course delivered at the regional research institute ......course delivered at the regional research...
TRANSCRIPT
An introduction to computable general equilibrium modelling for regional and environmental analyses
Course delivered at the Regional Research Institute, University of West Virginia, 10-12 August 2009
Dr Karen TurnerESRC Climate Change Leadership Fellow
Department of Economics, University of Strathclyde
ESRC ref: RES-066-27-0009
Course overview1. Introduction to multi-sectoral accounting and modelling – Input-Output analysis (Monday
9-11am)2. Introduction to CGE modelling (1) – Databases, calibration and specification (Monday
11.30am-12.30pm)3. Introduction to CGE modelling (2) - Application: the single region, 3-sector AMOS model
for Scotland, 1998, and introduction to Workshop 1 (Monday 1.30-3pm) 4. Workshop 1 – AMOS worksheet: simulating fiscal policy (Tuesday 9-11am)5. Review of AMOS fiscal policy simulations (Tuesday 11am-12.30pm)6. Extending and developing the AMOS framework to model increases in energy and labour
productivity and introduction to Workshop 2 (Tuesday 1.30-3pm) 7. Workshop 2 – Simulating increased labour efficiency (Wednesday 9am-12pm)8. Extending and developing the AMOS framework to model the interregional impacts of
increases in labour productivity (Wednesday 12-1pm)9. Using the AMOSUK interregional framework to demonstrate the value added from using
CGE analysis to model the impacts of a simple demand shock on economic and environmental indicators (Wednesday 2-3pm)
1. Introduction to multi1. Introduction to multi--sectoralsectoral accounting and accounting and modelling modelling –– InputInput--Output analysisOutput analysis
Monday 10 August, 9-11am
Input-output tables
• What is an Input-Output Table?– Records flows of inputs and outputs into individual sectors of the
economy
• Inputs– intermediate goods, imports, labour, other property services, taxes and
subsidies
• Outputs
– Intermediate demand, consumption, exports, tourism, government, investment, stock
1998 Scottish IO - 3-sector IxI Purchases by Sector Group (at basic prices): Final consumption expenditure:Manufacturing Non-Mfr Sheltered Sales to Total Final Gross output
Traded ID Local Capital External Demand (Total Demand)Manufacturing 3187.587 3896.852 775.760 7860.199 1545.663 387.604 28069.696 30002.963 37863.162Non-Manufacturing Traded 6072.304 18091.215 4726.423 28889.942 16506.298 5383.338 17609.990 39499.626 68389.568Sheltered 838.063 4669.769 5302.514 10810.346 20578.096 400.537 292.252 21270.885 32081.231
Total Intermediate Inputs 10097.955 26657.837 10804.696 47560.488 38630.057 6171.479 45971.938 90773.474 138333.962
Imports 13797.331 8496.241 3125.215 25418.787 17585.074 5024.481 363.471 22973.027 48391.813Net product & production taxes 1314.857 2671.122 754.917 4740.897 3457.135 588.111 1988.153 6033.399 10774.296Income from Employment 7076.821 19904.991 11414.186 38395.998 0.000 0.000 0.000 0.000 38395.998Other Value Added 5576.205 10659.378 5982.205 22217.789 0.000 0.000 0.000 0.000 22217.789
Total Primary Inputs 27765.214 41731.733 21276.522 90773.470 21042.210 5612.593 2351.624 29006.426 119779.896
TOTAL INPUTS 37863.169 68389.570 32081.219 138333.957 59672.267 11784.071 48323.562 119779.900 258113.858
Total employment (000 FTE) 346.144 1109.426 573.930Empoyment-output coefficients (th/£1m) 0.009 0.016 0.018
Composition of GDPComposition of GDP
• Sectoral gross value-added
• GDP at basic and market prices
• Reconciliation income, output and expenditure measures of GDP
Social Accounting Matrices (Social Accounting Matrices (SAMsSAMs))
• IO tables give us value of goods and services produced – i.e. generation of income
• But distribution of income?
• Extend IO tables to show income transfers – income actually accruing to local residents
• Total income and expenditure
• GNP and GDP
• GNP = GDP plus transfers of income earned abroad by local residents minus income earned locally by foreign residents
SCOTTISH SAM (1998, £million)Tourist Other Net. Commd.
Manuft. Non-manu.tSheltered HouseholdsCorporate GovernmentCapital Stocks Expend RUK ROW Labour Value Added Tax TotalManuft. 3187.6 3896.9 775.8 1545.7 0.0 0.0 371.2 16.5 80.9 9939.9 18048.9 0.0 0.0 0.0 37863.2Non-manu.t 6072.3 18091.2 4726.4 13239.4 0.0 3266.9 5058.2 325.2 1369.1 12619.2 3621.7 0.0 0.0 0.0 68389.6Sheltered 838.1 4669.8 5302.5 7068.7 0.0 13509.4 362.7 37.8 70.2 215.4 6.7 0.0 0.0 0.0 32081.2RUK 6901.0 7082.9 2413.3 10564.5 3266.0 3242.0 3628.3 110.7 251.8 0.0 0.0 0.0 0.0 0.0 37460.6 Goods and Services 6901.0 7082.9 2413.3 10564.5 0.0 0.0 3628.3 110.7 251.8 0.0 0.0 0.0 0.0 0.0 30952.6 Transfers 0.0 0.0 0.0 0.0 3266.0 3242.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 6508.0ROW 6896.3 1413.4 711.9 7020.5 3399.3 0.0 1246.4 39.1 111.6 0.0 0.0 0.0 0.0 0.0 20838.5 Goods and Services 6896.3 1413.4 711.9 7020.5 0.0 0.0 1246.4 39.1 111.6 0.0 0.0 0.0 0.0 0.0 17439.2 Transfers 0.0 0.0 0.0 0.0 3399.3 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3399.3Net commodity taxes 1314.9 2671.1 754.9 3373.6 0.0 83.6 570.2 17.9 121.7 1560.3 306.1 0.0 0.0 0.0 10774.3Labour 7076.8 19905.0 11414.2 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 38396.0Other Value Added 5576.2 10659.4 5982.2 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 22217.8Sales by final demand 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0Households 0.0 0.0 0.0 0.0 13882.8 10528.1 0.0 0.0 0.0 2180.1 0.0 38396.0 3036.0 0.0 68023.0Corporate 0.0 0.0 0.0 6138.5 0.0 1535.7 0.0 0.0 0.0 853.0 819.5 0.0 17891.6 0.0 27238.4Government 0.0 0.0 0.0 12950.0 5080.0 0.0 0.0 0.0 0.0 3005.5 0.0 0.0 1290.2 10774.3 33100.0Capital 0.0 0.0 0.0 6122.1 1610.2 934.3 0.0 0.0 0.0 6084.6 -2967.1 0.0 0.0 0.0 11784.1Total 37863.2 68389.6 32081.2 68023.0 27238.4 33100.0 11236.9 547.1 2005.3 36458.0 19835.8 38396.0 22217.8 10774.3 408166.6
History of inputHistory of input--outputoutput
• Francois Quesnay (1694-1774)– Physiocrat– Tableau Économique– Table showing how each part of the economy serves and is compensated
by each of the others
• Wassily Leontief (Harvard) – 1930s– Nobel Prize for Economic Science 1973– Modern SNA (UN)– Also, basis for accounting and modelling frameworks used for empirical
general equilibrium analyses (Leon Walras, 1834-1910)
InputInput--Output tables as regional and national accountsOutput tables as regional and national accounts
Links for IO (and SAM)
UN SNA 1993http://unstats.un.org/unsd/sna1993/toctop.asp.
UK National Statisticshttp://www.statistics.gov.uk/CCI/nscl.asp?ID=5871&x=8&y=6.
Scottish Government
http://www.scotland.gov.uk/Topics/Statistics/Browse/Economy.
Different formats of IO tablesDifferent formats of IO tables
• Supply and Use Tables (purchaser prices)
• Analytical IO tables – Producer prices (basic prices)– Symmetric – IxI or CxC– Appropriate for analytical purposes – multiplier analyses and modelling
Analytical IO tablesAnalytical IO tables
• Quantities in value terms• Basic/producer prices• Example of single entry accounting• 4 quadrants • Balancing identities:• Sectoral gross input = gross output• Total intermediate demand = total intermediate expenditure• Total primary demand = total final expenditure• Aggregate gross input = gross output
1998 Scottish IO - 3-sector IxI Purchases by Sector Group (at basic prices): Final consumption expenditure:Manufacturing Non-Mfr Sheltered Sales to Total Final Gross output
Traded ID Local Capital External Demand (Total Demand)Manufacturing 3187.587 3896.852 775.760 7860.199 1545.663 387.604 28069.696 30002.963 37863.162Non-Manufacturing Traded 6072.304 18091.215 4726.423 28889.942 16506.298 5383.338 17609.990 39499.626 68389.568Sheltered 838.063 4669.769 5302.514 10810.346 20578.096 400.537 292.252 21270.885 32081.231
Total Intermediate Inputs 10097.955 26657.837 10804.696 47560.488 38630.057 6171.479 45971.938 90773.474 138333.962
Imports 13797.331 8496.241 3125.215 25418.787 17585.074 5024.481 363.471 22973.027 48391.813Net product & production taxes 1314.857 2671.122 754.917 4740.897 3457.135 588.111 1988.153 6033.399 10774.296Income from Employment 7076.821 19904.991 11414.186 38395.998 0.000 0.000 0.000 0.000 38395.998Other Value Added 5576.205 10659.378 5982.205 22217.789 0.000 0.000 0.000 0.000 22217.789
Total Primary Inputs 27765.214 41731.733 21276.522 90773.470 21042.210 5612.593 2351.624 29006.426 119779.896
TOTAL INPUTS 37863.169 68389.570 32081.219 138333.957 59672.267 11784.071 48323.562 119779.900 258113.858
Total employment (000 FTE) 346.144 1109.426 573.930Empoyment-output coefficients (th/£1m) 0.009 0.016 0.018
IO attribution and modelling analysis (1)IO attribution and modelling analysis (1)
– Convert the IO accounts/tables into an economic model. – Assumptions– Regional IO models typically used for 3 primary forms of
analysis:
1. identification of the ultimate source of regional economic activity
2. measurement of the interdependencies and interactions 3. impact analyses
IO attribution and modelling analysisIO attribution and modelling analysis
• 1&2 - attribution analysis; 3 - modelling or marginal analysis
• IO makes a key distinction between sales to intermediate and final markets
– sales to final markets exogenous– sales to intermediate markets endogenous
• Particular causal sequence imposed
• Exogenous final demands drive endogenous intermediate demands
• Transmission mechanism – multipliers
IO multiplier analysesIO multiplier analyses
• Concept of multiplier – Keynes
• IO facilitates multiplier analysis at sectoral level
• Distinguish direct, indirect and induced effects (Type I and Type II)
• Construct Leontief inverse or multiplier matrix
• Assumptions about average technologies – input-output coefficients
Conventional demand-driven IO model (1)
Output in each sector = sum intermediate and final demand. For our 3x3 (ixj, i=j) table: (1) X1 = x1,1 + x1,2 + x1,3 + Y1
X2 = x2,1 + x2,2 + x2,3 + Y2
X3 = x3,1 + x3,2 + x3,3 + Y3
Input-output coefficients: ai,j = xi,j/Xj
Conventional demand-driven IO model (2)
(2) X1 = a1,1X1 + a1,2X2 + a1,2X3 + Y1
X2 = a2,1X1 + a2,2X2 + a2,3X3 + Y2
X3 = a3,1X1 + a3,2X2 + a3,3X3 + Y3
(3) (1-a1,1)X1 – a1,2X2 – a1,3X3 = Y1
a2,1X1 – (1-a2,2)X2 – a2,3X3 = Y2
a3,1X1 – a3,2X2 – (1-a3,3)X3 = Y3
Conventional demand-driven IO model (3)
In matrix notation: (3) [I-A]X = Y Key equation: (4) X = [I-A]-1.Y
[I-A]-1 is the Leontief inverse ⇒ column totals give us the OUTPUT MULTIPLIERS
A-MATRIX (input coefficients)Manufacturing Non-Manufacturing Sheltered Household
Traded ExpenditureManufacturing 0.084187023 0.056980216 0.02418111 0.04025583 3187.6/37863.2=0.08419Non-Manufacturing Traded 0.160374965 0.264531789 0.14732678 0.34481094Sheltered 0.022133989 0.068281888 0.16528406 0.18410089Income from Employment 0.186905139 0.291053027 0.35579028 0
I MATRIXManufacturing Non-Manufacturing Sheltered Household
Traded ExpenditureManufacturing 1 0 0 0Non-Manufacturing Traded 0 1 0 0Sheltered 0 0 1 0Income from Employment 0 0 0 1
I-A MATRIXManufacturing Non-Manufacturing Sheltered Household
Traded ExpenditureManufacturing 0.915812977 -0.056980216 -0.02418111 -0.04025583Non-Manufacturing Traded -0.160374965 0.735468211 -0.14732678 -0.34481094Sheltered -0.022133989 -0.068281888 0.83471594 -0.18410089Income from Employment -0.186905139 -0.291053027 -0.35579028 1
TYPE I LEONTIEF INVERSEManufacturing Non-Manufacturing Sheltered RECREATING BASE
Traded FINAL CONSUMPTION
ESTIMATED GROSS OUTPUT
ACTUAL GROSS OUTPUT
Manufacturing 1.108914406 0.090376289 0.04807579 Manufacturing 30002.96 37863.1619 37863.1622Non-Manufacturing Traded 0.25182482 1.402853201 0.25489779 TIMES Non-Manufacturing Traded 39499.63 EQUALS 68389.5691 68389.5682Sheltered 0.050004764 0.11715345 1.22013845 Sheltered 21270.89 32081.2334 32081.2312TYPE I OUTPUT MULTIPLIERS 1.411 1.610 1.523
TYPE II LEONTIEF INVERSEManufacturing Non-Manufacturing Sheltered Household RECREATING BASE
Traded Expenditure FINAL CONSUMPTION
ESTIMATED GROSS OUTPUT
ACTUAL GROSS OUTPUT
Manufacturing 1.143428224 0.14438635 0.10791737 0.11568333 Manufacturing 28457.3 37863.1623 37863.1622Non-Manufacturing Traded 0.472305202 1.747879033 0.63717635 0.73900558 Non-Manufacturing Traded 26260.27 68389.5718 68389.5682Sheltered 0.158877358 0.287526238 1.40890653 0.36491888 Sheltered 14202.15 32081.2348 32081.2312Income from Employment 0.407705489 0.638011074 0.70689766 1.36654621 Income from employment 0 38396.003 38395.9979TYPE II OUTPUT MULTIPLIERS (i) 1.775 2.180 2.154 1.220TYPE II OUTPUT MULTIPLIERS (ii) 2.182 2.818 2.861 2.586
Type I and II multiplier analysis Type I and II multiplier analysis
• Type I multipliers capture direct plus indirect effects– inter-industry (backward linkages)– rounds of multiplier
• Household consumption treated as an exogenously determined final demand expenditure
• Conventional to endogenise household consumption, by moving out of vector of final demand and into Leontief inverse
• Type II multipliers capture direct plus indirect plus induced effects– consumption and income effects
• Other possibilities – e.g. endogenise trade and capital formation; activity driven by local final consumption demand
IO attribution analysisIO attribution analysis
• IO – a snapshot picture of the economy for a given point in time (accounting period – 1 year)
• Use multipliers to carry out descriptive analysis of structure of economic activity during that period
• Based on average technologies in A-matrix
• E.g. Type I manufacturing multiplier of 1.41 tells us that for every £1 of final demand for manufacturing output, £1.41 of output was generated in all 3 production sectors
• What share of total output, GDP and/or employment was supported by demand for manufacturing outputs?
• What share of total output, GDP and/or employment was supported by export demand for manufacturing outputs?
Example of use of IO for attribution analysisExample of use of IO for attribution analysis
• Ecological or carbon footprints
• Common to extend IO tables for environmental variables– Pollution or resource use in physical units (as with employment)
• Partial application of Leontief (1970) pollution model– Full Leonfief pollution model examines resource implications of dealing with pollution – feasibility of
internalising negative externality of pollution?
• Attribution of pollution generation in local economy to final demands that drive activity
• But in a single region analysis some pollution attributed to external demand, no account taken of pollution embodied in imports
• Use of inter-regional or multi-region environmental IO analysis to account for emissions from a consumption perspective – e.g. carbon footprints; wide range of accounting perspectives
Average and marginal analyses (1) Average and marginal analyses (1)
Accounting – structure of the economy in the accounting period (IO as a ‘snapshot’)
E.g. Scottish production attributable to external consumption demand in the year the IO table are reported for:
(5) XE = [I-A]-1.YE
Multipliers as a tool for average analysis
Average and marginal analyses (2) Average and marginal analyses (2)
Modelling – impact of a change in exogenous final demand
E.g. An increase in external consumption demand:
(6) ΔX = [I-A]-1.ΔYE
Multipliers as a tool for marginal analysis
Latter subject to restrictive assumptions:1. Universal Leontief technology2. No supply constraints
Extended IO method Extended IO method –– e.g. employment e.g. employment Output-employment coefficients
(7) ei = Ei/Xi
where • Ei is the number of FTE workers employed in sector i
E.g. Employment, E, attributable to external consumption demand:
(8) EE = ei[I-A]-1.YE
Output-employment multipliers:
ei[I-A]-1
E.g. An increase in external consumption demand:
(9) ΔE = ei[I-A]-1.ΔYE
IO Modelling example IO Modelling example
• 20% increase in government final consumption (part of local exogenous final consumption demand) in 3-sector
• Here, for simplicity, assume same pattern as in base case
• 1999 12-sector Scottish IO tables
IO Modelling assumptions IO Modelling assumptions
• Passive supply – Supply-driven IO – demand passive
• Silent on prices– Price IO – silent on quantities
• Universal Leontief (fixed proportions) technology– Output rises by X%, use of all inputs rises by X%– No response to changes in relative prices
• Short-run excess capacity?• Long-run equilibrium (once all supply constraints relaxed)? Adjustment process?
IO as a longIO as a long--run equilibrium model (1) run equilibrium model (1)
• Relax assumptions of universal Leontief technology and passive supply –computable general equilibrium (CGE)
• Take a CGE and assume all constraints on supply are relaxed in long-run
• IO results may be recreated as a long-run equilibrium
• Crowding out and price changes in short-run, but, once all factor markets have adjusted, there will be no change in real prices
IO as a longIO as a long--run equilibrium model (2) run equilibrium model (2)
• Only where appropriate to assume that the supply-side is able to fully adjust through capital accumulation and factor mobility
• Appropriate assumption for regional economies?
• Only in cases of pure demand-side shocks
• If we have changes in supply-side behaviour/relationships, long-run impacts on prices
• Generally, IO not appropriate for modelling supply-side disturbances
Developing CGE models using IO/SAM databases (1)Developing CGE models using IO/SAM databases (1)
• We can augment the demand side provided by the I-O model to accommodate a corresponding supply-side (and more theory-consistent demand behavour) for each sector and for the economy as a whole, and incorporate dynamics
– SAM data plus additional information on labour market and investment demands
• A Computable General Equilibrium (CGE) model allows us to do this– can easily capture constraints on capacity of each sector in the short-run and on total
labour supply– wages and competitiveness explicitly modelled– dynamics through updating labour supply due to migration (if any) and sectoral capacity
updated by sectoral investment
Developing CGE models using IO/SAM databases (2)Developing CGE models using IO/SAM databases (2)
• The CGE therefore significantly adds to what I-O can do:
– supply and demand both captured and “matter”
– generally both simultaneously determine prices and quantities
– though I-O results would still apply under some circumstances (IO models are in fact a special case of CGEs, with simple supply and technology assumptions and so linearity)
– Theoretical basis – Walrasian general equilibrium theory, but no longer limited to classical economy (market clearing etc)
2. Introduction to CGE modelling (1) 2. Introduction to CGE modelling (1) –– databases, databases, calibration and specificationcalibration and specification
Monday 10 August, 11.30am-12.30pm
Indicative readingIndicative reading
• Pyatt, G. and J.I. Round (1985) eds. Social Accounting Matrices: A Basis for Planning, The World Bank, Washington, D.C., U.S.A.
• Reinert, K. A. and D. W. Roland-Holst (1997) ‘Social Accounting Matrices’, J. F. Francois and K. A. Reinert (eds), Applied Methods for Trade Policy Analysis: A Handbook, Cambridge University Press, Cambridge
• Devarajan, S., Go, D.S., Lewis, J.D., Robinson, S. and P. Sinko (1997), “Simple General Equilibrium Modelling” in J.F. Francois and K. A. Reinert eds. Applied Methods for Trade Policy Analysis, Cambridge University Press, Cambridge.
• Greenaway, D., Leybourne, S.J., Reed, G.V. and J. Whalley (1993), Applied General Equilibrium Modelling: Applications, Limitations and Future Developments, HMSO, London.
The basics and essentialsThe basics and essentials
• Applied/computable general equilibrium analysis involves simulating numerically the general equilibrium structure of the economy.
• Basic theoretical framework: the Walrasian general equilibrium system
Essential feature:• Supply equals demand in all markets at a set of relative prices that can be identified
• In practice, general equilibrium models are not restricted to the conventional Walrasian model of perfect competition.
• The essential assumption is that an equilibrium for the economy exists and it is unique
The key issues/steps in CGE model developmentThe key issues/steps in CGE model development
Greenaway et al (1993) give a succinct outline of the key issues in laying out the structure of a general equilibrium model (first step in any CGE analysis)
1. Dimensionality – the level of sectoral disaggregation of total economic activity (i.e. the number of products/production sectors and factors of production)
2. General specification of key relationships (including functional form) – supply and demand equations (including the interdependencies and interactions between sectors)
3. Collection of benchmark data – to model the benchmark case/initial equilibrium4. Calibration of the model’s parameters to that data set – while key parameter values
will be pre-specified, calibration involves choosing the remaining parameter values so that the model can reproduce the data set as an equilibrium solution
The key issues/steps in CGE model developmentThe key issues/steps in CGE model development
Once the structure is in place the model is solved for general equilibrium.
Then:• equilibrium relationships and interdependencies can be traced and examined • counterfactual equilibria can be computed for exogenous changes • questions of policy evaluation, including distributional effects, can be addressed.
The CGE framework is a very flexible one, so it is useful to go through these issues for a particular type of application – e.g. IO
The key issues/steps in CGE model developmentThe key issues/steps in CGE model development
Dimensionality• What and how many sectors from the IO accounts should be identified in the model?
Data collection. • Data on sectoral demand and supply, factor use and rewards, total GDP etc are collected to build the
transactions table – last week. This forms the benchmark equilibrium data set.
Choice of functional form • To move from accounting (transactions/IO tables) to IO modelling, the conventional basic IO assumption is that:
(1) Xij = aijXj
• where aij is a constant. What this tells us is that IO assumes fixed technical coefficients in production: Leontieftechnology.
• Thus, all IO relationships are of a fixed linear form, and are not subject to any supply constraints.
Calibration to benchmark equilibrium • Calibration to the base year equilibrium involves specifying the value of some parameters and running the
model so that it recreates base year, solving for all unknown parameters• This is done for IO through the key equation: X=[I-A]-1
Analysis using model. • With the model calibrated for the benchmark equilibrium, in the case of IO, exogenous demand changes (e.g.
20% increase in government expenditure) can be specified, counterfactual equilibria can be computed, and policy appraisal carried out based on comparison between the counterfactual and benchmark equilibria.
Core CGE database Core CGE database –– the Social Accounting Matrix, SAMthe Social Accounting Matrix, SAM
• From this morning, IO table augmented with information on income transfers between aggregate transactors (households, firms, government, external sector(s), capital account
• Construct a set of income-expenditure accounts
• Balance manually (retain integrity of initial, published, IO table)
Schematic Structure of a Basic SAM
Expenditure by Production Activities Institutions Factors of Production
(the I production sectors) H C G CF E Labour (L) Capital (K)
Income to
Production T U Activities
Institutions:
H
C
G V W X CF
E
Factors of
Production : L Y
(Value Added) K
Template for constructing incomeTemplate for constructing income--expenditure accountsexpenditure accountsIncome Expenditure
Income from employment (H)* IO final demand expenditure
Income from OVA* (incl expend taxes)*
Net commodity taxes (G)*
Payments from corporations** Payments to corporations **
Payments from government** Payments to government**
Payments from households** Payments to households**
Transfers from UK** Transfers to UK**
Transfers from REU** Transfers to REU**
Transfers from ROW** Transfers to ROW**
Payments to capital (savings)***
I N C O M E - E X P E N D I T U R E A C C O U N T S - S C O T L A N D 1 9 9 8
H o u s e h o l d sI n c o m e 6 8 , 0 2 3 . 0 0 E x p e n d i t u r e 6 8 , 0 2 3 . 0 0I n c o m e f r o m e m p l o y m e n t 3 8 , 3 9 6 I O e x p e n d i t u r e 4 2 , 8 1 2P r o f i t i n c o m e ( O V A ) 3 , 0 3 6 P a y m e n t s t o c o r p o r a t i o n s 6 , 1 3 9I n c o m e f r o m c o r p o r a t i o n s 1 3 , 8 8 3 P a y m e n t s t o g o v e r n m e n t 1 2 , 9 5 0I n c o m e f r o m g o v e r n m e n t 1 0 , 5 2 8 P a y m e n t s t o c a p i t a l 6 , 1 2 2T r a n s f e r s f r o m R U K 2 , 1 8 0 6 8 , 0 2 3 . 0 0
6 8 , 0 2 3 . 0 0 S E S t o t a l S c o t e x p 6 8 0 2 3 . 0 0
C o r p o r a t i o n sI n c o m e 2 7 , 2 3 8 . 3 5 E x p e n d i t u r e 2 7 , 2 3 8 . 3 5P r o f i t i n c o m e ( O V A ) 1 7 , 8 9 2 P a y m e n t s t o h o u s e h o l d s 1 3 , 8 8 3I n c o m e f r o m h o u s e h o l d s 6 , 1 3 9 P a y m e n t s t o g o v e r n m e n t 5 , 0 8 0I n c o m e f r o m g o v e r n m e n t 1 , 5 3 6 T r a n s f e r s t o R U K 3 , 2 6 6I n c o m e f r o m R U K 8 5 3 T r a n s f e r s t o R O W 3 , 3 9 9I n c o m e f r o m R E U n / a P a y m e n t s t o c a p i t a l ( s a v i n g s ) 1 , 6 1 0I n c o m e f r o m R O W 8 2 0 2 7 , 2 3 8 . 3 5
2 7 , 2 3 8 . 3 5
G o v e r n m e n tI n c o m e 3 3 , 1 0 0 . 0 0 E x p e n d i t u r e 3 3 , 1 0 0 . 0 0P r o f i t i n c o m e ( O V A ) 1 , 2 9 0 . 2 0 I O e x p e n d i t u r e 1 6 8 5 9 . 8 6N e t c o m m o d i t y t a x e s 1 0 , 7 7 4 . 3 0 P a y m e n t s t o c o r p o r a t i o n s 1 , 5 3 5 . 7 3I n c o m e f r o m h o u s e h o l d s 1 2 , 9 5 0 . 0 0 P a y m e n t s t o h o u s e h o l d s 1 0 5 2 8 . 1I n c o m e f r o m c o r p o r a t i o n s 5 , 0 8 0 . 0 0 T r a n s f e r s t o R U K 3 2 4 2I n c o m e f r o m R U K 3 , 0 0 5 . 5 1 T r a n s f t e r s t o R O W 0I n c o m e f r o m R O W 0 . 0 0 P a y m e n t s t o c a p i t a l ( s a v i n g s ) 9 3 4 . 3 1
3 3 , 1 0 0 . 0 0 3 3 1 0 0 . 0 0G E R S t o t a l e x p 3 3 1 0 0
C a p i t a lI n c o m e 1 1 , 7 8 4 . 0 7 E x p e n d i t u r e 1 1 , 7 8 4 . 0 7H o u s e h o l d s 6 , 1 2 2 . 0 7 I O e x p e n d i t u r e 1 1 7 8 4 . 0 7 1 2 7C o r p o r a t e 1 , 6 1 0 . 2 4G o v t 9 3 4 . 3 1R U K / R O W 3 , 1 1 7 . 4 5
1 1 , 7 8 4 . 0 7
E x t e r n a lU K i n c o m e f r o m S c o t l a n d 3 7 , 4 6 0 . 6 4 U K e x p e n d i t u r e i n S c o t l a n d 3 0 3 7 3 . 4 1 9 0 6G o o d s & S e r v i c e s 3 0 , 9 5 2 . 6 3 G o o d s & S e r v i c e s 2 4 3 3 4 . 8 0 6 6 5T r a n s f e r s 6 , 5 0 8 . 0 1 T r a n s f e r s 6 0 3 8 . 6 1 2 4 0 7
R O W i n c o m e f r o m S c o t l a n d 2 0 , 8 3 8 . 5 1 R O W e x p e n d i t u r e i n S c o t l a n d 2 2 8 0 2 . 9 4 5 9 6G o o d s & S e r v i c e s 1 7 , 4 3 9 . 1 9 G o o d s & S e r v i c e s 2 1 9 8 3 . 4 1 7 4 8T r a n s f e r s 3 , 3 9 9 . 3 2 T r a n s f e r s 8 1 9 . 5 2 8 4 8 5 4
T o u r i s t e x p e n d i t u r e i n S c o t l a n d 2 0 0 5 . 3 3 7 6 5 7
T o t a l i n c o m e 5 8 , 2 9 9 . 1 5 T o t a l e x p e n d i t u r e 5 5 , 1 8 1 . 7 0S u r p l u s / d e f i c i t 3 1 1 7 . 4 4 7 2 7
SCOTTISH SAM (1998, £million)Tourist Other Net. Commd.
Manuft. Non-manu.tSheltered HouseholdsCorporate GovernmentCapital Stocks Expend RUK ROW Labour Value Added Tax TotalManuft. 3187.6 3896.9 775.8 1545.7 0.0 0.0 371.2 16.5 80.9 9939.9 18048.9 0.0 0.0 0.0 37863.2Non-manu.t 6072.3 18091.2 4726.4 13239.4 0.0 3266.9 5058.2 325.2 1369.1 12619.2 3621.7 0.0 0.0 0.0 68389.6Sheltered 838.1 4669.8 5302.5 7068.7 0.0 13509.4 362.7 37.8 70.2 215.4 6.7 0.0 0.0 0.0 32081.2RUK 6901.0 7082.9 2413.3 10564.5 3266.0 3242.0 3628.3 110.7 251.8 0.0 0.0 0.0 0.0 0.0 37460.6 Goods and Services 6901.0 7082.9 2413.3 10564.5 0.0 0.0 3628.3 110.7 251.8 0.0 0.0 0.0 0.0 0.0 30952.6 Transfers 0.0 0.0 0.0 0.0 3266.0 3242.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 6508.0ROW 6896.3 1413.4 711.9 7020.5 3399.3 0.0 1246.4 39.1 111.6 0.0 0.0 0.0 0.0 0.0 20838.5 Goods and Services 6896.3 1413.4 711.9 7020.5 0.0 0.0 1246.4 39.1 111.6 0.0 0.0 0.0 0.0 0.0 17439.2 Transfers 0.0 0.0 0.0 0.0 3399.3 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3399.3Net commodity taxes 1314.9 2671.1 754.9 3373.6 0.0 83.6 570.2 17.9 121.7 1560.3 306.1 0.0 0.0 0.0 10774.3Labour 7076.8 19905.0 11414.2 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 38396.0Other Value Added 5576.2 10659.4 5982.2 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 22217.8Sales by final demand 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0Households 0.0 0.0 0.0 0.0 13882.8 10528.1 0.0 0.0 0.0 2180.1 0.0 38396.0 3036.0 0.0 68023.0Corporate 0.0 0.0 0.0 6138.5 0.0 1535.7 0.0 0.0 0.0 853.0 819.5 0.0 17891.6 0.0 27238.4Government 0.0 0.0 0.0 12950.0 5080.0 0.0 0.0 0.0 0.0 3005.5 0.0 0.0 1290.2 10774.3 33100.0Capital 0.0 0.0 0.0 6122.1 1610.2 934.3 0.0 0.0 0.0 6084.6 -2967.1 0.0 0.0 0.0 11784.1Total 37863.2 68389.6 32081.2 68023.0 27238.4 33100.0 11236.9 547.1 2005.3 36458.0 19835.8 38396.0 22217.8 10774.3 408166.6
Additional Data Requirements for CGEAdditional Data Requirements for CGE• The IO table and SAM contain information on which sectors’ outputs are used for the
purposes of capital formation. However, what they do not tell us is which sectors the demand for this capital formation comes from, so require (even estimated) investment demand data (we base on share other value-added, payments to capital, in base year)
• IO tables tend to also report FTE employment by sector
• This is sufficient labour market data for a demand-driven general equilibrium model like IO.
• However, for a more flexible CGE framework with an active supply-side, more information is required on supply conditions in the labour market.
• We also require data on on the structure of the aggregate labour market, such as base yearworking age population, participation rate and unemployment.
• Prices of labour, capital; some prices indexed to 1
CGE model specificationCGE model specification
CGE modelling more difficult to generalise than IO modelling:• CGEs characterized by huge variety of model types• Heterogeneity of views of supply-side, of technology• Nonetheless, it is possible to identify some common features:• Theoretial “roots” in Walrasian GE theory• Elements of model construction• Motivated by an attempt to shed light on key aspects of economic policy
Common theoretical antecedentsCommon theoretical antecedents
Schematic of a typical CGE reminiscent of simplest micro “circular flow” diagram, often characterized by
• disaggregation of households and industries/commodities• all commodity and factor markets clear simultaneously• factor prices and employment levels just sufficient to generate incomes and
demands that equal commodity supplies
Formal CGE theory initially:• assumed universal perfect competition• focused on providing existence of equilibrium in exchange and then production
economies
Contemporary CGE Contemporary CGE modellingmodelling
Existence of equilibrium typically taken as given
Solved routinely without Scarf-type algorithms – software developments C++, GAMS etc
CGEs – different degrees of imperfect competition now e.g. development; monopolistic competition, public economics
CGE requires simultaneous equilibrium in all markets
Flows? Source of supply and demand in each type of market?• can be complex• many goods and factor markets;• other transactors: government, ROW;• possible imperfections in markets
CGE model specification and calibrationCGE model specification and calibration
As with IO, we must decide• What sectors/activities/transactors we want to identify (dimensionality/specification)• For CGE also macroeconomic ‘closure’
– Development models often assume ‘Keynesian closure’ and involuntary unemployment (large excess capacity)
– Many models assume full employment - e.g. NAFTA studies– Closure matters!! – Models may have range of alternatives
• Degree of competition– Perfect competition still common in public economics– Monopolistic competition and returns to scale (Dixit/Stiglitz) commonly assumed in
trade models• Dynamic models
– Intertemporal or recursive dynamic• Single country/region and multi-country/region
Parameterisation and calibration (1)Parameterisation and calibration (1)
• Also require data to inform parameters/calibration• IO – just the A-matrix• CGE – more complex, but rooted in economic theory
– E.g. utility maximizing hhs and profit maximising/cost minimising firms– Need to specify/estimate functional forms and parameters (may involve non-
linearity)– Common use of hierarchical production and consumption structures
(separability assumptions)
Parameterisation and calibration (2)Parameterisation and calibration (2)
Three ways parameters are determined
1. By base year data (e.g. labour intensity = sectoral FTE employment /sectoraloutput) – structural parameters (change with structure of economy)
2. Exogenously imposed – e.g. estimated/informed elasticities of substitution in production, price elasticity of demand etc – key parameters
3. Calibrated – solve for unknown parameters
Parameterisation and calibration (3)Parameterisation and calibration (3)
• Calibration is normally to base year SAM• Values of ‘key parameters’ identified first
– Ideally, econometric estimation of individual relationships by modeler or external (secondary)
• Then all remaining parameters determined through reconciliation to base year SAM
Features of calibrated models – source of criticism• assume no errors or omitted variables and “estimate” parameters on a single
observation• no statistical measure of goodness of fit
Parameterisation and calibration (4)Parameterisation and calibration (4)
However: • Complete econometric estimation of CGE models not yet feasible and few ‘pure’
econometric models• Can conduct sensitivity analysis (not only parametric – also model closures)• Can accommodate benefits of econometric estimation where available
– Particularly key parameters (e.g. current work on AMOS Chicago, UK energy model)
• Model uses should reflect nature (recognise no precise point estimates)• Indeed, focussed and systematic sensitivity analysis can help develop
theoretical understanding (e.g. current work using Scottish and UK energy CGE models to understand ‘rebound’ effects of increased energy efficiency)
Model solutionModel solution
• Complex algorithms – won’t go into here but can purchase software such as GAMS for this
(choices of solver)
• Compare to IO – linearity means [I-A]-1 Leontief inverse is all required
Analysis using CGE modelsAnalysis using CGE models
Common motivating feature of CGEs: analyzing and evaluating the impacts of policies • But also other (non-policy) disturbances – e.g. energy CGEs in 1970s; resurgence from
90s to address climate change issues
Policies addressed include• regional, environmental, development and structural policies
Effects on• economy, energy, environment • often overall welfare impacts
Model dimensionality and specification will depend both on the target economy and the type of problem(s) it is designed to analyse
E.g. AMOS CGE modelling programme at FAI and Economics, University Strathclyde
3. Introduction to CGE modelling (2) 3. Introduction to CGE modelling (2) ––Application: the single region, 3Application: the single region, 3--sector sector
AMOS model for Scotland, 1998AMOS model for Scotland, 1998
Monday 10 August, 1.30-3pm
Basic readingBasic reading
AMOS summary paper (document ‘WVU course Aug 09_AMOS summary paper’), which is based on
• Harrigan, F., McGregor, P.G., Dournashkin, N., Perman, R., Swales, J.K. and Yin, Y.P. (1991) ‘AMOS: A macro-micro model of Scotland’, Economic Modelling, Vol.10, pp424-479.
Dimensionality/model specification (1)Dimensionality/model specification (1)
The initial AMOS framework incorporates: • 3 transactor groups – households, firms and government• 3 commodities and activities – manufacturing (M), non-manufacturing traded
(NMT), and non-traded/sheltered (NT) • 2 exogenous external transactors – for the case of Scotland (or any UK region),
these are the rest of the UK (RUK) and the rest of the world (ROW)
There are four main components of final demand:1. (Household) consumption, which is a linear homogenous function of real
disposable income.2. Investment is treated in various ways, dependent upon the particular model
closure that is chosen:
Dimensionality/model specification (2)Dimensionality/model specification (2)
• In the short- and medium-run the capital stock and its sectoral composition are fixed so that, even where investment is endogenous, capital stocks are not updated.
• In long-run equilibrium investment for each sector is endogenous and equal to depreciation with sectoral capital stocks set at their desired, cost-minimisinglevels.
• In the multi-period variant of the model, investment in each period is equal to depreciation plus some fraction of the gap between actual and desired capital stock
Dimensionality/model specification (3)Dimensionality/model specification (3)
3. Government expenditure, which is taken to be exogenous, or can be made endogenous linked to changes in income tax
4. Exports, where exports and imports are determined via an Armington link, making them relative price sensitive. The Armington trade substitution elasticity determined by the model user (or informed by econometric work).
User can choose between perfect and imperfect competition and between different macroeconomic and labour market closures
DatabaseDatabase
• IO tables augmented with other data for SAM
• Here (teaching model), 1998 Scottish IO tables (aggregated) augmented with other data for SAM
ParameterisationParameterisation/calibration/calibration
• Production and consumption structures fixed until recently (varying in current research with energy model)
• but option varies across application in terms of functional forms, parameter values etc
Figure 1 – Production Structure In The Basic (3-sector) AMOS Framework
GROSS OUTPUT INTERMEDIATES VALUE-ADDED ROW composite UK composite LABOUR CAPITAL
RUK composite LOCAL composite
Comm. j = 1…3
Production in AMOS (1)Production in AMOS (1)
• In all model configurations cost-minimisation in production is imposed• multi-level production functions
• generally of constant elasticity of substitution (CES) form
• so there can be input substitution in response to relative price changes
• but with Leontief and Cobb-Douglas (CD) available as special cases.
• In the CES functions, elasticities of substitution, σ, as with all parameter values, can be set for individual applications according to econometric or ‘best guess’ estimates.
Production in AMOS (2)Production in AMOS (2)
• The production inputs are labour (L), capital (K) and intermediates (J), with a choice between locally produced intermediate commodities and imports from RUK or ROW.
• The prices of the intermediate goods that make up the intermediate composite are required.
• All local input prices are endogenous to the system• All import prices are exogenous. • The precise nature of the intermediate composite, J, depends on relative
prices and the possibilities for substitution between different sources and types of intermediate input at each level.
• The precise form of the wage equation depends on what type of labourmarket regime is assumed to exist.
Regional labour markets in AMOS (1)Regional labour markets in AMOS (1)
One of the key features of the AMOS framework is that it incorporates five alternative labour market closures.
The specification of the wage equation in each is fairly standard:
1. Neo-classical/continuous market clearing. Here the wage adjusts so as to equate labour demand and labour supply.
2. Keynesian/national bargaining. The nominal wage is exogenously determined at the regional level. The motivation for this would generally be a national bargaining regime. The aggregate labour supply function is suspended up to full employment.
3. Real wage resistance. The real wage is fixed - i.e. the nominal wage is a mark up on the consumer price index.
Regional labour markets in AMOS (2)Regional labour markets in AMOS (2)
4. Exogenous labour supply. A fixed proportional relationship exists between employment and working population (this is often taken to be the closure for national CGE models).
5. Regional wage bargaining (also referred to as the bargained real wage, BRW, closure).
– The regional consumption wage is directly related to workers’ bargaining power and inversely related to the regional unemployment rate via a bargained real wage function.
– Note that this closure does imply that local wages are flexible in that they respond to the local excess demand for labour.
DynamicsDynamics
AMOS can be run as a recursive dynamic model
Capital stock updating• Each sector's capital stock is updated between periods via a simple capital stock
adjustment procedure• According to which investment equals depreciation plus some fraction of the gap between
the desired level of the capital stock and its actual level
Labour market/population• The regional economy is initially assumed to have zero net migration (as in the short-run
in the static model), and ultimately, net migration flows re-establish this (long-run) population equilibrium.
• Net in-migration in any one period is taken to be positively related to the regional real wage differential in the target region relative to the national economy and negatively related to the regional unemployment rate
• This migration model is based on that in Harris and Todaro (1970) (LNJ, 1991)
An illustrative practical application of AMOSAn illustrative practical application of AMOS-- lead into Workshop 1lead into Workshop 1
• A simple application – value-added from CGE relative to IO in modelling a demand disturbance
• Scotland: very small, very open economy• Shock: 20% step increase in total government expenditure, pattern unchanged
Model• Assume BRW closure• Migration possible and investment fully endogenous (i.e. no long-run supply
constraints)• Perfect competition and no macroeconomic constraints • i.e. BOP passive and, initially, no government budget constraint
Now go to the document titled ‘WVU course Aug 09_CGE_AMOS_workshop’
4. Workshop 1 4. Workshop 1 –– AMOS worksheet: AMOS worksheet: simulating fiscal policysimulating fiscal policy
Tuesday 11 August, 9-11am
Refer to worksheet in document ‘WVU course Aug 09_CGE_AMOS_workshop’
5. Review of AMOS fiscal policy simulations5. Review of AMOS fiscal policy simulations
Tuesday 11 August, 11am-12.30pm
An illustrative practical application of AMOSAn illustrative practical application of AMOS-- Fiscal policy simulationsFiscal policy simulations
• Start with a simple application – value-added from CGE relative to IO in modelling a demand disturbance
• Scotland: very small, very open economy• Shock: 20% step increase in total government expenditure, pattern unchanged
Model• Assume BRW closure (initially)• Migration possible and investment fully endogenous (i.e. no long-run supply constraints)• Perfect competition and no macroeconomic constraints • i.e. BOP passive and, initially, no government budget constraint• Then we introduce a binding constraint so that the income tax rate in Scotland has to rise
to fund the additional expenditure (i.e. income tax passive) – in the Workship 1 exercise you simulated an increase in income tax with government expenditure passive
Increase in government expenditure Increase in government expenditure -- scenariosscenarios
Take 2 basic cases:1. Leontief technology in all production and consumption functions
(except trade parameters – too inflexible: solution problem)2. Some substitution possible in all parameters (CES = 0.3 for local
and 2.0 for trade)
We also examine a third and fourth case3. As 2, but with a binding government budget constraint – income tax
rises to fund increased expenditure4. As 3, but assume national rather than regional wage bargaining in
UK
Cases 1 and 2 Cases 1 and 2 –– discussion of results (1)discussion of results (1)
• Both reach the same long run result – i.e. when all supply constraints are fully relaxed
• Increase in all quantities (sectoral outputs, employment and value added, total investment, consumption and GDP) except exports
• No change (relative to base) in prices – sectoral output prices, CPI, wages
But how did the economy adjust to this result?
Fig 1. Impact on sectoral output of a 20% increase in government expenditure after 5 years
-2
0
2
4
6
8
10
12
Cas
e 1
Man
ufac
turin
g
Cas
e 2
Man
ufac
turin
g
Cas
e 1
Non
-M
anu
Trad
ed
Cas
e 2
Non
-M
anu
Trad
ed
Cas
e 1
She
ltere
d
Cas
e 2
She
ltere
d
Sector/case
% c
hang
e fr
om b
ase
Cases 1 and 2 Cases 1 and 2 –– discussion of results (2)discussion of results (2)
• Initially, we observe crowding out• The sheltered (non-traded) sector, which receives the largest direct demand
effect, draws labour, capital and other inputs away from the other two sectors
• Capital and labour cannot adjust straight away (though they are elastic over time through investment and migration of labour)
• This is reflected through increases in prices – the economy only adjusts to long-run equilibrium when supply relaxes sufficiently to allow prices to adjust to their initial levels
Cases 1 and 2 Cases 1 and 2 –– discussion of results (3)discussion of results (3)
• With Leontief technology the negative effects of increase prices are smaller because producers cannot switch away from inputs that are relatively more expensive
• As a result, the economy adjusts to long-run equilibrium more quickly with Leontief technology
• but note that, here, even in the partial Leontief case, the economy only tends toward long-run equilibrium (zero change in prices) after a long time, still not quite making it after 30 years
Impact of a 20% increase in government expenditure on CPI and wages (Leontief case)
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50
Period (year)
% c
hang
e fr
om b
ase
Nominal before-tax wage Real T-H consumption wage Consumer price index
Impact of a 20% increase in government expenditure on wages and CPI (CES)
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50
Period (year)
% c
hang
e fr
om b
ase
Nominal before-tax wage Real T-H consumption wage Consumer price index
Cases 1 and 2 Cases 1 and 2 –– discussion of results (4)discussion of results (4)
• Full activity increase will not be realised until supply constraints are fully relaxed and there is no upward pressure on prices
• Note that during the period of adjustment, the rise in prices has negative competitiveness effects
• This causes export demand to fall – in the long run, when we see them at base year levels, this is due to the zero change in prices
This must be of interest to policymakers (whose timeframe is likely to be much less than 30-50 years)
Impact on total export demand from a 20% increase in government expenditure
-3.5
-3
-2.5
-2
-1.5
-1
-0.5
0
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50
Period (years)
% c
hang
e fr
om b
ase
% change in export demand (Leontief)% change in export demand (CES)
Cases 1 and 2 Cases 1 and 2 –– discussion of results (5)discussion of results (5)
• So, even with a pure demand shock, we are likely to get a supply-side response, except in conditions of huge excess supply
• Also, is the assumption of universal Leontief technology realistic – i.e. no price induced substitution in production and consumption?
• Elasticities of substitution may be difficult to measure – but is it better to assume zero?
• Not theory consistent!• But, leaving this aside, the assumption of passive supply is a very strong one• What if there is a direct supply-side disturbance?
• For example, so far we haven’t discussed how an increase in government expenditure would be financed….see Case 3
Case 3 Case 3 –– discussion of results (1)discussion of results (1)
• Case 3 assumes CES technology (but qualitatively results would be no different if we assumed Leontief)
• This time, we set the model up so that income tax is ‘passive’ – here it adjusts to ensure that the increase in government expenditure is matched by an equal increase in tax revenues (qualitatively the same as the worksheet simulation – only difference there is you have specified the tax rate and allowed expenditure to change endogenously)
• This will happen when a government budget constraint applies• Note that we do NOT get IO results in the long-run • i.e. even when all supply constraints are eventually relaxed, prices do not return to
their pre-shock levels• Why not? The supply side of the economy has changed• Workers seek to bargain to restore their previous real take home consumption
wage.• Therefore, a permanent increase in before tax wage (“tax-shifting” onto employers)
Impact of a 20% increase in goverment expenditure on real wages (balanced budget)
-4
-2
0
2
4
6
8
10
12
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50
Period (years)
% c
hang
e fr
om b
ase
Real B-Tx consumption wage Real T-H consumption wage
Case 3 Case 3 –– discussion of results (2)discussion of results (2)
• See full results for Case 3 and 4 below• For workers to maintain their original take-home real wage (purchasing
power), there is a lasting impact on the supply side of the economy • Note that while unemployment returns to its initial rate, the levels of
unemployment and employment fall• A balanced-budget “fiscal expansion” here actually leads to a big
contraction in activity!• Why? Adverse supply shock through the hike in pre-tax wages, but during
adjustment, fall in real wages induces outmigration. Note the fall in working population – Scotland’s real take home pay relative to the rest of the UK falls initially, so workers leave – there is net out-migration
• This outweighs beneficial demand effect associated with Keynesian BB multiplier
• However, note assumption of BRW closure
Case 4 Case 4 –– discussion of results (1)discussion of results (1)
• If instead assume national bargaining closure:• Essentially IO results (though combination of a reduction in C and increase
in T) – there is a fall in take-home wage, but all other prices return to base levels. In this case workers’ simply accept the cut in their real take home wage.
• no negative competitiveness effects, because no attempt to shift tax incidence
• net in-migration due to net increase in activity levels (gross out-migration due to increased taxes)
• fall in umemployment• rise in all aggregate activity except household consumption• Importance of labour market closure
Impact of a 20% increase in government expenditure on real w ages (national bargaing, balanced budget)
-15
-10
-5
0
5
10
15
20
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50
Period (years)
% c
hang
e fr
om b
ase
Real B-Tx consumption wage Real T-H consumption wage
Alternative fiscal simulations Alternative fiscal simulations –– Workshop 1 resultsWorkshop 1 results
• NB The rise in the income tax rate that would be required here is more than the + or – 3 pence in the pound that the Scottish Parliament can introduce
• It may be interesting to look at this problem from the other side – what change in government expenditure is possible within the powers of the Parliament – this is what you have simulated in Workshop 1
• There we noted that you could also take into account whether private households value government expenditure as a substitute for their own private expenditure – in contrast to what is assumed here
• Discussion of your simulation results?
6. Extending and developing the AMOS framework to model increases in energy and
labour productivity and introduction to Workshop 2
Tuesday 11 August, 1.30-3pm
Modelling increased energy efficiency in a CGE modelling framework: rebound, backfire and
disinvestment effectsSession 6 in Course delivered at the Regional Research Institute, University of West
Virginia, 10-12 August 2009 (Tuesday 11 August, 1.30-2.30pm)
Dr Karen TurnerESRC Climate Change Leadership Fellow
(ESRC ref: RES-066-27-0009), Department of Economics, University of Strathclyde
The research reported in this session has been funded under the ESRC First Grants Programme, Ref: RES-061-25-0010
Energy-Economy-Environment CGE modelling (1)
Policy concerns:• Energy supply issues (from 1970s); impacts of changes in economic conditions and/or
policy on environmental/sustainability indicators (renewed interest early 90s)• Sustainability and climate change – global and local concerns• UK – regional and national focus• Supply-side issues
• Modelling energy use in production (KLEM; energy-capital substitutability etc)• Energy and/or carbon taxation• Resource productivity, energy efficiency• Changes in technology
• Demand-side issues• Nature and structure of energy demand and use• Elasticity of energy demand
Energy-Economy-Environment CGE modelling (2)
Requirements for energy-economy-environment modelling:
• Multi-sectoral modelling• Different energy-use and pollution generation characteristics of different
production and consumption activities• Link physical energy use and pollution generation to economic activity
• System-wide• Interaction between different production and consumption activities
Energy-Economy-Environment CGE modelling (2)
• Bergman, L. (1988) Energy Policy Modelling: a Survey of General Equilibrium Approaches, Journal of Policy Modelling, 10, pp.377-399.
• Bergman, L. (2005) CGE Modelling of Environmental Policy and Resource Management, Chapter 24 in Mäler and Vincent (eds) Handbook of Environmental Economics, Volume 3: Economywide and International Environmental Issues, Elsevier, North Holland.
• Conrad, K. (1999) Computable General Equilibrium Models for Environmental Economics and Policy Analysis, in J.C.J.M. van den Bergh ed. Handbook of Environmental and Resource Economics, Edward Elgar Publishing Ltd, 1999.
• Greenaway, D. Leyborne, S., Reed, G., Whalley, J. (1993) ‘Applied General Equilibrium Modelling: Applications, Limitations and Future Developments’, HMSO, London.
AMOSENVI CGE modelling programme (1)
• Project commissioned by the States of Jersey from the Fraser of Allander Institute: ‘Programme to provide a regional database and IO and CGE models of Jersey’ (PIs: Peter McGregor and Kim Swales)
• Turner, K. (2002), Modelling the impact of policy and other disturbances on sustainability policy indicators in Jersey: an economic-environmental regional computable general equilibrium analysis, Ph.D. thesis, University of Strathclyde.
• ‘The importance of the regional/local dimension of sustainable development: an illustrative computable general equilibrium analysis of the Jersey economy’ by D. LEARMONTH, P.G. MCGREGOR, J.K. SWALES, K.R. TURNER and Y.P. YIN, Economic Modelling, Vol.24, pp. 15-41, 2007.
• ‘The additional precision provided by regional-specific data: the identification of fuel-use and pollution generation coefficients in the Jersey economy’ by K.R. TURNER, Regional Studies, Vol. 40, No.4, pp. 347-364, 2006.
AMOSENVI CGE modelling programme (2)
• ESRC-funded project, ‘Modelling the impact of “sustainability” policies in Scotland’, ESRC Grant No. R000223869 – PIs: P. G. McGregor, J. K. Swales and N.D. Hanley (University of Glasgow). 2002-2003
• ‘Incorporating sustainability indicators into a computable general equilibrium model of the Scottish economy’, by P.G. MCGREGOR, I. H. MCNICOLL, J.K. SWALES, K.R. TURNER and Y.P. YIN, Economic Systems Research, Vol. 17, No.2, pp. 103-140, 2005.
• ‘The impact of a stimulus to energy efficiency on the economy and the environment: a regional computable general equilibrium analysis’ by N.D. HANLEY, P.G. MCGREGOR, J.K.SWALES and K. TURNER, Renewable Energy, Vol. 31, pp. 161-171, 2006.
• ‘The impact of increased efficiency in the industrial use of energy: a computable general equilibrium analysis for the United Kingdom’, by G. ALLAN, N.D. HANLEY, P.G. MCGREGOR, J.K.SWALES and K. TURNER, Energy Economics, Vol. 29(4), pp. 779-798, 2007.
• ‘Do increases in energy efficiency improve environmental quality and sustainability?’ by N.D. HANLEY, P.G. MCGREGOR, J.K.SWALES and K. TURNER, Ecological Economics, 68, 692-709. 2009.
AMOSENVI CGE modelling programme (3)• Project under ESRC First Grants Initiative: ‘An empirical general equilibrium analysis of the factors that
govern the extent of energy rebound effects in the UK economy’. October 2007 – September 2010. [ESRC Reference: RES-061-25-0010]. PI: Karen Turner
• ‘Negative rebound and disinvestment effects in response to an improvement in energy efficiency in the UK Economy, by K. TURNER, Energy Economics, proofs in press doi:10:1016/j.eneco.2009.01.008. 2009. Results reported in this presentation are extracted from this paper.
• ‘Rebound and disinvestment effects in refined oil consumption and supply resulting from an increase in energy efficiency in the Scottish commercial transport sector’ by S. ANSON and K. TURNER, accepted to Energy Policy, April 2009.
• ‘Do productivity improvements move us along the Environmental Kuznets Curve?’ by K. TURNER, N.D. Hanley and J. De Fence, submitted to Environmental and Resource Economics, February 2009.
• ‘A computable general equilibrium analysis of the relative price sensitivity required to induce rebound effects in response to an improvement in energy efficiency in the UK economy’ by K. TURNER, submitted to Economic Systems Research, April 2009.
AMOSENVI – environmental component
• N-sector – identify sectors with distinct energy supply/use and/or pollution generation characteristics
• Initial version (Jersey project) – emphasis on pollution generation – Leontief ouput-pollution coefficients
– Captures changes in pollution due to scale and composition effects– But not due to input substitution and technology effects
Current version (ESRC projects):• KLEM production structure• Input- and output-CO2 coefficients
Figure1. Production structure of each sector i in the 25 sector/commodity AMOSENVI KLEMframework
GROSS OUTPUT
INTERMEDIATES VALUE-ADDED
ROW composite UK composite LABOUR CAPITAL
RUK composite LOCAL composite
ENERGY composite NON-ENERGY composite
Non-energy comm. j = 1……………………………….20
ELECTRICITY NON-ELECTRICITY composite composite
RENEWABLE NON-RENEWABLE(comm j=24) (comm j=25)
OIL NON-OIL (comm j=22) composite
COAL GAS (comm j=21) (comm j=23)
The rebound effect
• Jevons (1865) – “confusion of ideas” regarding productive use of fuel and diminished consumption – increase utility, impact on implicit prices
• Rebound and backfire effects – see e.g. Khazzoom 1980; Brookes 1990; Herring, 1999; Birol and Keppler, 2000; Saunders, 1992, 2000a,b; Schipper, 2000
• Initial UK Policy context – House of Lords (2005)
• Assessment of evidence by UKERC – Sorrell (2007)
• Direct, indirect and economy-wide rebound effects, production and consumption
General equilibrium analysis
• E.g. Semboja, 1994,; Dufournaud et al, 1994; Grepperud and Rasmussen, 2004; Glomsrød and Wei, 2005; Hanley et al, 2006, 2009; Allan et al, 2007
• ESRC funded project: ‘An empirical general equilibrium analysis of the factors that govern the extent of energy rebound effects in the UK economy’, Oct 2008-Sept 2010
– Main finding so far – importance of supply-side response to changes in energy prices
• Wei (2007) and Saunders (2008) – theoretical prediction of rebound effects that are bigger in long-run than in short run due to increased productive capacity
• Turner (2009), Anson and Turner (forthcoming Energy Policy, 2009) – negative rebound and disinvestment effects
– Wei (2009) – development of theoretical analysis emphasising importance of supply-side
Defining the rebound effect (1)
– If there is energy augmenting technical progress at a rate ρ, the relationship between the percentage change in physical energy use, and the energy use measured in efficiency units is given as:
– (1)
– Impact on price, measured in efficiency units:
– (2)
ε = Eρ +
ε Ep = p ρ−
Defining the rebound effect (2)
– With physical energy prices constant, we expect a fall in the price of energy in efficiency units to generate an increase in the demand for energy in efficiency units :
– (3)
– Change in energy demand in natural units can be found by substituting equations (2) and (3) into equation (1), giving :
– (4)
p= − εε η
E = ( 1)−η ρ
Defining the rebound effect (3)
– For an efficiency increase of , rebound, R, expressed in percentage terms, is defined as: :
– (5)
– (6)
– Defining boundaries of the efficiency change:
– (7)
1 100ER⎡ ⎤
= + ×⎢ ⎥⎣ ⎦ρ
100R = ×η
1 100TER⎡ ⎤
= + ×⎢ ⎥⎣ ⎦αρ
The energy efficiency improvement leads to an increase in the demand for energy in natural units that outweigh the reduction in demand from the efficiency improvement. Such a phenomenon is labelled as a ‘backfire effect’.
>100%> 1(elastic)
The reduction in energy demand from the efficiency improvement is entirely offset by increased demand for energy as prices fall.
100%1 (unitary elasticity)
Some of the energy efficiency improvement is reflected in a fall in the demand for natural energy units, but partly offset by increased (direct and derived) demand for energy as effective and/or actual energy prices fall.
0 – 100%0 to 1(relatively inelastic)
All of the energy efficiency improvement is reflected in a fall in the demand for natural energy units.
0%0(perfectly inelastic)
Implication for energy efficiency improvementRebound effectGeneral equilibrium price elasticity of demand for energy
General equilibrium responses to increased efficiency
1. Efficiency effect
2. Substitution effect
3. Output/competitiveness effect
4. Composition effect
5. Income effect
6. Disinvestment effect
The disinvestment effect• Previous rebound research – focus on demand response to changing energy prices
• Price falls – direct and derived demands for energy rise
• However, where actual energy prices affected (e.g. local energy supply)• Price falls – if quantity demanded does not rise sufficiently to offset decline in
revenue, profitability falls (more inelastic demand – larger drop in price)• Return on capital decreases in energy supply• Shedding of capital stock – disinvestment• Energy supply becomes more inelastic, energy prices rise• Constrains size of rebound effect
UKENVI
• 3 internal transactor groups (households, firms, government), plus Rest of World
• 25 commodities/sectors, including 5 energy sectors (coal, refined oil, gas, two electricity)
• Capital – labour – energy – materials (KLEM) production structure using multi-level CES production functions. Assumes cost minimisation.
• Calibrated to 2000 UK SAM
• Recursive dynamic – investment = depreciation plus proportion of difference actual and desired capital stock, where desired capital stock determined on cost minimisation criteria and reflects changing profitability at sectoral level
• Labour market – wage bargaining
Production structure of each sector in the 25 sector/commodity UKENVI framework
GAS(comm. j=23)
COAL(comm. j=21)
NON-OIL compositeOIL(comm. j=22)
NON-RENEWABLE (comm. j=25)
RENEWABLE (comm. j=24)
NON-ELECTRICITY composite
ELECTRICITY composite
Non-energy comm. j=1........20
NON-ENERGY compositeENERGY composite
CAPITALLABOURUK composite
INTERMEDIATES
GROSS OUTPUT
VALUE-ADDED
ROW composite
Energy efficiency shock
• Permanent, exogenous (and costless) 5% increase in energy augmenting technological progress
• Initially targeted at all 25 production sectors
• Rebound calculated as
• Distinguish electricity and non-electricity energy use, focus on local energy supply and total domestic use
1 100TERαρ
⎡ ⎤= + ×⎢ ⎥⎣ ⎦
30.954.7Non-electricity energy rebound effect (%)
-2.05-1.35Total consumption UK non-electricity energy
23.159.6Electricity rebound effect (%)
-2.76-1.45Total consumption UK electricity
0.000.00Total population (000's)
-2.61-2.44Unemployment rate (%)
0.210.20Total employment (000's):
-0.23-0.27Consumer price index
0.300.28Real T-H consumption wage
0.070.01Nominal before-tax wage
-0.21-0.23Imports
0.22-0.03Exports
0.140.06Investment
0.340.37Consumption
0.170.11GDP (income measure)
Long-run Short-run
(Percentage changes from base year equilibrum) - Base Case Scenario
in all production sectors in the UK economy
Table 2. The aggregate impact of a 5% increase in energy efficiency (locally supplied inputs)
Figure 4. Impact on capital rental rates in the UK energy supply sectors of a 5% increase in energy efficiency in all production sectors (% change from base)
-10
-8
-6
-4
-2
0
2
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
Period/year
Coal (Extraction)Oil processing and nuclear refiningGasElectricity - Renewable (hydro and wind)Electricity - Non-renewable (coal, nuke and gas)
Figure 5. Impact on UK energy supply sector capital stocks of a 5% increase in energy efficiency in all production sectors (% change from base)
-1.6
-1.4
-1.2
-1
-0.8
-0.6
-0.4
-0.2
01 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
Period/year
Coal (Extraction)Oil processing and nuclear refiningGasElectricity - Renewable (hydro and wind)Electricity - Non-renewable (coal, nuke and gas)
Figure 6 Percentage change in UK local energy supply prices in response to a 5% improvement in energy efficiency in all production sectors (applied to locally supplied energy) - (% change from base)
-25
-20
-15
-10
-5
0
0 3 6 9 12 15 18 21 24 27 30 33 36 39 42 45 48 51 54 57 60 63 66 69 72 75 78 81 84 87 90 93 96 99 102
105
108
111
114
117
120
123
126
129
132
135
138
141
144
147
150
Period/year
COAL (EXTRACTION)
OIL (REFINING & DISTR OIL AND NUCLEAR)
GAS
Electricity - Renew able (hydro and w ind)
Electricity - Non-renew able (coal, nuke and gas)
Disinvestment effects
• Wei (2007) and Saunders (2008) – rebound bigger in long run than in short run because positive supply shock leads to expansion in set of production possibilities
• Here, true for non-energy supply sectors
• Crucial for theoretical predictions – Wei (2007) assumes fixed/exogenous return on capital
• If profitability falls as local energy prices fall, return on capital in local energy supply falls, leading to contraction in capacity in these sectors
• However, if demand is sufficiently elastic, prices can fall without reducing profitability
• Use of UKENVI for analytical work to understand basic drivers of rebound – vary key assumptions one at a time; here focus on
– elasticities of substitution in production and trade elasticities (imports and exports)
Table 4. Results of sensitivity analysis of non-electricity energy rebound effects in the UK economy in response to a 5% exogenous improvement in energy efficiency in production (applied to use of locally supplied energy)
LONG RUN NON-ELECTRICITY REBOUND
Production 0.3 0.5 0.8 CD 1.1 1.3 1.5 1.8 2 3 4 5 0 (0.064) * * * -13.73 -12.26 -9.53 -6.96 -3.34 -1.05 9.44 18.79 27.32
0.1 * * -12.34 -9.14 -7.67 -4.89 -2.28 1.38 3.70 14.31 23.78 32.380.3 1.16 6.70 12.99 16.47 18.08 21.10 23.91 27.85 30.34 41.62 51.61 60.660.5 26.00 31.89 38.67 42.44 44.19 47.44 50.47 54.70 57.35 69.35 79.90 89.430.8 64.01 70.38 77.89 82.09 84.02 87.67 91.03 95.72 98.52 111.80 123.25 133.53
CD (0.999999) 91.29 96.15 102.26 105.86 107.57 110.81 113.89 118.26 121.05 133.83 145.24 155.631.1 102.91 109.76 117.97 122.61 124.76 128.76 132.49 137.66 140.89 155.27 167.71 178.821.3 129.37 136.52 145.21 150.14 152.42 156.70 160.66 166.17 169.61 184.86 198.00 209.711.5 * 163.70 172.87 178.08 180.50 185.05 189.27 195.12 198.76 214.95 228.83 241.161.8 * 205.30 215.16 220.83 223.46 228.42 233.02 239.41 243.40 261.03 276.10 289.45
2 * 233.60 243.92 249.89 252.68 257.91 262.78 269.55 273.77 292.41 308.32 322.40
Figure 7. Qualitative non-electricity energy rebound and disinvestment results for Scotland
TradeProduction 0 (0.064) 0.1 0.3 0.5 0.8 CD 1.1 1.3 1.5 1.8 2 3 4 5 0 (0.064)
0.10.30.50.8
CD (0.999999) A B C1.11.31.51.8
2
A. LR R > SR R Short run rebound dampened by weak competitiveness effects; long run rebound dampened by disinvestmentB. SR R > LR R Long run rebound dampened by disinvestmentC. LR R > SR R Long run rebound not constrained by disinvestment effects
Negative rebound effects
• Disinvestment a necessary but not sufficient condition for SR R > LR R
• Area A – in some cases efficiency improvement manifests as a negative supply shock
• Also, some cases of negative rebound effects
• Not ‘super-conservation’ (Saunders, 2008)
• Some downward pressure from disinvestment (LR), also some negative income effects
• But key here – negative multiplier effects in energy supply sectors
• Apparent when repeat simulation targeting only non-energy supply sectors
Figure 8. Percentage change in long run non-electricity energy consumption in UK production sectors in response to a 5% increase in energy efficiency in non-energy supply sectors (trade
parameters 1, production parameters 0.064) - % change from base
-6.00
-5.00
-4.00
-3.00
-2.00
-1.00
0.00
1.00
Agricu
lture,
fores
try an
d fish
ing
Other m
ining
and q
uarry
ing, in
cludin
g oil a
nd ga
s extr
actio
n
Mfr - Foo
d and
drink
Mfr - Tex
tiles
Mfr - Pulp
, pap
er an
d artic
les of
pape
r and
board
Mfr - G
lass a
nd gl
ass p
roduc
ts, ce
ramic
good
s and
clay
prod
ucts
Mfr - C
emen
t, lim
e plas
ter an
d artic
les in
conc
rete,
plaste
r and
ceme..
Mfr - Iro
n, ste
el firs
t proc
essin
g, an
d cas
ting
Mfr - O
ther m
etal p
roduc
ts
Mfr - O
ther m
achin
ery
Mfr - Elec
trical
and e
lectro
nics
Mfr - O
ther m
anufa
cturin
g
Wate
rCon
struc
tion
Distrib
ution
and t
ransp
ort
Commun
icatio
ns, fi
nanc
e and
busin
ess
Resea
rch an
d dev
elopm
ent
Public
admin
and e
duca
tion
Health
and s
ocial
work
Other s
ervice
sCoa
l (Extr
actio
n)
Oil proc
essin
g and
nucle
ar ref
ining
Gas Electric
ity - R
enew
able
(hydro
and w
ind)
Electric
ity - N
on-re
newab
le (co
al, nu
ke an
d gas
)
House
holds
Current/future research
• Relative importance of elasticities of substitution in production in determining rebound
• Current research - econometric estimations of KLEM production function (e.g. Kemfert 1998)
• Different production structures – energy as an intermediate or primary input?
• Sectoral level analysis – Anson & Turner (2009) focus on Scottish commercial transport sector and impacts on Scottish refined oil supply; IAEE 2009 paper with simulations of increased energy efficiency in energy use in non-renewable electricity supply
• Increased efficiency (total factor productivity) in energy production; increased efficiency in other factors of production: (i) Turner, Hanley and De Fence EKC paper also available in course material presented below; (ii) Turner and Hanley interregional analysis of increased labour productivity to be presented tomorrow in Session 8.
What happens if we target labour efficiency instead?• CO2 emissions are largely driven by energy use – a key policy concern in Scotland and the UK
• From above, improvements to energy efficiency can lead to falling CO2 emissions as GDP rises (moving down the EKC), but our research shows that this depends on general equilibrium price elasticities, structure of economy, which sectors are targeted, and migration.
• We have also used the AMOSENVI models to simulate the impacts of increased labour efficiency – idea that the substitution effect discussed above will lead to a shift in favour of labour away from energy (and other inputs)
• We find that, while increased labour efficiency won’t reduce aggregate CO2 emissions as GDP rises (because of the competitiveness and income effects, which boost all input use), it will typically result in an improvement in the CO2/GDP ratio, with CO2 growing more slowly than GDP
• However, again, quantitative results depend on whether energy sectors are targeted with this positive supply shock and the specification of key elasticities governing the substitution effect
Summary results for the CO2 intensity of GDP for a 5% increase in energy efficiency
Scotland UKLabour market: Real wage bargaining Real wage bargaining
Short run Long run Short run Long runMigration on Migration off Migration on Migration off
Key production parameters:Inelastic (0.8):All 25 sectors (EA1)
CO2/GDP 0.94% 1.78% 1.81% 0.29% -0.19% -0.25%CO2 level Rise Rise Rise Rise Rise Fall
20 non-energy supply sectors (EA3)CO2/GDP -0.22% -0.28% -0.27% -0.33% -0.33% -0.36%CO2 level Fall Fall Fall Fall Rise Fall
Elastic (1.1):All 25 sectors (EA2)
CO2/GDP 2.12% 2.77% 2.82% 1.70% 1.33% 1.29%CO2 level Rise Rise Rise Rise Rise Rise
20 non-energy supply sectors (EA4)CO2/GDP 0.22% 0.47% 0.25% 0.37% 0.40% 0.38%CO2 level Rise Rise Rise Rise Rise Rise
Summary results for the CO2 intensity of GDP for a 5% increase in labour efficiency
Scotland UKReal wage bargaining Real wage bargaining
Labour market: Short run Long run Short run Long runMigration on Migration off Migration on Migration off
Key production parameters:Inelastic (0.8):All 25 sectors
CO2/GDP -1.31% -1.02% -0.88% -0.17% 0.58% -0.04%CO2 level Rise Rise Rise Rise Rise Rise
20 non-energy supply sectorsCO2/GDP -1.71% -1.70% -1.55% -0.35% 0.47% -0.13%CO2 level Rise Rise Rise Rise Rise Rise
Elastic (1.1):All 25 sectors
CO2/GDP -1.46% -1.52% -1.46% -0.26% 0.10% -0.28%CO2 level Rise Rise Rise Rise Rise Rise
20 non-energy supply sectorsCO2/GDP -1.94% -2.18% -1.85% -0.42% 0.02% -0.36%CO2 level Rise Rise Rise Rise Rise Rise
Impacts of a 5% increase in labour productivity from the 1999 Scotenvi/AMOSENVI model
Scotland
Migration
Short-run Long-run (150) Long-run (150)
GDP (income measure) 3.46 5.70 8.27 Consumption 1.15 2.38 4.31 Investment 5.98 4.57 6.74 Exports 2.70 4.86 6.78 Imports 0.27 0.14 0.64
Nominal before-tax wage -0.59 -0.55 -2.57 Real T-H consumption wage 0.47 1.33 0.00 Consumer price index -1.05 -1.85 -2.57
Total employment (000's): 0.48 1.32 4.06 Unemployment rate (%) -4.05 -11.05 0.00 Total population (000's) 0.00 0.00 4.06
Scotland
No Migration
Introduction to Workshop 2Introduction to Workshop 2
• Labour efficiency results above extracted from ‘Do productivity improvements move us along the Environmental Kuznets Curve?’ by K. TURNER, N.D. Hanley and J. De Fence, Strathclyde Discussion Papers in Economics, No. 09-08. 2009.
• In Workshop 2, you should try introducing a similar shock to the 3 sector 1998 AMOS model– Configure model as in the main simulations in Workshop 1, except this time under the Technical
Progress menu, choose Harrod neutral (labour augmenting technical progress)– For the labour market, choose the regional bargaining closure– Set up a 50 period base model and run your simulations for 50 periods, as well as for a long-run
solution– Introduce a 5% increase in technological progress to all 3 sectors in the shock menu– Once you have run your simulations once with migration, run them again, but with migration turned
off.
7. Workshop 2 7. Workshop 2 –– Simulating increased labour Simulating increased labour efficiencyefficiency
Wednesday 12 August, 9am-12pm
Introduction to Workshop 2Introduction to Workshop 2
• Yesterday afternoon, we discussed some labour efficiency results above extracted from ‘Do productivity improvements move us along the Environmental Kuznets Curve?’ by K. TURNER, N.D. Hanley and J. De Fence, Strathclyde Discussion Papers in Economics, No. 09-08. 2009.
• Today, in Workshop 2, you should try introducing a similar shock to the 3 sector 1998 AMOS model– Configure model as in the main simulations in Workshop 1, except this time under the Technical
Progress menu, choose Harrod neutral (labour augmenting technical progress)– For the labour market, choose the regional bargaining closure– Set up a 50 period base model and run your simulations for 50 periods, as well as for a long-run
solution– Introduce a 5% increase in technological progress to all 3 sectors in the shock menu– Once you have run your simulations once with migration, run them again, but with migration turned
off.– In this session, we will discuss your results, and any extensions/sensitivity analysis that you may
wish to perform, individually and as a group towards the end of the morning.
8. Extending and developing the AMOS 8. Extending and developing the AMOS framework to model the interregional framework to model the interregional
impacts of increases in labour productivityimpacts of increases in labour productivity
Wednesday 12 August, 12-1pm
How do improvements in labour productivity in the Scottish economy affect the UK position on the Environmental Kuznets
Curve?
Session 8 in Course delivered at the Regional Research Institute, University of West Virginia, 10-12 August 2009 (Wednesday 12 August, 12-1pm)
Dr Karen TurnerESRC Climate Change Leadership Fellow, Department of Economics, University of Strathclyde
(co-authored with Nick Hanley, Department of Economics, University of Stirling)
The research reported in this session has been funded under the ESRC Climate Change Leadership Fellows Programme, Ref:RES-066-27-0009
ESRC Climate Change Leadership Fellowship
• Investigating the Pollution Content of Trade Flows and the Importance of ‘Environmental Trade Balances’ in Addressing the Problem of Climate Change
• 1 October 2008-31 December 2008 (27 months)
• Fellow: Karen Turner; Postdoctoral RF: Soo Jung Ha
• Developing from IO accounting to CGE modelling analyses of pollution trade balance issues – interregional analysis for the UK and US Midwest; possible international analysis using GTAP database
•
The Environmental Kuznets Curve• Empirical observation that as economic growth occurs, pollution (per capita?) first rises but
then falls, once the economy moves past a “turning point”
• Inverse U-shape
• Much disputed empirically, having generated a very large literature
• Panel, time series and cross section studies
• Seems to exist for some pollutants / environmental impacts, but not for others; seems to exist in some groupings of countries and not others; turning point also highly variable.
• See recent overview by Norman and Deacon (Land Econ., 2006)
Policy implication if true:
• We can grow without worrying about continually-increasing pollution
• Less of a trade-off between economic growth and environmental sustainability
• Countries can “grow their way” out of environmental problems.
Why should it happen?
Three main theoretical stories:
(1): structural change in an economy over time (Jaffe, 2003)(2): increasing real incomes translate into higher willingness to pay for environmental quality,
which translates into more voter pressure for stricter environmental legislation (Hokby and Soderquist, 2003)
(3): technological improvements, somehow correlated with economic growth, reduce the burden of each $ worth of economic activity on the environment (more fuel-efficient cars, more energy-efficient production systems..) (Bretschger, 2005).
WE FOCUS ON (3)• Johansson and Kristrom (2007) find technological progress to be key driver of the EKC in a
time series analysis for SO2 in Sweden.
EKC and CGE – previous research
• Project under ESRC First Grants Initiative: ‘An empirical general equilibrium analysis of the factors that govern the extent of energy rebound effects in the UK economy’. October 2007 –September 2010. [ESRC Reference: RES-061-25-0010].
• ‘Do productivity improvements move us along the Environmental Kuznets Curve?’ by K. TURNER, N.D. Hanley and J. De Fence, submitted to Environmental and Resource Economics, February 2009, also Strathclyde Discussion Papers in Economics, 09-08.
– Single region/nation analysis for Scotland and UK. – Compare impacts of (exogenous and costless) increases in energy and labour
augmenting technological progress
•
General equilibrium responses to increased efficiency1. Efficiency effect – reduce demand for input targeted with efficiency improvement (here,
labour)
2. Substitution effect – shift in favour of targeted input, labour, over others
3. Output/competitiveness effect – increase labour and other input demand (direct and derived)
4. Composition effect – shift in favour of more labour intensive activities
5. Income effect – increase labour and other input demand
6. ‘Disinvestment effect’ (energy efficiency – Turner, 2009); in case of labour, reduction in labour supply through out-migration
Conclusions of single region analysis
• Improvements to energy efficiency can lead to falling CO2 emissions as GDP rises (moving down the EKC), but this depends on general equilibrium price elasticities, structure of economy, which sectors are targeted, and migration.
• Improvements in labour efficiency never reduce aggregate CO2 emissions as GDP rises, but typically result in an improvement in the CO2/GDP ratio, as CO2 grows more slowly than GDP (moving along the EKC)
• vital to know the values of key parameters if we are to predict the effects of boosting factor productivity as part of climate change policy.
• but also issue of pollution leakage effects (Arrow et al, 1995; Stern, 1998; Suri and Chapman, 1997; Cole, 2004)
• Consider impacts of increased labour productivity in Scotland on UK position on EKC
AMOSRUK
• A simple 3-sector, 2 region CGE model of the Scottish and rest of UK (RUK) economies
• Allows for substitution between factor inputs
• All intermediate and final demands determined endogenously, responding to relative price changes induced by demand and/or supply disturbances
• Investment is endogenous and sectoral capital stocks are updated between periods
• Regional populations can also be endogenised – interregional migration in response to changes in relative real wage and unemployment rates
• Period-by-period (year-by-year) results – track adjustment process
• Environmental component – CO2 emissions linked to intermediate input use (including energy)
Simulation• Example of potential policy – Scotland aims to close labour productivity gap with rest
of UK in a period of 10 years • Labour productivity – GDP per employee (FTE)• Taking base year AMOSUK data (issues – 1999, quality of UK table), Scottish
GDP per employee £33,137; RUK £34,755; UK £34,618• Take default AMOSUK model parameterisation (incl. 0.3 key elasticities of
substitution – value added, gross output)• Estimate 5.8% increase in labour augmenting technological progress
(exogenous, costless) required to close gap in 10years (after introduction)• Actual policy analysis – quality of data for base year SAM and model
configuration
• Here, subject to sensitivity analysis regarding key parameter values (substitutabililtyof labour for other inputs) and model closure (labour market scenarios)
Figure 1. Impact of a 5.8% increase in labour augmenting technological progress on regional and national GDP per employee (default configuraiton: key paras 0.3; regional wage bargaining, interregional migration)
33,000
33,200
33,400
33,600
33,800
34,000
34,200
34,400
34,600
34,800
35,000
0 1 2 3 4 5 6 7 8 9 10
Scottish GDP (£) per employeeRUK GDP (£) per employee
UK GDP (£) per employee
Table 1. Impact of a 5.8% increase in labour efficiency in Scotland on regional and national GDP per employee (£)
Scotland RUK UKBase 33,127 34,755 34,618
10 years - 0.3 34,755 34,756 34,75610 years - 0.8 34,445 34,769 34,74110 years - 1.1 34,598 34,749 34,736
Table 2. Percentage change in regional and national GDP, employment and population over time from a 5.8% increase in labour efficiency in Scottish production (bargaining, interregional migration on)
KEY PARAS 0.3 1 2 3 4 5 10 15 20 30 50 75ScotlandGDP 2.774 3.266 3.625 3.927 4.198 5.254 5.974 6.474 7.069 7.511 7.633Employment -1.315 -0.960 -0.762 -0.597 -0.437 0.325 0.927 1.363 1.890 2.286 2.396Population 0.000 -0.440 -0.624 -0.676 -0.652 -0.096 0.531 1.022 1.637 2.108 2.240RUKGDP 0.013 0.026 0.039 0.048 0.054 0.058 0.039 0.014 -0.029 -0.069 -0.081Employment 0.019 0.038 0.053 0.063 0.068 0.057 0.026 -0.003 -0.048 -0.086 -0.098Population 0.000 0.041 0.059 0.064 0.061 0.009 -0.050 -0.096 -0.154 -0.198 -0.211UKGDP 0.235 0.287 0.327 0.360 0.388 0.476 0.516 0.534 0.543 0.541 0.540Employment -0.093 -0.046 -0.015 0.007 0.025 0.079 0.102 0.111 0.115 0.113 0.112Population 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000KEY PARAS 0.8 1 2 3 4 5 10 15 20 30 50 75ScotlandGDP 4.480 4.870 5.274 5.635 5.951 7.050 7.657 8.006 8.332 8.489 8.510Employment 1.161 1.304 1.558 1.812 2.047 2.955 3.512 3.848 4.169 4.325 4.346Population 0.000 0.414 0.726 1.013 1.283 2.373 3.078 3.513 3.934 4.139 4.167RUKGDP 0.027 0.012 0.000 -0.012 -0.024 -0.083 -0.129 -0.161 -0.194 -0.211 -0.213Employment 0.040 0.011 -0.010 -0.029 -0.047 -0.123 -0.176 -0.209 -0.242 -0.259 -0.261Population 0.000 -0.039 -0.068 -0.095 -0.121 -0.223 -0.290 -0.331 -0.370 -0.390 -0.392UKGDP 0.385 0.403 0.424 0.443 0.457 0.491 0.497 0.496 0.492 0.490 0.489Employment 0.134 0.119 0.122 0.126 0.129 0.136 0.135 0.132 0.129 0.127 0.126Population 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000KEY PARAS 1.1 1 2 3 4 5 10 15 20 30 50 75ScotlandGDP 4.946 5.267 5.679 6.054 6.383 7.526 8.156 8.516 8.851 9.008 9.029Employment 1.822 1.922 2.240 2.557 2.847 3.929 4.574 4.956 5.316 5.487 5.509Population 0.000 0.414 0.726 1.013 1.283 2.373 3.078 3.513 3.934 4.139 4.167RUKGDP 0.030 0.000 -0.022 -0.041 -0.060 -0.140 -0.197 -0.232 -0.267 -0.284 -0.287Employment 0.043 -0.009 -0.042 -0.071 -0.097 -0.199 -0.263 -0.302 -0.339 -0.357 -0.359Population 0.000 -0.039 -0.068 -0.095 -0.121 -0.223 -0.290 -0.331 -0.370 -0.390 -0.392UKGDP 0.425 0.424 0.437 0.449 0.459 0.477 0.476 0.472 0.466 0.464 0.463Employment 0.193 0.154 0.150 0.150 0.151 0.148 0.144 0.140 0.136 0.134 0.134Population 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000
Figure 2 Regional labour market impacts (% change) of a 5.8% increase in labour-augmenting technological progress in all Scottish production sectors
Real wages - bargaining, interregional migration on
-1
-0.5
0
0.5
1
1.5
2
0 1 2 3 4 5 6 7 8 9 10Scot 0.3Scot 0.8Scot 1.1RUK 0.3RUK 0.8RUK 1.1
Labour supply - bargaining, interregional migration on
-1
-0.5
0
0.5
1
1.5
2
2.5
3
3.5
0 1 2 3 4 5 6 7 8 9 10
Scot 0.3Scot 0.8Scot 1.1RUK 0.3RUK 0.8RUK 1.1
Unemployment rate - bargaining, interregional migration on
-25
-20
-15
-10
-5
0
5
10
15
20
0 1 2 3 4 5 6 7 8 9 10
Scot 0.3 Scot 0.8
Scot 1.1 RUK 0.3
RUK 0.8 RUK 1.1
Figure 3 Regional EKC impacts (% change) of a 5.8% increase in labour-augmenting technological progress in all Scottish production sectors
Scot EKC - key paras 0.8 (bargaining, interregional migration on)
-4
-2
0
2
4
6
8
0 1 2 3 4 5 6 7 8 9 10
GDPCO2CO2/GDPCO2/GDP per capitaPopulation
Scot EKC - Key paras 1.1 (bargaining, interregional migration on)
-4
-2
0
2
4
6
8
0 1 2 3 4 5 6 7 8 9 10
GDPCO2
CO2/GDPCO2/GDP per capita
Population
RUK EKC - key paras 0.8 (bargaining, interregional migration on)
-0.3
-0.2
-0.1
0
0.1
0.2
0.3
0 1 2 3 4 5 6 7 8 9 10
GDP
CO2
CO2/GDP
CO2/GDP per capita
Population
RUK EKC Key paras 1.1 (bargaining, interregional migration on)
-0.3
-0.2
-0.1
0
0.1
0.2
0.3
0 1 2 3 4 5 6 7 8 9 10
GDP
CO2
CO2/GDP
CO2/GDP per capita
Population
Scot EKC - key paras 0.3 (bargaining, interregional migration on)
-4
-2
0
2
4
6
8
0 1 2 3 4 5 6 7 8 9 10
GDPCO2CO2/GDPCO2/GDP per capitaPopulation
RUK EKC - key paras 0.3 (bargaining, interregional migration on)
-0.3
-0.2
-0.1
0
0.1
0.2
0.3
0 1 2 3 4 5 6 7 8 9 10
GDP
CO2
CO2/GDP
CO2/GDP per capita
Population
Figure 4 National EKC impacts (% change) of a 5.8% increase in labour-augmenting technological progress in all Scottish production sectorsUK EKC - Key parameters 0.8 (bargaining, interregional migration on)
-0.1
0
0.1
0.2
0.3
0.4
0.5
0.6
0 1 2 3 4 5 6 7 8 9 10
GDP
CO2
CO2/GDP
CO2/GDP per capita
Population
UK EKC - Key parameters 1.1 (bargaining, interregional migration on)
-0.1
0
0.1
0.2
0.3
0.4
0.5
0.6
0 1 2 3 4 5 6 7 8 9 10
GDP
CO2
CO2/GDP
CO2/GDP per capita
Population
UK EKC - Key parameters 0.3 (bargaining, interregional migration on)
-0.1
0
0.1
0.2
0.3
0.4
0.5
0.6
0 1 2 3 4 5 6 7 8 9 10
GDP
CO2
CO2/GDP
CO2/GDP per capita
Population
Figure 5. Labour market scenarios
Scotland RUK
Quasi IOFixed at the
regional level Fixed real wage Fixed real wage
Regional Bargaining
Fixed at the regional level Bargaining Bargaining
Flow Migration
Fixed at the national level Bargaining Bargaining
Regional Wage Setting
Population
Table 3. Regional impacts (% change) of a 5.8% increase in labour efficiency in Scotland on regional and national EKC indicators (a)
Table 3. Regional impacts (% change) of a 5.8% increase in labour efficiency in Scotland on regional and national EKC indicators (b)
Table 4. National impacts (% change) of a 5.8% increase in labour efficiency in Scotland on regional and national EKC indicators
UKLabour market scenario Quasi IO Bargaining Migration
Key parameters (el. sub VA and output) 10 years 75 years 10 years 75 years 10 years 75 years
0.3 GDP 0.541 0.983 0.473 0.561 0.476 0.540CO2 0.596 1.135 0.532 0.664 0.533 0.654CO2/GDP 0.054 0.150 0.058 0.102 0.057 0.114CO2/GDP per capita 0.054 0.150 0.058 0.102 0.057 0.114
0.8 GDP 0.888 1.021 0.505 0.524 0.491 0.489CO2 0.935 1.081 0.541 0.565 0.537 0.549CO2/GDP 0.046 0.059 0.036 0.041 0.045 0.060CO2/GDP per capita 0.046 0.059 0.036 0.041 0.045 0.060
1.1 GDP 0.973 1.038 0.494 0.504 0.477 0.463CO2 0.974 1.041 0.503 0.514 0.497 0.495CO2/GDP 0.001 0.003 0.008 0.009 0.021 0.032CO2/GDP per capita 0.001 0.003 0.008 0.009 0.021 0.032
Table 5. Difference in Scot/RUK productivity gap 10 years after shock(+ve overshoot; -ve undershoot)
Base case -4.68% (Scot gap relative to RUK)
Labour market scenarioQuasi IO Bargaining Migration
Key parameters (el. sub VA and output) 10 years 10 years 10 years
0.3 -0.10% 0.01% 0.00%0.8 -1.07% -0.52% -0.93%1.1 -1.66% -0.81% -1.44%
Conclusions
• As in our single region analysis (Turner, Hanley and De Fence, 2009), the results presented here demonstrate that what we assume about how labour markets function and the values associated with key elasticities of substitution are crucial in determining the economic and environmental effects (and their relative strength) of increases in labour augmenting technological progress
• What the analysis presented here adds is consideration of the importance of interregional spillover effects of increased technological progress in one region on others, particularly where this involves reallocation of the factor of production (labour) targeted with the efficiency improvement.
• However, even where there is no interregional migration of labour, we observe ‘pollution leakage’effects in that a positive supply shock in one region will lead to an indirect positive demand shock in other regions and increased trade flows will engender increased pollution generation in all regions
Current research
• Ongoing project under the ESRC Climate Change Leadership Fellows programme
• Update (2004) and improvement UK interregional IO data – Scottish Government, WERU, SEI
• UK interregional IO and CGE applications, expanding the number of sectors and regions modelled – ESRC/TSG collaborative studentship, WERU, BRASS, RESOLVE, University of Hertfordshire
• US interregional IO and CGE applications – 5 Midwest regions and RUS – REAL (Hewings); RRI (Jackson and Jensen)
• Econometric estimation key functions and parameter values (UK level)
• Future research - introduction KLEM production functions and consideration of increases in energy efficiency - crossover ESRC 1st Grant project on modelling ‘rebound effects’
9. Using the AMOSUK interregional framework to demonstrate the value added from using CGE analysis to
model the impacts of a simple demand shock on
economic and environmental indicators
Wednesday 12 August, 2-3pm
The added value from adopting a CGE approach to analyse The added value from adopting a CGE approach to analyse changes in environmental trade balances changes in environmental trade balances
Session 9 in Course delivered at the Regional Research Institute, University of West Virginia, 10-12 August 2009 (Wednesday 12 August, 2-3pm)
Dr Karen TurnerESRC Climate Change Leadership Fellow, Department of Economics, University of Strathclyde
(co-authored with Michelle Gilmartin, Kim Swales and Peter McGregor, EPSRC Supergen Marine Consortium)
The research reported in this session has been funded under the ESRC Climate Change Leadership Fellows Programme, Ref:RES-066-27-0009
ESRC Climate Change Leadership FellowshipESRC Climate Change Leadership Fellowship
• Investigating the Pollution Content of Trade Flows and the Importance of ‘Environmental Trade Balances’ in Addressing the Problem of Climate Change’
• 1 October 2008-31 December 2008 (27 months)
• Fellow: Karen Turner; Postdoctoral RF: Soo Jung Ha
• Developing from IO accounting to CGE modelling analyses of pollution trade balance issues – interregional analysis for the UK and US Midwest; possible international analysis using GTAP database
•
IO accounting: pollution attribution analysisIO accounting: pollution attribution analysis
• Production and consumption accounting principles – Munksgaard and Pedersen (2001)
• Ecological footprint literature – Wiedmann et al (2007) review – consumption accounting principle
• Regional (sub-national) analysis less common than international
• UK – devolution of responsibility for making and meeting sustainability objectives, but contribution to national Kyoto target
Interregional environmental IO analysisInterregional environmental IO analysis
• Input-output analysis of the CO2 trade balance between Scotland and the rest of the UK – McGregor et al 2008, apply Turner et al 2007
• [1]
• [2] Scot emissions in UK – production a/c principle:
• [3] Scot emissions in UK - consumption a/c principle:
•
y y11 12y y21 22
p pp p⎛ ⎞ ⎛ ⎞⎛ ⎞ ⎛ ⎞⎜ ⎟ ⎜ ⎟⎜ ⎟ ⎜ ⎟
⎝ ⎠⎝ ⎠ ⎝ ⎠⎝ ⎠
-1x11 121 11 12
x21 222 21 22
y ye 0 I-A -A=
y y0 e -A I-Ay y
1 11 12p = p +p
y y y1 11 21p =p +p
Table 1. The CO2 Trade Balance Between Scotland and RUK (tonnes, millions) - Type II Input-Output
Pollution supported by: Total regional
Scottish Govt Scottish Capital Scot-ROW RUK Govt RUK Capital RUK-ROW emissions of CO2Pollution generated in:Scotland 11.3 4.3 14.6 5.7 5.1 8.0 48.9RUK 8.1 6.3 10.8 144.5 117.7 228.0 515.4Total (UK) emissions supported by 19.3 10.6 25.4 150.2 122.7 236.1 564.3
Environmental trade balance:Scot pollution supported by RUK final demand 18.8RUK pollution supported by Scot final demand 25.2Scotland's CO2 trade balance -6.4
Illustrative demand disturbanceIllustrative demand disturbance
• Simulate a 10% increase in ROW demand for RUK Primary, Manufacturing and Construction sector
• Change in relevant vector within matrix Y on slide 3
• Universal Leontief technology; no supply constraints or prices
• Impact on CO2 trade balance limited to effects of increased ROW export demand for RUK production; no adjustment process
Table 3. Percentage change in key variables in response to a 10% increase in ROW export demand to the RUK Primary, Manufacturing and Construction sector
Output Value-added Employment Direct CO2Base (£million) % change Base (£million) % change Base (FTE, thousands) % change Base (tonnes, millions)
Scotland:PRIMARY, MFR and CONSTRUCTION 52471 0.99% 17134 0.99% 483 0.99% 12.4ELEC, GAS & WATER SUPPLY 5047 1.52% 1508 1.52% 14 1.52% 16.3SERVICES 83723 0.81% 43982 0.81% 1334 0.81% 9.6HOUSEHOLDS 40415 0.87% 10.7Total Scotland 62624 0.87% 1832 0.86% 48.9
RUK:PRIMARY, MFR and CONSTRUCTION 506584 4.46% 198046 4.46% 5581 4.46% 145.4ELEC, GAS & WATER SUPPLY 42067 2.91% 12896 2.91% 142 2.91% 128.9SERVICES 1031837 1.90% 504567 1.90% 16754 1.90% 109.0HOUSEHOLDS 453771.00 2.63% 132.3Total RUK 715508 2.63% 22477 2.54% 515.5
Total 2215914 2.60% 778132 2.49% 24309 2.41% 564.4
Table 4. Post-shock CO2 Trade Balance Between Scotland and RUK (tonnes, millions) - Type II Input-Output
Pollution supported by: Total regionalScottish Govt Scottish Capital Scot-ROW RUK Govt RUK Capital RUK-ROW emissions of CO2
Pollution generated in:Scotland 11.3 4.3 14.6 5.7 5.1 8.6 49.4RUK 8.1 6.3 10.8 144.5 117.7 243.8 531.2Total (UK) emissions supported by 19.3 10.6 25.4 150.2 122.7 252.4 580.6
Environmental trade balance:Scot pollution supported by RUK final demand 19.3RUK pollution supported by Scot final demand 25.2Scotland's CO2 trade balance -5.9
Table 5. Post-shock CO2 Trade Balance Between Scotland and RUK (% change from base) - Type II Input-Output
Pollution supported by: Total regionalScottish Govt Scottish Capital Scot-ROW RUK Govt RUK Capital RUK-ROW emissions of CO2
Pollution generated in:Scotland 0.00% 0.00% 0.00% 0.00% 0.00% 6.72% 1.10%RUK 0.00% 0.00% 0.00% 0.00% 0.00% 6.92% 3.06%Total (UK) emissions supported by 0.00% 0.00% 0.00% 0.00% 0.00% 6.91% 2.89%
Environmental trade balance:Scot pollution supported by RUK final demand 2.88%RUK pollution supported by Scot final demand 0.00%Scotland's CO2 trade balance -8.40%
CGE frameworkCGE framework
• Allows for a more comprehensive and theory consistent treatment of the issue:
• Overcomes technical limitations of passive supply side and assumption of universal Leontief technology
• Flexible structure allows consideration of model closures corresponding to different time periods and supply-side behaviour
AMOSRUKAMOSRUK
• For illustrative purposes, a simple 3-sector, 2 region CGE model of the Scottish and RUK economies
• Allows for substitution between factor inputs
• All intermediate and final demands determined endogenously, responding to relative price changes induced by demand disturbance
• Investment is endogenous and sectoral capital stocks are updated between periods
• Regional populations can also be endogenised
• Provides period-by-period (year-by-year) results – track adjustment process
Labour Market ScenariosLabour Market Scenarios
Scotland RUK
Quasi IO Fixed at the regional level Fixed real wage Fixed real wage
Regional Bargaining
Fixed at the regional level Bargaining Bargaining
Flow Migration
Fixed at the national level only Bargaining Bargaining
Regional Wage Setting
Population
Figure 1. Im pact on ROW export dem and for outputs of RUK production sectors in response to a 10% increase in ROW export dem and to the Prim ary, M anufacturing and Construction sector
(% changes from base year equilibrium )
-4
-2
0
2
4
6
8
10
12
0 3 6 9 12 15 18 21 24 27 30 33 36 39 42 45 48 51 54 57 60 63 66 69 72 75
Period/year
PRIMARY, MFR and CO NSTRUCTIO NELEC, G AS and W ATER SUPPLYSERVICES
Figure 2. Im pact on R O W export dem and for outputs o f S cottish production sectors in response to a 10% increase in R O W export dem and to the Prim ary , M anufacturing and C onstruction
sector (% changes from base y ear equ ilib rium )
-3
-2 .5
-2
-1 .5
-1
-0 .5
0
0 3 6 9 12 15 18 21 24 27 30 33 36 39 42 45 48 51 54 57 60 63 66 69 72 75
Pe riod/y e a r
P R IM A R Y, M FR an d C O N S TR U C TIO NE L E C , G A S an d W A TE R S U P P L YS E R V IC E S
F ig u re 3 . Im p act o n R U K G D P fro m a 10% in crease in R O W exp o rt d em an d to the R U K Prim ary , M an ufacturin g and C o nstru ctio n secto r (% ch an g e fro m b ase y ear eq u ilib riu m )
0
0 .5
1
1 .5
2
2 .5
3
0 3 6 9 12 15 18 21 24 27 30 33 36 39 42 45 48 51 54 57 60 63 66 69 72 75
Pe rio d /y e a r
Q u as i IOB arg ain in gF low m ig ra tion
Figure 4. Impact on Scottish GDP from a 10% increase in ROW export demand to the RUK Primary, M anufacturing and Construction sector (% change from base year equilibrium)
-1
-0.5
0
0.5
1
1.5
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60 62 64 66 68 70 72 74
Period/year
Quasi IOBargainingFlow m igration
Figure 5. Im pact on ROW export dem and for total RUK production in response to a 10% increase in ROW export dem and to the Prim ary , M anufacturing and Construction sector (%
changes from base year equilibrium )
0
1
2
3
4
5
6
7
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60 62 64 66 68 70 72 74
Period/year
Q uasi IO Bargain ing
Flow m igration
Figure 6. Impact on RUK exports to Scotland in response to a 10% increase in ROW export demand to the Primary, M anufacturing and Construction sector (% changes from base year
equilibrium)
-2
-1.5
-1
-0.5
0
0.5
1
1.5
0 3 6 9 12 15 18 21 24 27 30 33 36 39 42 45 48 51 54 57 60 63 66 69 72 75
Period/year
Quasi IO Bargain ing
Flow m igration
Figure 7. Im pact on R OW export dem and for to tal Scottish production in response to a 10% increase in R O W export dem and to the Prim ary , M anufacturing and C onstruction sector (%
changes from base y ear equilibrium )
-2 .5
-2
-1 .5
-1
-0 .5
0
0 3 6 9 12 15 18 21 24 27 30 33 36 39 42 45 48 51 54 57 60 63 66 69 72 75
Period /year
Q uasi IO B argain ing F low m ig ration
Figure 8. Im pact on Scottish exports to R UK in response to a 10% increase in R OW export dem and to the Prim ary , M anufacturing and C onstruction sector (% changes from base y ear
equilibrium )
0
0 .5
1
1 .5
2
2 .5
3
3 .5
0 3 6 9 12 15 18 21 24 27 30 33 36 39 42 45 48 51 54 57 60 63 66 69 72 75
Period /year
Q uasi IO B argain ing
F low m ig ration
Table 8. Post-shock CO2 Trade Balance Between Scotland and RUK (% change from base) - CGE Period 1 (Quasi IO)
Pollution supported by: Total regionalScottish Govt Scottish Capital Scot-ROW RUK Govt RUK Capital RUK-ROW emissions of CO2
Pollution generated in:Scotland -0.09% 0.05% -1.84% -0.34% 0.23% 3.04% -0.08%RUK -1.60% -2.37% -3.92% -0.78% -0.93% 2.43% 0.51%Total (UK) emissions supported by -0.72% -1.39% -2.72% -0.76% -0.88% 2.45% 0.46%
Environmental trade balance:Scot pollution supported by RUK final demand 1.26%RUK pollution supported by Scot final demand -2.79%Scotland's CO2 trade balance -14.63%
Table 5. Post-shock CO2 Trade Balance Between Scotland and RUK (% change from base) - Type II Input-Output
Pollution supported by: Total regionalScottish Govt Scottish Capital Scot-ROW RUK Govt RUK Capital RUK-ROW emissions of CO2
Pollution generated in:Scotland 0.00% 0.00% 0.00% 0.00% 0.00% 6.72% 1.10%RUK 0.00% 0.00% 0.00% 0.00% 0.00% 6.92% 3.06%Total (UK) emissions supported by 0.00% 0.00% 0.00% 0.00% 0.00% 6.91% 2.89%
Environmental trade balance:Scot pollution supported by RUK final demand 2.88%RUK pollution supported by Scot final demand 0.00%Scotland's CO2 trade balance -8.41%
Table 10. Post-shock CO2 Trade Balance Between Scotland and RUK (% change from base) - CGE Period 75 (Quasi IO)
Pollution supported by: Total regionalScottish Govt Scottish Capital Scot-ROW RUK Govt RUK Capital RUK-ROW emissions of CO2
Pollution generated in:Scotland -0.06% 0.93% -0.10% -0.60% 2.34% 6.09% 1.21%RUK -0.47% 0.55% -0.50% -0.50% 2.38% 6.43% 3.24%Total (UK) emissions supported by -0.23% 0.70% -0.27% -0.50% 2.38% 6.42% 3.06%
Environmental trade balance:Scot pollution supported by RUK final demand 3.06%RUK pollution supported by Scot final demand -0.23%Scotland's CO2 trade balance -9.83%
Table 13. Post-shock CO2 Trade Balance Between Scotland and RUK (% change from base) - CGE adjustment (alternative visions of the labour market)
Period/year after demand disturbance introduced:1 5 10 15 20 30 50 75
Scotland's CO2 trade balanceQuasi IO -14.63% -13.78% -12.59% -11.75% -11.16% -10.41% -9.90% -9.83%Bargaining -15.00% -14.86% -14.42% -14.06% -13.93% -13.84% -13.84% -13.84%Flow migration -15.00% -14.50% -13.59% -12.78% -12.31% -11.58% -11.03% -10.89%
Figure 9 Scotland's CO2 trade balance with RUK in the 10 years following the demand shock (millions of tonnes CO2)
-6.60
-6.40
-6.20
-6.00
-5.80
-5.60
-5.40
-5.20
-5.00
-4.800 1 2 3 4 5 6 7 8 9 10
Period/year
Quasi IO Bargain ing Flow m igration
Figure 10. CO2 embodied in gross interregional trade flows between Scotland and RUK in the 10 years following the demand shock (% change from base year equilibrium)
-0.04
-0.03
-0.02
-0.01
0.00
0.01
0.02
0.03
0 1 2 3 4 5 6 7 8 9 10
Period/year
Quasi - Scot po llution supported by RUK final demand Quasi - RUK po llution supported by Scot f inal demandBarg - Scot po llution supported by RUK final demand Barg - RUK po llution supported by Scot f inal demandFlow - Scot po llution supported by RUK final demand Flow - RUK pollution supported by Scot f inal demand
ConclusionsConclusions
• Results demonstrate limitations of IO modelling for the purposes of marginal analysis
• CGE framework is better placed to analyse the impact of marginal changes in economic activity and/or policy actions on interregional or international trade balance in pollutions,
• Qualifications numerical analysis presented here- Demand shock applied here is somewhat blunt and unrealistic- Data limitations due to estimated and experimental data embodied in
interregional IO and SAM databases- 3-sector, 2-region national framework may be too highly aggregated for
analysis of environmental issues
Current researchCurrent research
• ESRC Climate Change Leadership Fellowship programme
• Improvement UK interregional IO data – Scottish Government, WERU, SEI
• UK interregional IO and CGE applications – ESRC/TSG collaborative studentship, WERU, BRASS, RESOLVE, SEI, ISA
• US interregional IO and CGE applications – 5 Midwest regions and RUS – REAL
• Econometric estimation key functions and parameter values
• Introduction KLEM production functions, crossover ESRC project on modelling ‘rebound effects’
Regional and interregional CGE modelling for US states?Regional and interregional CGE modelling for US states?
• There are a number of examples of regional CGE models for US states
• Over the last few years, the AMOS team at Strathclyde have collaborated with the REAL team at Illinois to develop a CGE model of the Chicago economy
• Under the current ESRC Fellowship project, we are attempting to extend to a 6-region model of the MidWest economy (5 midwest states and rest of US)
• Starting point is IO tables, then extend to SAM, followed by identification of appropriate model specifications for economies to be modelled (including econometric estimation of key parameter values)
• AMOS can be applied to other economies, but can also build up models from scratch using software such as GAMS