industrial policy and the wto

8/19/2019 Industrial Policy and the WTO

1/50

UNITED NATIONS CONFERENCE ON TRADE AND DEVELOPMENT

POLICY ISSUES IN INTERNATIONAL TRADE AND COMMODITIES

STUDY SERIES No. 6

INDUSTRIAL POLICY AND THE WTO

by

Bijit BoraUNCTAD and Flinders University of South Australia

Peter J. LloydUniversity of Melbourne

Mari Pangestu

Centre for Strategic and International Studies, Jakarta


2/50

iii

NOTE

The views expressed in this study are those of the authors and do not necessarilyreflect the views of the UNCTAD secretariat.

The designations employed and the presentation of the material do not imply theexpression of any opinion whatsoever on the part of the Secretariat of the United Nationsconcerning the legal status of any country, territory, city or area, or of its authorities, or concerning the delimitation of its frontiers or boundaries.

Material in this publication may be freely quoted or reprinted, but acknowledgementis requested, together with a reference to the document number. A copy of the publicationcontaining the quotation or reprint should be sent to the UNCTAD secretariat:

Chief Trade Analysis Branch

Division on International Trade in Goods and Services, and CommoditiesUnited Nations Conference on Trade and Development

Palais des NationsCH – 1211 Geneva

UNCTAD/ITCD/TAB/7

UNITED NATIONS PUBLICATION

Sales No. E.00.II.D.26

ISBN 92-1-112497-2

ISSN 1607-8291

Copyright 8 United Nations 2000All rights reserved


3/50

iv

ACKNOWLEDGEMENTS

The authors would like to thank Bernard Hoekman, Will Martin, Joerg Weber andSusan Teltscher for helpful comments on an earlier draft. Helpful comments were alsoreceived from Rajesh Chand and Miguel Rodriguez-Mendoza, our discussants, when the

paper was presented at the World Bank Global Conference on the Millennium Round, heldat the World Trade Organization in Geneva on 20 and 21 September 1999. Manuela Tortora,the internal referee, greatly improved the paper with her incisive and helpful comments.


4/50

vi

CONTENTS

I. THE ISSUE.................................................................................................................1

II. THEORY OF INDUSTRIAL POLICY..................................................................4A. Second-best arguments for industry protection................................................................ 5B. Technology development................................................................................................ 7C. Strategic trade policy.................................................................................................... 11

III. EXPERIENCE WITH INDUSTRIAL POLICY IN EAST ASIA....................13

A. General trends .............................................................................................................. 13B. Industrial policy and implementation of Uruguay Round commitments......................... 14C. Changes since the Uruguay Round................................................................................ 15D. Policies towards foreign investors................................................................................. 17

IV. WTO RULES AND INDUSTRIAL POLICY .....................................................19

A. Import protection.......................................................................................................... 19B. Subsidies and export promotion.................................................................................... 20C. Agreement on Trade-related Investment Measures........................................................ 21D. Agreement on Trade-related Aspects of Intellectual Property Rights ............................. 22E. General Agreement on Trade in Services ...................................................................... 23F. Special and differential treatment.................................................................................. 23G. Implications of WTO rules for industrial policy ............................................................ 24

V. IMPLICATIONS FOR DEVELOPING COUNTRIES.....................................26

A. Import protection.......................................................................................................... 26B. Export promotion ......................................................................................................... 27

C. Competition policy ....................................................................................................... 27D. Market access for foreign investors............................................................................... 28E. Performance requirements for foreign investors ............................................................ 29F. Special and differential treatment.................................................................................. 31

VI. CONCLUSIONS......................................................................................................33

REFERENCES .....................................................................................................................35

ANNEX ..................................................................................................................................41

WTO Dispute Panel case on Indonesia: certain measures affecting the automobile industry ............................................................................................................................ 41Indonesia’s automobile policy............................................................................................... 41Objectives of the national automobile programme................................................................. 41Policies................................................................................................................................. 42WTO rules affected by the national automobile programme.................................................. 42Ruling................................................................................................................................... 43State of play.......................................................................................................................... 43Lessons................................................................................................................................. 44

Box 1 . Instruments of industrial and export promotion policies:Republic of Korea and Japan....................................................................... 15

Table 1 . Policies and measures to promote exports in Asia........................................ 39Table 2 . Evolution of industrial policies in East Asia................................................. 40Annex table 1 . Alleged breaches of multilateral trade agreements

by Indonesia’s national automobile programme ....... ....... ........ ....... ....... ....... 45


5/50

1

I. THE ISSUE

The general objective of promotingexports and achieving rapid structuralchange and economic growth has been anintegral part of development economicsand policy-making for many decades.There has been a succession of differentapproaches and thinking with regard tohow this objective can best be met, rangingfrom inward-looking or import substitutionindustrialization behind high protection, tooutward-oriented or export orientation and

promotion strategies considered to be part

of the success story of East Asia. The rangeof instruments used for conductingindustrial policy has also changed with theevolution of multilateral trading rules, aswell as unilateral liberalization, the latter occurring within a framework of structuraladjustment that is required in order to staycompetitive and in some cases to accessinternational finance. The combination of strategy and instruments used has been thesubject of numerous studies, with mixed

results on the value of interventions andtheir outcomes. There has also been a

plethora of studies which show thatindustrialization behind protective wallshas often extended beyond reasonable

periods of “infancy” and has led toinefficiency and welfare losses, andentrenched vested interests.

Despite the strong theoretical caseagainst activist industrial policy, it is stillwidely pursued in a number of countries. 1

1 A number of countries have pursuedinterventionist industrial policies with some degreeof success. In East Asia, the Republic of Korea,Taiwan Province of China and Japan are threeexamples of where government intervention in theform of activist policies was important for the paceand direction of development (Lall, 1994; Singh,1996; Asian Development Bank, 1999: 208-210).This intervention, however, was broad-based andnot confined to protection. It included aspects of

In the 1990s, however, the context inwhich it is pursued is different. Rapidtechnological change, shorter productcycles and developments in informationtechnology have combined with

privatization, and trade and foreigninvestment liberalization to produce aglobal economy that is distinctly different.In this context, developing countries arestriving to ensure that their industries arecompetitive by using industrial policy to

promote particular sectors.

It should be pointed out at theoutset that the term “industrial policy” isnot a well-defined one. It is ill-defined inrelation to the objectives, the industries thatare covered and the instruments that areused. The World Bank (1992) has

provided a working definition of industrial policy as “government efforts to alter industrial structure to promote productivity

based growth”. 2 This definition is useful

since it focuses on the objective of economy-wide factor productivity growthrather than on merely changing thestructure of industrial outputs.

With regard to objectives, manydeveloping countries have in mind the

potential for long-run productivityimprovements. However, in most casesindustrial policy is pursued with multipleobjectives, including short-termemployment, increased output, better targeted technological promotion, financing andskill development. In an effort to replicate thissuccess many developing countries have taken the

position that they too should be allowed to pursuesuch policies and not be restricted by multilateralrules.2 A recent paper by Martin and Mitra (forthcoming)shows that the productivity growth rate inagriculture is higher both on average and for groupsof countries at different stages of development.


6/50

2

income distribution and enhancingtechnological capacity. There are oftenalso, rightly or wrongly, non-economicobjectives of national pride and prestige, aswell as the perceived need to promote“strategic” domestic industries.

These objectives are further confused to the extent that manydeveloping countries have taken the viewthat ownership of assets matters. There is aconcern that foreign ownership may notalways fit in well with broader development objectives, includingenhancing domestic capabilities. 3 In somecases, foreign ownership could crowd outdomestic firms. Thus, even if the WorldBank definition is adopted and

productivity-based growth materializes, thefact remains that developing countries haveraised concerns about the source of growth.Growth in per capita GDP based ondomestic assets seems to be preferred togrowth based on foreign assets. The latter would not constitute “development” per se.Some countries may be prepared to trade

off a lower rate of growth in per capitaGDP combined with lower foreignownership against a higher rate of growthwith more foreign ownership. 4

The focus of “industries” almostinvariably seems to be on themanufacturing sector. This leaves outagriculture, services and mining, althoughthese sectors raise much the same issues.Processing of agricultural and mining

products occurs in the manufacturingsector, and the line between unprocessedand semi-processed products on the onehand and processed products on the other isarbitrary. Similarly, many services sector 3 For a discussion of how foreign ownership mattersin the context of development see UNCTAD(1999a).4 One possible reason for this could be the

perception that openness would increase thevulnerability of the country to external shocks.

industries add value to manufactures, andthey raise issues that parallel those of industrial development in manufacturingindustries. Restriction of the discussion tomanufacturing industries alonediscriminates against non-manufacturingindustries and leads to inefficiencies in the

production allocation of the economy.Although the growth of industry output andexports in some developing countries inAsia and elsewhere is concentrated inmanufactures, in others primary andservices sector development is animportant part of growth. In this paper,industrial policy is not restricted by sector.

With regard to instruments, thetraditional focus has been on tariffs or output-based subsidies or export subsidiesto industries as a way of rectifying allegedmarket failures due to externalities,missing markets or other failures (Lall,1994). These have also been used to directresources into certain sectors that may beconsidered more conducive to developmentsuch as those with high growth potential.

Recently, however, more attention has been devoted to factor markets, especiallyforeign direct investment (FDI). Here the

belief is that FDI is a bundle of assets thatcan contribute to economic development.At the same time, however, the use of theseassets by affiliates of transnationalcorporations (TNCs) can also hinder acountry’s development efforts.Government intervention is then requiredin order to alter the operations of foreign

affiliates so as to minimize their negativeeffects (UNCTAD, 1999b).

In reality, developing countrieshave used a mix of import protection,export promotion, foreign investmentrestrictions and performance requirements,tax incentives and other measures to

promote industrialization. The types of instruments used by developing economieshave changed, especially since the 1980s,


7/50

3

owing to increased restrictions on their usethrough multilateral and regionalagreements, as well as domestic regulatoryreforms initiated through structuraladjustment loans or domestic efforts torestructure their economies. The major changes faced by countries resulting frommultilateral rules are the various GATTCodes that emerged prior to the UruguayRound, particularly the GATT Code onSubsidies and Countervailing Duties of 1979, which restricted signatories’ use of export subsidies. The multilateral tradeagreements, agreed upon by WTOmembers as part of the Uruguay Roundnegotiations, have created new disciplineson the use of such policies. Meanwhilecommitments under the Uruguay Roundand regional agreements, and unilateralefforts to liberalize, have led to a decline inthe use of tariff and non-tariff measures.

The aim of this paper is to reviewthe objectives and instruments of industrial

policy in a changing global context andmultilateral rules and discipline. The

remainder of this paper is divided into four sections. In the next section an analyticalreview is undertaken of the objective of,

and justification for, industrial policy pursued by countries. The importance of having an analytical framework is that it

becomes the benchmark against whichobjectives, instruments and outcomes can

be measured. In the third section the useof different instruments for industrial

policy is reviewed. An attempt is made toassess whether changes have been due tocompliance with multilateral and/or regional commitments, or due to unilateralreform efforts. This section also discusseswhether new non-traditional instruments to

pursue protection were needed once the useof traditional instruments becamerestricted. The fourth section focuses onthe role of industrial policy in the post-Uruguay Round era with a view to the nextround of WTO negotiations. It examines

both the theoretical and the applied aspectsof industrial policy before surveying theextent to which existing WTO rules affecta member’s ability to pursue industrial

policy objectives. The possibilities andimplications of revising rules that affect theuse of industrial policy instruments are

discussed in the fifth section. The lastsection sets out a number of conclusions.


8/50

4

II. THEORY OF INDUSTRIAL POLICY

This section begins with a brief review of the traditional argument againstinfant industry protection. This argumentstill lurks behind most advocacy of government assistance for industrialdevelopment in developing (anddeveloped) countries. Moreover, anexamination of it highlights pitfalls in

policy development which apply equallyto other modern arguments since they areessentially variants of the old infantindustry argument.

The traditional infant industryargument justified a tariff, or a subsidy

based on the output of firms which havean equivalent effect on output, on the basisof some dynamic externality. Kemp(1964) provides probably the first carefulstatement of the argument. He identifiedlearning processes such as worker learning

by doing or on-the-job training as thesource of cost saving and distinguished

between learning processes which areinternal to the firm and those that areexternal. The former are appropriable bythe firm. Only those that are external tothe firm warrant assistance, and then onlyif the reductions in cost over timecompensate for the higher costs during the

period of assistance, 5 with all flowsappropriately discounted. The tax subsidyis temporary.

This argument immediately raisesa number of policy difficulties. It never

provides a justification for blanketassistance to all firms in an industry or even a sub-industry since the existence of an externality and the required cost savinghave to be demonstrated in every case.

5 If the instrument of assistance is a subsidy rather than a tariff, one should add the costs imposed bythe tax that funds the subsidy costs.

Baldwin (1969) raised a seconddifficulty. He pointed out that a tariff (or subsidy) provides no incentive per se for afirm to acquire more knowledge, becauseit is an output-based intervention. A firmwill increase output by the least costlymethod, not necessarily by acquiring moretechnology. The correct policy implied bythe argument, supposing that it isdemonstrated, calls for a subsidy related toknowledge creation, for example a subsidyon the particular workers who learn by

doing. Most knowledge or skillacquisition is process-, job- or product-specific. In these cases the correctivesubsidy will be confined to the process,

job or product, or whatever, and based onthe variable with which the externality isassociated. Thus, there are substantialqualifications to the infant industryargument.

This line of argument is in fact an

example of a much more general theme inthe literature of government intervention.Each externality or market failure calls for a tax subsidy whose base is the variablewhich generates the externality or failure,and the tax-subsidy rate will be the ratethat has the optimal effect. Bhagwati(1971) gives an early statement of the rule.Any tax subsidy other than the optimal taxsubsidy causes what Corden (1974) called

by-product effects that impose costs on theeconomy. Moreover, the tax-subsidy ratevaries across firms in an industry if thestrength of the effect justifyingintervention varies across firms. Evenwhen an intervention is called for, thechoice of a suboptimal instrument with

by-product effects reduces the net benefitsobtainable from the optimal instrumentand may in fact be welfare-reducing.


9/50

5

A. Second-best arguments for industryprotection

There are a number of variants of the infant industry argument based on the

presence of other tax subsidies or constraints that produce an argument for

protection of importable goods. Instead of deriving from the presence of someexternality that operates over time, thesearguments derive from the presence of other distortions in the economy, such astariffs or commodity taxes, which areconsidered unremovable and therefore

permanent. These are applications of thetheory of the second-best. In their famousarticle, Lipsey and Lancaster (1956)showed that if an optimization problem ismodified by the addition of newconstraints, the first-order conditions thatcharacterized the first-best optimizationare in general violated.

In the context of internationaltrade, this means that the standard findingthat free trade is the best policy for a smallcompetitive economy may not beapplicable. In the 1970s, a number of economists produced examples of thistheory. For instance, if a subset of importable goods is subject tounremovable distortions in the form of tariffs, the second-best optimum calls for tariffs or trade subsidies for some or all of the remaining goods whose prices are notfixed. The nature of the second-best set of tax subsidies can be characterized in terms

of the relationships of substitutability andcomplementarity between the two sets of goods (see Lloyd, 1974; Hatta, 1977).Again, as with the infant industryargument, the optimal governmentintervention and the associated rates of taxsubsidy depend on the exact nature of theconstraints. These arguments usually findthat a tariff is called for on the outputs of some industries, but this applies becausein the models that were used each industry

produces only one good. If there aremany-good industries, the second-best taxsubsidies will vary across the goods in anindustry.

These models also pose major difficulties for policy makers. In the first

place, it is not clear why we shouldconsider that some tariff rates or other government tax subsidy rates are

permanently totally unchangeable whileothers are freely changeable. Secondly, asthe second-best is the solution to acomplex constrained general equilibriumof an economy, the determination of thesecond-best requires perfect knowledge of all aspects of the economy. This includesall supply and demand parameters, in factthe determinants of the behaviour of allagents in the economy. This is grosslyunrealistic. Third-best interventions madein ignorance of the true values of some

behavioural parameters may be welfare-reducing.

Specific examples of this type of

theory are to be found in the recentliterature on protection in developingcountries. Some of the authors havechosen a trade-related investment measureas the form of government intervention. 6

One example is the use of localcontent programmes. Many developingcountries have introduced such

programmes in the automobile industry in particular – for example, India, Malaysia

and the Philippines in Asia and Mexico,Brazil and Argentina in Latin America(Moran, 1998). This has been donechiefly in the belief that any policy thatincreases the local content of a unit of 6 There is an older literature which makes a similar case for export subsidies on the ground that tariffsand other policies which promote importsubstitution have discriminated against exports; seeHarris and Schmitt (1999) for a review of thisliterature.


10/50

6

output, i.e. the proportion of domesticvalue added in the production of thegoods, must be beneficial. This naive

belief ignores the effects whichencouraging the production of oneindustry has on other industries; the greatstrength of general equilibrium theory isthat it brings out the economy-wide effectsof intervention in one industry. This

belief also ignores the volume effectswithin the industry. Dixit and Grossman(1982) made an elegant analysis of content

plans in a general equilibrium model witha continuum of stages. Although there isno FDI in their model, the main effects of a content plan are highlighted. They showthat a content plan will raise the cost of intermediates to downstream and final

producers and thereby lower their effective protection. It will also increasethe range of goods produced but may beanti-protective in terms of the aggregatelabour employed and value added in the

protected industry. Moreover, it iswelfare-reducing.

There are second-best theorymodels that purport to show that content plans are welfare-improving for developing countries. These modelsintroduce additional constraints due to the

presence of unemployed factors generated by a fixed minimum wage or a tariff onthe final good.

Chao and Yu (1993) put forwardan argument in favour of protection by

means of content plans. They constructeda model of a dual economy with urbanunemployment. The urban sector

produces a processed good and the ruralsector produces an agriculturalcommodity. The production of the

processed good requires intermediateinputs that can be imported or sourcedlocally. In the tradition of the Harris-Todaro model, a fixed minimum wage isset institutionally in the urban sector

giving rise to urban unemployment, whilethere is a flexible wage in the agriculturalsector. It is easily shown that urbanunemployment makes the marginal ratesof transformation of the manufacture for the agricultural good exceed the relative

price ratio. This gap implies an allocativeinefficiency in the economy. Essentially,the agricultural output is under-produced

because at the margin the social value of this good relative to that of themanufacturing good is greater than therate of transformation.

This is a variant of the standardmodel of international trade with a factor market distortion which has been studiedextensively as an argument for protection.The twist in favour of content protectioncomes from adding structure to the

production sector of the model. Theassumption made is that the domesticmaterials are produced by the agriculturalsector and used in the manufacture of theurban product, and that the materialssector is an infant industry whose

domestic price is higher than the world price. As in the Dixit–Grossman model,the content protection increases thecontent per unit but reduces the aggregateoutput of the protected industry. Thisleads to a decrease in demand for theoutput of the urban sector, which causes areversal of migration to the urban sector and reduces the gap between the rate of transformation and the relative price.Chao and Yu (1993) claim that the effect

is welfare-enhancing, although in their earlier paper they found that it waswelfare-reducing.

Richardson (1993) examines theeffects of a content plan in a second-bestsituation due to the existence of a tariff ona final manufacturing good. He adds FDIin both the final manufacture and itscomponent suppliers. The content planapplies only to the foreign investor and is


11/50

7

therefore a trade-related investmentmeasure (TRIM). A content plan hasoffsetting negative allocation effects

because it increases output and reducesimports in the components sector, and

positive revenue effects because itincrease imports of the final manufactureand reduces payments to foreign capitalwhich is specific to final manufactures. Adomestic content policy set at low

percentages is welfare-increasing in thiscontext.

Rodrik (1987) and Greenaway(1992) have made a similar second-bestcase for TRIMs in the form of exportshare requirements. Rodrik considers aneconomy which is host to FDI by amultinational corporation. The prior distortions are tariffs, which inducedforeign investment or oligopolistic

behaviour in the industry. As is wellknown, foreign investment isimmiserizing if the import industry iscapital-intensive; it exacerbates the loss of tariff revenue. In this situation, an export

requirement may partially offset thenegative effect of foreign investment bylowering the profitability of the latter,reducing output and increasing imports or lowering the rate of return on foreigninvestment. In the case of oligopolisticinteraction between the multinationalcorporation and the local firms, an exportrequirement reduces the output of themultinational and shifts profits to the localsuppliers. Greenaway (1992) extends this

line of argument to a range of TRIMs.

Morrissey and Rai (1995) alsomake a case for TRIMs based on the prior existence of a range of restrictive business

practices by multinational corporations.These practices include some whichrestrict exports from the host country,such as international market allocation andintra-company restrictions on exports byaffiliates.

All these second-best argumentsare subject to the same objections as theinfant industry argument. If, as in themodel of Chao and Yu, the cause of thedistortion is a policy-induced minimumwage, this should be abolished and nofurther action is required. If, as in themodels of Richardson and Rodrik, the

prior distortion is a tariff on the finalmanufacture, the first-best plan is toreduce the tariff. If, as in the models of Rodrik and Morrissey and Rai, it is ananti-competitive form of business conduct,the appropriate instrument is acompetition law remedy. First-bestreform avoids the by-product loss of welfare due to the increase in the price toconsumers of the output of the protectedsector and the distortions of the productionstructure of the industry. This isrecognized by some of the authors, butthey accept that a tariff on the final goodis unchangeable while an implicit tariff onits components is feasible. Again, the

performance requirements have negative

effects if they are pushed too far. Acontent plan provides no incentive for upstream firms to acquire moreknowledge. The incentive is merely toincrease their outputs.

These arguments have an old-fashioned ring. They ignored the benefitsof technology and management fromforeign investments, and they belong to anera when there was no pressure on

developing countries to lower their existing tariffs and not to introduce newtrade-restricting measures.

B. Technology development

Some writers have advocated border assistance on the grounds of technology development. For example,Balasubramanyam (1991) bases his


12/50

8

argument for content plans on theincentives provided to multinationalcompanies to develop technology in thematerials sector of the host economy.However, the previous section makes itclear that government interventions todevelop technology should be technology-

based, not output-based.

With few exceptions, thetechnologies used by producers indeveloping countries are not the latest or most advanced in the world. Generallyspeaking, these countries can be classifiedas low- or middle-technology countries. 7

When new industries or products areestablished they will, in most cases, use atechnology already developed in someother more advanced or industrializedcountry.

In this context, FDI will beimportant as a vehicle for the transfer of technology. The literature has recognizedthat technology may be transferred in twoways:

• Foreign technology has a higher total factor productivity, which istransferred initially to theenterprise with foreign investmentin the host economy.

• The knowledge of the enterpriseswith foreign investment spills over to other firms in the sameindustries.

Each of these effects occurs withinindustries (appropriately defined); that is,they are intra-industry effects. The firstwill be called the direct technologytransfer effect and the second the spillover effect.

7 See section 4.4 on the role of the Agreement onTrade-related Aspects of Intellectual PropertyRights (TRIPS) in enhancing technology transfer to developing countries.

The direct technology transfer effect derives from an old argument thatforeign investors have a superior technology of production, which istransferable to foreign affiliates anddomestic firms. In recent years this effecthas been incorporated into a number of models of technology catch-up or technology ladders. They view thetechnology differences across nations asgiven. They are the result of past researchand development (R&D) or other

processes of technology acquisition andno attempt is usually made to explainthese differences.

Another model has been developedsince 1990 by a number of authors toendogenize technology transfer by linkingFDI to technological improvement in theform of new varieties of capital input (for a textbook treatment, see Barro and Sala-i-Martin (1995, chapter 6)). In this model,an economy produces a single good, y,using a Cobb–Douglas value added

technology:y = AH αK 1-α (1)

where A is an efficiency parameter, Hdenotes the human capital input and K the

physical capital input. The functionexhibits constant returns to scale withrespect to the two inputs. Physical capitalis a composite of different varieties of thecapital good, K j:

K = j = i

N

∑ [(K j1− α )]1/1-α (2)

The total number of varieties of capital good, N, is produced either bydomestic firms, n, or by foreign firms, n*.Thus, N = n + n*. K may be regarded asunits of effective capital. With thistechnology, the total effective capital


13/50

9

stock increases as the number of varietiesincreases, for a given number of units of capital, e.g. machines. This is a specialform of capital-augmenting technologicalchange.

Expansion in the number of varieties occurs as a result of R&D. Thecost of production of a new variety,therefore, has fixed set-up costs, F, andconstant marginal costs. It is assumed thatthese fixed costs are a decreasing functionof n* and of n/n*, that is F = F(n*, n/n*)where the partial derivatives with respectto both arguments are negative. The firstof these variables captures the idea thatforeign firms have an advantage in

producing new capital goods because of their accumulated knowledge. The secondcaptures the idea that countries that aremore backward technologically, asrepresented by the number of capitalgoods produced domestically, have anadvantage in catching up. Moreover, the

production function in equation (1)ensures that the marginal product of an

increase in the stock of capital is anincreasing function of the stock of humancapital in the economy. This model yieldsa regression equation for the economygrowth rate which contains the level of FDI and an interaction variable, FDIxH,with a positive coefficient. In anempirical study of cross-country rates of growth, Borensztein, de Gregorio and Lee(1998) found that FDI itself has aninsignificant effect on growth rates, but

the interaction terms are significant and positive.

An alternative model introducestechnological change through the capitalinput again but in a different way. Thismodel assumes that there are a fixednumber of capital input varieties, but eachvariety lies on a quality continuum and issubject to quality improvement over time(see Barro and Sala-i-Martin (1995,

chapter 7)). This yields a positiverelationship between FDI and growth,again with a positive interaction withhuman capital.

These models are simplistic withno international trade in goods and onlyone sector, but they are suggestive. Theyintroduce two key ideas. The first is arelationship between the variety of thecapital inputs and output, and the secondis a complementarity effect between FDIand human capital accumulation in thehost economy. With international trade,the composite capital input adds newvarieties through the importation of newcapital goods since the foreign varietiescan be supplied only by foreign producers.This is one vehicle of technologyimprovement.

The notion that knowledge spillsover from one firm to others has become

popular in recent years. Blomstrom(1989) provides an early and influentialstatement, although the idea was put

forward much earlier by Findlay (1978),who called it “contagion”. This notion hasan intuitive appeal, but it has beenmodelled in different ways.

One strand of the literature makesthe total factor productivity of a firm afunction of cumulative industry output

because of learning-by-doing. Grossmanand Helpman (1995, section 2) surveythese models. Suppose that there are only

two countries, the home (= host) countryand the foreign (= source) country. Thenthe output of a good (= industry), industryi, in some country is given by the

production function

Yi = A i (⋅)φI(v i) (3)

Vi is a vector of primary andintermediate inputs, and A i (⋅) is the indexof technical knowledge or know-how in


14/50

10

the country. The subscript for the firm isomitted on the assumption that the samevalue of the coefficients A i and a i apply toall producers of the good in one country. 8

Knowledge is transmitted among producers by making A i some function of cumulative outputs. The index A i can be afunction of the cumulative national outputof the good or of the cumulative worldoutput, or sometimes of the output of agroup of related outputs in an industry atthe national or international level; that is,Ai = α iYi + β iYi* where Y denotes thecumulative output in the country, *denotes the foreign country and δ i is aconstant. This generates a family of learning-by-doing spillover models.

In these models, spillovers occur asa function of cumulative aggregateindustry output at some level. It does notmatter whether the output in a country is

produced by home or foreign firms.Alternatively, one could make the shiftfactor A i (⋅) some function of FDI or of theactivities of the source enterprises.

Thus, there is some factor-disembodied kind of transfer among firms

but the actual mode of transfer is notspecified. Van and Wan (1999) introducethe idea that much technology transfer takes place through the establishment of new domestic firms staffed by workerswho were previously employed by foreignfirms and acquired work and productionskills and knowledge of the technology of

production from this employment. Thenew firms may be subcontractors for theforeign firm or competitors, or even

produce other goods that use a similar technology. New domestic firms have atechnology that is superior to old domestic

8 In these models it has usually been assumed thatlabour is the only factor of production but thespecification is easily generalized.

firms. This might be called the new firmseffect.

These models do not provide muchsupport for government subsidies or other interventions to promote industrialdevelopment. In the direct technologytransfer models, where the technology isassumed to be in the form of some firm-specific blueprint or asset and istransferred costlessly, the firms

purchasing the new capital inputsappropriate the benefits of increased

productivity. The same applies where thetechnology transfer occurs through theintroduction of new varieties or newqualities of investment goods. The policyimplication of this model is thatGovernments should not impose tariffs or other restrictions on imports of thesecapital goods.

In the spillover models, where theimprovements in total factor productivityare associated with learning-by-doing,there is a kind of externality. However,the host country Government has little

possibility of capturing the benefits if theyare associated with the aggregate industryoutput in the rest of the world or of thewhole world. If there are spilloversassociated instead with FDI in someindustries, this is an added benefit of FDI,

but such benefits will be industry-specificand the Government would have to knowthe mechanism by which the technology istransferred in order to capture them. For example, if they are due to the new firms

effect, they will happen in the markets, provided that there are no obstacles to theformation of new firms. FDI incentiveswill be a crude and generally ineffectiveway of capturing the benefits of FDI asthese benefits will be firm- or process-specific, where they exist. R&D subsidiesare not warranted unless there is somegeneral uniform externality associatedwith R&D. The best way to maximize the

benefits of technology transfers associated


15/50

11

with FDI is through the adoption of generic measures aimed at improving theoverall regulatory and economicenvironment by enhancing competitionand improving human capital skills andtechnological capacity.

The critical question in this debateis the nature of the technology transfer.Grossman and Helpman (1995: 1334)concluded their survey of technology andtrade as follows: “…what can the Southdo to encourage technology transfer toindigenous agents without causing the

Northern innovators to take their businesselsewhere? To answer these questions wewill need models that pay closer attentionto how knowledge is transmitted withinand between firms.”

C. Strategic trade policy

Strategic trade policy is a set of cases developed in the 1980s thatsupposedly justify government

interventions. The distinguishing featureof this body of theory is that thearguments hinge on the existence of strategic interdependence among a smallnumber of firms. Brander (1995) surveysthe theory.

The standard example is a modelof two duopolists, each from a differentcountry, competing in a third-countrymarket. If there are no home market

effects, an export subsidy granted by theGovernment of one of them may improvenational welfare by allowing the domestic

producer to earn additional profits in theexport market that exceed the amount of the subsidy payment. Note that in thisthird-country market model an outputsubsidy is equivalent to an export subsidy.The model can be extended to a fixednumber of competitors greater than two.It can also be extended to markets in

which the firms make domestic sales,although in this case the gain is less sincethe increase in profits is offset by the lossof consumer surplus and the optimalintervention by one country is now a tariff.

However, the outcome isextraordinarily fragile as it is subject to anumber of conditions. It requires thatfirms’ decision variables be strategicsubstitutes, in that greater output by thesubsidized firm reduces the profits of itscompetitor. It is not robust with respect tothe assumptions of the model; if theconditions of entry, the choice of decisionvariable (Cournot or Bertrandcompetition, or whatever) or the presenceof economies of scale or other features of the model change, the nature of theoptimal intervention changes. Theselection of the optimal level of the(optimal) instrument imposes substantialinformation requirements on theintervening Government. This kind of intervention also lends itself to politicalmanipulation, with the possibility that the

subsidies will go to producers that seek to protect themselves from foreigncompetitive pressures rather than to the

producers who can shift profits. Finally,such strategic policy interventions may bein the interests of one country interveningalone but if the second country retaliates,there is a prisoners’ dilemma – both lose.This is because the result stems from

profit shifting between markets. It is notefficient from the point of view of the

world economy to have any intervention.

Other strategic policy cases withdifferent assumptions about the nature of competition are subject to the sameobjections. Brander (1995: 1446), himself one of the architects of strategic trade

policy theory, concluded his survey withthe remark that “It seems natural to expectthat strategic trade policy can only expandthe scope for socially wasteful transfer-


16/50

12

seeking…Even if free trade does notemerge as an optimal policy in normativestrategic trade policy models, once

political economy considerations are takeninto account, perhaps it is the best we cando.”

There is, therefore, nothing instrategic trade theory to recommend it todeveloping countries. In most markets,these countries are price takers. If themarket conditions should conform to somemodel of strategic trade policyintervention, developing countries do notsatisfy the informational and politicaleconomy requirements for successfulintervention. They would do better to takeaction to increase competition.

There is another even more powerful objection to the theory. Itignores completely the rules of the WTO.These exclude export subsidies for

manufactures and severely constrain thelevels of tariff rates as most industrialtariffs are now bound. These rules haveevolved over 50 years precisely in order toconstrain national beggar-thy-neighbour

policies. Moreover, as with the second- best argument for tariffs, one should nottake the conditions of imperfectcompetition as given. The WTO has

become more concerned with competitionin world markets in recent years. Mostsmall-number competition markets are theresult of government-sanctioned restraintson entry (such as the tolerance of exportcartels) rather than natural monopoly.Action in the WTO and other intergovernmental forums such as theOrganisation for Economic Co-operationand Development (OECD) should be toremove barriers to entry and cross-border competition and thereby make marketscompetitive.


17/50

13

III. EXPERIENCE WITH INDUSTRIAL POLICY IN EAST ASIA

The purpose of this section is to provide an insight into the practical use of industrial policies. The focus is on EastAsia, but this does not mean that LatinAmerica or industrial country experienceis not relevant. In the Latin Americancase, import-substitution policies were thehallmark of development strategy. But, asis widely recognized, the region’seconomic performance did not match thatof East Asia. This may reflect a lateadoption of liberalization policies, or

possibly even difficulties in implementingthe appropriate policies. Similarly, muchof the development of Western Europeand other industrial countries can beattributed to interventionist policies. Thissection points out, through the example of East Asia, that active industrial policieswere relevant. The same conclusionapplies to Latin America and theindustrialized countries.

A. General trends

It has always been an issue towhether government interventions usingvarious instruments ranging from creditand export subsidies, protection andexport promotion measures were effectivein the spectacular growth experienced byEast Asia. Some interventions weresuccessful, and part of the success was

because interventions, especially export promotion measures, were performance- based or contest-based (World Bank 1992), unlike government interventionswithout any performance requirements.

Table 1 summarizes the types of industrial policy pursued by countries inEast Asia. These policies have evolvedover the past three decades starting withimport substitution, which depending on

the size of the country has evolved intoexport orientation. 9 Export orientationnormally begins with assembly or originalequipment manufacturing (OEM) typemanufacturing as well as light industries,and over time the value added of exportsincreases. Often, export-oriented policieswill run parallel to import substitution

policies as protection is only removedgradually. However, beginning in themid-1980s, the increased need to attractFDI for technology and market access,

pressures from major markets such as theUnited States and Europe for marketopening, reducing trade surpluses,

becoming signatories to codes andimplementation of the Uruguay Round,and unilateral moves to reform domesticeconomies as the need to adjust wasrecognized – all these led to greater deregulation and liberalization in all theEast Asian economies. The economicliberalization and deregulation trends in

China and Viet Nam also provided agreater push to liberalize. Furthermore,the Republic of Korea and TaiwanProvince of China have both been subjectto a great deal of United States pressure toliberalize. Countries wanting to join theWTO have faced the same set of

pressures.

In general, the South-East Asiancountries have adopted a more liberal andmarket-oriented policy regime(Masuyama, Vanderbrink and Chia, 1997)compared with the North-East Asiancountries at the same level of 9 If countries have large internal markets, for example Brazil, China and India, the importsubstitution programmes allow for economies of scale for the protected firms. Country size,however, does not mean that the sectors will beefficient. See Bruton (1998) for a discussion of theevolution of import-substitution policies.


18/50

14

development. This is evidently becausethe South-East Asian economies followedlater in the export-orientedindustrialization strategy in the late 1970sand early 1980s. For instance, Indonesiahad to abandon its export-subsidy schemein the mid-1980s. It consisted of a dutydrawback provision which, because of theway it was calculated, provided both asubsidy and an export credit to exporters,the latter at interest rates below marketrates. To the extent that developingcountries still pursue industrial policy, thetypes of instruments will have to vary.

B. Industrial policy andimplementation of Uruguay Roundcommitments

Singh (1996) provides a usefulsummary of the types of export promotionand import restrictions used by theRepublic of Korea and Japan, which areoften identified as industrial policyinstruments (box 1), and analyses which

ones would no longer be valid under theimplementation of the Uruguay Roundcommitments. Table 1 provides a similar categorization for a number of Asiancountries.

Many of those falling under theexport promotion and import restrictionswould not now be allowed under multilateral rules, and some of the other

policies would violate the new

Agreements, especially under TRIMs,subsidies and TRIPS. Only instrumentssuch as government provision of information to exporters and changes inthe exchange rate would still be allowedunder the present rule structure. Export

promotion agencies such as the JapanExternal Trade Organization (JETRO) arestill allowed, as long as their task is onlyto provide information and not to provideexport guarantees or insurance elements.

Indeed, all the Asian countries in table 1have an export promotion agency, but of course its quality and effectiveness varyacross countries.

Another major instrument for subsidizing interest rates and preferentialcredit allocation used by the Republic of Korea was ruled out when linked toexports under the GATT Code onSubsidies and Countervailing Duties, andmore generally by the Uruguay RoundAgreement. Foreign investmentconditional on domestic content or trade

balancing would be in violation of TRIMs.

A number of instruments such asgeneral fiscal concessions, provision of subsidized R&D up to a certain level, andmeasures to promote corporate investmentand discourage payment of dividends arestill acceptable.

Since the mid-1980s most South-East Asian countries have adopted export-oriented policies maintaining competitive

real exchange rates (after 1995, with sharpmovements of the yen, this changedsomewhat) and promoting inward FDI.Most of the countries deregulatedrestrictions on FDI, Indonesia being oneof the boldest ones with a dramatic FDIliberalization being introduced in 1994,the year Indonesia was host to APEC.Apart from FDI, South-East Asiancountries have also been more flexible andopen about other factors of production

such as labour. Singapore has an open policy towards employing skilled andsemi-skilled labour from outsideSingapore, and Malaysia has had todepend on foreign workers (many of whom are Indonesians) to meet theshortage of labour domestically(Masuyama, Vanderbrink and Chia,1997).


19/50

15

Box 1Instruments of industrial and export promotion policies:

Republic of Korea and Japan

Export promotion and import r estri ctions • Import restrictions, both general and specific;• Favouring particular sectors for export promotion, in some cases particular firms for that

purpose;• Seeking compliance for subsidies given to exporters by means of export targets for

specific firms (Republic of Korea);• Interest rate subsidies and the availability of credit and foreign exchange to favoured

firms that meet export targets;• General export promotion through JETRO (Japan) and KOTRA (Republic of Korea);• Provision of infrastructure, including human capital, in support of exports;• Taxation relief on imported inputs and on R&D expenditures;

• Allowing favoured conglomerates to import capital goods and foreign technology and toraise cheaper finance on international markets.

I ndustri al poli cy measur es • Lax enforcement of competition policy, including the extensive use of cartels;• Government creation and promotion of conglomerates (Republic of Korea);• Tax concessions to corporations to increase investment;• Promotion of a close, long-term relationship between finance and industry which was

critical to the implementation of the industrial policy;• Labour repression to ensure labour peace in a period of structural change (Republic of

Korea)• Establishment of State industries to enhance industrial development (Republic of Korea)• Extensive administrative guidance

Source : Singh (1996: 163).

Table 2 and other informationindicate first that prior to the crisis therewas a strong trend towards economicliberalization, and second that there has

been a greater emphasis on

complementary policies, which will pushindustrial restructuring. Trade andinvestment liberalization alone are notsufficient. A variety of policies haveemerged to upgrade industrial structure.The types of policies that fall into thiscategory include R&D, infrastructure,development of strategic industries, and

policies aimed at forming industrialclusters. The Republic of Korea andTaiwan Province of China, for instance,

have emphasized government subsidiesfor R&D. Singapore has attempted todevelop services, and then knowledgeindustries, by providing fiscal incentives.

C. Changes since the Uruguay Round

There has been a general decline inthe use of tariffs for import protection. Inthe Uruguay Round, trade weighted-boundmost-favoured-nation (MFN) tariff averages declined substantially for Japan(from 3.9 to 1.7 per cent), the Republic of Korea (from 18 to 8.3 per cent) andThailand (from 37.3 to 28.0 per cent), but


20/50

16

only slightly for the Philippines (from23.9 to 22.2 per cent) and Malaysia (from10.2 to 9.1 per cent). On the other hand,they increased for Indonesia (from 20.4 to36.9 per cent) because of binding at amuch higher level than applied rates.Even though the percentage boundincreased substantially, the rate at which itwas bound remained high for somecountries. As expected, tarrification of agriculture led to a jump in the averagetariffs of the East Asian countries after theUruguay Round.

More recent data on applied tariffsfor countries of the Asia-PacificCooperation Forum (APEC), based onUNCTAD Trade Analysis andInformation System (TRAINS) data for 13APEC economies, show that averagetariffs for APEC have come down over time apart from those in agriculture as awhole, which rose in 1995 and then fell in1998. Average tariffs in agriculture arenow higher than those in manufacturing.At the two-digit level, the agriculture-

related sectors of food, beverages andtobacco, agriculture and hunting, textilesand fisheries have above average tariffs.

The incidence of core non-tariff measures (NTMs) for the APECeconomies halved over the 1995-1998

period. However, the NTMs in agricultureand hunting, and in chemicals went upslightly over the full period, those sectorsand are among the sectors with a high

incidence of NTMs, together with themanufacture of food, and forestry andlogging.

Thus, while the average use of tariffs and NTMs has decreased, everycountry has a sensitive sector or a sector itwishes to promote for various reasonswhich retains peak tariffs or core NTMs.

Not all NTMs are captured; for instance,tariff quotas are not included. This sector

typically not only has high tariffs but alsois subjected to other types of policyintervention. The automotive sector is agood example where tariffs are still highand there are often content requirementsstill in place.

Tariffs are the most transparentmeans of protection and, as is well known,as tariffs come down, non-tariff barriers or other measures emerge to the extent thatcountries wish to prolong the protection.Examples abound. The MalaysianGovernment has recently announced awide range of financial incentives to

promote its domestic automobile industry, justifying it on the grounds of stronglinkage effects (Mody, 1999).

Typically, there has also been acorresponding increase in anti-dumpingduties. Although, as is well recognized,the amount of trade subject to anti-dumping duties is not large, the simplethreat of duties being imposed on a

product is sufficient for there to be a

substantial impact on exports. There has been an increase in the number of anti-dumping actions by both developed anddeveloping economies since 1995, and in1998 an estimated 300 cases weresubmitted compared with 225 in 1997.

The evidence on the effectivenessof export subsidies and promotion is notconclusive; East Asia is often being citedas an example of their effectiveness.

Export promotion that uses incentives toencourage exports works in the same wayas import protection and can be subject tothe same abuses. Export subsidies can beabused through over-invoicing, falseshipments, and ad hoc subsidies that donot necessarily go to the most competitiveexporters. In the World Bank East AsianMiracle study (World Bank, 1992), therelative success of export promotioninterventions in East Asia is because with


21/50

17

exports, the criteria for performance wasmuch easier to measure. In studies thatlook at what determines exports, exportsubsidies showed little change in the years

prior to the boom of East Asian exports(Rodrik (1993), as quoted in Mody(1999)). Furthermore, some of the EastAsian economies, such as Indonesia, hadin fact moved away from using exportsubsidies since the mid-1980s.

Obviously, there are other factorsthat influence exports. Supportingmeasures that complement exportsubsidies could be important. In theRepublic of Korea, for instance, detailedsectoral and firm-specific export targetswere identified and given access to exportcredits. There were also benefitsassociated with agencies that developednew markets and testing and standardsorganization.

With regard to TRIMs, four members of the Association of South-EastAsian Nations (ASEAN) notified local

content policies in their automotivesectors to the WTO. 10 In addition,Indonesia submitted local contentrequirements for fresh milk and soybeancake production, and Thailand submittedlocal content requirements for manufactured goods. Notification meansthat they will phase out the measures infive years. However, there have beendivergences in application, as is evidentfrom the national automobile policy of

Indonesia described in the Annex.

Examination of rules under theWTO and the East Asian experience so far indicate that there are many instrumentsthat are no longer valid. However, thereare still instruments that could be used,

10 See the annex for a description of Indonesia’sWTO experience regarding its local content

policies in the automotive sector.

and the usual caveats about specificity of policy in relation to the objective or targetapply. Furthermore, the instrument must

be implemented in a transparent way, have built-in performance requirements andhave a clear exit point. Before the crisisthere were many derogations from such

basic concepts. These were compounded by a close relationship between businessand government in channelling capital andcredit to specific sectors. 11

D. Policies towards foreign investors

World FDI flows have grown atvery impressive rates. Between 1986 and1991 the annual average was US$ 159.3

billion. In 1999 the estimated total is US$800 billion, which is an increase of 25 per cent over the previous year. 12 While asignificant proportion of the increase isdue to the developed countries, amongstdeveloping countries East Asia’s

performance has been particularlynoteworthy. Between 1986 and 1997 nine

developing countries13

in the region haveaccounted for more than 50 per cent of total flows to developing countries. Theincrease in the importance of China as ahost country is particularly noteworthy.

The contribution of FDI flows todevelopment continues to be debated;however, the focus seems to be more ondegree. Without a doubt the access toexternal capital and the performance of

foreign affiliates in East Asian host 11 In the post-crisis era a number of issuesregarding corporate governance and restructuringare emerging as critical for a sustainable recovery.12 Revised figures and more detail can be found inUNCTAD (1999a) and at www.unctad.org.13 These are the four newly industrializingeconomies – Hong Kong (China), the Republic of Korea, Singapore and Taiwan Province of China – and China, Indonesia, Malaysia, the Philippinesand Thailand.


22/50

18

economies have had an impact on theregion’s economic growth. But thesuccess was achieved through a range of

policies aimed at attracting andchannelling FDI into specific sectors. 14

Consider first the attraction of FDI.The approach adopted by some economies

–Hong Kong (China) and Singapore – wasone of liberalization. By selectivelyreducing equity restrictions and

performance requirements they were ableto attract significant quantities of FDI.These policies were complemented by anextensive array of incentives, whichincluded duty-drawback provisions, taxholidays and subsidies. They wereimplemented in an aggressive and targetedmanner, so much so that their goal of industrial restructuring using FDI has byand large been successful. Even thesecond-tier Asian countries such asMalaysia and Indonesia have been able toincrease their manufacturing capacity,relative to primary industries, with someassistance from FDI. China in particular,

with its State-owned enterprise system, iswell aware of the role that foreignaffiliates can play in helping some of its

privatized industries to be competitive.

Each of the East Asian countries atone time or another made extensive use of

performance requirements. These includedlocal content, export performance targets,foreign exchange restrictions, licensingrestrictions, mandatory local participation

14 See Hill and Arthukorala (1998) for a recentsurvey of the contribution of FDI to developmentin the region.

and trade balancing. In the past decade,some countries have removed theserestrictions unilaterally, but others havechosen to retain certain policies. Theresult is a policy landscape for FDI in EastAsia that is rather mixed.

While all countries welcome FDI,there are still a range of impediments toflows and policies designed to alter

behaviour (PECC, forthcoming). Theseimpediments are distortionary and, asshown in section 3, can lead to welfarelosses even if certain non-economicobjectives are achieved. All the EastAsian countries are members of the APEC

process and have committed themselves to“free and open trade and investment” by2010 for developed members and 2020 for developing countries. Furthermore, themembers of ASEAN have signed aFramework Agreement for an ASEANInvestment Area, which includes anegative list liberalization of equityrestrictions by 2010 for members and2020 for non-members. These

arrangements would seem to indicate thatthere is much support for investmentliberalization, but some caution aboutincluding it in the WTO. The issue, then,is not liberalization, but to what extent

perceived gains can be achieved throughhaving binding rules on market access for foreign investors, including licensing,

joint ventures and performancerequirements. This is an issue taken up insection 5.


23/50

19

IV. WTO RULES AND INDUSTRIAL POLICY

A number of provisions in theWTO rules deal with various measuresthat member States can use to protectdomestic suppliers and promote exportsand technology transfer. Articles I and IIIof GATT 1994 lay down MFN andnational treatment for imported goods.However, up to the bound rate (if a tariff item has been bound), tariffs can still beused to protect infant industries anddevelop domestic capacity. Tariffs areoften complemented by other tools of

industrial policy such as subsidies, whichare used to both promote particular firmsand industries and to penetrate foreignmarkets. In this section we examine howthe WTO rules have constrained theflexibility of member States in the choiceof instruments that may be used to pursueindustrial policy objectives.

A. Import protection

Tariffs, non-tariff measures andsubsidies protect domestic firms fromimport competition. Although tariff

protection has declined, there continue to be peak tariffs in some industries in bothdeveloped and developing countries.Also, the dispersion of protection remainshigh in many countries. 15

One particular policy, which usedto be quite common, is local content

protection. This policy was the hallmark of a number of countries as they tried todevelop large-scale industries withexternalities. In particular, automotive

15 It is useful to distinguish between sunset andinfant industries. The former are industries that aredeclining. The latter are industries that areexpanding and, owing to market failures, require

protection from competition.

industries around the world have beenheavily dependent on local content

protection. 16 There was much discussionduring the Uruguay Round as to whether local content protection had an impact ontrade. Local content policies wereformally included in the MultilateralTrade Agreement (MTA) through theAgreement on Trade-related InvestmentMeasures (TRIMs). The effect of this isthat such policies have to be phased out byWTO members at the latest by 1 January

2002, unless an extension can begranted. 17

Import protection can also beachieved by challenging the fairness of thecompetition by using anti-dumping or safeguard measures. In the context of industry policy both measures have often

been used in declining industries. TheFinal Agreement on Implementation of Article VI had a few additional provisions

in favour of developing countries as theytry to develop their exports. Moreimportant, there are still systemic issues inthe implementation of Article VI that canfrustrate legitimate attempts by developingcountries to exploit export markets. 18

16

Australia, for example, had a national car programme in place immediately after the end of the Second World War. The first “Australian car”rolled off the production line in 1948. However,

protection in the industry only started to declinesubstantially in the early 1990s (Bora and Pomfret,1995).17 A WTO Dispute Panel decision on Indonesia’slocal content protection policies clearly stated thatlocal content policies disadvantage imports. Seethe Annex for more details.18 See Laird (1997) for a discussion of anti-dumping.


24/50

20

B. Subsidies and export promotion

The Agreement on Subsidies andCountervailing Measures (SCM) increasesdisciplines on the use of subsidies andcountervailing measures to offset anyinjury caused by subsidized imports. Itapplies to non-agricultural products; thereare separate (and more comprehensive)disciplines on agricultural products in theAgreement on Agriculture.

The SCM covers financialcontributions 19 made by or at the directionof a Government 20 that provides a

benefit. 21 It defines three areas of specificity which would bring a subsidyunder its rules. These are:

• Enterprise specificity – a particular company or companies is targeted;

• Industry specificity – a particular sector is targeted;

• Regional specificity – a particular region is targeted.

If a subsidy fits the specificitydefinition, it is placed in one of threecategories: prohibited, actionable or non-actionable. The prohibited categorycomprises subsidies for exports andsubsidies for the use of domestic in

preference to imported inputs. In additionto non-specific subsidies, the category of non-actionable subsidies includes thefollowing exemptions:

19 The Agreement contains a list of types of measures that would be considered to be financialcontributions: grants, loans, equity infusions, loanguarantees, fiscal incentives and the provision of goods and services.20 Since this is defined to include any public bodywithin the territory of a member, sub-nationalGovernments, public bodies and State-ownedcompanies are covered by the Agreement.21 The definition of a benefit has not been fullyresolved in cases where indirect financialcontributions are involved.

• Disadvantaged region initiatives;• Research and development;• Environment.

The bulk of the subsidies fit intothe categories of actionable and non-actionable rather than prohibitedsubsidies. Action under the SCM relies on

providing proof that subsidies are having anegative effect on the trade of another member. This is done by showing thatthere is harm to another member in theform of injury, 22 serious prejudice 23 or impairment 24 and nullification of benefits.Once this has been proved, the subsidy

must be removed or changed to conformto WTO regulations.

The implications of the SCM for industrial policy are considerable. 25 In thefirst instance, developing countries whichdid not sign the Subsidies Code during theTokyo Round are now bound by theAgreement. Second, the SCM nowextends to sub-national Governments.

However, the discipline onsubsidies based on production is weak. Itdoes not apply to agricultural and servicestrade. The main message of the SCM isthat subsidies that have a direct effect ontrade are explicitly prohibited. This rulesout the possibility of developing a sector

22 Injury to a domestic industry caused bysubsidized imports into the territory of thecomplaining member.23 Adverse effects in the market of the subsidizingmember or in a third market.24 This can arise when improved market access dueto a bound tariff reduction is undercut by thesubsidy.25 Take, for example, the case of the Republic of Korea, which has been notorious for its use of targeted subsidies. Prior to 1995 it had offered 26different types of subsidies with an annual total of 2.5 trillion won. In 1995 it reduced this to onesubsidy to small and medium-sized enterprises of only 15.2 billion won (WTO, 1996b).


25/50


26/50

22

a mechanism for promotingindustrialization were held to beincompatible with the rules of themultilateral trading system. This couldlimit attempts to build up domesticcapacity and increase the transmission anddiffusion of technology. 27

Another concern about restrictingTRIMs is the second-best argument that

policies such as export restrictions andlocal content are needed to defend againstanti-competitive practices (UNCTAD,1999b). This argument has somesupporters, but it raises a number of concerns, not the least of which isimplementation. The fundamental

problem is, as stated in section 2, that if itis anti-competitive practices that are theissue, effort needs to be directed towardsdeveloping national and multilateralcompetition rules.

D. Agreement on Trade-relatedAspects of Intellectual Property

RightsThe value of the Agreement on

Trade-related Aspects of IntellectualProperty Rights (TRIPS) for developingcountries continues to be hotly debated.The Agreement consists of three parts:standards, enforcement and disputesettlement. It involved, perhaps more thanany of the other Agreements, substantialchanges in national legislation. These

changes are designed to strengthen the protection of intellectual property rights(IPR) and to have a positive impact onlocal innovation, FDI and technologytransfer. However, at the same time anumber of negative impacts, at least as far

27 Moran (1998) cites evidence of the policyhaving a positive impact on the development of local capacity. In doing so, however, he does notmention the costs of the policies.

as developing countries are concerned,were predicted (UNCTAD, 1996). Theseincluded higher prices for protectedtechnologies and products, and restricted

possibilities for diffusion through reverseengineering. Also, new legislation indeveloping countries required further examination of the balance between thedegree of protection required for innovation and the restricted diffusion of technologies.

The strengthened protection hasimplications for industrial policyobjectives. In the case of domestic firmsit means that there is an incentive toinnovate and compete dynamically. TheSCM allows R&D subsidies, and theoutput of this process can then be

protected through the TRIPS Agreement.For foreign firms it means that, where

permitted, market access through acommercial presence may now be viablesince they have better IPR protection.Developing countries do not in generalhave a comparative advantage in

innovation. Therefore, attempts todevelop certain sectors within the contextof the WTO mean that that they will haveto rely heavily on the transfer anddiffusion of technology from foreigncountries instead of on domesticinnovation.

An important provision for developing countries is Article 66.2,which requires developed countries to

provide incentives to promote technologytransfer to least developed countries. Sofar, little is known about the extent towhich this provision has beenimplemented (UNCTAD, 1999b). It isaccompanied by a transition periodallowance for developing countries.


27/50

23

E. General Agreement on Trade inServices

The legally enforceable rulescovering the international trade in servicescontained in the General Agreement onTrade in Services (GATS) can also affectindustrial policy initiatives. Relative todisciplines on goods trade, the Agreementas a whole is much less effective in termsof liberalization. It contains a positivelisting of sectoral commitments on marketaccess and national treatment. It allowssectoral bindings on four modes of supply:cross-border supply, consumption abroad,commercial presence and presence of natural persons.

Through the inclusion of commercial presence as a mode of supply,rules on foreign investment in services arenow part of the multilateral tradingsystem. Members, therefore, can useforeign investment liberalization as tool of industrial policy. This has happened to

some extent with bindings in tourism, butnot in other sectors.

As with other forms of liberalization, the effect of the GATS istwofold. First, market access makes it

possible to develop export sectors.Second, bindings have the effect of inducing competition in home markets.Developing countries have an exportinterest in a limited range of sectors such

as tourism and professional services. Inthese and other cases a key issue is themovement of natural persons (or mode 4).Horizontal barriers in this area make itdifficult for developing countries to buildexport competitiveness in their comparative advantage areas.

The competition effects of liberalization in the area of infrastructureare extremely important for developing

countries. As discussed above, disciplinesin the area of subsidies and performancerequirements are forcing developingcountries to think of more neutral ways todevelop export capacity (Laird, 1997).One of these is infrastructure and, in

particular, telecommunication, financialand transport services.

In addition to market accesscommitments, the GATS has a provisionrelated to performance requirements.However, the prohibition of thesemeasures relates only to a sector that has

been inscribed. This means thatdeveloping countries that have taken theview that performance requirements needto be maintained in service industries, as aquid pro quo for liberalization, cannotmaintain them. It is important to note theconsistency of this provision with theSCM, where the quid pro quo argumenthas also been ignored.

F. Special and differential treatment

In GATT 1947 Article XVIII madespecific mention of contracting parties intheir “early stages of development” andallowed them to “maintain sufficientflexibility in their tariff structure”.Section B of that Article provideddeveloping countries with flexibility toimpose trade measures to protect their

balance of payments, and Section Cenabled them to take such measures for

the protection of an infant industry. In1966, Part IV was added to the GATT,and the Tokyo Round adopted an enablingclause which gave “special anddifferential” (S&D) treatment todeveloping countries. The concept of S&D allowed non-reciprocal tariff

preferences as implemented throughschemes, under the Generalized System of Preferences (GSP).


28/50

24

Prior to the Uruguay Round, littleuse had been made of Article XVIII,section C (infant industry), because theuse of this provision requires the paymentof compensation. As a result, since 1967no country has specifically invoked it.Instead, numerous countries have madeuse of section B (protection for balance-of-payments reasons), which does notrequire compensation.

Infant industry protection, byinvoking Article XVIII for infant industryor balance-of-payments protection, is still

possible under the WTO, but the new procedures for balance-of-paymentsconsultation are likely to constrain thescope and duration of such exceptions(Singh , 1996: 166). Furthermore, duringthe Uruguay Round a new approach toS&D developed, which essentiallyamounts to allowing for flexibility innominating sectors for liberalization andin most cases an extended transition

period for meeting obligations under theAgreements (Youssef, 1999).

The scope for S&D treatment inthe application of industrial policy existsin each of the Agreements referred toabove, for example:

• Delays in implementation (TRIMs,SCM, Safeguards Agreement,TRIPS);

• Preferential disciplines (SCM,Safeguards Agreement);

• Flexibility to increase protection(temporarily) (Article XVIII);

• Flexibility in approachingliberalization (GATS).

There is mounting concern on the part of developing countries that these provisions do not promote their interests(UNCTAD, 1999b). Two issues need to

be distinguished. The first is the existence

of S&D provisions and second is their relevance. Take, for example, localcontent protection. The difficulties faced

by developing countries in making thetransition from this instrument arerecognized by allowing them a longer

period. However, some countries arguethat this allowance is not significant.

G. Implications of WTO rules forindustrial policy

The above review has identified anumber of WTO rules that discipline theuse of government intervention to promote

particular industries. Different countrieshave different objectives, and wouldtherefore require different sets of policytools. As a result, the impact of the WTOrules on countries would differ accordingly. There are, however, somecommon features of the Agreements,which deserve to be highlighted.

First, each of the Agreements takes

a trade, not a balance-of-payments,financing approach to disciplining policies. Since Articles I and III are thecornerstones of the rules-based system,any non-border policy that has an effecton the trade in goods and services is under discipline or has had an exemptionnegotiated.

Second, the rules are ownership-neutral. Apart from the GATS and TRIPS,

where a national treatment standard isapplied, policies such as subsidies andlocal content protection do not distinguish

between foreign affiliates and domesticenterprises. What is important is the“trade effect” of the instrument. Thismeans that countries seeking to apply a

particular policy to foreign-owned firmsmust first find a provision in theAgreement that allows the use of the


29/50


30/50

26

V. IMPLICATIONS FOR DEVELOPING COUNTRIES

The present WTO rules haverestricted the industrial policy instrumentsavailable to WTO members, as discussedabove. The use of border measures hasdeclined, and the disciplines on the use of subsidies together with contingent

protection and intellectual property ruleshave been strengthened. The direction of any revision of WTO rules depends on thenet assessment of these changes. On theone hand, there is evidence to show that anumber of policies that distort trade are

still allowed under existing rules. On theother hand, the added discipline imposed by the WTO rules has reduced theflexibility of national Governments to

pursue development objectives. In thissection we examine some of the issuesarising from the possible revision of WTOrules as they relate to the pursuit of industrial policy objectives by developingcountries.

A. Import protection

The scope for import protectioncontinues to diminish. Tariffs aredeclining, local content protection is onthe verge of being prohibited andcontingent protection is now somewhatmore disciplined. What scope is there for continued import protection? The answer is that it depends on the ability of developing countries to negotiate a

provision that will allow for greater discretion for protection.

It is important to distinguish between import protection for the purposes of protecting a “sunset” or declining industry, and protection to

promote an infant industry or newlyexpanding industry, which is being

protected because of some perceived

externality. Efforts to develop transparentand objective rules for both types of

protection are needed.

However, the issue related toindustrial policy that we wish to addresshere concerns local content protection andrules of origin in preferential tradingagreements. These are prevalent only inlarge-scale industries where there is scopefor significant linkages with domesticindustries. 28 There is evidence that this

policy has been proved to be successful inestablishing some industries in somecountries. Australia, for example, usedlocal content policies in the establishmentof some manufacturing industries.However, it was an inefficient industry.The policy has recently been abandonedand tariff rates have been reduced (Boraand Pomfret, 1995). The result is anindustry that is more competitive, albeitafter some structural adjustment. This

abandonment vindicates those who arguethat the policies are not required for efficient competitive industries (Pursell,1999).

However, it is arguable that acontent-protected industry would not existat all if local content policies had not beenused in the first place. This appears to bethe position of the developing countries inthis regard – that as newly industrializingeconomies they have not had the 40-year grace period that, say, the Australianautomotive industry had, nor have theyhad the degree of protection and market

28 However, Barbados, Colombia, Cyprus, India,Indonesia, Pakistan, Peru, Romania, South Africa,Thailand and Uganda notified local content in theagriculture industries (Ministry for Trade andIndustry, 1999).


31/50

27

access afforded to Canada through theCanada–United States Autopact.

Given these considerations, what isthe best way to proceed? If developingcountries have as their objective efficientnational production, content plans distortthe production within an industry andthereby lead to inefficiencies, as noted insection 2. If there is a dynamic learning or cost reduction process, a tariff or subsidytemporarily assisting the processes inwhich the cost reduction occurs is much

preferable to a continuation of content plans for the whole industry.

If developing countries have astheir development objective somethingother than efficiency, as in the case of theIndonesian automobile industry (seeAnnex), or as argued by Venezuela, 29

there will be an inconsistency with thefundamental rules of the trading system.For example, in the automobile case, whena completely built-up unit is imported for assembly it will usually be feasible to

source at least 20 per cent of the productslocally. 30 This means that in order toincrease local content further imports willhave to be displaced. Here, one mustquestion whether the developmentobjective is compatible with efficientlong-run allocation of resources, or

29 Venezuela argues that there are policies which“not only induce growth of their traditional flows,

but also promote the structural transformation of their economies and the possibility to add morevalue to their exports”. It argues that they are“development policy issues” and that it may be

possible to identify instruments, which could beused to promote development but which areconsistent with the principles of non-discrimination (WT/GC/W/279, 29 July, 1999).30 These would be generic components for whichtransport costs would be high, such as tyres andoil. See the references in the Annex for details of the cost structure in the Indonesian automobilescase.

whether it favours some group at theexpense of national development.

B. Export promotion

As has been seen, directintervention by Governments to boostexports is being increasingly restricted bythe WTO rules. This leaves developingcountries little room for manoeuvre in thearea of export subsidies for industrial

products. Nevertheless, there is a widerange of alternatives that are still pursued

by Governments. These include exportcredit and insurance schemes belowmarket rates, concessional tax and duty

provisions and export processing zones.While some of these remain WTO-consistent, developing countries need toreassess the extent to which other policieswhich discriminate in favour of

industrial policy and the wto

Documents