Lvp5 AX-733
POLICY
RESEARCH
WORKING
Trade Policy Reform
and Poverty Alleviation
PAPER
2733
How to implement trade
liberalization as part of a
strategy for alleviating poverty
in developing countries.
BernardHoekman
Constantine Michalopoulos
Maurice Schiff
David Tarr
The World Bank
DevelopmentResearchGroup
Trade
December2001
I
POLICYRESEARCH
WORKINGPAPER2733
Summary findings
In this paper, developed as part of the World Bank's
Poverty Reduction Strategy Sourcebook, Hoekman,
Michalopoulos, Schiff, and Tarr examine how to
implement trade liberalization as part of a strategy for
alleviating poverty in developing countries. They discuss
trade policy instruments, institutions, complementary
policies, sector issues, adjustment policies, and safety nets
in an integrated approach to trade policy as a tool for
poverty alleviation.
The authors examine the patterns or models of trade
policy that have been successful in alleviating poverty.
They discuss the role of tariffs, nontariff barriers,
contingent protection (such as safeguards and
antidumping), special import regimes (such as duty
drawback), export taxes, export subsidies, and traderelated institutions (such as standards, marketing, export
finance, customs clearance, and regional trade
arrangements).
The authors also discuss policies that complement
successful trade reform, including macroeconomic
stability, a competitive exchange rate, flexible labor
markets, competitive product markets, and policies that
do not discriminate against foreigners in investment.
They suggest approaches to policies and institutions in
services and agriculture, key sectors in poverty reduction.
They explain the roles of retraining and safety nets in
dealing with the adjustment costs of trade liberalization.
Finally, the authors elaborate guidelines for
implementing trade reform and explain tools for
assessing whether trade reform will help or harm the
poor in particular sectors in the short run.
This paper-a product of Trade, Development Research Group-is part of a larger effort in the group to.understand the
links between trade reforms and poverty reduction. Copies of the paper are available free from the World Bank, 1818 H
Street NW, Washington, DC 20433. Please contact Rebecca Martin, room MC3-308, telephone 202-473-9065, fax 202522-1159, email address rmartinl@worldbank.org. Policy Research Working Papers are also posted on the Web at http:/
/econ.worldbank.org. The authors may be contacted at bhoekman@worldbank.org, cmichalopoulos@worldbank.org,
mschiff@worldbank.org, or dtarr@worldbank.org. December 2001. (44 pages)
The PolicyResearcbWorkingPaperSeoesdisseminatesthe findingsof work in progressto encouragethe exchangeof ideasabout
development issues.An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished The
paperscarry the names of the authors and sbould be cited accordingly. The findings, interpretations, and conclusionsexpressed in this
paper are entirely those of the authors. They do not necessarilyrepresent the view of the World Bank, its Executive Directors, or the
countriesthey represent.
Produced by the Policy Research Dissemination Center
Trade Policy Reform
and Poverty Alleviation
*
Bernard Hoekman **
Constantine Michalopoulos *
Maurice Schiff **
David Tarr **
Preparedby the DevelopmentResearchGroup,InternationalTradeTeam of the World Bankfor the
PovertyReductionStrategysourcebookas a guide to help the least developedcountries employtrade
policy and institutionsin their poverty reductionstrategies.Pilot studies in Cambodia,Mauritania,and
Madagascarwere beingundertakenduring the summerand fall of 2001. This paper was preparedwith
assistancefrom CarstenFink, FaezehForoutan,AadityaMattoo,MarceloOlarreaga,Garry Pursell,and
the staff of the IMFPolicy Developmentand ReviewDepartment-in particular,GeoffreyBannister,Anne
McGuirk,and KamauThugge.The report benefitedfrom helpfulcommentsand suggestionsfrom Brian
Ames, Ataman Aksoy, Ishac Diwan, Philip English, Simon Evenett, Mike Finger, Jeni Klugman, Will
Martin, Brad McDonald,Richard Newfarmer,Jayanta Roy, and John Wilson. Maria Kasilag provided
excellentsupportin the preparationof the manuscript.
DevelopmentResearchGroup, InternationalTradeTeam,WorldBank
SeniorEconomicAdvisor,WorldBank
Table of Contents
1
1.
Introduction..........................................
2.
TradePolicyReform,Growthand PovertyAlleviation..........................................3
3.
TradePolicyInstrumentsand Institutions
..........................................
4.
ComplementaryPoliciesto a TradeReform.........................................
14
5.
SectorIssues.........................................
19
6.
Copingwith Uncertainty:The Roleof SafetyNets......................................... 28
7.
TradePolicyReform:GuidelinesFor PovertyAlleviation......................................
30
Appendix.Impactof Trade Reformon the Poor:A SimpleFramework.36
iii
6
1 Introduction
Trade and Growth. There is a preponderanceof cross-countryevidence that trade
liberalizationand opennessto trade increasesthe growth rate of incomeand output.' In
addition,numerous individualcountry studiesover the past three decadessuggest that
"trade does seem to create,even sustainhighergrowth"(Bhagwatiand Srinivasan1999).
A country'strade policy is the key link in the transmissionof price signalsfrom the world
market to the national economy. Undistorted price signals from world markets, in
combinationwith the exchangerate,allow resourceallocationconsistentwith comparative
advantage, thereby increasing productivity.An open trade and investment regime
encouragesintegrationinto the globaltradingenvironmentand the import of diverseand
2
modemtechnologiesthat are importantfor productivityimprovements.
Trade,Growthand the Poor.Growthin incomesof the poor is stronglyrelatedto
overallgrowth in the economy-althoughthe relationshipdifferssubstantiallyfrom country
to country.The link of overall growthto povertyalleviationhas beendemonstratedboth in
3 Trade
cross country analyses (Dollar and Kraay, 2000), and for individual countries.
liberalizationcan therefore be expectedto help the poor overall, given the positive
associationbetweenopennessand growth. As important,the political economyof trade
policyis such that in practiceit is unlikelythat the poor havemuch influenceon the pattern
of protection that prevails in a country. As trade policy inherently has the effect of
redistributingincome, this implies that trade policy often acts to tax the poor, and that
liberalizationcan thereforedo muchto improveincomesof the poor. However,in the short
run liberalizationmay have a negativeimpact on some of the poor, dependingon their
sources of incomeand the impact on prices of goods and servicesthe poor consume.
Thus,there is a needto examinethe impactof trade liberalizationin some detailto helpto
design policies that protect those among the poor who may be adversely effected,
especiallyin the shortrun.
Characteristicsof the Poor. Knowledgeof how the poor obtain and how the
poor spend their income is important in designing pro-poor policies. As explained at
greater length in the Appendixto this chapter,trade policy will affect the welfare of the
poor through what they consume and what they produce. Food is by far the most
important item of expenditure.In Bangladesh,for example,food representsabout 73
percent of the total expendituresof the poor. Fuels, housing and clothing combined
4 The sale of unskilledlabor tends to be the
accountfor about21 percentof expenditures.
most important source of income for the poor, complementedby the value of "own
consumption,"that is, the imputed value of what the poor consume from their own
5
production.
' Sachsand Wamer (1995),Dollar (1992),Edwards(1993, 1998),Ben David(1993)and Frankeland Romer
(1999).
Coe,Helpmanand Hoffmaister(1997)for evidenceand (Romer,1994)for a furtherdiscussion.
Forexample,Srinivasan(2000)foundthat of the 17 percentagepoint reductionin the populationbelowthe
povertyline over some 40 years (between1951-55and 1993-94),a 15 percentagepoint reductionis to be
attributedto growthand 2 to redistributivepolicies.
the 1991HouseholdExpenditureSurveyfor Bangladesh.
4From
5 In Mexico,wages from unskilledlabor represents40 to 45 percentof the incomeof the poor, and own
consumptionis the next largestat 15 percent(lanchovichinaet al., 2000).
3See
In general,the impacton the sourceof incomeof the poorwill be a more important
determinantof the effect of liberalizationthan the effect on the pricesof their consumption.
The reasonis that trade reformwill affectmanyrelativeprices, someof whichwill move in
offsetting directions. In contrast, as the poor generally have limited assets, the most
importantof whichis low-skilledlabor,the impacton wagesandemploymentdominates.
ComplementaryPolicies and Institutions.While a liberal trade policy is
necessaryfor growthand povertyreduction,it alone is not sufficientfor growth of trade.
When trade reformhas been implementedin an unstablemacroeconomicenvironmentor
without efforts at strengthening trade-related domestic institutions and appropriate
complementarypolicies,it has ofteneither been reversedor failed to stimulategrowth. In
this chapter, we will also discuss important "behind the border' reforms that should
accompany trade liberalizationto effectively allow a country to integrate into world
markets. This includesa regime that encouragesinvestmentand competition,including
openness to foreign direct investment(FDI), so that business services are supplied at
competitive prices, and macroeconomicpolicies that encourage stable prices and a
competitivereal exchangerate.6 Althoughthe poor are very diverse,they frequentlywork
in the rural sector and in the urban informalsector.Thus, policiesthat affect agricultural
and labormarketsare importantcomplementsto trade reformfor the poor.
Political Economyof Protectionand the Poor. Even when trade reform will
benefit the poor and the economybroadly,it will often be resisted.The sectors with the
highest protectionknow they receive concentratedgains from protectionand they will
oppose the reform. The expansionof exports following reform is likely to be spread
throughoutthe economy,often with new and sometimesunexpectedindustriesarising.It
is oftendifficult to identifyfutureexportsand exporters.Thus,the employmentand income
gains from reform are likely to be diffuse. The same is true for the consumerswho will
gain from the reform through lower prices and greater choice.The diffuse nature of the
gainsto consumersand producersexplainswhy thosewho opposeliberalizationoftenare
dominant in the political lobbying.Redistributiveeffects of trade reformcan be a major
factor impedingthe launch of welfare-improvingpolicychanges(Rodrik,1998).
Abstractingfrom fiscal revenueconsiderations,barriersto trade are typicallyput in
place to protect domestic producersfrom internationalcompetition,and usually benefit
powerfulinterest groups,not the poor. Non-tariffbarriersare especiallyperniciousin this
regard as they result in the transferof rents from consumersbroadly,includingthe poor,
to licenseholders.Such rents arise becausethe restrictionon importsresultsin domestic
prices that are above the world price. Whereas in the case of a tariff the government
collects the revenue that is implied by the differencebetweenthe world price and the
tariff-inclusivedomestic price, in the case of non-tariff barriers this implicit revenue is
capturedby those who havethe right (holdthe license)to import.There is evidencethat in
developing countries such rents are a major source of inefficiencyfrom rent-seeking
activity(Krueger,1974),i.e., from the spendingof real resourcesto obtainimport licenses
and influencepolicy in general. These license holders are often among the wealthiest
members of society. Thus, in addition to the inefficiency costs of trade protection,
protectionwill often transfer incometowardthe rich and away from the poor. Indeed, in
6 One view, representedby Rodriguezand Rodrik (2000), is that there is a strong associationbetween
exportsand growth,but that we are unableto be sure whetherthis associationis a consequenceof exports
causingoutputgrowth,or of the two beingjointly determinedby the strengthof countries'institutions.Evenif
we were to acceptthis agnosticview, the resultingpolicy prescriptionis a relativelycomprehensiveapproach
to developmentthat includesimprovementof a range of institutionsalongwith tradeliberalization.
2
percentageterms the effect of trade reformon the incomesof those who currentlygain
and those who currentlylose from reform can easily be a multipleof the economy-wide
welfaregainsfrom liberalization.This is becausetrade policyis inherentlya redistributive
policy.
Structureof this Chapter.In the next sectionof this chapterwe describe briefly
the experiencewith successfultrade policyreformand discussadjustmentcosts and the
implicationsfor the poor. Section III discussesand evaluatesthe principaltrade policy
instrumentsand institutions.In sectionIV the most importantcomplementarypoliciesare
discussedand evaluated.Giventhe importanceof agricultureand businessservicesfor a
poverty reduction strategy that employs trade, these sectors are accorded special
treatment in section V. General and trade policy specific safety nets are discussed in
section VI. We brieflysummarizethe principalpoints in a successfulstrategyfor using
trade for poverty reductionin the last section.The appendixprovides a frameworkfor
thinkingaboutthe impacton the poorfrom a trade reformin the shortand the longrun.7
2 Trade Policy Reform, Growth and Poverty Alleviation
2.1 Models of Successful Trade Integration and Poverty Reduction
In the last 40 years of the 20Qh
century,severalcountrieshave been highly successfulin
increasingincomesand reducingpoverty.Most notable,has beenthe experienceof the
South-EastAsian economies,especiallySingapore,Hong Kong, Japan,Taiwan (China)
and Korea. In the final 15 years of the century,Chile and Mauritiusalso saw remarkable
increasesin income.All of these countriesdramaticallyincreasedtheir exports (andtrade
to GDP ratio), raisedincomesand reducedpovertyand are now active participantsin the
global trading environment.On the other hand,there are no examplesof countriesthat
have significantlyreduced povertywithoutsignificantlyincreasingtheir exports.Although
export expansionis the commonelementto all the successstories of poverty reduction,
there are considerabledifferencesin the modelsof trade policythat these countrieshave
adopted.The successstoriesmay be groupedinto three broadcategories:
* Economy-wideTrade Liberalization.Countries like Hong Kong, Singapore and
*
Chile adoptedvery liberal trade regimes.These countriesavoidednon-tariff barriers.
HongKongand Singaporepracticedfree trade (zerotariffs),while Chile employedlow
uniformtariffs.
Protectionwith OffsettingPoliciesfor Exporters.Somecountriesthat experienced
rapid growthin trade and GDP did so in the contextof trade regimescharacterizedby
significantimportcontrolson the domesticmarket.Korea,Taiwan(China) and Japan
(in the early stages) are the main examples. The key to understandingthese
experienceshas to do with lookingat all the factors that affect competitivenessand
the incentivesto producersto sell in the domesticversusthe export market.Protection
' Spaceconstraintspreventa detaileddiscussionof manyof the subjectsaddressedin this
chapter.Morein-depthtreatments
of tradepolicyfroma development
perspective
canbe foundon
the websitewww.worldbank.org/trade
and in two recentcompilations:
McColluch,Wintersand
Cirera(forthcoming)
-andHoekman,Englishand Mattoo(forthcoming).
TheWorldBanksite also
offersaccessto toolsthatcan be usedto assistin thedesignof tradereforms,andcross-country
dataon tradepoliciesthatare usefulfor benchmarking
purposes.
Specifictoolsthatfocuson the
impactof reformonthepoorarestillbeingdeveloped
andwillbepostedonceavailable.
3
*
creates incentivesto produceand sell to the domesticmarket,i.e., protectioncreates
a bias against exports. Indeed, protectionof intermediateproducts and services
creates a serious handicapto export industriesbecauseit raisestheir costs to levels
that are higherthan those of their potentialcompetitorsin world markets,and they are
discouragedthrough the impact of protectionon the real exchange rate. But these
countries managedelaborate systemsthat offset the bias against exports (Box 1).
The experiencesof nearlyall other developingcountrieswith high protectionof their
domestic markets are quite different. Notably,the institutionsneeded for effective
implementationof duty drawbacksystemshave been shownto be weak in most subSaharaAfrican countries(WorldBank,2000c).This suggeststhat the Koreanmodelis
not relevantfor most developingcountriesin that it is too difficult to manageand too
easily taken over by special interestswhowill obtainrents.
Protection with Export ProcessingZones. In a protected trade regime that
discouragesexports, export processingzones (EPZs) may be used to partly place
exporterson a footing equal to producersfor the domesticmarket.While very many
countries have introducedEPZs,few have actuallysucceededin stimulatingexports
substantiallyand on a sustainablebasis through this mechanism(see below). The
experienceof Mauritiusin the last 15 years of the 20t century,however,provides an
exampleof a countrythat expandedexportssignificantlyand reducedpovertythrough
EPZs in a trade regimethat was not liberaloverall.Exportersin an EPZ havetariff free
access to intermediateinputs and often have fewer regulatory constraintson their
actions. As a result they may are able to compete on a footing almost equal to
producersfor the domesticmarket.Althoughthere is a role for EPZs in countrieswith
institutional constraints (Watson, 2000), there are potential piffalls in establishing
successfulEPZs (see below).
Box 1. How KoreaOvercamethe Anti-ExportBias of ProtectiveTradePolicies
Duringthe 1960sand 1970sespecially,Korea had relativelyhigh importtariffs and
also used non-tariffbarriers,but both exportsand the economygrew very rapidly.In
principle,tariffs imposea bias againstexports by making productionfor the home
market more profitable.Korea managedan elaboratesystem that offset the bias
against exports.Two of the most importantmechanismsused by Koreawere dutyfree access to intermediatesused in productionfor export (temporaryadmission)
and capital subsidies to exporters. Korea managed a very efficient temporary
admission system (supplementedin the 1980s by an efficient duty drawback
system) which allowedtariff free access to intermediateinputs for exporters (and
'indirect' exporters).Also importantwas that exporters(and indirectexporters)had
preferred access to working capital at interest rates that were considerablylower
than the interest rates paid by firms supplying the domestic market. The
developmentand enforcementof these policiesover a long period requiredstrong
political will supported by a broad national consensus, and a highly competent
administration.
Source: Based on Westphal and Kim (1977), Westphal (1991) and Rhee, RossLarson and Pursell (1984). These studies provide a full treatment of the
mechanismsemployedby Korea to provideexporterswith incentivesto offset the
bias of the tariffs.
In practice, the most practicalway of stimulating trade and opening up to the
internationaleconomy is through liberal trade regimes, rather than through a complex
4
structure of protection and export incentives.As a practical matter, duty drawback
mechanismsare ineffectivein most of the least developedcountries.Thus, a regimewith
high protectionwill diminish exports and growth. Moreover,differentiatedstructuresof
protectionand subsidizationcreate opportunitiesfor elites and powerfulproducergroups
to capturetrade policyfor their specialinterests.This lobbyingfor protectionand subsidies
engenderscorruptionand inefficiencieswhich, in the end hurt the poor. These problems
can be avoided by simple and transparentprotectionregimesof low uniformtariffs. For
low income countries with weak trade related institutions, EPZs can be an effective
vehicle to promoteexport production,not only because they allow duty-free access to
importedinputs, but becausethey providea meansto deal with infrastructureand public
sectorservice-relatedweaknessesthat impedeinvestment(Watson,2000).
Z;;s
=Costs
of Trade Liberalization: Impact on the Poor
Trade reform may not be implementeddue to politicaldynamicsor a fear of adjustment
costs. A recent review of over 50 empiricalstudies of adjustmentcosts by Matusz and
Tarr (2000)found that all the evidencesupportsthe view that the adjustmentcoststo the
economyare very small in relationto the benefitsof trade liberalization.For example,a
retrospectivestudy of trade reform in 19 developingcountriesby Papageorgiou,Choksi
and Michaely (1990) concluded that trade liberalizationdid not generally result in
decreased employment,even in the short run. Compared with the pre-liberalization
period, manufacturingemploymentwas larger one year subsequentto the completionof
liberalizationin all but one of the twelve countriesfor which data was reported. In fact,
manufacturingemploymentwas higher in twelve of thirteencasesduring the liberalization
period comparedwith the levels registeredprior to liberalization.Parker et al. (1995) and
Harrison and Revenga(1995) performedsimilar but less extensivestudies with similar
resultsfor marketeconomies.
Althoughfurther evidenceis neededto generalizethese resultsfully to developing
countries-most of the literature has focused on industrializedcountry experienceLiedholmand Meade (1995)found there were very high-start up rates (twentypercent)
for mediumand small enterprises(MSEs) in six African economiesand the Dominican
Republic,where MSEs account(as they do elsewhereas well) for a significantportionof
total employment.The high start-uprates,wherethe businessenvironmentis supportive,
suggestthat entrepreneursin these countriesare quick to respondto new opportunities,
making speedy adjustmentto trade reform quite likely. Looked at another way, the
magnitudeof dislocationcausedby liberalizationis unlikelyto be significantlylarger than
dislocationsassociatedwith the everydayworkingsof the economyin manycountries.
Private adjustmentcosts can be substantialor very small dependingon whether
displacedworkerswere earningrents in their initial positionand whethermarketsfunction
reasonablywell. The poorare not likelyto be earningrentsor they would not be poor. The
evidence shows that adjustment costs are typically short term and terminate when
workers find a job, while the benefits of trade reformcan be expectedto grow with the
economy.In developingeconomies,trade liberalizationshould favor labor since exports
will typically be labor intensive.Significantwithin-industryshiftstypicallyoccur after trade
liberalization,which tend to minimizethe dislocationof factors of production.Moreover,
the duration of unemploymentfor most industriesis not high, especiallywhere workers
were not earning substantialrents in the originaljob. Finally, in many industries normal
labor turnover exceeds dislocationfrom trade liberalization,so that downsizing where
necessarycouldbe accomplishedwithoutcausingmuch unemployment.
5
It is difficult to disentanglethe effects of trade liberalizationfrom other events
occurringsimultaneously,but generallymanufacturingemploymentincreasedsubsequent
to the trade liberalization.Thus, while it is necessaryto apply caveats to most of the
individualstudies, the collectiveweightof their evidencesuggeststhat adjustmentcosts
are low relativeto the gainsfrom trade liberalization.Nonetheless,the extreme poor may
be incapableof sustainingeven short periodswith adverseadjustmentcosts, and there
may be selectedpoor groupsthat do not gain, while some of those that gain may lose in
the short run. In order to minimizeadjustmentcosts and help make trade reform work
most effectively for the poor, complementarypolicies are necessary. One important
complementarypolicyfor the poor is an efficientsocial safety net (discussedbelow).In
general, attaining and sustaining a high rate of economic growth is a key factor in
improvingoutcomesfor the poor over time.
NKM9-.6t
,
Emma
The first step in designinga strategyto use trade policyfor growthand poverty alleviation
is to understandhow the trade regimeworks.This involveslearningof the importanceof
non-tariff barriers,includingwho gets the licenses and permits.What is the structure of
the tariff and includingits dispersion,exemptionsand rebates?Howmuch revenuecomes
from tariffs? What policies are in place which may tax or subsidizeexports?Are traderelated institutions, such as standards organizations,export finance and marketing
facilities adequateto support an expansionof exports? Does the pattern of protection
favor the incomeof the poor or a segmentof the poor, and may policies be designedto
assistthe poorduringthe transition?
In this section we describe the principal trade policies and institutions that are
employedto influencethe flow of goodsand servicesand suggesthow to evaluatethese.
We discusscomplementarypoliciesin later sections.
Non-tariff barriersinclude mechanismssuch as quotas, licensesand monopolyrights to
import.When these mechanismsare in placefor reasonsother than for health or safety
they are the most perniciousof trade barriersin terms of their harmto growth and poverty
alleviation.Partly this is because non-tariff barriers encouragecompeting interests to
lobbyto obtainthe valuablelicensesto import.This competinglobbyingactivity(knownas
urent-seeking")
wastesvaluableresources. Non-tariffbarriersalso lack transparency,and
thereby may allow protectionto go relativelyunnoticed.As discussedabove, the political
economyof protectionsuggeststhat importcontrols(and sometimesexport controls)are
usuallyput in placeto benefitpowerfulinterestgroups,not to help the poor.
Most low-incomedevelopingcountrieshave differentiatedtariff structureswith significant
tariff escalation(Michalopoulos,1999).The main reasonsincludefiscal objectives,import
substitution motivationscombined with the political weight of vested interests. Tariff
escalationis a problemsince it affordshigh"effective"protectionto final goods producers,
thereby discouraging the development of intermediate industries. Exporting of
intermediateproductsis an importantway for developingcountriesto participatein the
modernglobaleconomy;butthese activitiesare discouragedby the escalationof tariffs.
6
A uniformtariff conveysa numberof advantages(Tarr, 2001),the most important
of whichis that if the tariff is uniform,the gainsto industrylobbyingare much smaller(and
may be negative), creating a kind of free-rider problem for the lobbying industry and
dramaticallyreducesthe incentiveto lobbyfor protection.Chile,which has had a uniform
tariff since 1979,is a dramaticcase in point.In Chile,in 1998,the legislatureconsidereda
progressivereductionof the uniformtarifffrom 11 to 6 percent,to be accomplishedby one
percent per year reductionsthough2003. Chileanindustrygroups supporteda reduction
of the tariff, which passed the Chilean legislature. Evidently, uniform tariffs led
industrialiststo concludethat a reductionwas in their collectiveinterest.
A uniformtariff greatlysimplifiescustomsoperations,eliminatesa numberof ways
used to avoid paying the tariff, and should help reduce corruption (which may have
positivespillovereffectsinto other dimensionsof governmentactivity)and save on scarce
administrativeresources.There will also be a direct saving of resourcesfrom reduced
lobbying for higher protection, and an associated gain from encouraging scarce
entrepreneurialtalent to be employedmore productivelyin the creation of better and
cheaper products.Overall,the level of protectionis likely to be lower as the incentiveto
lobby for higher tariffs is attenuated.Many of these factors are pro-poor as they greatly
reducethe scopefor the exerciseof powerand rent-seeking.
Customs and other officials in low-incomecountries have tended to argue that
althoughuniformtariffs look good in theory,implementingsuch a structureis not feasible
in practice. This is frequentlya covert argumentfor continued protectionof particular
sectors. In fact, in addition to Chile, El Salvador and the Kyrgyz Republic have tariff
structureswith very small dispersion,and Hong Kong, Singaporeand Estonia have a
uniformtariff of zero.
Uniformitydoes not implythat that there can be no exemptionsfor productsthat are
deemed to be of great social importancesuch as essentialmedicinesor mosquitonets
(Bannister and Thugge, 2001). However,care should be taken that such exceptions
targetonly productsthat are criticalto attain socialand publichealthobjectives.
3.3 Emergency Protection, Antidumping and Other Trade Remedies
Liberalizationof trade will give rise to adjustment pressures as import competition
intensifies.For this reasonreform programsoften reducetariffs and other trade barriers
gradually and are ideally complemented by measures to facilitate adjustment
(complementaryreforms-see below).However,in some instancesimportsmay expand
so rapidly that governmentsmay want to be able to interveneby raising trade barriers.
Import surges that create serious difficultiesfor a domestic industry may also occur
independentof liberalization.In such cases trade interventionmay be the most effective
instrumentavailable to governments.If so, the appropriateaction to take is a so-called
safeguardmeasure-the impositionof a temporaryduty againstall importsof the product
concerned.A necessaryconditionfor interventionshould be that the domesticindustryis
seriouslyinjuredby importcompetition,and a determinationby economicdecisionmakers
that protectionis in the nationalinterest(i.e., the associatedcosts to consumersare less
than the benefits that would accrue to producers). Safeguard actions should be
temporary,lasting no more than 2 or 3 years, as the objectiveis not to raise protection
permanentlybut to facilitateadjustmentby the industryto increasedcompetition.It is also
importantthat tariffs be used,not quotas,for transparencyand politicaleconomyreasons.
7
Many countriestend to use antidumpingas a safeguard instrument.This is ill
advised. Antidumpingis a trade policy instrumentthat allows duties to be imposed on
imports that are sold for less than what is chargedin the exportershome market.That is,
it can be invoked to offset price discriminationacross markets.Such differentialpricing
usuallyreflectseconomicconditions,and is not detrimentalto welfare.As antidumpingis
an instrument that is easily captured by industriesto raise the price of imports, and
requires the use of scarce administrativeresources,it is counterproductiveto economic
developmentand povertyreduction.If there is a needto raise protectionbecauseimports
injure domesticindustry,it is preferableto useWTO consistentsafeguardactions, as they
allow the country to consider the impact of taking action on the economyas a whole,
8
includingthe poor, as opposedto simply the industrythat confrontsimport competition.
Antidumpingprocedurestypically replace considerationsof the national interest or the
impact on the poor with an assessmentof pricing practicesof foreignfirms. That is, the
question of whether it is in the nation's interestto offer protectionto the industry under
considerationfor a duty is neverasked.
The best policy optionfrom a developmentperspectivein this area is to have no
antidumpinginstrument.If antidumpingproceduresare adopted,efforts should be made
to establishproceduresthat allow for the nationalinterestand the impact on the poor to
be taken into accountpriorto the impositionof an antidumpingduty.
*-I.
___-
.
Tariffs on imports hurt exports in various ways. Tariffs reduce the demand for foreign
exchange, and that tends to appreciatethe real exchange rate. The appreciatedreal
exchange rate hurts exportersand competitivenessof nationalfirms as they must pay
higher pricesfor their importedintermediates.Prices of labor and capital are also bid up
by importcompetingsectorsthat are favoredby the tariff.
Programslike duty drawbackand temporaryadmission,if properlyadministered,
allow exporters duty-free access to imported intermediates. Although the exchange rate
9 provisionof tariff-freeaccessto
bias againstexportsthat resultsfrom protectionremains,
imported intermediatesfor exportersis crucial. This is also the case for countries with
uniformtariffs,althoughthe needfor duty-freeaccessdeclinesas the averagelevelof the
tariff falls.
The principalproblem with duty drawbackschemes (involvingthe repaymentof
duties paid on imported inputs that are embodiedin subsequentexports) is that their
administrationcan be very costly, and leadto cumbersomeproceduresand delays when
tariffs are high. Exporterscomplainof delays and lack of re-paymentin many countries.
When tariffs are high there is also the risk of fraudulentclaims. The empiricalevidence
suggeststhat in countrieswithout well-functioningpublic administrations,duty drawback
becomesineffective.They are very difficultto administerat tariff rates in excess of 15 or
20 percent becauseof leakage,delays in paymentand fraudulentclaims (Mitra, 1992).
UndertheWTO,safeguards
alsorequirecompensation
to be offeredto exportingcountriesif the
actionlastsmorethan3 years.Thisis a usefulmechanism
to ensurethatprotectionis temporary.
SeeHoekman
andKostecki(2001)for moredetaileddiscussion.
8
In addition,to avoid anti-exportbias, dutydrawbackschemeswouldhaveto be extendedto indirect
exportersas well(i.e.,firmswhichdo not exportthemselves
but whichsell to exporters).
Administration
of
suchmechanisms
arequitecomplicated
in practice
formostdeveloping
countries..
9
8
Delays are particularlydetrimentalto small and medium enterprisesand small farmer
organizations.
Temporary admission may be more effective at allowing tariff free access to
intermediateinputs for exporters in these situationsas they do not involve payment of
duties on importedinputs,but rathera requirementthat firmsdocumentex postthat these
inputs have been used in the productionof exports.The major potentialproblemwith this
approachin low-incomecountrieswith weak administrativecapacityis so-calledleakage
of goods into the economy (that is, the goods are not used for export production).A
frequentlyused option to control such leakageis the bondedwarehouse,or on a greater
scale, an export processingzone. These are specificterritories that are controlled by
customsand where imports are not taxed on entry, but goods are taxed if sold into the
domesticmarket.Exportprocessingzonesare discussedfurtherbelow.
3.5 Export Subsidies
Export subsidiesmay be appropriateto offset marketfailures, for example,information
problems associated with breaking into new markets by individualfirms. In practice,
to promoteexports and to offset the anticountrieshave often used them indiscriminately
export bias createdby other policies.While theycan stimulateexports,they maydo so at
a huge cost to the budget.From a poverty perspectivetheir impact on the poor may be
detrimentalas they may result mostlyin rentsto relativelyrich exporters.Using subsidies
to offset the negative effect of other policies (such as protection) on exports is
inappropriate.Such situations call for the removal of protection and the adoption of
instrumentssuch as drawbackor EPZs.
DevelopingcountryWTO membershavebecomemore constrainedin the use of
export subsidies,but least developedcountriesand countrieswith per capita incomes
below $1000 are permittedto use them. This provideslegal coverto pursuepoliciesthat
aim to offset informationalasymmetriesand relatedmarketfailures. To be effectiveand
not distortincentives,such policiesshouldbe 'horizontal'in nature,not sector-specific.
The main problemwith export subsidiesis their use by high-incomecountriesfor
agriculturalcommodities.This has a destabilizingeffect on world prices and is highly
detrimentalto producersin developingcountriesof the productsconcerned(see below).
3.6 Export Taxes
Developingcountriesoften imposeexport taxes on primarycommodityexports.In some
casesthese are imposedinsteadof royaltiesfor the extractionof minerals;in others they
are used in an attemptto exercisemarketpoweror to supportlocalprocessingindustries.
In such casesthere is oftenan adverseeffect on the poor (Box2).
Export taxes result in a lower price for the farmers than the prevailing price in
world markets for their commodities.Eliminationof the tax will raise the incomes of
farmers, the majority of whom tend to be poor, and reduce the profitability of the
establishedprocessing facilities. The latter may employ poor urban labor. Examples
includecocoaand coffee processorsin West Africathat obtaincocoa and coffeebeansat
prices belowthe export price;textile firms in Pakistan,Indiaand FrancophoneWest Africa
that obtaindomesticcotton at favorablepricesfrom parastatalexport monopoliesor as a
result of export taxes or restrictions;leather processingfirms in India that buy local and
9
partly processedhides at low prices as a result of cascadingexport taxes; cashewnut
processors in Mozambiquethat benefit from export taxes on raw cashew nuts; and
cashmerewool processorsin Mongoliathat benefitfrom the price reducingeffect of an
export tax on raw cashmere. In all these cases the export taxes and restrictionsare
economically inefficient because they reduce the incentive to produce the primary
agriculturalproductrelativeto the incentiveto processit.
In most cases,the net effect on povertyof removingexport taxes and restrictions
will be stronglypositive.Thereare manymore poorfarmersthat wouldbenefitfrom export
tax removalthan there are industrialworkerswho might lose,and the highestincidenceof
the most severepovertyis usuallyfound in rural areas.For example,the cashmereexport
tax in Mongoliareducesthe incomesof about250,000goat-herdingfamilies,the majority
of which are poor, with the poorest heavily dependenton incomefrom cashmere. In
contrast, there are only about 2,000 workers employed in cashmere processing.
Moreover,part of the incomethe export tax transfersfrom the goat herdersgoes to the
owners of the processingfirms (Filmer,2001; and Takacs, 1994).1iAs is generallythe
case, safety net policiesfor those adverselyimpactedby exporttax eliminationshould be
carefullyconsidered,especiallywhenthe impacton processorsis severe.
Box 2: Madagascar's Marketing Board for Vanilla Exports:Taxing the poor
In 1960the internationalmarketfor naturalvanillawas dominatedby Madagascar,
the World's lowest cost-producerof high-quality bourbon vanilla. From its dominant
position of 60 percent of world exports of natural vanilla, Madagascarorganized the
bourbon vanilla cartel with Comoros and Reunion, sefting high export prices and
restrictingsupplyby regulatingits domesticmarketthrougha MarketingBoardwhichfixed
low producerpricesand requiredlicensesto grow, prepareand exportvanilla.
If the MarketingBoardwas to be assessedby the effect it had on export prices of
vanilla from Madagascar,it was a clear success.The export price of vanilla increased
from US$10in the late 60s to more than US$ 65 per kilo in the early 90s. However,the
share of Madagascarin world marketsdeclinedto 30 percent as Indonesia(outsidethe
cartel)took advantageof highworld pricesto developa highexport capacity.The entry of
Indonesia into world markets left the total value of Madagascar'sexports constant
throughoutthe 1970sand 80s. The Board's interventionsin the domesticmarket had a
similareffect on producerprices,whichfluctuatedaroundUS$5per kilo duringthe 1980s.
Who benefitedfrom the bourbonvanillacartel and the marketingboard'sdomestic
policies? Indonesianproducers were the winners, Madagascar'sproducersthe losers.
These were small producers (60,000),with an average productionof 130 kilos and an
average incomeof US$650per plantation.Estimatesof producerprices that would have
prevailed in Madagascarhad the marketingboard been abolishedare close to US$26,
well above the US$5 price fixed by the board. Taking into account the increase in
production that such a change in prices would have generated, abolitionwould have
increasedvanillaproducersurpluseight-fold.
'° Muchof cashmereis smuggledout of thecountry.Whilethis avoidsthe tax, herdersstill lose
capturemuchof thedifferencebetween
worldanddomesticprices.
incomeasmiddlemen
10
An alternative to free-trade would have been to eliminate interventionin the
domesticmarket,but to continueto exploitmarketpowerin internationalmarketsthrough
an export tax. Estimatessuggestthat the optimalexport tax would have been close to
US$25per kilo insteadof the US$61implicittax that was imposedon producers.
Why where these policiesnot adopted?A possibleexplanationis that taxation of
small producersgeneratedan importantincomeredistributionfrom the rural poor to the
urbanelite.
Source:De Melo,Olarreagaand Takacs(2000).
3.7 Export Processing Zones (EPZs)
EPZs are "enclaves"for export production,especiallynon-traditionalexports, that are
often used in contextswhereeconomy-widetrade reformis impededor infrastructureand
regulatoryrequirementscannotbe met on a nationalbasis. EffectiveEPZs combineclear
private property rights and investmentregulations,no restrictionson foreign exchange,
tariff free imports for export production, moderate levels of taxation, streamlined
administrativeproceduresand privatesectormanagement.Some of the most successful
EPZs are in Mauritius where, in 1994, EPZs generated 71% of gross exports and
employed 16.6%of the work force. Mexico has also had successfulEPZs. Such zones
can not only expand exports, but may also have a significanteffect in raising female
employment.Bangladeshzones employ a majorityof women (70% of ChittagongEPZ
employeesare women,a much higher ratio than the nationalaverage).In the Dominican
Republic,EPZ employmentwas an importantfactor in decreasingthe share of female
poorfrom 22.6 percentto 15.8 percentover the 1986-93period(Madani,1999).
Experience has shown that development of EPZs, including provision of
infrastructureand management,should be privately handled. Publicly developed and
managed EPZs have typically been unsuccessful.Attractinginvestmentinto EPZs is a
functionof many factors, includingsome that are national-such as politicalstabilityand
sound macroeconomicmanagement.An overvaluedexchange rate will discourage
exporting from EPZs just as much as from the rest of the economy. In general, the
conditionsthat make EPZs successfulinvolvethe same complementarypoliciesthat are
required to make a trade reform successful(see below). Consequently,EPZs are best
regardedas transitionalmechanismsin the pursuitof an overallliberaltrade regime.
3.8 Other Trade Related Institutions
In additionto the commercialpolicy instrumentsdiscussedabove,there are a numberof
other trade-relatedinstitutionsthat can have importantimplicationsfor the impactof trade
reforms.These includecustomsclearance,exportfinance,productstandardsand access
to informationon marketopportunities.
Customs Clearanceefficiencyand transparencyis an important determinantof
the costs associatedwith trade. Burdensomeand redundantprocedures-red tape-can
give rise to substantial uncertaintyand are often associated with rent seeking and
corruption. Minimizing discretion by simplifying as much as possible the clearance
process,includingthrough adoptionof internationalstandardsfor classificationof goods,
elimination of most exemptions (see below) and providing officials with training and
appropriateinformationtechnologyare importantdimensionsof trade reform.
11
Export finance is one of the major constraintsinhibiting exports in many low
income developingcountries.Inadequaciesmay result from the overall weaknessof the
financial sector or may reflect difficultiesin assessing creditworthinessof traders or
becausetradersdo not havesufficientassetsto be judgedcreditworthy.To the extentthat
the poor are involvedin trading activities,they may face special difficultiesin obtaining
accessto the trade credit they need-just as they face such difficultiesin accessingother
parts of the financial sector. While ensuringavailabilityof trade finance is a matterthat
needs to be left to the private sector, any effort to expand exports and to promote
increasedopportunitiesfor the poor in the export sector needs to investigatewhether
credit is a problem. Remediesshould be sought as part of overall efforts to increase
accessto financefor low-incomeproducers.These are discussedelsewherein the PRSP
sourcebook
Product standards based on international norms are important to poverty
reduction.When standardsare used appropriatelyin internationaltrade (as well as in
domestictransactions),they facilitateinterconnectionof goodsand informationexchange,
and guaranteesafety, health,qualityor the environment.As the poor have less accessto
informationand do not have the resourcesto buy higherqualitygoods and services,they
are more dependenton efficientstandardizationand consumerprotectionregimes.
In the area of trade,developingcountriesface a dual challenge:on the one hand
they need to reformregulationsand establishefficienttesting, certification,and laboratory
accreditationrequirementsto minimizethe impact on trade while still attaining sanitary,
phytosanitary,and product standards. Low-income developing countries need both
technicaland financial assistanceto meetthis objective.Institutionalarrangementsneed
to be developedto ensurethat poorfirms (such as farmers,smallproducersand artisans)
have access to standards organizations-through co-operativesand similar collective
organizations-and are not unduly penalized by the use of labor-intensiveproduction
technologies.
Marketing of exports internationallyis a challenging task for all low-income
countries.They have to overcomeproblemsof lack of information,product and country
recognition,and concerns about quality. Foreign partners and FDI can be helpful in
providingthe necessarycontactsand advice;but frequentlyit is necessaryto organizea
local association of exporters or producers.When the producers tend to be a large
numberof small poor farmers, co-operativesand similarventures can be very helpful in
ensuringthat the potentialbenefitsfrom exports are realizedby the poor. But competition
in these institutionsis importantor poorfarmersmayreceivelower pricesfor their outputs
Most developing countries have formed or are members of regional integration
agreements(RIAs). From an efficiencyof resource allocationviewpoint, RIAs between
developing countries (so-called South-Southagreements) are likely to hurt member
countries(WorldBank,2000b).The reasonis that the trade preferencestypicallydisplace
efficiently produced low-priced imports from non-partner countries with inefficiently
1 A solutionto this problem is for
producedhigh-pricedproductsfrom partnercountries."
member countriesto lower their externaltrade barriers,thereby reducingthe inefficient
11 The impactof North-Southor North-Northagreementstends to also involvemoreefficientlypricedimports
frompartnercountriesand thus maybe beneficial.
12
displacementof non-partnercountry imports.In fact, a harmfulRIA can be turned into a
beneficialone by loweringexternaltrade barrierssufficiently.
From the viewpoint of poverty, the effects of an agreement among developing
countries are likely to be asymmetric, with the poorer member countries losing
significantlyand the higher incomememberslikelyto gain (World Bank,2000b).This is, in
part, due to the fact that the relatively developed country produces protected
manufacturesthat are importedon balanceby the less developedcountry.With a RIA, the
more advancedcountry sells its protectedmanufacturesat a higherprice since its exports
are no longer subject to a tariff in the poorercountry'smarkets.This entails a terms-oftrade loss for the poorer country, since it pays a higher price to its partner country for
imports it could buy more cheaplyfrom the rest of the world. This problemof asymmetric
distributionof gains and losses emerged in the 1960sin the East African Community-where Kenya gained and Tanzania and Uganda lost--and in the Central American
CommonMarket-where El Salvadorgained and Honduraslost. In both cases,the more
advancedcountry resistedextemal tariff reduction,which led to resentmentin the less
developedmembercountriesand a breakdownin implementationof the agreements.
Loweringexternaltrade barriersis beneficialboth from the viewpointof efficiency
and from the viewpointof minimizingthe losses for the poorermembercountry. In fact,
the losses from regionalism,both in overall efficiency and in adverse distributional
consequences, can be reduced by lowering external trade barriers. In terms of
implementation,in the case of a free trade agreement(such as the plannedFree Trade
Agreement of the Americas or the ASEAN Free Trade Agreement),each country can
lowerits tariffs independently,and can thus protectitselffrom adverseconsequences.For
a customs union (such as the West African Economic and Monetary Union or
MERCOSUR),the member countriesmust agreeto jointly lower the common extemal
tariff. Thus, customs unions among developingcountries should be approached very
cautiously,since tariff policyis taken out of the handsof nationalpolicy-makers.
Space constraints prohibit a detailed discussionof the WTO, which has many
provisions that constrain the use of trade-related policies by members, including
and www.wto.orgfor more
developingcountries (Box 3). See www.worldbank.org/trade
informationand referencesto the literature.A major dimensionof the WTO that makes it
relevant from a poverty perspective is that it provides a potent mechanism for the
reciprocal liberalizationof trade barriers. Although most major industrializedmarkets
provide preferential access for exports from developing countries, especially least
developednations,trade restrictionsremainprevalentespeciallyfor agriculturalproducts,
appareland other labor-intensiveproducts.
Box3. WTO and PovertyAlleviation
The WTO agreementsinvolvelegal commitmentsby governmentsregardingthe rules and
policiesthey follow in the conduct of internationaltrade in goods and services. As such,
the Agreementscontain no explicit referencesor direct links to policiesaffecting poverty.
Their impact derivesfrom the trade policiesthat governmentshave committedto pursue
based on their WTO commitments.More than 100 of the 140 WTO members are
developingcountries(althoughthe WTO does not define what constitutesa developing
nation-this is largelyleft to self-selection).Almostall the 30+ countriesseekingto accede
are also developingcountriesor economiesin transition.
13
Within the WTO agreements, developing countries are supposed to receive
special and more favorable treatment(usually referredto as "Special and Differential"
Treatment)aimed at addressingtheir particularcircumstancesin internationaltrade. The
poorest of the developingcountries-the 49 countrieson the UN list of least developed
countries (LDCs) are provided with further additional more favorable treatment. The
Specialand DifferentialTreatmentprovisionsinvolvefive sets of measures:(a) developing
countriesare given moreflexibilityin termsof their own trade policies-e.g. they have not
had to reducetheir tariffs significantlyin previousmultilateraltrade negotiationsand they
do not have to 'bind' all their tariffs; (b) their exports can be charged lower preferential
tariff rates in developedor other developingcountry markets; (c) they are given longer
'transition' periods to implement provisionscontained in some WTO Agreements; (d)
developedcountriesare supposedto providethemwith technicalassistanceto help them
implement the commitmentsthey have assumed; and (e) developed countries are to
implement their WTO commitments in a manner 'favorable' to developing country
interests.
WTO provisions in general, and those which apply to developing countries in
particular,tend to be quite permissivein terms of the trade policies an individualcountry
can pursue-and hencedo not tend to constrainthese policiesin ways that would harm
the poor. On the contrary,they are on the whole supportiveof trade and foreign direct
investment policies that would tend to promote poverty alleviation. Moreover, the
availabilityin the WTO of an effectivedisputesettlementmechanismand the guaranteeof
MFN treatmentfor all its memberscan be of specialimportanceto developingcountriesin
their dealingswith more powerfuldevelopedcountrytradingpartners.Thus, it is important
for all developingcountriesto be membersof the WTO.
At the same time, the permissivenessof the WTO Special and Differential
Treatment provisions has been a problem for LDCs. This has enabled developing
countriesto maintain higher levels of domesticprotection,which has harmedtheir own
economies.And, export subsidies--often used to offset the disincentivesof protectionare a drain to the budget, and hence are not affordable,and can not be relied upon to
provide sustainable,laborintensiveexportgrowththat helpsalleviatepoverty.
Source: For more on the WTO, see http://www.wto.org/.Hoekmanand Kostecki (2001)
provide a general introduction; McCulloch,Winters and Cirera (2001) focus on the
poverty-relateddimensionsof manyWTO agreements.
00.
IEIIKO
0=-
.0
M
0"
Trade liberalizationinvolves reducing discriminationagainst foreign suppliers of goods
and services. This is achieved not simply by eliminatingquotas and reducing average
tariffs and dispersionacross tariffs, but also by strengtheningtrade-relatedinstitutions,in
particularcustomsand standardsbodies.In addition,complementarypoliciesare needed
to supporttrade reform.
-
09
0
0
-.
1*
-R
To be sustainedand contributeto an efficientallocationof resources,trade liberalization
must be supported by a stable macroeconomicenvironment and a competitive real
exchangerate. Trade reform works throughthe transmissionof price signals.These are
14
concealed in a regime of high and variable inflation. Thus, macro-stabilityis a key
complementarypolicy.12
A competitivereal exchangerate is also crucialto create conditionsfor continued
support of liberal trade policies. With an overvaluedexchangerate, import competing
industriesare at a competitivedisadvantage,and this generatespolitical pressuresfor
protectionthat are difficult to withstandin the face of rising trade deficits and declining
foreignexchangereserves(ShatzandTarr, 2001providea moregeneraltreatment).
Initially,trade liberalizationis likely to lead to a trade deficit because the rise in
importstends to occur faster than the export supplyresponse.A depreciationof the real
exchangerate will help to restorea balancebetweenexports and importssince it makes
imports more expensiveand exports are more profitablein domesticcurrency. Under a
flexible exchangerate regime,the real exchangerate will adjust throughmarketforces.
Undera fixed exchangerate regime,significanttrade reformshouldbe accompaniedby a
devaluationof the domesticcurrency.The requireddepreciationis largerthe greater the
extent of trade liberalizationand the greater the lags in the supply response (see
Appendix).
Without recommendingany particulartype of exchangerate regime,we note that
some developingcountriesare participantsin a currencyzone, which limitstheir capacity
to devalue-an example is providedby the countriesin West and CentralAfrica who are
membersof the CFA zone. Due primarilyto currencyovervaluation,in the CFA zone
between1986and 1994there was no economicgrowth,whenother Sub-SaharanAfrican
countrieswere growingat 2.5 percentannually(Clement,1994).In fact, for some of the
CFA countries (e.g., Cameroon, Cote d'lvoire) the overvaluationinduced an output
contractionbetween 1986 and 1994comparableto the Great Depressionin the United
States(Shatzand Tarr, 2001).A numberof CFAcountriesalsosufferedlargeincreasesin
poverty (Devarajanand Hinkle, 1994).After the devaluationof 1994, growth strongly
resumed. Some countries seek to maintain a fixed exchange rate through the
establishmentof a currency board or some other arrangement(e.g., Argentina and
Estonia). In such countries, to contain output losses, trade liberalizationmay have to
proceedat a paceconsistentwith the feasiblerate of real exchangeratedepreciation.
4.2 Fiscal Revenueand the Design of Tariff Reform
In 1990, collectedtrade taxes as a percentof GDP averaged0.6 percentamong OECD
13 Policy makers in low-income
countries and 4.4 percent among non-OECDcountries.
countriesmay be concernedthat reformmight substantiallyreducegovernmentrevenue,
yieldinglargerfiscal deficits and inducinginflation.While these concernshavesome merit,
trade reform need not entail diminished revenues.Many countries have implemented
successfultrade reformprogramswithoutsignificantloss of revenue.For example,in the
12 The relativepriceof tradable
to non-tradable
(or home)goods,Pt/Ph,is typicallyreferredto as
the real exchangerate(RER)and is usedas a measureof the competitiveness
of the tradable
sector.Thereasonis thatseveralof thedeterminants
of Ph (suchaswages)affecttheproduction
costsof tradables.Oversimplifying
for purposes
of illustration,
supposea govemment
runsa fiscal
deficitandfinancesit by printingmoney.Theresultinginflationraisespricesin the non-tradable
sector.If tradablepricesdo notincreasebythesameamount,inflationlowersPt/Phandthusthe
realexchangerate,therebyreducing
thecompetitiveness
of thetradablesector.SeetheAnnex.
13 Trade taxes as a percent of GDP were: 5.3 percentin Africa,4.4 percentin Asia, and 3.5 percentin the
MiddleEast.Ebrillet al. (1999).
15
1990s, Ghana, Kenya, Senegal and Malawi implementedtrade reforms without a
significantloss in revenueas a percentof GDP(Ebrillet al., 1999).
Perhapsthe foremostreasonwhy trade reformneednot lead to a loss of revenue
is that developingcountrieshave traditionallyrelied heavilyon quantitativerestrictionsof
imports.Governmentrevenueincreaseswhen quantitativerestrictionsare convertedinto
tariffs. Reductionof tariffs should start with the highestones. Also, where tariff rates are
very high initially,they will generatelittle or no revenue.Reductionsof the tariffs to more
moderatelevels will increaseimports and thus revenues,as the incentiveto smuggle is
reduced.Moreover,exemptionsoften are a significantsourceof revenue loss and their
reductionincreasestariff collections.Finally,an exchangeratedepreciation,whichshould
accompanysignificanttariff reduction,will raise the local currencyvalue of imports and
thus tariff revenue.
Box4. Summaryof RevenueImpactsof TradeLiberalization
ExpectedRevenueImpact
TradeReform
ReplaceNTBswith tariffs
Eliminatetariff exemptions
Eliminatetrade-relatedsubsidies
Reducetariff dispersion
Eliminatestatetradingmonopolies
Positive
Positive
Positive
Ambiguous/Positive
Ambiguous/Positive
Reducehigh averagetariffs
Lowermaximumtariff
Reducemoderateor low averagetariffs
Eliminateexport taxes
Ambiguous
Ambiguous
Negative
Ambiguous/Negative
Source: Shareret al. (1998).
When tariffs rates are already uniform and in the moderateto low range, then
further tariff reduction is likely to result in revenue loss. Only in this latter case does
revenue representa genuine problemfor trade liberalization.Alternate broad-basedand
nondiscriminatoryrevenue sources should be sought, and trade reform sequencedto
coincidewith the availabilityof these alternaterevenuesources. Such altemative,broad
based tax instrumentswill be more efficient (much less distorting)than trade taxes. For
many products,such as alcohol,tobacco and petroleum,collectionof taxes on domestic
productionas well as imports will have very low additionaladministrativecosts, and will
reduce the incentives to develop inefficient import-substitutingfirms. Even very poor
countriessuch as Cambodiahave beenableto introducebroad-basedconsumptiontaxes
that reduce dependence on customs duties, and raise much-needed revenues for
developmentexpenditures.
1111
-M. "I.
.0
Where labor market flexibility is low, reformsto improvethe operationof labor markets
shouldaccompanytrade reformsin orderto enhancelabor mobility.For instance,in Peru
in the 1980s, a trade reform failed to generateany supply responsebecause of severe
labor market rigidities. Labor legislationprohibitedfirms to shed labor, close plants or
even change activities.This led to many bankruptcies,contributedto foreign exchange
and financialcrises and a failureof the reform(Nogues,1991).
16
The poor are often concentratedin the informalsector and reformswhichincrease
labor mobility in the formal sector can have a powerfuleffect on reducingpovertywhen
combinedwith trade liberalizationby openingup additionaljobs in the formal sectorfor
workerspreviouslyin the informalsector.This was the case in Panama,for example,see
World Bank (1999). The mutuallysupportiverelations between trade, macroeconomic,
labor market and other policiesmay then serve to increasethe credibilityand payoffsof
each.
Property rights for land (and water) are also important in coping with needed
adjustmentsor taking advantageof export opportunitiesthat may be beneficialto the
poor. For example,Egyptcould producelabor-intensivecrops (fruits and vegetables)for
export to the EU, as is done in other Mediterraneancountries,and the rural poor could
gain significantly(Barresand Valdes,2000).But becauseof the lack of existenceof land
(and water) markets,Egypthas not taken advantageof this opportunityand continuesto
grow traditionalcrops.
4.4 Competition Policies
In manufacturing,pricespaidby poor consumersare affectedby competitiveconditionsin
the country. It is possibleto increasethe competitivenessof marketsfor manufactured
goods in developingcountries,that is, reducethe markupover the cost of productionby
loweringexternalbarriersto internationalcompetitionas well as by reducinggovernment
imposedbarriersto entryagainstdomestic firms, Both are importantin providinggoodsto
poor consumerswith low monopolymarkups. In small countriesan open trade regimeis
especiallycrucial, since there are fewer domesticfirms contestingthe market. In large
countries, administrativeand legal barriers againstdomestic entry are relatively more
important than in small countries since there are more potential entrants blocked by
domesticbarriersto entry(Hoekman,Kee and Olarreaga,2001).
Box5. RuralPoverty,Cottonand Parastatalsin WestAfrica
In most of FrancophoneAfrica, the nationalparastatalcompanyis the sole authorized
purchaserand processorof the farmers' seed cotton, and the sole seller of the joint
productsof ginning: cotton lint and cottonseed. It also suppliesinputs to the farmers,
owns and operatesall the cottongins, and providesthe transportneededto supplythe
inputsto the farms, the seed cottonfromthe farmsto the gins, the lint to the portsor to
local textile firms, and the seed to local oil mills or to other markets. It is also
responsiblefor the sale of the lint, nearlyall of whichis exported.
Cotton has provedto be an economicallyvery efficientcrop and has made major
contributionsto the developmentof rural areas,exports,povertyalleviationand growth.
The share of cottonlint exports increasedfrom about4 percentof world trade in cotton
in the mid-1970sto about 15 percentin 1997.However,the absenceof any competition
in the purchaseof seed cotton from farmers has impliedthat farmers have been paid
pricesfor their seed cottonthat tend to be far below competitivelevels. In terms of lint
equivalent,seed cotton prices in the FrancophoneAfrican countries have generally
been withina range of 40 to 50 percentof the export price of cottonlint, comparedto
ratios averagingalmost90 percentin Indiaand around80 percentin Zimbabwe.
Detailedcomparativeanalysesof costs show that only a small part of the low
Francophoneseed cotton prices can be explainedby higher transport and operating
costs. More importantare high governmenttaxes (includingspecialtaxes on parastatal
profits and export taxes on cotton lint), and implicittaxes from sale of lint to domestic
textile firms at low prices and periodic costly managementmistakes and corrupt
17
practices. Simulations suggest that removing this taxation would generate large
percentageincreases in seed cotton prices for farmers, varying from 45 percent in
Cameroon to 87 percent in Burkina Faso, expand output, and increase real farm
incomes.Another problemwith parastatalsis that they typically impose pan-territorial
prices wherebythey pay the same price for all locations(and times of the crop year).
These subsidiesfor transportand storageeliminatedany incentivefor privateprovision
of these services, and resulted in production in socially inefficient locations. The
parastatalmonopolyalso preventedthe developmentof privatecredit services.
Allowing private firms to provide the services of the parastataland permitting
farmersto contractwith privatefirms can be expectedto greatly improveefficiencyand
have distributionaladvantagesfor poor farmersin the long run. However,it is important
to pay attentionto supportingservicesand markets(such as credit and transportation)
duringthe transitionto a marketbasedsystem.It is also possiblethat the eliminationof
pan-territorialpricing will have a negativeeffect on poor farmers that are located in
isolated regions.As is frequentlythe case, not all will necessarilygain, pointingto the
needfor a socialsafety net.
Source: Pursell(1998).
An important sector in the context of trade reform is distribution. If there are
barriers to entry into distribution,those who control this sector may be the primary
beneficiariesof trade liberalization,pocketingmuch of what used to be collectedas tariff
revenue and not passing the tariff cut on to consumers. In agriculture, parastatal
marketingboardsoften stronglyrestrict competitionfor the productsof poor farmers and
restrain their incomes (Box 5). Eliminationof the these boards, while paying attention
during the transitionthat key ancillaryservices like transportationand credit that these
parastatals may be providing continue to be supplied, should benefit the poor. But
competitionamongprivatefirms is also important.Thus, exclusivegovernmentlicensesto
the private sector should also be avoided so that poor farmers do not have to pay
excessiveprices for their inputs or receive monopsonisticallydepressedprices for their
outputs. The prescription applies to import monopolies and exclusive distribution
arrangements.
FDI is an important channel of technology transfer across national boundaries.
Multinational corporations account for a large share of the world's research and
development,and more than 80 percentof royaltypaymentsfor technologytransfersflow
from subsidiariesof foreign companiesto their parentfirms (UNCTAD1997). Econometric
evidencetends to supportthe view that developingcountriesreceivingFDI perform better
in terms of productivitythan their counterpartsthat are not FDI recipients.
What mattersfrom a poverty reductionperspectiveis whether and to what extent
FDI has a positiveeffect on the incomes(employment)of the poor and on the prices of
what they consume.As far as employmentis concerned,it is importantfrom a short run
viewpointthat FDI involve labor-intensiveproductionand result in the transfer of skills
throughtraining. Improvementsin communications,
transport and informationtechnology,
togetherwith globaltrade policyreforms,have madeit much moreattractiveto companies
to engage in so-calledoutsourcingand processingtrade, wherethe labor-intensiveparts
of production are located in developingcountries. In the 1970s such FDI focused in
particular on textiles, more recently it has also includedthe electronicsand auto parts
18
sectors. Such investmentcan be an importantcatalyst for the creation of low skilled
employment-as exemplifiedby Mexicoand South-EastAsianeconomies.
Given the importanceof low trade costsfor suchactivity,policymakersshouldavoid
offering trade protectionto foreign investors,since this will attract the 'wrong' type of
investmentfrom an employmentcreationperspectiveby deprivingthe host countryof the
benefitsfrom participationin internationalproductionand distributionnetworks.Protection
may also result in lossesto the hosteconomyby providingrentsto foreigninvestorsat the
expense of domestic consumers.Lall and Streeten (1977:172-174)studied some 90
foreign investments,using a cost-benefitmethodology,and found that more than 33%
reduced nationalincome;this was mainlydue to excessivetariff protectionthat allowed
high cost firms to producefor the local marketat very highprices,even thoughthey could
have been imported much more cheaply. Encarnationand Wells (1986) found that
between25-45% of 50 projects studied(dependingon analyticalassumptions)reduced
nationalincome;againthe mainculpritwas highprotection.
Intellectualpropertyright protectioncan be importantin attractingFDI in sectorsthat
rely extensivelyon patent protection,helpingto tilt the focusof investmentprojectstoward
manufacturingand away from distribution (Smarzynska,2000). Intellectual property
protection is of direct importance to the poor in developing countries, especially for
products and sectors that rely heavily on traditional knowledgeand culture, including
ethno-botanicals,as well as activitiessuch as writingand performingmusic.The absence
of effectiveprotectionof intangibleassetsand intellectualpropertycan havevery adverse
consequencesfor the poorthat are either producers(e.g., handicrafts)or the beneficiaries
of assetsthat have been built up over time (e.g.,traditionaldesigns,plant varieties).The
costs of intellectualproperty rights include the price.increasingeffect of protection.This
can have harmfuleffectson the poor by preventingaccessto drugsand keepingpricesfar
above the cost of production. As discussed at greater length by Maskus (2001),
intellectual property right protectionmust be complementedby trade and competition
policyinstrumentsthat serve to offsetthe marketpowergrantedto right holders.
5 Sector Issues
In many developingcountries the manufacturingsector has been most protectedas a
means of promotingindustrializationand longer term development.A great deal of the
analysis regardingtrade reform in earlier sectionsfocuses implicitly on reforms in that
sector.However,trade reformsin agricultureand servicesprovidesignificantopportunities
for employmentand incomes of the poor as well as reductionsin the costs of their
consumption.
5.1 The importance of agriculture
Becausethe poor in developingcountriesare oftenlocatedin ruralareas and employedin
agriculture,how trade reform affects agriculturewill criticallyaffect its overall impact on
poverty alleviation.Agriculturalimportables(mainlystaples)are typicallyprotected,while
agriculturalexportablesare often subject to export taxes. Agriculture as a whole has
tendedto be taxed indirectlythroughprotectionof the manufacturingsector,overvaluation
of the real exchangerate (Schiffand Valdes, 1992,2000),and the operationof marketing
boards or similar parastatals (see above). Trade reform, which should cover both
agricultureand manufacturing,will thereforetypicallyraiseagriculture'sdomesticterms of
trade and help the rural poor. The same is true for improvedmacroeconomicand fiscal
19
policiesthat result in a real exchangerate that is closer to a sustainableequilibrium.In
other words, the rural poor will generallybenefitfrom trade reform,evenwhen it includes
agriculture.
Some of the rural poor,however,maylose. Foremostamongthese to considerare
those who are employedor producein sectorsthat are highly protected,low-productivity
sectors. Examples include maize in Mexico, wheat in Morocco, and various importcompetingcrops in manydevelopingcountries.If the mobilityof these rural pooris limited,
then a reductionin the tariffs in that sectoris likelyto hurtthat sub-group,especiallyin the
short term, as prices of their output fall. Over time these losses can be minimizedas
farmerschangetheir outputmix and producemore of the crops whosepricesdid not fall.
In a study on the Northeastof Brazil,one of the world's poorerregions,switchingto more
profitable crops after trade liberalizationand devaluationgreatly reduced negative real
income effects for small farmers (Lopez and Romano,2000). Thus, trade liberalization
accompaniedby devaluation(seesub-section4.1) can help dampenthe short-termeffect
on incomesof the rural poor.
The abilityto shift to the productionof new products(possiblyexportableswhose
price has risen) may depend importantlyon necessarycomplementaryreforms.Shifting
crops may require restructuredlandarrangements,additionalcapitalor accessto water. If
marketsfor these factors are poorlydeveloped,farmersmay be unableto take advantage
of new opportunities.Thus, complementaryreformsthat focus on these marketsmay be
necessaryto help poor farmers. Given that poor farmers may be hurt in the short run,
compensationpoliciessuch as improvingrural infrastructure,researchand development,
and creating educational opportunitiesfor children are important. These options will
almost alwaysbe betterthan protection,as only if farmer'sreturns outsideagricultureare
increased will poverty be reduced in the longer term. As far as trade reform itself is
concerned, one option to consider is to phase down tariffs in the vulnerable sector
graduallyfollowing a pre-announcedschedule.This is what was agreed for the maize
sector in the NAFTA negotiationsbetweenthe US and Mexico. Experienceshows that
unlessa scheduleof tariff reductionsis pre-announcedand actuallyimplementedfrom the
start, reforms with long transition periods (typicallyover five years) lack credibilityand
providelobbyistswith time to defeatreform.
Fluctuatingworld prices.The liberalizationof a country'sagriculturaltrade policies
meansnot only adjustingto loweror highergeneralprice levelsprevailingin international
markets,but alsoto the fluctuationsin worldpricesof agriculturalcommodities.In many
countries,the impactof these fluctuationson the domesticeconomyis a major motivation
for interventionsthat de-linkdomesticandinternationalmarkets.Exportand import
controls,governmentcontrolledparastatalsor marketingboards,and variabletariffs all
reducethe size of the internationalmarketin whichthe commodityis freely tradedand
thus increasethe sensitivityof internationalpricesto exogenousshockssuch as bumper
harvestsand crop failures.For example,internationaltrade in rice is equivalentto only 5
percentof world production.Moreimportant,the extentto whichsome of the largest
producingand consumingcountries(e.g.China,India, Indonesia,and Brazil)importor
export is subjectto discretionarygovernmentcontrols.Policiesby these countriesaimed
at stabilizingdomesticpricescan havea very largeimpacton prices in the narrow
internationalmarket,in turn leadingto pressuresto maintainor increaseborder
interventionselsewhere.For this reason,all countrieshave a commoninterestin reducing
the instabilityof world pricesby openingtheir domesticmarketsand removingpolicies
whichde-linkdomesticfrom worldmarkets.
20
Althoughpoliciesto open domesticmarkets,especiallyin large countries-both
14 --havethe potentialto substantiallyreducethe amplitudeof
developedand developing
fluctuationsin worldcommodityprices,these pricesare inherentlyless stablethan the
5 In any case, most
world pricesof manufacturedgoodsand the pricesof services."
individualsmall countriescannoton their own have any perceptibleeffect on the levelor
volatilityof world commodityprices,and needto find the bestway to live with the price
swingsthat they mustby and largetake as a given.Consequently,trade liberalizationin
any individualcountry,if fully implemented,will often(but by no meansalways)mean that
domesticcommodityprices,includingpricesat farm level,will becomemore unstable.16
Farmersin many countrieshave repeatedlyshownthat they can adjustrapidlyand
efficientlyto relativeprice changes.For thereto be an overallwelfareimprovementfrom
shelteringfarmersfrom pricefluctuations,the extraincomeof the groupsthat benefitfrom
interventionsmust outweighthe net economiclossesconsequenton distortedconsumer
pricesand changesin the government'sfiscal position.If all changesin income,including
governmentincome,are weightedequallyin the decision-makingprocess,the costs of
interveningwill alwaysexceedthe benefits.It is generallybetterto targetthe low income
groupsthat are hurt with actionsthat increasethe abilityof farmersand workersto switch
betweencrops,complementedwith welfareand workfareprogramsuntil a reasonable
degree of flexibilityand responsivenessis achieved.
In consideringpricestabilizingpolicies,it importantto recognizethat once such
policiesare introduced,it is very difficultif notimpossibleto stopthe politicallypowerful
from usingthem to shift the averageprice.The commoditymarketingboardsand
parastatalmonopoliesin developingcountriesdiscussedpreviouslyoftenstabilizedprices
at very low levelsrelativeto worldprices,implyingheavytaxationof smallfarmersand
resourcetransfersto governmentsand local processinginterests.This may seriously
retardthe growthof economicallyefficientruralindustriesand slow downthe contribution
that their expansioncouldmaketo the alleviationof rural poverty.For importsubstitutes,
the politicaleconomyforcesgenerallypointin the oppositedirection,and intervention
originallyjustifiedin the nameof price stabilization(e.g. price bandschemesintroducedin
LatinAmericaduring the 1990s)very oftenbecomede facto price supportschemesthat
developa momentumof their own.
OECDintervention. Exportsubsidieson productssuch as meat, dairy products
and grains, depressprices on world markets,and thus also on the domestic marketsof
developingcountries.Some countriessuch as Japan maintainvery high barriersagainst
imports of wheat, rice and other agriculturalcommodities.As mentioned,such policies
if the marketsfor rice,wheat,sugar,milkproducts,sugarandcottonwereopenin
"4 Forexample,
Chinaand India,that wouldgreatlyexpandthe sizeof the worldmarketfor theseproductsand
shocks(TyersandAnderson,1992).
reducetheimpactof variousexogenous
from fragmentedworld markets,commodityprice instabilityis associatedwith
15 Abstracting
etc.
weatherconditions
dueto seasonalproduction,
inelasticdemand,lagsin supplyadjustments
markets(e.g.for sorghumandmaizein inlandareasof Africaand
commodity
16 Pricesin domestic
of high domesticand
from worldmarketsby a combination
SouthAsia) that are disconnected
marketingcostsand/orrestrictivetradepolicies,mayfluctuatemorethanworldpricesbothwithin
and acrossseasons.Linkingthemto world marketsby openingimportscould increaseprice
stability,e.g.,bycuttingoff pricepeaksresultingfromcropfailures.Theuseof measuresto control
mayalsolead
marketconditions
importsandexportsin waysthatdo notadjustflexiblyto domestic
to higherdomesticpriceinstabilitythan in worldmarkets,or worsestill, to costlyand wasteful
A significantportion
marketingorganizations.
of excessstocksheldbygovernment
accumulations
of suchstocksoftenarelostto pestsandweather.
21
contribute to world market price instability, constrain exports and increase import
competition.While most of the poorest developingcountries produce temperatezoneproductssuch as fruits, nuts and vegetablesthat do not competewith subsidizedand
protectedmeat, milk and grains,for some low-incomecountries(e.g., meat productionin
Mali or Burkina Faso), and, of course,for middle incomedevelopingcountriessuch as
Argentina or Brazil, EU export subsidies are a major factor constrainingagricultural
exports.
What is the appropriateresponseto such foreign policies? Insofaras prices are
forced downward, consumerstend to gain. The impact of the policies on producers
dependson whetherfarmerscompetewith subsidizedimports,whetherthe subsidiesvary
significantlyover time, whetherfarmers are on average net buyers or net sellers of the
commoditiesconcemed,the relativeimportanceof subsistencefarming, and the relative
sizes of the rural and urban poor population.If there is no domestic productionof the
agriculturalproductsconcerned,there is no needfor intervention,as subsidizedpricesare
beneficialto consumers.Mattersare more complexin instanceswherethere is domestic
productionof the commoditiesconcerned.It is importantto note that householdsthat are
net buyers of the products will gain from subsidies,while subsistencefarmers will be
unaffected.The urbanpoor will generallygain insofaras the subsidieslowerprices. Thus,
what matters from a policy point of view is whetherthe gains to the non-ruralpoor are
largerthan the losses incurred by rural householdsthat are net sellers. To answerthis
questionrequirescollectinginformationon the numberof householdsthat are engagedin
subsistenceproduction,those that are net buyers,and those that are net sellers.Another
importantfactor concernsthe distributionof the laborforce engagedin productionin farms
that are net sellers.Those that ownthe landmay lose from lowerprices, but the extent of
such losses are bounded(as they may shift to subsistencefarming).While this can imply
a severe reductionin real incomes,landlesslaborersworking for net sellers confrontthe
most severe potential loss in that they may lose their source of incomewithout having
subsistencefarmingto fall back on.
In cases of agriculturalsectorswith significantnumbersof householdsthat are net
sellersof commoditiesthat are subsidizedby OECDcountries,it is sometimesarguedthat
higher tariffs on agriculturalimports are appropriate.This increasesthe domestic price,
thus helping to offset the effect of the subsidy and allowing domestic producers to
compete.This is akin to 'countervailing'the effect of foreign policies. From the viewpoint
of efficiency,whensubsidiesare permanent,such a policyis not desirableas it movesthe
structureof protectionawayfrom uniformity,and distortsproducerand consumerchoices.
Given the long-lastingnature of interventionby high-incomecountries, countriesshould
regardthis as part of the externalenvironment.The implicationis that importingcountry
governments should not impose countervailingtariffs.17 However, with the Uruguay
Round, multilateral disciplines were strengthened, and it is clear that there will be
significant pressureon the EU and other high-incomecountriesto substantiallyreduce
protectionand graduallyeliminateexport subsidiesin the comingdecade.Thus, over time
17 A key issueis whetherthe exportsubsidies
are permanentor transitory.If the subsidiesare
permanent,
it willbe in thecountry'sinterestto ignorethe subsidyfor policypurposesbecausethe
economiccostsof protectionto the economyoutweighthe benefitsto domesticproducers.If the
foreignsubsidyis permanent,thenthe nationalinterestis unaffected
by whetherthe priceis low
becauseof foreigncomparative
advantage
or exportsubsidies.If thesubsidiesaretransitory(and
thereare capitalmarketor otherimperfections),
thereis a casefor temporary
protectionfor those
tariff lines subjectto competitionfrom subsidizedexportsso that adjustmentcosts can be
minimized-see below.
22
the external environment is likely to become less distorted. Insofar as developing
countries currently maintain above averagetariffs on the relevant agriculturalimports,
there is then a potentialcase not to reducethese as muchor as rapidlyas tariffs on other
8 The reason is that liberalizationcould
productsas part of an overalltrade policyreform."
give rise to inefficientreallocationof resourcesif there are adjustmentcosts associated
with the downsizingof agriculturalproductionfollowingliberalizationand the subsequent
expansionas the exportsubsidiesand OECDprotectionare reduced.
In practice, however,OECD liberalizationcan be expectedto take a long time,
witnesseffortsto do so over the past30 years.Moreover,it will be impossibleto precisely
countervailthe impactof foreignpolicieson worldprices;the informationrequirementsfor
such a policyare substantial.Accountshould also be taken of the risk that protectionmay
become permanent even if the foreign protectionand export subsidiescome off, and,
more fundamentally,that trade interventionis a secondbest approachto dealingwith the
problem. The best policy is to push for the phase-outof export subsidies and OECD
protectionof agriculturein the WTO context,and ensurethat any assistanceprovidedto
domestic farmers is designed primarily to target domestic poverty concerns. More
generally,whetheror not foreigncountriesintervenein agriculturalmarketsshould not be
the focus of policy. Instead, the focus should be on determining how large is the
populationthat may lose from trade liberalizationin the short run, that is, net sellers of
protectedcommodities.
While trade policy is not the appropriatelong-terminstrumentthroughwhich to
pursuerural povertyreductionobjectives,it can play a role in three situations.The first is
to deal with temporary import surges that have a significant negative impact on the
livelihood of poor farmers. In such cases a special safeguard mechanism can be
considered, under which temporary protection can be sought if domestic prices of
productsthat are of importanceto the poor in terms of productionand employmentfall
significantlyin a short period of time due to imports.Whetherthese are subsidizeddoes
not matter;what mattersis that there is a seriousdetrimentalimpacton poorfarmers.The
generaldesign of any such safeguardmechanismshould conformto the principlesset out
in section 3.3 above. The second situationis if the overall policy regime discriminates
against agriculture,i.e., policy favors industrialproductionand/or urban consumersof
food.While this can be used as an argumentin favor of higher protectionfor poor farmers
in low-incomecountries,-theappropriatepolicy is insteadloweringprotectionfor industry
and offsetting any prevailingpolicy bias against the rural poor. In contextswhere this
cannotbe achieved,or whereit is pursuedgradually,there may be a casefor maintaining
higher rates of protectionon agriculturalcommoditieson second-bestgrounds.Third, in
situationswhere complementarypolicies(safety nets) are inadequateand a significant
numberof the poor are engagedin productionof commoditiesthat are sold domestically,
agriculturaltrade policy reformshould be gradual,involvinga pre-announcedscheduleof
tariff reductions.
Potentialnegativeimpactsof reformin this sectoron net sellerscan be attenuated
by accompanyingthe trade reform with a devaluationand implementingcomplementary
reformsin marketsfor land, credit and water to enable farmersto take advantageof the
new opportunities,adjustto changedincentivesand benefitfrom the reform.Govemment
has an important role to play in fostering agriculturaland rural development,including
thereis no rationale
18 If agriculture
importbarrierscurrentlyare nothigherthanfor manufactures,
for raisingthemgiventhatadjustment
willalreadyhavetakenplace.
23
through encouraging absorption of new technologies, education and providing
infrastructure.Such complementarypolicies are crucial in that they can be critical in
addressingthe needsof the rural poor. In all casesthere is a needfor carefulanalysisof
the prevailingsituationbefore reformsare pursued.This should focus on identifyingthe
balance betweenrural net buyers, rural net sellers,and the urban poor, as well as the
importanceof subsistencefarming.
A final remark.For the least developedcountries,a key challengeis meetingthe
standards and rule of origin requirementsimposed under preferential arrangements
granting duty-freeaccess.By allowinggoodsproducedin these countriesto receivehigh
prices in major export markets,but goodsto be importedat world prices,they allow the
poorest countriesa double benefit. Preferentialaccess providesa positiveincentiveto
produce these productsfor sale in the highly protectedmarketsof the OECD countries
that grant such access. In effect, the preferentialaccess provides protection to LDC
producers in the developed country markets that offer preferentialaccess. As these
markets are both vastly larger than their own, and offer very inflated prices due to
protection,this mayallowfarmersin LDCsto increasetheir incornesby producingmore of
19
these productsfor export.
IMa
17-;
0S
Services-which include activities such as transport of goods and people, financial
services (banking, insurance) telecommunications,distribution, tourism (hotels and
restaurants),construction,as well as educationand health care, account for a rising
portion of GDP, in even in the lowest-incomecountries.The importanceof an efficient
servicesectorgoes beyondthe contributionof the servicessector itself to the balanceof
payments,because the efficiencyof many service sectors is a key determinantof the
competitivenessof firms. Key sectors that influencethe ability of firms to participatein
world trade are telecommunications,
transportation,financialservicesand other business
servicessuch as accountingand legal services.
The gainsfrom eliminatingbarriersto competitionin the variousbusinessservices
can be very large and fundamentallyeffect the country's comparativeadvantage and
pattern of trade (Markusen,Rutherfordand Tarr, 2000). As nations reduce barriers to
trade in goods, it has become apparent that in many countries, inefficient business
services sectors have become the principal barrier to effective integration in world
markets. Inefficient provision of trade-supportservices acts as a tax on exporters of
merchandisegoods and ultimatelyon growthand poverty reduction.Establishingpolicies
that encouragecompetitiveand efficient services sectors,such as allowing entry where
possibleand encouragingforeign direct investmentshouldtherefore be a major element
of global integrationand povertyreductionstrategies.In particular,projectsthat address
restructuringand reform of service sectors should take into accountthe importance of
establishinga regulatoryenvironmentthat is contestable.
Liberalizationof servicessectorsmeansreductionor eliminationof barriers,where
barriers tend to take the form of prohibitions,quantitativerestrictions and regulations
(UNCTADand World Bank, 1994). For example,such restrictionsmay prohibit foreign
Farmersin middleincomecountriesmayfacedifferentincentives,as theywill suffera decline in demandfor
their productsin the EUbecauseof preferentialaccessprovidedonly to the LDCs.
t9
24
direct investmentin certainsectors,limit the share of ownershipof foreign firms in these
sectors, limit the numberof expatriatesthat can be employed,or restrict the amount of
imports of a particular service. Frequentlythe restrictionsapply to both domestic and
foreignsuppliersand result in publicsectormonopoliesin the provisionof services,e.g. in
air and maritime transport (inclusive port services), telecommunications,or financial
services. In many such cases, eliminationof public sector monopoliesmay need to be
accompaniedwith openingup of servicemarketsto foreigndirect investment(e.g.through
relaxing of provisionsregardingthe right of establishmentof foreign firms) becauseit is
the foreign providersthat may be able to provide significantimprovementsin efficiency
and cost reductions.
There is evidenceof a positiverelationshipbetweenprivate competitiveprovision
of telecommunicationservicesand the availabilityof telephonelines at affordable rates.
This is especiallytrue in countrieswhere initial conditionsare characterizedby a low
teledensityor servicerationing(longwaitinglistsfor obtainingconnections).Simply letting
the marketwork can substantiallyimproveaccessin an environmentwhere serviceshave
beentraditionallyprovidedby inefficientpublicmonopolies-even in the poorestcountries
and amonglow incomeconsumers.
For many intemationallytraded goods, the cost of internationaltransportationis
higher than the applicabletariff on imports.For a small economyconfrontinggiven world
pricesof traded goods, highertransportcosts reduceexport prices and increasesprices
of delivered imports. Hence, exporting industrieswith higher transport costs must pay
lower wagesor accept lower returnson capital.Amjadiand Yeats (1995)find that freight
rates for Sub-SaharanAfrican countriesoften are considerablyhigher than on similar
goods originatingin other countriesand thus have contributedto the region'spoor trade
performanceover the last decades.High transportationcosts are due in part to anticompetitivepolicies,such as cargo reservationschemesmaintainedby a large numberof
African countriesand internationalshippingcartels and legislatedmonopolyprovidersin
both industrializedand developing countries (Francois and Wooton, 2001; Fink and
Mattoo,2001).
Box6. LiberalizingTradein TransportServices
Shipping deregulationin Chile started in 1979 and aimed at dismantlinga system
basedon cargo reservationprotectionfor the nationalfleet. The core of the reformwas
eliminatingrestrictionson foreignshippingcompaniesto supply internationaltransport
services. In addition,monopolyrights in cargo loadingoperationsin Chileanseaports
were eliminated,allowingthe entry of new privateoperatorsand the developmentof a
competitivemarketfor cargo handlingservices.Moreover,the governmenttripledport
capacityby obtaininglaborflexibility.It did this by payingthe unionfor the right to use
workers not under union contractwhich allowedit to move from a union-constrained
eight hour day to 24 hour cargo handling.This led to a significantcost reductionand
an increase in exports, particularly the time-sensitiveagricultural exports that are
intensive in rural labor. Demand increased especiallyfor female labor involved in
picking,sorting,selectingand packingof fruits and vegetables
The new status quo provedbeneficialduring the 1981-82economiccrisis, as
Chile's exportershad "untrammeledaccessto the internationalshippingmarkets"and
were able to supply more price-competitiveproducts.Chileanfruit exportersentered
into long-term contracts with international shipping enterprises to ensure regular
service, guaranteeing rapid transit and certain timing of delivery, essential
requirementson the highlycompetitiveexportfruit market.
25
Denial of access to foreign-ownedtransport companiesto domestic inland
routes can also be important in cutting off a region to internationaltrade. China
provides an example: exportersfrom inner China typically send goods to ports in a
break-bulkform by truck or railway,where the goods are containerizedand shipped.
Multinationalcompanies,however,requireaccessto containertransportfrom door-todoor. This multimodal transport requirement cannot be met given low inland
penetrationof containers,so that most export industrieslocateon the coast and often
import raw materialsand intermediatesfrom abroadrather than sourcefrom mainland
China. Lack of competitionwas a major factor: two state-ownedcompanies(SOEs)
had around 80% of the marketfor freight forwardingand shippingagencies.Although
the governmentallowsmajorforeignshippinglinesto establishfreightforwardingarms
in China, they may only handle parent companies'products.Restrictionson shipping
agenciesare even more sweeping:governmentrules bar foreignor joint venturefrom
enteringthe shippingmarket,limitingit solelyto SOEs.
Source: Bennathan(1989)andWorld Bank (1996).
The case of Zimbabwe'svegetableexports illustratesthe enormouspossibilities
when transport and related services are efficient. In the early 1990s, farmers near the
capital have been supplying fresh vegetablesto the London market by picking them,
immediatelytruckingthem to the airport, andflying them ovemightto Londonwhere they
are on the shelves ready for sale in the morning.This requires cheap and reliable air
transport and moderntelecommunications
becausethe shipmentsare deliveredto order
(Krugman, 1998). In the Maghreb,Amiot and Salama(1996) estimate that eliminating
restrictions on ship registry in Morocco could lead to 25% savings on the maritime
transport costs for clothingand footwear.All Algerianand Moroccanports, as well as the
main ports in Tunisia, are operated by inefficient government-ownedcargo-handling
companies,which charge 30% in excess of the rates that would be quoted by a private
independentoperator. Licenses required by exporters and importersto charter vessels
also create barriersto internationaltrade,since permissionto charteris grantedonly if the
national shipping line withdraws.This is estimatedto cost an additional 20% over the
usualinternationalfreight rates.
Services reformand the poor.Althoughthe poor spend less of their incomeon
services than the non-poor, significantbenefits can accrue to the poor from increased
efficiency of services markets. Services such as transport, education, and access to
communicationsand finance are vital determinantsof the ability of the poor to find
employment and market their production.The incomes of the rural poor are strongly
dependenton marketingand transportationcosts, and on the efficiencyof transportation
networks.Hightransportor marketingcostslowerthe pricesreceivedby poor farmersand
raise the prices of food to poor consumers.Competitionin these sectorsis very important
to poverty reduction, as are resources devoted to "trade facilitation" to improve the
efficiencyof servicenetworksand reducecorruptionand relatedtransactionscosts.
Recent research based on household surveys found that farmers' access to a
public telephoneis positivelyrelatedto the price they receivein district marketsfor their
farm output. Decreasingthe distanceto a telephoneby 10 percentwould lead to a 1.6
percent increase in local prices (Larson, 2000). In Bangladesh villages, women
entrepreneursprovide pay phone services at a profit, using mobile cellular technology.
Even though rural villagers cannot afford a phone individually,they can afford one
collectively(Lawson and Meyenn,2000).A key aspect of servicesliberalizationis that it
26
often involves the movementof factors of productionbecausethe services concerned
cannotbe traded.Giventhe structureof factor pricesin poor countries,an inflowof capital
through FDI would tend to be to the advantage of the unskilled poor-increasing
employmentopportunitiesand wages.
Liberalizationof servicesand the resultantcompetitionare likely to lead to lower
prices, greater availabilityand improvedquality of services. In so far as the poor are
consumersof theseservices,they are likelyto benefit.Butthere is a twist. Frequently,the
prices pre-liberalization
are not determinedby the marketbut set administratively,and are
kept artificiallylowfor low incomeend-users.Thus, rural borrowersmay pay lowerinterest
rates than urbanborrowers,and prices of localtelephonecalls and public transport may
be keptlowerthan the cost of provision.Thisstructureof pricesis often sustainedthrough
cross-subsidizationwithin public monopoliesor through governmentfinancial support.
New entrants may focus on the most profitable market segments("cream-skimming"),
such as urban areas, where the cost or service provision may be lower and incomes
higher. Privatizationcould mean the end of governmentsupport.The result is that even
though the sector becomesmore efficientand averagepricesdecline, the pricesfor low
20
incomehouseholdsmayactuallyincreaseand/oravailabilitydecline.
Universal service or access goals are not contradictorywith liberalizationof
service markets.The handicapof providingservicesto low income householdscan in
principlealso be imposedon new entrantsin a non-discriminatory
way. Thus, universal
service obligations can be part of the license conditionsfor new entrants into fixed
network telephony and transport. But recourseto fiscal instrumentshas proved more
successful than direct regulation-for example, through universal service funds or
subsidiesfor providingservicesin ruralareas.Anothereffectivemechanismis to fund the
consumerratherthan the providerthroughvouchers,as has been the case for education
and energyservicesin a numberof countries(Mattoo,2001).
Assessingservices policy and performance.A careful evaluationof services
trade policy requires analysis of the conditionsof competition in a particular sector,
notably restrictionson entry; ownershiplimitations,private and foreign; and regulation,
especially elements designed to achieve pro-poor outcomes in competitivemarkets.
Relevantquestionsfor policy-makersinclude:Howmuch greaterwould the benefitsbe if
privatization were accompanied by the introductionof competition?Are there good
reasonsto limit entry by policy?What institutionalfeaturespromotethe effectivenessof a
sector regulator?What shouldthe regulatorregulate?What are the costs and benefitsof
restrictionson foreignownership?
In many countries, an assessment of policy and performance in services is
frustrated by a dearth of data. Main performanceindicators include price and quality
variables and measures of access and availabilityof services to the poor. Detailed
templatesfor an assessmentof policyand performancein servicesin a particularcountry
2" These
have recentlybeendevelopedby the World Bank for three key servicessectors.
can be usedto benchmarkcountriesagainstinternationalexperience.
20
In the caseof agriculturalexportables,
increaseddomestictransportefficiencyto the portwill
typicallyraise inlandfarmgate prices.This could worsenthe welfare of low incomeconsumersof
that productin inlandareas. .
air and maritimetransport,andfinancialservicesand are availableat
21 Theycovertelecommunications,
http:./wwwl
.worldbank.org/wbiep/trade/services.html.
27
*
-aIn
0O-reSI
One of the most importantcomplementarypoliciesfor the pooris an efficientsocialsafety
net. It is highly recommendedthat a programto establisha social safety net be in place
independentof the needs relatedto trade liberalization.The best outcomesfor the poor
can be expectedwhen as a result of the overall reformprocess,of whichtrade is a part,
growthacceleratesin the economyas a whole.Especiallyin the short run, however,there
are boundto be some effectson somegroupsof poor who maybe incapableof sustaining
even short periods with adverse adjustmentcosts. One needs to be especiallycareful
regarding the effects of any reform on the poor, as they are least able to bear risks
becausethey do not havethe savings.
An important issue that confronts all trade policy reforms then relates to the
hardshipfaced by poor workersin import competingactivitieswho lose incomefollowing
import liberalization.The policychoicesbroadlydefinedare: employgeneralsocialsafety
nets; establish safety nets targetedto those who are harmedby the trade reform; and
selectivelimitationof the reformsor interventionin marketsfor the purposeof limitingthe
impactof marketreformson the poor.
One type of market interventionis based on the view that opening up to world
marketswill increase risk becauseworld marketsare typicallyunstable.However,poor
countriesare also subjectto largedomesticshocksand it is unclearwhetheropennessto
trade increasesrisk. Governmentattemptsto reducethe risk of trade opennessthrough
marketingboardsand similarinstitutionsaimedto cushionthe impact of internationalprice
fluctuations on producers,especiallythe poor, have often been counterproductive.As
discussedabove, many of these agencieshave imposedsignificanttaxes on the poor.
Experiencewith these institutionsrevealsthe risksfor the poor of efforts by governments
to limit the scope of market reforms,even when these limitationsare intendedfor their
benefit. A fundamental problem in using government interventionsthat limit market
reforms ostensiblyfor the benefit of the poor is that these interventionsare subject to
political lobbying.The poor typicallylack political power, so that political interventionin
market processeswill typically result in outcomesthat are even worse for the poor (see
above). A variety of efforts are underwayin many countriesto replace parastatalsand
2 2 In low-incomecountries such
similar bodies with more efficient, privatesector entities.
alternativesmay not be a viable option,however.As notedearlier,complementaryactions
may be called,such as improvingand reducingthe cost of educationto poor households.
Specializedsafety nets linked to trade reform have a spotty history. In practice
difficult to distinguishworkers who are harmeddue to trade reformfrom those who are
harmed due to normal turnoveror displacementin an economy.The United States has
been providingtrade adjustmentassistance(TAA) to workers displacedby international
trade since 1962 (not linked to poverty). The US program provides both monetary
compensation(called Trade ReadjustmentAllowances,TRA) and retraining.In the early
years of the program,it was found that incomesupportwas typicallyprovidedto workers
who were not permanentlyseparatedfrom their employers,i.e., the programwas not well
targeted (Corsonand Nicholson,1981).However,changesin the design and monitoring
For example,the InternationalTask Force on CommodityRisk Managementseeks to establish marketbased price insurance schemesthat would reduce the price risk farmers (especiallythe poor) have to
shoulderin exportingcashcrops.
22
28
of the programin 1982and 1988have resultedin bettertargeting(Deckerand Corson,
1995).
The resultsof retrainingprogramsappearto be mixed.When retrainingis required,
as in the U.S., it may be ineffective. More generally,the effectivenessof retraining
programstends to increase if they are demand driven, so, for example, subsidized
apprenticeshipsin the privatesectormay work better than governmentprovidedtraining
23
programs.
An alternateapproachto requiringretrainingis to requirethe participationin a
job search program.This appearsto increasethe likelihoodof employmentand reduce
unemploymentbenefits among recipients (Johnsonand Klepinger, 1991; Decker and
Corson,1995).
Fundamentally,it is difficult to morallyjustify safety net programsto poor people
who are harmeddue to trade reformand denyassistanceto other poor peoplewho suffer
equivalentharm from fluctuationssuch as technologicaldisplacementor price changes
due to domesticdemandshifts. Consequently,it is bestto employgeneral,country-wide,
safety nets to deal with problemslinked to trade reform, rather than establish special
safety net programsfor trade relatedproblems.As the main need for the poor during a
difficult transitionperiod is likelyto be food, one approachis a time limitedfood subsidy
and distributionprogram.Targetinga food subsidy,however,is difficult and where it has
been done, it has been subjectto abuse.24 An alternativeis an untargetedsubsidyon
inferiorgoods, as has been pursuedin Egypt(Adams,2000), althoughEgypt's program
has beenpermanent,not temporary.
Direct income support tends to be the most efficient type of social safety net,
providedit can be administrativelyarranged.A problemis that it is very hard to identify
who actuallyneeds the money and even harderto get it to all those who need it. One
approach,whichwas employedsuccessfullyin Jordan,is to providea moneypaymentto
all households initially. The programwas subsequentlynarrowedto middle and lowincomefamiliesand finally,to only low-incomefamilies.Becausedistinguishingthe poor
from the non-poormay be difficult,workfareprogramsmay be more generallyapplicable,
and have been proven effective under certain circumstances(Ravallion, 1999), as
individualscan self-identifyfor these programs.The chapterin this Sourcebookon social
protectionprovidesmore guidanceon safetynets.
In general, as trade reforms are undertaken,the groups of poor which may be
adverselyaffectedneedto be identifiedas accuratelyas possibleand provisionsshould
be madefor their enrollmentin whateversafetynet programsare availablefor as long as
necessary.Clearlysafetynetsare neededto supportthe poorduring a periodof transition
which may vary in durationand severitydependingon their age, skills, mobilityand other
similarfactors. If there are no general safety netsavailable,they should be installed,not
because trade reform demands them, but because they should be an essential
componentof a sustainablepoverty reductionstrategy.A practicalproblemis that some
poor countriesmay not be able to afford a full-fledgedsafetynet. For the design of trade
policyreformthis strengthensthe need both for up front analysis of wherethe poor are
locatedin termsof production(income)and consumption,assessingwhichgroupsmay be
O'Leary(1995)discussesthe measurementof the effectivenessof labormarket programsin Hungaryand
Poland.
24 Subsidiesto consumptionof essentialcommoditieswill often result in a transferof income to the middle
and upperclassessincethewealthybuy morein absoluteamountsof essentialcommodities.
23
29
seriously detrimentallyaffected,and determiningwhat types of complementaryreforms
would best offsetthese potentiallosses.
1-|
*I
-
Bi
-
l
-
*1
-
l
As discussed above and in the Appendix, in tracing through the impact of trade
liberalizationon the poor, it is importantto considerboth the pattern of expendituresand
sourcesof incomesof the poor. However,what mattersmost in the shortrun is the impact
on incomes.The effectof trade reformon povertyin the longerterm hingesvery much on
the growth process-which in turn will dependon a varietyof complementarypoliciesand
institutions.In low incomecountries,the key complementarypolicies/institutions
that need
to be analyzedfall into the following major areas: (a) macro-economic,and especially
exchangerate policy; (b) the operationof the marketfor labor, since the poor are often
concentratedin the informalsector;(c) the operationof the marketsfor agriculture-which
is both a major source of income and accountsfor a large portion of the household
expendituresof the poor; (d) access of the poor to trade relatedservices-for example,
credit,marketing,transportation;and (e) accessto safety nets. There are of course other
issues, such as governance,which are importanthere as well as in other reformefforts.
Tools are being developedby the World Bankthat can assist policymakersin identifying
26 What follows provides a "checklist" of
the impact of trade liberalizationon the poor.
questionsand issuesthat shouldbe consideredin the designand pursuitof trade reform.
The basicelementsof a goodtrade policyregimeinvolvepredictability,transparency,and
uniformity.The following bullets describe in a nutshell what constitutes a liberal trade
policy regime. These provide a benchmarkagainstwhich to judge the prevailingtrade
regimeand provideguidancefor the directionof reforms.
* No licensing,or other approvals,exceptfor health,safetyand environmentalreasons,
and automaticlicensingusedfor statisticalpurposes;no otherQRs.
* Low and uniformtariffs. If the tariff is not uniform,it should have little dispersion,with
onlya smallnumberof bands.A few sectorswith very high tariffs shouldbe especially
avoided.
* If tariffs are importantfor revenuegeneration,uniformityimpliesthat the overall level
of the tariff should be such as to generatethe revenue required. However,some
products-such as alcohol and tobaccoproducts-may be subjectedto high dutiesto
raise revenue, as long as equivalent excise taxes are imposed on domestic
production.
* An efficient customs clearance process with little red tape, that ensurestariff-free
accessto intermediateimportsfor exporters.
* Only one instrument of contingent protection-a safeguard provision. No antidumping.
A moredetailedpresentationof theseguidelinesmaybe foundin Michalopoulos(2001).
For a synthesisof availablemethodssee the proceedingsfrom a October2000 conference"Povertyand
the IntemationalEconomy",availableat: http:I/wwwl.worldbank.orglwbiep/trade/povertyconf.html.
Tools will
be postedon this site as they becomeavailable.
25
26
30
*
Contestableservice markets-measures to ensure competitionprevails,there is no
discriminationagainst foreign suppliers that seek to establish a presence in the
market,andthat appropriateregulationis in place.
7.3 Is Trade Reform Needed?
A number of questionsare relevantin determiningwhether trade reform should be a
priorityand whetherthis will benefitthe majorityof the poor:
*
*
i
*
*
*
What is the impactof status quo trade policieson the poor? It is importantto
determinethe effect on the poor of the existing pattern of protection/subsidization.
Such effects may be positiveor negative,and affectparticularproductsconsumedby
the poor or the incomesof a significantnumberof the poor, whethercountry-wideor in
a particular region. Taxes or supports for important food staples or inputs to
agriculture,in particular,should be identifiedand their incidenceexamined.In those
cases wherethe structureof protectionis not beneficialto the poor there is a prima
facie case for reform.In those caseswheresome of the poor benefit,an assessment
should be made of the relativemagnitudeof the potential lossesand the economywide gainsfrom reform.
Are there non-tariff barriersfor reasons other than for health, safety and the
environment?To the extentthat significantnon- tariff barriersare present,there is
again a primafacie case of a needfor reform-starting with conversionto tariffs.This
reform is likely to benefit the poor more than the non-poorsince license recipients
typicallycollectrents, are unlikelyto be poor (almostby definition)andthe competition
for licenseswastesresourcesthat can be usedproductively.
What is the average tariff and how dispersedis it? The more dispersion,the
greaterthe differencein treatmentof differentsectorsand segmentsof societyis likely
to be, and the greater the urgency for reform. Dispersion often generated by
exemptionsand tariff escalationwill lead to high effectiverates of protectionand is
likelyto entailsignificantinefficiencies.
Is there discriminationagainst agriculture? The overall policy stance affecting
agricultureshouldbe determined,startingwith an assessmentof the 'effective'rate of
protectionfor this sector comparedto manufacturing-see Schiff and Valdes (1992)
for a descriptionof a methodologyto do this. The importanceof agriculturelies in the
fact that the rural poor are likelyto accountfor a largeshareof the poor.
How well do critical service markets function? Do the poor have access to
importantancillaryservicessuch astransport?Do policiesdiscriminateagainstforeign
suppliersand leadto highcost, low qualitydomesticsupply?Is entrypossiblein laborintensivesectorssuch as tourism or back-officeservices?Doescompetitionprevail in
key 'backbone'sectorssuchas transport,financeand communications?
Is appropriate
pro-competitiveregulationin place?
How efficientis customs?How long does it take to clear a containeror air freight
shipment?How doesthis compareto neighboringcountriesandto best practice?How
large are 'unofficial'trade facilitationpayments?Is there a functioningdrawbackand
temporaryadmissionmechanism?
7.4 Getting There
The overall analysis of the trade regime should yield a preliminaryjudgment on the
desirabilityof trade reform.Analysisof both the impactof the statusquo policyset and the
likelyeffect of alternativereformson the pooris important.The tools to undertakesuch an
31
analysiscan be constructedfor mosteconomies;the basicrequirementsincludedetailed
data on imports and exports,the trade barriersthat apply to those goods, household
survey information on the consumptionpattem of the poor and the sources of their
income,and data on the basic structureof the economy(ideallyan input-outputtable or
social accountingmatrix). The basic elementsof a frameworkfor the analysis that is
requiredis laid out in the Appendix-practicaltools to undertakesuch analysesare being
as they
developedby the World Bank and will be posed on www.worldbank.ora/trade
27
becomeavailable.
Thisjudgmentshouldbe then reviewedin the lightof the potentialshort-termeffects of
trade reformon the poor. If there are possiblenegativeeffects,it is importantto identify
the relevant products and sectors early on, in order to help design arrangementsfor
dealing with adverse impacts of the reform and develop strategies for developing
consensusin their support.
Strong government commitmentto the reform is critical. The govemmentshould
attemptto explainthe desirabilityfor reformand obtainthe supportfrom some parts of
civil society. Obtaining a broad consensusmay be a difficult task. The benefits from
reform are likely to be dispersed,uncertainand spreadover time; whereas the private
costs of sectorsthat will be facing increasedcompetitionfrom imports will be obvious,
nearterm and likelyto be concentratedin powerfulpoliticalgroups.
It might appeartemptingto design a pro-poortrade reformby identifyingsectorsthat
are importantto the poor-either on the consumptionside or the income side-and
singlingout these sectorsfor differentiatedcuts in protection.If, for instance,many poor
peopleproducemaize,as in Mexico,it mightseem sensibleto excludethis productfrom a
tariff reduction.There are at least two problemswith this approach.One is fundamental
and the other relatesto politicaleconomy. Thefundamentalproblemis that trade policyis
a single instrument,and a fundamentalprincipleof economicpolicyformulationis that a
single instrumentcan not be expectedto addressmultipletargets.The politicaleconomy
problem is that, once a highly differentiatedtrade regime is adopted, it is essentially
impossibleto stop special interests building a case that their sector deserves special
treatmentfor one reasonor another.Returningto the exampleof maize, if we decideto
maintainor raise protection,we are likelyto find that there is anotherimportantgroup of
poorpeoplefor whom maizeis an importantexpenditureitem.
A better approachis to focus on developingtwo different sets of instruments-one,
trade policy, focused on providingthe incentivesappropriatefor efficient productionand
use of goods and services, and another, distributionalpolicy, focused on alleviating
poverty. With this assignmentof instruments,trade policy can be designed using the
simple, comprehensible,guidelines for trade policy formulation.A set of distributional
instruments will necessarily have a much wider range of dimensions, including
investmentsin expandingaccessto education,the provisionof safetynets, and a rangeof
infrastructureinvestmentsneededto allow peoplein poorerregionsaccessto the markets
and other amenitiesenjoyedby relativelyadvantagedpeople.
27 On this websitethe reader may find the initial efforts whichincludeHarrison,Rutherfordand Tarr (2001)
andother papersfromthe Stockholmconferenceon Povertyandthe InternationalEconomy.
32
7.5 Complementary Policies
Trade policyreformand institutionalstrengtheningmust be implementedin the contextof
a varietyof complementarypolicies.Someof these are generaland some are focusedon
makingthe trade policyreformmore likelyto benefitthe poor.
Macro-economicstabilizationand a competitiveexchange rate are essential to
support greater integrationinto the world markets. Exchange rate depreciationin the
beginningof major trade reform programswill facilitateadjustment.Great care must be
taken to avoid real exchangerate appreciationat a time of import liberalization.As this is
largely determinedby macroeconomicand fiscal policies, it is importantthat these are
managedto be consistentwith the trade reform.
Markets.If marketsare not competitiveor are missing,trade reformsmay not benefit
the poor. Critical obstacles to the operation of market signals must be identified.
Questionsthat shouldbe posedinclude:
* In agriculture,are prices passed on to farmers,or are there governmentor private
intermediarieswhich makelargeprofitsin the sale of farm productsor farm inputs?
* More generally,is the reformlikely to destroyexistingmarketsthat are significantfor
the poor? Will it allow poor consumerto obtainnew goods? (For further discussion,
seeWinters,2000,2001).
* Are there seriousimpedimentsof variouskinds(legal,cultural,transportation)to labor
mobility? Labor market restrictions,such as prohibitionson firing of workers, often
result in an informal labor market sector,with the poor concentratedin the informal
sector.Reductionin the restraintsin the labormarket,especiallycombinedwith trade
reform,can result in an expansionof the formal sector.This can have a strongimpact
on poverty reductionsince the poor will move from the informalto the formal sector
with the expansionof demandin the formalsector.
- Are there serious financing obstaclesto participationin trade? In the long run,
developingan effective financialsystem is a key to development.In the short run,
trade-focusedinstrumentssuch as the use of back-to-backlettersof credit may allow
some of the most pressingobstaclesto tradeto be alleviated.
- Are there serious obstaclesin setting up a business?Competitionmay be impeded
because it is difficult to obtain a license to start a new business or to make an
investment.Or foreigninvestmentmaybe impeded.The reductionin barriersto entry,
especially those imposed by governmentsat various levels, can be expected to
improvecompetitionand allow entrepreneursto sell new cheaperimportsto the poor
or to provideservices necessaryto bring goods to market. The latter is particularly
importantin cases where reformaffects entitiesthat provideancillaryservicesto the
poor and may ceaseto do so after reformoccurs.An examplewould be the provision
of transport,storageand distributionservicesto farmersby a marketingboard.If there
are barriersto entry into such service activities,or entry is unlikelyto occur because
the market cannot sustain operations,continuedgovernmentinvolvementmay be
necessary.
* Are there transportobstaclesto trade? Hightransportcosts make it difficultto engage
in trade. Landlockedeconomiesthat are far away from marketsmay have little that
they can influence,both becauseof the absenceof direct links and the difficultiesof
establishingtransit arrangementsbut sometimesthere are governmentpolicies or
institutionsthat can be modifiedwhich would lowertransportcosts. Here again a key
test shouldbe whetherentryis feasible.
33
*
*
In those instanceswhere analysissuggeststhat the marketwill not supplythe needs
of the poor or those locatedin outlying regions,there may be a need for universal
service regulation. Experience suggests that explicit subsidies to achieve such
objectivescan be efficient.
Are there serioustrade obstaclesin enteringmajorexternalmarketsor in competition
from abroad? Most low income developingcountriesface relatively low traditional
trade barriers in external markets,given a variety of preferenceschemes; but there
are specificproblemsrelatedto sanitaryand phytosanitarycontrolsin many markets,
the threat of contingentprotectionand competitionfrom subsidizedexports in third
markets.
Trade Related Institutions.Success of trade policy reforms involves a variety of
institutionsboth public and private.On the governmentside, an effectiveand non-corrupt
Customs Authority is critical to the success of reforms. Other institutions to which
particularattentionneedsto be paid in orderto ensurethat trade reformsbenefitthe poor
include marketingand export finance.Both are necessaryfor export expansion.In order
for the poor to benefit,it maybe usefulto establishorganizationssuch as a co-operatives
which can put together large enough shipmentsfrom individualproducers to supply
foreign markets; and to be able to obtain financing linked to their exports-which
individualpoor farmerscan not. There is a lot of internationalexperienceon these issues,
including through the InternationalTrade Centre and UNCTAD in Geneva. Bilateral
donorsand multilateraldevelopmentbanks can provide assistancein the design of such
programs,as well as financing.
Agricultureand the Rural Poor. Key issues where complementaryactions may be
required for poor farmers include the availability and cost of education for children;
researchand development;and infrastructure,especiallytransport and communications.
Unless the opportunitycosts of farm residentscan be raised, they will stay trapped in
poverty.
11
1"*
.
-"11
Existing interest groups, often entrenchedelites benefitingfrom the status quo, may
opposereform.And if it is agreed,there will be pressuresto postponeadjustmentas long
as possible. Recognizingthat this is the likely environmentin which trade reforms are
usually proposed, it is nonetheless,extremelyimportant to investigate in advance the
impactof reformson specificgroupsof the poor and to design programsto addressthem.
In such a case, the timing of the implementationof trade reformsneeds to be closely
linkedto the establishmentof the programsthat dealwith their impacton the poor. Some
pointsworth notingaboutsequencingare the following:
* If a reformis preannouncedto be implementedover a few years and it is a credible
reform, then normal market adjustmentand attrition can be used to eliminate or
greatly reduceadjustmentcosts. However,this may come at the cost of the threat of
reversalof the trade reforms,as entrenchedinterestswill be grantedtimeto mobilize
opposition.A stagedreformthat is scheduledto take morethan five years is not likely
to be credible unless it is anchoredin WTO commitmentsor a far-reachingregional
trade agreement.
* It is important, as noted earlier, to address non-tariff barriersand high tariff peaks
earlier ratherthan later.
* It is also importantto reducetariffs across the boardduring each stage of a gradual
reform.If insteada target is set basedon the tariff average,the tendencywill be to cut
34
*
*
tariffs only where they cause no immediatedifficultyand leaveall the adjustmentto
last.
Broadtrade reformsfrequentlymeet with much less politicalresistancethan cuts in
protectionto individualsectors.Broadreformshelpthe winnersfrom reformrecognize
their potential gains, and tend to reduce the costs even for industries that lose
protectionon their output.
Waiting for some importantinfrastructureproject to be completed,such as a port
facilityor a road, is not usuallya goodreasonto delayreforms.
Summingup, not everythingwill be perfectat the start,but some minimum,especially
macro-economicstabilityand a competitiveexchangerate, should be in place.The best
outcomesfor the poor can be expectedwhen,as a resultof the overallreformprocess,of
which trade reformis a part, growthacceleratesin the economyas a whole.That said, it
must be recognizedthat the poor are least able to bear risks, and that In the short run
there will be lossesfor some groups.Analysisof the status quo and the likely impact of
reform on the poor is thereforevery important.Generalsafety nets may not exist or be
inadequate in many low-incomecountries. In such situations reforms should not be
postponed, but rather should be implementedgradually, following a pre-announced
schedule, and complementedby actions to minimize adverse consequencesto the
poorestin society.In many casesthis can be doneby directlytargetingtrade policiesthat
are currentlyclearlydetrimentalto the interestsof the poor, and ensuringthat the reform
processalso considersthe needfor actionin ancillaryareassuchas servicemarkets.
35
*
MA.g[
, M.
0r*
-
A-
-
S
In order to providea schematicoverviewof the variouspossibleeffects of a trade policy
reform on the poor, it is useful to distinguishbetween three types of sectors: those
producing importable or import-substitutegoods (M), exportablegoods (X) and nontradableor home goods (H), as well as two factorsof production,labor and capital.We
assumehere that the only assetof the poor consistsof labor,while the asset owned by
the non-poor is capital. The effects of trade policy reform on the poor depend on the
consumptionand productionof the poor in these three sectors.The effects also differ in
the short and long run. In the short run we assumethat the factors of productionare
immobile,while they are mobilein the longrun.
^jlA
uj
--
-
We assumethat the countryconcernedis small--ithas no powerto affect worldprices of
tradedgoods-and that labormarketsfunctionwell in the sensethat that nominaland real
wagesare flexible.Domesticpricesof M (Pm) andX (Px) dependon their worldprice and
on policy variablessuch as the exchangerate and importtariffs. On the other hand, the
price of H (Ph) is determinedfundamentallyby domesticsupplyand demand.Allocationof
resourcesdependson these three prices.In the long run, resourceallocationdependson
28
relativepricesonly, such as Px/Pmand PxIPh.
Trade liberalization(a reductionin tariffs) raises Px/Pm,and laborand capitalhave
an incentiveto move from M to X. Whether Pm falls or Px rises makes an enormous
differencein the short run and is likely to determinethe successof the reform.This is
wherecomplementarypoliciesplaya crucialrole, includingexchangeratepolicy.
Suppose the nominal exchangerate (ER) remains unchangedfollowing a tariff
reduction.Then Pm falls while Px remainsunchanged,and labor and capital in sector M
are hurt in the short run. The groupsthat are hurt are likelyto lobbyfor a policyreversal.
Also, though in the long run both imports and exports increasewith a tariff reduction,
importstend to increasefaster thanexports,with a likelydeficitin the balanceof trade that
may be unsustainable.Boththe pressurefrom short-termlosers and the balanceof trade
problemmay result in a failure of the reform.This outcomecan be avoidedor its effects
are mitigated by depreciation of the domestic currency. This raises the price of
importablesrelative to non-tradables,and helps dampen both the increase in import
demand and the decline of labor and capital's nominalincomein sectorM. On the other
29
hand,laborand capitalin sectorX benefitfrom the devaluationsince Px increases.
Thus, a policypackageof tariff reductionand currencydepreciationshould make it
easier for the factorsof productionin sectorM in the short run and during the transition
period, and should dampen the resistanceto the reform. In countrieswith a flexible or
floating exchangerate policy, the lower tariff will raise the demand for imports and for
foreignexchange.This will raise the priceof foreign exchangeor lowerthe value of the
domestic currency. In other words, the exchange rate will depreciate (more units of
Withthreenominalprices,thereareonlytwoindependent
relativeprices.Forinstance,
choosing
Px/PmandPx/Ph,thethirdrelativeprice(Pm/Ph)
is obtained
bydividingPxlPhbyPxlPm.
29 A devaluation
has no impacton the relativepricePx/Pmbecausebothpricesincreasein the
sameproportion.
28
domesticcurrencyper unit of foreigncurrency).Thisis similarto a devaluationexceptthat
it is determinedby the marketand not by the monetaryauthorities.30
The effect of trade reformon the poor also dependson the second relative price
PxlPh.That relativepricedependsnot only on policybut also on consumerreactionto the
policy since it is determinedby supply and demand. PxlPh also rises following a tariff
reduction,thoughless than Px/Pm.
When the value of the nominalexchangerate cannotbe changed,a tariff reduction
has no impacton Px but lowersPm. This leadsto a shift in consumptionfrom H and X to
M, and thus to a reductionin Ph (thoughless than the reductionin Pm). This implies an
increase in PxIPh. With a full devaluationequivalentto the tariff reduction,Pm remains
unchangedand Px rises by the magnitudeof the depreciation,shiftingconsumptionfrom
X to M and H, raising Ph. Px/Ph rises by the exact same amount as in the absence of
devaluation. Finally, with flexible exchange rates, the depreciationis less than the
reductionin the tariff, so Pm falls, while Px rises.Consumptionshiftsfrom X to H and M,
and from H to M, so the net effect on the demandfor H is ambiguous,as is the effect on
Ph. Note,however,that Px/Phrisesexactlyas in the othertwo cases.
A.2 Effects on real income in the short run
The impact of trade reform on the poor in the short run will critically depend on their
location in terms of consumptionand production(income),in particularwhetherthey are
employed in tradable or nontradableactivities. There are three cases to consider that
indicatethetypes of effectsthat may arise:
i) Poor employedin the exportablesector.The relativeprice of sectorX increases.
Thus, in the short run, as factors are not mobile acrosssectors, the wage rate of labor
employedin X increases.On the consumptionside, labor (andthe poor, by assumption)
would gain as long as they consumeeither someM or some H or both (sincetheir prices
fall). Thus, labor's real income must improve; and the higher the proportionthe poor
spend on H and M, the largerthe gains.Thus,the real incomeof labor in X must rise, or
remainunchangedin the unlikelycircumstancethat the poor spendtheir entire incomeon
the exportableX.
ii) Poor employedin the importablesector.If, on the other hand,the poor produce
in the importablesector, a tariff reductionwould leadto a declinein the wage of the poor
(labor)employedin the importablesector.Howmuch they would lose then would depend
on the consumptioneffect:if they spendall their incomeon importables,the incomeand
consumptioneffectswould cancelout andthe net effect of trade liberalizationon their real
income is zero. However,if they also consumeX and H, they will lose. The expected
result is that the poor lose in the shortrun, but their loss is smallerthan the declinein their
wages, becauseof the gainsfrom the effect of trade liberalizationon the pricesof things
they consume.
iii) Poor produceonly in the non-tradablesector.With the declinein the price of H,
the wage rate in that sector also declinesby about the same percentage.On the other
Thissectionemploysa stylizedandsimpleframework
thatabstractsfromthecomplexities
which,
in practice,are often important.For example,the existenceof unemployment,
the presenceof
intermediates
products
etc.
30
37
hand, labor in H also benefitsfrom the lowercost of consumingM and H. It is possible
that the impact on the real incomeof the poorrises becausethe cost of the consumption
bundlefalls more than their wages. In general,the impacton the real incomeof laborin H
is ambiguousand dependson the sharesof M, X and H in the consumptionbasket,and
on the responseof the price of H to trade liberalization.The largerthe share of M in the
consumptionbasketof the poor, the greaterthe likelihoodthat they will gain. They must
gain if they only consume M, they must lose if they only consume X, and they are
unaffectedif theyonly consumeH.
These results are summarized in the matrix below. Each cell in the matrix
representsthe "location"of the poor in termsof productionand consumption.The first sign
representsthe effect of trade liberalizationon the incomeof the poor, i.e. the return to
their assets (labor).The second sign representsthe effect on their real income due to
changesin the cost of their consumptionbasket.Thus a "+" after the "I' sign meansthat
the cost of their consumptionbasket has fallenfollowingtrade liberalization.The sign in
parenthesisgives the net effect of changes in their nominal income and cost of their
consumptionbasketson their real-incomein different"locations".To summarize,the best
outcomeis when the poor are employedprimarilyin the exportablesectorX and consume
importablegoods M. And the worst outcomeoccurs if the poor are primarilyemployedin
sector M and consumeprimarilyexportablegoodsX.
Table 1: Locationof the poorand effectsof trade liberalizationin the short-run
M
Production
M
X
H
TotaP
-/+(0)
Consumption
X
H
-I()
-I+(-)
Totala
-/+(-)
+/+(+)
+/-(O)
+/+(+)
+/+(+
-I+(+)
?/+(+)
-I-(?/-(-)
-I+ (°)
?/+(?)
-+?
a Thiscolumngivesthe effectsfor the poorthatreceivetheirincomefrom production
in onlyone
fromthethreesectors.
sector,buttheirconsumption
basketincludesproducts
t Thisrow givesthe effectsfor the poorthatconsumeproductsfromonlyonesector,butreceive
theirincomefromthethreesectors.
Although the discussionhas focused on trade reform involvingtariffs, in practice
reforms often involve the abolition of quantitativerestrictions (QRs) such as import
licenses. As discussed above, due to rent-seeking,shifting from QRs to tariffs could
significantlyhelpthe poor.
In the long run, labor and capital are mobile across sectors. Then, trade liberalization
results in a contractionof sector M and an expansionof sectorX. If, as is likely for most
low incomedevelopingcountries,M is on average capital intensivewhile X is relatively
labor intensive,then, in the new output configurationresults in an increaseddemandfor
labor and a higher nominalwage rate. As the prices of M and H fall, labor's real income
rises as well. Consequently,while in the short run some labor employedin M loses from
trade liberalizationand the impact on labor in H is ambiguous,when factors are mobile,
labor in both sectorsgain. Of course,for this to applyto all the poor, labor marketsneed
to be integrated.If they are segmented,then some poor could lose, especiallyif they are
38
employedin the import competingsectorand are unableto move.In orderto ensurethat
the poor are betteroff followingtrade liberalization,the conditionsaffectingthe functioning
of the labormarketare thereforecritical.
In the analysispresentedabove,it is assumedthat all factors are fully employed
and changesin trade policyare reflectedin changesin relativefactor prices. In practice,
and for many of the countriesfor whichPRSPsare being prepared,there may be a large
supply of unskilledlabor in the subsistencesector that can be employedat a fixed real
wage in the modernsector. Trade reform may have a positiveimpact in this case, not
throughincreasein the wagesof the unskilledworkers but ratherby reducingthe amount
of unemployedor underemployedin the subsistencesectorand inducingan expansionof
the output of the modern sector. Indeed, following the Indian trade reform in 1991,
manufacturingemploymentincreasedfaster while wages increasedslower than before
the reform(Winters,2000). In mostcases,one can expect a lastingtrade policyreformto
have a mixtureof quantityand price effectson the labormarkets.But no matterwhat the
situation, labor mobility is essential in order ensure movementof workers from the
contracting and expandingsectors.
A.4 Sector-specific issues
The aboveframeworkis highlystylizedand abstractsfrom manyfactorsthat are important
in determiningthe impact of reform on the poor. Such factors include the existenceof
imperfect competition and inter-sectoral dependencies.For example, although the
agriculturalsectoris generallymadeup of smalland competitivefarms,this is typicallynot
the case for marketing and distributionservices. In a number of LDCs, marketing is
organizedby public agencies or parastatals,who usually fix producerprices at levels
belowworldpricesand do not alwayschangethemin responseto changesin world prices
or in exchangerates. For instance,in some of the countrieswhere the devaluationfrom
50 to 100 CFA Francsto the FrenchFranctook place,farm pricesdid not at first increase
by the full amountof the devaluation,while the prices of their importedinputs often did,
and some of the productsfarmers consumedid as well. This may have had a negative
impact on the real income of farmers in the short run. Thus, though the framework
developed above is useful in order to understandthe likelyfirst-orderimpacts of trade
reform on the real income of the poor, the specificsof each case needto be taken into
account.
To continue with the agriculturalexample, an issue to take into account is the
degree to which farmers consume their own output. The greater the share of own
consumption,the smallerthe impact of the reformon the real income of the farmers. If
farmers consumeexactly what they produce,then the real incomeeffect of trade reform
on them is nil. If farmersare net buyers,it is often arguedthat in that case farmers lose
from an increasein the price of the productthey produce.This maywell be the case, but
one must also consider that in order to be net buyers, they need to obtain additional
income. If this additionalincome is obtainedby working on other farms, real incomeof
these farmersneed not decline giventhat nominalrural wageswill tend to increasewith
the price of farm products(or increasewith trade reformin the long run).
39
Note:a numberof these papersas well as many othersdealingwith trade policyissues
can be downloadedfromwww.worldbank.org/trade.
Adams,R. (2000), "Self-targetedsubsidies:the politicaland distributionalimpact of the
Egyptianfood subsidy system,"EconomicDevelopmentand Cultural Change,
49(1):115-36(October).
Amiot, F. and 0. Salama (1996),"LogisticalConstraintson IntemationalTrade in the
Maghreb,"PolicyResearchWorkingPaper1598(May),TheWorld Bank.
Bannister,G. and K. Thugge(2001),"InternationalTrade and PovertyAlleviation,"IMF
WorkingPaper01/54(www.imf.org).
Barres,J.F. and A. Valdes (2000),"Egypt:AgriculturalExport-OrientedStrategy,"World
Bank (MNSRE).
Ben David, D. (1993), "Equalizing Exchange: Trade Liberalization and Income
Convergence,"QuarterlyJournalof Economics,108(3).
Bennathan,E. (1989),"Deregulationof Shipping:What is to Be Learnedfrom Chile,"
WorldBank DiscussionPapersNo. 67, (Washington,DC:TheWorld Bank).
and Development:Are
Bhagwati,J. and T.N. Srinivasan(1999), "Outward-Orientation
RevisionistsRight," Yale UniversityEconomicGrowthCenter DiscussionPaper
No. 806.
Cadot, O., J. de Melo and M. Olarreaga (2000), "Can Duty Drawbacks Have a
ProtectionistBias,"The World Bank,mimeo.
Clarette,R. (1989), "The EconomicEffectsof Trade Liberalizationon Philippine
Agriculture,"PhilippineReviewof Economicsand Business,vol. 26 (2), 208-235.
Coe, D. and E. Helpman (1995), "InternationalR&D Spillovers,"European Economic
Review,39, 859-887.
Coe, D. E. Helpman and E. Hoffmaister(1997), "North-SouthR&D Spillovers," The
EconomicJournal,107,134-149.
Corson,W. andW. Nicholson(1981),"TradeAdjustmentAssistancefor Workers:Results
of a Survey of Recipientsunder the Trade AdjustmentAssistanceAct of 1974,"
Researchin LaborEconomics,4:417-469.
De Melo, J., M. Olarreagaand W. Takacs (2000),"Pricing Policy Under Double Market
Power:Madagascarand the InternationalVanilla Market,"Reviewof Development
Economics.
Decker, P. and W. Corson (1995), InternationalTrade and Worker Displacement:
Evaluationof the Trade AdjustmentAssistanceProgram," Industrial and Labor
RelationsReview,48(4):758-774.
Devarajan,S. (1997),"Real ExchangeRateMisalignmentin the CFA Zone,"Journal of
AfricanEconomies,Vol. 6, No. 1, March,pp. 35-53.
Devarajan,S. and L. Hinkle (1994), "The CFA Franc Parity Change: an opportunityto
restoregrowthand reducepoverty," Africa Spectrum,29(2), 131-151.
Dollar, D. (1992), "Outward-OrientedDeveloping Economies Really Do Grow More
Rapidly: Evidence from 95 LDCs, 1976-1985," Economic Development and
CulturalChange,40(3),April, 523-544.
Dollar, D. and A. Kraay (2001). "Growth Is Good for the Poor," World Bank Policy
ResearchWorkingPaperNo. 2587.
Ebrill, L., J. Strotskyand R. Gropp (1999),"RevenueImplicationsof Trade Liberalization,"
InternationalMonetaryFundOccasionalPaperNo. 180.
Edwards,S. (1989), "ExchangeRate Misalignmentin DevelopingCountries," The World
BankResearchObserver,Vol. 4, No. 1, January,pp. 3-21.
40
Edwards, S. (1993), "Openness, Trade Liberalization,and Growth in Developing
Countries,"Journalof EconomicLiterature,31(3),September,1358-1393.
Edwards,S. (1998), "Openness,Productivityand Growth:What Do We Really Know?,"
Economic Journal, 108(March),383-398.
Edwards, S. and D. Lederman (1998), "The Political Economy of Unilateral Trade
Liberalization:The Caseof Chile",NBERWorkingPaper, no.6510,April 1998.
Elbadawi,I. and N, Majd(1996),"Adjustmentand EconomicPerformanceUndera Fixed
ExchangeRate: A ComparativeAnalysisof the CFA Zone,"WorldDevelopment,
Vol. 24, No. 5, May,pp. 939-951.
Encarnation,D. and L. Wells (1986), In TheodoreMoran(ed.),Investingin Development;
NewRolesfor PrivateCapital?WashingtonDC:OverseasDevelopmentCouncil.
Filmer,R. (2001), TheMongolianCashmereIndustry.Consultantreportpreparedfor the
World Bank,mimeo(March).
Finger, J.M. and P. Schuler(2000), "Implementationof Uruguay Round Commitments:
The DevelopmentChallenge,"WorldEconomy,April.
Frankel, J. and D. Romer (1999), "Does Trade Cause Growth?"American Economic
Review,89(3), 379-399.
Ghura,D. and T. Grennes(1993),"The RealExchangeRate and Macroeconomic
Performancein Sub-SaharanAfrica," Joumalof DevelopmentEconomics,Vol.42,
No. 1, October,pp. 155-174.
Harrison,A. and A. Revenga(1995),"FactorMarketsand Trade Policy Reform,"World
Bank manuscript.
Harrison,G., T. Rutherfordand D. Tarr (2001),"TradeLiberalization,Povertyand Efficient
Equity."Availableat: http:l/wwwl.worldbank.org/wbiep/trade/povertyconf.html.
Hinkle,L. and P. Montiel(eds.)(1999), ExchangeRate Misalignment:Conceptsand
Measurementfor DevelopingCountries,NewYork: OxfordUniversityPress.
Hoekman,B. and M. Kostecki(2001),ThePoliticalEconomyof the WorldTrading
System.Oxford:OxfordUniversityPress.
Hoekman,B., H.L. Kee and M. Olarreaga(2001),"Markups,Entry Regulationand Trade:
DoesCountrySize Matter?,"TheWorld Bank,mimeo.
lanchovichina,E., A. Mattoo and M. Olarreaga(2001), "UnrestrictedMarketAccessfor
Sub-SaharanAfrica: How much is it worth and who pays?,"World Bank Policy
ResearchWorkingPaperNo. 2595, Washington,DC (April).
lanchovichina,E., A. Nicita and 1.Soloaga(2000), "Implicationsof Trade Reformon the
Distributionof Incomein Mexico,"World Bank,mimeo.
Johnson, T. and D. Klepinger (1991), "Evaluationof the Impacts of the Washington
AlternativeWork SearchExperiment,"UnemploymentInsuranceOccasionalPaper
No. 91-4,WashingtonD.C.: U.S. Departmentof Labor, Employmentand Training
Administration.
Krueger, A. (1974), "The Political Economy of the Rent-SeekingSociety," American
EconomicReview,64(3):291-303(June).
Krueger,A. (1983), Trade and Employmentin Developing Countries: Synthesisand
Conclusions,Chicago:NationalBureauof EconomicResearchand The University
of ChicagoPress.
Krueger,A. (1998),"WhyTrade Liberalizationis Goodfor Growth,"EconomicJoumal, 108
(September),1513-1522.
Krugman,P. (1998), TheAccidentalTheorist, Norton:NewYork.
Lall, S. and P. Streeten(1977),ForeignInvestment,Transnationalsand Developing
Countries.
Larson,D. (2000),"Crop Marketsand HouseholdParticipation,"World Bank,mimeo.
41
Lawson, C. and N. Meyenn (2000), "BringingCellularPhone Service to Rural Areas,"
Viewpoint,Note204, TheWorld Bank.
Liedholm,C. and D Mead (1995),"The DynamicRole of Micro and Small Enterprisesin
the DevelopmentProcess,"GEMINIActionResearchProgramI: Final Report.
Lopez, R., Nash, J. and Stanton, J. (1995), "Adjustmentand Poverty in Mexican
Agriculture: How Farmers'Wealth Affects Supply Response",Policy Research
Workingpaper,no. 1494.TheWorld Bank,Washington,D.C.
factor
Lopez, R. and C. Romano (2000), "The determinantsof farm revenuesand
returns of poorfarmersin Brazil,"mimeo,Universityof Maryland,CollegePark.
Lopez,R. and M. Schiff(1998),"Migrationand the Skill Compositionof the Labor Force:
The Impact of Trade Liberalizationin LDCs," CanadianJoumal of Economics
(May).
Madani,D. (1999)," A Reviewof the Roleand Impactof ExportProcessingZones",World
Bank,mimeo.
Matusz,S. and D. Tarr (2000),"Adjustingto Trade Liberalization,"in Anne Krueger(ed.)
EconomicPolicyReform,Chicago:Universityof ChicagoPress,forthcoming.
Markusen, J. (1983), "Factor Movementsand CommodityTrade as Complements,"
Journalof InternationalEconomics14:341-356.
Markusen,J, T. Rutherfordand D. Tarr (2000),"ForeignDirect Investmentin Servicesand
the DomesticMarketfor Expertise,"NBERWorkingPaperNo. 7700, (May);Policy
and ResearchWorkingPaperNo.2413,TheWorld Bank (August).
Maskus,K. (2001),IntellectualPropertyin the GlobalEconomy.WashingtonDC: Institute
for InternationalEconomics.
McCulloch,N., L.A. Winters and XavierCirera (2001),Handbookon Trade Liberalization
and Poverty.London:CEPR,forthcoming.
Michalopoulos,C. (1999), "'Trade Policy and Market Access Issues for Developing
Countries,"PolicyResearchWorkingPaperNo. 2214,World Bank.
Michalopoulos,C. (2001),"Rules of Thumbfor Trade Policy," in B. HoekmanP. English
and A. Mattoo (eds.), Trade Policy, Economic Developmentand M,ultilateral
Negotiations:A Sourcebook,Washington:World Bank,forthcoming.
Mitra, P. (1992), "The CoordinatedReform of Tariffs and Indirect Taxes," World Bank
ResearchObserver,7:195-218.
Moran, T. (1998). Foreign Direct Investment and Development,Washington, D.C.,
Institutefor InternationalEconomics.
Nogues,J. (1991), "Trade,Employment,and Growthin Peru SinceWWII" in Liberalizing
Foreign Trade,DemetrisPapageorgiou,Michael Michaely,and Armeane Choski
(eds.),Basil Blackwell.
O'Leary, C. (1995), "PerformanceIndicators:A ManagementTool for Active Labour
Programmesin Hungaryand Poland,"InternationalLabourReview,134: 729-751.
Panagariya,A and D. Rodrik(1993),"Political-Economy
Argumentsfor a UniformTariff,"
InternationalEconomicReview.
Papageorgiou,D., A. Choksi, and M. Michaely (1990), LiberalizingForeign Trade in
DevelopingCountries: TheLessonsof Experience,The WorldBank.
Parker, R., R. Riopelle, and W.F. Steel (1995), "Small Enterprises Adjusting to
Liberalizationin FiveAfrican Countries,"WorldBank DiscussionPapersNo. 271.
Pursell,G. (1998),CottonPoliciesin FranconphoneAfrica,World Bank,mimeo.
Ravallion,M. (1999),"Protectingthe Poorin Crisis" PREMNote no. 12. The World Bank,
WashingtonD.C.
Ravallion,M. and D. Van de Walle, (1991), "The Impact on Poverty of Food Pricing
Reforms:WelfareAnalysisfor Indonesia",Journalof PolicyModeling,13:281-99..
42
Rhee, Y., B. Ross-Larson,and G. Pursell(1984),Korea's CompetitiveEdge: Managing
Entryinto WorldMarket.Washington,D.C.:WorldBank.
Rodriguez,F. and D. Rodrik (1999), "Trade Policy and EconomicGrowth: A Skeptic's
Guide to the Cross-NationalEvidence,"Centre for Economic Policy Research
WorkingPaperNo. 2143,London.
Rodrik, D. (1998), "Why is Trade Reform So Difficult in Africa," Joumal of African
Economies,7,10-36.
Rodrik, D. (2000), "Trade Policy Reform as InstitutionalReform," in B. Hoekman, P.
English and A. Mattoo (eds.), Trade Policy, Economic Development and
MultilateralNegotiations:A Sourcebook.Washington:World Bank,forthcoming.
Romer,P. (1994),"NewGoods,Old Theoryand the WelfareCostsof TradeRestrictions,"
Journal of Development Economics," 43(1): 5-38 (February).
Sachs, J. and A. Warner (1995), "Economic Reforms and the Process of Global
Integration,"BrookingsPaperson EconomicActivity, 1-118.
Schiff, M. and A. Valdes (1992), The PoliticalEconomyof AgriculturalPricingPolicy.
Volume 4. A Synthesis of the Economicsin DevelopingCountries, Johns
HopkinsUniversityPress.
Sen, A. (1981), Poverty and Famines: An Essay on Entitlementand Deprivation,
OxfordUniversityPress,Oxford.
Sharer,R. and others.(1998),"TradeLiberalizationin IMF-SupportedPrograms,"IMF,
World Economicand FinancialSurveys,WashingtonDC.
Schatz,H. and D. Tarr (2001),"ExchangeRate Managementand Trade Protection,"in B.
Hoekman,P. Englishand A. Mattoo(eds.), TradePolicy, EconomicDevelopment
and Multilateral Negotiations: A Sourcebook, Washington: World Bank,
forthcoming.
Smarzynska,B. (2000), "Compositionof ForeignDirect Investmentand Protectionof
IntellectualPropertyRightsin TransitionEconomies,"TheWorld Bank,mimeo.
Srinivasan,T.N. (2000), Eight lectures on India's economicreforms, Oxford University
Press.
Tarr, D. (2001), "On the Design of Tariff Policy: Argumentsfor and Against Uniform
Tariffs,"in B. Hoekman,P. EnglishandA. Mattoo(eds.), TradePolicyReformand
MultilateralNegotiations:A Sourcebook,Washington:WorldBank,forthcoming.
Takacs, W.E. (1994), The Economic Impact of Export Controls:An Application to
Mongolian Cashmere and Romanian Wood Products. World Bank, Policy
ResearchWorkingPaperNo. 1280(March).
Tyers,R. andAnderson,K. (1992),Disarrayin WorldFood Markets:A Quantitative
Assessment.CambridgeUniversityPress.
UNCTAD(1997),World InvestmentReport:TransnationalCorporations,MarketStructure,
and CompetitionPolicy,UnitedNations:NewYork.
UNCTC (1992), The Determinants of Foreign Direct Investment. A Survey of the
Evidence,UnitedNations:NewYork.
Watson,P.. 2000.'ExportProcessingZones:HasAfricaMissedthe Boat? Notyet!,'World
Bank,mimeo.
Westphal,L.E. (1990), "IndustrialPolicyin an Export-propelledEconomy: Lessonsfrom
South Korea'sExperience,"Joumalof EconomicPerspectives,4: 41-60.
Westphal,L.E. and K. S. Kim (1977),IndustrialPolicyand Developmentin Korea.World
Bank StaffWorkingPaper No 263. Condensedversionsubsequentlypublishedin
Bela Balassa and others; (1982), DevelopmentStrategies in Semi-Industrial
Economies.JohnsHopkinsUniversityPress.
Winters, L. A. (2000), "Trade and Poverty:is there a connection?,"in WTO: Special
Studies# 5: Trade,incomedisparityand poverty,Geneva.
43
Winters, L. A. (2001)."Trade Policiesfor PovertyAlleviationin DevelopingCountries,"in
B. Hoekman, P. English and A. Mattoo (eds.), Trade Policy, Economic
Developmentand MultilateralNegotiations:A Sourcebook,Washington:World
Bank,forthcoming.
World Bank.(1996)."ContainerTransportServicesand Trade:Frameworkfor an Efficient
ContainerTransportSystem."ReportNo. 15303-CHA.(TheWorld Bank).
World Bank (1999), PanamaPovertyAssessment:Prioritiesand Strategiesfor Poverty
Reduction,Human DevelopmentDepartment,Latin Americanand the Caribbean
Region,Washington,DC.
World Bank (2000a),World DevelopmentReport:Attacking poverty, Washington,DC.
World Bank (2000b),Trade Blocs. A PolicyResearchReport,Washington,DC (June).
World Bank (2000c), GlobalEconomicProspects,Washington,DC.
44
Policy Research Working Paper Series
Title
Author
Date
Contact
for paper
WPS2717Bridgingthe EconomicDividewithin RajaShankar
Nations:A Scorecardon the
AnwarShah
Performanceof RegionalDevelopment
Policiesin ReducingRegionalIncome
Disparities
November2001
A. Santos
31675
WPS2718 LiberalizingBasic
Telecommunications:
The Asian
Experience
CarstenFink
AadityaMattoo
RandeepRathindran
November2001
R. Simms
37156
WPS2719Is There a PositiveIncentiveEffect
from PrivatizingSocialSecurity?
EvidencefromLatin America
TrumanG. Packard
November2001
T. Packard
89078
WPS2720InternationalMigrationand the Global AndresSolimano
EconomicOrder:An Overview
November2001
A. Bonfield
31248
WPS2721Implicationsfor SouthAsianCountriesSanjayKathuria
of Abolishingthe Multifibre
WillMartin
Arrangement
AnjaliBhardwaj
November2001
M. Kasilag
39081
WPS2722Japan'sOfficialDevelopment
Assistance:Recent Issuesand
FutureDirections
November2001
J. Mendrofa
81885
WPS2723UsingDevelopment-Orineted
Equity HeloMeigas
Investmentas a Tool for Restructuring
TransitionBankingSectors
November2001
S. Torres
39012
WPS2724TropicalBubbles:Asset pricesin
LatinAmerica,1980-2001
SantiagoHerrera
GuillermoPerry
November2001
R. lzquierdo
84161
WPS2725BankRegulationand Supervision:
WhatWorks Best?
James R. Barth
GerardCaprioJr.
RossLevine
November2001
A. Yaptenco
38526
WPS2726Applyingthe DecisionRights
Approachto a Caseof Hospital
InstitutionalDesign
FlorenceEid
November2001
A. Santos
31675
WPS2727HospitalGovernanceand Incentive
Design:TheCase of Corporatized
PublicHospitalsin Lebanon
FlorenceEid
November2001
A. Santos
31675
WPS2728EvaluatingEmergencyPrograms
WilliamF. Maloney
December2001
A. Pillay
88046
December2001
R. Vo
33722
MasahiroKawai
ShinjiTakagi
WPS2729InternationalEvidenceon the Value Luc Laeven
of Productand GeographicDiversity
Policy Research Working Paper Series
Title
Contact
for paper
Author
Date
J. MichaelFinger
FrancisNg
SonamWangchuk
December2001
R.Simms
37156
WPS2731An AlternativeTechnicalEducation GladysLopez-Acevedo
Systemin Mexico:A Reassessment
of CONALEP
December2001
M. Geller
85155
WPS2732The UnbalancedUruguayRound
Outcome:The NewAreasin Future
WTO Negotiations
December2001
R. Simms
37156
WPS2730Antidumpingas SafeguardPolicy
J. MichaelFinger
Julio J. Nogues