecause of tighter restrictions on immigration) and their production diminishes ifand when a UK-EU27 FTA is in place

Even with a UK-EU27 FTA in place EU27 wine exporters export US$56 millionless wine in 2025 thanks to Brexit Chile and South Africa export $88 million lesswine and wine exports from the rest of the world are little different in aggregatealthough slightly larger from Australia

Clearly there will be great uncertainly for some time over the possible policy out-comes to flow from Brexit and their consequent sequential impacts on UK house-hold disposable incomes foreign exchange rates and bilateral wine tariffsMeanwhile the above projections under explicit assumptions provide some idea ofhow wine markets might be affected by the most-likely first two stages of theBrexit process (agreeing on a new tariff schedule at the WTO and agreeing to andimplementing a UK-EU27 FTA) In particular they make clear that nontrivialinitial impacts could affect the domestic wine market impacts that are likely to belarger than just the direct impact of changes in bilateral tariffs If the UnitedKingdom succeeds in getting countries to agree to transition arrangements thatdelay the changes in tariffs until new FTAs are signed and ratified the initialeffects will be less dramatic than in our first scenarios but it remains to be seenwhether any such agreements can be reached In any event the net effect of Brexiton the welfare of the worldrsquos consumers and producers of wine as a whole will benegative not just initially but permanently unless new trade policy commitmentsby the United Kingdom with major wine-exporting countries are sufficiently moreliberal than current arrangements

Kym Anderson and Glyn Wittwer 245

Appendix Key Equations in the Global Wine Markets Model

Amodel of the worldrsquos wine markets was first published byWittwer et al (2003) andmuch improved by Anderson and Wittwer (2013) Several significant enhancementshave been made to that original model Wine types have been disaggregated intomore types ndash namely nonpremium (including generic bulk) commercial-premiumand superpremium still wines and sparkling wines8 As in the original modelthere are two types of grapes premium and nonpremium Nonpremium wines usenonpremium grapes exclusively superpremium wines use mainly premium grapesand commercial-premium and sparkling wines use both types of grapes In theregional dimension the number of countries and country groups has expandedfrom 10 in the original model to 51 44 individual nations and 7 compositeregions The modelrsquos database is calibrated to 2014 for this paper based on datain Anderson et al (2017) trade data for which are downloaded from httpscomtradeunorg The modelrsquos supply-and-demand equations are based on the orig-inal ORANI modelrsquos theory (see Dixon Parmenter Sutton and Vincent 1982) andthe model is implemented using GEMPACK software (Harrison Horridge Jerieand Pearson 2014)

An enhancement of importance to the present study is the inclusion of exchange-rate variables in the model which allow a distinction between price impacts asobserved in local currency units and those observed in 2014 US dollars

In the model the grape and wine sectors minimize costs of intermediate inputssubject to weak constant elasticity of substitution (CES) substitutability betweeninputs By assumption no intermediate inputs are imported from other countries9Hence

Xcid frac14 f ethX1id CESfrac12P

cid=P1id $THORN eth1THORN

P1id X1id frac14X

c

Xcid P

cid eth2THORN

whereXcid is the quantity demanded of commodity c by grape or wine industry i in

region d Pcid is the corresponding price and X1id and P1id are the respective inter-

mediate composite quantities and prices

Two primary factors are employed in the sector labor (the quantity of which isendogenous with perfectly elastic supply) and capital Capital is usually treated asexogenous in quantity with rates of return bearing all the adjustment in thevarious scenarios This reflects the fact that grapes (a perennial crop) and wine-

8Commercial-premium still wines are defined by Anderson and Nelgen (2011) to be those priced betweenUS$250 and $750 per litre pre-tax at a countryrsquos border or wholesale9An exception concerns bulk wine imports used mainly in the bottling of commercial premium wine in theUK USA Canada Japan and Germany

246 UK and Global Wine Markets by 2025 and Implications of Brexit

plant capacity adjust slowly to market signals

Lid frac14 f ethFid CESfrac12Wid=PFid $THORN eth3THORN

Kid frac14 f ethFid CESethRid=PFid $THORN eth4THORN

PFid Fid frac14 Lid Wid thorn Kid Rid eth5THORN

Grape and wine producers are assumed to minimize costs subject to CES substitu-tion between capital and labor Equations (3) to (5) show primary factor demandsfor the labor composite Lid and capital Kid subject to a composite factor demandFid by industry i in region d The factor prices are Wid for labor Rid for capitalrentals and PFid for composite factor prices

Appendix Table A1Cumulative Consumption and Population Growth Rates and Changes in the Real Exchange Rate

(RER) Relative to the US Dollar 2014 to 2025 without Brexit ()

Aggregateconsumption Poprsquon RER

Aggregateconsumption Poprsquon RER

France 18 4 minus11 Australia 35 11 minus17Italy 11 2 minus9 New Zealand 32 9 minus26Portugal 14 0 minus9 Canada 27 8 minus18Spain 26 8 minus9 United States 31 8 0Austria 19 4 minus7 Argentina 7 10 109Belgium 20 7 minus9 Brazil 16 8 minus29Denmark 22 2 minus9 Chile 55 8 minus2Finland 21 3 minus7 Mexico 42 12 minus8Germany 14 minus2 minus11 Uruguay 45 3 1Greece 22 minus1 minus14 Other Latin America 60 10 minus5Ireland 42 12 minus9 South Africa 36 12 minus1Netherlands 21 4 minus9 Turkey 50 8 20Sweden 24 9 minus13 North Africa 53 11 0Switzerland 18 8 minus6 Other Africa 109 18 84United Kingdom 32 6 1 Middle East 52 18 minus12Other West Europe 21 10 minus1 China 79 3 5Bulgaria 41 minus7 7 Hong Kong 42 3 2Croatia 20 minus2 minus1 India 134 13 17Georgia 35 0 23 Japan 11 minus3 minus24Hungary 25 minus3 minus11 Korea 38 1 minus9Moldova 49 minus11 13 Malaysia 62 15 minus16Romania 45 minus4 22 Philippines 75 18 7Russia 18 minus2 minus8 Singapore 44 21 minus22Ukraine 22 minus5 14 Taiwan 29 1 minus13Other East Europe 40 minus5 48 Thailand 47 3 minus9

Other Asia 99 10 10

Source Authorsrsquo compilation from projections by various international agencies and from global economy-wide modeling by Anderson andStrutt (2016)

Kym Anderson and Glyn Wittwer 247

The composite factor demand Fid is proportional to total output Qid subject to aprimary factor using technology Aid Hence

Fid frac14 Qid Aid eth6THORN

The perfectly competitive zero-pure-profit condition is that total revenue valued atthe output price P0s

i multiplied by Qid equals the total production cost

P0si Qid frac14

X

cPcid X1cid thorn

X

oWo

id Loid thorn Rid Kid eth7THORN

Household demands follow a linear expenditure system in each region We reducethe optimizing problem for household consumption of each commodity subjectto a budget constraint to equations describing subsistence and discretionarydemands Aggregate subsistence expenditure WSUBd depends only on consumerprices P3cd for each commodity and the number of households N as per-capita sub-sistence quantities XSUBcd subject to given preferences are constant

WSUBd frac14X

c

P3cd XSUBcd Nd eth8THORN

Discretionary expenditures for each commodity (the left-hand side of equation (9))are equal to the marginal budget share (βcd) of aggregate discretionary expenditureThis aggregate is the bracketed term on the right-hand side of equation (9) whereW3TOTd is aggregate nominal expenditure

P3cdethX3cd XSUBcd NdTHORN frac14 βcdethW3TOTd WSUBdTHORN eth9THORN

Because real aggregate consumption is usually exogenous in our partial equilibriumsimulations the linear expenditure system determines the consumption shares ofindividual final commodities (ie the five wine types plus a composite of all otherconsumption items) driven by changes in relative prices as faced by domestic con-sumers The income elasticity of demand for each commodity is equal to the mar-ginal budget share divided by the expenditure share This number varies from 05for nonpremium wine to 17 for superpremium still wine The income elasticity ofdemand for other consumption is very close to 10 because wine accounts for anaverage of only 03 of aggregate expenditures globally and no more than 11in any country (Anderson and Nelgen 2011 Table 166)

A feature of our revised model of world wine markets is the inclusion of nominalexchange rates These appear directly in the equation linking retail prices (P3scd ) toproducer prices by country of origin (P0s

c ) where c denotes the wine type

P3 scd frac14 P0s

cfd

fsTtarcd T

taxcd thorn Pm

cd eth10THORN

248 UK and Global Wine Markets by 2025 and Implications of Brexit

The exchange rates in the consuming and producing regions are fd and fs respec-tively expressed as local currency units per $US Ttar

cd is the power of the tariff in theconsuming region and Ttax

cd is the power of the domestic consumption (or excise) taxprior to any generic value-added or goods-and-services tax Pm

cd is the price of marginm assumed to be locally supplied nontradable and therefore unaffected by theexchange rate

A given level of consumption for wine type c (X3cd) is satisfied using theArmington (1969) assumption in which wines from different countries of originare imperfectly substitutable First domestic wine is imperfectly substitutable witha composite of imports

X3sscd frac14 f ethX3cd CESethP3sscd=P3cdTHORNTHORN ss frac14 domestic imports eth11THORN

and then imports by origin (X3scd ) are determined in a second CES equation

X3scd frac14 f ethX3ssfrac14importscd CESethP3scd=P3

ssfrac14importscd THORNTHORN eth12THORN

The model enables us to show how changes in international competitiveness affectthe worldrsquos wine markets A crucial part of this exercise is explaining how pricesdetermined outside the grape and wine markets influence these markets Becausethe model is in partial equilibrium to depict the impacts of changes in internationalcompetitiveness outside price changes need to be imposed as shocks on the modelThe price of intermediate inputs other than grapes and wine shown in equations (1)and (2) is set equal to the price of GDP (Pg

d ) multiplied by a shifter Fcd

Pcid frac14 Fc

dPgd eth13THORN

If no specific price observations are available the shifter Fcd remains exogenous and

unshocked with the change in price being determined by a shock to the price ofGDP If observations are available for specific input price movements the shifterFc

dbecomes endogenous with Pc

id now exogenous and shocked

Wid frac14 Fwd P

gd eth14THORN

Wage rates are treated similarly In equation (14) if the wage shifter Fwd is exogenous

changes in wage rates Wid are determined by changes in the price of GDP If wagerate data are available Fw

d becomes endogenous andwage rates are shocked directly

Pmcd frac14 Fm

d Pgd eth15THORN

The prices of trade and transport margins are also determined by the price of GDP ifthe shifter Fm

d in equation (15) is exogenous

Changes in international competitiveness depend on changes in relative pricelevels and changes in nominal exchange rates In equation (16) fR

s denotes real

Kym Anderson and Glyn Wittwer 249

exchange-rate movements relative to the US dollar

fRs frac14 Pg

s =frac12PgUSA

fs$ eth16THORN

In equation (16) the nominal exchange rate for the United States is alwaysunchanged because nominal and real exchange rates are expressed relative to theUS currency

Changes in international market conditions may have impacts in one direction onproducer prices as expressed in US dollars P0s

i and potentially in the opposite direc-tion in local currency units Hence we calculate real producer pricesP0s

iloc in localcurrency terms (ie the price most relevant to domestic producers)

P0siloc frac14 P0s

ifs=P

gs eth17THORN

To obtain real price changes in local currency terms we convert US dollar prices(P3 s

cd for the source-specific price and P3cd for the source-composite price) to reallocal currency prices (P3 s

cdloc and P3cdloc) using the consumer price index (CPI)(Pc

d ) as the deflator

P3 scdloc frac14 P3 s

cdfd=P

cd eth18THORN

and

P3cdloc frac14 P3cdfd=Pcd eth19THORN

References

Anderson K and Nelgen S (2011) Global Wine Markets 1961 to 2009 A StatisticalCompendium Adelaide University of Adelaide Press Also available as an e-book atwwwadelaideeduaupresstitlesglobal-wine

Anderson K Nelgen S and Pinilla V (2017) Global Wine Markets 1860 to 2016 AStatistical Compendium Adelaide University of Adelaide Press (forthcoming) Also tobe available as an e-book at wwwadelaideeduaupress

Anderson K and Pinilla V (with the assistance of A J Holmes) (2017)Annual Database ofGlobal Wine Markets 1835 to 2016 Wine Economics Research Centre University ofAdelaide to be posted in October at wwwadelaideeduauwine-econdatabases

Anderson K and Strutt A (2016) Impacts of Asiarsquos rise on African and Latin Americantrade Projections to 2030 World Economy 39(2) 172ndash194

Anderson K andWittwer G (2013) Modeling global wine markets to 2018 Exchange ratestaste changes and Chinarsquos import growth Journal of Wine Economics 8(2) 131ndash158

Anderson K and Wittwer G (2015) Asiarsquos evolving role in global wine markets ChinaEconomic Review 35 1ndash14

250 UK and Global Wine Markets by 2025 and Implications of Brexit

Armington P A (1969) A theory of demand for products distinguished by place of produc-tion IMF Staff Papers 16 159ndash178

Baldwin R E (ed) (2016) Brexit Beckons Thinking Ahead by Leading Economists LondonCentre for Economic Policy Research

Campos N F (2016) Lousy experts Looking back at the ex ante estimates of the costs ofBrexit In R E Baldwin (ed) Brexit Beckons Thinking Ahead by Leading Economists(ch 3) London Centre for Economic Policy Research

Dixon P Parmenter B Sutton J and Vincent D (1982) ORANI A Multisectoral Modelof the Australian Economy Contributions to Economic Analysis 142 Amsterdam North-Holland

European Commission (2015) Trade for All Towards a More Responsible Trade andInvestment Policy Brussels European Commission

Harrison J Horridge M Jerie M and Pearson K (2014) GEMPACK ManualMelbourne GEMPACK Software

HM Treasury (2017) Spring Budget 2017 London HM TreasuryHolmes A J and Anderson K (2017a)Annual Database of National Beverage Consumption

Volumes and Expenditures 1950 to 2015 Wine Economics Research Centre University ofAdelaide posted at wwwadelaideeduauwine-econdatabases

Holmes A J and Anderson K (2017b) Convergence in National Alcohol ConsumptionPatterns New Global Indicators Journal of Wine Economics 12(2) 117ndash148

Ludington C C (2013) The Politics of Wine in Britain A New Cultural History BasingstokeUK Palgrave Macmillan

Office of National Statistics (2017) Analysis of real earnings London Office of NationalStatistics Available at httpswwwonsgovukemploymentandlabourmarketpeopleinworkearningsandworkinghoursarticlessupplementaryanalysisofaverageweeklyearningslatest

Rollo J Borchert I Dawar K Holmes P and Winters L A (2016) The World TradeOrganisation A safety net for a post-Brexit UK trade policy Briefing Paper 1UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-1-final-1pdf

Smith A (2017) Brexit Hard truths and hard choices UKTPO blog 19 June Available athttpsblogssussexacukuktpo20170619brexit-hard-truths-and-hard-choices

Swinbank A (2017) World trade rules and the policy options for British agriculture post-Brexit Briefing Paper 7 UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-7pdf

Viner J (1950) The Customs Union Issue New York Carnegie Endowment for InternationalPeace

Wine Australia (2015) Export Market Guide European Union Adelaide Wine AustraliaWittwer G Berger N and Anderson K (2003) A model of the worldrsquos wine markets

Economic Modelling 20(3) 487ndash506

Kym Anderson and Glyn Wittwer 251

Page 23: U.K. and Global Wine Markets by 2025, and Implications of ... · U.K. and Global Wine Markets by 2025, and Implications of Brexit* Kym Andersona and Glyn Wittwerb Abstract The United

in 2025 than in the ldquosmallrdquo scenario ndash that is they regain two-thirds of the value theywould have reached without Brexit-related uncertainty

This subsequent development in the Brexit process reverses most of but not allthe initial effects of Brexit by 2025 because of our assumption that the lost growthin the initial years of uncertainty following the Brexit vote are only partly recoveredby 2025 following the implementation of a UK-EU27 FTA Moreover the longer ittakes before this FTA is finalized and implemented the longer the estimated initialadverse effects persist and the larger the cumulative cost of Brexit to UK wine con-sumers and to grape and wine producers in wine-exporting countries

Table 6 summarizes the subsequent trade effects for 2025 It suggests that all butone-tenth of the loss in value of world trade in wine from the initial ldquosmallrdquo impact isrestored most of it because of a smaller reduction in the United Kingdomrsquos wineimports following the FTA Most of that improved outcome is because of recoveredimports from the EU27 commensurate with the latterrsquos high share of UK importsEven though Chile and South Africa are assumed in this scenario to have not yetsigned an FTA with the United Kingdom they export slightly more to the UnitedKingdom (and even more to all other countries) than in the initial ldquosmallrdquo scenario

Within the United Kingdom this FTA brings down the local currency consumerprice of wine by 9 largely offsetting the 11 rise in the initial ldquosmallrdquo Brexit sce-nario and it raises the volume consumed in the United Kingdom by 18 fully off-setting the 17 fall in the initial ldquosmallrdquo Brexit scenario

Table 6Difference in 2025 Bilateral Wine-Import Volumes and Values from Key Exporters by the

United Kingdom and the Rest of the World (RoW) as a Result of Implementing a UK-EU27FTA Following Initial ldquoSmallrdquo Brexit Shock (ML and 2014US$ Million Difference Relative to

Initial ldquoSmallrdquo Brexit Scenario)a

Volume (ML) Value (2014US$ million)

UK RoW WORLD (ldquosmallrdquo)a UK RoW WORLD (ldquosmallrdquo)a

EU27 212 minus112 100 (minus96) 750 0 750 (minus806)Chile 1 3 4 (minus18) 3 40 43 (minus92)South Africa 3 2 5 (minus14) 3 18 21 (minus60)USA 3 3 6 (minus5) 26 21 47 (minus50)Australia 1 1 2 (minus5) 8 38 46 (minus33)Argentina 1 4 5 (minus6) 4 20 24 (minus28)New Zealand 4 minus3 1 (minus1) 61 minus25 36 (minus46)Others 1 6 7 (minus9) 5 minus46 minus41 (92)WORLD 226 minus96 130 (minus154) 860 66 926 (minus1022)

a Numbers in parentheses are the world trade differences between the ldquosmallrdquo initial Brexit scenario and the baseline scenario copied fromcolumns 3 and 7 of Table 5(b)

Source Authorsrsquo model results

Kym Anderson and Glyn Wittwer 243

In short this subsequent step in the Brexit process can be expected to restore muchof the initial adverse effects in the United Kingdom of the Brexit vote by 2025 Thewine exporters in the EU27 Chile and South Africa also lose less in this scenarioas compared with the initial ldquosmallrdquo scenario while other key exporters sell nearlyas much in 2025 in this scenario as they are projected to in the baseline ndash that isthey are beneficiaries of the reduced discrimination in global wine markets in thiscase where Chile and South Africa no longer enjoy preferential access to the UKmarket (and even though EU27 producers are again allowed to do so) Keep inmind though that although the estimated losses in the initial scenarios are reportedjust for 2025 they are expected to be felt throughout the preceding years of uncertaintyTherefore the longer it is before uncertainty abates and the United Kingdomrsquos currentpreferential trading arrangements are replaced by newones such as a UK-EU27 FTAthe higher the cumulative cost of the Brexit vote to the wine trade

VII Caveats and Conclusions

The above Brexit simulations are just a few of many scenarios that could be modeledObvious additional ones could also assume that FTAs are reached between theUnited Kingdom and other trading partners including Chile and South Africaand such countries as Australia and New Zealand Some of these other countriesare already exploring the ECrsquos invitation (see European Commission 2015) to nego-tiate an FTA with what will be the EU27 The sequence in which FTAs are signedand the speed with which they are implemented will matter (as was also the casewith the sequential signing over the past decade of bilateral FTAs with NortheastAsian countries by Chile Australia and New Zealand see Anderson andWittwer 2015) Even if the United Kingdom were able to sign additional bilateralFTAs and begin implementing them before 2025 it would have little effect on theabove results (because wine tariffs are such a minor contributor to them) unlessthose FTAs were to accelerate the United Kingdomrsquos economic growth and therise in the British poundrsquos value before 2025

We assume above that no changes are made to alcohol excise duties in the UnitedKingdom following Brexit even though they are scheduled to be progressively raisedwith inflation and may be raised even more for wine relative to spirits to offset theopposite effects of Brexit on those two domestic industries Nor do we make anyallowance for UK increases in consumer wine prices that may be needed to coverthe higher cost of clearing customs on imports from EU27 countries (which maybe more or less offset by increased smuggling and duty-free purchasing)Importers of fine wines including individuals who buy Bordeaux en primeur maywell reduce their demand because of concerns that delays on docks will affect thequality of their wines or because of greater currency uncertainty

To summarize Brexit is costly initially to UK consumers of wine (and of manyother tradable products) because the domestic retail price in local currency tax-

244 UK and Global Wine Markets by 2025 and Implications of Brexit

inclusive terms is 22 higher than otherwise in 2025 and the volume of wine con-sumed domestically is 28 lower in the ldquolargerdquo Brexit scenario (or 11 higher pricesand 17 lower quantities in the ldquosmallrdquo case) by 2025 unless a UK-EU27 FTA isdeveloped before then Even if such an FTA does get signed ratified by all 28 par-liaments and implemented by 2025 the slower income growth in the interim meansa smaller UK wine market in 2025 than would otherwise have been the case Thevolume reduction is a blow to many participants in UK wine-bottling transportingstoring wholesaling and retailing businesses in addition to restaurants and pubsVery little of that initial impact occurs because of higher import tariffs most impor-tant is the assumed fall in UK real incomes

The small but growing number of local UK vignerons (who supply less than 05of all domestic-wine sales and less than 4 of sparkling-wine sales) is initially pro-jected to sell less wine domestically because of reduced demand and to export only alittle extra abroad while the British pound is devalued Overall their production isonly 3 higher in 2025 even in the ldquolargerdquo Brexit scenario (not taking intoaccount the fact that their casual labor is likely to be more expensive under Brexitbecause of tighter restrictions on immigration) and their production diminishes ifand when a UK-EU27 FTA is in place

Even with a UK-EU27 FTA in place EU27 wine exporters export US$56 millionless wine in 2025 thanks to Brexit Chile and South Africa export $88 million lesswine and wine exports from the rest of the world are little different in aggregatealthough slightly larger from Australia

Clearly there will be great uncertainly for some time over the possible policy out-comes to flow from Brexit and their consequent sequential impacts on UK house-hold disposable incomes foreign exchange rates and bilateral wine tariffsMeanwhile the above projections under explicit assumptions provide some idea ofhow wine markets might be affected by the most-likely first two stages of theBrexit process (agreeing on a new tariff schedule at the WTO and agreeing to andimplementing a UK-EU27 FTA) In particular they make clear that nontrivialinitial impacts could affect the domestic wine market impacts that are likely to belarger than just the direct impact of changes in bilateral tariffs If the UnitedKingdom succeeds in getting countries to agree to transition arrangements thatdelay the changes in tariffs until new FTAs are signed and ratified the initialeffects will be less dramatic than in our first scenarios but it remains to be seenwhether any such agreements can be reached In any event the net effect of Brexiton the welfare of the worldrsquos consumers and producers of wine as a whole will benegative not just initially but permanently unless new trade policy commitmentsby the United Kingdom with major wine-exporting countries are sufficiently moreliberal than current arrangements

Kym Anderson and Glyn Wittwer 245

Appendix Key Equations in the Global Wine Markets Model

Amodel of the worldrsquos wine markets was first published byWittwer et al (2003) andmuch improved by Anderson and Wittwer (2013) Several significant enhancementshave been made to that original model Wine types have been disaggregated intomore types ndash namely nonpremium (including generic bulk) commercial-premiumand superpremium still wines and sparkling wines8 As in the original modelthere are two types of grapes premium and nonpremium Nonpremium wines usenonpremium grapes exclusively superpremium wines use mainly premium grapesand commercial-premium and sparkling wines use both types of grapes In theregional dimension the number of countries and country groups has expandedfrom 10 in the original model to 51 44 individual nations and 7 compositeregions The modelrsquos database is calibrated to 2014 for this paper based on datain Anderson et al (2017) trade data for which are downloaded from httpscomtradeunorg The modelrsquos supply-and-demand equations are based on the orig-inal ORANI modelrsquos theory (see Dixon Parmenter Sutton and Vincent 1982) andthe model is implemented using GEMPACK software (Harrison Horridge Jerieand Pearson 2014)

An enhancement of importance to the present study is the inclusion of exchange-rate variables in the model which allow a distinction between price impacts asobserved in local currency units and those observed in 2014 US dollars

In the model the grape and wine sectors minimize costs of intermediate inputssubject to weak constant elasticity of substitution (CES) substitutability betweeninputs By assumption no intermediate inputs are imported from other countries9Hence

Xcid frac14 f ethX1id CESfrac12P

cid=P1id $THORN eth1THORN

P1id X1id frac14X

c

Xcid P

cid eth2THORN

whereXcid is the quantity demanded of commodity c by grape or wine industry i in

region d Pcid is the corresponding price and X1id and P1id are the respective inter-

mediate composite quantities and prices

Two primary factors are employed in the sector labor (the quantity of which isendogenous with perfectly elastic supply) and capital Capital is usually treated asexogenous in quantity with rates of return bearing all the adjustment in thevarious scenarios This reflects the fact that grapes (a perennial crop) and wine-

8Commercial-premium still wines are defined by Anderson and Nelgen (2011) to be those priced betweenUS$250 and $750 per litre pre-tax at a countryrsquos border or wholesale9An exception concerns bulk wine imports used mainly in the bottling of commercial premium wine in theUK USA Canada Japan and Germany

246 UK and Global Wine Markets by 2025 and Implications of Brexit

plant capacity adjust slowly to market signals

Lid frac14 f ethFid CESfrac12Wid=PFid $THORN eth3THORN

Kid frac14 f ethFid CESethRid=PFid $THORN eth4THORN

PFid Fid frac14 Lid Wid thorn Kid Rid eth5THORN

Grape and wine producers are assumed to minimize costs subject to CES substitu-tion between capital and labor Equations (3) to (5) show primary factor demandsfor the labor composite Lid and capital Kid subject to a composite factor demandFid by industry i in region d The factor prices are Wid for labor Rid for capitalrentals and PFid for composite factor prices

Appendix Table A1Cumulative Consumption and Population Growth Rates and Changes in the Real Exchange Rate

(RER) Relative to the US Dollar 2014 to 2025 without Brexit ()

Aggregateconsumption Poprsquon RER

Aggregateconsumption Poprsquon RER

France 18 4 minus11 Australia 35 11 minus17Italy 11 2 minus9 New Zealand 32 9 minus26Portugal 14 0 minus9 Canada 27 8 minus18Spain 26 8 minus9 United States 31 8 0Austria 19 4 minus7 Argentina 7 10 109Belgium 20 7 minus9 Brazil 16 8 minus29Denmark 22 2 minus9 Chile 55 8 minus2Finland 21 3 minus7 Mexico 42 12 minus8Germany 14 minus2 minus11 Uruguay 45 3 1Greece 22 minus1 minus14 Other Latin America 60 10 minus5Ireland 42 12 minus9 South Africa 36 12 minus1Netherlands 21 4 minus9 Turkey 50 8 20Sweden 24 9 minus13 North Africa 53 11 0Switzerland 18 8 minus6 Other Africa 109 18 84United Kingdom 32 6 1 Middle East 52 18 minus12Other West Europe 21 10 minus1 China 79 3 5Bulgaria 41 minus7 7 Hong Kong 42 3 2Croatia 20 minus2 minus1 India 134 13 17Georgia 35 0 23 Japan 11 minus3 minus24Hungary 25 minus3 minus11 Korea 38 1 minus9Moldova 49 minus11 13 Malaysia 62 15 minus16Romania 45 minus4 22 Philippines 75 18 7Russia 18 minus2 minus8 Singapore 44 21 minus22Ukraine 22 minus5 14 Taiwan 29 1 minus13Other East Europe 40 minus5 48 Thailand 47 3 minus9

Other Asia 99 10 10

Source Authorsrsquo compilation from projections by various international agencies and from global economy-wide modeling by Anderson andStrutt (2016)

Kym Anderson and Glyn Wittwer 247

The composite factor demand Fid is proportional to total output Qid subject to aprimary factor using technology Aid Hence

Fid frac14 Qid Aid eth6THORN

The perfectly competitive zero-pure-profit condition is that total revenue valued atthe output price P0s

i multiplied by Qid equals the total production cost

P0si Qid frac14

X

cPcid X1cid thorn

X

oWo

id Loid thorn Rid Kid eth7THORN

Household demands follow a linear expenditure system in each region We reducethe optimizing problem for household consumption of each commodity subjectto a budget constraint to equations describing subsistence and discretionarydemands Aggregate subsistence expenditure WSUBd depends only on consumerprices P3cd for each commodity and the number of households N as per-capita sub-sistence quantities XSUBcd subject to given preferences are constant

WSUBd frac14X

c

P3cd XSUBcd Nd eth8THORN

Discretionary expenditures for each commodity (the left-hand side of equation (9))are equal to the marginal budget share (βcd) of aggregate discretionary expenditureThis aggregate is the bracketed term on the right-hand side of equation (9) whereW3TOTd is aggregate nominal expenditure

P3cdethX3cd XSUBcd NdTHORN frac14 βcdethW3TOTd WSUBdTHORN eth9THORN

Because real aggregate consumption is usually exogenous in our partial equilibriumsimulations the linear expenditure system determines the consumption shares ofindividual final commodities (ie the five wine types plus a composite of all otherconsumption items) driven by changes in relative prices as faced by domestic con-sumers The income elasticity of demand for each commodity is equal to the mar-ginal budget share divided by the expenditure share This number varies from 05for nonpremium wine to 17 for superpremium still wine The income elasticity ofdemand for other consumption is very close to 10 because wine accounts for anaverage of only 03 of aggregate expenditures globally and no more than 11in any country (Anderson and Nelgen 2011 Table 166)

A feature of our revised model of world wine markets is the inclusion of nominalexchange rates These appear directly in the equation linking retail prices (P3scd ) toproducer prices by country of origin (P0s

c ) where c denotes the wine type

P3 scd frac14 P0s

cfd

fsTtarcd T

taxcd thorn Pm

cd eth10THORN

248 UK and Global Wine Markets by 2025 and Implications of Brexit

The exchange rates in the consuming and producing regions are fd and fs respec-tively expressed as local currency units per $US Ttar

cd is the power of the tariff in theconsuming region and Ttax

cd is the power of the domestic consumption (or excise) taxprior to any generic value-added or goods-and-services tax Pm

cd is the price of marginm assumed to be locally supplied nontradable and therefore unaffected by theexchange rate

A given level of consumption for wine type c (X3cd) is satisfied using theArmington (1969) assumption in which wines from different countries of originare imperfectly substitutable First domestic wine is imperfectly substitutable witha composite of imports

X3sscd frac14 f ethX3cd CESethP3sscd=P3cdTHORNTHORN ss frac14 domestic imports eth11THORN

and then imports by origin (X3scd ) are determined in a second CES equation

X3scd frac14 f ethX3ssfrac14importscd CESethP3scd=P3

ssfrac14importscd THORNTHORN eth12THORN

The model enables us to show how changes in international competitiveness affectthe worldrsquos wine markets A crucial part of this exercise is explaining how pricesdetermined outside the grape and wine markets influence these markets Becausethe model is in partial equilibrium to depict the impacts of changes in internationalcompetitiveness outside price changes need to be imposed as shocks on the modelThe price of intermediate inputs other than grapes and wine shown in equations (1)and (2) is set equal to the price of GDP (Pg

d ) multiplied by a shifter Fcd

Pcid frac14 Fc

dPgd eth13THORN

If no specific price observations are available the shifter Fcd remains exogenous and

unshocked with the change in price being determined by a shock to the price ofGDP If observations are available for specific input price movements the shifterFc

dbecomes endogenous with Pc

id now exogenous and shocked

Wid frac14 Fwd P

gd eth14THORN

Wage rates are treated similarly In equation (14) if the wage shifter Fwd is exogenous

changes in wage rates Wid are determined by changes in the price of GDP If wagerate data are available Fw

d becomes endogenous andwage rates are shocked directly

Pmcd frac14 Fm

d Pgd eth15THORN

The prices of trade and transport margins are also determined by the price of GDP ifthe shifter Fm

d in equation (15) is exogenous

Changes in international competitiveness depend on changes in relative pricelevels and changes in nominal exchange rates In equation (16) fR

s denotes real

Kym Anderson and Glyn Wittwer 249

exchange-rate movements relative to the US dollar

fRs frac14 Pg

s =frac12PgUSA

fs$ eth16THORN

In equation (16) the nominal exchange rate for the United States is alwaysunchanged because nominal and real exchange rates are expressed relative to theUS currency

Changes in international market conditions may have impacts in one direction onproducer prices as expressed in US dollars P0s

i and potentially in the opposite direc-tion in local currency units Hence we calculate real producer pricesP0s

iloc in localcurrency terms (ie the price most relevant to domestic producers)

P0siloc frac14 P0s

ifs=P

gs eth17THORN

To obtain real price changes in local currency terms we convert US dollar prices(P3 s

cd for the source-specific price and P3cd for the source-composite price) to reallocal currency prices (P3 s

cdloc and P3cdloc) using the consumer price index (CPI)(Pc

d ) as the deflator

P3 scdloc frac14 P3 s

cdfd=P

cd eth18THORN

and

P3cdloc frac14 P3cdfd=Pcd eth19THORN

References

Anderson K and Nelgen S (2011) Global Wine Markets 1961 to 2009 A StatisticalCompendium Adelaide University of Adelaide Press Also available as an e-book atwwwadelaideeduaupresstitlesglobal-wine

Anderson K Nelgen S and Pinilla V (2017) Global Wine Markets 1860 to 2016 AStatistical Compendium Adelaide University of Adelaide Press (forthcoming) Also tobe available as an e-book at wwwadelaideeduaupress

Anderson K and Pinilla V (with the assistance of A J Holmes) (2017)Annual Database ofGlobal Wine Markets 1835 to 2016 Wine Economics Research Centre University ofAdelaide to be posted in October at wwwadelaideeduauwine-econdatabases

Anderson K and Strutt A (2016) Impacts of Asiarsquos rise on African and Latin Americantrade Projections to 2030 World Economy 39(2) 172ndash194

Anderson K andWittwer G (2013) Modeling global wine markets to 2018 Exchange ratestaste changes and Chinarsquos import growth Journal of Wine Economics 8(2) 131ndash158

Anderson K and Wittwer G (2015) Asiarsquos evolving role in global wine markets ChinaEconomic Review 35 1ndash14

250 UK and Global Wine Markets by 2025 and Implications of Brexit

Armington P A (1969) A theory of demand for products distinguished by place of produc-tion IMF Staff Papers 16 159ndash178

Baldwin R E (ed) (2016) Brexit Beckons Thinking Ahead by Leading Economists LondonCentre for Economic Policy Research

Campos N F (2016) Lousy experts Looking back at the ex ante estimates of the costs ofBrexit In R E Baldwin (ed) Brexit Beckons Thinking Ahead by Leading Economists(ch 3) London Centre for Economic Policy Research

Dixon P Parmenter B Sutton J and Vincent D (1982) ORANI A Multisectoral Modelof the Australian Economy Contributions to Economic Analysis 142 Amsterdam North-Holland

European Commission (2015) Trade for All Towards a More Responsible Trade andInvestment Policy Brussels European Commission

Harrison J Horridge M Jerie M and Pearson K (2014) GEMPACK ManualMelbourne GEMPACK Software

HM Treasury (2017) Spring Budget 2017 London HM TreasuryHolmes A J and Anderson K (2017a)Annual Database of National Beverage Consumption

Volumes and Expenditures 1950 to 2015 Wine Economics Research Centre University ofAdelaide posted at wwwadelaideeduauwine-econdatabases

Holmes A J and Anderson K (2017b) Convergence in National Alcohol ConsumptionPatterns New Global Indicators Journal of Wine Economics 12(2) 117ndash148

Ludington C C (2013) The Politics of Wine in Britain A New Cultural History BasingstokeUK Palgrave Macmillan

Office of National Statistics (2017) Analysis of real earnings London Office of NationalStatistics Available at httpswwwonsgovukemploymentandlabourmarketpeopleinworkearningsandworkinghoursarticlessupplementaryanalysisofaverageweeklyearningslatest

Rollo J Borchert I Dawar K Holmes P and Winters L A (2016) The World TradeOrganisation A safety net for a post-Brexit UK trade policy Briefing Paper 1UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-1-final-1pdf

Smith A (2017) Brexit Hard truths and hard choices UKTPO blog 19 June Available athttpsblogssussexacukuktpo20170619brexit-hard-truths-and-hard-choices

Swinbank A (2017) World trade rules and the policy options for British agriculture post-Brexit Briefing Paper 7 UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-7pdf

Viner J (1950) The Customs Union Issue New York Carnegie Endowment for InternationalPeace

Wine Australia (2015) Export Market Guide European Union Adelaide Wine AustraliaWittwer G Berger N and Anderson K (2003) A model of the worldrsquos wine markets

Economic Modelling 20(3) 487ndash506

Kym Anderson and Glyn Wittwer 251

Page 24: U.K. and Global Wine Markets by 2025, and Implications of ... · U.K. and Global Wine Markets by 2025, and Implications of Brexit* Kym Andersona and Glyn Wittwerb Abstract The United

In short this subsequent step in the Brexit process can be expected to restore muchof the initial adverse effects in the United Kingdom of the Brexit vote by 2025 Thewine exporters in the EU27 Chile and South Africa also lose less in this scenarioas compared with the initial ldquosmallrdquo scenario while other key exporters sell nearlyas much in 2025 in this scenario as they are projected to in the baseline ndash that isthey are beneficiaries of the reduced discrimination in global wine markets in thiscase where Chile and South Africa no longer enjoy preferential access to the UKmarket (and even though EU27 producers are again allowed to do so) Keep inmind though that although the estimated losses in the initial scenarios are reportedjust for 2025 they are expected to be felt throughout the preceding years of uncertaintyTherefore the longer it is before uncertainty abates and the United Kingdomrsquos currentpreferential trading arrangements are replaced by newones such as a UK-EU27 FTAthe higher the cumulative cost of the Brexit vote to the wine trade

VII Caveats and Conclusions

The above Brexit simulations are just a few of many scenarios that could be modeledObvious additional ones could also assume that FTAs are reached between theUnited Kingdom and other trading partners including Chile and South Africaand such countries as Australia and New Zealand Some of these other countriesare already exploring the ECrsquos invitation (see European Commission 2015) to nego-tiate an FTA with what will be the EU27 The sequence in which FTAs are signedand the speed with which they are implemented will matter (as was also the casewith the sequential signing over the past decade of bilateral FTAs with NortheastAsian countries by Chile Australia and New Zealand see Anderson andWittwer 2015) Even if the United Kingdom were able to sign additional bilateralFTAs and begin implementing them before 2025 it would have little effect on theabove results (because wine tariffs are such a minor contributor to them) unlessthose FTAs were to accelerate the United Kingdomrsquos economic growth and therise in the British poundrsquos value before 2025

We assume above that no changes are made to alcohol excise duties in the UnitedKingdom following Brexit even though they are scheduled to be progressively raisedwith inflation and may be raised even more for wine relative to spirits to offset theopposite effects of Brexit on those two domestic industries Nor do we make anyallowance for UK increases in consumer wine prices that may be needed to coverthe higher cost of clearing customs on imports from EU27 countries (which maybe more or less offset by increased smuggling and duty-free purchasing)Importers of fine wines including individuals who buy Bordeaux en primeur maywell reduce their demand because of concerns that delays on docks will affect thequality of their wines or because of greater currency uncertainty

To summarize Brexit is costly initially to UK consumers of wine (and of manyother tradable products) because the domestic retail price in local currency tax-

244 UK and Global Wine Markets by 2025 and Implications of Brexit

inclusive terms is 22 higher than otherwise in 2025 and the volume of wine con-sumed domestically is 28 lower in the ldquolargerdquo Brexit scenario (or 11 higher pricesand 17 lower quantities in the ldquosmallrdquo case) by 2025 unless a UK-EU27 FTA isdeveloped before then Even if such an FTA does get signed ratified by all 28 par-liaments and implemented by 2025 the slower income growth in the interim meansa smaller UK wine market in 2025 than would otherwise have been the case Thevolume reduction is a blow to many participants in UK wine-bottling transportingstoring wholesaling and retailing businesses in addition to restaurants and pubsVery little of that initial impact occurs because of higher import tariffs most impor-tant is the assumed fall in UK real incomes

The small but growing number of local UK vignerons (who supply less than 05of all domestic-wine sales and less than 4 of sparkling-wine sales) is initially pro-jected to sell less wine domestically because of reduced demand and to export only alittle extra abroad while the British pound is devalued Overall their production isonly 3 higher in 2025 even in the ldquolargerdquo Brexit scenario (not taking intoaccount the fact that their casual labor is likely to be more expensive under Brexitbecause of tighter restrictions on immigration) and their production diminishes ifand when a UK-EU27 FTA is in place

Even with a UK-EU27 FTA in place EU27 wine exporters export US$56 millionless wine in 2025 thanks to Brexit Chile and South Africa export $88 million lesswine and wine exports from the rest of the world are little different in aggregatealthough slightly larger from Australia

Clearly there will be great uncertainly for some time over the possible policy out-comes to flow from Brexit and their consequent sequential impacts on UK house-hold disposable incomes foreign exchange rates and bilateral wine tariffsMeanwhile the above projections under explicit assumptions provide some idea ofhow wine markets might be affected by the most-likely first two stages of theBrexit process (agreeing on a new tariff schedule at the WTO and agreeing to andimplementing a UK-EU27 FTA) In particular they make clear that nontrivialinitial impacts could affect the domestic wine market impacts that are likely to belarger than just the direct impact of changes in bilateral tariffs If the UnitedKingdom succeeds in getting countries to agree to transition arrangements thatdelay the changes in tariffs until new FTAs are signed and ratified the initialeffects will be less dramatic than in our first scenarios but it remains to be seenwhether any such agreements can be reached In any event the net effect of Brexiton the welfare of the worldrsquos consumers and producers of wine as a whole will benegative not just initially but permanently unless new trade policy commitmentsby the United Kingdom with major wine-exporting countries are sufficiently moreliberal than current arrangements

Kym Anderson and Glyn Wittwer 245

Appendix Key Equations in the Global Wine Markets Model

Amodel of the worldrsquos wine markets was first published byWittwer et al (2003) andmuch improved by Anderson and Wittwer (2013) Several significant enhancementshave been made to that original model Wine types have been disaggregated intomore types ndash namely nonpremium (including generic bulk) commercial-premiumand superpremium still wines and sparkling wines8 As in the original modelthere are two types of grapes premium and nonpremium Nonpremium wines usenonpremium grapes exclusively superpremium wines use mainly premium grapesand commercial-premium and sparkling wines use both types of grapes In theregional dimension the number of countries and country groups has expandedfrom 10 in the original model to 51 44 individual nations and 7 compositeregions The modelrsquos database is calibrated to 2014 for this paper based on datain Anderson et al (2017) trade data for which are downloaded from httpscomtradeunorg The modelrsquos supply-and-demand equations are based on the orig-inal ORANI modelrsquos theory (see Dixon Parmenter Sutton and Vincent 1982) andthe model is implemented using GEMPACK software (Harrison Horridge Jerieand Pearson 2014)

An enhancement of importance to the present study is the inclusion of exchange-rate variables in the model which allow a distinction between price impacts asobserved in local currency units and those observed in 2014 US dollars

In the model the grape and wine sectors minimize costs of intermediate inputssubject to weak constant elasticity of substitution (CES) substitutability betweeninputs By assumption no intermediate inputs are imported from other countries9Hence

Xcid frac14 f ethX1id CESfrac12P

cid=P1id $THORN eth1THORN

P1id X1id frac14X

c

Xcid P

cid eth2THORN

whereXcid is the quantity demanded of commodity c by grape or wine industry i in

region d Pcid is the corresponding price and X1id and P1id are the respective inter-

mediate composite quantities and prices

Two primary factors are employed in the sector labor (the quantity of which isendogenous with perfectly elastic supply) and capital Capital is usually treated asexogenous in quantity with rates of return bearing all the adjustment in thevarious scenarios This reflects the fact that grapes (a perennial crop) and wine-

8Commercial-premium still wines are defined by Anderson and Nelgen (2011) to be those priced betweenUS$250 and $750 per litre pre-tax at a countryrsquos border or wholesale9An exception concerns bulk wine imports used mainly in the bottling of commercial premium wine in theUK USA Canada Japan and Germany

246 UK and Global Wine Markets by 2025 and Implications of Brexit

plant capacity adjust slowly to market signals

Lid frac14 f ethFid CESfrac12Wid=PFid $THORN eth3THORN

Kid frac14 f ethFid CESethRid=PFid $THORN eth4THORN

PFid Fid frac14 Lid Wid thorn Kid Rid eth5THORN

Grape and wine producers are assumed to minimize costs subject to CES substitu-tion between capital and labor Equations (3) to (5) show primary factor demandsfor the labor composite Lid and capital Kid subject to a composite factor demandFid by industry i in region d The factor prices are Wid for labor Rid for capitalrentals and PFid for composite factor prices

Appendix Table A1Cumulative Consumption and Population Growth Rates and Changes in the Real Exchange Rate

(RER) Relative to the US Dollar 2014 to 2025 without Brexit ()

Aggregateconsumption Poprsquon RER

Aggregateconsumption Poprsquon RER

France 18 4 minus11 Australia 35 11 minus17Italy 11 2 minus9 New Zealand 32 9 minus26Portugal 14 0 minus9 Canada 27 8 minus18Spain 26 8 minus9 United States 31 8 0Austria 19 4 minus7 Argentina 7 10 109Belgium 20 7 minus9 Brazil 16 8 minus29Denmark 22 2 minus9 Chile 55 8 minus2Finland 21 3 minus7 Mexico 42 12 minus8Germany 14 minus2 minus11 Uruguay 45 3 1Greece 22 minus1 minus14 Other Latin America 60 10 minus5Ireland 42 12 minus9 South Africa 36 12 minus1Netherlands 21 4 minus9 Turkey 50 8 20Sweden 24 9 minus13 North Africa 53 11 0Switzerland 18 8 minus6 Other Africa 109 18 84United Kingdom 32 6 1 Middle East 52 18 minus12Other West Europe 21 10 minus1 China 79 3 5Bulgaria 41 minus7 7 Hong Kong 42 3 2Croatia 20 minus2 minus1 India 134 13 17Georgia 35 0 23 Japan 11 minus3 minus24Hungary 25 minus3 minus11 Korea 38 1 minus9Moldova 49 minus11 13 Malaysia 62 15 minus16Romania 45 minus4 22 Philippines 75 18 7Russia 18 minus2 minus8 Singapore 44 21 minus22Ukraine 22 minus5 14 Taiwan 29 1 minus13Other East Europe 40 minus5 48 Thailand 47 3 minus9

Other Asia 99 10 10

Source Authorsrsquo compilation from projections by various international agencies and from global economy-wide modeling by Anderson andStrutt (2016)

Kym Anderson and Glyn Wittwer 247

The composite factor demand Fid is proportional to total output Qid subject to aprimary factor using technology Aid Hence

Fid frac14 Qid Aid eth6THORN

The perfectly competitive zero-pure-profit condition is that total revenue valued atthe output price P0s

i multiplied by Qid equals the total production cost

P0si Qid frac14

X

cPcid X1cid thorn

X

oWo

id Loid thorn Rid Kid eth7THORN

Household demands follow a linear expenditure system in each region We reducethe optimizing problem for household consumption of each commodity subjectto a budget constraint to equations describing subsistence and discretionarydemands Aggregate subsistence expenditure WSUBd depends only on consumerprices P3cd for each commodity and the number of households N as per-capita sub-sistence quantities XSUBcd subject to given preferences are constant

WSUBd frac14X

c

P3cd XSUBcd Nd eth8THORN

Discretionary expenditures for each commodity (the left-hand side of equation (9))are equal to the marginal budget share (βcd) of aggregate discretionary expenditureThis aggregate is the bracketed term on the right-hand side of equation (9) whereW3TOTd is aggregate nominal expenditure

P3cdethX3cd XSUBcd NdTHORN frac14 βcdethW3TOTd WSUBdTHORN eth9THORN

Because real aggregate consumption is usually exogenous in our partial equilibriumsimulations the linear expenditure system determines the consumption shares ofindividual final commodities (ie the five wine types plus a composite of all otherconsumption items) driven by changes in relative prices as faced by domestic con-sumers The income elasticity of demand for each commodity is equal to the mar-ginal budget share divided by the expenditure share This number varies from 05for nonpremium wine to 17 for superpremium still wine The income elasticity ofdemand for other consumption is very close to 10 because wine accounts for anaverage of only 03 of aggregate expenditures globally and no more than 11in any country (Anderson and Nelgen 2011 Table 166)

A feature of our revised model of world wine markets is the inclusion of nominalexchange rates These appear directly in the equation linking retail prices (P3scd ) toproducer prices by country of origin (P0s

c ) where c denotes the wine type

P3 scd frac14 P0s

cfd

fsTtarcd T

taxcd thorn Pm

cd eth10THORN

248 UK and Global Wine Markets by 2025 and Implications of Brexit

The exchange rates in the consuming and producing regions are fd and fs respec-tively expressed as local currency units per $US Ttar

cd is the power of the tariff in theconsuming region and Ttax

cd is the power of the domestic consumption (or excise) taxprior to any generic value-added or goods-and-services tax Pm

cd is the price of marginm assumed to be locally supplied nontradable and therefore unaffected by theexchange rate

A given level of consumption for wine type c (X3cd) is satisfied using theArmington (1969) assumption in which wines from different countries of originare imperfectly substitutable First domestic wine is imperfectly substitutable witha composite of imports

X3sscd frac14 f ethX3cd CESethP3sscd=P3cdTHORNTHORN ss frac14 domestic imports eth11THORN

and then imports by origin (X3scd ) are determined in a second CES equation

X3scd frac14 f ethX3ssfrac14importscd CESethP3scd=P3

ssfrac14importscd THORNTHORN eth12THORN

The model enables us to show how changes in international competitiveness affectthe worldrsquos wine markets A crucial part of this exercise is explaining how pricesdetermined outside the grape and wine markets influence these markets Becausethe model is in partial equilibrium to depict the impacts of changes in internationalcompetitiveness outside price changes need to be imposed as shocks on the modelThe price of intermediate inputs other than grapes and wine shown in equations (1)and (2) is set equal to the price of GDP (Pg

d ) multiplied by a shifter Fcd

Pcid frac14 Fc

dPgd eth13THORN

If no specific price observations are available the shifter Fcd remains exogenous and

unshocked with the change in price being determined by a shock to the price ofGDP If observations are available for specific input price movements the shifterFc

dbecomes endogenous with Pc

id now exogenous and shocked

Wid frac14 Fwd P

gd eth14THORN

Wage rates are treated similarly In equation (14) if the wage shifter Fwd is exogenous

changes in wage rates Wid are determined by changes in the price of GDP If wagerate data are available Fw

d becomes endogenous andwage rates are shocked directly

Pmcd frac14 Fm

d Pgd eth15THORN

The prices of trade and transport margins are also determined by the price of GDP ifthe shifter Fm

d in equation (15) is exogenous

Changes in international competitiveness depend on changes in relative pricelevels and changes in nominal exchange rates In equation (16) fR

s denotes real

Kym Anderson and Glyn Wittwer 249

exchange-rate movements relative to the US dollar

fRs frac14 Pg

s =frac12PgUSA

fs$ eth16THORN

In equation (16) the nominal exchange rate for the United States is alwaysunchanged because nominal and real exchange rates are expressed relative to theUS currency

Changes in international market conditions may have impacts in one direction onproducer prices as expressed in US dollars P0s

i and potentially in the opposite direc-tion in local currency units Hence we calculate real producer pricesP0s

iloc in localcurrency terms (ie the price most relevant to domestic producers)

P0siloc frac14 P0s

ifs=P

gs eth17THORN

To obtain real price changes in local currency terms we convert US dollar prices(P3 s

cd for the source-specific price and P3cd for the source-composite price) to reallocal currency prices (P3 s

cdloc and P3cdloc) using the consumer price index (CPI)(Pc

d ) as the deflator

P3 scdloc frac14 P3 s

cdfd=P

cd eth18THORN

and

P3cdloc frac14 P3cdfd=Pcd eth19THORN

References

Anderson K and Nelgen S (2011) Global Wine Markets 1961 to 2009 A StatisticalCompendium Adelaide University of Adelaide Press Also available as an e-book atwwwadelaideeduaupresstitlesglobal-wine

Anderson K Nelgen S and Pinilla V (2017) Global Wine Markets 1860 to 2016 AStatistical Compendium Adelaide University of Adelaide Press (forthcoming) Also tobe available as an e-book at wwwadelaideeduaupress

Anderson K and Pinilla V (with the assistance of A J Holmes) (2017)Annual Database ofGlobal Wine Markets 1835 to 2016 Wine Economics Research Centre University ofAdelaide to be posted in October at wwwadelaideeduauwine-econdatabases

Anderson K and Strutt A (2016) Impacts of Asiarsquos rise on African and Latin Americantrade Projections to 2030 World Economy 39(2) 172ndash194

Anderson K andWittwer G (2013) Modeling global wine markets to 2018 Exchange ratestaste changes and Chinarsquos import growth Journal of Wine Economics 8(2) 131ndash158

Anderson K and Wittwer G (2015) Asiarsquos evolving role in global wine markets ChinaEconomic Review 35 1ndash14

250 UK and Global Wine Markets by 2025 and Implications of Brexit

Armington P A (1969) A theory of demand for products distinguished by place of produc-tion IMF Staff Papers 16 159ndash178

Baldwin R E (ed) (2016) Brexit Beckons Thinking Ahead by Leading Economists LondonCentre for Economic Policy Research

Campos N F (2016) Lousy experts Looking back at the ex ante estimates of the costs ofBrexit In R E Baldwin (ed) Brexit Beckons Thinking Ahead by Leading Economists(ch 3) London Centre for Economic Policy Research

Dixon P Parmenter B Sutton J and Vincent D (1982) ORANI A Multisectoral Modelof the Australian Economy Contributions to Economic Analysis 142 Amsterdam North-Holland

European Commission (2015) Trade for All Towards a More Responsible Trade andInvestment Policy Brussels European Commission

Harrison J Horridge M Jerie M and Pearson K (2014) GEMPACK ManualMelbourne GEMPACK Software

HM Treasury (2017) Spring Budget 2017 London HM TreasuryHolmes A J and Anderson K (2017a)Annual Database of National Beverage Consumption

Volumes and Expenditures 1950 to 2015 Wine Economics Research Centre University ofAdelaide posted at wwwadelaideeduauwine-econdatabases

Holmes A J and Anderson K (2017b) Convergence in National Alcohol ConsumptionPatterns New Global Indicators Journal of Wine Economics 12(2) 117ndash148

Ludington C C (2013) The Politics of Wine in Britain A New Cultural History BasingstokeUK Palgrave Macmillan

Office of National Statistics (2017) Analysis of real earnings London Office of NationalStatistics Available at httpswwwonsgovukemploymentandlabourmarketpeopleinworkearningsandworkinghoursarticlessupplementaryanalysisofaverageweeklyearningslatest

Rollo J Borchert I Dawar K Holmes P and Winters L A (2016) The World TradeOrganisation A safety net for a post-Brexit UK trade policy Briefing Paper 1UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-1-final-1pdf

Smith A (2017) Brexit Hard truths and hard choices UKTPO blog 19 June Available athttpsblogssussexacukuktpo20170619brexit-hard-truths-and-hard-choices

Swinbank A (2017) World trade rules and the policy options for British agriculture post-Brexit Briefing Paper 7 UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-7pdf

Viner J (1950) The Customs Union Issue New York Carnegie Endowment for InternationalPeace

Wine Australia (2015) Export Market Guide European Union Adelaide Wine AustraliaWittwer G Berger N and Anderson K (2003) A model of the worldrsquos wine markets

Economic Modelling 20(3) 487ndash506

Kym Anderson and Glyn Wittwer 251

Page 25: U.K. and Global Wine Markets by 2025, and Implications of ... · U.K. and Global Wine Markets by 2025, and Implications of Brexit* Kym Andersona and Glyn Wittwerb Abstract The United

inclusive terms is 22 higher than otherwise in 2025 and the volume of wine con-sumed domestically is 28 lower in the ldquolargerdquo Brexit scenario (or 11 higher pricesand 17 lower quantities in the ldquosmallrdquo case) by 2025 unless a UK-EU27 FTA isdeveloped before then Even if such an FTA does get signed ratified by all 28 par-liaments and implemented by 2025 the slower income growth in the interim meansa smaller UK wine market in 2025 than would otherwise have been the case Thevolume reduction is a blow to many participants in UK wine-bottling transportingstoring wholesaling and retailing businesses in addition to restaurants and pubsVery little of that initial impact occurs because of higher import tariffs most impor-tant is the assumed fall in UK real incomes

The small but growing number of local UK vignerons (who supply less than 05of all domestic-wine sales and less than 4 of sparkling-wine sales) is initially pro-jected to sell less wine domestically because of reduced demand and to export only alittle extra abroad while the British pound is devalued Overall their production isonly 3 higher in 2025 even in the ldquolargerdquo Brexit scenario (not taking intoaccount the fact that their casual labor is likely to be more expensive under Brexitbecause of tighter restrictions on immigration) and their production diminishes ifand when a UK-EU27 FTA is in place

Even with a UK-EU27 FTA in place EU27 wine exporters export US$56 millionless wine in 2025 thanks to Brexit Chile and South Africa export $88 million lesswine and wine exports from the rest of the world are little different in aggregatealthough slightly larger from Australia

Clearly there will be great uncertainly for some time over the possible policy out-comes to flow from Brexit and their consequent sequential impacts on UK house-hold disposable incomes foreign exchange rates and bilateral wine tariffsMeanwhile the above projections under explicit assumptions provide some idea ofhow wine markets might be affected by the most-likely first two stages of theBrexit process (agreeing on a new tariff schedule at the WTO and agreeing to andimplementing a UK-EU27 FTA) In particular they make clear that nontrivialinitial impacts could affect the domestic wine market impacts that are likely to belarger than just the direct impact of changes in bilateral tariffs If the UnitedKingdom succeeds in getting countries to agree to transition arrangements thatdelay the changes in tariffs until new FTAs are signed and ratified the initialeffects will be less dramatic than in our first scenarios but it remains to be seenwhether any such agreements can be reached In any event the net effect of Brexiton the welfare of the worldrsquos consumers and producers of wine as a whole will benegative not just initially but permanently unless new trade policy commitmentsby the United Kingdom with major wine-exporting countries are sufficiently moreliberal than current arrangements

Kym Anderson and Glyn Wittwer 245

Appendix Key Equations in the Global Wine Markets Model

Amodel of the worldrsquos wine markets was first published byWittwer et al (2003) andmuch improved by Anderson and Wittwer (2013) Several significant enhancementshave been made to that original model Wine types have been disaggregated intomore types ndash namely nonpremium (including generic bulk) commercial-premiumand superpremium still wines and sparkling wines8 As in the original modelthere are two types of grapes premium and nonpremium Nonpremium wines usenonpremium grapes exclusively superpremium wines use mainly premium grapesand commercial-premium and sparkling wines use both types of grapes In theregional dimension the number of countries and country groups has expandedfrom 10 in the original model to 51 44 individual nations and 7 compositeregions The modelrsquos database is calibrated to 2014 for this paper based on datain Anderson et al (2017) trade data for which are downloaded from httpscomtradeunorg The modelrsquos supply-and-demand equations are based on the orig-inal ORANI modelrsquos theory (see Dixon Parmenter Sutton and Vincent 1982) andthe model is implemented using GEMPACK software (Harrison Horridge Jerieand Pearson 2014)

An enhancement of importance to the present study is the inclusion of exchange-rate variables in the model which allow a distinction between price impacts asobserved in local currency units and those observed in 2014 US dollars

In the model the grape and wine sectors minimize costs of intermediate inputssubject to weak constant elasticity of substitution (CES) substitutability betweeninputs By assumption no intermediate inputs are imported from other countries9Hence

Xcid frac14 f ethX1id CESfrac12P

cid=P1id $THORN eth1THORN

P1id X1id frac14X

c

Xcid P

cid eth2THORN

whereXcid is the quantity demanded of commodity c by grape or wine industry i in

region d Pcid is the corresponding price and X1id and P1id are the respective inter-

mediate composite quantities and prices

Two primary factors are employed in the sector labor (the quantity of which isendogenous with perfectly elastic supply) and capital Capital is usually treated asexogenous in quantity with rates of return bearing all the adjustment in thevarious scenarios This reflects the fact that grapes (a perennial crop) and wine-

8Commercial-premium still wines are defined by Anderson and Nelgen (2011) to be those priced betweenUS$250 and $750 per litre pre-tax at a countryrsquos border or wholesale9An exception concerns bulk wine imports used mainly in the bottling of commercial premium wine in theUK USA Canada Japan and Germany

246 UK and Global Wine Markets by 2025 and Implications of Brexit

plant capacity adjust slowly to market signals

Lid frac14 f ethFid CESfrac12Wid=PFid $THORN eth3THORN

Kid frac14 f ethFid CESethRid=PFid $THORN eth4THORN

PFid Fid frac14 Lid Wid thorn Kid Rid eth5THORN

Grape and wine producers are assumed to minimize costs subject to CES substitu-tion between capital and labor Equations (3) to (5) show primary factor demandsfor the labor composite Lid and capital Kid subject to a composite factor demandFid by industry i in region d The factor prices are Wid for labor Rid for capitalrentals and PFid for composite factor prices

Appendix Table A1Cumulative Consumption and Population Growth Rates and Changes in the Real Exchange Rate

(RER) Relative to the US Dollar 2014 to 2025 without Brexit ()

Aggregateconsumption Poprsquon RER

Aggregateconsumption Poprsquon RER

France 18 4 minus11 Australia 35 11 minus17Italy 11 2 minus9 New Zealand 32 9 minus26Portugal 14 0 minus9 Canada 27 8 minus18Spain 26 8 minus9 United States 31 8 0Austria 19 4 minus7 Argentina 7 10 109Belgium 20 7 minus9 Brazil 16 8 minus29Denmark 22 2 minus9 Chile 55 8 minus2Finland 21 3 minus7 Mexico 42 12 minus8Germany 14 minus2 minus11 Uruguay 45 3 1Greece 22 minus1 minus14 Other Latin America 60 10 minus5Ireland 42 12 minus9 South Africa 36 12 minus1Netherlands 21 4 minus9 Turkey 50 8 20Sweden 24 9 minus13 North Africa 53 11 0Switzerland 18 8 minus6 Other Africa 109 18 84United Kingdom 32 6 1 Middle East 52 18 minus12Other West Europe 21 10 minus1 China 79 3 5Bulgaria 41 minus7 7 Hong Kong 42 3 2Croatia 20 minus2 minus1 India 134 13 17Georgia 35 0 23 Japan 11 minus3 minus24Hungary 25 minus3 minus11 Korea 38 1 minus9Moldova 49 minus11 13 Malaysia 62 15 minus16Romania 45 minus4 22 Philippines 75 18 7Russia 18 minus2 minus8 Singapore 44 21 minus22Ukraine 22 minus5 14 Taiwan 29 1 minus13Other East Europe 40 minus5 48 Thailand 47 3 minus9

Other Asia 99 10 10

Source Authorsrsquo compilation from projections by various international agencies and from global economy-wide modeling by Anderson andStrutt (2016)

Kym Anderson and Glyn Wittwer 247

The composite factor demand Fid is proportional to total output Qid subject to aprimary factor using technology Aid Hence

Fid frac14 Qid Aid eth6THORN

The perfectly competitive zero-pure-profit condition is that total revenue valued atthe output price P0s

i multiplied by Qid equals the total production cost

P0si Qid frac14

X

cPcid X1cid thorn

X

oWo

id Loid thorn Rid Kid eth7THORN

Household demands follow a linear expenditure system in each region We reducethe optimizing problem for household consumption of each commodity subjectto a budget constraint to equations describing subsistence and discretionarydemands Aggregate subsistence expenditure WSUBd depends only on consumerprices P3cd for each commodity and the number of households N as per-capita sub-sistence quantities XSUBcd subject to given preferences are constant

WSUBd frac14X

c

P3cd XSUBcd Nd eth8THORN

Discretionary expenditures for each commodity (the left-hand side of equation (9))are equal to the marginal budget share (βcd) of aggregate discretionary expenditureThis aggregate is the bracketed term on the right-hand side of equation (9) whereW3TOTd is aggregate nominal expenditure

P3cdethX3cd XSUBcd NdTHORN frac14 βcdethW3TOTd WSUBdTHORN eth9THORN

Because real aggregate consumption is usually exogenous in our partial equilibriumsimulations the linear expenditure system determines the consumption shares ofindividual final commodities (ie the five wine types plus a composite of all otherconsumption items) driven by changes in relative prices as faced by domestic con-sumers The income elasticity of demand for each commodity is equal to the mar-ginal budget share divided by the expenditure share This number varies from 05for nonpremium wine to 17 for superpremium still wine The income elasticity ofdemand for other consumption is very close to 10 because wine accounts for anaverage of only 03 of aggregate expenditures globally and no more than 11in any country (Anderson and Nelgen 2011 Table 166)

A feature of our revised model of world wine markets is the inclusion of nominalexchange rates These appear directly in the equation linking retail prices (P3scd ) toproducer prices by country of origin (P0s

c ) where c denotes the wine type

P3 scd frac14 P0s

cfd

fsTtarcd T

taxcd thorn Pm

cd eth10THORN

248 UK and Global Wine Markets by 2025 and Implications of Brexit

The exchange rates in the consuming and producing regions are fd and fs respec-tively expressed as local currency units per $US Ttar

cd is the power of the tariff in theconsuming region and Ttax

cd is the power of the domestic consumption (or excise) taxprior to any generic value-added or goods-and-services tax Pm

cd is the price of marginm assumed to be locally supplied nontradable and therefore unaffected by theexchange rate

A given level of consumption for wine type c (X3cd) is satisfied using theArmington (1969) assumption in which wines from different countries of originare imperfectly substitutable First domestic wine is imperfectly substitutable witha composite of imports

X3sscd frac14 f ethX3cd CESethP3sscd=P3cdTHORNTHORN ss frac14 domestic imports eth11THORN

and then imports by origin (X3scd ) are determined in a second CES equation

X3scd frac14 f ethX3ssfrac14importscd CESethP3scd=P3

ssfrac14importscd THORNTHORN eth12THORN

The model enables us to show how changes in international competitiveness affectthe worldrsquos wine markets A crucial part of this exercise is explaining how pricesdetermined outside the grape and wine markets influence these markets Becausethe model is in partial equilibrium to depict the impacts of changes in internationalcompetitiveness outside price changes need to be imposed as shocks on the modelThe price of intermediate inputs other than grapes and wine shown in equations (1)and (2) is set equal to the price of GDP (Pg

d ) multiplied by a shifter Fcd

Pcid frac14 Fc

dPgd eth13THORN

If no specific price observations are available the shifter Fcd remains exogenous and

unshocked with the change in price being determined by a shock to the price ofGDP If observations are available for specific input price movements the shifterFc

dbecomes endogenous with Pc

id now exogenous and shocked

Wid frac14 Fwd P

gd eth14THORN

Wage rates are treated similarly In equation (14) if the wage shifter Fwd is exogenous

changes in wage rates Wid are determined by changes in the price of GDP If wagerate data are available Fw

d becomes endogenous andwage rates are shocked directly

Pmcd frac14 Fm

d Pgd eth15THORN

The prices of trade and transport margins are also determined by the price of GDP ifthe shifter Fm

d in equation (15) is exogenous

Changes in international competitiveness depend on changes in relative pricelevels and changes in nominal exchange rates In equation (16) fR

s denotes real

Kym Anderson and Glyn Wittwer 249

exchange-rate movements relative to the US dollar

fRs frac14 Pg

s =frac12PgUSA

fs$ eth16THORN

In equation (16) the nominal exchange rate for the United States is alwaysunchanged because nominal and real exchange rates are expressed relative to theUS currency

Changes in international market conditions may have impacts in one direction onproducer prices as expressed in US dollars P0s

i and potentially in the opposite direc-tion in local currency units Hence we calculate real producer pricesP0s

iloc in localcurrency terms (ie the price most relevant to domestic producers)

P0siloc frac14 P0s

ifs=P

gs eth17THORN

To obtain real price changes in local currency terms we convert US dollar prices(P3 s

cd for the source-specific price and P3cd for the source-composite price) to reallocal currency prices (P3 s

cdloc and P3cdloc) using the consumer price index (CPI)(Pc

d ) as the deflator

P3 scdloc frac14 P3 s

cdfd=P

cd eth18THORN

and

P3cdloc frac14 P3cdfd=Pcd eth19THORN

References

Anderson K and Nelgen S (2011) Global Wine Markets 1961 to 2009 A StatisticalCompendium Adelaide University of Adelaide Press Also available as an e-book atwwwadelaideeduaupresstitlesglobal-wine

Anderson K Nelgen S and Pinilla V (2017) Global Wine Markets 1860 to 2016 AStatistical Compendium Adelaide University of Adelaide Press (forthcoming) Also tobe available as an e-book at wwwadelaideeduaupress

Anderson K and Pinilla V (with the assistance of A J Holmes) (2017)Annual Database ofGlobal Wine Markets 1835 to 2016 Wine Economics Research Centre University ofAdelaide to be posted in October at wwwadelaideeduauwine-econdatabases

Anderson K and Strutt A (2016) Impacts of Asiarsquos rise on African and Latin Americantrade Projections to 2030 World Economy 39(2) 172ndash194

Anderson K andWittwer G (2013) Modeling global wine markets to 2018 Exchange ratestaste changes and Chinarsquos import growth Journal of Wine Economics 8(2) 131ndash158

Anderson K and Wittwer G (2015) Asiarsquos evolving role in global wine markets ChinaEconomic Review 35 1ndash14

250 UK and Global Wine Markets by 2025 and Implications of Brexit

Armington P A (1969) A theory of demand for products distinguished by place of produc-tion IMF Staff Papers 16 159ndash178

Baldwin R E (ed) (2016) Brexit Beckons Thinking Ahead by Leading Economists LondonCentre for Economic Policy Research

Campos N F (2016) Lousy experts Looking back at the ex ante estimates of the costs ofBrexit In R E Baldwin (ed) Brexit Beckons Thinking Ahead by Leading Economists(ch 3) London Centre for Economic Policy Research

Dixon P Parmenter B Sutton J and Vincent D (1982) ORANI A Multisectoral Modelof the Australian Economy Contributions to Economic Analysis 142 Amsterdam North-Holland

European Commission (2015) Trade for All Towards a More Responsible Trade andInvestment Policy Brussels European Commission

Harrison J Horridge M Jerie M and Pearson K (2014) GEMPACK ManualMelbourne GEMPACK Software

HM Treasury (2017) Spring Budget 2017 London HM TreasuryHolmes A J and Anderson K (2017a)Annual Database of National Beverage Consumption

Volumes and Expenditures 1950 to 2015 Wine Economics Research Centre University ofAdelaide posted at wwwadelaideeduauwine-econdatabases

Holmes A J and Anderson K (2017b) Convergence in National Alcohol ConsumptionPatterns New Global Indicators Journal of Wine Economics 12(2) 117ndash148

Ludington C C (2013) The Politics of Wine in Britain A New Cultural History BasingstokeUK Palgrave Macmillan

Office of National Statistics (2017) Analysis of real earnings London Office of NationalStatistics Available at httpswwwonsgovukemploymentandlabourmarketpeopleinworkearningsandworkinghoursarticlessupplementaryanalysisofaverageweeklyearningslatest

Rollo J Borchert I Dawar K Holmes P and Winters L A (2016) The World TradeOrganisation A safety net for a post-Brexit UK trade policy Briefing Paper 1UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-1-final-1pdf

Smith A (2017) Brexit Hard truths and hard choices UKTPO blog 19 June Available athttpsblogssussexacukuktpo20170619brexit-hard-truths-and-hard-choices

Swinbank A (2017) World trade rules and the policy options for British agriculture post-Brexit Briefing Paper 7 UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-7pdf

Viner J (1950) The Customs Union Issue New York Carnegie Endowment for InternationalPeace

Wine Australia (2015) Export Market Guide European Union Adelaide Wine AustraliaWittwer G Berger N and Anderson K (2003) A model of the worldrsquos wine markets

Economic Modelling 20(3) 487ndash506

Kym Anderson and Glyn Wittwer 251

Page 26: U.K. and Global Wine Markets by 2025, and Implications of ... · U.K. and Global Wine Markets by 2025, and Implications of Brexit* Kym Andersona and Glyn Wittwerb Abstract The United

Appendix Key Equations in the Global Wine Markets Model

Amodel of the worldrsquos wine markets was first published byWittwer et al (2003) andmuch improved by Anderson and Wittwer (2013) Several significant enhancementshave been made to that original model Wine types have been disaggregated intomore types ndash namely nonpremium (including generic bulk) commercial-premiumand superpremium still wines and sparkling wines8 As in the original modelthere are two types of grapes premium and nonpremium Nonpremium wines usenonpremium grapes exclusively superpremium wines use mainly premium grapesand commercial-premium and sparkling wines use both types of grapes In theregional dimension the number of countries and country groups has expandedfrom 10 in the original model to 51 44 individual nations and 7 compositeregions The modelrsquos database is calibrated to 2014 for this paper based on datain Anderson et al (2017) trade data for which are downloaded from httpscomtradeunorg The modelrsquos supply-and-demand equations are based on the orig-inal ORANI modelrsquos theory (see Dixon Parmenter Sutton and Vincent 1982) andthe model is implemented using GEMPACK software (Harrison Horridge Jerieand Pearson 2014)

An enhancement of importance to the present study is the inclusion of exchange-rate variables in the model which allow a distinction between price impacts asobserved in local currency units and those observed in 2014 US dollars

In the model the grape and wine sectors minimize costs of intermediate inputssubject to weak constant elasticity of substitution (CES) substitutability betweeninputs By assumption no intermediate inputs are imported from other countries9Hence

Xcid frac14 f ethX1id CESfrac12P

cid=P1id $THORN eth1THORN

P1id X1id frac14X

c

Xcid P

cid eth2THORN

whereXcid is the quantity demanded of commodity c by grape or wine industry i in

region d Pcid is the corresponding price and X1id and P1id are the respective inter-

mediate composite quantities and prices

Two primary factors are employed in the sector labor (the quantity of which isendogenous with perfectly elastic supply) and capital Capital is usually treated asexogenous in quantity with rates of return bearing all the adjustment in thevarious scenarios This reflects the fact that grapes (a perennial crop) and wine-

8Commercial-premium still wines are defined by Anderson and Nelgen (2011) to be those priced betweenUS$250 and $750 per litre pre-tax at a countryrsquos border or wholesale9An exception concerns bulk wine imports used mainly in the bottling of commercial premium wine in theUK USA Canada Japan and Germany

246 UK and Global Wine Markets by 2025 and Implications of Brexit

plant capacity adjust slowly to market signals

Lid frac14 f ethFid CESfrac12Wid=PFid $THORN eth3THORN

Kid frac14 f ethFid CESethRid=PFid $THORN eth4THORN

PFid Fid frac14 Lid Wid thorn Kid Rid eth5THORN

Grape and wine producers are assumed to minimize costs subject to CES substitu-tion between capital and labor Equations (3) to (5) show primary factor demandsfor the labor composite Lid and capital Kid subject to a composite factor demandFid by industry i in region d The factor prices are Wid for labor Rid for capitalrentals and PFid for composite factor prices

Appendix Table A1Cumulative Consumption and Population Growth Rates and Changes in the Real Exchange Rate

(RER) Relative to the US Dollar 2014 to 2025 without Brexit ()

Aggregateconsumption Poprsquon RER

Aggregateconsumption Poprsquon RER

France 18 4 minus11 Australia 35 11 minus17Italy 11 2 minus9 New Zealand 32 9 minus26Portugal 14 0 minus9 Canada 27 8 minus18Spain 26 8 minus9 United States 31 8 0Austria 19 4 minus7 Argentina 7 10 109Belgium 20 7 minus9 Brazil 16 8 minus29Denmark 22 2 minus9 Chile 55 8 minus2Finland 21 3 minus7 Mexico 42 12 minus8Germany 14 minus2 minus11 Uruguay 45 3 1Greece 22 minus1 minus14 Other Latin America 60 10 minus5Ireland 42 12 minus9 South Africa 36 12 minus1Netherlands 21 4 minus9 Turkey 50 8 20Sweden 24 9 minus13 North Africa 53 11 0Switzerland 18 8 minus6 Other Africa 109 18 84United Kingdom 32 6 1 Middle East 52 18 minus12Other West Europe 21 10 minus1 China 79 3 5Bulgaria 41 minus7 7 Hong Kong 42 3 2Croatia 20 minus2 minus1 India 134 13 17Georgia 35 0 23 Japan 11 minus3 minus24Hungary 25 minus3 minus11 Korea 38 1 minus9Moldova 49 minus11 13 Malaysia 62 15 minus16Romania 45 minus4 22 Philippines 75 18 7Russia 18 minus2 minus8 Singapore 44 21 minus22Ukraine 22 minus5 14 Taiwan 29 1 minus13Other East Europe 40 minus5 48 Thailand 47 3 minus9

Other Asia 99 10 10

Source Authorsrsquo compilation from projections by various international agencies and from global economy-wide modeling by Anderson andStrutt (2016)

Kym Anderson and Glyn Wittwer 247

The composite factor demand Fid is proportional to total output Qid subject to aprimary factor using technology Aid Hence

Fid frac14 Qid Aid eth6THORN

The perfectly competitive zero-pure-profit condition is that total revenue valued atthe output price P0s

i multiplied by Qid equals the total production cost

P0si Qid frac14

X

cPcid X1cid thorn

X

oWo

id Loid thorn Rid Kid eth7THORN

Household demands follow a linear expenditure system in each region We reducethe optimizing problem for household consumption of each commodity subjectto a budget constraint to equations describing subsistence and discretionarydemands Aggregate subsistence expenditure WSUBd depends only on consumerprices P3cd for each commodity and the number of households N as per-capita sub-sistence quantities XSUBcd subject to given preferences are constant

WSUBd frac14X

c

P3cd XSUBcd Nd eth8THORN

Discretionary expenditures for each commodity (the left-hand side of equation (9))are equal to the marginal budget share (βcd) of aggregate discretionary expenditureThis aggregate is the bracketed term on the right-hand side of equation (9) whereW3TOTd is aggregate nominal expenditure

P3cdethX3cd XSUBcd NdTHORN frac14 βcdethW3TOTd WSUBdTHORN eth9THORN

Because real aggregate consumption is usually exogenous in our partial equilibriumsimulations the linear expenditure system determines the consumption shares ofindividual final commodities (ie the five wine types plus a composite of all otherconsumption items) driven by changes in relative prices as faced by domestic con-sumers The income elasticity of demand for each commodity is equal to the mar-ginal budget share divided by the expenditure share This number varies from 05for nonpremium wine to 17 for superpremium still wine The income elasticity ofdemand for other consumption is very close to 10 because wine accounts for anaverage of only 03 of aggregate expenditures globally and no more than 11in any country (Anderson and Nelgen 2011 Table 166)

A feature of our revised model of world wine markets is the inclusion of nominalexchange rates These appear directly in the equation linking retail prices (P3scd ) toproducer prices by country of origin (P0s

c ) where c denotes the wine type

P3 scd frac14 P0s

cfd

fsTtarcd T

taxcd thorn Pm

cd eth10THORN

248 UK and Global Wine Markets by 2025 and Implications of Brexit

The exchange rates in the consuming and producing regions are fd and fs respec-tively expressed as local currency units per $US Ttar

cd is the power of the tariff in theconsuming region and Ttax

cd is the power of the domestic consumption (or excise) taxprior to any generic value-added or goods-and-services tax Pm

cd is the price of marginm assumed to be locally supplied nontradable and therefore unaffected by theexchange rate

A given level of consumption for wine type c (X3cd) is satisfied using theArmington (1969) assumption in which wines from different countries of originare imperfectly substitutable First domestic wine is imperfectly substitutable witha composite of imports

X3sscd frac14 f ethX3cd CESethP3sscd=P3cdTHORNTHORN ss frac14 domestic imports eth11THORN

and then imports by origin (X3scd ) are determined in a second CES equation

X3scd frac14 f ethX3ssfrac14importscd CESethP3scd=P3

ssfrac14importscd THORNTHORN eth12THORN

The model enables us to show how changes in international competitiveness affectthe worldrsquos wine markets A crucial part of this exercise is explaining how pricesdetermined outside the grape and wine markets influence these markets Becausethe model is in partial equilibrium to depict the impacts of changes in internationalcompetitiveness outside price changes need to be imposed as shocks on the modelThe price of intermediate inputs other than grapes and wine shown in equations (1)and (2) is set equal to the price of GDP (Pg

d ) multiplied by a shifter Fcd

Pcid frac14 Fc

dPgd eth13THORN

If no specific price observations are available the shifter Fcd remains exogenous and

unshocked with the change in price being determined by a shock to the price ofGDP If observations are available for specific input price movements the shifterFc

dbecomes endogenous with Pc

id now exogenous and shocked

Wid frac14 Fwd P

gd eth14THORN

Wage rates are treated similarly In equation (14) if the wage shifter Fwd is exogenous

changes in wage rates Wid are determined by changes in the price of GDP If wagerate data are available Fw

d becomes endogenous andwage rates are shocked directly

Pmcd frac14 Fm

d Pgd eth15THORN

The prices of trade and transport margins are also determined by the price of GDP ifthe shifter Fm

d in equation (15) is exogenous

Changes in international competitiveness depend on changes in relative pricelevels and changes in nominal exchange rates In equation (16) fR

s denotes real

Kym Anderson and Glyn Wittwer 249

exchange-rate movements relative to the US dollar

fRs frac14 Pg

s =frac12PgUSA

fs$ eth16THORN

In equation (16) the nominal exchange rate for the United States is alwaysunchanged because nominal and real exchange rates are expressed relative to theUS currency

Changes in international market conditions may have impacts in one direction onproducer prices as expressed in US dollars P0s

i and potentially in the opposite direc-tion in local currency units Hence we calculate real producer pricesP0s

iloc in localcurrency terms (ie the price most relevant to domestic producers)

P0siloc frac14 P0s

ifs=P

gs eth17THORN

To obtain real price changes in local currency terms we convert US dollar prices(P3 s

cd for the source-specific price and P3cd for the source-composite price) to reallocal currency prices (P3 s

cdloc and P3cdloc) using the consumer price index (CPI)(Pc

d ) as the deflator

P3 scdloc frac14 P3 s

cdfd=P

cd eth18THORN

and

P3cdloc frac14 P3cdfd=Pcd eth19THORN

References

Anderson K and Nelgen S (2011) Global Wine Markets 1961 to 2009 A StatisticalCompendium Adelaide University of Adelaide Press Also available as an e-book atwwwadelaideeduaupresstitlesglobal-wine

Anderson K Nelgen S and Pinilla V (2017) Global Wine Markets 1860 to 2016 AStatistical Compendium Adelaide University of Adelaide Press (forthcoming) Also tobe available as an e-book at wwwadelaideeduaupress

Anderson K and Pinilla V (with the assistance of A J Holmes) (2017)Annual Database ofGlobal Wine Markets 1835 to 2016 Wine Economics Research Centre University ofAdelaide to be posted in October at wwwadelaideeduauwine-econdatabases

Anderson K and Strutt A (2016) Impacts of Asiarsquos rise on African and Latin Americantrade Projections to 2030 World Economy 39(2) 172ndash194

Anderson K andWittwer G (2013) Modeling global wine markets to 2018 Exchange ratestaste changes and Chinarsquos import growth Journal of Wine Economics 8(2) 131ndash158

Anderson K and Wittwer G (2015) Asiarsquos evolving role in global wine markets ChinaEconomic Review 35 1ndash14

250 UK and Global Wine Markets by 2025 and Implications of Brexit

Armington P A (1969) A theory of demand for products distinguished by place of produc-tion IMF Staff Papers 16 159ndash178

Baldwin R E (ed) (2016) Brexit Beckons Thinking Ahead by Leading Economists LondonCentre for Economic Policy Research

Campos N F (2016) Lousy experts Looking back at the ex ante estimates of the costs ofBrexit In R E Baldwin (ed) Brexit Beckons Thinking Ahead by Leading Economists(ch 3) London Centre for Economic Policy Research

Dixon P Parmenter B Sutton J and Vincent D (1982) ORANI A Multisectoral Modelof the Australian Economy Contributions to Economic Analysis 142 Amsterdam North-Holland

European Commission (2015) Trade for All Towards a More Responsible Trade andInvestment Policy Brussels European Commission

Harrison J Horridge M Jerie M and Pearson K (2014) GEMPACK ManualMelbourne GEMPACK Software

HM Treasury (2017) Spring Budget 2017 London HM TreasuryHolmes A J and Anderson K (2017a)Annual Database of National Beverage Consumption

Volumes and Expenditures 1950 to 2015 Wine Economics Research Centre University ofAdelaide posted at wwwadelaideeduauwine-econdatabases

Holmes A J and Anderson K (2017b) Convergence in National Alcohol ConsumptionPatterns New Global Indicators Journal of Wine Economics 12(2) 117ndash148

Ludington C C (2013) The Politics of Wine in Britain A New Cultural History BasingstokeUK Palgrave Macmillan

Office of National Statistics (2017) Analysis of real earnings London Office of NationalStatistics Available at httpswwwonsgovukemploymentandlabourmarketpeopleinworkearningsandworkinghoursarticlessupplementaryanalysisofaverageweeklyearningslatest

Rollo J Borchert I Dawar K Holmes P and Winters L A (2016) The World TradeOrganisation A safety net for a post-Brexit UK trade policy Briefing Paper 1UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-1-final-1pdf

Smith A (2017) Brexit Hard truths and hard choices UKTPO blog 19 June Available athttpsblogssussexacukuktpo20170619brexit-hard-truths-and-hard-choices

Swinbank A (2017) World trade rules and the policy options for British agriculture post-Brexit Briefing Paper 7 UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-7pdf

Viner J (1950) The Customs Union Issue New York Carnegie Endowment for InternationalPeace

Wine Australia (2015) Export Market Guide European Union Adelaide Wine AustraliaWittwer G Berger N and Anderson K (2003) A model of the worldrsquos wine markets

Economic Modelling 20(3) 487ndash506

Kym Anderson and Glyn Wittwer 251

Page 27: U.K. and Global Wine Markets by 2025, and Implications of ... · U.K. and Global Wine Markets by 2025, and Implications of Brexit* Kym Andersona and Glyn Wittwerb Abstract The United

plant capacity adjust slowly to market signals

Lid frac14 f ethFid CESfrac12Wid=PFid $THORN eth3THORN

Kid frac14 f ethFid CESethRid=PFid $THORN eth4THORN

PFid Fid frac14 Lid Wid thorn Kid Rid eth5THORN

Grape and wine producers are assumed to minimize costs subject to CES substitu-tion between capital and labor Equations (3) to (5) show primary factor demandsfor the labor composite Lid and capital Kid subject to a composite factor demandFid by industry i in region d The factor prices are Wid for labor Rid for capitalrentals and PFid for composite factor prices

Appendix Table A1Cumulative Consumption and Population Growth Rates and Changes in the Real Exchange Rate

(RER) Relative to the US Dollar 2014 to 2025 without Brexit ()

Aggregateconsumption Poprsquon RER

Aggregateconsumption Poprsquon RER

France 18 4 minus11 Australia 35 11 minus17Italy 11 2 minus9 New Zealand 32 9 minus26Portugal 14 0 minus9 Canada 27 8 minus18Spain 26 8 minus9 United States 31 8 0Austria 19 4 minus7 Argentina 7 10 109Belgium 20 7 minus9 Brazil 16 8 minus29Denmark 22 2 minus9 Chile 55 8 minus2Finland 21 3 minus7 Mexico 42 12 minus8Germany 14 minus2 minus11 Uruguay 45 3 1Greece 22 minus1 minus14 Other Latin America 60 10 minus5Ireland 42 12 minus9 South Africa 36 12 minus1Netherlands 21 4 minus9 Turkey 50 8 20Sweden 24 9 minus13 North Africa 53 11 0Switzerland 18 8 minus6 Other Africa 109 18 84United Kingdom 32 6 1 Middle East 52 18 minus12Other West Europe 21 10 minus1 China 79 3 5Bulgaria 41 minus7 7 Hong Kong 42 3 2Croatia 20 minus2 minus1 India 134 13 17Georgia 35 0 23 Japan 11 minus3 minus24Hungary 25 minus3 minus11 Korea 38 1 minus9Moldova 49 minus11 13 Malaysia 62 15 minus16Romania 45 minus4 22 Philippines 75 18 7Russia 18 minus2 minus8 Singapore 44 21 minus22Ukraine 22 minus5 14 Taiwan 29 1 minus13Other East Europe 40 minus5 48 Thailand 47 3 minus9

Other Asia 99 10 10

Source Authorsrsquo compilation from projections by various international agencies and from global economy-wide modeling by Anderson andStrutt (2016)

Kym Anderson and Glyn Wittwer 247

The composite factor demand Fid is proportional to total output Qid subject to aprimary factor using technology Aid Hence

Fid frac14 Qid Aid eth6THORN

The perfectly competitive zero-pure-profit condition is that total revenue valued atthe output price P0s

i multiplied by Qid equals the total production cost

P0si Qid frac14

X

cPcid X1cid thorn

X

oWo

id Loid thorn Rid Kid eth7THORN

Household demands follow a linear expenditure system in each region We reducethe optimizing problem for household consumption of each commodity subjectto a budget constraint to equations describing subsistence and discretionarydemands Aggregate subsistence expenditure WSUBd depends only on consumerprices P3cd for each commodity and the number of households N as per-capita sub-sistence quantities XSUBcd subject to given preferences are constant

WSUBd frac14X

c

P3cd XSUBcd Nd eth8THORN

Discretionary expenditures for each commodity (the left-hand side of equation (9))are equal to the marginal budget share (βcd) of aggregate discretionary expenditureThis aggregate is the bracketed term on the right-hand side of equation (9) whereW3TOTd is aggregate nominal expenditure

P3cdethX3cd XSUBcd NdTHORN frac14 βcdethW3TOTd WSUBdTHORN eth9THORN

Because real aggregate consumption is usually exogenous in our partial equilibriumsimulations the linear expenditure system determines the consumption shares ofindividual final commodities (ie the five wine types plus a composite of all otherconsumption items) driven by changes in relative prices as faced by domestic con-sumers The income elasticity of demand for each commodity is equal to the mar-ginal budget share divided by the expenditure share This number varies from 05for nonpremium wine to 17 for superpremium still wine The income elasticity ofdemand for other consumption is very close to 10 because wine accounts for anaverage of only 03 of aggregate expenditures globally and no more than 11in any country (Anderson and Nelgen 2011 Table 166)

A feature of our revised model of world wine markets is the inclusion of nominalexchange rates These appear directly in the equation linking retail prices (P3scd ) toproducer prices by country of origin (P0s

c ) where c denotes the wine type

P3 scd frac14 P0s

cfd

fsTtarcd T

taxcd thorn Pm

cd eth10THORN

248 UK and Global Wine Markets by 2025 and Implications of Brexit

The exchange rates in the consuming and producing regions are fd and fs respec-tively expressed as local currency units per $US Ttar

cd is the power of the tariff in theconsuming region and Ttax

cd is the power of the domestic consumption (or excise) taxprior to any generic value-added or goods-and-services tax Pm

cd is the price of marginm assumed to be locally supplied nontradable and therefore unaffected by theexchange rate

A given level of consumption for wine type c (X3cd) is satisfied using theArmington (1969) assumption in which wines from different countries of originare imperfectly substitutable First domestic wine is imperfectly substitutable witha composite of imports

X3sscd frac14 f ethX3cd CESethP3sscd=P3cdTHORNTHORN ss frac14 domestic imports eth11THORN

and then imports by origin (X3scd ) are determined in a second CES equation

X3scd frac14 f ethX3ssfrac14importscd CESethP3scd=P3

ssfrac14importscd THORNTHORN eth12THORN

The model enables us to show how changes in international competitiveness affectthe worldrsquos wine markets A crucial part of this exercise is explaining how pricesdetermined outside the grape and wine markets influence these markets Becausethe model is in partial equilibrium to depict the impacts of changes in internationalcompetitiveness outside price changes need to be imposed as shocks on the modelThe price of intermediate inputs other than grapes and wine shown in equations (1)and (2) is set equal to the price of GDP (Pg

d ) multiplied by a shifter Fcd

Pcid frac14 Fc

dPgd eth13THORN

If no specific price observations are available the shifter Fcd remains exogenous and

unshocked with the change in price being determined by a shock to the price ofGDP If observations are available for specific input price movements the shifterFc

dbecomes endogenous with Pc

id now exogenous and shocked

Wid frac14 Fwd P

gd eth14THORN

Wage rates are treated similarly In equation (14) if the wage shifter Fwd is exogenous

changes in wage rates Wid are determined by changes in the price of GDP If wagerate data are available Fw

d becomes endogenous andwage rates are shocked directly

Pmcd frac14 Fm

d Pgd eth15THORN

The prices of trade and transport margins are also determined by the price of GDP ifthe shifter Fm

d in equation (15) is exogenous

Changes in international competitiveness depend on changes in relative pricelevels and changes in nominal exchange rates In equation (16) fR

s denotes real

Kym Anderson and Glyn Wittwer 249

exchange-rate movements relative to the US dollar

fRs frac14 Pg

s =frac12PgUSA

fs$ eth16THORN

In equation (16) the nominal exchange rate for the United States is alwaysunchanged because nominal and real exchange rates are expressed relative to theUS currency

Changes in international market conditions may have impacts in one direction onproducer prices as expressed in US dollars P0s

i and potentially in the opposite direc-tion in local currency units Hence we calculate real producer pricesP0s

iloc in localcurrency terms (ie the price most relevant to domestic producers)

P0siloc frac14 P0s

ifs=P

gs eth17THORN

To obtain real price changes in local currency terms we convert US dollar prices(P3 s

cd for the source-specific price and P3cd for the source-composite price) to reallocal currency prices (P3 s

cdloc and P3cdloc) using the consumer price index (CPI)(Pc

d ) as the deflator

P3 scdloc frac14 P3 s

cdfd=P

cd eth18THORN

and

P3cdloc frac14 P3cdfd=Pcd eth19THORN

References

Anderson K and Nelgen S (2011) Global Wine Markets 1961 to 2009 A StatisticalCompendium Adelaide University of Adelaide Press Also available as an e-book atwwwadelaideeduaupresstitlesglobal-wine

Anderson K Nelgen S and Pinilla V (2017) Global Wine Markets 1860 to 2016 AStatistical Compendium Adelaide University of Adelaide Press (forthcoming) Also tobe available as an e-book at wwwadelaideeduaupress

Anderson K and Pinilla V (with the assistance of A J Holmes) (2017)Annual Database ofGlobal Wine Markets 1835 to 2016 Wine Economics Research Centre University ofAdelaide to be posted in October at wwwadelaideeduauwine-econdatabases

Anderson K and Strutt A (2016) Impacts of Asiarsquos rise on African and Latin Americantrade Projections to 2030 World Economy 39(2) 172ndash194

Anderson K andWittwer G (2013) Modeling global wine markets to 2018 Exchange ratestaste changes and Chinarsquos import growth Journal of Wine Economics 8(2) 131ndash158

Anderson K and Wittwer G (2015) Asiarsquos evolving role in global wine markets ChinaEconomic Review 35 1ndash14

250 UK and Global Wine Markets by 2025 and Implications of Brexit

Armington P A (1969) A theory of demand for products distinguished by place of produc-tion IMF Staff Papers 16 159ndash178

Baldwin R E (ed) (2016) Brexit Beckons Thinking Ahead by Leading Economists LondonCentre for Economic Policy Research

Campos N F (2016) Lousy experts Looking back at the ex ante estimates of the costs ofBrexit In R E Baldwin (ed) Brexit Beckons Thinking Ahead by Leading Economists(ch 3) London Centre for Economic Policy Research

Dixon P Parmenter B Sutton J and Vincent D (1982) ORANI A Multisectoral Modelof the Australian Economy Contributions to Economic Analysis 142 Amsterdam North-Holland

European Commission (2015) Trade for All Towards a More Responsible Trade andInvestment Policy Brussels European Commission

Harrison J Horridge M Jerie M and Pearson K (2014) GEMPACK ManualMelbourne GEMPACK Software

HM Treasury (2017) Spring Budget 2017 London HM TreasuryHolmes A J and Anderson K (2017a)Annual Database of National Beverage Consumption

Volumes and Expenditures 1950 to 2015 Wine Economics Research Centre University ofAdelaide posted at wwwadelaideeduauwine-econdatabases

Holmes A J and Anderson K (2017b) Convergence in National Alcohol ConsumptionPatterns New Global Indicators Journal of Wine Economics 12(2) 117ndash148

Ludington C C (2013) The Politics of Wine in Britain A New Cultural History BasingstokeUK Palgrave Macmillan

Office of National Statistics (2017) Analysis of real earnings London Office of NationalStatistics Available at httpswwwonsgovukemploymentandlabourmarketpeopleinworkearningsandworkinghoursarticlessupplementaryanalysisofaverageweeklyearningslatest

Rollo J Borchert I Dawar K Holmes P and Winters L A (2016) The World TradeOrganisation A safety net for a post-Brexit UK trade policy Briefing Paper 1UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-1-final-1pdf

Smith A (2017) Brexit Hard truths and hard choices UKTPO blog 19 June Available athttpsblogssussexacukuktpo20170619brexit-hard-truths-and-hard-choices

Swinbank A (2017) World trade rules and the policy options for British agriculture post-Brexit Briefing Paper 7 UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-7pdf

Viner J (1950) The Customs Union Issue New York Carnegie Endowment for InternationalPeace

Wine Australia (2015) Export Market Guide European Union Adelaide Wine AustraliaWittwer G Berger N and Anderson K (2003) A model of the worldrsquos wine markets

Economic Modelling 20(3) 487ndash506

Kym Anderson and Glyn Wittwer 251

Page 28: U.K. and Global Wine Markets by 2025, and Implications of ... · U.K. and Global Wine Markets by 2025, and Implications of Brexit* Kym Andersona and Glyn Wittwerb Abstract The United

The composite factor demand Fid is proportional to total output Qid subject to aprimary factor using technology Aid Hence

Fid frac14 Qid Aid eth6THORN

The perfectly competitive zero-pure-profit condition is that total revenue valued atthe output price P0s

i multiplied by Qid equals the total production cost

P0si Qid frac14

X

cPcid X1cid thorn

X

oWo

id Loid thorn Rid Kid eth7THORN

Household demands follow a linear expenditure system in each region We reducethe optimizing problem for household consumption of each commodity subjectto a budget constraint to equations describing subsistence and discretionarydemands Aggregate subsistence expenditure WSUBd depends only on consumerprices P3cd for each commodity and the number of households N as per-capita sub-sistence quantities XSUBcd subject to given preferences are constant

WSUBd frac14X

c

P3cd XSUBcd Nd eth8THORN

Discretionary expenditures for each commodity (the left-hand side of equation (9))are equal to the marginal budget share (βcd) of aggregate discretionary expenditureThis aggregate is the bracketed term on the right-hand side of equation (9) whereW3TOTd is aggregate nominal expenditure

P3cdethX3cd XSUBcd NdTHORN frac14 βcdethW3TOTd WSUBdTHORN eth9THORN

Because real aggregate consumption is usually exogenous in our partial equilibriumsimulations the linear expenditure system determines the consumption shares ofindividual final commodities (ie the five wine types plus a composite of all otherconsumption items) driven by changes in relative prices as faced by domestic con-sumers The income elasticity of demand for each commodity is equal to the mar-ginal budget share divided by the expenditure share This number varies from 05for nonpremium wine to 17 for superpremium still wine The income elasticity ofdemand for other consumption is very close to 10 because wine accounts for anaverage of only 03 of aggregate expenditures globally and no more than 11in any country (Anderson and Nelgen 2011 Table 166)

A feature of our revised model of world wine markets is the inclusion of nominalexchange rates These appear directly in the equation linking retail prices (P3scd ) toproducer prices by country of origin (P0s

c ) where c denotes the wine type

P3 scd frac14 P0s

cfd

fsTtarcd T

taxcd thorn Pm

cd eth10THORN

248 UK and Global Wine Markets by 2025 and Implications of Brexit

The exchange rates in the consuming and producing regions are fd and fs respec-tively expressed as local currency units per $US Ttar

cd is the power of the tariff in theconsuming region and Ttax

cd is the power of the domestic consumption (or excise) taxprior to any generic value-added or goods-and-services tax Pm

cd is the price of marginm assumed to be locally supplied nontradable and therefore unaffected by theexchange rate

A given level of consumption for wine type c (X3cd) is satisfied using theArmington (1969) assumption in which wines from different countries of originare imperfectly substitutable First domestic wine is imperfectly substitutable witha composite of imports

X3sscd frac14 f ethX3cd CESethP3sscd=P3cdTHORNTHORN ss frac14 domestic imports eth11THORN

and then imports by origin (X3scd ) are determined in a second CES equation

X3scd frac14 f ethX3ssfrac14importscd CESethP3scd=P3

ssfrac14importscd THORNTHORN eth12THORN

The model enables us to show how changes in international competitiveness affectthe worldrsquos wine markets A crucial part of this exercise is explaining how pricesdetermined outside the grape and wine markets influence these markets Becausethe model is in partial equilibrium to depict the impacts of changes in internationalcompetitiveness outside price changes need to be imposed as shocks on the modelThe price of intermediate inputs other than grapes and wine shown in equations (1)and (2) is set equal to the price of GDP (Pg

d ) multiplied by a shifter Fcd

Pcid frac14 Fc

dPgd eth13THORN

If no specific price observations are available the shifter Fcd remains exogenous and

unshocked with the change in price being determined by a shock to the price ofGDP If observations are available for specific input price movements the shifterFc

dbecomes endogenous with Pc

id now exogenous and shocked

Wid frac14 Fwd P

gd eth14THORN

Wage rates are treated similarly In equation (14) if the wage shifter Fwd is exogenous

changes in wage rates Wid are determined by changes in the price of GDP If wagerate data are available Fw

d becomes endogenous andwage rates are shocked directly

Pmcd frac14 Fm

d Pgd eth15THORN

The prices of trade and transport margins are also determined by the price of GDP ifthe shifter Fm

d in equation (15) is exogenous

Changes in international competitiveness depend on changes in relative pricelevels and changes in nominal exchange rates In equation (16) fR

s denotes real

Kym Anderson and Glyn Wittwer 249

exchange-rate movements relative to the US dollar

fRs frac14 Pg

s =frac12PgUSA

fs$ eth16THORN

In equation (16) the nominal exchange rate for the United States is alwaysunchanged because nominal and real exchange rates are expressed relative to theUS currency

Changes in international market conditions may have impacts in one direction onproducer prices as expressed in US dollars P0s

i and potentially in the opposite direc-tion in local currency units Hence we calculate real producer pricesP0s

iloc in localcurrency terms (ie the price most relevant to domestic producers)

P0siloc frac14 P0s

ifs=P

gs eth17THORN

To obtain real price changes in local currency terms we convert US dollar prices(P3 s

cd for the source-specific price and P3cd for the source-composite price) to reallocal currency prices (P3 s

cdloc and P3cdloc) using the consumer price index (CPI)(Pc

d ) as the deflator

P3 scdloc frac14 P3 s

cdfd=P

cd eth18THORN

and

P3cdloc frac14 P3cdfd=Pcd eth19THORN

References

Anderson K and Nelgen S (2011) Global Wine Markets 1961 to 2009 A StatisticalCompendium Adelaide University of Adelaide Press Also available as an e-book atwwwadelaideeduaupresstitlesglobal-wine

Anderson K Nelgen S and Pinilla V (2017) Global Wine Markets 1860 to 2016 AStatistical Compendium Adelaide University of Adelaide Press (forthcoming) Also tobe available as an e-book at wwwadelaideeduaupress

Anderson K and Pinilla V (with the assistance of A J Holmes) (2017)Annual Database ofGlobal Wine Markets 1835 to 2016 Wine Economics Research Centre University ofAdelaide to be posted in October at wwwadelaideeduauwine-econdatabases

Anderson K and Strutt A (2016) Impacts of Asiarsquos rise on African and Latin Americantrade Projections to 2030 World Economy 39(2) 172ndash194

Anderson K andWittwer G (2013) Modeling global wine markets to 2018 Exchange ratestaste changes and Chinarsquos import growth Journal of Wine Economics 8(2) 131ndash158

Anderson K and Wittwer G (2015) Asiarsquos evolving role in global wine markets ChinaEconomic Review 35 1ndash14

250 UK and Global Wine Markets by 2025 and Implications of Brexit

Armington P A (1969) A theory of demand for products distinguished by place of produc-tion IMF Staff Papers 16 159ndash178

Baldwin R E (ed) (2016) Brexit Beckons Thinking Ahead by Leading Economists LondonCentre for Economic Policy Research

Campos N F (2016) Lousy experts Looking back at the ex ante estimates of the costs ofBrexit In R E Baldwin (ed) Brexit Beckons Thinking Ahead by Leading Economists(ch 3) London Centre for Economic Policy Research

Dixon P Parmenter B Sutton J and Vincent D (1982) ORANI A Multisectoral Modelof the Australian Economy Contributions to Economic Analysis 142 Amsterdam North-Holland

European Commission (2015) Trade for All Towards a More Responsible Trade andInvestment Policy Brussels European Commission

Harrison J Horridge M Jerie M and Pearson K (2014) GEMPACK ManualMelbourne GEMPACK Software

HM Treasury (2017) Spring Budget 2017 London HM TreasuryHolmes A J and Anderson K (2017a)Annual Database of National Beverage Consumption

Volumes and Expenditures 1950 to 2015 Wine Economics Research Centre University ofAdelaide posted at wwwadelaideeduauwine-econdatabases

Holmes A J and Anderson K (2017b) Convergence in National Alcohol ConsumptionPatterns New Global Indicators Journal of Wine Economics 12(2) 117ndash148

Ludington C C (2013) The Politics of Wine in Britain A New Cultural History BasingstokeUK Palgrave Macmillan

Office of National Statistics (2017) Analysis of real earnings London Office of NationalStatistics Available at httpswwwonsgovukemploymentandlabourmarketpeopleinworkearningsandworkinghoursarticlessupplementaryanalysisofaverageweeklyearningslatest

Rollo J Borchert I Dawar K Holmes P and Winters L A (2016) The World TradeOrganisation A safety net for a post-Brexit UK trade policy Briefing Paper 1UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-1-final-1pdf

Smith A (2017) Brexit Hard truths and hard choices UKTPO blog 19 June Available athttpsblogssussexacukuktpo20170619brexit-hard-truths-and-hard-choices

Swinbank A (2017) World trade rules and the policy options for British agriculture post-Brexit Briefing Paper 7 UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-7pdf

Viner J (1950) The Customs Union Issue New York Carnegie Endowment for InternationalPeace

Wine Australia (2015) Export Market Guide European Union Adelaide Wine AustraliaWittwer G Berger N and Anderson K (2003) A model of the worldrsquos wine markets

Economic Modelling 20(3) 487ndash506

Kym Anderson and Glyn Wittwer 251

Page 29: U.K. and Global Wine Markets by 2025, and Implications of ... · U.K. and Global Wine Markets by 2025, and Implications of Brexit* Kym Andersona and Glyn Wittwerb Abstract The United

The exchange rates in the consuming and producing regions are fd and fs respec-tively expressed as local currency units per $US Ttar

cd is the power of the tariff in theconsuming region and Ttax

cd is the power of the domestic consumption (or excise) taxprior to any generic value-added or goods-and-services tax Pm

cd is the price of marginm assumed to be locally supplied nontradable and therefore unaffected by theexchange rate

A given level of consumption for wine type c (X3cd) is satisfied using theArmington (1969) assumption in which wines from different countries of originare imperfectly substitutable First domestic wine is imperfectly substitutable witha composite of imports

X3sscd frac14 f ethX3cd CESethP3sscd=P3cdTHORNTHORN ss frac14 domestic imports eth11THORN

and then imports by origin (X3scd ) are determined in a second CES equation

X3scd frac14 f ethX3ssfrac14importscd CESethP3scd=P3

ssfrac14importscd THORNTHORN eth12THORN

The model enables us to show how changes in international competitiveness affectthe worldrsquos wine markets A crucial part of this exercise is explaining how pricesdetermined outside the grape and wine markets influence these markets Becausethe model is in partial equilibrium to depict the impacts of changes in internationalcompetitiveness outside price changes need to be imposed as shocks on the modelThe price of intermediate inputs other than grapes and wine shown in equations (1)and (2) is set equal to the price of GDP (Pg

d ) multiplied by a shifter Fcd

Pcid frac14 Fc

dPgd eth13THORN

If no specific price observations are available the shifter Fcd remains exogenous and

unshocked with the change in price being determined by a shock to the price ofGDP If observations are available for specific input price movements the shifterFc

dbecomes endogenous with Pc

id now exogenous and shocked

Wid frac14 Fwd P

gd eth14THORN

Wage rates are treated similarly In equation (14) if the wage shifter Fwd is exogenous

changes in wage rates Wid are determined by changes in the price of GDP If wagerate data are available Fw

d becomes endogenous andwage rates are shocked directly

Pmcd frac14 Fm

d Pgd eth15THORN

The prices of trade and transport margins are also determined by the price of GDP ifthe shifter Fm

d in equation (15) is exogenous

Changes in international competitiveness depend on changes in relative pricelevels and changes in nominal exchange rates In equation (16) fR

s denotes real

Kym Anderson and Glyn Wittwer 249

exchange-rate movements relative to the US dollar

fRs frac14 Pg

s =frac12PgUSA

fs$ eth16THORN

In equation (16) the nominal exchange rate for the United States is alwaysunchanged because nominal and real exchange rates are expressed relative to theUS currency

Changes in international market conditions may have impacts in one direction onproducer prices as expressed in US dollars P0s

i and potentially in the opposite direc-tion in local currency units Hence we calculate real producer pricesP0s

iloc in localcurrency terms (ie the price most relevant to domestic producers)

P0siloc frac14 P0s

ifs=P

gs eth17THORN

To obtain real price changes in local currency terms we convert US dollar prices(P3 s

cd for the source-specific price and P3cd for the source-composite price) to reallocal currency prices (P3 s

cdloc and P3cdloc) using the consumer price index (CPI)(Pc

d ) as the deflator

P3 scdloc frac14 P3 s

cdfd=P

cd eth18THORN

and

P3cdloc frac14 P3cdfd=Pcd eth19THORN

References

Anderson K and Nelgen S (2011) Global Wine Markets 1961 to 2009 A StatisticalCompendium Adelaide University of Adelaide Press Also available as an e-book atwwwadelaideeduaupresstitlesglobal-wine

Anderson K Nelgen S and Pinilla V (2017) Global Wine Markets 1860 to 2016 AStatistical Compendium Adelaide University of Adelaide Press (forthcoming) Also tobe available as an e-book at wwwadelaideeduaupress

Anderson K and Pinilla V (with the assistance of A J Holmes) (2017)Annual Database ofGlobal Wine Markets 1835 to 2016 Wine Economics Research Centre University ofAdelaide to be posted in October at wwwadelaideeduauwine-econdatabases

Anderson K and Strutt A (2016) Impacts of Asiarsquos rise on African and Latin Americantrade Projections to 2030 World Economy 39(2) 172ndash194

Anderson K andWittwer G (2013) Modeling global wine markets to 2018 Exchange ratestaste changes and Chinarsquos import growth Journal of Wine Economics 8(2) 131ndash158

Anderson K and Wittwer G (2015) Asiarsquos evolving role in global wine markets ChinaEconomic Review 35 1ndash14

250 UK and Global Wine Markets by 2025 and Implications of Brexit

Armington P A (1969) A theory of demand for products distinguished by place of produc-tion IMF Staff Papers 16 159ndash178

Baldwin R E (ed) (2016) Brexit Beckons Thinking Ahead by Leading Economists LondonCentre for Economic Policy Research

Campos N F (2016) Lousy experts Looking back at the ex ante estimates of the costs ofBrexit In R E Baldwin (ed) Brexit Beckons Thinking Ahead by Leading Economists(ch 3) London Centre for Economic Policy Research

Dixon P Parmenter B Sutton J and Vincent D (1982) ORANI A Multisectoral Modelof the Australian Economy Contributions to Economic Analysis 142 Amsterdam North-Holland

European Commission (2015) Trade for All Towards a More Responsible Trade andInvestment Policy Brussels European Commission

Harrison J Horridge M Jerie M and Pearson K (2014) GEMPACK ManualMelbourne GEMPACK Software

HM Treasury (2017) Spring Budget 2017 London HM TreasuryHolmes A J and Anderson K (2017a)Annual Database of National Beverage Consumption

Volumes and Expenditures 1950 to 2015 Wine Economics Research Centre University ofAdelaide posted at wwwadelaideeduauwine-econdatabases

Holmes A J and Anderson K (2017b) Convergence in National Alcohol ConsumptionPatterns New Global Indicators Journal of Wine Economics 12(2) 117ndash148

Ludington C C (2013) The Politics of Wine in Britain A New Cultural History BasingstokeUK Palgrave Macmillan

Office of National Statistics (2017) Analysis of real earnings London Office of NationalStatistics Available at httpswwwonsgovukemploymentandlabourmarketpeopleinworkearningsandworkinghoursarticlessupplementaryanalysisofaverageweeklyearningslatest

Rollo J Borchert I Dawar K Holmes P and Winters L A (2016) The World TradeOrganisation A safety net for a post-Brexit UK trade policy Briefing Paper 1UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-1-final-1pdf

Smith A (2017) Brexit Hard truths and hard choices UKTPO blog 19 June Available athttpsblogssussexacukuktpo20170619brexit-hard-truths-and-hard-choices

Swinbank A (2017) World trade rules and the policy options for British agriculture post-Brexit Briefing Paper 7 UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-7pdf

Viner J (1950) The Customs Union Issue New York Carnegie Endowment for InternationalPeace

Wine Australia (2015) Export Market Guide European Union Adelaide Wine AustraliaWittwer G Berger N and Anderson K (2003) A model of the worldrsquos wine markets

Economic Modelling 20(3) 487ndash506

Kym Anderson and Glyn Wittwer 251

Page 30: U.K. and Global Wine Markets by 2025, and Implications of ... · U.K. and Global Wine Markets by 2025, and Implications of Brexit* Kym Andersona and Glyn Wittwerb Abstract The United

exchange-rate movements relative to the US dollar

fRs frac14 Pg

s =frac12PgUSA

fs$ eth16THORN

In equation (16) the nominal exchange rate for the United States is alwaysunchanged because nominal and real exchange rates are expressed relative to theUS currency

Changes in international market conditions may have impacts in one direction onproducer prices as expressed in US dollars P0s

i and potentially in the opposite direc-tion in local currency units Hence we calculate real producer pricesP0s

iloc in localcurrency terms (ie the price most relevant to domestic producers)

P0siloc frac14 P0s

ifs=P

gs eth17THORN

To obtain real price changes in local currency terms we convert US dollar prices(P3 s

cd for the source-specific price and P3cd for the source-composite price) to reallocal currency prices (P3 s

cdloc and P3cdloc) using the consumer price index (CPI)(Pc

d ) as the deflator

P3 scdloc frac14 P3 s

cdfd=P

cd eth18THORN

and

P3cdloc frac14 P3cdfd=Pcd eth19THORN

References

Anderson K and Nelgen S (2011) Global Wine Markets 1961 to 2009 A StatisticalCompendium Adelaide University of Adelaide Press Also available as an e-book atwwwadelaideeduaupresstitlesglobal-wine

Anderson K Nelgen S and Pinilla V (2017) Global Wine Markets 1860 to 2016 AStatistical Compendium Adelaide University of Adelaide Press (forthcoming) Also tobe available as an e-book at wwwadelaideeduaupress

Anderson K and Pinilla V (with the assistance of A J Holmes) (2017)Annual Database ofGlobal Wine Markets 1835 to 2016 Wine Economics Research Centre University ofAdelaide to be posted in October at wwwadelaideeduauwine-econdatabases

Anderson K and Strutt A (2016) Impacts of Asiarsquos rise on African and Latin Americantrade Projections to 2030 World Economy 39(2) 172ndash194

Anderson K andWittwer G (2013) Modeling global wine markets to 2018 Exchange ratestaste changes and Chinarsquos import growth Journal of Wine Economics 8(2) 131ndash158

Anderson K and Wittwer G (2015) Asiarsquos evolving role in global wine markets ChinaEconomic Review 35 1ndash14

250 UK and Global Wine Markets by 2025 and Implications of Brexit

Armington P A (1969) A theory of demand for products distinguished by place of produc-tion IMF Staff Papers 16 159ndash178

Baldwin R E (ed) (2016) Brexit Beckons Thinking Ahead by Leading Economists LondonCentre for Economic Policy Research

Campos N F (2016) Lousy experts Looking back at the ex ante estimates of the costs ofBrexit In R E Baldwin (ed) Brexit Beckons Thinking Ahead by Leading Economists(ch 3) London Centre for Economic Policy Research

Dixon P Parmenter B Sutton J and Vincent D (1982) ORANI A Multisectoral Modelof the Australian Economy Contributions to Economic Analysis 142 Amsterdam North-Holland

European Commission (2015) Trade for All Towards a More Responsible Trade andInvestment Policy Brussels European Commission

Harrison J Horridge M Jerie M and Pearson K (2014) GEMPACK ManualMelbourne GEMPACK Software

HM Treasury (2017) Spring Budget 2017 London HM TreasuryHolmes A J and Anderson K (2017a)Annual Database of National Beverage Consumption

Volumes and Expenditures 1950 to 2015 Wine Economics Research Centre University ofAdelaide posted at wwwadelaideeduauwine-econdatabases

Holmes A J and Anderson K (2017b) Convergence in National Alcohol ConsumptionPatterns New Global Indicators Journal of Wine Economics 12(2) 117ndash148

Ludington C C (2013) The Politics of Wine in Britain A New Cultural History BasingstokeUK Palgrave Macmillan

Office of National Statistics (2017) Analysis of real earnings London Office of NationalStatistics Available at httpswwwonsgovukemploymentandlabourmarketpeopleinworkearningsandworkinghoursarticlessupplementaryanalysisofaverageweeklyearningslatest

Rollo J Borchert I Dawar K Holmes P and Winters L A (2016) The World TradeOrganisation A safety net for a post-Brexit UK trade policy Briefing Paper 1UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-1-final-1pdf

Smith A (2017) Brexit Hard truths and hard choices UKTPO blog 19 June Available athttpsblogssussexacukuktpo20170619brexit-hard-truths-and-hard-choices

Swinbank A (2017) World trade rules and the policy options for British agriculture post-Brexit Briefing Paper 7 UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-7pdf

Viner J (1950) The Customs Union Issue New York Carnegie Endowment for InternationalPeace

Wine Australia (2015) Export Market Guide European Union Adelaide Wine AustraliaWittwer G Berger N and Anderson K (2003) A model of the worldrsquos wine markets

Economic Modelling 20(3) 487ndash506

Kym Anderson and Glyn Wittwer 251

Page 31: U.K. and Global Wine Markets by 2025, and Implications of ... · U.K. and Global Wine Markets by 2025, and Implications of Brexit* Kym Andersona and Glyn Wittwerb Abstract The United

Armington P A (1969) A theory of demand for products distinguished by place of produc-tion IMF Staff Papers 16 159ndash178

Baldwin R E (ed) (2016) Brexit Beckons Thinking Ahead by Leading Economists LondonCentre for Economic Policy Research

Campos N F (2016) Lousy experts Looking back at the ex ante estimates of the costs ofBrexit In R E Baldwin (ed) Brexit Beckons Thinking Ahead by Leading Economists(ch 3) London Centre for Economic Policy Research

Dixon P Parmenter B Sutton J and Vincent D (1982) ORANI A Multisectoral Modelof the Australian Economy Contributions to Economic Analysis 142 Amsterdam North-Holland

European Commission (2015) Trade for All Towards a More Responsible Trade andInvestment Policy Brussels European Commission

Harrison J Horridge M Jerie M and Pearson K (2014) GEMPACK ManualMelbourne GEMPACK Software

HM Treasury (2017) Spring Budget 2017 London HM TreasuryHolmes A J and Anderson K (2017a)Annual Database of National Beverage Consumption

Volumes and Expenditures 1950 to 2015 Wine Economics Research Centre University ofAdelaide posted at wwwadelaideeduauwine-econdatabases

Holmes A J and Anderson K (2017b) Convergence in National Alcohol ConsumptionPatterns New Global Indicators Journal of Wine Economics 12(2) 117ndash148

Ludington C C (2013) The Politics of Wine in Britain A New Cultural History BasingstokeUK Palgrave Macmillan

Office of National Statistics (2017) Analysis of real earnings London Office of NationalStatistics Available at httpswwwonsgovukemploymentandlabourmarketpeopleinworkearningsandworkinghoursarticlessupplementaryanalysisofaverageweeklyearningslatest

Rollo J Borchert I Dawar K Holmes P and Winters L A (2016) The World TradeOrganisation A safety net for a post-Brexit UK trade policy Briefing Paper 1UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-1-final-1pdf

Smith A (2017) Brexit Hard truths and hard choices UKTPO blog 19 June Available athttpsblogssussexacukuktpo20170619brexit-hard-truths-and-hard-choices

Swinbank A (2017) World trade rules and the policy options for British agriculture post-Brexit Briefing Paper 7 UKTPO University of Sussex Available at httpblogssussexacukuktpofiles201701Briefing-paper-7pdf

Viner J (1950) The Customs Union Issue New York Carnegie Endowment for InternationalPeace

Wine Australia (2015) Export Market Guide European Union Adelaide Wine AustraliaWittwer G Berger N and Anderson K (2003) A model of the worldrsquos wine markets

Economic Modelling 20(3) 487ndash506

Kym Anderson and Glyn Wittwer 251