D. Trade openness and the broader socio-economic context · the broader socio-economic context. ... necessarily reflected in this trend. In Sub-Saharan Africa, for instance, Nigeria

world trade report 2013

220

Section C showed how fundamental economic factors – demography, investment, technology, natural resources, transportation and institutions – can affect the future of trade. But trade takes place within a broader socio-economic context. This context matters for trade and trade policy. Historically, social and macroeconomic concerns have repeatedly influenced decisions in trade policy matters. Section B of this report provided examples of such situations. Both themes are currently high on the political agenda and will undoubtedly affect policy-makers’ views and positions in the area of trade reform in the future. A third factor relates to environmental concerns, an issue that has rapidly been gaining prominence in the national, regional and global policy debate. It has also been repeatedly linked to trade, notably in the context of a number of high-profile WTO disputes, in the context of regional trade agreements and as an element of the on-going Doha Development Agenda.

D. Trade openness and the broader socio-economic context

II – Factors shapIng the Future oF world trade

221

II D. Tr

aD

e o

pe

nn

es

s a

nD

Th

e b

ro

aD

er

so

cIo

-e

co

no

mIc

co

nTe

xT

Contents 1 Socialconcerns:inequalityandunemployment 222

2 Environmentalconcerns 237

3 Macroeconomicandfinancialconcerns 251

4 Conclusions 262

Some key facts and findings

• Successful integration into global markets requires the constant need for individuals and societies to cope with changes in the competitive environment. These adjustments can put labour markets under strain and can shape attitudes towards trade openness. Economies with a well-trained workforce and a business-friendly environment tend to be better placed to adjust successfully.

• Societies’ transition to a sustainable development path requires careful management of the multi-faceted relationship between trade and the environment in order to avoid “green protectionism” and to maximize the environmental benefits that open trade can bring.

• The expansion of trade needs to be supported by a stable financial and monetary system – delivering a sufficient volume of trade finance at an affordable cost, particularly for developing countries, and macroeconomic policies that promote exchange rate stability.


222

1. Socialconcerns:inequalityandunemployment

Employmenthasbeenhighonpolicy-makers’agendasin recent years. This is the case for countries acrossdifferent income levels although the reasons forconcern differ. In many industrialized countries,unemploymenthassoaredduringtherecentcrisisandin some a recovery of the labour market is not yet insight.1 Emerging economies with large populations,suchasChinaandIndia,donotfinditstraightforwardto absorb large numbers of rural workers into formallabour markets, even when the economy is booming.The important role of jobs for economic and socialdevelopment in least-developedcountries (LDCs)hasbeenhighlightedinthemostrecentWorld Development Report(WorldBank,2012b).

Incomes of those who do work are also a matter ofconcern, especially in light of increasing incomeinequalitywithincountries.Inanumberofindustrializedcountries,incomeinequality,measuredintermsoftheshare in total wealth of the 1 per cent wealthiestindividuals, is close to the levels prevalent in the1920s,andmorethandoublethelevelsofthe1970s.In many middle-income countries, income inequalityhasalsoincreasedsharplysincetheearly1990s.

Asaresultofthesedevelopments,policyreformsneedto perform well on the employment and distributionalfrontinordertoobtainpublicsupport.Thissub-sectionprovides an overview of the patterns of inequalitywithin and across countries, and of unemploymentlevels across countries. It then proceeds to discusswhether,andtowhatextent,tradehasplayedarolein

driving theobservedpatterns.Lastly, this sectionwilldiscusswhether theobserved labourmarketpatternsarelikelytoaffectattitudestowardstradeopennessorits effects. The sub-section concludes by venturinginto a discussion of the expected labour marketchallengesthatdifferentcountrieswillfaceinthenearfuture and how those challenges may relate to theirtradeperformance.

(a) Incomedistributionandunemployment:recenttrends

Thetwodecadesprecedingtherecenteconomiccrisiswere characterized by significant increases in tradeandcapitalflows.Incomeinequalityincreasedinmostcountriesandregionsduringthesameperiod.Dataonthelong-runevolutionofinequalityindicatethattherehasbeena clear change in the late1980sandearly1990s.

Figure D.1 illustrates that in a set of countries calledthe “U-shaped” group by Atkinson et al. (2011),inequality–measuredasthepercentageshareofthe1 per cent richest households in total wealth – hasrisen quite dramatically in recent years and hasachievedpostFirstWorldWar levels. It shows that intheUnitedStates,therichest1percentofhouseholdsheld 19.6 per cent of national wealth in 1928. Thatshare fell to a low of 7.7 per cent in 1973 and thensteadily increased again to reach 18.3 per cent in2007,i.e.beforethestartoftheGreatRecession.Theshare of income of the wealthiest households fellduringtheRecessionbutisnowagainontheincrease.

Theevolutionof inequality followsasimilarpattern inthe other countries illustrated in Figure D.1. In the

FigureD.1:Share of the 1 per cent richest households in total wealth: “U-shaped” countries, 1910-2010

United States Canada Australia New Zealand United Kingdom Ireland

0

5

Top

perc

entil

e sh

are

19

10

19

15

19

20

19

25

19

30

19

35

19

40

19

45

19

50

19

55

19

60

19

65

19

70

19

75

19

80

19

85

19

90

19

95

20

00

20

05

20

10

10

15

20

25

Source:WorldTopIncomeDatabase.


223

II D. Tr

aD

e o

pe

nn

es

s a

nD

Th

e b

ro

aD

er

so

cIo

-e

co

no

mIc

co

nTe

xT

United Kingdom, inequality was higher than in theUnited States in the few years for which data areavailable in the inter-war period and just followingthe Second World War. Inequality then fell below6percenttowardstheendofthe1970sandsteadilyincreased thereafter to reach 15.4 per cent in 2007.Atkinsonetal.(2011)identifyanothersetofcountrieswith an L-shaped evolution of inequality. Thesecountries, includingGermany,FranceandJapanwerecharacterized by very high levels of inequality in theinter-war period. Inequality dropped sharply after theSecond World War and remained constant thereafteruntil the second half of the 1990s, when the incomeshareofthetop1percentofincomeearnersstartedto increase, although significantly less than incountriesillustratedinFigureD.1.

Figure D.2 reflects this evolution of inequality forJapanandthreeotherAsianeconomiesforwhichdataareavailable.Inallfourcountries,inequalitystartedtoincrease in the 1990s, with the increase beingsharpestinSingapore.Therichest1percentinChina,India, Japan and Singapore, however, own a smallershare of national income than their counterparts inCanada, Ireland, the United Kingdom or the UnitedStates.2

AnotherfrequentlyusedvariabletomeasureinequalityistheGinicoefficient(explainedinSectionB.2).Usinginformation on Gini coefficients during the pastdecade,FigureD.3indicatesthatinequalityishighestin much of South America and Sub-Saharan Africa.Both Brazil and South Africa have Gini coefficientsabove50percent.ChinaandtheRussianFederationfall into the40-49percent range.TheUnitedStatesfalls into that same group. India’s Gini coefficient is

lower and falls into the 30-39 per cent group.FigureD.3alsoshowsthatmanyofthecountrieswithvery low inequality, i.e.Ginicoefficientsbelow30percent, can be found in Europe, e.g. Germany and theScandinaviancountries.

Income distribution within countries is expected toundergofurtherchangesinthenearfuture.Oneofthemost important trends affecting future incomeinequalityisthechangeinthesizeofthemiddleclass.While the middle class is expected to increase andbecomericherinemergingeconomies,notablyinAsia,itappearstobeshrinkingintheUnitedStatesandtheEuropeanUnion(seeSectionC.1).

A phenomenon that emerged during the economiccrisis and that remains a challenge in many high-income countries is increased unemployment. TheInternational Labour Office (ILO, 2012) highlights ageneraldividebetweenthedevelopedanddevelopingregions,withunemploymentratesremainingfarabovehistorical averages in a group of countries that theycall “Developed Economies and the European Unionregion” (8.6 per cent in 2012 versus an average of6.9 per cent between 1998 and 2007), whileunemployment rates in 2012 were below historicalaveragesinmostdevelopingregions.

These patterns are reflected in Table D.1, whichshows unemployment rates in 2007 and 2010 for aselectednumberofcountries.Itillustratesthatrecentincreases in unemployment have been sharpest inindustrializedcountries,withanumberofcountriesinEurope and North America experiencingunemployment increasesabove4percentagepoints.However,thisphenomenoncannotbegeneralized.In

FigureD.2:Share of income of top 1 per cent income earners in selected Asian countries, 1922-2010

China India Japan Singapore

0

5

19

22

19

26

19

30

19

34

19

38

19

42

19

46

19

50

Per

cen

t

19

54

19

58

19

62

19

66

19

70

19

70

19

74

19

78

19

82

19

86

19

90

19

94

20

10

10

15

20

25

19

98

20

02

20

06

Source:WorldTopIncomeDatabase.


224

FigureD.3:World inequality map based on Gini coefficients

Above 60 50-59 40-49 30-39 Less than 30 Missing

Sources:HumanDevelopmentIndexdatabase2011andCIAWorld Fact Book ,2011.

Note:Valuesrepresentthemostrecentdatapointavailableintheperiod2000-11.TheGinicoefficienttakesvaluesbetween0and100inthischart,withhighervaluesreflectinghigherlevelsofinequality.

Anotherelementthatislikelytoaffecttradeflowsandpoliciesinthenearfutureisthedistributionofincomeacross countries, notably because relative cross-country income levels will determine what countriesconsume and what they produce. Section B.2 of thisreporthasdiscussed thephenomenonof “newglobalplayers”. This group of countries is commonlyconsidered to include the BRICS (Brazil, the RussianFederation, India, China and South Africa), most ofwhichhaveexperiencedhigh levelsof incomegrowthinthepasttwodecades.

Averageannualrealgrossdomesticproduct(GDP)percapita growth was above 10 per cent in China from1990 to 2011 and above 6 per cent in India in thesameperiod.GDPpercapitagrowthwasmoremodestbutsignificantinBrazil(2.8percent)andSouthAfrica(2.6 per cent), while it was relatively sluggish in theRussian Federation (0.7 per cent). This reflects acertain level of “catching up” in terms of GDP percapitabetween fourof theBRICSand thewealthiestregionsintheworld,giventhatGDPpercapitagrowthintheUnitedStateswas2.4percent,intheEuroarea1.7percentandinJapan1.1percentduringthesameperiod.3 This, together with the growth of the middleclassobservedinanumberofemergingeconomies,isprobablyoneofthemaindriversbehindthefindinginMilanovic (2012) that global income inequality hasdecreased in recent decades (i.e. in the period1988-2008). Low-incomecountries are, however, notnecessarily reflected in this trend. In Sub-SaharanAfrica,forinstance,Nigeriaexperiencedagrowthrateof close to 5 per cent, while countries such as the

Poland, for instance, unemployment remainedunchanged, while it decreased in Germany. Otherdeveloped countries, such as Japan, Mexico, theRepublicofKoreaandTurkey,experiencedmoderateincreases inunemployment.Unemploymentdataareonly available for a small number of developingcountries.TableD.1illustratesthatcountriessuchasColombia, Indonesia or the Philippines experiencedstableunemploymentratesorevenreductionsinthe2007-10period.

TableD.1:Levels and changes in unemployment, 2007-2010, selected countries (percentage)

2007 2010 Difference

Lithuania 4.3 17.8 13.5

Spain 8.3 20.1 11.8

Ireland 4.6 13.5 8.9

UnitedStates 4.6 9.6 5.0

Greece 8.3 12.5 4.2

Mexico 3.4 5.3 1.9

Turkey 10.3 11.9 1.6

RussianFederation 6.1 7.5 1.4

Japan 3.9 5.0 1.1

Korea,Rep.of 3.2 3.7 0.5

Poland 9.6 9.6 0.0

Philippines 7.4 7.4 0.0

Colombia 12.0 11.6 -0.4

Germany 8.6 7.1 -1.5

Indonesia 9.1 7.1 -2.0

Source:WorldBank.


225

II D. Tr

aD

e o

pe

nn

es

s a

nD

Th

e b

ro

aD

er

so

cIo

-e

co

no

mIc

co

nTe

xT

Democratic Republic of the Congo or ZimbabweexperiencednegativeGDPpercapitagrowthbetween1990 and 2011. The latter two countries and otherswith similar recent growth patterns therefore run therisk of “staying behind” while the rest of the worldgrowsmorecloselytogether.

Notwithstanding the above phenomenon of “catchingup”, the distribution of income remains very unequalacrosscountries,asillustratedinFigureD.4.GDPpercapitainCanada,theUnitedStates,Europe,AustraliaandJapancontinues tosignificantlyexceedGDPpercapitainSouthAmerica,AfricaandmostofAsia.Thisdifference will have an effect on comparativeadvantageandthusthelabourmarketeffectsoftrade.

Poorercountrieswithalargelabourforcearelikelytohave a comparative advantage in labour-intensivegoods and services. Countries such as China, Indiaand possibly also Indonesia and Mexico are likely tobasetheirexportsonanadvantageinlabourcosts.Asillustrated in Figure D.4, Indonesia’s labour force isroughlysimilarinsizetothelabourforceintheUnitedStates. The same is true for Mexico’s labour forcewhen compared to that of Germany. Yet, wages inGermanyaremorethansixtimestheaveragewagesinMexico,whileUSwagesexceedIndonesia’sbyafactorof20.AveragewagesinChinaandIndiacurrentlyalsoexceed those in Indonesiabut are significantly belowthoseinMexicoandwellbelowwagesinGermanyandtheUnitedStates.At the same time,Chinaand Indiaarecharacterizedbyavery large labour force thatbyfaroutweighsthoseofothercountries.

Although theproductivity levelsandskill compositionofthelabourforcediffersignificantlyacrosscountries,Figure D.4 suggests that the labour cost advantagemayremaininlowandmiddle-incomecountries–and,inparticular,Asianones–forawhile,evenifwagesincountriessuchasChinaareontheincrease(e.g.Lietal.,2012).Someofthefuturelabourmarketchallengesthat countries at different income levels are likely tofaceinthecontextofincreasedglobalintegrationarediscussedinmoredetailbelow.

(b) Tradeandlabourmarkets:atwo-wayrelationship?

Theeconomic literatureon the labourmarketeffectsoftradereformhastraditionallyfocusedontheeffectsof trade on relative factor incomes. A significantnumberofstudies inthe1980sand1990sexaminedthe impact of trade on the relative wage of high- tolow-skilled labour, arguably driven by the observationthat the relative wage of the highly skilled wasincreasing in a number of industrialized countries. Adifferent strandof literature focusedon theeffect oftradereformonunemploymentlevels. Inrecentyears,researchers have analysed the relationship betweenglobalization and the wage share of GDP, a measurethat has the advantage of combining information onwage levels and job numbers.4 Neither of thesemeasures is likely to captureeverything that isgoingon in labourmarketsasa resultof trade reform,5buttogethertheycanprovideageneralpictureofthemainmechanismsatwork.

FigureD.4:GDP per capita across countries, 2008 (US$atmarketexchangerates)

above 80,000 ($) 60,000 - 80,000 ($) 50,000 - 59,999 ($) 40,000 - 49,999 ($) 30,000 - 39,000 ($)

20,000 - 29,000 ($) 10,000 - 19,999 ($) 1,000 - 9,999 ($) less than 999 ($)

0

0 1,000 2,000 3,000

0 1,000 2,000 3,000

0 1,000 2,000 3,000

MexicoGermany

MexicoGermany

IndiaChina

IndonesiaUS

IndonesiaUS

IndiaChina

500 1,000

0 500 1,000

0 500 1,000

Labour force (in millions)

Labour force(in millions)

Labour force(in millions)Nominal wage (US$)

Nominal wage (US$)

Nominal wage (US$)

Source:YearlyaveragenominalmonthlywagesfromGlobalWageDatabase,ILO;exchangeratesfromInternational Financial Statistics ,InternationalMonetaryFund;labourforceparticipationratefor15to64year-olds,totalpopulationandpercapitaGDPfromtheWorldBank.Alldataarefrom2008,thelatestyearforwhichwagedatawereavailable.


226

(i) Does trade affect inequality within countries?

Much has been written on the relationship betweentradereformand incomedistributionwithincountries.The trade literature has focused on a variety ofquestions.Whatarethechannelsthroughwhichtradeaffects income distribution within countries? What isthe relative importance of the effect of trade onincome distribution when compared with otherpossible determinants of distribution? To what extentdoestheeffectoftradeonincomedistributiondependoncountrycharacteristics?

According to traditional trade theory based oncomparativeadvantage, a countryexportsgoods thatare intensive in the use of its relatively “abundant”factor and imports goods that intensively use itsrelatively “scarce” factor. This implies that tradeopening would increase demand for the abundantfactor relative to the scarce factor. Hence, trade isexpected to reduce income inequality in developingcountriesbycreatingnew jobsand increasingwagesfor unskilled labour but increase income inequality inindustrializedcountriesviaa lossof jobsforunskilledlabour due to imports from more cost-efficientproducersoverseas.

A number of empirical studies have analysed therelationship between trade and wage differences.6

Evidence on the determinants of change in wagesindicates that trade isonlyoneofmanydeterminantsof wage inequality. Other determinants includetechnologicalchange,de-unionization,anerodingrealminimum wage and changing consumer tastes. Ingeneral, skill-biased technologicalchange is found tobethemaindeterminantofchangesinwageinequality,whiletradeisonlyfoundtoaccountforaminorshare.

Morerecentliteraturedeviatesfromtraditionalwaysofanalysing the trade-inequality nexus in a number ofways.First,thereisanincreasinginterestinexaminingthe determinants of the evolution of top incomes, i.e.themeasureofinequalitydescribedinFiguresD.1andD.2. Theoretical contributions analysing thisphenomenon refer to the concept of “super-stareconomics” (Rosen, 1981) and the notion that onlyvery limitednumbersof individualsreapa largeshareof the gains from increases in market size (ManasseandTurrini,2001;CostinotandVogel,2010).

Atkinsonetal. (2011)conjecture that recentchangesin the evolution of the income share of top-incomeearnersmaybe linkedto thefact that “theexpansionof scale associated with globalization and withincreasedcommunicationopportunitieshasraisedtherentsofthosewiththeveryhighestabilities”.Haskeletal. (2012) propose a theoretical framework thatexplains how the combination of globalization andinnovation can end up boosting the real and relativeearningsof“superstars”.

This lastpointhintsat theseconddeviation from thetraditionalanalysesoftheglobalizationand inequalitynexus. While the early empirical literature tried toseparate the distributional effects of trade from theeffects of technological change, the more recentliterature focuses on the fact that trade andtechnologicalchangemaygohandinhand.Oneoftheimplications of this literature is that trade affectsincome distribution through its effect on technologychoice and productivity. It also implies that positivegrowth effects of trade may be more systematicallyaccompanied by inequality increases than thought inthe past, and that development of new activitiesbecomes increasingly important to guaranteesufficient jobcreationandavoid joblessgrowth traps(e.g. Burstein et al., 2011; Newfarmer andSztajerowska,2012).

Recentresearchhasalsobeguntofocusondifferentforms of inequality, notably the relationship betweentrade and increased wage inequality across firms forworkers with otherwise similar characteristics (e.g.AmitiandDavis,2011;Friasetal.,2012;Krishnaetal.,2011). One of the findings of the research is thatworkers in exporting firms earn more than workerswithotherwisesimilarcharacteristicsinnon-exportingfirms.7 This suggests that firms, which adjustsuccessfully to globalization, pay higher wages andoffer better working conditions (Newfarmer andSztajerowska,2012).

Finally, there is increased interest in how othercomponents of globalization affect the distribution ofincome within countries. Foreign financial flows havebeenidentifiedasapossibledeterminantofinequalitythatmayoftenactincombinationwithtradeflows.8

FDI inflows may, for instance, increase inequality inlow-income countries because they increase therelative demand for skilled labour (Feenstra andHanson,1997).Increasedmobilityofcapitalthatisnotmatchedbyasimilarincreaseingloballabourmobilityis also likely to have an impact on the relativebargaining power of workers and capital owners andon taxation choices. Both have an effect on incomedistribution within countries (e.g. Boix, 2011). Recentempirical studies have found that increases in globalfinancialflowshavecontributedtoincreasedinequality(e.g. ILO, 2011; IMF, 2007; Jayadev, 2007; OECD,2011).

Overall, therefore, evidence seems to indicate thattrade in goods and services is unlikely to have had asignificantimpactoninequalitythroughthetraditionalchannels of shifting relative demand for productionfactors (Haskel et al., 2012). However, there is someevidence that trade, combined with technologicalchange or with FDI, can significantly affect incomedistributionwithincountries.Atthesametime,allthreefactors are among the major drivers of economicgrowth.


227

II D. Tr

aD

e o

pe

nn

es

s a

nD

Th

e b

ro

aD

er

so

cIo

-e

co

no

mIc

co

nTe

xT

(ii) Is trade openness a determinant of unemployment levels?

Tradeopeningtriggerseconomicchange.Itallowswell-performingfirmstoexpandtheiractivitiesandtoexport.It also puts under-performing firms under increasedcompetitive pressure from abroad. As a consequence,those firms may shrink or even close down. Thesechanges brought about by trade opening are thus aninevitableanddesirablepartof theprocess that leadsto improvements in economic performance andsubsequently to increases in wealth. However, duringtheperiodofchange,jobsarecreatedinsomepartsofthe economy and lost in others. Policy-makers,therefore, often consider such periods of transition ascritical.9Theyarealsoconcernedabouttheemploymentimplicationsofchangesinthecompetitiveenvironmentoncetheireconomyisopen.ThisisreflectedinthefactthatWTOagreementscontainsafeguardmeasuresthatallowgovernmentstointerveneundercertainconditionsif unexpected surges in imports have significantnegativeemploymenteffects.10

Economic research provides policy-makers withinformationonthedirectionofchangefollowingtradeopening.Inparticular,itprovidesinformationonwhichparts of the economy are most likely to suffer joblosses. In traditional trade theory, the reshuffling ofresourcesandeconomicactivitywasexpectedtotakeplace across sectors, with jobs being created inexporting sectors andbeing lost in import-competingsectors.Morerecenttrademodelsshowthatfirm-leveladjustment following trade opening leads to jobcreationandjoblossinallsectors,duetothefactthathigh-productivityfirmsfarebetterinbothnet-exportingandnet-importingsectors,whilelow-productivityfirmsfare worse (e.g. Bernard et al., 2007). The latterimplies that part of the adjustment process followingtradeopeningtakesplacewithinsectors(JansenandLee, 2007), which probably makes adjustment easierthancross-sectoraladjustment.

In general, the focus of economic research in recentdecades has not been on understanding theadjustmentprocessfollowingtradeopeningbutratheronassessingwhethertradeopeninghasaneffectonlong-run unemployment rates. In this context, it isworthnotingthat tradeopeningwouldhavenoeffectonunemploymentratesifmarkets–particularlylabourmarkets – function smoothly. In theoretical modelsanalysing the relationship between trade andunemployment,economistsassumeimperfectmarketswhere wages do not reach the market-clearing level.This may be because minimum wages put a lowerbound on wage levels (e.g. Brecher, 1974; Davis,1998),becauseworkers’effortsonthejobdependonwhethertheyconsiderwagestobefair(e.g.EggerandKreickemeier,2009),11orbecauselabourmarketsarecharacterized by matching or search frictions (e.g.Jansen and Turrini, 2004; Helpman et al., 2010;Felbermayretal.,2011b).12

The impact of trade opening on unemployment willdepend on whether increased trade is expected toexacerbate the impact of already existing frictions orwhether it is expected to reduce the strain on aparticular friction. If, for instance, trade opening putspressureonwagesofworkersthatarealreadyworkingatminimumwagelevels, increasedforeigncompetitioncan lead tohigher unemployment if demand for thoseworkers is reducedfurther. If,on theotherhand, tradeopening allows companies to take advantage of scaleeconomies, minimum wages and search frictions willbecome less binding and unemployment levels will godowninthelong-run.Astradereformandopennessareexpectedtotriggeracombinationofdifferenteffects,itis difficult to predict the effect of trade on long-runemploymentonthebasisofeconomictheoryalone.Theempiricalliteratureprovidesmoreclear-cutinsightsintothe long-runeffectsof tradeonunemployment,aswillbehighlightedlaterinthissection.

One reasonwhy theoretical researchhas focusedonthe long-run rather than the short- to medium-runeffects of changes in trade flows may be thateconomistsexpectadjustmentphasestobeshortandnotverycostly.Earlystudiesattemptingtoassesstheeconomic costs of adjustment following trade reformconcludedthatthosecostswerelowandaround5percent of the total benefits of trade (Magee, 1972;Baldwinetal.,1980).

It is therefore reasonable to assume that changes intradeflowsdonotnecessarilyhaveasignificantimpacton macroeconomic measures, such as the overallemployment rates, when the value of trade is smallcompared with the size of the overall economy. In acountrysuchastheUnitedStates,theratioofimportstoGDPwasaround15percentintheyearsprecedingtheeconomiccrisis.Whenmeasuredintermsofvalueadded,i.e.ifonlytheforeignvalueaddedembodiedinimports is taken into account, imports representedless than 14 per cent of the United States’ GDP in2008andaround11percentin2009.13

Figure D.5, however, suggests that an increase incompetitionmayaffectacountry’seconomicstructure,even in economies as large as the United States. Itreflectsstructuralchange,measuredby theso-calledstructuralchangeindex(SCI),whichcaptureschangesin the relative size of sectors. The index ranges fromzero to 100, with higher values reflecting moreimportant changes in the relative size of sectors,changes that typically go hand in hand with thereshufflingof resources.The individualbars inFigureD.5 illustrate the extent to which the economicstructure of an economy in a given year differs fromtheeconomicstructuretenyearsearlier.Theperiodoften years has been chosen because such a periodwould typically be expected to cover two businesscycles. Structural changes in terms of the sectoralcomposition of value-added and in terms ofemploymentarereflectedseparately.


228

Figure D.5 shows that observed levels of structuralchange differ over time, with some periods beingcharacterized by high levels of change and others bylower levels. It also demonstrates that the measuredrate of structural change has increased over time inthe United States, as both the peaks and troughsreflected in the chart increase over the years. FigureD.5 even indicates that structural change may havesomething to do with trade. The two lines reflect theshare of world exports of Japan and China,respectively. It isstrikingtoseehowtheriseofthesetrading powers, for example, has coincided withperiods of increased structural change in the UnitedStates.Thefigureillustratesthatadjustmentsintermsofemploymentappear to takeplace later thanvalue-added adjustments, perhaps reflecting that firm-levelproductivity adjustments pre-date adjustments intermsofstaffing.Furthermore,thedifferenceintimingbetween labour and production adjustments is largerinrecentdecadesthaninthe1980sandearly1990s.The employment adjustments are also significantlylarger in the more recent period. This discussion,however, cannot claim to reflect any causal linkbetween increases inexports frommajor tradersandstructural change in their trading partners. It is also

the case that the nature of the reshuffling processfollowing trade reform is likely to be country specificand to depend on the timing of trade reform and thenatureoftradeshocks(e.g.Haltiwanger,2011).

The story reflected in Figure D.5 seems to be in linewithfindingsinrecenttradeliterature:theadjustmentphase following trade shocks can be challenging.14Cosar (2011), for instance, points out that it may beparticularlydifficultforolderworkerstoadjust.15Autoret al. (2012) highlight that the medium-run efficiencylossesassociatedwithadjustmenttotradeshockscanbesignificant.DavidsonandMatusz(2004b)showinatheoretical set-up that unemployment levels followingtradeshockscanbelowerorhigherdependingontheadjustmentpaththataneconomytakes.

Theexistingempiricalliteratureonthedeterminantsofunemployment finds that trade opening is likely todecrease unemployment in the long-run, while it maylead to increases in unemployment in the short-run.Using information on 92 countries for the period1990-2000,Duttetal.(2009)findthatunemploymentincreases immediately after trade opening, butdecreases in the first and second year after trade

FigureD.5:The rise of new competitors and structural change in the United States, 1979-2010

19

79

19

80

19

81

19

82

19

83

19

84

19

85

19

86

19

87

19

88

19

89

19

90

19

91

19

92

19

93

19

94

19

95

19

96

19

97

19

98

19

99

20

00

20

01

20

02

20

03

20

04

20

05

20

06

20

07

20

08

20

09

20

10

0

0.05

0.1

0.15

0.2

0.25

0

1

2

3

4

5

6

Sha

re o

f US

impo

rts

from

indi

cate

d co

untr

y

Ave

rage

SC

I for

the

10

yea

rs p

rece

ding

the

indi

cate

d ye

ar

USA SCI value added USA SCI employment China Japan

Sources:UnitedNations,NationalAccountsmainaggregates(grossvalueaddedbyeconomicactivity,currentUSdollars),LABORSTA,UNCOMTRADE.

Note:TheStructuralChangeIndex(SCI)iscomputedusingfiveyearaveragesatthebeginningandtheendofeachten-yearperiod.Itistypicallymeasuredashalfthesumoftheabsolutevalueofthedifferencesinvalue-addedsectoralsharesovertime:

SCI=

DifferentlevelsofsectoraldisaggregationcanbeusedwithhigherdisaggregationleadingtohigherSCIvalues.TheSCIinthisfigurewasconstructedusingadisaggregationintofivesectors:agriculture,hunting,forestryandfishing;miningandutilities;manufacturing;services;andotheractivities.TheSCIwaspreviouslyusedinProductivityCommission(1998)andBacchettaandJansen(2003).Francoisetal.(2011)alsorefertothemeasure.


229

II D. Tr

aD

e o

pe

nn

es

s a

nD

Th

e b

ro

aD

er

so

cIo

-e

co

no

mIc

co

nTe

xT

opening.Thedeclinemorethanoutweighstheoriginalsurge in unemployment.16 Görg (2011) echoes thisfinding in a literature overview where he concludesthatglobalizationmayleadtohigherjobturnoverintheshort run but that there is no indication that trade oroffshoring lead to higher unemployment overall. In apaper focusing on the long run, Felbermayr et al.(2011a)findthatgreatertradeopennessisconsistentlyassociated with a lower structural rate ofunemployment. The empirical literature thus appearstosuggestthatinthelongrun,tradedoesnotincreaseunemployment or it contributes to reducing it. Wheretrade contributes to increased unemployment, this islikelytobeonlyashort-runphenomenon.

Actual or expected economic difficulties duringphases of transition may, however, create importantchallengesforpolicy-makers,particularlyiftheyaffectthe public’s attitude towards trade or put long-runstructural adjustment processes at risk. Fear of jobloss may, for instance, affect voters’ attitude towardstrade reform independent of whether relevantindividualsactuallyendup losing their job.This issuewillbediscussed in thenextsub-section.Adjustmentperiodsfollowingtradeopeningmayalsoaffect long-run trade and growth patterns (e.g. Mussa, 1978;Davidson and Matusz, 2004b; Francois et al., 2011),becausetheysetthestagefortheresultingstructuralcomposition of employment and production. Thechallengesthatdifferenttypesofcountriesfaceinthiscontextarediscussedbelow.

(iii) Effect of unemployment and income distribution on trade policy

Thissub-sectiondiscusseshowtheactualorperceivedimpactoftradeonjobsandthedistributionofincomeislikelytoaffectpolicy-making,withaparticularfocusontradepolicy.Italsodiscusseshowincomeinequalitywithin countries can affect the benefits that thesecountriesdrawfromtradeopening.

Perceived impact of trade on labour markets and potential rises in protectionism

Incomeinequalityhasrisenwithinmostcountriesandregions over the past two decades. Since this periodhas also been characterized by unprecedentedinternational trade, it is often perceived that thebenefits of rising living standards associated withglobalization have not been shared equally across allsegmentsof thepopulation.There isa risk that suchconcerns may translate into protectionist sentimentandultimatelyaffecttradepoliciesandtradeflows.

Basedontraditionaltradetheory,itwouldbeexpectedthat individuals employed in import-competingindustries are sceptical about trade opening. Peoplewithaskillthatwillbelessindemandafterreformarealso likely to lose and, according to the most recentliterature,thoseemployedinsmallfirmsaremorelikely

toexperiencenegativeconsequencesfromtradethanthoseemployedinlargefirms.17

In industrializedcountries (i.e.countrieswell-endowedwith high-skilled labour), low-skilled labour hastraditionally been expected to lose (in relative terms)fromtrade.Econometricanalysisofsurveyinformationhasconfirmedthatattitudestowardstradeopeningareindeed in line with theoretical predictions. Mayda andRodrik(2005)findthatindividualsworkinginnon-tradesectors tend to be the most pro-trade, while those inimport-competingsectorsarethemostprotectionist.18They also find that individuals with higher levels ofeducational attainment oppose trade restrictions incountrieswell-endowedwithhumancapital.Basedonthe above, standard political economy considerationswould predict that policy decisions would take aprotectionist tendency if sufficiently large numbers ofindividualsconsiderthemselvestobelosersfromtradeopening (Boix, 2011; Mayer, 1987; Dutt and Mitra,2002and2006).Ifthedistributionofgainsfromtradeis sufficiently skewed, protectionist sentiments mayprevailevenwhentheoveralleffectontheeconomy’swelfareispositive.

While most economic analyses of this question havefocused on the actual distributional effects of tradeon income, perceptions and uncertainty aboutindividual outcomes matter. Individuals who sufferfrom job or income loss may be hostile to tradeopeningiftheyperceivetradetobethesourceoftheirproblems independentofwhether this is actually thecase. Individuals may also nurture protectionistsentimentsiftheyfearthelossoftheirjobsasaresultoftradeopeningeveniftheyendupkeepingtheirjobor finding a better one. The latter phenomenon hasbeen examined in Fernandez and Rodrik (1991) whoshowthatindividualsprefertomaintainthestatusquoif they do not know in advance who is going to beaffected by possible negative consequences ofreform.

Ananalysisofsurveydata,collectedintheyear2000,suggests that individuals may evaluate differentlyindividual employment perspectives and theemploymentperspectivesfortheircountryasawhole.In thesurvey,AsianandEuropean interviewees in18countrieswereaskedabouttheirviewsregardingtheirpersonalworksituation,unemploymentintheircountryand the need to limit imports of foreign products.Intervieweeswereasked:

• whether they believe that globalization has a badeffectonjobsecurity(globalization1)

• whether they believe that globalization has a badeffectonstandardsofliving(globalization2)

• whethertheyagreedthattheircountryshouldlimitimportsofforeignproducts(trade)


230

• whether they were worried about their personalworksituation(job1)

• whethertheywereworriedaboutunemploymentintheircountry(job2).

FiguresD.6andD.7 reflecthowconcernsabout jobsare related to views on trade or globalization byrepresenting relevant correlations across individualreplies for the 18 countries. Only statisticallysignificant correlations are presented. Figure D.6illustrates that individuals who are concerned abouttheir personal work situation also believe thatglobalization is bad for job security and standards ofliving. This pattern holds across countries, withcorrelationsbeingsomewhathigher inEuropeanthanin Asian countries. Individuals who are concernedabout their personal work situation also tend to havestronger protectionist views when it comes to trade.Thisagainholdsforallcountriesinbothregions,withtheexceptionoftheUnitedKingdom.

This pattern of correlations is reversed whenindividuals are asked about employment outcomes atthe national level. Figure D.7 reveals that in allcountries, individual responses regarding concernsabout globalization are systematically negativelycorrelated with individual concerns aboutunemployment in their country.19 Combining theinformation shown in Figures D.6 and D.7 suggeststhatintervieweeswhotendtobelievethatglobalization

hasnegativeeffectsonjobsecurityandthestandardoflivingfeartheirownemploymentperspectiveswhileacknowledging that the country’s overall employmentperspectivesmaybepositive.

The above distinction between overall and individualeffectsisalsoreflectedinothersurveyevidencefromEuropean countries. This indicates that a majority ofrespondents believe that globalization providesopportunities for economic growth but increasessocial inequalities.20FigureD.8 looksmorecloselyatsurvey informationregarding inequalityandcomparescountryresponsesregardingtwoquestions:

• Do you consider that differences in incomes aretoolarge?

• Do you consider that globalization represents athreattonationalemploymentandcompanies?

The first question was asked in the Eurobarometersurvey of 2009, while the second question wascontained in the Eurobarometer survey of 2012.Figure D.8 reflects the percentage of respondentswhorepliedpositivelytothetwoquestionsinindividualcountries. It reflects a positive correlation betweenconcerns about inequality and concerns aboutglobalization. This correlation may reflect a Europe-specificphenomenonassimilarexercisesforadatasetcovering global attitudes did not find positivecorrelationsofacomparablesignificance.21

FigureD.6:Attitude towards job insecurity (personal work situation) (correlationwithattitudestowardstradeandglobalization)

-0.1

-0.05

0

0.05

0.1

0.15

0.2

0.25

Japa

n

Chi

na

Chi

nese

Taip

ei

Kor

ea,

Rep

. of

Sin

gapo

re

Mal

aysi

a

Indo

nesi

a

Thai

land

Phi

lippi

nes

Uni

ted

Kin

gdom

Irela

nd

Fran

ce

Ger

man

y

Sw

eden

Italy

Spa

in

Por

tuga

l

Gre

ece

Globalization 1 Globalization 2 Trade

Source:Authors’calculationsbasedonInoguchi(2001).

Note:Globalization1–Doesglobalizationhaveabadeffectonjobsecurity?Globalization2–Doesglobalizationhaveabadeffectonstandardsofliving?Trade–Shouldyourcountrylimitimportsofforeignproducts?Onlystatisticallysignificantcorrelationsarerepresentedinthefigure.


231

II D. Tr

aD

e o

pe

nn

es

s a

nD

Th

e b

ro

aD

er

so

cIo

-e

co

no

mIc

co

nTe

xT

Overall,thesurveyevidencediscussedaboveseemstosuggest that individuals perceive globalization asentailing risks for their personal job situation. Aspredicted inFernandezandRodrik (1991), individuals

may take these risks very seriously even when theyacknowledgethelikelypositiveeffectsofglobalizationfor the overall job market. One of the possibleconsequences is that individuals may increasingly“vote”againstglobalization.

Inequality and the benefits of trade reform

Actual or perceived inequality may not only affecttrade through its possible impact on trade policydecisions; inequality is also likely to affect tradedirectly through its effect on consumption andproduction patterns. Wealthy individuals consumedifferent products from poor individuals and high-income countries produce different goods from low-incomecountries.22

Asconsumersbecomewealthier,theywillspendlowershares of their income on so-called necessity goodssuch as food and increasing shares of income ongoodssuchashousehold furnitureandservicessuchaseducation.Astheir incomeincreasesfurther,otherluxurygoodssuchasjewelleryandcarsstarttoplayamore important role in their consumption basket.Economists refer to thisphenomenonas the “incomeelasticity of demand”: as consumers become richer,they spend more money on goods that have a high-income elasticity of demand. Businesses use thisconcept to predict future sales of their productsdepending on expected changes in income levels orincomedistributioninthemarketstheyserve.

FigureD.7:Attitudes towards job insecurity (unemployment in the country) (correlationwithattitudestowardstradeandglobalization)

Japa

n

Chi

na

Chi

nese

Taip

ei

Kor

ea,

Rep

. of

Sin

gapo

re

Mal

aysi

a

Indo

nesi

a

Thai

land

Phi

lippi

nes

Uni

ted

Kin

gdom

Irela

nd

Fran

ce

Ger

man

y

Sw

eden

Italy

Spa

in

Por

tuga

l

Gre

ece

Globalization 1 Globalization 2 Trade

-0.3

-0.25

-0.2

-0.15

-0.1

-0.05

0

Source:Authors’calculationsbasedonInoguchi(2001).

Note:Globalization1–Doesglobalizationhaveabadeffectonjobsecurity?Globalization2–Doesglobalizationhaveabadeffectonstandardsofliving?Trade–Shouldyourcountrylimitimportsofforeignproducts?Onlystatisticallysignificantcorrelationsarerepresentedinthefigure.

FigureD.8:Attitudes towards globalization (2012) and inequality (2009) in Europe (percentageofrespondentsthatagreeoragreestronglywiththestatementsshownontheaxes)

60

70

80

90

100

Diff

eren

ces

in in

com

e ar

e to

o la

rge

Globalization represents a threat to nationalemployment and companies

20 30

DK

40 50 60 70

NLMT

SE

UKLU

ES

PT CZ

BE

CYEL

FR

LVHUSI

PL

BG

RODE

SK

FL

EELT

IE

AT

IT

Source:Authors’calculationsbasedonEurobarometer2009and2012.


232

Inlinewiththeabove,Grigg(1994)showsthatspendingonfoodrangedfrom64percentofhouseholdincomein Tanzania to less than 15 per cent in Australia andNorthAmericaintheearly1980s.23BoxD.1illustrateswhy such differences in consumption patterns mayaffect countries’ positions in trade negotiations.Consumption patterns do not only differ acrosscountries, theyalsodifferwithincountriesanddependon income distribution within countries. Dalgin et al.(2008)find that importsof luxurygoods increasewithcountries’levelofinequality.

Income inequality is also likely to affect production.Linder (1961) argues that proximity to a largeconsumermarketforhigh-qualitygoodsgivesfirmsinwealthy countries a comparative advantage inproducing those goods. When exporting, these firmsfind larger markets for high-quality goods in otherhigh-income countries. Accordingly, Linder predictedthat trade volumes are larger among countries withsimilar income levels. More recent research hasconfirmed that the richer and the more similarcountriesare, themore they tradeamongeachother,and the larger theshareof intra-industry trade– thatis, the larger the share of differentiated goods (e.g.Bergstrand,1990).

As low- and middle-income countries grow richer, theyare likely to consume, produce and export increasinglysophisticatedgoods.Ifinequalitycontinuestoincreaseinlinewithrecentlyobservedtrends,thiswillbeparticularlybeneficialfortradein“luxurygoods”,i.e.goodswithhigh-incomeelasticitiesofdemand.Fieler(2011),forinstance,predicts that continued growth due to productivityincreasesinChinawillgohandinhandwithasignificantincreaseintheconsumptionofluxurygoods.

In Fieler’s model, China’s production of luxury goodsalsoincreasesbutbymuchless,becausethecountry’scomparative advantage remains in the production ofless sophisticated goods.24 Indeed, Chineseproductionof lesssophisticatedgoodsisexpectedtotake significant advantage of the productivityincreases mentioned before. Fieler (2011) predictsthat world prices of “basic” goods will decrease inrelativetermsasaconsequenceofChineseincreasesinsupply.Relativepricesofluxurygoodsarepredictedto increase because of the demand surge in China.According to Fieler, rich countries that are netexportersofluxurygoodswouldtakeadvantageofthischange. Poor countries that are large consumers of“basic”productswouldtakeadvantageofthedecreasein their prices. Middle-income countries that are netimportersofluxurygoodscouldbenegativelyaffectedbysuchchangesinrelativeprices.

Fieler (2011)doesnotprovideaseparateanalysisforconsumersofdifferentincomegroupswithincountries.Inlinewiththediscussioninpreviousparagraphs,itisnot unreasonable to expect that the relative pricechanges may end up benefiting lower-incomehouseholds who consume more basic goods and thehighest-income households if they are owners ofproductionfactorsinvolvedinproducingluxurygoods.Fieler’spaperandmostof the literaturediscussedsofarassumesthatglobalandnationalmarketsfunctionrelatively smoothly. If this is not the case, inequalitywithin countries can have a significantly strongereffect on the distribution of gains from trade withinandacrosscountries.25

The work by Foellmi and Oechslin (2010; 2012)illustratesthatiffinancialmarketsarecharacterizedby

BoxD.1: Food security versus food safety

Cross-countrydifferencesinconsumptionpatternsmayaffecttradepolicy-makers’negotiationpositions.Inrecent debates related to agricultural trade, for instance, representatives of low-income countries havetendedtoemphasizetheneedforsecureaccesstofoodwhilerepresentativesofindustrializedcountriesarepayingincreasingattentiontotheneedtoguaranteethequalityoffood.

Thelastdecadehasbeencharacterizedbyhighlevelsofpricevolatilityinagriculturalcommodities.Giventhatpoorhouseholdstendtospendalargeshareoftheir incomeonfood,theyareparticularlyvulnerabletopricehikesinagriculturalcommodities.TheWorldBank(2011)hasestimatedthatrisesinfoodpricesbetweenJuneandDecember2010pushedanadditional44millionpeoplebelowtheUS$1.25povertyline.Asaconsequence,concerns about food security have been high on the agenda of policy-makers, particularly in developingcountries. In order to ensure adequate food supply at acceptable prices, policy-makers have turned tointerventionssuchassubsidizingtheconsumptionoffoodorrestrictionsontheexportsoflocallyproducedfood.

Althoughvolatilityincommoditypriceshasalsoaffectedindustrializedcountryconsumers,concernsaboutfoodsafetyratherthanfoodsecurityhavebeenprominentinthepublicdebateinthosecountries(Cheongetal.,2013).Indeed,asconsumersbecomerichenoughnottoworryaboutbasicaccesstofood,thequalityoffoodstartsgaining in importanceforthem.Thebovinespongiformencephalopathy(BSE)crisis inthe late1990s, the 2011 E-coli outbreak and bird flu transmission through poultry trade in recent decades havecontributedtoconcernsaboutthesafetyofimportedfood.Inthiscontext,demandsforstricterfoodsafetyregulation have increased and have resulted in new forms of non-tariff measures (NTMs) or in privatelabellingschemes.


233

II D. Tr

aD

e o

pe

nn

es

s a

nD

Th

e b

ro

aD

er

so

cIo

-e

co

no

mIc

co

nTe

xT

market imperfections, inequality may exclude someparts of the economy from trade. In particular, theyshowthattradeopeningmaywidenincomedifferencesamong firm owners in LDCs since shrinking profitmargins make it difficult, if not impossible, for theseentrepreneurs toaccesscredit (FoellmiandOechslin,2010). If investments in new technologies arenecessary to be competitive at the global level,imperfectfinancialmarketsare likely to impedesmalland medium-sized enterprises (SMEs) in LDCs frommaking the necessary investments (Foellmi andOechslin, 2012). This suggests that policies targetedtowards facilitating access to credit would allow alargerpartofanLDC’seconomytoparticipateintradeandwouldthusincreasethegainsfromtrade.26

Intheset-upproposedbyFoellmiandOechslin(2012),targeting companies that are relatively less creditconstrained, i.e. the larger companies among thecredit-constrained ones, is likely to bring the largestbenefits in terms of increases in trade. As such, thefindings of Foellmi and Oechslin (2012) may provideuseful insights for the debate on strengtheningdevelopingcountries’supplyresponsetotradereforminthecontextofAidforTrade.

(c) Tradeandlabourmarkets:differentchallengesatdifferentstagesofdevelopment

Whether and how inequality will affect future tradeflowsandtheresultingeconomicbenefits,andwhetherconcerns about jobs or inequality will affect tradenegotiators’ future decisions, is to a large extentdependentoncountry-specificsituations.Asdescribedabove,GDPpercapitastilldifferssignificantlyacrosscountries and these cross-country differences willcontribute to differences in national consumptionpatterns. They will also determine comparativeadvantage and thus the global distribution ofproduction.

What the future brings will very much depend onwhether fast growing developing countries willcontinue to catchup, andwhether thosewhodidnotmanagetocatchupinthepastaremoresuccessfulinthe future. Another factor will be how current high-income countries cope with the emergence of newcompetitors.Muchhasbeenwrittenabouthowpolicy-makers can influence the development path ofindividualeconomies.27ThesimulationsinSectionB.3providedsomeillustrationofhowpolicydecisionscaninfluence future tradeandgrowthpaths. This sectionadds to this discussion by examining how labourmarketchallengesmayinteractwithgrowthchallengesencounteredbydifferenttypesofcountries.

Inparticular,thissub-sectiondiscussesthreestagesof integration in global markets that may determinethe future challenges to be faced by individual

countries in their labour markets. First, numerouslow-income countries, in particular LDCs, have notmanaged to successfully integrate into globalmarketsandneed to findways toovercomeexistingobstacles.Secondly,with increasingGDPpercapitaand wages, a number of low and middle-incomecountries that have successfully integrated intoglobalmarketsasprovidersoflow-wageexportsmaywish to move into the production of higher value-added segments. Thirdly, in the past two decades,advancedeconomieshavehad to copewith the riseof new competitors, which has often put labourmarketsunderstrain.Ifthenewcompetitorsmanageto move into new product niches in the near future,furtherlabourmarketadjustmentsmaybenecessaryintheindustrializedworld.

(i) Not staying behind

Developing economies, especially in Africa, areabundantinrawmaterialsandhavebecomeimportantsources of global supply to meet the strong demandarisingfromtherapidurbanizationandindustrializationinotherdevelopingcountries,suchasChinaandIndia.Will an increase in the terms of trade resulting fromthese demand increases spur income growth forcommodityexporters, notably inSub-SaharanAfrica?Or will the dependence on commodity exports leadsuchcountries to fall furtherbehindother regions, inparticularAsia?

One of the consequences of the recent increase incommodity prices has been an increased level ofspecialization in commodity exports by many Africancountries, including the LDCs among them (seeSection B.2). It is likely that this increased economicimportance of commodity exports has gone hand inhandwithincreasesinGDPintheshortrun.However,evidenceshowsthatover timecountries thatbecomericheralsotendtodiversifytheirexports(Cadotetal.,2011). It suggests that a focus of policy-makers onincreasing economic diversification in low-incomecountriesappearstobejustified.

The debate about ways to promote diversification canprovide useful insights for the multilateral tradingsystem and in particular for the capacity-buildingactivities within this system. Brenton et al. (2009)emphasize two points that may be relevant for policy-makers concerned about diversification. First, low-incomecountriesexperienceamuchhigher“deathrate”of new exports than middle or high-income countries.Thisindicatesthatfirmsinthesecountriesfindithardtosustain and increase exports of potentially viableproducts. Policy-makers may therefore want to payattention tomarket failures, institutionalobstaclesandpolicyshortcomingsthatarestranglingproductlinesintheir infancy. Secondly, low-income countries tend toserveonlyasmallportionofpotentialoverseasmarketsfor the products that they already export. A proactiverole of governments to help introduce prospective


234

exporters to foreignmarketsor link themup toglobalsupply chains can therefore have a high pay-off.Lederman et al. (2009) describe the role that exportpromotion agencies can play in this context andemphasizetheimportanceoftheinstitutionalset-upofsuchagenciesfortheirsuccess.

Anyeffort to increasediversificationwillhave to takeinto account one important labour marketcharacteristic of LDCs. On average, agriculturalemployment represents 72 per cent of totalemployment inLDCs, compared toonly4percent inhigh-income and highly diversified economies. Anymoves towards more diversification and betterintegration in world markets is therefore likely to gohandinhandwithmigrationfromruraltourbanareas(WorldBank,2012).Measurestofacilitateintegrationinurbanareascould includeproviding informationonaccommodation or job opportunities (Cheong et al.,2013).Becauseofthesizeoftheagriculturalsector,itcould also make sense to direct at least some efforttowardsthestrengtheningofcertainbranchesofthatsector in order, for instance, to follow a number ofLDCs that managed to enter high value-added nichemarkets with their agricultural exports. The successstories of Kenyan cut flower exports and Rwandancoffee exports are among the best-known examples.Aid for Trade has the potential to play an importantroleinthiscontext,asillustratedinBoxD.2.

(ii) Catching up with the front-runners

Global supply chains have increased trade betweendeveloped and developing economies and have alsorekindledinterestintheroleofcomparativeadvantagein international production.Countriesexport differenttypesofpartsandcomponentsatdifferentstagesofdevelopment, with developing countries completinglow-skill labour-intensive tasks. It is the advancedeconomies, where skill and capital-intensive tasks,that capture most of the value-added trade arecompleted.Intra-firmtrade,facilitatedbyinvestmentinthe establishment of subsidiaries overseas, isbecomingincreasinglyimportantinthiscontext.

The transfer of technology and knowledge facilitatedthrough such trade and FDI has made it possible fordevelopingcountriestomoveuptheproductladderintermsofcapitalintensityandqualitymorerapidlythanin the past. A number of emerging economies havebeen rather successful in taking advantage of theseopportunities (see Sections C.2 and C.3). For China,there is already some evidence of deepeningproductive capacity and of a move up the productladders(Rodrik,2006;Hausmannetal.,2007).India’sexportbasket is sophisticated relative to itsGDPpercapita, albeit to a lesser extent than China’s(Hausmannetal.,2007;TianandYu,2012).

TableD.2indicatesthatthelabourmarketchallengesthat emerging economies will face in this contextdiffer significantly across countries. While thesectoral distribution of employment in the RussianFederation and South Africa already resembles thatin industrialized countries, China and India stillemploy the majority of their labour force in theagricultural sector. Both countries will therefore beable to rely on a sizeable quantity of cheap andrelatively low-skilled labour. Yet, if they manage toshift production to a higher-technology and higher-qualitymix,itistherelativedemandforskilledlabourthat will rise. Both countries will therefore face thedual challenge of absorbing large numbers ofrelatively low-skilled rural workers into the labourmarkets and of educating large numbers of workerstopreparethemforthenextgenerationofjobsinthemanufacturing and services sectors. Given thedivergentdemographicpatternsinthetwocountries,thischallengemaybemoreformidableforIndiathanforChina.

(iii) Adjusting to new competitors

Withtheriseofnewmajorglobalplayers(forexample,Brazil,ChinaandIndia;seeSectionB.2),othertradershad to adjust to a new situation in global markets.OECDimportsfromChinahaveincreasedsignificantlyoverthepasttwodecades,replacing localproductionand imports from other countries. At the same time,

BoxD.2: Entering niche markets with agricultural exports: Rwandan coffee

“Rwandahasa ‘NationalCoffeeStrategy’.Rwandanspecialtycoffee iswinning internationalcompetitions,commands some of the world’s highest prices and is sought out by Starbucks, Green Mountain Coffee,Intelligentsia,andCounterCultureCoffee.Thereispreliminaryevidencethatthecoffeeindustryiscreatingjobs,boostingsmallfarmerexpenditureandconsumptionandpossiblyevenfosteringsocialreconciliationbyreducing“ethnicdistance”amongtheHutusandTutsiswhoworktogethergrowingandwashingcoffee.

Howdidthishappen?First,theRwandangovernmentloweredtradebarriersandliftedrestrictionsoncoffeefarmers. Second, Rwanda developed a strategy of targeting production of high-quality coffee, a specialtyproductwhosepricesremainstableevenwhenindustrial-qualitycoffeepricesfall.Third,internationaldonorsprovidedfunding,technicalassistanceandtraining,creatingprogramssuchastheUSAID-fundedSustainingPartnershipstoEnhanceRuralEnterpriseandAgribusinessDevelopment(SPREAD).SPREAD’spredecessorstartedthefirstRwandancoffeecooperativeasanexperimentin2001,andtheprojectcontinuesitsworkimprovingeachlinkinnewly-identifiedhigh-valuecoffeesupplychains.”

Source:EasterlyandFreschi,AidWatch,May2010.


235

II D. Tr

aD

e o

pe

nn

es

s a

nD

Th

e b

ro

aD

er

so

cIo

-e

co

no

mIc

co

nTe

xT

access to the Chinese market has also providedopportunitiesforOECDexporters.Theresulthasbeena repositioningofnumerousproducersandexporterswithregardtoglobalmarkets.

Most OECD countries have lost in terms of globalmarket share in the past two decades. Table D.3provides details for Germany, Japan and the UnitedStates,thethreemainexportpowersbeforetheriseofChina. Employment patterns have also changed overthis period, with industrial employment decliningmarkedly everywhere. In 2010, while industrialemployment represented more than a quarter of totalemployment in Germany and Japan, it had shrunk toaround 17 per cent in the United States. In all threeOECDcountries,servicesemploymentnow representsbyfarthelargestshareinemployment.Someobserversconsiderthattheextentandspeedatwhichthischangehas taken place is a matter for concern (e.g. Spence,2011),notablybecauseoftheroleofmanufacturingasa driver of innovation. Pisano and Shih (2012), forinstance, argue that production and research anddevelopment(R&D)activitiesinmanufacturingneedtotakeplace in thesameor inneighbouring locations inorder for R&D to be effective. According to theirargument, countries losing their production base inmanufacturing would also run the risk of losing theirinnovativecapacity(seeSectionC.3).

Employment in advanced economies is beingincreasingly concentrated in the services sector. It

consists of a number of sub-sectors that areassociated with high qualifications and high pay (e.g.finance, legal affairs) and others where employmenttendstobeassociatedwithlowqualificationsandlowpay(e.g.retail,hospitality,construction,daycare).Thelattergroupofsub-sectorsisalsooftencharacterizedbyhighlevelsofinformalemployment.

In order to understand whether restructuring ofemployment leads to better or worse employmentoutcomes, it would be necessary to understand whattypes of services jobs are created. Until recently,studies analysing labour market changes followingtradereformonlytookintoaccountthemanufacturingsectorand failed toaccount for theservices industryor the informal economy. The availability of newdatasetsmakesitpossibletoanalysepotential labourflows out of manufacturing. Ebenstein et al. (2009)findthatworkerswholeavemanufacturing,asaresultof trade reform or offshoring, to take jobs in theservicessectorsufferfromawagedeclineofbetween6and22percent.28Thegrowthoftheservicessectorin terms of employment may therefore be one of thedriversoftheobservedpatternsofinequalityincreaseinindustrializedeconomiesthatwasdiscussedabove.

The current employment structure in industrializedcountrieshasarisen following theemergenceofnewplayers in global markets. As mentioned above,emerging economies may seek to climb up the valuechaininordertosustaingrowth.IfChinaandpossibly

TableD.2:Evolution of the share of employment per sector, BRICS (percentage)

Share of world exportsShare of employment

Agriculture Industry Services

Brazil1995 0.9 26.1 19.6 54.3

Brazil2009 1.2 17.0 22.1 60.7

China1995 2.9 52.2 23.0 24.8

China2008 8.9 34.6 27.2 33.2

India1994 0.6 61.9 15.7 22.4

India2010 1.5 51.1 22.4 26.5

RussianFederation1995 1.6 15.7 34.0 50.0

RussianFederation2008 2.9 8.6 28.9 62.4

SouthAfrica2000 0.5 15.6 24.2 59.4

SouthAfrica2009 0.5 5.7 25.7 68.6

Source:ILOKILM(Keyindicatorsofthelabourmarket)databaseandWTO.

TableD.3:Evolution of the share of employment per sector, major OECD exporters (percentage)

Share of world exportsShare of employment

Agriculture Industry Services

Germany1995 10.1 3.2 36.0 60.8

Germany2010 8.2 1.6 28.4 70.0

Japan1995 8.6 5.7 33.6 60.4

Japan2010 5.0 3.7 25.3 69.7

UnitedStates1995 11.3 2.9 24.3 72.8

UnitedStates2010 8.4 1.6 16.7 81.2

Source:ILOKILM(Keyindicatorsofthelabourmarket)databaseandWTO.


236

other large emerging economies such as Brazil andIndiadoso,itispossiblethatindustrializedeconomieswill have to withstand another wave of major labourmarket adjustments. It cannot be excluded that thisadjustment would contribute to a further polarizationwithin labour markets, whereby there is growth inemployment in the highest- and lowest-skilledoccupations,withdecliningemployment in themiddlerange of the skill distribution.29 With greatercompetitiveness across all skill levels, matching firmlevelstrategieswitheducationandtrainingpolicies islikelytobecomeincreasinglyimportant(seeBoxD.3).

(d) Conclusions

Employmenthasbeenhighonpolicy-makers’agendasinthepastyearsandmayremainthereforawhile.Thereasons for this differ across countries. Somecountries need to find ways to absorb a growingpopulationintothelabourmarketortoabsorbalargeruralpopulationintoformalandurbanactivities.Othercountries have been plagued by high unemploymentrates since the recent economic crisis. In addition,there appears to be a growing discomfort across theglobewith increasing levelsof income inequality.Theshareofincomeofthe1percentwealthiestindividualshasincreasedsignificantlyinmanycountriessincethe1990s. While there is no conclusive evidence thattrade contributes significantly to changes in long-rununemployment or in inequality, public perceptionsimplythatpolicyreformshavetobeseentodowellon

these two fronts inorder to receivepublic support inthecomingyears.

In the public debate, “globalization” has often beenassociated with increases in inequality observed inrecentdecades.Researchhastriedtodisentangletheeffect of different components of globalization onincome distribution and has also tried to understandwhether different components of globalization actjointly. Available evidence suggests that trade isunlikely tohavehadasignificant impacton inequalitythrough the traditional channels of shifting relativedemandforproductionfactors.However,thereissomeevidence that trade goes hand in hand withtechnologicalchangeandthatthecombinationofthetwo contributes to increased inequality. Recentevidencealsosuggeststhatglobalfinancialflowsmayplay a role in explaining observed increases ininequality. Given that FDI, technological change andtrade are among the main drivers of growth, thistherefore hints at a possible challenge for policy-makers to ensure that growth is maintained but alsobalancedintermsofincomedistribution.

The relationship between trade and employment hasreceivedasignificantamountofattentionfrompolicy-makers in recentyears.Evidencesuggests that tradeopening can contribute to job creation. At the sametime,asittendstogohandinhandwiththeadoptionof new technologies and productivity increases,successful integration in terms of export growth may

BoxD.3: Relevance of education and training policies for integration in global markets

Intoday’srapidlychangingandhighlyintegratedworld,skillsatalllevelsofthefirmbecomeextremelycriticalforperformanceandglobalcompetitiveness.Accesstoaskilledlabourforcewillmakeiteasierforfirmstoenternewmarketsabroad,tointegrateintoglobalsupplychains,tosurviveandthriveinthedomesticmarketandtoadjusttochangingconditionsinglobalmarkets(e.g.Greggetal.,2012;Froyetal.,2012).

Educationandskillspoliciesalsohavethepotentialtocontributetotwoobjectivesdiscussedinthissection:job creation (in particular for the young) and reducing inequality (because the highly skilled tend to copebetterinmoderneconomiesthanthelowskilled).

Onewaytopreparetheyoungforthechallengesoftheirfutureworkingenvironmentistoensurethattheyhaveagoodbasicknowledgeandtheabilitytoemploythisknowledgeindifferentsettings(Almeidaetal.,2012;Woessmann,2011).However,ensuringthatyoungpeoplehaveaskillsetthatmakesthem“adaptable”isunlikelytobeenoughforthemtofindajob,inparticularwhentheyfirstenterthejobmarket.

Inhiringprocesses,employersaretypicallylookingforcandidatesthathaveasetofskillsspecifictothejob,ortothesectororsub-sectorinwhichthecompanyisactive.Giventhateducationandtrainingdecisionsaretakenwell–oftenyears–aheadof themomentof labourmarketentry, time-inconsistencyproblemsmayarise,i.e.situationswhereeducationandtrainingdecisionstodaydonotmatchtheskillsdemandoftomorrow(e.g.Almeidaetal.,2012).

Tominimizethisproblemandtoreducetheoccurrenceofskillsmismatches,itwillbeincreasinglyimportantforgovernmentstostrengthenskillanticipationmechanismsintheireconomy.Thisimpliesstrengtheningthecollectionofinformationaboutcurrentandpossiblefutureskilldemandbyemployersandtoensurethatthisinformation is passed on to students. It also implies that the transmitted information should influenceeducation and training supply. Employers are well placed to know about current and possible future skilldemand and their involvement is therefore likely to be important for skills anticipation policies to besuccessful.


237

II D. Tr

aD

e o

pe

nn

es

s a

nD

Th

e b

ro

aD

er

so

cIo

-e

co

no

mIc

co

nTe

xT

not lead to large-scale jobcreationunless thesupplyresponse is significant in exporting firms and thedomesticsupplychainsupportingthem.

More generally, globalization facilitates the spread ofideas and innovations, which is likely to contribute toan increased speed of technological change. Thelatterimpliesthatfirmsandworkersneedtoconstantlyadjusttonewtechnologies.Thosecompetinginglobalmarketsalsoneed toconstantlyadjust tochanges inthe competitive environment, as has been evidentduring the past two decades that have beencharacterizedbytheriseofnewmajorplayersinglobalmarkets.

Inordertofarewellinanincreasinglyintegratedworld,economies need to have a strong capacity to adjust.This is true for many aspects of the economy but inparticularfortheirlabourmarkets.Thenatureandtheextent of labour market challenges will differ acrosscountries.Formanylow-incomecountriesnotyetwellintegrated into global markets, successful integrationwill imply significant economic restructuring, mostlikely from agricultural to industrial and servicesemployment.

A number of emerging economies may face the dualchallengeof having toemploy largenumbersof ruralworkers while simultaneously moving into highervalue-addedactivities.Inordertosuccessfullydothis,these countries will need to maintain exports in anumberoflow-skilledactivitieswhileatthesametimeexpanding employment rapidly in new, higher value-added activities. If growth in emerging economies issustained and the relative weight of individualexporters continues to change, labour markets inindustrialized countries may continue to be underpressuretoadjust.

Survey evidence reveals that individuals inindustrialized and emerging economies alike areconcerned about their individual employmentperspectives, even when they have optimistic viewsabouttheeconomicperspectivesfortheircountryasawhole in a globalized world. Taking into account theextent of such fears and the fact that they oftencoincidewithconcernsaboutthedistributionaleffectsof globalization, negative views of globalization maypersistorevenincrease.Itwouldbeveryriskytoreachapointwheredeterioratingperceptionsofjobsecurityand income distribution within countries become apressurepointforcountriestoresorttoprotectionism.

In the short run, policy-makers can address thesefears by providing social protection to individualsduring periods of unemployment.30 In the medium tolong run, education and training policies are likely toplay an important role for all countries. Access to askilled labour force will make it easier for firms toaccessnewmarketsandforfirmsandworkersaliketoadjust to changing market conditions. Education and

training policies can also play a role in addressingpossible distributional concerns, as high-skilledworkers will find it easier to take advantage of newopportunitiesthanlow-skilledworkers.

In addition, active labour market policies that helpdisplaced workers to find new jobs can contribute toreducingfearsaboutjobloss.Policiesthatstrengthentheenablingenvironmentforenterprisescanpositivelycontribute to job creation. Initiatives to strengthendomestic financialmarkets canhaveparticularly highpay-offs,totheextentthattheysucceedinfacilitatinginvestmentsnecessarytoraisefirms’competitiveness.Moregenerally, initiatives, suchasAid forTrade, thataim to strengthen supply response in developingcountriescancontributeinthisregard.

2. Environmentalconcerns

Open trade and environmental protection are keyelements of sustainable development.31 This wasrecognized at the Rio Earth Summit 20 years ago,when the international community emphasized theimportance of cooperation “to promote a supportiveand open international economic system that wouldleadtoeconomicgrowthandsustainabledevelopmentin all countries, to better address the problems ofenvironmental degradation”.32 When the WTO wasestablishedafewyearslater,WTOmembersaffirmedtheir commitment to sustainable development andidentifiedenvironmentalprotectionandthesustainableuseoftheworld’sresourcesasessentialgoalsofthemultilateraltradingsystem.33

Bothopentradeandsoundenvironmentalpoliciesshouldwork to utilize existing resources better (UNEP, 2013).Openness to trade allows countries to specialize in theproductive activities in which they have a comparativeadvantage. It also extends the market for domesticproducers,enablingthemtoexploiteconomiesofscale.Environmentalpolicyseekstoincreasetheefficiencyofthe economic system by ensuring that the full costs ofproduction and consumption, including environmentalcosts,arereflectedineconomicdecisions.

Beyondthisbroadlevelofcommonality,tradeandtheenvironment interact in complex ways, with multiplelinks and feedback effects between them. Thus,managing the interface between trade and theenvironment poses multiple challenges, including forthe WTO. Based on this recognition, the sectionexploresselectedaspectsoftherelationshipbetweentrade and the environment, and identifies possiblefuturechallenges.Specifically, itexaminesthe impactoftradeopennessontheenvironment,andtherelatedquestion of how trade may be influenced by publicperceptionsofitsenvironmentalimpacts.Thissectionthen explores the multiple interactions betweenenvironmental policy and trade, as illustrated by twosets of climate change policies (border carbon


238

adjustments and incentive schemes for renewableenergy). The analysis reveals that, if not managedcarefully, the interaction between trade and theenvironmentmaygiverisetotradeandothertensions,whichmayundermine the futurecontributionof tradeopenness to economic growth and sustainabledevelopment.

(a) Patternsofenvironmentaldegradation

The environment and the economy are twointerdependentsystems. In recentdecades, thescaleof growth has led to significant environmentaltransformations and problems (Dittrich et al., 2012)(seeFigureD.9).For instance,morethan50percentof the planet’s land surface has been modified byhuman activities (Hooke and Martín-Duque, 2012).Increased pressure on biodiversity has led tosignificant loss of wild species and increased risk ofextinction(SecretariatoftheConventiononBiologicalDiversity,2012).

Environmental degradation is complex and can takemultiple (non-mutually exclusive) forms, such as airpollution (e.g. outdoor and indoor air pollution,stratospheric ozone depletion, climate change), waterpollution (e.g. groundwater depletion, freshwaterpollution,marinepollution,coralloss),changesinlanduse(e.g.soilerosion,desertification,drought,wetlandsloss),biodiversity loss (e.g. speciesextinction,naturalhabitat loss, invasive species, overfishing), orchemicals and waste pollution (heavy metals,persistent organic pollutants, radioactive waste)(UnitedNationsEnvironmentProgramme,2012).

Environmentaldegradationposesdirect risksto long-termeconomicdevelopment(Delletal.,2012),nationalsecurity (Matthew, 2000) and political stability(O’Loughlin et al., 2012). In addition, environmentalpollution has various harmful and adverse healtheffects.Ithasbeensuggestedthatalmostone-quarterofalldeathsandthetotaldiseaseburden(uptoone-third of the disease burden of children) can beattributedtoenvironmentalriskfactors(WorldHealthOrganization,1997;Prüss-ÜstünandCorvalán,2006).

Economic theory suggests that environmentaldegradation is the result of market failures, such asthe difficulty to define, allocate and enforce propertyrights of environmental resources. Environmentaldegradation is a typical negative externality, whicharises when producers or consumers who useenvironmentalresourcesandgeneratepollutiondonottakeintoaccounttheharmfuleffectsoftheiractivitieson the rest of society, which leads to social costs inexcessofprivatecosts.

Depending on their scope and range, environmentalproblemsmaybelocal,regionalorglobal(Ramanathanand Feng, 2009). If the polluting activity and itsassociated environmental impact occur in the samegeographical location, pollution is considered local.Examples include water pollution, emissions ofparticulate matter, and land degradation. Regionalpollution (e.g. emissions of sulphur dioxide (SO2) orcontamination of large rivers) involves a pollutingactivity whose effects straddle an entire region andpossibly multiple jurisdictions. Global pollution (e.g.emissionsofcarbondioxide(CO2)orozone-depleting

FigureD.9:Trends in output, trade and pollution, 1970-2008

0

50

100

150

200

250

300

350

400

19

70

19

71

19

72

19

73

19

74

19

75

19

76

19

77

19

78

19

79

19

80

19

81

19

82

19

83

19

84

19

85

19

86

19

87

19

88

19

89

19

90

19

91

19

92

19

93

19

94

19

95

19

96

19

97

19

98

19

99

20

00

20

01

20

02

20

03

20

04

20

05

20

06

20

07

20

08

Living Planet Index SO2 Emissions Index GDP (constant 2000 US$) Index

CO2 Emissions Index NOX Emissions Index Merchandise Trade (% GDP) Index

Sources:WTOSecretariat,basedonEuropeanCommissionJointResearchCenter(2011),WWF(2012)andWorldBank(2012c).

Note:Datahavebeentransformedintoindices(1970baseyear).


239

II D. Tr

aD

e o

pe

nn

es

s a

nD

Th

e b

ro

aD

er

so

cIo

-e

co

no

mIc

co

nTe

xT

substances) refers to a polluting activity withworldwide impacts. Global pollution does notnecessarily entail homogenous consequences; forexample,oneofthefeaturesofclimatechangeisthatallcountrieswillbeaffected,butnotinthesameway.

Duetothemulti-facetednatureofcountries’impactontheenvironment, it isdifficulttodefineasinglesetofenvironmental indicators comparableacross timeandcountries. Several sets of indicators have beendeveloped; some integrate economic and socialvariables to reflect sustainability. Besides data onemissions of CO2, SO2 and nitrogen oxide (NOX),34andbiodiversitymeasures,35thereareseveralsetsofbroadenvironmentalindicators,includingAdjustedNetSaving, Environmental Performance Index, EcologicalFootprint,andEnvironmentalImpact.36

A descriptive analysis of these data indicates thatcountries’impactontheenvironmentattheworldwidelevel is characterized by strong heterogeneity andasymmetry (see Figure D.10). Developed anddeveloping countries perform differently in terms ofenvironmental impact, yet both can perform better orpoorly. For example, in terms of emissions, eightdeveloped and 11 developing countries37 generatemore than three-quarters of global CO2 emissions.Similarly,17countries(13developingcountriesamongthem)38 generate more than three-quarters of globalSO2emissionswhilesevencountries39representmorethan half of global emissions of NOx. The strongvariationintherankingscouldpartlyreflectdifferencesin the methodologies and environmental coverage ofthe indices and measures as well as different

FigureD.10:Comparison of environmental performance indices

-20

10

4 5 6

20

30

40

-10

0

Adj

uste

d sa

ving

s pe

r cap

ita 2

01

0

log GDP per capita (constant 2000 US$)

7 8 9 10 11 12

20

50

4 5 6

60

70

80

30

40

Env

ironm

enta

l per

form

ance

20

10


7 8 9 10 11 12

0

6

4 5 6

8

10

12

2

4

Eco

logi

cal f

ootp

rint 2

00

8


7 8 9 10 11 12

0

15

4 5 6

20

25

30

5

10

Bio

capa

city

20

08


7 8 9 10 11 12

50

110

4 5 6

130

150

190

70

90

170

Pro

port

iona

l env

ironm

enta

l im

pact


7 8 9 10 11 12

50

110

4 5 6

130

150

190

70

90

170

Abs

olut

e en

viro

nmen

tal i

mpa

ct


7 8 9 10 11 12

Sources:WTOSecretariat,basedonWorldBank(2012c),YaleCenterforEnvironmentalLawandPolicyandCenterforInternationalEarthScienceInformationNetwork(2012),Boruckeetal.(2013)andBradshawetal.(2010).

Note:ThelogarithmofGDPpercapita(constant2000US$)wasusedtomitigateskewedness.


240

environmental challenges faced by countries. Severalcountries, developed and developing, are outliers (i.e.significantly above or below the average) in terms ofpositive or negative impact on the environment,reflecting thespecific relationshipbetweeneconomicdevelopmentandenvironmentalperformance.

The so-called environmental Kuznets curvehypothesizes an inverted-U shape relationshipbetween environmental degradation and economicgrowth: environmental degradation increases witheconomic growth for low levels of income per capitabut eventually improves beyond a given threshold(Grossman and Krueger, 1993). This pattern ofpollution and income may be related to consumers’income-elastic demand for environmental protection,which entails an improvement in pollution abatementpolicies and environmental quality as incomeincreases. The empirical evidence in support of theenvironmentalKuznetscurveremainscontroversial, inpart due to the lack of appropriate data and severaleconometricissues.Whilesomestudiesfindevidenceof an inverted U-shape curve for specific pollutants(e.g. SO2 emissions), others produce inconsistentresults(e.g.CO2emissions).

Recent empirical estimates that rely on morerepresentativedatasets,higher-qualitydataandmoreappropriate econometric techniques suggest thatcountries’ environmental performance depends notonlyonthelevelofeconomicdevelopmentbutalsoonseveral factors that are related to income, includingpolitical institutions, good governance and thediffusionoftechnologicalinnovation.Regardingtrade,thekeyquestionistowhatextentproduction,transportand consumption related to trade contributes toadditional pressure on the environment. This isdiscussedinthenextsection.

(b) Trade,theenvironmentandpublicperceptions

As noted, international trade flows have increaseddramatically during the past three decades, a periodthat has coincided with significant environmentaldegradation. This has raised concerns about thepossible contribution of trade to environmentaldegradation,andhassparkedalargeliteratureonthesubject of whether trade is good or bad for theenvironment. The answer to this question hasimportant implications for the future of internationaltrade, reflecting in part the two-way relationshipbetween trade and the environment, and the multiplefeedback effects between these two interconnectedsystems. The following discussion illustrates possiblepolicychallengesthatarisefromthisdualrelationshipby examining the effect of trade on the environmentandtheinfluenceontradeofpublicperceptionsaboutitsenvironmentalconsequences.

(i) How are trade and the environment linked?

One way in which economists have examined howtradeaffectstheenvironmentisbybreakingdowntheimpact of a marginal change in trade into three“effects”:scale,compositionandtechnique(Grossmanand Krueger, 1993). The magnitude and sometimesthe direction of the individual effects depend on theparticular circumstances of each country, and mustthereforebedeterminedempirically.Thenet resultofthe threeeffectsprovides theoverall impactof tradeopening on the environment in a given economy. Inwhat follows, this framework is used to help uncoverthemain“drivers”underlyingtherelationshipbetweentrade and environmental conditions, a necessary firststep in examining the future evolution of thisrelationship.

Scale effect

Thescaleeffect refers to the increase in the levelofeconomicactivityduetotradeopening,anditsimpacton the environment. Unless production becomescleaner and less resource intensive, and consumerschangetheirbehaviour,forexamplebybecomingmorewilling to recycle waste, the increase in the level ofproduction, transport and consumption associatedwithtradeopeningleadstoenvironmentaldegradation.

The contribution of transport to the scale effect oftrade has received considerable attention. Reflectingthe heavy reliance of transport on petroleum as asource of energy, much of the attention has focusedontheimpactoftransportonclimatechange.Althoughthe bulk of international trade is transported by sea,whichisthemostefficientmodeoftransportintermsofcarbonemissionsandrepresentsarelativelyminorshare of worldwide carbon emissions, trade-relatedtransport activity is projected to increase sharplyduring the next few decades, as are emissions fromtransport.

Ithasbeenestimatedthatemissionsfrominternationalshipping represent approximately 3 per cent of worldCO2 emissions from fossil fuel combustion(InternationalTransportForum,2010).40Regardingthemore CO2-intensive modes of transport, theircontribution has been estimated at 1.4 per cent ofglobal carbon emissions from fuel combustion for airand17percentforroad.Thesenumbersoverestimatethe contribution of trade, given that they includeemissions generated by the transport of peoplebesides freight. In addition, the figure for roadtransport comprises both domestic and internationaltransport.

Relativetoemissionsgeneratedbytrade(i.e.emissionsfrom the production and transport of goods tradedinternationally), transport is estimated to representapproximatelyone-thirdofworldwidecarbonemissions


241

II D. Tr

aD

e o

pe

nn

es

s a

nD

Th

e b

ro

aD

er

so

cIo

-e

co

no

mIc

co

nTe

xT

(Cristea et al., 2011). This average masks largedifferences in the contribution of different economicsectors and countries to trade-related transportemissions.Forexample,theshareoftransport-relatedemissionsintotalemissionsfromexportsrangesfrom14 per cent for South Asia to 55 per cent for NorthAmerica (see Figure D.11). This largely reflects theheavy reliance of North American exports on air androad transport.41 Regarding economic sectors, theshareoftransport-relatedemissionsintotalemissionsfrom exports of agricultural, mining and other bulkproductsthatrelyonmaritimeshippingareoftenlessthan 5 per cent, compared with 75 per cent fortransportequipment,electronicequipment,machineryandsomemanufacturedgoods.

Between 2010 and 2050, carbon emissions frominternationalfreighttransportareprojectedtoincreasebyafactoroffouroutsidetheOECD,andbyafactorof1.5 in the OECD area, assuming that the modalcompositionremainsconstant(InternationalTransportForum, 2012). These projections highlight theimportance of multilateral efforts that seek tointernalize environmental costs in the prices ofinternational air and maritime transport, includingthroughtaxesandothermarket-basedmeasures.

The future evolution of the trends identified inSectionB.2mayaffecttheprojectedpatternsofCO2emissions from transport. For example, the growingimportanceof cross-regional insteadof intra-regionalpreferentialtradeagreementsshouldmaketrademoreglobalized, implying a shift towards more distanttrading partners. This could result in higher levels ofgreenhouse gas emissions from trade-relatedtransport, especially if accompanied by heavier

relianceonairinsteadofseatransport.Thespreadofregionalinsteadofglobalsupplychains,asisoccurringinAsia(seeSectionB.2(e)),wouldworkintheoppositedirection.

Composition effect

Thecompositioneffectreferstothechangesinrelativeprices and levels of pollution brought about by tradeopening.Tradeopeningcausessomesectorstoexpandand others to contract, in line with a country’scomparative advantage. Holding constant the scale ofeconomic activity and methods of production, tradeopening reduces domestic pollution if the expanding(export) sector is less pollution-intensive than thecontracting(import-competing)sector.Thecompositioneffectworksagainstthescaleeffectincountrieswithacomparativeadvantage incleansectors,while the twoeffects reinforce each other in countries with acomparativeadvantageindirtysectors.

What determines whether a country specializes inclean or dirty production? The so-called “factorendowments hypothesis”, which is based on theHeckscher-Ohlin model, predicts that trade openingwill cause capital-abundant (developed) countries tospecialize in theproductionofcapital-intensive (dirty)goods,whiledevelopingcountriesspecializeinlabour-intensive(clean)production.42

An alternative hypothesis, known as the “pollutionhaven hypothesis”, is based on the idea thatenvironmentalpolicyisthemainsourceofcomparativeadvantage. The hypothesis posits that trade openingwill lead to the relocation of pollution-intensiveproductionfromcountrieswithstringentenvironmental

FigureD.11:Transport-related emissions from exports, 2004 (percentage)

0

10

North America Europe East Asia MiddleEast/Africa

South America South AsiaOceania CentralAmerica

20

30

40

60

50

Transport emissions share Transport emissions intensity

Source:WTOSecretariat,basedonCristeaetal.(2011).

Note:“Transportemissionsintensity”referstotheshareoftransport-relatedemissionsintotalemissionsfromexports.Forperspective,thefigurealsodisplaystransportemissionsforeachregionasashareoftotaltransportemissions(“transportemissionsshare”).


242

policy to countries with relatively lax environmentalpolicy (Taylor,2005).43This implies thespecializationof developing countries (which are assumed to havelower than average levels of environmental policystringency) in dirty production, while developedcountries specialize in clean production.44 Whetherthe factor endowments hypothesis prevails over thepollutionhavenhypothesisorviceversainaparticulareconomyisanempiricalquestion(seeSectionD.2(c)).

Technique effect

The technique effect refers to the improvements inenvironmental conditions that result from trade-inducedchanges in themethodsbywhichgoodsandservices are produced.45 It implies a reduction in thepollution intensity of individual firms as a result oftrade opening (see Box D.4). Although the impact ofthe technique effect on the environment is alwayspositive, nothing in the literature suggests that thetechniqueeffectwillhappenbycompellingnecessity,orthatitsmagnitudewillbesufficientlylargetooffsetthe negative environmental impact of the scale (andpossibly, composition) effects. One reason that hasbeen advanced is that without proper incentives,private agents are unlikely to adopt the technologiesneeded to improve production methods (Copeland,2012). Given the key role of environmental policy inproviding incentives for the adoption of newtechnologies, it would appear that the magnitude ofthe techniqueeffectwill depend in largemeasureontheexistenceandadequate implementationof soundenvironmentalpolicy.

The economics literature has identified at least twowaysinwhichtrademayimproveproductionmethods.First, the increase in per capita income associatedwithopentradecangiverisetogreaterdemandbythepublic for a cleaner environment. Provided that thepolitical process is not “captured” by pollutingindustriesorotherwisecompromised, thedemand forimprovedenvironmentalqualityshouldresultinamore

stringent environmental policy that entices producerstoreducethepollution intensityofoutput (NordströmandVaughan,1995).

Secondly, eliminating tariffs and other trade barrierstendstoincreasetheavailabilityandlowerthecostofenvironmentally friendly technologies embodied inimported capital goods or in the form of knowledge-based processes diffused by the movement ofpersonnel. For example, a study cited in Dutz andSharma (2012) finds that if the top 18 developingcountriesrankedbygreenhousegasemissionswouldeliminate tariffs and non-tariff barriers on renewableenergygoods,their importswould increaseby63percentforenergyefficientlighting,23percentforwindpower generation, 14 per cent for solar powergeneration and close to 5 per cent for clean coaltechnology.

Thepotentialenvironmentalbenefitsoftradehighlightthe critical importance of the first ever WTOnegotiations on trade and environment. The mandateof these negotiations stipulates “the reduction, or asappropriate,eliminationoftariffandnon-tariffbarriersto environmental goods and services” (WTO, 2001).These negotiations could result in fewer and lowerbarrierstotradeinenvironmentalgoodsandservices,therebyimprovingmarketaccessconditionsworldwideto more efficient, diverse and less expensive greentechnologies embodied in such goods and services(seeSectionE.2).

A closely related issue is the possibility that foreigndirect investment (FDI) will result in “environmentalspillovers”. It has been argued that multinationalenterprises may impose particular environmentalrequirements on their supply chain subsidiaries andexternalsuppliers(forexample,duetoconcernsabouttheirreputationoreconomiesofscale),inducingthemto adopt environmentally friendly technologies(Albornoz et al., 2009). The movement of trainedworkers from foreign to domestic firms has been

BoxD.4: The technique effect in a world where trade is concentrated in a few global companies

Asdiscussed inSectionB.2(f), empiricalevidencesuggests that trade ismainlydrivenbya fewbigfirmsacrosscountries.ArecentstudybyKreickemeierandRichter(2012)explorestheimplicationsofthisfindingfortheenvironmentalimpactoftradethroughthetechniqueeffect.

Theauthorsdevelopatrademodelwithmonopolisticcompetitionandheterogeneousfirmsthatshowshowopenness to trade reallocates production from the least productive (and more pollution-intensive) firms,whichareforcedtoexitthemarketduetoincreasedcompetitionfromabroad,tothemoreproductive(andleastpollution-intensive)firms.The increase inaggregateproductivitycausedby tradeopening leads toareductioninoveralldomesticpollutioniffirm-specificpollutionintensitydecreasesstronglywithincreasingfirmproductivity.

ThelinkmadeinKreickemeierandRichter’smodelbetweentheproductivityoffirmsandtheirenvironmentalefficiency finds some support in the empirical literature. For example, Cole et al. (2008) use data on15industriesinChinaduringtheperiod1997-2003andfindevidenceofanegativelinkbetweentotalfactorproductivity and emissions of three air pollutants. Mazzanti and Zoboli (2009) find a positive relationshipbetweenlabourproductivityandemissionsefficiencyforseveraltypesofairemissionsinItaly.


243

II D. Tr

aD

e o

pe

nn

es

s a

nD

Th

e b

ro

aD

er

so

cIo

-e

co

no

mIc

co

nTe

xT

identifiedasanadditionalchannel throughwhichFDImay encourage the dissemination of environmentallyfriendlyknowledgeandtechnologies.

In sum, the scale-composition-technique frameworkhas revealedat least threedriversof the relationshipbetween trade and the environment that seem to beparticularlyimportant.Thefirstistheroleoftechnologyin minimizing or possibly offsetting any negativeenvironmental effects that occur through the scaleand possibly the composition effects. Adequatelydesigned and enforced environmental policy and anopen trade regime have been identified as keyconditions to facilitate the adoption of newtechnologies necessary to improve productionmethods.

The second driver is trade-related transport, and itscontribution to greenhouse gas emissions. Whetherthis becomes a pressure point for the multilateraltrading system depends in large measure on theresults of cooperative efforts to internalizeenvironmental costs in international transport prices,andonthefutureevolutionofthegeographicalscopeof preferential trade agreements and supply chains,among other factors. An additional driver pertains tothe significance of environmental policy (relative to“traditional” factors, such as endowments of capitaland labour) in determining countries’ comparativeadvantage in dirty or clean production. Derivingappropriate policy conclusions from this aspect ishamperedbythedivergenceofviewsonthedifficultyof analysing the relationship between environmentalpolicyandtrade(seeSectionD.2(c)).

(ii) What is the net effect of trade on the environment?

Starting with Grossman and Krueger (1993), severaleconometricstudieshaveexaminedtheenvironmentalimpactoftrademostlybyseekingtoassessempiricallythenetresultofthescale,compositionandtechniqueeffects(Antweileretal.,2001;ColeandElliott,2003).Theambiguityintheresultsofthisliteraturemaypartlyreflect differences in underlying conceptualframeworks, data sources and proxies, andeconometric methodologies. Broadly, the studiessuggest that totalpollutionmay increaseordecreasedependingonwhether the techniqueeffectoverridesthe scale effect. The type of pollutant is among thefactors that influence the net result. For a globalpollutantsuchasCO2,itappearsthatthescaleeffecttends to dominate the technique and compositioneffects.Forsomelocalpollutants,thetechniqueeffectis likely to exceed the scale effect. Moreover, somestudiesfinddifferences in the impactof tradeon theenvironment depending on countries’ income levels(Managi,2012).

Given that trade separates production andconsumption across space, a set of studies have

developed concepts describing how trade distributesenvironmental degradation between countries. Thesestudiesremainlargelydescriptiveanddonotimplyanycausalitybetween international tradeandevidenceofspecialization in “dirty” production. Because of largedatarequirementsintermsofcomparableinput-outputtables and environmental impact measures, theyusually consider a single country (Weber andMatthews, 2007; Jungbluth et al., 2011) or a smallgroupofcountries(Nakanoetal.,2009).Whileseveralof these studies assess the environmental impactsembodied in trade forairpollutantssuchasCO2andSO2 (Antweiler, 1996), others do this for water(Hoekstra and Hung, 2005), land use (Hubacek andGiljum, 2003), material extraction (biomass, fossilfuels,metals,andindustrialmineralsandconstructionminerals) (Bruckner et al., 2012), and pressure onbiodiversity (Lenzen et al., 2012). None of thesestudies takes into account the energy and emissionsassociatedwithinternationalfreighttransport.

Although empirical evidence suggests that mostdevelopedcountrieshaveincreasedtheirconsumption-based environmental impacts faster than theirdomestic environmental impacts, making developingand emerging countries net exporters of embodiedenvironmental impacts, there isnosinglepatternwithrespecttoembodiedenvironmentalimpacts(GhertnerandFripp,2007;PetersandHertwich,2008;Bruckneretal.,2012;DouglasandNishioka,2012;Lenzenetal.,2012; Peters et al., 2012). In fact, several developedand developing countries appear to be net exportersand importers, respectively. As noted, environmentalimpacts embodied in trade depend on many factors,includingfactorendowments,productiontechnologies,tradebalance,energyintensityandtradespecialization(Jakob and Marschinski, 2012). Institutional factorsmayplayanimportantroletoo.

Pressures on the environment caused by the rapidlygrowingweightofemergingeconomiesininternationaltrade has attracted much attention, and will probablycontinue to do so in the future. Looking ahead, acriticalissueseemstobetheextenttowhichthescaleeffectsassociatedwiththerapidrateoftradegrowthin emerging economies will be accompanied bychangesinproductionmethodsthatlowerthepollutionandenergyintensityoftheirproductionandtrade.Oneadditional issue is the extent to which FDI intoemergingeconomiesresults inpositiveenvironmentalspillovers. Some empirical studies have sought toexamine these factors with respect to particularcountries. For example, Dean and Lovely (2008) findthat China’s increasing participation in global supplychains generated composition and technique effectsthat may have played a role in reducing the pollutioncontentofitstrade.


244

(iii) How do environmental concerns affect trade?

In the absence of robust empirical evidence on theenvironmental effects of trade, there is a risk thatpublicperceptionsoftheenvironmentalconsequencesof trade will be shaped predominantly by narrowexamples of harmful effects, which tend to carrystrong emotional appeal and attract considerablemedia attention. This raises concerns that theperceivednegativeimpactoftradeontheenvironmentcould exacerbate existing protectionist sentimentcaused by economic uncertainty and the perceivedcontribution of trade openness to growingunemployment and income disparities within andacross countries (see Section D.1(b)). The followingdiscussion reviews the available evidence on publicperceptions of the environmental consequences oftrade,andtheirpossibleinfluenceontradepolicy.

Data on how individuals perceive the environmentaleffectsoftradearescarce.A2007surveyofattitudesabout globalization in 18 economies (comprisingaround56percentoftheworld’spopulation)revealssignificant differences in attitudes across countries,with individuals inopeneconomiesdisplayingamorepositive perception about the environmental impactsof trade than those in less open ones (seeFigure D.12). Differences across countries are alsoevident in a survey of the 27 EU member states(European Commission, 2010). The share ofrespondents in the EU who consider environmentalharmtobeoneofthemajornegativeconsequencesofinternationaltraderangedfrom4percentinBulgariato 42 per cent in Austria. For the EU as a whole,environmental harm ranks well below unemploymentamong the perceived harmful consequences ofinternationaltrade.

Relatively few studies have dealt with the interactionbetweenenvironmentalconcernsandpublicattitudestowardstrade.Amongthem,Bechteletal. (2011)findthat individuals in Switzerland who express a higherlevelofconcern for theenvironment tend toperceiveglobalization negatively, favour the use of trade-restrictive measures to protect jobs endangered byimport competition, and consider factors that gobeyondpriceandqualityintheirpurchasingdecisions.Additional empirical research is needed to determinewhethertheseresultsapplymorewidely.

Ina relatedsetof studies,politicaleconomistshavedrawnalinkbetweenthegrowingtendencytoincludeenvironmental provisions in regional tradeagreementsand individuals’environmentalconcerns.Specifically,thisliteraturearguesthatsomecountriesmayberelyingonenvironmentalprovisionsinregionaltrade agreements to appease voters who fear thenegativeeffectsoftradeontheenvironment(Bechtelet al. , 2011). This argument would seem to beparticularly relevant for one type of environmental

provisionfoundinagrowingnumberofregionaltradeagreements, namely the commitment by the partiesto not lower environmental standards as a meanstogaina tradeadvantageorattract investment (seeBoxD.5).

Theoverallconclusionsinthisstrandoftheliteraturemay be applicable to other features of trade policybesides environmental provisions in regional tradeagreements.Forexample, theproliferationofprivatestandards could be seen as an indication of theinfluence on trade policy of concerns about theenvironmentaleffectsoftrade.

Privatestandardsaredevelopedbynon-governmentalentities, including businesses, trade associations,consumer groups and standardizing bodies, inresponse to rising consumer demand for informationabout the environmental and other characteristics ofgoodsandservices.Theyseektostrengthenmarketsforgoodsandserviceswhoseenvironmentallyfriendlyattributesare “invisible” toconsumers.Whilesomeofthesestandardsfocussolelyonenvironmentalcriteria(related to a good or the way it is produced), othersmayincorporatefoodsafetyandsocialcriteria.

Privatestandardshavebeendevelopedwithrespecttoseveral environmental indicators, including carbon

FigureD.12:Public perceptions about the environmental impact of trade (percentageofrespondentsthatagreewiththestatementshownontheverticalaxis)

0

10

0 40 60 80 120 140 160

20

30

40

60

80

Trade openness (trade to GDP ratio – %)

50

70

20

Trad

e is

goo

d fo

r the

env

ironm

ent

100

United States

Argentina

India

Armenia Mexico

Korea, Rep. of

Ukraine

Thailand

China Poland

Israel

France

Russian Federation

Source:Authors’calculationsbasedontheChicagoCouncilonGlobalAffairsandWorldPublicOpinion.org(2007).

Note:Theverticalaxisshowstheshareofsurveyrespondentsthatanswered“good”tothefollowingquestion:“Overall,doyouthinkthattradeisgoodorbadfortheenvironment?”,weightedbytheshareofpopulationthatansweredeither“good”orbad”.


245

II D. Tr

aD

e o

pe

nn

es

s a

nD

Th

e b

ro

aD

er

so

cIo

-e

co

no

mIc

co

nTe

xT

BoxD.5: Environmental provisions of regional trade agreements

PriortotheentryintoforceoftheNorthAmericanFreeTradeAgreement(NAFTA)in1994,environmentalprovisionsinregionaltradeagreements(RTAs)wererelativelylimitedinscope.ReferencestotheenvironmentappearedmostlyinthepreambletoRTAsandtherulesongeneralexceptions,largelymirroringtheapproachfollowedatthemultilaterallevel.

SinceNAFTA,thescopeofenvironmentalprovisionsinRTAshasexpanded.Ofthe194RTAsthathavebeennotifiedtotheWTO(uptoend-2010),closeto60percentcontainenvironmentalprovisionsotherthanthosein the preamble or the general exceptions. Of the RTAs with more extensive environmental provisions,55 per cent involved RTAs between developed and developing countries, 38 per cent RTAs betweendevelopingcountries,and7percentRTAsbetweendevelopedcountries.

FigureD.13:Environmental provisions of regional trade agreements

RTAs with only environmentalexceptions/rights, 75, (36%)

RTAs with environmental provisions beyond exceptions/rights, 131, (62%)

RTAs with only environmentalexceptions/rights and preamblelanguage, 10, (8%)

RTAs with environmental provisionsbeyond exceptions/rights and preamblelanguage, 121, (92%)

RTAs with environmental provisionsbut no environmental

exceptions/rights, 1, (0.5%)

RTAs with no environmentalprovisions, 4, (2%)

Source:WTOSecretariat.

The expansion of the “environment dimension” of RTAs notably comprises the decision to includecommitmentsonthelevelandstringencyofdomesticenvironmentallawsandstandards.OfallRTAsnotifiedto the WTO, 46 contain such provisions. These provisions take the form of commitments to not lowerenvironmental standards as a means to attract investment, to enforce domestic environmental laws andstandardseffectively,toraiseenvironmentalstandardsoreventoharmonizethemamongthepartiestotheRTA.AlthoughprovisionsonenvironmentallawsandstandardsaremostcommonlyfoundinRTAsbetweendeveloped and developing countries, several agreements between developing countries contain suchprovisionstoo.

Cooperation on environmental matters is an additional subject often covered in RTAs. Some 70 notifiedRTAscoverenvironmentalcooperation.Thereissignificantvariationinapproachestothisissue.WhilesomeRTAs outline general principles, others identify specific issues or sectors for cooperation. Provisions onenvironmental cooperation found in RTAs between developed and developing countries tend to focus onbuilding capacity and strengthening the design and implementation of environmental laws, while RTAsinvolvingonlydevelopingcountriesstresstheneedtotacklecommonenvironmentalproblems.

Parties to RTAs sometimes establish institutional arrangements specifically targeted at facilitating theimplementationofenvironmentalprovisionscontained in theagreement.Thesearrangementsmay includereview and monitoring bodies, dialogue and consultation mechanisms, or formal dispute settlementprocedures.Increasingly,RTAsprovideopportunitiesforpublicparticipation(e.g.accesstoinformationanddocuments, representation in committees, and submissions on enforcement matters) as part of theimplementationarrangementsrelatedtoenvironmentalprovisions.

footprints, food miles and embodied water use. Theyspan a widening range of products, including food,householdappliances, forestryproductsandservicessuchastourism.Althoughthesestandardsarecastas“voluntary” in nature (because they are imposed byprivateentities),theymayneverthelesshavesignificant

impactsontrade,anissuethathasbeenofparticularconcerntodevelopingcountries(WTO,2012b).

Insum,thereisinsufficientevidencetodrawadefinitiveconclusionaboutthefuture interactionbetweenpublicperceptions of the environmental consequences of


246

trade on the one hand and trade policy on the other.More research is needed on the question of howwidespread negative perceptions are about theenvironmental effects of trade, and whether suchperceptionscouldresultinaprotectionistbacklash.

Pending more robust empirical results, the availableevidence would seem to suggest that people’sconcernsabouttradeandtheenvironmentdoexertaninfluence on trade policy, albeit in other ways thanoutright protectionism. For example, concerns abouttheenvironmental impactoftradearereflectedinthegrowing tendency to includeenvironmentalprovisionsin regional trade agreements; such concerns couldalso be seen as partly fuelling the proliferation ofprivate standards that seek to respond to risingconsumerdemandforinformationontheenvironmentalandothercharacteristicsofgoodsandservices.

(c) Environmentalpolicyandtrade

Akeyaspectoftherelationshipbetweentradeandtheenvironmentistheimpactoftheenvironmentontradethrough the channel of environmental policies. Whenreferring to “stringent” environmental policies, thefollowingdiscussionassumesthatthestringencylevelin law is matched by correspondingly stringentenforcement. In practice, this may not always be thecase,asadequateimplementationmaybeimpairedbyweak institutional capacity in many countries(PoelhekkeandPloeg,2012).

The interaction between environmental policy andtrade is probably two-way. In other words, efforts totackle impact on the environment throughenvironmentalpoliciesmayinfluencethedirectionandcomposition of trade flows, while openness to trademayaffect thewillingnessandabilityofgovernmentsto adopt environmental policies. The complex andmultidirectional interaction between environmentalpoliciesandtradeposeschallengesforthemultilateraltradingsystem.Akeyquestionfromtheperspectiveofthisreportiswhetherthesechallengeswillintensifyinthe future. The following discussion considers thisquestion by examining two sets of policies related toclimate change. Given that concerns aboutcompetitiveness strongly permeate policy andacademicdiscussionsontheeffectsofenvironmentalpolicies and trade, we begin with an overview of thisissue.

(i) The competitiveness consequences of environmental policy

Whenevaluatingtheimpactofenvironmentalpolicyontrade, a useful starting point is to recognize that themagnitude of the impact can vary quite drasticallydepending on whether the measure in question istargeted at pollution from consumption or production(Copeland, 2012). An environmental policy measurethat is targeted at pollution from consumption will

normallyincreasecostsforbothdomesticandforeignproducers. One example would be a requirementsetting a minimum level for the energy efficiency ofhousehold appliances. If the cost of complying withsuch a requirement is higher for domestic producersthan foreign producers, imports of householdappliances would increase and domestic productiondecrease.

By contrast, environmental policy measures that aretargeted at pollution from production instead ofconsumption(e.g.alimitonwaste-waterdischargestosurface waters by domestic producers of chemicals)could negatively affect the competitiveness ofdomesticproducersbydrivinguptheircostsrelativetoforeign producers. This could result in a decrease inoutput of domestic chemicals, or put domesticproducers of chemicals at a disadvantage relative toforeignproducers.

Much of the economics literature assumes thatenvironmentalpoliciesentailcostsforparticularfirmsand sectors (Pasurka, 2008). However, someobservers have argued that these costs need notalways result in competitiveness losses for affectedfirms and industries.46 Porter and Linde (1995)postulate thatproperlydesignedenvironmentalpolicycan lead to “innovation offsets” that will not onlyimprove environmental performance but also partiallyorfullyoffsettheadditionalcostofthosepolicies.ThisisknownasthePorterhypothesis,whichhassparkedan abundant empirical literature. Following Ambec etal. (2011), it appears that the “weak” version of thehypothesis(i.e.stricterpolicyleadstomoreinnovation)isfairlywellsupportedbythedata,whiletheempiricalevidence on the “strong” version (i.e. stricter policyenhancesbusinessperformance)ismixed.

Assessing the competitiveness consequences ofenvironmentalpolicy is fraughtwithdifficulties,partlybecauseofthelackofdataonthecostsofcompliancewith such policies. The United States has publishedsome relevant data, which suggests that the directcost of pollution control is relatively minor. In 2005(latest year available), US industry spent close toUS$21billiononoperationalcoststoreducepollution.For most industries, pollution abatement operationcosts represent 0.5 per cent or less of revenue (seeFigureD.14).This sharecan reachup to1.1per centfor pollution-intensive industries, such as primarymetal and paper. Regarding capital expenditures toreduce pollution, the share of these expenditures intotal new capital expenditures is usually below 5 percent except for a handful of pollution-intensiveindustries, such as paper, petroleum and coal,chemicals and primary metals. The amount spent byUS industry on pollution abatement capitalexpenditures in 2005 totalled around US$ 6 billion,slightly less than one-third of the total spent onpollutionabatementoperatingcoststhatsameyear.


247

II D. Tr

aD

e o

pe

nn

es

s a

nD

Th

e b

ro

aD

er

so

cIo

-e

co

no

mIc

co

nTe

xT

Economistshavesoughttoassessthecompetitivenessconsequences of environmental policy by testingwhether the “pollution haven” hypothesis holds inpractice. As noted, the pollution haven hypothesispredicts that trade (and capital) openness results inthe relocation of pollution-intensive production fromcountries with stringent environmental policy tocountries with lax environmental policy. Although thehypothesis is relatively simple, empirical studiesdesigned to test it have yielded conflicting evidence,partly reflecting the use of different conceptualframeworks, data sources and proxies, andeconometric methodologies. Following Copeland andTaylor (2004), recent studies in this area have foundthat differences in the degree of stringency ofenvironmental policies tend to influence thedistribution of “dirty” production across countries,suggesting that more stringent environmental policyhas a “deterrent effect” on the production of “dirty”goods. Nonetheless, there is no robust evidence thatthis deterrent effect is “strong enough to be theprimary determinant of the direction of trade orinvestmentflows”.

This general result implies that there may be negativecompetitivenesseffectsassociatedwithmorestringent

environmental policy for particular firms and sectors,depending on their pollution intensity and degree ofgeographicalmobility,amongotherfactors.Thisraisesthe possibility that affected firms and sectors (andpossibly others) may appeal to (real or perceived)competitiveness concerns when lobbying againstenvironmental policies. Governments may in turnrespondbyincorporatingtrade-restrictiveelementsintoenvironmentalpolicies(“greenprotectionism”)asawayboth to compensate affected sectors and overcomeresistancetoenvironmentalpolicyreform.

Ithasbeensuggested that thegrowing fragmentationof the production process along global supply chains(see Section B.2(e)) could ease competitivenessconcerns associated with stringent environmentalpolicy.Withglobalsupplychains, it ispossible foronlycertainpartsoftheproductionprocesstoberelocatedin response tostringentenvironmentalpolicyathome.UsingalargesampleofJapanesemanufacturingfirms,Coleetal.(2011)findevidencethatfirmsoutsourcethedirtypartoftheirproductionprocess.Theysuggestthat“increases in environmental regulations can increasethe health of local citizens without the massive joblossesassociatedwithwholesale relocationorclosurepredictedbyindustrylobbygroups”.

FigureD.14:Pollution abatement costs in the United States, 2005 (percentage)

0.0 2.0 4.0 6.0 8.0 16.0

Furniture

Transport equipment

Electrical equipment

Computer and electronic

Machinery

Fabricated metal

Primary metal

Non-metallic mineral

Plastics and rubber

Chemical

Petroleum and coal

Printing

Paper

Wood

Leather and allied

Textile product mills

Textile mills

Beverage and tobacco

Food manufacturing

All industries

10.0 12.0 14.0

PAOC PACE

Source:Authors’calculations,basedonUSCensusBureau(2008).

Note:PAOC–Pollutionabatementoperationcosts–asashareofthevalueofshipments.PACE–pollutionabatementcapitalexpenditures–asashareofnewcapitalexpenditures.


248

Nonetheless,theprospectofgovernmentsusinggreenprotectionisminexchangeforpoliticalsupportformorestringent environmental policies remains a possibility,which could, if realized, complicate future efforts toopentrade.Asillustratedinthefollowingdiscussionofthetradeimplicationsofspecificenvironmentalpolicies,muchwilldependonwhethercompetitivenessconcernsassociated with environmental policy promptgovernmentstoseekcooperativesolutionstocommonenvironmental problems instead of resorting to greenprotectionism.

(ii) Interaction between environmental policies and trade

Coping with carbon leakage

Some of the more complicated and contentiousenvironmental issues are global in nature – that is,they involve countries whose economic activitiespollute or reduce a common resource, damaging allothercountries.Theabsenceofcooperativesolutionstocorrectthesecross-bordereffectsposeschallengesfor tackling global environmental problems andmanaging the interface between environmentalpoliciesandtrade.Challengesareparticularlyevidentintheareaofclimatechange,wheretheemergenceofa patchwork of regional, national and sub-nationalclimate change regimes may lead to concerns aboutthe loss of competitiveness of certain firms andsectors, and the possibility of “carbon leakage”. Thisreferstoasituationinwhichreductionsofgreenhousegas emissions by one set of countries (“constrained”countries) are offset by increased emissions incountries which do not take mitigation actions(“unconstrained” countries). Openness to trade andinvestment are two of the channels through whichcarbonleakagecanoccur.47

Untilnow,therehasbeennorobustevidenceofcarbonleakage,inpartbecausemanycarbonpricingpoliciesare relatively recent, and carbon prices relatively low.The available literature, most of which relies oncomputablegeneralequilibriummodels,suggeststhatcarbonleakageismodestoverall.48Nonetheless,withhigh carbon prices, leakage would become relativelylarge for sectors that are energy and emissionsintensiveandexposedtointernationaltrade.

Most governments that have put in place carbonemissions trading schemes have sought to allaycompetitivenessfearsandreducetheperceivedriskofcarbon leakage by allocating emissions allowancesfreely to participants.49 Examples include Australia,the European Union, the Republic of Korea and NewZealand.Akeyissuewithimplicationsforthefutureoftrade and the multilateral trading system is howcountries will manage the perceived threats ofcompetitiveness losses and carbon leakage duringfuture implementation stages of their emissionstrading schemes. Among the possible instruments to

manage these twin threats are border carbonadjustments, which involve the extension of carbonpricing to imports and have therefore receivedsignificantattentionintradepolicycircles.

Nocountrywithanemissionstradingschemehasyetput in place border adjustments but some proposalshavebeenconsidered.Mostofthemfocusonimportsandwouldtaketheformofataxonimportedgoods,ora requirement for importers to purchase emissionpermitsorallowancesfortheirimports.Ontheexportside, border adjustments could take the form of anexport rebate, where exporters shipping items tounconstrainedcountriesarecompensatedforthecostofcomplyingwithemissionrequirements.

Although border adjustments may be justified as asecond-best measure to complement carbon pricingschemesifnoagreementonglobalcarbonpricingcanbe reached, their practical implementation may be asource of trade friction. There are several practicaldifficulties involved in the implementationofabordertaxadjustmentinrelationtoacarbonpricingscheme,and further difficulties in designing a mechanism toadjust the cost of emission allowances and calculatetheproperlevelofborderadjustment(WTOandUNEP,2009). The main challenges relate to the difficulty inassessing product-specific emissions and thefluctuationsofthecarbonprice(orallowanceprice)inthe context of an emissions trading scheme. Anadditionaldifficultymayariseincaseswhereimportedproductsaresubject, inthecountryoforigin,tootherclimate change regulations, such as technicalregulations, rather than price mechanisms such astaxes.Compliancewithcertain regulations,suchasafuel efficiency standard, may involve a cost (e.g.investmentinmoreenergy-efficienttechnologies)thatmay be complex to evaluate and transform into anadjustablepriceora“comparableaction”.

Furthermore,although there iswidespread interest inreducing carbon leakage and countries can haveenvironmental reasons for using trade measures toprevent such leakage, there is a risk that thesemeasures may be used to manipulate the terms oftrade and protect domestic producers. The possible“dual use” of anti-leakage policies may blur thedistinction between policies that seek to pursue alegitimate policy objective (e.g. tackling climatechange) and those that are used as a means ofsupporting competitively challenged domestic firmsandindustries.Thisunderlinesthepracticalchallengeofdistinguishingbetween“legitimate”andprotectionistmotivations for anti-leakage measures and ofidentifyinginstanceswheretheycreatetradebarriers.

Promoting green technologies

Agrowingnumberofgovernmentshaveputinplaceorare considering incentives for green technologies, inparticular renewable energy. Some of these policies


249

II D. Tr

aD

e o

pe

nn

es

s a

nD

Th

e b

ro

aD

er

so

cIo

-e

co

no

mIc

co

nTe

xT

(often referred to as “market-pull policies”) seek tocreate demand for these technologies while othersaim to increase supply or foster innovation (so-called“technology-push policies”). Common market-pullinstrumentsincludequotasystems,feed-intariffsandpremiums, and a wide range of tax incentives.50Technology-push policies usually take the form ofsupportforresearchanddevelopmentatearlystagesof innovation, and for product development andmanufacturingatlaterstages,mostlythroughfinancialandtaxincentives.

Incentives to promote innovation and adoption ofrenewableenergytechnologieshavebeenjustifiedonthe basis of particular hurdles facing renewableenergy, including those related to the energy marketstructure, infrastructure, learning curves and futureclimate policy uncertainty (Popp, 2012; Serres et al.,2010). Nonetheless, the empirical evidence on theeffectiveness of incentives for renewable energyremainsrelativelylimited,inpartduetoissuesofdataavailabilityaffectingsomerecentlyenactedincentives(Fischer and Preonas, 2010). A study assessing therole of incentives in promoting the deployment ofrenewableenergytechnologiesin35countriesduringtheperiod2000-05foundthatonlyasmallsubsetofcountrieshadimplementedpoliciesthatsucceededinaccelerating deployment during that period(International Energy Agency, 2008). Several otherstudies on the effectiveness of incentives forrenewable energy focus on selected instruments,limiting theextent towhichpolicyalternativescanbecompared(e.g.Klaassenetal.,2005).Moreover,someof the studies that examine the experience withincentives tend to avoid the question of whether thebenefits of these policies exceed their costs (e.g.Buen,2006).

Themotivationsforrenewableenergyincentivesreachwell beyond purely environmental concerns.Governments increasingly cast renewable energyincentives within the broader framework of “greencompetitiveness”–thatis,astoolsnotonlytoachieveenvironmental (e.g. climate change mitigation) goalsbut also to stimulate economic growth, spur jobcreation and promote exports and diversification. Forexample, in 2007, the European Union put in placepolicies to increase theshareof renewableenergy intotal energy consumption as a way to diversify EUenergy supply and create new industries, jobs,economicgrowthandexportopportunities (EuropeanCommission,2012).

The Republic of Korea has identified several energytechnologies as “new growth engines” in its NationalStrategy for Green Growth. The decision whether toincludeaparticulartechnologyinthelistwasbasedonits potential contribution to economic growth andenvironmental sustainability, and its “strategicimportance” (OECD, 2010b). The 27 technologiesselectedarethefocusofincreasedpublicspendingon

research and development. Under its strategicroadmap for 2011-15 (known as the 12th Five-YearPlan),Chinahasidentifiedforsupportseveralstrategicexport industries, including clean energy technologyandcleancars.

Therapidproliferationofrenewableenergyincentivesin both developed and developing countries and thegrowing tendency to link these measures to greencompetitiveness pose significant challenges for themanagement of the interface between trade and theenvironment. Although the trade effects of incentivemeasures will vary according to the magnitude ofsupport provided and the measure’s design features,including the question of how close the supportedresearchorgoodsaretocommercialapplication,sometypesofincentivepoliciesmayassistdomesticfirmsintaking market share and profits away from moreefficient foreigncompetitorsormayotherwisedistorttrade so the risk of regulatory “capture” cannot beeasilydismissed(WTO,2012b).

The intertwining of environmental and greencompetitiveness objectives could increase thevulnerability of renewable energy incentives topowerful lobbies and rent-seeking behaviour. It couldalso result in flawed policy design due to theinsufficient information toachievemultiple (andoftenvaguely defined) policy objectives (World Bank,2012b). Some evidence lends support to concernsabout the possible influence of political economyconsiderations in the design of renewable energyincentives.Forexample,Fischeretal. (2012)conductsimulations of the electricity sector and find that themagnitude of observed renewable energy technologyincentives exceeds, probably by a wide margin, whatwouldbejustifiedbythepositiveeffectsgeneratedbyresearch and development or learning by doing. Thisresult holds even assuming high spillover rates intoother areas from learning by doing. Other work findsthat certain renewable energy policies are driven by,among other factors, the presence of well-organizedinterest groups instead of environmentalconsiderations(LyonandYin,2010).

A relatedconcernabout renewableenergy incentivesstems from the decision by some countries toincorporate into such incentive schemes trade-restrictivepolicies,suchaslocalcontentrequirements.For example, under some national or sub-nationalprogrammes, participation in a feed-in tariff scheme(or theavailabilityofadditionalbenefitsundersuchascheme) is contingent on the use of domesticallymanufactured components of renewable energytechnologies. These local content requirements maychannel the additional demand for inputs created bytherenewableenergyincentivetodomesticproducers,possibly at the expense of more efficient foreignproducers,causingdistortionstointernationaltrade.


250

(d) Conclusions

Tradeopennessandenvironmentalprotectionarekeyelements of sustainable development, and policies inboth fields should work to utilize existing resourcesbetter. There is no inherent conflict between tradeopenness and environmental sustainability and therearemanyinstanceswheremeasuresthatpromoteoneobjectivepromote theother too. Thiswas recognizedatthetimeoftheWTO’screationin1995,whenWTOmembersstressedthattradeandtheenvironmentcanand should be mutually supportive, and that tradepolicies should be conducted in accordance withsustainable development objectives and should seektoprotectandpreservetheenvironment.

Tradeandtheenvironmentinteractincomplexways,withmultiplefeedbackeffectsbetweenthem.Ifnotmanagedcarefully, the multifaceted relationship between thesetwo interconnected systems may give rise to tensions,which can weaken the positive contribution of trade toeconomicgrowthandsustainabledevelopment.

The dramatic increase in world trade during the pastthree decades, which has coincided with a period ofsignificant environmental transformations andproblems, has drawn attention to the growing scaleeffects of trade. Most of this attention has beendirected at large emerging economies, given theirrapidly growing weight in international trade.Conclusive empirical evidence on the global, averageor net effect of trade opening on the environment,however,remainselusive.

Looking ahead, there remain many unexploitedopportunities tobolster theenvironmentalgains fromtrade. Trade has the potential to prompt changes inproductionmethods,whichoffsetthescaleeffectsoftrade.However,theso-calledtechniqueeffectmaybecontingent on many conditions, including an opentraderegime,soundenvironmentalpoliciesandsoundinstitutions. This highlights the importance andurgency of the first-ever multilateral negotiations ontradeand theenvironment,whereWTOmembersareseeking to reduce or eliminate the barriers affectingtradeingreengoodsandservices.

Transport, which is an additional driver of the scaleeffects of trade, has come under increased scrutinytoo, largely because of its contribution to carbonemissions. Although the bulk of trade relies onmaritimetransport,whichisthemostefficientmodeoftransportation in terms of carbon emissions, trade-related transport is projected to increase sharplyduringthenext fewdecades,asare transport-relatedemissionscosts.Theseprojectionsmaybeaffectedbythe future importance of cross-regional relative tointra-regional preferential trade agreements and theevolutionof international supply chains.Anadditionalfactor tobeconsidered is theoutcomeofmultilateralefforts to align the private and social costs of

international transport, including through market-basedmeasuresthat incorporateenvironmentalcostsininternationalairandmaritimetransportprices.

In principle, the scale effects of trade could play animportant role in shaping public perceptions andattitudes towards trade, thereby heighteningprotectionist sentiment. Nonetheless, drawing adefinitiveconclusiononhowpublicperceptionsrelatedto the environment are affecting trade must awaitadditional research. The available evidence wouldseem to suggest that individuals’ concerns about theenvironment do exert an influence on trade policy inwaysotherthanprotectionism.Forexample,concernsabout the environmental impact of trade have beenimplicated in the growing tendency to includeenvironmentalprovisionsinregionaltradeagreementsorintheproliferationofprivatestandards.

Besidesthescaleeffectsoftrade,thecompetitivenesseffects of environmental policy are sometimesperceivedasholdingbackenvironmentalpolicyreform.Thereisnorobustempiricalevidenceontheso-calledpollutionhavenhypothesis.Still,environmentalpoliciesinevitablyaffectproductionandconsumptionpatterns,and may therefore have adverse effects on thecompetitiveness of particular firms or sectors, whichmaythenappealto(realorperceived)competitivenessconcerns when lobbying against environmentalpolicies. Governments, in turn, may respond byincorporating trade-restrictive elements intoenvironmentalpolicies.

In an effort to boost “green competitiveness”, agrowing number of governments have put in placeincentive packages for green technologies, with afocus on renewable energy. From the perspective offuture trade and the multilateral trading system, theriskisthattheintertwiningofenvironmentalandgreencompetitiveness objectives may increase thevulnerability of renewable energy incentives topowerful lobbies. It may also result in flawed policydesign due to the lack of sufficient information toachieve multiple (and often vaguely defined) policyobjectives.Thiscouldexacerbatethepossibleadversetrade effects of some types of incentive measures,andunderminetheirenvironmentaleffectiveness.

Global environmental problems are likely to poseadditional challenges for trade and the multilateraltradingsystem.This isparticularlyevident in theareaof climate change, where the emergence of apatchwork of regional, national and sub-nationalclimate change regimes may lead to concerns aboutthe loss of competitiveness of energy-intensive andtrade-exposed firms and sectors, and the relatedpossibility of “carbon leakage”. Although borderadjustments may be seen as a second-best measureto prevent these effects, if no agreement on globalcarbon pricing can be reached, their practicalimplementation may be a source of trade friction. In


251

II D. Tr

aD

e o

pe

nn

es

s a

nD

Th

e b

ro

aD

er

so

cIo

-e

co

no

mIc

co

nTe

xT

addition, there is a risk that these measures may beusedformanipulatingthetermsoftradeandprotectingdomesticproducers.Thepossible “dualuse”of thesemeasures underlines the practical challenge ofdistinguishing between “legitimate” and protectionistmotivations for anti-leakage measures and ofidentifyinginstanceswheretheycreatetradebarriers.

The individual and collective decisions by openeconomiesinmanagingtherelationshipbetweentradeand theenvironment carry significant implications forthe future of international trade and the WTO.The future evolution of this relationship will dependin large measure on how governments respondto competitiveness and other concerns related toenvironmental pressures and policies. Collectiveeffortsthatresultinagreedpolicyapproachestowardsglobalenvironmentalproblemsmaylimitthescopefornegative competitiveness consequences, therebyreducing the risk that individual countries will favourprotectionist policies. This underlines the importanceof improved multilateral cooperation at the WTO asmuch as within the global environmental governanceregime.

3. Macroeconomicandfinancialconcerns

As destabilizing as they may be in the short run,macroeconomicandfinancialshocksonlyaffectlong-termtrendswhentheunderlyingfactorsofgrowtharenegativelyanddurablyaltered,forexampleintheformofseverechangesinthesupplyof labourandcapital.The2008-09financialcrisis,anditsrelativesimilarityto the 1929 crisis (a financial crisis having globaleffects on production and trade), is a reminder thatmacroeconomicandfinancialshockscanhavestrong,recurrent and global, if not durable, effects on tradeandgrowth.

TwochannelsofparticularinteresttotheWTOthroughwhich macroeconomic and financial shocks influenceinternational trade are credit crunches, which reducethe amount of finance available to traders, andexchange rate shifts, which divert trade flows andstraintraderelations.Whileexchangerateseventuallyadjust and credit crunches are generally associatedwith the “purge” of over-leveraged financial sectors,they may nonetheless derail both the trend and therateofexpansionofworld trade.Thissectionreviewstheproblemsthatexcessexchangeratevariabilityandshortages of trade finance pose to trade expansionanddiscusseshowtheymaycontributetoshapingthemacro-financial environment impacting trade in thefuture.

(a) Tradefinance

Financeisthe“oil”ofcommerce.Mosttradetransactionsaresupportedbyshort-termtradecredit–accordingto

IMF-BAFT(2009),80percentofalltradefinanceisintheformofeitherstructuredfinance(lettersofcreditorsimilar commitments, using the merchandise ascollateral) or open account liquidity supplied againstreceivables.51 In principle, while the commercial risksinvolvedinaninternationaltradetransactionseemtobelargerthaninadomestictradetransaction(riskofnon-payment, risk of loss or alteration of the merchandiseduring shipment, exchange rate risk), trade finance isgenerally considered to be a particularly safe form offinance, as it is underwritten by strong collateral anddocumentedcreditoperations.

AccordingtotheInternationalChamberofCommerce’s(ICC’s) “trade finance loss register”, the averagedefault rateonshort-term international tradecredit isno larger than 0.2 per cent, of which 60 per cent isrecovered (ICC, 2011). Despite trade finance being aroutine task, it is vital for trade. Until the financialcrisesofthe1990sandof2008-09,tradefinancehadbeen taken for granted. The crisis periods createddistortions in the trade finance market which madepolicy interventions necessary. Below is a discussionofthelinkbetweentradeandtradefinanceandofkeydriversoftrendsintradefinance.

(i) Clarifying the link between trade finance and trade

During the Asian financial crisis, policy institutions,suchastheIMFandtheWTO,hadrevealedelementsofmarketfailureinexplainingthetradecreditcrunch.Theseincludedherdbehaviour,increasedgapbetweenthe level of risk and its perception, marketconcentration and confusion between country andcounterpartyrisk(IMF,2003;WTO,2004b).Academicresearchontheroleoftradefinancehasgrowninthecontext of the 2008-09 financial crisis and thesubsequent economic downturn, when global tradeoutpacedthedeclineinrealGDPbyafactorof12–afigure much larger than anticipated under standardmodels.AssummarizedbyEichengreenandO’Rourke(2012), “the roots of this trade collapse remain to befullyunderstood,althoughrecentresearchhasbegunto shed light on some of the causes (see Baldwin,2009; Chor and Manova, 2012)”. While most authorsagree that the fall in demand has been largelyresponsible for the decline in trade flows, the debatehas focused on the extent to which other potentialfactors, such as trade restrictions, a lack of tradefinance,verticalspecializationandthecompositionoftrade,mayhaveplayedarole.

Empirical work on trade finance has been limited bythe lack of a comprehensive dataset, despite theavailability of qualitative information provided bysurveys on market trends and structure (ICC 2009;IMF-BAFT2009).However, progresshasbeenmadeinhighlightingsomelinksbetweenfinancialconditions,tradecreditsandtradeatthefirmlevel.


252

Amiti and Weinstein (2011) establish causalitybetween firms’ exports, their ability to obtain creditand the health of their banks. Using firm-level datafrom1990to2010,theysuggestthatthetradefinancechannel accounted for approximately 20 per cent ofthe decline in Japan’s exports during the financialcrisis of 2008-09. The authors show that exportersare more reliant on trade credit and guarantees thandomestic producers, and that firms working withtroubled banks saw their foreign sales drop by morethanthatoftheircompetitors.Multinationalenterprisesseem less affected, notably because a large part ofmultinational trade is intra-firm, which exhibits lessrisk. Multinationals are also able to optimize theproduction-to-tradecycle, therebyminimizingworkingcapitalneeds:theshorterthelagbetweenproductionandpayment,thelessfinanceisaproblem.

In the same vein, Bricongne et al. (2012) find thatsectors highly dependent on external finance havebeen most severely hit by the financial crisis andexperiencedthelargestdeclineintheirexportactivity.Usingmonthlydata for individualFrenchexportersatthe product and destination level, the authors alsofoundthatbothsmallandlargefirmshadbeensimilarlyaffectedbythecrisis.UsingdataonUSimports,Chorand Manova (2012) find that credit conditions wereone channel through which the crisis led to thecollapseintrade.Countrieswithtightercreditmarkets,measured by their interbank interest rates, exportedless to the United States during the financial crisis.This effect was especially strong for financiallyvulnerableindustries.Theseindustries,categorizedbyChor and Manova (2012) as those that requireextensive external financing, had limited access to

tradecredit.Accesswasespecially limitedduring thepeak of the financial crisis. Some studies, however,have not found any significant role played by tradefinance in the “Great Trade Collapse” (for example,Paravisinietal.,2011andLevchenkoetal.,2010).

At themacro level,Korineketal. (2010)findastrongstatistical relationship between insured short-termtrade credit, as a proxy for total trade finance, andtradeflows.Whenextendingthesamedatasetoverafullbusinesscycle,2005-12,thisstrongcorrelationisconfirmed(seeFigureD.15).

(ii) Key drivers of recent developments in trade finance markets

Risk aversion and market failures during recent financial crises

As indicated above, trade finance is one of the safestfinancialactivities,withlowcreditdefaultrates.However,asashort-termlendingactivity,itisheavilyreliantonre-financingontheinter-bankmarketforwholesale,short-termfunds.Inthe2009financialcrisisandintheAsianfinancial crisis of 1997-99, the overall tightening ofliquidityon inter-bankmarketsappears tohavehadanimpactontradecreditsupplythroughacontagioneffect:notonlywasliquidityinsufficienttofinanceallrequestsforlendingbuttradelendingwasadditionallyaffectedbythe general re-assessment of risk linked to theworseningofglobaleconomicactivities.

In the last quarter of 2008, notably at a time whencentral banks injected large amounts of liquidity, the

FigureD.15:Relation between imports and insured trade credits, 2005-2012 (US$million,averagedoverallcountries)

Real imports Insured trade credits

0

10,000

20,000

30,000

40,000

50,000

60,000

Rea

l im

port

s (U

S$

mill

ion)

Time

Insu

red

trad

e cr

edits

(US

$ m

illio

n)

20

05

q1

20

05

q2

20

05

q3

20

05

q4

20

06

q1

20

06

q2

20

06

q3

20

06

q4

20

07

q1

20

07

q2

20

07

q3

20

07

q4

20

08

q1

20

08

q2

20

08

q3

20

08

q4

20

09

q1

20

09

q2

20

09

q3

20

09

q4

20

10

q1

20

10

q2

20

10

q3

20

10

q4

20

11

q1

20

11

q2

20

11

q3

20

11

q4

20

12

q1

0

1,000

2,000

3,000

4,000

5,000

6,000

Source:AuboinandEngemann(2012).


253

II D. Tr

aD

e o

pe

nn

es

s a

nD

Th

e b

ro

aD

er

so

cIo

-e

co

no

mIc

co

nTe

xT

G20 discussed whether a specific, tailor-made tradefinance“package”wasrequiredtoaddressthelackoftradefinance.Theproblemwastwo-fold.Alargeshareoftheadditional liquidityprovidedbycentralbanksatthetimewasnotintermediatedintonewloans.Hence,it did not finance “new” trade transactions. Secondly,the liquidity injection by the central banks did notresolve the growing problem of risk aversion, as thecrisisspread.

The perception of risk of non-payment increaseddisproportionately relative to the actual level of risk.Thismanifesteditselfinasharpincreaseinthedemandby traders for short-term trade credit insurance orguarantees.TheG20 responded to thisbycommittingto supply greater “capacity” through export creditagencies.ThequestionarosewhethertheG20packagecarriedanelementof “moralhazard”, i.e.whethersuchinsurancemightleadtoimprudentlendingdecisions.

The 2008-09 financial crisis revealed many marketfailures, starting with the failure by credit ratingagenciesandallothermarketsurveillancemechanismsto detect early signs of deterioration of the generalsoundnessofbanks, inparticularthemultiplicationofoff-balance sheet operations and the subsequentdeteriorationof their riskprofiles.Anotherfailurewastheabsenceofaproper“learningcurve”toallowforabetter differentiation between “ill” market segmentsand“healthy”ones.

By the time of the London G20 Summit in April 2009,surveysbytheInternationalChamberofCommerce(ICC)Banking Commission and the Bankers’ Association forFinance and Trade (BAFT) on current trade financemarket trends (ICC, 2009; IMF-BAFT, 2009) hadprovided confirmation of the sharp deterioration (lowervolumes, higher prices) of markets and evidence ofshortagesinsomeregions.52ThispromptedtheG20toprovide US$ 250 billion of trade finance for two years(Auboin,2009;ChauffourandMalouche,2011).

The G20 package provided temporary trade financesupport inaway thatwouldnot result in the long-rundisplacement of private market activity. This packagecomprised a mix of instruments that allowed forgreaterco-lendingandriskco-sharingbetweenbanksand public-backed international and nationalinstitutions. The working group that was establishedby the G20 to monitor the implementation of thepackage found that after one year, some US$ 150billionofthefundinghadbeenused.

Specific problems of low-income countries

Theproblemsfacedbytradersinlow-incomecountriesin accessing affordable trade finance are to a largeextent structural and have worsened since the crisis.For example, a recent survey by the Netherlands’Centre for thePromotionof Imports fromDevelopingCountries (CBI) revealed that a majority of small and

medium-sized enterprise (SME) exporters in Africaconsiderthattradefinancecostshaveincreasedinthelast threeyears,andthataccesstotradefinancehasbecomemoredifficult(CBI,2013).

SMEsindevelopingcountriesaregenerallyfacedwithamixof“structural”constraints.Theserangefromalackofknow-howinlocalbankstoalackoftrust,reflectedin traders being required to set aside very largecollateralagainstatrade loanandtopayhighfeesforthese loans. This is despite the fact that the rate ofdefault on trade payments in low-income countries isnotmuchhigherthaninotherpartsoftheworld.

Multilateral development banks have developed anetwork of trade finance facilitation programmesaimed at supporting trade transactions at the “lowerend” of such trade finance markets – transactionsrangingfromafewthousanddollarstoamaximumofafewmillion.Theseprogrammesprovideriskmitigationcapacity (guarantees) to both issuing and confirmingbanks toallow, inparticular, for rapidendorsementoflettersofcredit–amajor instrumentused tofinancetradetransactionsbetweendevelopingcountryplayersand between developed and developing countries. Inthemidstof thefinancial crisis,during theautumnof2008, and with support from the WTO and the G20,the guarantee limits of these programmes wereincreased to support trade transactions in thesemarketsandtoreducethe“structuralconfidencegap”between the existing level of risk and its perception.The demand for these programmes continues to bestrong.

Challenging regulatory requirements

The expansion of world trade depends on the stableandpredictablefunctioningofthefinancialsystem.Asa result, the strengthening of prudential rules isbeneficial for both the financial system and for worldtrade. Ina joint lettersent toG20 leaders inSeoul inNovember2010,theheadsoftheWorldBankandtheWTO raised the issue of the potential unintendedconsequences of new global prudential rules (so-calledBaselIIandIIIframeworks)ontheavailabilityoftradefinanceinlow-incomecountries.

While trade finance received preferential regulatorytreatmentunder theBasel I framework in recognitionof its safe, mostly short-term character, theimplementation of some provisions of Basel II proveddifficult for trade. Basel III added a leverage tax onletters of credit to these requirements. The letterpointed out that given the systemic importance oftrade finance for trade and development, theapplication of excessively stringent regulatoryrequirements could reduce incentives in the financialsector to engage in trade finance (relative to otherassetcategories).

Asaresult,theseissueswerediscussedbytheBaselCommittee on Banking Supervision Policy


254

Development Group and the institutions concernedwith trade finance, notably the WTO, the World Bankand the ICC. Based on proposals made by the WTOandtheWorldBank, theBaselCommitteedecided inOctober 2011 to waive the obligation to capitalizeshort-term letters of credit for one full year as theaverage maturity is between 90 and 115 days. Thismeasure has the potential to unblock hundreds ofmillions of US dollars to finance more tradetransactions. In January 2013, the Basel Committeealso decided to alter the liquidity requirements forshort-term lending, particularly those used bydevelopingcountrytraders.

(iii) Challenges for trade finance in a context of financial deleveraging

Akeyquestioniswhetheradownsizingofthefinancialsectorasawholecouldpotentiallyleadtoareductionin the supply of trade finance as well – and hencehamperthefutureexpansionoftrade.

The 2012 Annual Report of the Bank for International Settlements (BIS) indicates that the European andUS banking sectors are currently undergoing a periodof “de-leveraging” of bank balance sheets that mightresult ina “welcomedownsizingof thebankingsectorover the long run” (BIS, 2012). This may lead to moresustainableandsoundfinancialconditionsintheglobaleconomy.Consideringthat theexpansionof theglobalfinancial industry in the2000s (measuredby itsshareinGDPortheshareoftotalcredittoincomes)hadbeenencouragedbyexcessive“leveraging”ofbanksandrisk-taking,aperiodofcreditmoderationandmorerealisticreturns on capital would yield substantial economicbenefits. These include more prudent lending policies,

decliningdebt to incomeratiosanda return toamoreusualallocationofcapitalresources,withlessdivertedfromothersectorsbecauseofartificiallyhighreturnsinthefinancialsector.

However, financial crises, when triggered by thebursting of asset bubbles (real estate or financialassets),mayleadtosignificantandlengthycorrectionsinthefinancialsector,withlong-lastingeffectsontheeconomy. Downsizing can be a long and bumpyprocess, which may also lead to adverse macro- andmicro-economic consequences. Figure D.16 showsthatafter thecredit crunch in2008-09, year-on-yeargrowth in claims on non-financial sectors remainedmainly negative from 2010 to the beginning of 2012for theEuroareaaswell as foradvancedeconomiesmoregenerally,i.e.thatbankswerelendinglesstotherealeconomy.Onlyemergingcountrybanksincreasedtheirlendingactivitiesoverthisperiod.

At themacro level, financial crisescanhavenegativespillovers in several ways. Banks may reduce thesupplyofnewcredittoeconomicagentsinanefforttocontain, or even reduce, the size of their assets inordertomeetprudentialratios.53Existing,over-valuedassetsmayhavetobewrittenofforsoldataloss,withthe effect of reducing bank profitability.54 Secondly,thecombinationofreducedprofitabilityonbankassetsand reduced new lending may be a source ofcontraction for the economy’s overall investment rate–bothforthefinancialsectorandfortheeconomyasa whole (through reduced lending). If capitalaccumulation were to be impaired for an extendedperiodoftime,potentialoutputwouldbereduced.

AccordingtoIrvingFisher’sdebt-deflationmechanism,financialcrisesusuallyleadtoacollapseincreditand

FigureD.16:Year-on-year growth of claims on non-financial sectors, 2006-2012 (percentage)

-0.3

2006q1 2007q1 2008q1 2009q1 2010q1 2011q1 2012q1

Per

cen

t

-0.2

-0.1

0

0.1

0.2

0.3

0.4

Euro area Advanced economies Emerging economies

Source:Authors’calculationsbasedonBISLocationalBankingStatistics.


255

II D. Tr

aD

e o

pe

nn

es

s a

nD

Th

e b

ro

aD

er

so

cIo

-e

co

no

mIc

co

nTe

xT

adeclineinpricelevels,hencedeflation(Fisher,1933).Both high debt ratios and deflation generally causedepressions because debt burdens become evenhigher in real terms. As Fisher (1933) put it, “eachdollarofdebtstillunpaidbecomesabiggerdollar,andif the over-indebtedness with which we started wasgreatenough,theliquidationofdebtscannotkeepupwiththefallofpriceswhichitcauses.”

During the recent financial crisis, high debt and highleverage ratios have been the main topics ofdiscussion, with deflation being discussed less.Figure D.17 shows that annual growth in consumerprices decreased during this period, but it was onlynegative in 2009 for the United States and China,while for Europe it remained positive. In 2010 and2011, consumer prices for the United States, Chinaand Europe rose again. Central banks provided thenecessary liquidity to allow banks to deleverage.However, theproblemof longdeleveragingperiods isnot necessarily deflation but a misallocation ofresources.Newloansaredisplacedbyoldloans,whichmay induce a long period of credit crunch leading tostagnation.

At the micro level, a long period of financialretrenchment may also lead to substantial negativeeffects,inparticularfortradefinanceandhencetrade.Explicitly, the allocation of capital resources may notimproveifthereislesscredit.

Long periods of credit crunch can affect certaincategoriesofeconomicagentsorcredit,suchastradecredit,disproportionately–despitetheirgoodcreditorsafetyrecord.AmitiandWeinstein(2011)arguethatthedecade-long downward adjustment of the Japanesefinancial industry has not had a neutral effect on thefinancing of Japanese exporters. Firms working withtroubledbankssawtheirexportperformancedeclineinabsolute terms. SMEs, in particular exporting ones,

were the most affected because they were the mostdependentontradecredit.

ThequestionarisesastowhethertheaccessofSMEsto credit in general, and to trade credit in particular,will be negatively affected in a context of increasedcompetitionwithinthecreditcommitteesofbankswhoarbitrate on the different categories of loans. Onepotential pitfall of a process of greater “selectivity ofrisk” is the possible allocation by banks of scarcecapital resources to the most profitable creditsegments,therebyreducingtheirinvolvementinlower-profitabilityproductssuchasshort-termtradefinance.Another pitfall is that banks may focus on their mostprofitable customers – the larger ones. Hence, adownsizing of the financial sector and greaterselectivity in risk-taskingmaynotactautomatically infavour of an improved allocation of resources in thefinancialindustry.

Tradefinancemaybeusedasaprime instrument forreducing the size of a bank’s balance sheet, henceachieving rapid deleveraging. Because of its short-term, roll-over nature, most trade credit lines expireafter90days,theaveragedurationoftransactions.Bynotrenewing(rolling-over)orbyreducingthesecreditlines, banking intermediaries can achieve a quickreductionoftheirlending(deleveraging)whenneeded.Attheendof2011,afewEuropeanbanksannounceda reduction of trade credit lines in an effort torestructuretheirbalancesheets.Thisapproachprovedtobeshort-lived.

Tradefinancemayalsobenegativelyaffectedifthere-scaling of the financial sector is accompanied by “re-nationalisation” of lending activities at the expense ofcross-border lending. Many international banks havealreadyscaledbackinternationalactivities.Asindicatedby theBIS, “inaddition towritedownsofcross-borderassets during the crisis, the more expensive debt andequity funding also led to reductions in the flow ofcross-border credit. As a result, credit to foreignborrowershasfallenasashareofinternationallyactivebanks’ total assets (see Figure D.18). For Europeanbanks,thesharehasdeclinedbyalmost30percentagepoints since early 2008. Not all banks have reducedforeignactivities,withnotableexceptionsbeingAsian-based banks and banks in other emerging countries.However a re-composition of the banking landscape,withshiftsinmarketshare,maybeatplay.

(iv) Looking ahead

The future direction of the international bankingindustry is difficult to predict, although somereduction of its share in GDP, at least in advancedeconomies, may be expected. Much depends on theincentives provided by a new, reformed financialsystem.Normally,banklendingshouldbere-orientedtowardsmoresustainableformsoffinance.Ifbalancesheet shrinkage works at the expense of “leverage

FigureD.17:Annual inflation – year-on-year change in consumer prices of all items, 2005-2011 (percentage)

2005 2006 2007 2008 2009 2010 2011-2

-1

0

1

2

3

4

5

6

7

United States OECD - Europe China

Source:OECD.


256

finance and off-balance sheet toxic investment”,traditional forms of finance might benefit. In thatcase, lending would be re-oriented towards realeconomy financing, including trade finance, which isan important factor of trade, not only in periods ofcrisis (Auboin and Engemann, 2012). At the sametime,ifrationalizationofthesectorworksinfavourofhigher-yieldingformsoflending,asopposedtocross-border lending, the question as to whether to stayengaged in trade finance will be posed by manyfinancialintermediaries.

Thequestionofenteringorexitingtradefinanceisnotan easy one to answer. Trade finance bears “fixedcosts” of doing business, particularly costs oforigination of trade finance transactions (investing inback offices, customers and sales relations, openingforeign branches, being acquainted with internationaltrade finance procedures). Of course, the decision tostayengaged intradefinancedepends largelyonthedemand for real trade transactions – and hence thecontinuationofproductionsharingandtraderelations.Multilateral agencies will need to remain engaged intradefinance,atleasttohelpfillthegapatthe“lowerend” of the market, notably in low-income countries.Dialogue with regulatory agencies will need to bepursuedtoensurethattradefinanceisrecognizedasadevelopment-friendlyandlow-riskformoffinance.

(b) Currencymovements

Exchange rates can influence international trade inmanyways.Realexchangerates(therelativepricesoftradable to non-tradable products) can affect theincentive to allocate resources between sectors

producing such tradable and non-tradable goods. Thetradeimpactofexchangeratescanbeanalysedthroughtwo effects: the fluctuations of exchange rates, whichcanbea sourceof frustration for individual producersandtraders,astheymayimposehighuncertaintycosts;and prolonged deviations of currencies from theirequilibrium levels – so-called misalignments – whichare regarded as important distortions in internationalpricecompetition.Intheshortrun,theymaynegativelyimpacttheallocationofresourcesbetweencountries.Inthelongerrun,thesituationislessclear.

Economictheorysuggeststhatwhenmarketsarefreeof distortions, anexchange ratemisalignmenthasnolong-run effect on trade flows as it does not changerelativeprices.However,long-runeffectsarepredictedin models that assume market distortions, such asinformationproblemsorproductmarketfailures.Intheshort run,whensomeprices in theeconomyare lesslikelytoadjust,movementsinnominalexchangeratescanalter relativepricesandaffect international tradeflows, although this depends on several factors,including thepricingstrategyof tradingfirmsandtheimportanceofglobalproductionnetworks.Thus,theseshort-runtradeeffectsarenotstraightforward.

Even if longer-run economic effects of currencymisalignments cannot be fully established for allcountries and in all circumstances, persistentmisalignmentsinexchangeratesareasystemicirritantininternationaltradebecausetheyfuelperceptionsofunfair monetary competition. This, in turn, createspressureontheuseoftradepolicymeasures,suchastariffs and trade “defence” instruments, to redressperceived monetary imbalances. For this reason inparticular, the world trading system needs aninternational monetary system promoting exchangerate stability and adjustments. Progress in monetarycooperation is uneven, however. The exchange rateissue can, therefore, be expected to remain with theworldtradingsystemforsometime.

FiguresD.19andD.20showtrendsinthevolatilityandlevels,respectively,ofrealeffectiveexchangeratesforselectedregionsandcountriesovertime.

(i) Exchange rate volatility and trade

Afteraperiodof30yearsof relativestabilityofbothnominal and real exchange rates under the BrettonWoods system, increased volatility of exchange ratesfrom theearly1970s triggereda livelydebateon thechannelsthroughwhichsuchincreasedvolatilitycouldaffecttherealeconomy(FigureD.19).

Particularly strong concerns were expressed by thetrading community, which had negotiated substantialreductions in border protection when the GoldExchangeStandarddeterminedexchangerates.Attherequest of the then Director-General of the GeneralAgreementonTariffsandTrade (GATT) (onbehalfof

FigureD.18:Ratio of foreign to total assets, 2006-2011

2006q2 2007q2 2008q2 2009q2 2010q2 2011q2

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

All banks European banks Non-European banks

Source:IMFInternational Financial Statistics ,BISinternationalbankingstatistics,WTOcalculations.

Note:TotalforeignclaimsofBISreportingbanksheadquarteredinCanada,France,Germany,Italy,Japan,theNetherlandsandtheUnitedStatesasapercentageoftotalassets,wheredomesticclaimsareobtainedfromIMFInternational Financial Statistics .


257

II D. Tr

aD

e o

pe

nn

es

s a

nD

Th

e b

ro

aD

er

so

cIo

-e

co

no

mIc

co

nTe

xT

theGeneralCouncil),theInternationalMonetaryFund(IMF) examined the effects of greater exchange ratevolatilityonglobaltrade.

While concluding that the evidence concerning anegativeeffectoftheincreasedvolatilityofexchangerates on global trade was slim, the IMF (1984)highlightedtheroleofexchangeraterisk.Thiswasinline with earlier (1970s and 1980s) theoreticalanalyses of the relationship between exchange ratesand international trade. These studies focusedprimarilyonthecommercialriskinvolvedinconducting

international transactions and the uncertaintygenerated by short-term or long-term exchange ratevolatility.Thequestionofhowthisuncertaintyaffectedthe decision to trade, its expected profitability andeventually the allocation of resources betweentradableandnon-tradablegoodsandserviceswasthemainfocusofthesestudies.

A simple explanation of how exchange rate volatilityaffectstradeisprovidedbyClark(1973).Heexplainshow the uncertainty about future exchange ratestranslates directly into uncertainty about future

FigureD.19:Volatility of real effective exchange rates, 1975-2012

Developing Asia Sub-Saharan Africa Western Hemisphere Middle East and North Africa United States

0

0.02

0.04

0.06

0.08

0.1

0.12

RE

ER

yea

rly v

olat

ility

19

75

19

77

19

79

19

81

19

83

19

85

19

87

19

89

19

91

19

93

19

95

19

97

19

99

20

01

20

03

20

05

20

07

20

09

20

11

20

12

FigureD.20:Levels of real effective exchange rates, 1995-2012 (index,1995=100)

China Japan Korea, Rep. of United States Euro zoneBrazil

0

20

40

60

80

100

120

140

160

19

95

19

96

19

97

19

98

19

99

20

00

20

01

20

02

20

03

20

04

20

05

20

06

20

07

20

08

20

09

20

11

20

12

20

10

Source:BruegelInstitute–RealEffectiveExchangeRateDatabase.

Source:IMFInternational Financial Statistics .

Note:Volatilityiscomputedastheyearlystandarddeviationofmonthlygrowthinrealexchangerates.


258

receipts in domestic currency and how producersadjust production and exports downwards to reduceexposure to exchange rate risk. The view that anincrease inexchange ratevolatilitywillhaveadverseeffects on the volume of international trade isrelativelywidespreadinstudiesconductedthroughoutthe 1970s and 1980s (see also Baron, 1976;Cushman, 1983; De Grauwe and Verfaille, 1988;Giovannini, 1988; Bini-Smaghi, 1991). However,theseconclusionsrestonrelativelyfirmassumptions,which have been scrutinized and adapted by otherauthors – notably the assumption of perfectcompetition, the large role of the invoicing currency,theabsenceof imported inputs, thehighaversion torisk and the absence of exchange rate hedgingfinancial instruments. This has led to moresophisticated multi-country models with diversifiedfirms, in which the relationship between exchangerates, the supply of goodsand thedecision to tradebecomemoreambiguous(seeBoxD.6).

Reflecting the relatively inconclusive state of earlytheoretical models regarding the effects of exchangerate variability on trade, the vast empirical workproduces equally ambiguous results. As argued byTaglioni (2002), “it is customarily presumed that theadverseeffectofexchangeratevolatilityontradeflows,if it exists, is certainly not large”. This conclusion isgenerallysharedbyOzturk(2006),whorevealsawiderangeofempiricalevidence,somesupportingandsomecontradicting the hypothesis of a negative relationshipbetweenexchangeratevolatilityandtrade.

As aptly summarized by Coric and Pugh (2010), “onaverage, exchange rate variability exerts a negativeeffect on international trade. Yet, […] this result ishighly conditional. […] Average trade effects are notsufficiently robust to generalize across countries.”Results are conditional for the same reasons asidentified in theoretical models: the existence ofhedginginstruments,thepresenceofimportedinputs,the possibility of invoicing in local currency and thecapacity to absorb losses due to exchange ratechangesandotherfactorsinprofitmargins.Thefirmsmost sensitive to exchange rate volatility may not bethe largeonesbut rather thesmallerfirms (asshowninSectionB.2(f)).Inaddition,empiricalstudiestendtofindasignificanteffectmainlyinthecaseoftradewithclose neighbours, particularly in the case of veryintegratedeconomies.

(ii) Exchange rate misalignments

Thedebateonthetradeimpactofexchangerateshasresurfaced in the past decade, with the build-up oflarge external global imbalances. After the 2009global recession, concerns about unemployment andslowrecoveryhaveincreasedthesuspicionthatsomecountriesareseeking to “export” theirwayoutof thecrisis at the expense of trading partners. Hence, thepolicydebatehasshiftedfromtheimpactofexchange

ratevolatilityontradetowardstheeffectsofsustainedcurrencymisalignment.Thismeansthattheemphasisis less on the effects of variability and more on thelevelofrealexchangerates.

Exchangeratescandepartfromtheirequilibriumlevelfortworeasons.First, thismaybeduetogovernmentintervention directly aimed at altering the realexchangerate(currencymanipulation).Inthisrespect,governmentsand/orcentralbankspossessanumberofpolicy instruments thatcanaffect therealvalueoftheexchangerate,includingtheintroductionofcapitalcontrols or targeted intervention in foreign exchangemarkets. Secondly, misalignments can be theunintended side effect of macroeconomic policiesaimedatachievingdomesticobjectivesortheresultofdistortionsintheinternationalfinancialarchitectureorindomesticstructuralconditions.

The academic and policy debate on currencymisalignments concerns two important points. Thefirst istheextenttowhichtherealexchangerateisavariable that policy-makers can influence(Eichengreen, 2007; Rodrik, 2008). The consensusview is that the realexchange rate,being the relativepriceof traded tonon-tradedgoods, isnotunder thedirectcontrolofpolicy-makers.However, its levelcanbe influenced by policy in the short to medium term.Eichengreen (2007) provides an illustration byoutlining the experience of the Republic of Korea inthe1960s,whereanominaldevaluationwascombinedwithfiscalconsolidation.The latterhelpedmaintainalowerleveloftherealexchangerate.

The second point relates to the measurement of theequilibrium exchange rate. Ascertaining theequilibrium exchange rates – hence, the cause of acurrency misalignment – is difficult in theory and inpractice.Itisdifficultintheorybecausetheexchangerate is a variable determined by a variety ofmacroeconomic, financial and trade factors. It isdifficult in practice because there are a number ofdifferent methodologies measuring equilibriumexchangerates,withnonebeing“better”thanothers.Themainmethodologiesused for theassessmentofexchange ratesarebasedon thecompetitivenessofthe tradable goods sector (so-called purchasingpower parity), general equilibrium model calculationsand estimates required to achieve the equilibrium ofthe balance of payments (so-called fundamentalequilibrium exchange rates). They may lead to arelatively wide range of estimates. The IMF uses arangeofsuchestimatestomakeitsownassessmentofequilibriumexchangerates.55

Another question concerns trade effects in the longrun versus the short run. According to standardeconomictheory, longrunpricesarefullyflexibleandadjust to any policy change (or any other shock). Inparticular, when markets have no distortions, anexchangeratemisalignment–suchasadevaluationof


259

II D. Tr

aD

e o

pe

nn

es

s a

nD

Th

e b

ro

aD

er

so

cIo

-e

co

no

mIc

co

nTe

xT

thecurrency–hasnolong-runeffectontradeflowsoron real economic activity, as it would not changerelativeprices.Theshortrunandmediumrun,however,canbedifferent.Thereason is that, ifsomeprices inthe economy take time to adjust (i.e. are “sticky”),movements in nominal exchange rates can alter

relative prices and affect both the allocation ofresourcesbetweennon-tradableandtradablesectorsandinternationaltradeflows.

Open macro-economy models embed the short-termeffectsofexchangeratemisalignments(e.g.Krugman

BoxD.6: Overview of the literature on exchange rate volatility and trade

There are five main strands of literature extending the model of Clark (1973), which only focused on theuncertaintycreatedbyexchangeratevolatility.

First,theeffectofincreasedvolatilityofexchangeratesontradedependsheavilyonthelevelofriskaversionoftraders(DeGrauwe,1988;DellasandZilberfarb,1993).Risk-neutraltradersareunlikelytobeaffectedbyexchangerateuncertaintybutrisk-aversetraderswillbeaffected,albeittodifferentdegrees.AsindicatedbyDe Grauwe (1988), for very risk-averse traders, paradoxically, exporting more could be a response toincreasedvolatilityinordertocompensatefortheexpectedfallinrevenueperexportedunit.TheexistenceofsucharelationshipwaslaterconfirmedtheoreticallybyBrollandEckwert(1999).

A second set of studies account for the possibility of firms hedging against exchange rate risks. Theavailabilityoffinancialhedgingthroughforwardexchangemarketshelpsreducetheuncertaintygeneratedbyfluctuationsofnominalexchangeratesalthoughfirmshaveunequalaccesstohedgingfacilitiesandmaydisplaydifferentbehaviouraccordingtowhichsideofthehedgingpositiontheystand.ViaeneanddeVries(1992)suggestthatforwardmarketscreate“losers”and“winners”amongexportersandimportersthatareon opposite sides of the forward transactions. Besides, as noted by the IMF (1984), foreign exchangehedgingcontractsarenotnecessarilyavailableinallcountriesandtoallcategoriesoffirms.Contractsaretypicallyrelativelylarge,maturitiesshortandfeeshigh.Inaddition,theyonlycoveralimitedshareofpossiblefluctuationsduring theproposedmaturitiesas it ishard,bydefinition, toanticipate themagnitudeofsuchfluctuations.Hence,itisgenerallyacceptedthatlargerexportingfirmsareinabetterpositionthansmallerfirmstobenefitfromexchangeratehedging.

Athirdextensionoftheliteraturefocusesonadjustmentcosts.Theassumptionthatexchangeratesaffecttradebecausefirmscannotadjustinputsaccordingtoexchangeratefluctuationshasalsobeenrelaxedbyseveral authors. De Grauwe (1992) has worked with a wider spectrum of cases than those described byClark (1973). If firmscanadjust factorsofproductionupwardsanddownwardsaccording toworldprices,they are likely to sell more when international prices in foreign currency are high (with a limit set by theproductioncapacityofthe“flexible”factor)andlesswhensuchpricesarelow.However,thiswilldependonriskaversiontowardsprofituncertainty.Themoreriskaversefirmsare,thelesslikelytheyaretoexportmorein lightofhigherprofitvariancefromexchangeratevolatility.Ontheotherhand, lessrisk-aversefirmswillsellmoreevenwithprofituncertaintybecausetheopportunityfrompricevariabilitycanoffsettheuncertainty.

Afourthsetofstudiesanalysestheeffectofexchangeratevolatilityonthecompositionoftraderatherthanits gross volume. Some models focus on the extensive rather than the intensive margin of trade (i.e. thenumber of products traded rather than the volume of trade of a given number of products). Specifically,modelsofpersistenceorpathdependence inglobal tradeshowthat thehighvariabilityofexchangeratesandassociateduncertaintycan influence thedecision toenterorexit foreign trademarkets (inparticular,Dixit,1989;Krugman,1986;Franke,1991).

Finally,agroupofstudiesremovestheassumptionthatexchangerateuncertaintyisexogenous.BacchettaandWincoop(2000)examinetheimpactofvolatilityonthelevelsoftradeandwelfareinacontextofbothfixedandflexiblearrangements.One interestingoutcome illustratingthecomplexityof theexchangerate-traderelationshipisthatmonetarystimulusinonecountrymayleadtothedepreciationoftheexchangerateofthatcountrywithoutmucheffectontrade.Thisisbecausethedepreciationoftheexchangeratemayontheonehandreduceimportswhileontheotherhandtheincreaseindomesticdemandwouldboostimportsinanoffsettingmovement.Ofcourse,theneteffectwilldependonawholesetofvariables,fromdemandelasticitiesfor importstosupply-sidefactors,suchasthedesireorabilityofdomesticproducerstoadjustpricestothedepreciationofthecurrency.

Since2000,empiricalworkonthetradeimpactofexchangeratevolatilityhascontinued,notablywithcross-countryanalysis.Forexample,anOECDstudy(Huchet-BourdonandKorinek,2012)examinestheimpactofexchangeratevolatilityontradeintwosmallopeneconomies,ChileandNewZealand.Thestudyconcludesthatsmaller,openeconomiestendtobemoreaffectedthanlargeronesbyexchangeratechanges.


260

and Obstfeld, 2009). In particular, when prices are“sticky”, anominaldepreciationof thehomecurrencyresults in a real depreciation of the exchange rate,thereby increasing the price of the foreign goodrelative to that of the domestic one. This change inrelative prices prompts the home economy to importless, as home consumers switch to less expensivedomestic goods, and export more, as foreignconsumersturntolessexpensivehomegoods.Underthese standard macroeconomic models, all otherthings being equal, the trade balance of the homecountries would improve, with increasing exports andfalling imports, as a function of this short-termdepreciationoftheexchangerate.

Thereare two importantassumptionswhichmustholdinthiscase:thenominaldepreciationoughttoresultina real depreciation, thus raising the price of foreigngoodsrelativetohomegoods;therelativepricechangemusthaverapideffectsonthequantitiesimportedandexported,andhenceon the tradebalance. Inpractice,however, the short-term effects of exchange ratemisalignments may be more complex, as these twoassumptions may not always apply (depending, forexample,on thedemandorsubstitutionelasticities foreachgood).Inaddition,thetradebalancedependsonawide range of others factors, such as income in thehomecountryandtherestoftheworld.

On the first assumption, recent literature shows thatthese effects depend, among other things, on thecurrency in which domestic producers invoice theirproducts. This is, in part, because the currency ofdenomination affects the extent to which a nominaldepreciation results (or not) in a real depreciation oftheexchangerate.Forexample, ifproducerssettheirprice in thehomecurrency (asgenerally assumedbystandardtrademodels),thereisagood“pass-through”fromchangesinthenominalexchangeratetotherealexchange rate, and an unanticipated devaluationlowersthepriceofdomesticgoodsrelativetoforeigngoods,asnotedabove.

However, the trade effect of a devaluation would bedifferentifdomesticproducersweretosettheirpriceinthebuyers’currencyorinavehiclecurrency,suchasthe US dollar or the euro. This is because the pass-through effect would be less than “perfect” in thesecases.Forexample, the theory suggests thatwhile adevaluation would still have real effects, suchconsequences would not be equivalent to exportpromotionbutrathertoimportrestrictions(StaigerandSykes,2010).

The second assumption, i.e. the short-term impact ofcurrencymisalignments,canbequestioned.Underthe“J-curveeffect”,thedepreciationoftherealexchangerate is often synonymous with an immediatedeterioration of the trade balance and a subsequent(rapid) improvement.Partofthismechanismassumesthatthedevaluationisunexpected(hencethechange

inprices isunanticipated)and thatacertainshareoftradeispre-ordered(someshareofimportsandexportorders are placed several months in advance). Thevalue of the pre-contracted level of imports rises intermsofdomesticproducts,whichimpliesthatthereisan initial fall in the trade balance. The increase inimport prices may be partly or fully offset by thesubstitution, if available, of imported goods by localgoodsbutthisimpliesanadjustmentinthecapacityofdomestic firms which requires time. When thesechanges have taken place, a real exchange ratedepreciationwill improvethetradebalancerelativetoits pre-depreciation level. In brief, understanding theshort-termimpactofanexchangeratedevaluationontrade flows and the trade balance in the short andmedium run is conceptually more complex than itinitiallyappears.

The above argument does not take into account thepossibility of market failures. For instance, in thepresence of information problems (e.g. the quality ofexportgoods isunknowntoforeignconsumers), ithasbeenarguedthatthelevelofexportsmaybeinefficientlylow(BagwellandStaiger,1989;Bagwell,1992).Ahigh-quality exporter may need to signal quality, which iscostly.Firmsmayalsohavecommonuncertaintyabouttheprofitabilityofexporting(FreundandPierola,2010).Inthiscontext,theundervaluationoftheexchangeratemayhavelong-runeffectsifitallowsexporterstoenterforeignmarkets,thusovercomingtheinitialinefficiency.If this logic is correct, one would expect currencydepreciation to be associated with entry into newmarketsandnewproductlines(i.e.theextensivemarginoftrade),andforitnot(ornotcompletely)tobeundoneinthelongrunwhenpricesadjust.Moreover,asmarketfailuresareconsideredtoplayamoreprominentroleindeveloping as opposed to developed economies, oneshould expect that these long-run effects are weakerforthelatter.

(iii) Looking ahead

Whetherexchangeratevolatilityandmisalignmentcanhavearealeffectontradeintheshortandlongrunisan empirical question. And the empirics yield mixedfindings.Asindicatedabove,acurrencyundervaluationis sometimes found to have a positive impact onexports but the presence, size and persistence ofthese effects are not consistent across differentstudies.AsdescribedinSectionB.2(e),thecomplexityofthisrelationshipisnotlikelytobereducedasglobalproduction networks become more prominent ininternational trade and as business cycles betweencountriesbecomeincreasinglyinterdependent.

Fortheworldtradingsystem,exchangeratesarelikelytoremainasystemicissue.GATT/WTOmembershaveconsistently argued that an international monetarysystempromotingthestabilityofexchangeratesiskeyto establishing an enabling environment for worldtrade(seeBoxD.7).


261

II D. Tr

aD

e o

pe

nn

es

s a

nD

Th

e b

ro

aD

er

so

cIo

-e

co

no

mIc

co

nTe

xT

ThedebateonthetradeimpactofexchangerateshassurgedagainrecentlyintheWTO,56andislikelytodosoeach time that it is felt, rightlyorwrongly, that thepresent state of international monetary cooperationdoes not allow for orderly exchange rate adjustmentreflecting balance of payments positions and doesallow a particular member, or several members, toenjoy competitive advantages as a result of such alack of cooperation. While the influence ofmacroeconomicandstructuralpolicies indeterminingexchangeratesisacknowledged(Eichengreen,2007),the world trading system must regularly “deflect”tensionsassociatedwiththeperceivedtradeimpactofexchange rates. This has become more frequent inrecent years, as growing international inflows andoutflows of foreign exchange have the potential todestabilizedomesticeconomicpoliciesandreducetheefficacyof traditionalcontrols (notably restrictionsoncapitalmovements).

The question for the WTO is also systemic becauseexchange rate shifts increase or weaken the desiredorperceivedlevelofprotectionofdomesticoperators–andthusseemtohavearoleinthedefinitionoftradepolicy.Atthemultilaterallevel,theerraticmovementofexchange rates is frustrating the desired levels ofprotection that are negotiated by WTO membersthrough long-term commitments – precisely becausepoliciesareaimedatsettingpredictableconditionsofaccess for producers and traders. In turn, membersmay seek a way to address cyclical development orexchangeratechangesinthetradepolicytoolkit.

Someempiricalstudiessuggestthatcontingenttrademeasures are used in response to trading partners’currency depreciations. For example, Knetter andPrusa(2003)andNielsandFrancois(2006)establisha linkbetweenanti-dumpingcasesandtheexchangerates for a number of countries: the number of anti-dumpingcomplaintsagainstpartnerstendstoincreasewhen the local currency appreciates relative to thepartner’s currency and when the current accountdeficitwidens.

Inaworldoflargecapitalmarkets,aproblemarisesfortradersandpolicy-makersalikewhenexchange rates

behave in a disorderly way and do not adjust toeconomic fundamentals. During the Bretton Woodsera,partoftheinternationaltradingcommunityfoundasystemoforderlyadjustmentofrealexchangerates.The system was not ideal. However, the internationaltrading community felt that there was a system,providing, at least in the early stages, for a sense oforganized governance in the international monetarysystem.

The need for greater coherence for trade andexchangeratepolicieswas included in theGATTrulebook at the outset (see Section E.3(c)). The IMF andGATT were created in response to a lack ofcoordination of economic policies during the GreatEconomicDepression– thesenew institutionsaimedatdealingwith tradeandexchange ratepoliciesasamatter of common interest, with the introduction ofdisciplines to avoid competitive devaluations, tomaintainexchangeratestability, toreducebalanceofpayments crises and to fight protectionism. From theoutset,theinternationalmonetaryandtradingsystemswere linked by a coherent set of rules aimed at theprogressive opening of trade and payments. GATTprovisions on coherence reflected two things: theattachment of the trade community to exchange ratestability; and the need for that community to ensurethat the trading system was not frustrated by theundisciplineduseofexchange restrictionsormultipleexchange rates. The institutional set-up remains verymuchoneofcoherence–andnotofconflict–betweenthetwosystems.

In the 1994 Ministerial Declaration on the WTO’sContribution to Coherence in Global Policy Making,WTO ministers “recognized, however, that difficultiestheoriginsofwhich lieoutside the tradefieldcannotbe redressed through measures in the trade fieldalone”.Thismeansthatthetradingsystemcannotbearexcessive expectations with respect to ensuring orcontributing to stable and cooperative monetary andmacroeconomic conditions. A number of institutionsand policy processes are in place to enforce bettersurveillance of exchange rates and to reduce globalimbalances (for example, the G20 and the IMF’sMutual Assessment Process). The question as to

BoxD.7: Coherence in global policy-making

At the end of the Bretton Woods arrangement for fixed but adjustable exchange rates, GATT ministersindicated in the Declaration opening the Tokyo Round in 1973 that “the policy of liberalizing world tradecannot be carried out successfully in the absence of parallel efforts to set up a monetary system whichshieldstheworldeconomyfromtheshocksandimbalanceswhichhavepreviouslyoccurred.TheMinisterswillnot losesightof the fact that theeffortswhicharemade in the tradefield implycontinuingefforts tomaintainorderlyconditionsandtoestablishadurableandequitablemonetarysystem”.

Thesewordsaremirrored in the1994MinisterialDeclarationon theWTO’sContribution toCoherence inGlobalPolicyMaking: “successful cooperation ineachareaofeconomicpolicy contributes toprogress inother areas. Greater exchange rate stability, based on more orderly underlying economic and financialconditions,shouldcontributetowardstheexpansionoftrade,sustainablegrowthanddevelopment,andthecorrectionofexternalimbalances.”


262

TableD.4:Currency distribution of foreign exchange transactions, 2001-2010 (percentagesharesofaveragedailyturnoverinAprilofthe20mostusedcurrencies)Currency 2001 2004 2007 2010

USdollar 89.9 88.0 85.6 84.9

Euro 37.9 37.4 37.0 39.1

Japaneseyen 23.5 20.8 17.2 19.0

Poundsterling 13.0 16.5 14.9 12.9

Australiandollar 4.3 6.0 6.6 7.6

Swissfranc 6.0 6.0 6.8 6.4

Canadiandollar 4.5 4.2 4.3 5.3

HongKongdollar 2.2 1.8 2.7 2.4

Swedishkrona 2.5 2.2 2.7 2.2

NewZealanddollar 0.6 1.1 1.9 1.6

Koreanwon 0.8 1.1 1.2 1.5

Singaporedollar 1.1 0.9 1.2 1.4

Norwegiankrone 1.5 1.4 2.1 1.3

Mexicanpeso 0.8 1.1 1.3 1.3

Indianrupee 0.2 0.3 0.7 0.9

Russianrouble 0.3 0.6 0.7 0.9

Chineserenminbi 0.0 0.1 0.5 0.9

Polishzloty 0.5 0.4 0.8 0.8

Turkishlira 0.0 0.1 0.2 0.7

SouthAfricanrand 0.9 0.7 0.9 0.7

Source:BISTriennialCentralBankSurvey2010.

Note:Thesumofpercentagesharesofindividualcurrenciesequals200percentastwocurrenciesareinvolvedineachtransaction.

whether conditions will be met to set up a morecooperativeorbindingsystemofexchangeratesattheinternational level remains open. Some authors(Mundell, 1961; Williamson, 2009) have suggestedthat international cooperation is enhanced in optimalcurrencyareasand/orwhenregionalcurrenciesreachcomparable weights and status. At present,internationaltradeandfinancialmarketsaredominatedby the US dollar and the euro much more than othercurrencies(Auboin,2012).

ThecurrentprudentexpansionoftheChineserenminbi(RMB) as a trade currency raises the probability thatthe Chinese currency will play an important role intomorrow’s international monetary system. China hasmadesteadyprogresstopromotetheinternationaluseof the RMB lately, particularly in international tradetransactions.Tosomeextent, thewillingnessof tradepartnerstoadopttheRMBmayreflectapreferencetoreduce reliance on the US dollar and the euro withintheglobalmonetarysystemaswellasanopportunitytocuttransactioncostsforbilateraltrade.However,aspointed out in Section B.2(a), there is still a largediscrepancy between China’s importance in worldtradeflows(some11percentof imports)and thatoftheRMBintradepayments(some1percent).

Tobesuccessfulattheinternational level, themarketfor a currency needs to be large, liquid and global –even if used for trade purposes only (internationaltradersaskroutinelyforcurrencyhedging,requiringadevelopment of the derivatives markets). In 2010,according to the BIS, the US dollar accounted for

85percentofforeignexchangetransactionsfollowedbytheeurowhichaccountedfor40percent.TheRMBaccounted for only 0.9 per cent of global foreignexchange transactions in 2010 (see Table D.4; BIS,2010). There is little doubt, however, that in themedium to long term the RMB will expand as one ofthe world’s key currencies. The gradual lifting ofrestrictionsontheuseoftheRMBinothertransactions– in particular, inward and outward investment andinternational fundraising–willalsopushforward theRMB’sinternationalization.

4. Conclusions

The futureofglobal tradewill, toasignificantextent,dependonthesocio-economiccontextatthenational,regional andglobal level. This sectionhasprovidedadetailed discussion of the possible environmental,social and macroeconomic challenges that lie ahead.The nature of the challenges differs substantiallyacross the three domains and their relationship withthemultilateral tradingsystemhasevolveddifferentlyover time.Yet,eachof theseconcernshasbeenhighon policy-makers’ agendas in many countries. It ispossible that public perceptions of the relationshipbetween trade and these concerns can turn into apressure point for protectionism. Decisions in thesethree areas will therefore undoubtedly affect tradeflows in the future, either directly or through theireffectontradepolicydecisions.

In this rapidly changing global environment, it will becrucial to ensure that policies in socio-economic


263

II D. Tr

aD

e o

pe

nn

es

s a

nD

Th

e b

ro

aD

er

so

cIo

-e

co

no

mIc

co

nTe

xT

domains, such as labour markets, environment andfinance,arewellalignedwith tradepolicies.Opennessis likely to generate greater benefits in economiescharacterized by a strong enabling environment forenterprises and well-designed education and trainingpolicies. Individualsfind iteasiertocopewithchangesin thecompetitiveenvironment ineconomiesequippedwith social protection systems. Well-designedenvironmental policies can be both economically andenvironmentally beneficial. Open economies flourishwhen appropriate regulation guarantees stability infinancialmarketsandwhenaccesstofinance,includingtradefinance,isfacilitated,particularlyforSMEs.

In many areas, the desirable alignment across policiescan be achieved by cooperation between relevantministriesatthenational level.However, inotherareas,greatercooperationatthegloballevelmaybenecessary.In the area of environment, for instance, collectiveefforts that result inagreedpolicyapproaches towardsglobal environmental problems may limit the scope forenvironmentalpoliciestounnecessarilydistorttrade.

The interconnectedness between trade, the labourmarketandmacroeconomicpolicieswasalreadyonthemindofnegotiatorswhen theoriginalGATT legal textsweredesigned.57Indeed,theneedforgreatercoherencebetween trade and exchange rate policies has beenexplicitly reflected in the GATT legal texts. Currentconcerns about jobs and inequality, macroeconomicstabilityandenvironmentalsustainabilitymaygiveanewrelevancetothisinterconnectedness.

The recent period of economic turmoil has alsocreatedasituationthatmayleadtoincreasedcallsforprotectionism. This highlights the importance of theWTO’smonitoringandsurveillance roleaswell as itsplace in the institutional framework of globalgovernance,aswillbefurtherdiscussioninSectionE.

1 SeeJansenandVonUexkull(2010)foradiscussionoftheemploymenteffectsoftradeinlowandmiddle-incomecountriesduringtheGreatRecession.

2 WhenthisReportwaswritten,theWorldTopIncomesDatabasecontainedinformationontheincomeshareofthetop1percentofincomeearnersfor27countries,ofwhichmostwereOECDcountries.China,IndiaandSouthAfricaweretheonlyBRICScountriesrepresentedinthedatabase.InformationonChinaandIndiaisreflectedinFigureD.2.InSouthAfrica,theshareofthetop1percentincomeearnersincreasedfrom9.9percentin1990to16.6percentin2010.

3 Authors’calculationsbasedonWorldBank,WorldDevelopmentIndicators.

4 See,forinstance,MilbergandWinkler(2011).

5 Autoretal.(2012),forinstance,findevidenceregardinganincreaseintheuseofinvalidityinsuranceinreactiontoincreasedimportsfromChina.

6 SeeGoldbergandPavcnik(2007)andPavcnik(2011)foroverviewsoftherelevantliterature.

7 Thepresenceofanexportwagepremiummaydependonworkers’skilllevels.Kleinetal.(2010),forinstance,findthatlow-skilledworkersinGermanmanufacturinghaveanexportwagediscount,whilehigher-skilledworkershaveanexportwagepremium.

8 Differentcomponentsofglobalizationmayalsointeractthroughpoliticaleconomymechanisms.Peters(2012),forinstance,arguesthatincreasedpossibilitiestoinvestabroadandoffshoreproductionhasreducedUSfirms’lobbyingeffortstofacilitateimmigration.

9 SeetheopeningquoteinBacchettaandJansen(2003).

10 SeeWTO(2009)andspecificallyadjustmenttotradereforminBacchettaandJansen(2003).

11 Underthisso-called“fairwage”hypothesis,workersareassumedtoexpecthigherwagesfromfirmsthatareeconomicallysuccessful.

12 Thesefrictionsoccurwhenlookingforjobsandorlookingforpotentialemployeesisassociatedwithcosts.

13 BasedontheOECD-WTOTradeinValueAdded(TIVA)databaseavailableat:http://stats.oecd.org/Index.aspx?DataSetCode=TIVA_OECD_WTO.

14 SeeFrancoisetal.(2011)andDavidsonandMatusz(2010)foroverviews.

15 Individualsaffectedbyjoblossduringadjustmentperiodsarelikelytosufferbothintheshortrun(unemployment)andpossiblyinthemediumtolongrun(lowerwages).See,forinstance,theworkbyKletzer(2000;2001)andbyEbensteinetal.(2009).

16 Hasanetal.(2012)usestateandindustry-leveldataforIndia.Theydonotfindanyevidenceofunemploymentincreasingfollowingtradereform.Theirstate-levelanalysisrevealsthaturbanunemploymentdeclineswithtradeopeninginstateswithflexiblelabourmarketsandlargeremploymentsharesinnetexporterindustries.

17 Theexpectationthatoppositiontoopeningisdeterminedbythesectorofemploymentstemsfromtheso-called“specificfactormodel”.TheHeckscher-Ohlinmodelpredictsthatfactorsthatarerelativelyscarceinacountrylosefromtradeandmaythusopposeopening.Firmsizemattersinthe

Endnotes


264

so-callednew-newtrademodelsandoneofthespecificitiesinthesemodelsisthatfactorsemployedinrelativelysmallfirmsaremorelikelytolosefromtradeopening.

18 MaydaandRodrik(2005)useinformationcontainedintheInternationalSocialSurveyProgramme(ISSP)thatcoversmorethan20,000individualsin23countries.

19 Othersurveyevidenceindicatesthatindividualsmayalsobeconcernedaboutthecountry-wideemploymenteffectsofglobalization.AndersonandGascon(2007),forinstance,reportthat75percentofrespondentsinaUSsurveyrepliedthat“outsourcinghurtsAmericanworkers”.AnothersurveyshowsthatabouthalfofNorthAmericansandEuropeansthinkthat“freertrade”resultsinmorejobdestructionthanjobcreation(GermanMarshallFund,2007).

20 SeeevidencefromEurobarometerdiscussedinBacchettaandJansen(2011).

21 Authors’calculationsbasedonalargersetofcountriesinthePEWResearchGlobalAttitudesProjectdatabase.

22 SeeBertolaetal.(2006)foranextensivediscussionoftheroleofinequalityinmacroeconomicframeworks.

23 Grigg(1994)ascitedinFieler(2011).

24 Thismaychangeinthefutureasthecountry’sproductionofhigh-endmanufacturedgoodsiswellunderway.

25 InthelightofdisruptionsinfinancialmarketsinthewakeofandduringtheGreatRecession,numerousrecentresearchprojectshavedrawnattentiontotheroleofinequalityinaworldwherefinancialmarketsareimperfect.Ithas,forinstance,beenarguedthatinequalitycanactasanamplificationchannelfortrade-relatedaggregateshocks(PothierandPuy,2012).KumhofandRanciere(2010)illustratethatinequalitycanevenbethemaintriggerofmajorfinancialcrises.

26 ThisisinlinewithHaltiwanger(2011)whoemphasizestheimportanceoffunctioningcreditmarketstoensuresmoothadjustmenttotradereform.

27 Lin(2010)andPisanoandShih(2012)areonlytwoexamplesofavastrecentliteratureontheroleofpublicandprivatesectoractorsindetermininggrowthpaths.WhileLin(2010)focusesondevelopingcountrychallenges,PisanoandShih(2012)analyserelevantchallengesfromanindustrializedcountry’spointofview.Bourguignonetal.(2006)examinewhetherinternationalpoliciesonaid,tradeorfactormovementscanaffecttheinternationaldistributionofincome.Theyfindthattradeopeninginhigh-incomecountriescancontributetoabetterdistributionofincomeatthegloballevel.

28 CasacubertaandGandelman(2010)andMuendler(2010)showthattradeopeninginUruguayandBrazilresultedinhigherjobdestructionthanjobcreation.Displacedworkerswerenotabsorbedbythemostcompetitiveindustriesbutmovedintonon-tradingsectorsoroutofformalemployment.

29 SeeGoosandManning(2007)andAutoretal.(2006)onjobpolarization.

30 SeealsoMitraandRanjan(2011),Pacietal.(2009)andJansenandvonUexkull(2010)ontheroleofsocialprotectioninopeneconomies.

31 Theso-calledBrundtlandReportdefinessustainabledevelopmentasprogressthat“meetstheneedsofthepresentwithoutcompromisingtheabilityoffuturegenerationstomeettheirownneeds”(WorldCommissiononEnvironmentandDevelopment,1987).

32 Principle12ofUnitedNationsConferenceonEnvironmentandDevelopment(1992).

33 PreambletotheMarrakeshAgreementEstablishingtheWorldTradeOrganization.Availableat:www.wto.org.

34 EmissionsdataaretakenfromEuropeanCommissionJointResearchCentre(2011).Morerecentemissionsdataarenotyetavailableattheworldwidelevel.

35 BiodiversitymeasuresaretakenfromWorldBank(2012c).

36 AdjustedNetSaving,anindicatordevelopedbytheWorldBank,capturestheeconomy’struesavingsrateonceinvestmentsinhumancapital,depletionofnaturalresourcesanddamagecausedbypollutionaretakenintoaccount.TheEnvironmentalPerformanceIndex,developedbyYaleandColumbiaUniversitiesintheUnitedStates,isacompositeindexthatcovers22variables,includingchildmortality,SO2emissionspercapita,pesticideregulation,forestlossandCO2emissionspercapita(YaleCenterforEnvironmentalLawandPolicyandCenterforInternationalEarthScienceInformationNetwork,2012).TheEcologicalFootprint,developedbytheGlobalFootprintNetwork,measurestheratiooflandandwaterrequirementstoavailableresourcesinordertosustainthelivingstandardofacountry.The(absoluteandproportional)EnvironmentalImpactindices,developedbyBradshawetal.(2010),measuretheproportionalandabsoluteenvironmentalimpactwithrespecttoeachcountry’s(andtheworld’s)availableresourcesintermsofnaturalforestloss,naturalhabitatconversion,marinecapture,fertilizeruse,waterpollution,threatenedspecies,andcarbonemissions.Inordertofacilitateinterpretation,theEnvironmentalImpactindicesmeasureshavebeentransformedasfollows:-EnvironmentalImpact+200.

37 ThetopCO2emittersin2008wereChina,UnitedStates,theRussianFederation,Indonesia,India,Japan,DemocraticRepublicoftheCongo,Germany,Brazil,Canada,UnitedKingdom,theRepublicofKorea,Mexico,Italy,Australia,France,KingdomofSaudiArabia,IranandSouthAfrica.

38 ThetopSO2emittersin2008wereChina,UnitedStates,India,theRussianFederation,Australia,Kazakhstan,Indonesia,Japan,SouthAfrica,Canada,KingdomofSaudiArabia,Brazil,Mexico,Chile,Turkey,ChineseTaipeiandPeru.

39 ThetopNOxemittersin2008wereChina,UnitedStates,India,theRussianFederation,CentralAfricanRepublic,Brazil,andSudan.

40 Datafor2007,whichisthelatestavailableyear.

41 Cristeaetal.(2011)notethatonekilogrammeofcargoflownonekilometreonaplanegeneratesbetween50and200timestheemissionsofthatsamekilogramme/kilometreonabulkcargocarrier.

42 Thefactorendowmentshypothesisisbasedonthefollowingstylizedfacts:developedcountriestendtobecapitalabundantrelativetodevelopingcountries;andthepollutionintensityofaneconomicsectortendstogohandinhandwithitscapitalintensity.

43 Therelocationofpollution-intensiveproductioncouldentailarelocationofpollution-intensiveindustriesfromcountrieswithstringentenvironmentalpolicytocountrieswithlaxenvironmentalpolicies,oranincreaseintheproduction(andnetexports)ofpollution-intensivegoodsincountrieswithlaxenvironmentalpolicies.

44 Underthepollutionhavenhypothesis,thehigheracountry’spercapitaincomelevel,themorestringentitsenvironmentalpolicy,seeCopelandetal.(2003).


265

II D. Tr

aD

e o

pe

nn

es

s a

nD

Th

e b

ro

aD

er

so

cIo

-e

co

no

mIc

co

nTe

xT

45 Themethodor“technique”usedinproductioncanbelooselydefinedaspollutionperunitofoutput.

46 Forexample,McAusland(2004)usesapartialequilibriummodeloftradetoillustrateaso-called“Californiaeffect”,throughwhichdomesticfirmsseeanincreaseintheirrentsfollowingarequirementtousecleanerinputs.Evenifthestricterrequirementdoesnotapplyoverseas,foreignfirmshaveanincentivetocomplywithit.Sinceproductionissubjecttoincreasingreturnsanddemandinthecountrywiththestricterrequirementislarge,itischeaperforforeignfirmstocomplywiththerequirementtousecleanerinputs.Nonetheless,usingcleanerinputsdrivesuptheproductioncostsofforeignfirmsbymorethanthoseofdomesticfirmsandshiftsworlddemandtowardsdomesticfirms.

47 A“fossilfuelchannel”hasbeensuggestedasanadditionalchannelforcarbonleakage.Thisinvolvesadeclineininternationalfossilfuelprices(duetoadecreaseindemandforfossilfuelsinconstrainedcountries),whichmaytriggeradditionalenergydemandandemissionsinunconstrainedcountries(Morgensternetal.,2007).

48 StudiesoftenexpresscarbonleakageasaratiooftheincreaseinCO2emissionsofunconstrainedcountriesandthereductionintheemissionsofconstrainedcountries.Mostestimatesofcarbonleakagerangefrom5percentto20percent(Elliottetal.,2010).

49 Themostcommonformofemissionstradingschemeisknownascap-and-trade.Underthissystem,anoveralllimitontheamountofcarbonemissionsissetbyacentralauthority,whichthenissuespollutionallowancesorpermitsequivalenttothatceiling.Thepermitsareallocatedtoentitieswhoseactivitiescontributetoemissionsinaccordancewithspecificrulesandconditions,andmaybetradedamongparticipants.Emissionsallowancesmaybeauctionedoffordistributedatnocosttotherecipient.SeeSerresetal.(2010).

50 Feed-intariffschemesofferaguaranteedpriceorpremium(onthemarketpriceforelectricity)foreachunitofelectricityfedintothegridandproducedfromrenewableenergy.Underquotasystems,governmentstypicallyestablishanobligationonautilitycompanyorgroupofcompaniestoprovideapre-determinedminimumshareofrenewableenergyofeitherinstalledcapacityorelectricitygenerated.Quotasystemsarevariouslyknownasrenewableportfoliostandards,renewableelectricitystandards,renewablesobligationandrenewableenergytargets.PreviousdisputesettlementcasesshowthatWTOrulesgivespaceforcountriestoadoptawiderangeofpoliciesinpursuitoflegitimateenvironmentalobjectives.Atthesametime,suchspaceisdisciplinedbyspecificconditionsaimedatensuringmeasuresarenotappliedarbitrarilyandarenotdisguisedrestrictionsoninternationaltrade.Forinstance,theuseofadomesticcontentrequirementintheoperationofafeed-in-tariffwasfoundtobeinconsistentwithWTOMembers’obligationsundertheTRIMSAgreementintherecentCanada–RenewableEnergy/FITProgramcases(WT/DS412/AB/R;WT/DS426/AB/R,Canada–CertainMeasuresAffectingtheRenewableEnergyGenerationSector–Canada–MeasuresRelatingtotheFeed-InTariffProgram–AB-2013-1–ReportsoftheAppellateBody,6May2013).

51 Short-termfinanceisnecessaryformostinternationaltradetransactionbecauseatime-lagexistsbetweentheproductionofthegoodsandtheirshipmentbytheexporter,ontheonehand,andreceptionbytheimporter,ontheother.Generally,exporterswouldrequirepayment,atthelatest,uponshipment(attheearliestuponordering),whileimporterswouldexpecttopay,attheearliest,uponreception.Thistimelaggenerallyjustifiestheexistenceofacreditoraguaranteeofpayment.Thecreditcaneitherbeextendeddirectlybetweenfirms–asupplierorabuyer’scredit,orbybankingintermediaries,whichmayoffertheexporterortheimportertocarryforthempartofpaymentrisk(andsomeotherrisksinvolvedintheinternationaltradetransaction)forafee.Forexample,underaletterofcredit,thebankofthebuyerprovidesaguaranteetothesellerthatitwillbepaidregardlessofwhetherthebuyerultimatelyfailstopay.Theriskthatthebuyerwillfailtopayishencetransferredfromthesellertotheletterofcredit’sissuer.

52 Forexample,the2009IMF-BAFTsurveycoveringtheperiodfromthethirdquarterof2008tothefirstquarterof2009indicatesthattheflowsofsecuredorunsecuredtradefinancetodevelopingcountrieshadfallenmorethantheflowsoftradein2008,calculatedonayear-on-yearbasis.

53 InitsFinancialStabilityReviewofJune2012,theEuropeanCentralBank(ECB)isconcernedthatthedeleveragingprocesscouldadverselyaffectthesupplyofcredittotherealeconomyintheEuroarea.AccordingtotheECB,suchconcernsaremorerelevantfortheEuroareathanforotherlargeeconomies,owingtothepredominantroleofbanksinthefinancingoftheeconomy.

54 Inits2010Annual Report,theBISestimatedthatinthetwoyearsbetweentheonsetofthefinancialcrisisandthepublicationofthatreport,internationalbankshadexperiencedcumulatedlossesonwrite-downsofassetsofsomeUS$1.3trillion,metbytotalrecapitalizationofUS$1.2trillion.Sincethen,theBISnolongerreportsthisfigurebutitislikelytohaveincreased.

55 MoredetailsontheIMFnormativeestimatesofequilibriumexchangeratescanbeobtainedontheIMFwebsite(www.imf.org).Ofparticularinterestisthe2012methodologicalnote:“ExternalBalanceAssessment(EBA):TechnicalBackgroundofthePilotMethodology”.

56 SeeinparticulartheseminarheldbytheWTOWorkingGrouponTrade,DebtandFinance,availableathttp://www.wto.org/english/news_e/news12_e/devel_27mar12_e.htm.

57 SeealsoSectionD.7ofWTO(2007)onthisissue.

Documents

D. Trade openness and the broader socio-economic context · the broader socio-economic context. ... necessarily reflected in this trend. In Sub-Saharan Africa, for instance, Nigeria