UNITED NATIONS CONFERENCE ON TRADE AND DEVELOPMENT

POLICY ISSUES IN INTERNATIONAL TRADE AND COMMODITIES

STUDY SERIES No. 23

SHIFTING SANDS: SEARCHING FOR A COMPROMISE

IN THE WTO NEGOTIATIONS ON AGRICULTURE

by

Ralf Peters and David Vanzetti

Trade Analysis Branch Division on International Trade in Goods and Services, and Commodities

United Nations Conference on Trade and Development Geneva, Switzerland

UNITED NATIONS

New York and Geneva, 2004

ii

NOTE

The purpose of this series of studies is to analyze policy issues and to stimulate discussions in the area of international trade and development. The series includes studies by UNCTAD staff, as well as by distinguished researchers from academia. In keeping with the objective of the series, authors are encouraged to express their own views, which do not necessarily reflect the views of the United Nations.

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

Material in this publication may be freely quoted or reprinted, but acknowledgement is requested, together with a reference to the document number. It would be appreciated if a copy of the publication containing the quotation or reprint were sent to the UNCTAD secretariat:

Chief

Trade Analysis Branch Division on International Trade in Goods and Services, and Commodities

United Nations Conference on Trade and Development Palais des Nations CH-1211 Geneva

Series Editor: Sam Laird

Officer-in-Charge, Trade Analysis Branch DITC/UNCTAD

UNCTAD/ITCD/TAB/23

UNITED NATIONS PUBLICATION Sales No. E.04.II.D.4 ISBN 92-1-112611-8

ISSN 1607-8291

© Copyright United Nations 2004 All rights reserved

iii

ABSTRACT

The WTO negotiations on agriculture remain deadlocked after four years of discussion, and

efforts to find a solution at Cancún failed. Analysis shows that the recent draft Cancún text offers more flexibility than earlier proposals, and such flexibility most likely implies a lower level of ambition overall. However, developing countries are less able to take advantage of this flexibility and their bound tariffs will be reduced to levels at or below applied rates. The reduction in levels of intervention and the expiry of the Peace Clause make it more likely that there will be greater resort in the future to safeguards and countervailing measures.

Analysis of the various proposals using the UNCTAD/FAO Agricultural Trade Policy

Simulation Model (ATPSM) shows that most of the benefits from liberalization accrue to developed countries, which currently have the highest levels of intervention. The group of developing countries, which include both exporters and importers of agricultural products, gain from liberalization, but those gains are small and unevenly distributed. In fact, net-food importing developing countries tend to lose because of higher world prices. Within developing countries, producers tend to gain from higher world prices at the expense of consumers. Since negotiating positions suggest that governments attach a higher weight to producer than consumer benefits, a possible solution to the impasse lies in switching support in developed countries from border measures to less-trade-distorting measures such as direct income support. Providing compensation to current beneficiaries of European Union support in ACP countries for the erosion of preferences may also assist in the search for a compromise.

iv

ACKNOWLEDGEMENTS

The contribution of funding from United Kingdom DFID to further develop ATPSM is gratefully acknowledged.

v

CONTENTS

INTRODUCTION................................................................................................................ 1 I. THE STATE OF PLAY........................................................................................... 2 a. Negotiations on Agriculture ............................................................................. 2 b. Development Box ............................................................................................. 4 II. PROPOSALS FOR REFORM................................................................................ 6 a. Market Access .................................................................................................. 6 b. Domestic Support ............................................................................................. 7 c. Export Subsidies ............................................................................................... 8 d. Non-trade Concerns and Other Issues .............................................................. 9 III. MARKET ACCESS: THE BLENDED FORMULA .......................................... 11 IV. MODELLING AGRICULTURAL REFORM.................................................... 17

a. Data ............................................................................................................... 18 b. Scenarios......................................................................................................... 19

V. RESULTS – CONFLICTING INTERESTS AND IMPACTS .......................... 23

a. The EU and the United States, Two Major Players........................................ 23 b. Trade Flows .................................................................................................... 24 c. Price Changes and Distribution Effects.......................................................... 25 d. Wider Implications of Welfare Estimates....................................................... 28 e. Special and Differential Treatment................................................................. 29 f. Declining Ambitions....................................................................................... 31 g. Sensitivity Analysis ........................................................................................ 31

VI. CONCLUSIONS .................................................................................................... 33 REFERENCES................................................................................................................... 35 APPENDIX Table A1. Country coverage in ATPSM................................................................. 37 Table A2. Commodities in ATPSM........................................................................ 38 The ATPSM modeling framework........................................................................... 38 Data ........................................................................................................................ 41

vi

Tables

1. Bound and applied tariffs on agricultural products.................................................... 2 2. Applying the blended market access formula .......................................................... 14 3. Alternative liberalization scenarios.......................................................................... 22 4. Impacts of alternative proposals on the EU and United States ................................ 23 5. Consumer surplus impacts from Cancun scenarios ................................................. 26 6. Producer surplus impacts from Cancun scenarios ................................................... 26 7. Government revenue impacts from Cancun scenarios............................................. 27 8. Welfare impacts from Cancun scenarios.................................................................. 27 9. Export revenue impacts from Cancun scenarios...................................................... 28 10. Results from an ambitious Cancun scenario ............................................................ 32

Figures 1. Interests and alliances ................................................................................................ 3 2. Market access formula from various proposals ....................................................... 11 3. Impact of the blended formula on tariffs ................................................................. 12 4. New bound rates after application of the blended formula with varying distributions ........................................................................................ 16 5. Percentage change in import cost in the Cancun and Harbinson scenario .............. 24 6. Percentage change in export revenue in the Cancun and Harbinson scenario......... 25 7. Special and differential treatment: Impact on developing countries ....................... 29 8. Special and differential treatment: Impact on developed countries......................... 29 9. Impact of liberalization on developing countries..................................................... 30 10. Impact of liberalization on developed countries ...................................................... 30

1

Intransigence, shifting alliances andmiscalculation have plagued and at leasttemporarily derailed the WTO sponsoredmultilateral trade negotiations. The CancúnMinisterial Conference, as part of the DohaWork Programme, ended in failure. The 2001Doha Ministerial Declaration had launchednew negotiations on a range of subjects,including agriculture, on which negotiationshad begun earlier under the “built-in agenda”of the Uruguay Round. Agriculture was madepart of Doha’s Single Undertaking in whichvirtually all the linked negotiations weresupposed to end by January 2005. After thefirst deadline for a commitment on“modalities” in March 2003 was missed,Ministers discussed in Cancún a framework forthese modalities. Discrepancies about how toreform the agricultural trading sector were,together with the so-called Singapore issues,mainly responsible for the breakdown of thenegotiations. The road ahead seems like atrackless waste.

What can be done to put thenegotiations back on track? There are a largenumber of issues and options, with far-reaching but barely predictable consequences.Some restructuring of the negotiations seemsto be necessary in order to exploit thepossibil it ies of an agreement withoutendangering developing and less-developedcountries on issues such as food security.Recognition of developing country concernswas emphasised at the Doha MinisterialMeeting in November 2001, which putdevelopment issues at the centre of the WTOwork programme. The impact on developing

countries of the various proposals underconsideration is therefore a central focus of thisstudy.

The study attempts to provide arigorous quantitative assessment of variousoptions being discussed in the WTOagriculture negotiations. In particular, itfocuses on the Framework for EstablishingModalities in Agriculture, which was an annexof the draft Cancún Ministerial Text, SecondRevision, submitted on the 13th of September.It analyses the positions of some of the keyplayers and the joint EC-United Statesproposal. Earlier, the Chairman of the WTOCommittee on Agriculture, Mr. Harbinson,had put forward, and subsequently revised, acompromise proposal.

In order to quantify the economiceffects of these proposals, the study uses acomputable global trade model, the AgricultureTrade Policy Simulation Model (ATPSM).1

This model is a deterministic, comparativestatic, partial equilibrium trade model designedto analyse trade policy issues. A goal of thestudy is to show that a trade model can be usedto assist in the preparation and evaluation ofnegotiating positions.

The study is laid out as follows. Thenext two chapters describe the negotiatingcontext and the key proposals. In Chapter IVthe computable model is described in somedetail. In Chapter V the draft Cancún text andother recent proposals are analysed. ChapterVI deals with policy implications, limitationsand conclusions.

INTRODUCTION

1 The ATPSM modelling framework was initially developed by UNCTAD and further refined by FAO andUNCTAD.

2

(a) Negotiations on Agriculture

The WTO Agreement on Agriculturewas a significant step towards reformingagricultural trade. It brought agriculturalproducts under more effective multilateral rulesand paved the way for further liberalization ofagricultural production and trade. The DohaMinisterial Declaration of 2001 launched newnegotiations on a range of subjects, include the‘built-in’ negotiations on agriculture which hadalready begun in 2000 under the MarrakeshAgreement. Agriculture is now part of theSingle Undertaking in which virtually all thelinked negotiations are to end by January 2005.

The Uruguay Round Agreement onAgriculture “tariffied” and bound many non-tariff barriers and some progress was made inreducing tariffs on fast-growing, high-value-added products. However, much remains to bedone, including reducing tariff peaks and tariffescalation. Tariffs in agriculture are stillsignificant, even high in some product areas.

Table 1 shows average applied and bound ratesfor country groups.

Before and at Cancún, countriesexpressed their disappointment with the draftMinisterial text. Developed Cairns Groupmembers want to see a less flexible and moreambitious round, whereas countries includingJapan, Norway and Switzerland want moreflexibility, particularly in the areas of non-tradeconcerns, and less ambition. Most developingcountries want the developed countries toliberalize, but, at this stage, for reasons of ruraldevelopment and food security are reluctantto open their own markets. Some developedcountries such as the European Union do notwant at this time to eliminate export subsidies,although such subsidies constitute one of themost trade-distorting policy instruments.Essentially, the positions differ concerning thetwo dimensions of ambition and the degree ofspecial and differential treatment. Figure 1shows the positions of some WTO members.

I. THE STATE OF PLAY

2 These are simple averages at the four digit level of ad valorem tariff equivalents for commodities listed inTable A1. Applied rates are set equal to bound rates if not specified. Tariffs are averaged over 142 countries for whichdata are available.

Table 1Bound and applied tariffs on agricultural products

(in per cent)

MFN bound tariffs Applied tariffs

Developed Countries 51 48

Developing Countries 57 20Least-Developed Countries 79 17

Source: UNCTAD2

3

The horizontal axis describes the levelof ambition and the vertical axis the degree ofspecial and differential treatment toaccommodate developing country needs. Anytwo-dimensional picture can only provide arough overview because there are differentdimensions in each of the pillars - marketaccess, export subsidies, domestic support andnon-trade concerns. Nonetheless, the diagramsuggests that, as a group, developing countrieshave interests that overlap with the EuropeanUnion, the United States and the friends of“multifunctionality” that focus on non-tradeconcerns. A problem is that developingcountries are not a homogeneous group withcommon interests. Some are food importers,some exporters, while others have preferentialaccess to consider.

In an attempt to guide the variousparties to a mutually acceptable agreement, theChairman of the Committee on Agriculture,

Mr. Stuart Harbinson, circulated in March2003 a revised version of his first draft ofmodalities for the further commitments,submitted in February 2003. Many memberson either side of the agricultural tradeliberalization spectrum found the Harbinsonrevised draft inadequate. As a result,negotiations were deadlocked for months andonly very limited progress was made. The firstdeadline for the agreement on modalities,agreed at Doha, was missed. In mid-August2003, the EU and the United States jointlyproposed a modalities framework for furtherreform of agriculture but developing countriesexpressed their disappointment at theframework. However, the EC-United Statesinput galvanised the process such that severalcountries and country groups tabled alternativetexts that modify the EC-United States draft.Among these texts is a counter-proposalsubmitted by 16 developing countries that hasalso found support from four other developing

EUEU

JapanJapanSwitzerlandSwitzerland

Norway Norway Rep. of Rep. of KoreaKorea

Moderate, flexible tariff cuts and

flexible domestic support

Eliminateexport

subsidies

“Multi-functionality”

Cuts in domestic support

Substantial tariff cuts

Reduceexport credits

Developing Developing countriescountries

Special and differential treatment

United StatesUnited States

EUEU

JapanJapanSwitzerlandSwitzerland

Norway Norway Rep. of Rep. of KoreaKorea

Moderate, flexible tariff cuts and

flexible domestic support

Eliminateexport

subsidies

“Multi-functionality”

Cuts in domestic support

Substantial tariff cuts

Reduceexport credits

Developing Developing countriescountries

Special and differential treatment

United StatesUnited States

Figure 1 Interests and Alliances

Source: UNCTAD

4

countries. By the end of August 2003 a reviseddraft Cancún Ministerial Text from the WTOGeneral Council was circulated and inSeptember a second revision was released.Annex A of this draft Text is a framework forfurther reform of agriculture. The draft CancúnMinisterial Text covers the three pillars of theAgreement on Agriculture, i.e. market access,domestic support and export competition, andin this regard is comprehensive. It containsformulae, rules and special and differentialtreatment provisions on each of the three pillarsbut without specifying the level of ambition.The document does not contain specific figuresor ranges for reductions, and many issues areleft for further negotiations. The documentcontains a section for “other” issues for whichthe Harbinson revised draft is to serve as areference document.

(b) Development Box

The Doha declaration gave differenttreatment of developing countries a centralposition in the current round of negotiations.

“We agree that special and differentialtreatment for developing countries shall be anintegral part of all elements of the negotiationsand shall be embodied in the schedules of conces-sions and commitments and as appropriate in therules and disciplines to be negotiated, so as to beoperationally effective and to enable developingcountries to effectively take account of their de-velopment needs, including food security andrural development.” (Paragraph 13, Doha Dec-laration)

A large part of the current negotiationsis focused on the degree of differentialtreatment. There is a narrow and a broadnotion of a Development Box. The narrownotion is a box of measures that would beadded to the green box and comprises various

special provisions for developing countries inaddressing food security, rural poverty, etc. Thewider notion of a development box describesall concepts addressing the specific problemsof developing countries such as hunger andpoverty in food-insecure, low-income regions.Developing countries submitted variousproposals aimed at protecting and enhancingtheir food production capacity, particularly inkey staples, safeguarding employmentopportunities for the rural poor, and protectingsmall farmers from cheap imports.3 The mostprominent mechanisms in a potentialdevelopment box that are discussed in thenegotiations on agriculture include:

• Lower reduction commitmentsconcerning tariffs and domestic supportmeasures such as “de minimis” payments.

• Longer implementation periods.

• Expanded government measures ofassistance like domestic support toencourage agricultural and ruraldevelopment (Article 6.2, Agreement onAgriculture).

• Expanded access to green box exemptmeasures.

• Different formulas for tariff reductions.

• Expanded tariff-rate quotas administeredby developed countries.

• Special Products (SP).

• Special Agricultural SafeguardMechanisms (SSM).

• Preferential access to developed countrymarkets.

3 Informal paper from Cuba, Dominican Republic, El Salvador, Honduras, Kenya, Nigeria, Pakistan, SriLanka and Zimbabwe (2002). Here obtained from Braun, Wobst and Grote (2002).

5

• Special provisions for least-developedcountries and net-food-importingdeveloping countries.

Under the Special Product provisions,a limited number of sensitive products wouldbe exempt from reduction commitments, soas to enable developing countries to takeaccount of their food security, ruraldevelopment and livelihood security concerns.The selection of the products turns out to becontroversial in the negotiations because theadditional flexibility waters down the level ofambition and threatens the growth in South-South trade. The intention with this provisionis not to protect against temporary price shocksor import surges. For this purpose the SpecialAgricultural Safeguard Mechanism provides atime-limited safeguard against imports whenthey threaten to disrupt domestic production.It is supposed to be invoked in reaction toexceptional market conditions. It was debated

whether the mechanism should be restrictedto a limited number of food security crops likecereals or broadened to include particular cropsimportant for the livelihood of many poorpeople in developing countries. The potentialcriteria to be used in the identification ofeligible products could be based on numerousfactors, each favouring some countries at theexpense of others.4 Agreement on suitablecriteria has yet to be worked out.

One difficulty in the currentnegotiations is the determination of countriesthat can benefit from certain development boxmeasures. Developing countries are diversewith respect to their resource endowments,level of development, degree of integration intothe world economy, and their current povertyand food security situation. Hence, it isdifficult to get agreement as to which countriesshould receive special and differentiatedtreatment.

4 For more details see e.g. Ruffer and Vergano (2002).

6

(a) Market Access

The early United States proposal foraddressing market access issues is to reduceapplied tariffs according to a harmonising Swissformula by which higher tariffs are reducedmore than proportionately (USDA 2002).Under this formula the maximum final tariffis proposed to be 25 per cent.5 This implies,for example, that a tariff of 100 per cent wouldbe reduced by 80 per cent while an initial tariffof 10 per cent would be reduced by about 30per cent. Other elements of the proposalinclude elimination of in-quota tariffs and a20 per cent expansion of import quotas. Sinceit focuses on applied tariffs, it has a significantimpact on developing countries as it makesirrelevant the often substantial differencebetween applied and bound tariffs. Developingcountries would be obliged to makeproportionally greater cuts from their boundrates than developed countries. As countrieswith the same initial rates are treated similarly,the approach does not recognize special anddifferentiated treatment for developingcountries as agreed in the Doha Declaration.However, the harmonizing formula isparticularly appropriate in reducing tariff peaksand tariff escalation.

Tariff peaks and escalation are notspecifically mentioned in the EC proposal,which is a continuation of the Uruguay Roundapproach, a 36 per cent average cut in boundtariffs with a minimum 15 per cent cut in eachtariff line (EC, 2002). While the EU proposalmentions but does not specify the special and

differentiated conditions that apply todeveloping countries, they are interpreted assimilar to the Uruguay Round conditions,whereby developing countries implementedtwo-thirds of these reductions over a longerimplementation period. This approachcontains an inherent flexibility that may allowhigh tariffs to be maintained on sensitiveproducts. For example, a reduction in tariffson a sensitive product from 100 to 85 per centcould be offset by reducing a 10 per cent tariffto 4.3 per cent to give the required simpleaverage cut of 36 per cent. While import-competing producers may feel this to beadvantageous, exporters would likely beconcerned that it does not adequately addresstariff peaks nor provide them with sufficientimprovement in market access.

The Harbinson Proposal is acompromise between the harmonizing and theflexible approach (WTO Committee onAgriculture, 2003). Out-of-quota bound tariffswould be reduced by a simple average for allagriculture products, subject to a minimumreduction per tariff line. The formula includesbands where, depending on the initial tariff,average and minimum reductions are higherfor higher tariffs. For developed countries theproposed average reduction is between 40 and60 per cent and the minimum between 25 and45 per cent. For developing countries thereductions are between 25 and 45 per cent witha minimum between 15 and 30 per cent. Tariffquota quantities would be expanded to 10 percent of current domestic consumption indeveloped and 6.6 per cent in developing

II. PROPOSALS FOR REFORM

5 The proposed Swiss formula is new tariff=(initial tariff*25)/(initial tariff+25).

7

countries. Least-developed countries would notbe required to undertake any reductioncommitments.

The EC-United States joint proposalis to apply the Uruguay Round approach to acertain, as yet unspecified, share of tariff lines,the Swiss formula to a further share of tarifflines, and the provision of duty-free access tothe remainder. The first group would mostlikely include the more sensitive products.Furthermore, a maximum tariff or anequivalent additional market access isproposed. Developed countries would provideduty-free access for a certain percentage ofimports from developing countries.Concerning special and differential treatment,the proposal is that developing countries mayreduce tariffs by a smaller amount.

The draft Cancún Ministerial Text,second revision, put forward by the GeneralCouncil of the WTO adopts the EC-UnitedStates blended formula (WTO GeneralCouncil 2003). No maximum tariff is proposedfor developing countries (although it wouldremain under negotiation) and the duty-freepart of the blended formula is replaced by thealternatives to reduce tariffs to zero or 5 percent. Tariff reductions are to be lower andimplementation periods longer in developingcountries.

Both the Harbinson and the draftCancún Text foresee Special Products fordeveloping countries. Harbinson proposed toreduce the corresponding tariff lines by anaverage of ten and a minimum of 5 per cent.The creation of a category of Special Productsis a major demand by developing countriesincluding the so-called group of 33 countriesthat want these products to be exempt fromany reductions.

The draft Cancún text calls fordeveloped countries to accept duty-free allimports from least-developed countries and acertain percentage of imports from developingcountries. The EC had already proposed to

provide duty-free access for 50 per cent of allimports from developing countries and 100 percent for least-developed countries. Thus, theEuropean Union itself substantially meets thiscriterion. Among the major importers Japanwould have the most difficulty meeting thisstandard as only a quarter of its agriculturalimports from developing countries are duty-free.

(b) Domestic Support

Support levels are still significantdespite various declarations of intent. Forexample, in the OECD countries totalagricultural production in 2000 was valued atthe farm gate at US$632 billion, but toencourage this production, producers receivedsupport of US$323 billion (OECD 2002). Themajor beneficiaries of this largesse areproducers in the European Union (35 per centof OECD receipts), the United States (27 percent) and Japan, and most of the estimatedgains from liberalization stem from reform inthese regions. Most developing countriescannot afford and do not grant substantialdomestic support.

Domestic support measures indeveloped countries appear to increase globalproduction, forcing down world prices. Thisbenefits consumers in net food importingdeveloping countries at the expense of netexporters. Since producers in both groups ofcountries face lower prices as a result ofdomestic support in developed countries, mostdeveloping countries are demanding thereduction of domestic support.

In WTO terminology, domesticsupport in agriculture is classified by “boxes”.Green-box support must be only minimallytrade distorting, whereas amber-box supportmeasures are considered to distort productionand trade and, as such, are subject toreductions. The blue box is for direct paymentsthat are tied to programmes that limitproduction, for example, payments are basedon historical land area or number of livestock.

8

The United States proposal fordomestic support reductions is to reduce overfive years the non-exempt support as definedby the Aggregate Measurement of Support(AMS) (amber box) as well as production-limited (blue box) support to at most 5 percent of the average value of agriculturalproduction. By some later date, all non-exemptdomestic support would be eliminated. Deminimis payments, i.e. support not exceedingfive per cent of the total value of production,would be excluded from reductions andsubsequent elimination. Developing countrieswould have special conditions to enable themto provide additional support to facilitatedevelopment and food security.

The EC proposal involves maintainingthe amber, blue and green boxes essentiallyunchanged and reducing the amber boxAggregate Measurement of Support by 55 percent. The green-box criteria would be expandedto encompass so-called ‘non-trade concerns’such as rural development, the environmentand animal welfare. This is in contrast to theUnited States proposal whereupon the green-box criteria would not be expanded. At presentthe EU’s AMS expenditure is not a bindingconstraint, but could become so in the future,depending on movements in world prices. Aflexible green-box allows support to beswitched from the non-exempt amber to theexempt green-box, as decided in June 2003 bythe EC in the reform of the its CommonAgricultural Policy (CAP) by increasing directincome support. Finally, the European Unionproposes eliminating the de minimis provisionin developed countries. The European Unionmakes less use of this provision than the UnitedStates.

The Harbinson proposal on domesticsupport is to maintain green-box supportmeasures unchanged and to reduce blue boxpayments by 50 per cent in developed and 33per cent in developing countries. The amberbox Aggregate Measurement of Support wouldbe reduced by 60 per cent in developed and40 per cent in developing countries. The de

minimis level of five per cent would be reducedto 2.5 per cent in developed and would remainunchanged at ten per cent in developedcountries.

The EC-United States joint proposalalso envisages leaving green-box supportmeasures unchanged but broadening andweakening the definition of direct blue boxpayments. These “new blue box” paymentswould have to fulfil several requirements butwould no longer have to be production-limiting. Under the proposal they would notexceed 5 per cent of the total value ofagriculture production. The “most trade-distorting domestic support” and de minimispayments would be reduced in a certain range,with countries having the higher trade-distorting support making greater efforts. Thesum of amber and “new blue” box and deminimis support would be capped at the sumof the amber and blue box and de minimissupport level in 2004.

The draft Cancún text adopted the EC-United States proposal to modify and expandthe blue box but required a linear cut of thecorresponding payments. Green-box paymentswould remain under negotiation, whichprobably means that there would not be anychanges in the next few years. As in the EC-United States proposal, amber box and deminimis payments would be reduced within acertain range.

(c) Export Subsidies

Of the current 148 WTO members, 25countries have export subsidy commitments,volume and budgetary outlay constraints, forvarious groups of products. Almost 90 per centof all agricultural export subsidies are providedby the European Union. It is, therefore,perhaps not surprising that the United Statesproposes to eliminate export subsidies over fiveyears whereas the European Commissionsuggests a modest reduction of an average 45per cent in expenditure. As with tariff cuts,averaging provides flexibility by permitting

9

large cuts in lightly traded or lightly protectedproducts.

Between 1995 and 2000 the EU’saverage subsidies were US$5.5 billion, only 20per cent lower than its final bound expenditurelevel of US$6.8 billion. But in 2000 and 2001,outlays decreased to US$2.5 and US$2.3billion respectively and, could thereforeaccommodate a reduction of more than 60 percent in the total expenditure. However, severalindividual commodities are currently upagainst expenditure or volume constraints,including rice, sugar, cheese and other milkproducts, poultry, fresh fruits and vegetablesand incorporated products.

The United States proposes, in additionto the elimination of export subsidies, thatdisciplines would be placed on officiallysupported export credits, food aid and otherforms of export support without specifyingquantitative limits. Globally, most exportcredits are provided by the United States totheir farmers. The EU proposes that the trade-distorting elements of export credits foragricultural products should be identified andsubjected to strict disciplines.

The Harbinson Proposal involveseliminating export subsidies in both developedand developing countries, although the latterwould have a longer implementation period.Export credits would be subject to disciplines.

In their joint paper, the EC and theUnited States propose to eliminate exportsubsidies for as yet unspecified products thatare of particular interest to developingcountries, and to reduce export subsidies forthe remaining products. Trade-distortingelements of export credits should be treated inthe same manner as export subsidies.

In the draft Cancún text, the WTOGeneral Council adopted the EC-United Statesapproach with a view to eventually phasing outall export subsidies and trade-distortingelements of export credits.6 Most developingcountries, including the Group of 20, areseeking the elimination of all forms of exportsubsidies as an outcome in the currentnegotiations. The failure to meet the objectivein the draft Cancún text was one of the majorconcerns of developing countries and especiallynet food exporting countries.

(d) Non-trade Concerns and Other Issues

The agriculture negotiations providescope for governments to pursue “non-trade”concerns such as the environment, ruraldevelopment, labour standard and foodsecurity. However, not all countries are readyto negotiate these “non-trade” issues. TheUnited States does not mention this issue atall and favours a narrow round excluding theseissues.

The European Commission proposesthat measures aimed at achieving certainsocietal goals such as the protection of theenvironment, traditional landscapes, ruraldevelopment and animal welfare should beaccommodated in the agreement onAgriculture. For example, payments tocompensate for the additional cost of meetinghigher animal welfare standards would beexempt from reduction commitments underthe proposal. Other non-trade concernsinclude geographical indicators, such as‘Champagne’, and restrictions on imports ofgenetically modified organisms. TheHarbinson Proposal acknowledges non-tradeconcerns such as structural adjustments andanimal welfare. Payments should be time-limited.

6 Agreed disciplines on export credits would address appropriate provisions for differential treatment infavour of least-developed and net food-importing developing countries.

10

So-called non-trade concerns aredifficult to specify and hence cannot readilybe modelled with models that are used to assessthe economic effects of alternative scenarios.However, this doesn’t imply they areunimportant. They may have an importantbearing on the negotiations. Countries that arepushing strongly for these concerns to beaddressed may well be prepared to compromisein other areas, such as export subsidies ordomestic support. While many of these non-trade concerns may be understandable from adeveloped country’s perspective, they tend toincrease standards and the amount oflegitimate farm support to levels that most

developing countries cannot afford, and thusimpose a disadvantage against developingcountries in international trade.

These non-trade concerns add to thecomplexity of the negotiations on agriculture.Both the EC-United States joint paper and thedraft Cancún text list a number of issues onwhich there is no agreement and on whichfurther work on modalities is necessary. Amongthese are several non-trade concerns, the peaceclause7 and flexibility for certain groupings.These issues are not covered in greater detailin this study.

7 Under the WTO Agreement on Agriculture, it was agreed (Article 13) that members would refrain fromthe use of countervailing measures until the end of 2004. This was known as the ‘peace clause’.

11

The differences among the fourproposals are reflected in the market accessapproaches. Each proposal contains a formulawith which new tariffs can be calculated fromthe initial tariffs. The approaches differaccording to average levels of reduction,flexibility inherent in the formula and, theextent to which tariff escalation and tariff peaksare addressed. A so-called harmonizingapproach cuts higher tariffs by a higherpercentage than lower tariffs and thus addressestariff escalation and tariff peaks. This is a

particular characteristic of the Swiss formulathat the Uruguay Round approach does notpossess to the same degree. Countriesconcerned to protect their own agriculturalmarkets tend to reject a harmonizing approach,whereas agricultural exporters like CairnsGroup members tend to believe it will opengreater market opportunities for them. TheHarbinson approach is a compromise betweenthe Swiss formula and the Uruguay Roundformula (see Figure 2).

III. MARKET ACCESS: THE BLENDED FORMULA

Figure 2Market access formula from various proposals

Source: UNCTAD

0

10

20

30

40

50

60

70

80

90

0 20 40 60 80 100 120 140

Old Rate (%)

New

Rat

e (%

) EU Harbinson

United States

12

The blended formula in the Cancúntext gives countries a degree of flexibility suchthat if countries take advantage of theirpotential clearance, it can even be anti-harmonizing. This can be seen in Figure 3where final tariffs are plotted againsthypothetical initial values. It is assumed thatbound tariffs are equally distributed between0 and 150 per cent and that high tariffs aresubject to the Uruguay Round approach and,within this category the highest are reducedusing the minimum requirement, intermediatetariffs are reduced using the Swiss formula and,small tariffs are reduced to zero. For developingcountries the concept of Special Products hasbeen taken into account and small boundtariffs are reduced to five rather than zero percent. The issue of tariff escalation in developed

countries is however, addressed since it isproposed in the Cancún text to apply an as yetunspecified factor to the tariff reduction of theprocessed product in cases where its tariff ishigher than the tariff for the product in itsprimary form. This would increase thecommitments for developed countries butcould not be taken into account in our analysis.

Since the draft Cancún Ministerial textis a framework that does not contain specifictargets, it is necessary to make assumptionsabout plausible values in order to analyse theblended formula. The following box containsthe proposed formula where the assumed italicnumbers replace the empty brackets in theCancún Ministerial text (Paragraphs 2.1 and2.7 of Annex A of the Cancún Ministerial text).

Figure 3Impact of the blended formula on tariffs

Source: UNCTAD

0

20

40

60

80

100

120

140

160

0 20 40 60 80 100 120 140 160

Developing Countries

Developed Countries

13

It is important to identify theconsequences of the blended formula fordeveloped and developing countries. Toil lustrate the effects of the possiblespecifications in Box 1, the three largestdeveloped countries were chosen and 15developing countries from different regionsand with different tariff schemes were chosenarbitrarily. The blended formula was appliedwith the assumptions of reductions specifiedin Box 1 to bound tariff rates at the HS six-digit classification level. The final bound tariffrates were calculated and it was checkedwhether applied rates would have to be reducedor not. The general rule, which was appliedwhen choosing which tariff lines fall under

which part of the blended formula, was guidedby the objective that applied tariffs are to beas little affected as possible.8 Initial tariffs areranked by the height of bound rates fordeveloped countries. Since in most developingcountries there is a significant differencebetween bound and applied rates, the rankingis according to the percentage differencebetween bound and applied rates in developingcountries. Where the difference is high andbound tariffs are low, the bound rates arereduced to five rather than zero per cent.9

Table 2 shows the initial bound andapplied rates as well as the final tariff rates.10

It can be observed that the percentage

8 By choosing the reductions for each of the more than 600 tariff lines strategically a single country couldbetter achieve the objective of reducing applied rates as little as possible. However, a transparent rule that can bereadily applied to all countries was chosen instead. With this procedure the average reduction of bound rates is notminimized. A maximum tariff was not taken into account.

9 For more details and the line of reasoning concerning these rules see also Section IV Modelling agriculturalreform, Subsection (b) Scenarios.

10 The percentage change of the bound and applied tariff rates in the last two columns do not match withthe difference of tariffs shown in the table as the difference is the more accurate average of percentage differences atthe six digit level.

Box 1Possible tariff reduction specifications under the Cancún proposal with assumed numbers

The formula applicable for tariff reduction by developed countries shall be a blended formula under whicheach element will contribute to substantial improvement in market access for all products. The formula shall be asfollows:(i) 40% of tariff lines shall be subject to a 36% average tariff cut and a minimum of 15%; for these import-

sensitive tariff lines market access increase will result from a combination of tariff cuts and TRQs.(ii) 40% of tariff lines shall be subject to a Swiss formula with coefficient 25.(iii) 20% of tariff lines shall be duty-free.

The formula applicable for tariff reductions by developing countries shall be as follows:(i) 50% of tariff lines shall be subject to a 24% average tariff cut and a minimum of 10%; for these tariff lines

market access increase will result from a combination of tariff cuts and TRQs. Within this category, develop-ing countries shall have additional flexibility under conditions to be determined to designate Special Products(SP) which would only be subject to a linear cut of a minimum of 5% and no new commitments regardingTRQs; however, where tariff bindings are very low (below 30%) there shall be no requirement to reducetariffs.

(ii) 40% of tariff lines shall be subject to a Swiss formula with a coefficient of 50.(iii) 10% of tariff lines shall be bound between 0 and 5%, taking into account the importance of tariffs as a source

of revenue for developing countries.

14

reductions in bound rates under the blendedformula are 22, 16 and 20 per cent in the threedeveloped countries, the EU, the United Statesand Japan respectively, much lower than thereductions of between 32 and 52 per cent indeveloping countries (column (5)). This is thecase even though the commitments in each ofthe components of the blended formula werechosen to be higher for developed countriesthan for developing countries. Developedcountries for example do not have SpecialProducts and the Swiss coefficient of 25 muchmore ambitious in terms of liberalisation thanthe corresponding coefficient of 50 fordeveloping countries. The reason for this is thatbound tariff rates for developed countries are

quite heterogeneous and therefore thesecountries can take advantage of the flexibilityinherent in the blended formula. This is notthe case for developing countries where boundrates for the various tariff lines are morehomogenous and on average higher than indeveloped countries. Thus, concerning boundrates the blended formula includes higherobligations for developing than for developedcountries even if special and differentialtreatment through smaller numbers in each ofthe components is taken into account. In termsof simple average bound rates, the blendedformula appears to better suit the needs ofdeveloped countries. In effect, this would be areverse special and differential treatment.

Table 2Applying the blended market access formula

Source: UNCTAD calculations

-1 -2 -3 -4 -5 -6 -7 -8 -9 -10

% % % % % % % % % %

EU 17 17 11 11 22 22 8 8 35 34United States 6 6 4 4 16 16 3 3 31 31Japan 21 18 14 13 20 18 9 9 34 31Bolivia 40 10 25 10 37 5 28 10 30 0Cameroon 78 23 43 23 44 2 51 23 35 1Cuba 37 11 24 11 35 0.2 26 10 29 2Domi.Rep. 40 23 25 22 36 4 28 21 30 6Egypt 96 52 20 19 32 14 61 50 28 9Guatemala 52 11 26 11 39 0.2 35 11 31 1Honduras 32 11 21 11 34 0.2 23 11 29 1India 115 42 48 36 49 12 72 38 36 5Indonesia 46 7 27 6 38 0.1 32 5 30 0.5Kenya 100 24 56 24 43 0 65 24 35 0.1Nigeria 150 33 75 32 50 5 90 32 40 0.2Pakistan 97 23 49 22 46 3 63 21 35 3Peru 31 18 20 16 34 9 22 16 30 7Sri Lanka 49 21 29 20 39 0.2 34 20 30 1Venezuela 31 16 20 16 34 2 22 16 30 0

Bou

nd

tari

ff

Bou

nd

tari

ff

Bou

nd

tari

ff

Percentagereduction

Percentagereduction

Bou

nd

tari

ff

App

lied

tari

ff

App

lied

tari

ff

App

lied

tari

ff

App

lied

tari

ff

Blended formula Harbinson approachFinal values

Final values

Initialvalues

Bou

nd

tari

ff

App

lied

tari

ff

15

While WTO negotiations are a legalprocess concerned with the negotiation ofreductions in bound rates, it is also instructiveto look at applied tariff rates to ascertain likelydirect economic effects of possible negotiatedoutcomes. With cuts between 16 and 22 percent, reductions in the applied rates tend tobe deeper in developed countries, whereas thereduction of applied rates in most developingcountries are more moderate (Column 6). Innine out of the 12 developing countriesreductions in applied rates are less than 5 percent. Only in Egypt and India are cuts higherthan 10 per cent. In developing countries, thereis in general a degree of “water in the tariffs”(the gap between bound and applied rates),and, as seen in Table 2, the implication of theproposals is to squeeze out this water bybringing bound rates to the same level or lowerthan applied rates. This process completelyremoves any latitude for increasing appliedrates, for example in response to import surgesor subsidized imports. This suggests that, inthe future (also with the passing of the PeaceClause on countervailing actions), there maybe greater resort to safeguards or countervailingactions than in the past.

Comparing the blended formula withnumbers as in Box 1 with the Harbinsonapproach, it can be seen in Table 2 that thereductions of bound rates are higher under theblended formula for all developing countries(columns (5) and (9)). Developed countrieshowever, have lesser obligations under theblended formula. New bound and applied ratesare lower with the Harbinson approach thanwith the blended formula. In some developingcountries new applied rates are lower under theHarbinson approach smaller than under theblended formula. The reason for this is that

advantage was not taken of the flexibility thatthe Harbinson approach gives countries whenthey calculate the new bound rates. As in theUruguay Round component of the blendedformula, bound tariffs can be reduced by acertain minimum whenever the requiredaverage is met. Furthermore, althoughsuggested by Harbinson, Special Products werenot taken into account when calculating newtariffs under the Harbinson approach.Therefore, although it is theoretically possibleit is practically unlikely that the new appliedrates in any developing country would besmaller under the Harbinson approach thanunder the blended formula with assumptionsas in Box 1 (columns (6) and (10)). Taken theflexibility of the Harbinson approach andSpecial Products into account, new appliedrates are likely to be smaller under the blendedformula in all developing countries.

However, whether or not obligations ofdeveloping countries are higher with theblended formula depend on the numbers thatare assumed. With the specified assumptionsthe average reduction in applied rates indeveloping countries is about 8 per cent.11 Ina much more ambitious Cancún scenario,where reductions in each of the componentsof the blended formula are higher and wherethe distribution of tariff lines among thecomponents is stricter, the average reductionof applied tariffs is 14 per cent. This compareswith the 10 per cent average reduction ofapplied tariffs in developing countries in theHarbinson scenario.12 Reductions of appliedrates in all developed countries are 24 per centwith the above stated assumptions, 32 per centin the ambitious Cancún scenario and 26 percent in the Harbinson scenario.

11 Based on the 99 developing countries that are in UNCTAD’s ATPSM database (see Appendix for a list ofcountries).

12 Note, that the numbers for all 99 developing and all 20 developed countries are calculated at the ATPSMaggregation level, which is roughly the 4-digit level. The ambitious Cancún scenario is described in greater detail inSection 6, Subsection Sensitivity analysis.

16

Tariff reductions are highly sensitive tothe bands between components. The sensitivityof the average reduction in bound tariff ratesin developing countries to changes in thedistribution of tariff lines between componentsof the blended formula is shown in Figure 4.Since the Swiss formula component and thereduction to 5 per cent reduce relatively highbound rates considerably, the overall reductionincreases with the number of tariff lines thatare subject to these two components. Thereduction within each category remains asassumed in Box 1. If for example, a countryhas all initial tariffs at 100 per cent and, if 80per cent of the tariff lines are subject to theUruguay Round component, 10 per cent tothe Swiss formula and 10 per cent reduced to5 per cent, then the new bound rates are onaverage 65 per cent and the average reductionis 35 per cent. However, if instead of 80/10/10 the distribution of tariff lines acrosscomponents is 33.3/33.3/33.3, the averagereduction increases dramatically from 35 to 62per cent.

Improved access to developed countrymarkets is important for developing countries.Assuming that the major developed countrieswould wish to preserve the sectors that are mostvulnerable to import competition, they wouldput sugar, dairy products, beef meat and somecereals under the Uruguay Round component.Tariffs on sugar, which is an important productfor developing countries since they can producea substitute to sugar beet (which is producedin temperate regions) would only be reducedminimally. Also, processed chocolate, oil seedsand some fruits and vegetables, including sometropical fruits, would be put into the UruguayRound component with low reductions.Products for which tariffs are already ratherlow, like tea, vegetable oils, coffee beans andcacao, are likely to be made subject to the Swissformula or duty free component. Thus, theflexibility inherent in the Cancún proposalseems likely to water down substantially thelevel of ambition.

Figure 4New bound rates after application of the blended formula with varying distributions

Source: UNCTAD

0

20

40

60

80

100

120

Initial Value Harbinson 80-10-10 60-30-10 50-40-10 40-40-20 34-33-33

17

UNCTAD’s Agricultural Trade PolicySimulation Model (APTSM) is used toestimate the potential impact of variousproposals for reforming the agricultural tradesector, assuming their implementation is asspecified.13 The static, partial-equilibrium,global, agricultural-trade model is able toestimate the economic effects of changes inwithin-quota, applied and out-quota tariffs,import quotas, export subsidies and domesticsupport on production, consumption, prices,trade flows, trade revenues, quota rents,producer and consumer surplus and welfare.A more detailed description of ATPSM can befound in the Appendix to this paper.

The Uruguay Round reformsintroduced several policies that presentdifficulties in quantification. Quotas and tariffson imports and export subsidies generate quotarents of some US$10 billion.14 It is assumedhere that all the rents (over and above “normal”profits) generated by the EU and United Statessugar policies and half of the rents frombananas are initially allocated to producers inexporting countries according to thedistribution of trade.15 Rents from theremaining products are initially allocated to theimporting countries.

A second difficult modelling issueconcerns the decoupling of domestic support,that is, the production effects of changes insupport. This is a complex issue concerningthe method of administration, perceptions ofrisk, the wealth effects of direct payments andthe likelihood of changes in governmentpolicies. The approach taken here is to assumethat most of the domestic support is decoupledor is conflated with border support.16 Thus,the additional effects of removing domesticsupport are minimal in most cases. Thisassumption may give a downwards bias to theestimated benefits from liberalization.

Several modelling assumptions are alsoimportant to note. For example, the ATPSMallows two-way trade. This requires anadditional equation to specify either exportsor imports. In this version of the model thechange in imports is determined through an“Armington” elasticity (determining thesubstitutability of differentiated products fromsuppliers at home and abroad), which is set at2.2. Consumers first decide how much theyconsume and, in a second step, how muchdomestic and foreign products they want tobuy. Exports are determined so as to clear themarket, that is, supply plus imports equalsdemand plus exports.

IV. MODELLING AGRICULTURAL REFORM

13 An operational version of the model, associated database and documentation are available free of chargefrom UNCTAD (http://www.unctad.org/tab).

14 This estimate assumes import quotas are filled. To the extent that import quotas are unfilled, the estimateis inflated.

15 In a previous application of the model, reported in Vanzetti and Sharma (2002), it was assumed that therent from all products went to producers, although this assumption is difficult to justify. The allocation of rentsaffects the distribution of gains from liberalization.

16 See de Gorter (1999) for a discussion of the methodological issues involved in measuring domestic support.

18

Where producers receive rents they donot respond by changing quantities produced.This implies for example, that changes inwithin-quota tariffs change only quota rents,not quantities, prices or global welfare. Theshifting of quota rents is a zero sum game. Thisis described in more detail in the Appendix,

The model does not have a specifictime-dimension. The general interpretation isthat the economic effects are of a medium-termnature, with the impacts taking three to fiveyears to work through.

A final observation relates tolimitations on modelling preferential access.Data on bilateral tariffs are not included in thedatabase, although bilateral trade flows areavailable. Thus, it is not possible to liberalizeon a bilateral basis and directly capture theeffects of preference erosion as MFN rates arebrought down closer to preferential rates heldby many developing and all least-developedcountries. However, much of the effect ofdiminishing preferences is captured by thedepletion of quota rents allocated initially toexporters. The model structure does not allowfor trade diversion from changes in rents, butwhere the quotas are filled this effect will beminimal, at least for small changes in prices.

The present version of the model covers175 countries of which the current 15European Union members form a singleregion. Countries designated here as‘developed’ are defined by the World Bank ashigh-income countries with per capita GNPin excess of US$9,266 (World Bank, 2001).Another group is the 49 least-developedcountries as defined by the United Nations.There are 36 commodities in the ATPSM dataset, covering most of the agricultural sector.This includes many tropical commodities ofinterest to developing countries, although

many of these have relatively little trade bycomparison with some of the temperate-zoneproducts. Meat, dairy products, cereals, sugar,edible oils, vegetables, fruits, beverages,tobacco and cotton are among the productscovered by the ATPSM database.

(a) Data

The data in the model come fromdifferent sources, including OECD (AMAD),FAO, UN, WTO and UNCTAD (seeAppendix).17 The year 2000 represents the baseyear for the model.

An indicator of the degree of distortionis the revenue raised or governmentexpenditure outlaid on each commodity. Mostof the global protection in agriculture is ontemperate products, particularly beef, wheat,maize, dairy products, vegetables oils andoilseeds. According to the ATPSM database,tariff revenues and rents for the products inthe model amount to around US$45 billion,with export subsidies and productiondistorting domestic support accounting for anadditional US$11 billion. Among the productsthat can be grown in tropical regions, tobacco,sugar and poultry attract substantialprotection. These products or close substitutescan also be grown in temperate regions. Thereis relatively little tariff revenue raised ontropical products such as beverages (exceptchocolate) and cotton. For the productscovered by ATPSM, tariff revenues amount to17 per cent of import costs.

The European Union and Japan raisethe largest amounts of agricultural tariffrevenue (over US$4 billion each) but severalother countries collect over US$1 billionannually. These are Mexico, the Republic ofKorea, the United States, the United ArabEmirates, Egypt and Turkey. In fact, 50

17 The OECD maintains the Agricultural Market Access Database (AMAD), a cooperative effort betweenseveral international and national institutions (including the European Commission). The database is publicly availableonline.

19

countries gather in excess of US$100 millionannually in agricultural tariff revenues. Thisillustrates the scope for global reform.

The major commodities attractingexport subsidies are wheat, beef, dairy productsand sugar. Of the US$4.6 billion attributed tocommodities in the database, US$4 billion ispaid by the European Union, with Switzerland,Norway and the United States responsible formuch of the remainder. The European Union(US$2.3 billion) and Japan (US$1.9 billion)also provide most of the domestic support thatis considered in the ATPSM database to beproduction distorting. Once again, the UnitedStates accounts for most of the remainder.Tobacco leaf, cotton, fresh milk and beefaccount for the largest slices of domesticsupport.

(b) Scenarios

Four simulations are undertaken toanalyse the proposals on agriculture. Anambitious scenario now seems off the agendabut indicates the opportunity cost of foregonegains. A conservative proposal is the UruguayRound continuation that was once consideredthe least that could be gained from the currentround. The third scenario is the Harbinsonproposal that was the basis for discussion formonths before the EC-United States proposalwas submitted. Finally, the fourth scenario isthe WTO’s response to the EC-United Statesproposal. The four scenarios are listed in Table3 and described in more detail in this section.

The ambitious scenario is relativelystraightforward. It is based on the United Statesproposal of a Swiss cut with a maximum tariffof 25 per cent. Tariff cuts are based on appliedrather than bound rates, an importantdistinction for developing countries where thegap between bound and applied rates can besignificant.

The second scenario is the conservativeproposal, an extension of the Uruguay Roundapproach. Developed countries make tariff cutsaveraging 36 per cent with a minimum of 15per cent and in developing countries theaverage must be 24 per cent with a minimumof 10 per cent. In modelling this, an attempthas been made to specify which products wouldbe selected for the minimum tariff reductions.Concerning the flexibility that countries mighthave in selecting the reduction in single tarifflines, the following reasoning was used, whichwas also mentioned briefly in Chapter 3, wherethe blended formula was described. WTOmembers usually want other countries toliberalize whereas they want to protect theirown markets or, at least they want to have theflexibility to protect them by not having strongbinding commitments. For developedcountries, it was assumed that products withhigh bound tariffs are (politically) sensitiveproducts and, therefore, countries want toreduce these tariffs as little as possible. For theproducts with the highest tariffs, the minimumpossible reduction is implemented and lowertariffs are reduced such that the requiredaverage is met. For developing countries, thesame rule was applied but, the most sensitiveproducts are defined as having the smallestpercentage difference between out-quotabound and applied tariff rates.18 Since mostdeveloping countries have much higher boundthan applied tariff rates, this rule means thatdeveloping countries’ applied rates are onlyslightly affected. The reduction for eachcountry was not chosen such that applied ratesare affected as little as possible but rather anumber of tariff lines were specified that arereduced according to the minimum reductionand the remaining tariffs reduced such thatrequired averages were met.

18 In developed countries the difference between bound and applied rates is small so that for these countriesthis rule does not make sense.

20

In the Harbinson proposal, cuts dependon the initial tariff level. Although countrieshere also have the flexibility to apply withinthree or four bands minimum reductionswhenever an average is met, the dependencyof the average and the minimum on the initialtariff reduces the flexibility. Therefore, whenanalysing this proposal, the specified selectionrule for sensitive products is not applied butrather all tariffs in each of the bands by theproposed average are cut.

The next simulation is the Cancúnproposal. The main concern in specifying thissimulation is the lack of detail in the proposal.For example, the draft Cancún Ministerial textdoes not contain specific targets, but phrasessuch as “[…]% of tariff lines would be subjectto a […]% average tariff cut and a minimumof […]%, …”. In order to assess theseproposals, assumptions on the numbers haveto be made. Therefore, it is not possible topredict the precise nature of the tariffs or eventhe distribution of tariff cuts.

Prior to Cancún the EC and the UnitedStates tried to bridge their differences andagreed on a compromise that combines theharmonising Swiss formula and theconservative Uruguay Round approach. Sincethe blended market access formula was adoptedby the General Council for the Cancúnproposal from the EC-United States proposal,a first natural approach is to assume for theUruguay Round part the numbers that hadbeen proposed by the EC in its own, earlierproposal (36 per cent average reduction with aminimum cut of 15 per cent) and for the Swissformula component, a coefficient of 25 thathad been initially proposed by the UnitedStates. However, it is assumed that the Swissformula would be applied on out-quota boundrather than applied tariffs. Furthermore, it isassumed that the share of tariff lines subject toboth the Uruguay Round and Swiss formula istwice as high as the share of duty-free tarifflines, which gives a distribution of 40-40-20.For two reasons a maximum tariff is notmodelled. First, because it is proposed that

developed countries can alternatively ensureeffective market access through a request-offerprocess that could include tariff rate quotas.Second, it is proposed that developed countriesmay have the additional flexibility for a verylimited number of products to be designatedas of non-trade concern that would not besubject to a maximum tariff.

For developing countries, it is assumedthat 10 per cent of the products can bedenominated as Special Products with areduction of 5 per cent. For the UruguayRound part, it is again assumed that theUruguay Round numbers, i.e., an averagereduction of 24 per cent and a minimum of10 per cent, where the latter is applied to the10 per cent most sensitive products, those withthe highest tariffs, within the group of the 40per cent of tariff lines subject to the UruguayRound formula. The next 40 per cent mostsensitive products are subject to a Swiss formulacut with a coefficient of 50, which wasproposed earlier in the post-Doha negotiationsby the Cairns Group. The remaining 10 percent of tariff lines are reduced to 5 per cent.This approach differs slightly from the draftCancún Text since it does not include theSpecial Products in the Uruguay Round set ofreductions. This approach gives almost thesame results as the approach where the SpecialProduct category is included in the UruguayRound part if this part is extended to 50 percent of tariff l ines. Furthermore, sincedeveloping countries, unlike the developedcountries, do not already have a large numberof tariff lines at zero rates, a reduction to 5 percent would be a strong commitment, andtherefore, it might be considered to be unlikelythat this number of tariff lines is as large asthe zero rate tariff-line number for developedcountries. Therefore, the total number of tariffswas increased, subject to a linear reduction indeveloping countries, to 50 per cent comparedto 40 per cent in developed countries. The tariffcuts are applied to the 36 ATPSMcommodities, broadly the four-digit level,rather than the tariff-line level.

21

The export subsidy and domesticsupport reductions in the draft Cancún text,were assumed to be the same as for theHarbinson proposal. Since average reductionsare proposed, this flexibility causes similarmodelling problems as the flexibil ityconcerning tariff reductions. Here, it isassumed in all four scenarios that the rates arebinding and that countries do not takeadvantage of flexibility to vary the reductionsacross different commodities. This assumptionthus overstates the likely impacts from reform,as is the case in all simulations where subsidiesare not eliminated entirely. Additionally, thedraft Cancún text distinguishes betweensubsidies on products of particular interest todeveloping countries and other products, anelement that is not captured here.

Simulating the proposals isproblematic. In some of them many details arespecified that cannot be captured by ATPSM.For example, the Harbinson proposal includesthe flexibility to reduce export subsidies indifferent steps. Similarly, the proposedexpansion of import quotas depends on currentquotas averaged across several commodities,

whereas within ATPSM the changes are appliedto each of the 36 commodity groups.Furthermore, a possibil ity to preservepreferential schemes and several other specialand differential treatment issues are proposed.Additionally, the approaches provide flexibilityfor countries to self-select tariff and subsidyreductions. Although attempts were made topredict the countries’ behaviour with the rulesdescribed above, countries may in practice actdifferently. Since it is not possible to modelall the elements of the proposals, the simulationresults can only provide a rough picture of thepossible economic effects. Nonetheless, sincethe simulations reflect the majorcharacteristics, the implications drawn from itare likely to be quite robust.

Tariff reductions and averages arecalculated at the 4-digit commodity level ofthe ATPSM database. For developed countries,most sensitive products are defined as thosewith the highest out-quota bound tariff rates,and for developing countries they are theproducts with the highest percentage differencebetween bound and applied rates.

22

Table 3Alternative liberalization scenarios

Label Description

Ambitious A reduction in applied out-quota tariffs according to the Swiss formula t1=(t0*25)/(t0+25), elimination of in-quota tariffs, a 20% expansion of import quotas, elimi-nation of domestic support and export subsidies in all countries and all commodi-ties.

Conservative A reduction in bound out-quota tariffs of the 10% most sensitive products of15%, a 44.1% reduction of remaining products, a 55% reduction in domesticsupport and 45% reduction of export subsidy equivalent in developed countrieswith two-thirds of these cuts in developing countries. No reductions in least-de-veloped countries.

Cancún Developed countries: 40% of tariff lines are subject to the Uruguay Round for-mula, where bound out-quota tariffs of the four most sensitive products are re-duced by 15% and the next 10 most sensitive products by 44.4% (average 36%),40% of tariff lines are subject to the Swiss formula with a coefficient of 25, 20%of tariff lines with the lowest initial bound values are reduced to zero; exportsubsidies are reduced by 80% and domestic support by 60%.

Developing countries: 10% most sensitive tariff lines are reduced by 5% (SpecialProducts), next 40% most sensitive products are subject to Uruguay Round for-mula, where bound out-quota tariffs of the 4 most sensitive products are reducedby 10% and the next 10 most sensitive products by 26.7% (average of last twocategories 24%), 40% of tariff lines are subject to the Swiss formula with a coeffi-cient of 50, while the remaining 10% are reduced to 5%; export subsidies arereduced by 70%, domestic support reduced by 20%.Least-developed countries: no reductions.

Harbinson A reduction in bound out-quota tariffs of 60% where the initial tariff is higherthan 90%, 50% (initial tariff between 15 and 90), or 40% (initial tariff smallerthan 15); an 80% reduction in export subsidies; and a 60% reduction of domesticsupport in developed countries. In developing countries: a 40% reduction wherethe initial tariff are higher than 120%, 35% (initial tariff between 60 and 120),30% (initial tariff between 20 and 60) and 25% (initial tariff smaller than 20); a70% reduction of export subsidies; and a 20% reduction of domestic support. A20% expansion of import quotas in developed and developing countries. Nochanges in least-developed countries.

23

(a) The EU and the United States, Two Major Players

It seems unlikely that there will be anagreement on agriculture unless the UnitedStates and the European Union agree.19 Thus,the impact of any proposal on these countriesis particularly important. The initial proposalsin the Doha negotiations from the two majorplayers were very different in their level ofambition. The United States proposal soughta greater degree of liberalization than that ofthe EU. As a competitive exporter of majoragricultural products, the United States hasmore to gain from an ambitious proposal thanthe EU. The results of the analysis confirm this.

Considering the impact on the tworegions, Table 4 shows that the moreconservative proposals have fewer producerlosses for the EU than the more ambitiousapproaches, while the opposite is the case forthe United States. Furthermore, in the UnitedStates the increase in export revenue is muchhigher under an ambitious liberalizationscenario. The reason is the different structureof the agriculture sectors. Although bothcountries protect their markets and providetrade-distorting subsidies, the United Statesagriculture sector is, overall, more competitivein world markets than that of the EU. Theinitial export revenue in the United States isoverall about twice as high as in the EU, even

V. RESULTS - CONFLICTING INTERESTS AND IMPACTS

Cancún Harbinson Conservative Ambitious

European UnionConsumer Surplus $ 19.2 23.9 13.3 29.2Producer Surplus $ -19.6 -22.4 -13.8 -26.8Government Revenue % 4.1 4.1 3.2 3.9Export Revenue % -6.9 -6 -4.5 -4.3Welfare $ 3.8 5.6 2.8 6.3

United StatesConsumer Surplus $ -4.5 -5 -2.2 -9.9Producer Surplus $ 4.9 5.8 2.2 11.4Government Revenue % 0.1 -0.05 0.3 0.1Export Revenue % 3.2 3.9 1.9 8.6Welfare $ 0.6 0.8 0.3 1.6

Table 4 Impacts of Alternative Proposals on the EU and the United States

(in US$billion)

Source: ATPSM simulations.

19 Under the WTO consensus principle for decision-making all member countries have an effective veto,but the larger trading countries are able to bring political pressure to bear.

24

though subsidies in the EU are much higher.Overall welfare gains for the EU result fromremoving border support to farmers, whichleads to lower production, lower consumerprices and reduced export subsidyexpenditures. That the EC favours aconservative approach reflects the politicalweight attached to transfers to producers.

The EU and the United States cannotpredetermine the outcome. The impacts of thevarious proposals on the other players arediscussed here. It should be kept in mind thatall groups are rather heterogeneous, andalliances may shift according to the particularissue.

(b) Trade Flows

The purpose of trade liberalization isultimately to improve the economic well-being(growth/development) of participants but, inthe immediate effect of removing trade barriersfor agricultural goods will increase imports andexports. However, the estimates show that thechanges resulting form current proposals aremodest in the aggregate. Most of the benefits,in terms of export revenues at least, are mainlyconcentrated in the area of temperate-zoneproducts.

The trade-weighted average increase inexport revenues is only about 5 per cent in theCancun and Harbinson scenarios. Highlyprotected developed countries import more ofalmost all products, but especially temperateproducts cereals, fruits, oilseeds and vegetables.Because world food prices increase, least-developed countries import lower volumes ofmost products in the scenarios where they donot liberalize themselves. However, theirimport bills rise. Figure 5 shows the percentageincrease of total import costs by variouscountry groups.

In developed countries, the highestpercentage increases in imports are for fruitsand vegetables while, in developing countries,the highest increases in imports are invegetables and dairy products and, in least-developed countries, dairy products andcereals. The EU and other developed countrieswith relatively high protection in agriculture,export less of almost every commodity, butmost other countries export more agriculturalproducts. Largest trade increases are in fruitsand vegetables, dairy products and meat.Developing and least-developed countries anddeveloped Cairns group members can increasetheir exports of products within these groupsconsiderably. Exports of tropical beverages like

Figure 5Percentage change in import cost in the Cancun and Harbinson scenario

Source: UNCTAD

0

2

4

6

8

10

12

14

Developed Cairns Group Developing ex. Cairns LDC

Cancun Harbinson

25

tea and coffee, which are important exportproducts for many least-developed countries,are almost unchanged. Oilseeds are in theproduct groups with the highest absoluteincreases in export revenues for developingcountries, but the percentage increases inexports are modest. The total export revenueis slightly reduced or almost unchanged inpercentage terms for developed countries andslightly higher for Cairns Group members, fordeveloping countries that are not Cairns Groupmembers and for the least-developed countries(see Figure 6).

The marked increase of export revenuesin the least-developed countries comeshowever, from a very low base. Since the initialexport revenues of the Cairns Group are almosttwice as high as the initial revenues fordeveloping countries excluding Cairns Groupmembers, the per capita increase in CairnsGroup countries is the highest among all othergroups. Thus, as expected, Cairns Groupmembers’ exporters would strongly benefitfrom trade liberalization as would exportersfrom other developing countries and least-developed countries, albeit to a lesser extent.However, since the blended formula givescountries the flexibility to not reduce applied

tariffs considerably, the global increase ofexport revenues of US$12 billion in theCancun scenario compares with US$19 billionin the Harbinson scenario and US$44 billionin the ambitious scenario.

(c) Price changes and distribution effects

Average price changes provide anindication of the level of ambition of thevarious proposals. A greater price increasemeans a higher level of ambition. The trade-weighted price increases are 2.4 per cent in theCancún simulation and 3.1 per cent in theHarbinson simulation. This compares with aweighted increase of 1.4 per cent in theConservative and 5.8 per cent in the Ambitiousscenario. Comparing world prices acrosscommodities confirms that the more highlyprotected sectors such as dairy products, sheepmeat, sugar, beef and vegetable oils are mostaffected in all scenarios. Price changes are lowerfor tropical than temperate products. In theCancún scenario, the trade-weighted averageincrease for tropical products is 1.3 per centcompared with 3.1 per cent for temperateproducts. About half of the world price increasecan be attributed to the reduction of exportsubsidies in this scenario.

Figure 6Percentage change in export revenue in the Cancun and Harbinson scenario

Source: UNCTAD

-5

0

5

10

15

20

25

30

35

Developed Cairns Group Developing ex. Cairns LDC

Cancun Harbinson

26

The changes in world prices have majordistributional effects, both between countriesand among different economic agents.Exporters gain from increasing world prices,whereas importers are hurt by rising prices.Net-food-exporting countries benefit frommore ambitious liberalization scenarios and thereduction of subsidies. On the other hand, netfood importing developing countries areexpected to face higher prices from a moreambitious agreement. In fact, 83 per cent ofall net food-importing countries, according tothe ATPSM dataset, face net welfare lossesfrom the Cancún scenario. The net foodimporting countries which do not lose fromtrade liberalization are mostly those developedcountries that benefit from reduced subsidies.

The distributional impacts on groupsof consumers, producers and taxpayers differamong various country groups. In developedcountries, consumers gain and producers losefrom reductions in domestic prices (Tables 5and 6). However, this result is stronglyinfluenced by the EU numbers. Due todecreasing domestic prices in the EU there is ahuge consumer surplus that exceeds thenegative producer impacts.

The third component of the totalwelfare, government revenue (Table 7) is inaggregate positive in developed countriesbecause of a reduction in export subsidyexpenditure in the EU. In many developingcountries government revenue is decreasing.

Cancún Harbinson Conservative Ambitious

US$ m US$ m US$ m US$ m

Developed -16 543 -24 403 -12 358 -27 222Developing 17 707 19 204 5 239 14 486Least Developed 1 600 2 230 1 200 -2 625World 2 764 -2 970 -5 918 -15 361Group of 20 11 481 12 097 3 532 8 753Cairns 7 266 8 900 3 789 12 933

Source: ATPSM simulations.

Table 5Consumer surplus impacts from Cancún scenarios

Cancún Harbinson Conservative Ambitious

US$ m US$ m US$ m US$ m

Developed 20 032 34 735 13 452 44 866Developing -14 529 -18 023 -3 681 667Least Developed -1 760 -2 455 -1 295 4 141World 3 743 14 256 8 476 49 674Group of 20 -11 123 -11 558 -2 966 -1 675Cairns -5 954 -7 090 -2 949 -7 962

Source: ATPSM simulations.

Table 6 Producer surplus impacts from Cancún scenarios

27

Global total welfare gains increase withthe level of ambition. Table 8 shows that thegains are about between US$5 billion andUS$22 billion. Using the standard welfaremeasure (consumer and producer surplus plusgovernment revenue) to identify winners andlosers, the analysis shows that there are quite anumber of losers behind the world’s overallwelfare gains. Welfare increases in only 49 outof the 161 countries in the Cancún scenario.In the Harbinson scenario this number is 55and in the Ambitious scenario 77. Thus, thenumber of winners increases with the level ofambition. However, since the amount thatwould have to be put aside to compensate alllosing countries also increases with the levelof ambition, the latter’s losses are greater underthe Ambitious scenario.20

In developing and least-developedcountries, consumers lose as a group andproducers gain because the rise in world priceslifts domestic prices. The government revenuechanges only slightly, between 1 and 2 per cent.In less ambitious scenarios, the total welfarefor developing countries is positive but smalland for least-developed countries it is negative.This is influenced to a large degree by thereduction of export subsidies, which increasesprices for temperate products, and somewhatby a reduction in quota rents on sugar receivedby a significant number of developingcountries. Developing countries gain morefrom more ambitious scenarios that requirethem to make significant reductions in appliedtariffs, giving rise to allocative efficiency gains.

Table 7Government revenue impacts from Cancún scenarios

Cancún Harbinson Conservative Ambitious

US$ m % US$ m % US$ m % US$ m %

Developed 3 730 31 1 652 14 3 972 33 -2 735 - 23Developing -3 014 - 15 - 140 - 1 - 816 - 4 -9 401 - 46Least Developed 19 1 26 2 12 1 - 470 - 30World 735 2 1 538 4 3 167 9 -12 605 - 37Group of 20 - 162 - 2 381 4 66 1 -2 814 - 33Cairns - 84 - 3 217 8 170 6 - 437 - 16

Source: ATPSM simulations.

Cancún Harbinson Conservative Ambitious

US$ m US$ m US$ m US$ m

Developed 7 220 11 983 5 066 14 910Developing 163 1 040 742 5 752Least Developed - 141 - 199 - 83 1 045World 7 242 12 824 5 725 21 707Group of 20 196 920 631 4 264Cairns 1 228 2 027 1 009 4 535

Source: ATPSM simulations.

Table 8Welfare impacts from Cancún scenarios

20 A total of US$1.5 billion would need to be put aside to compensate all the losing countries in the Cancunscenario. In the Compromise and Ambitious scenario it is US$1.7 and US$2.4 billion, respectively.

28

Least-developed countries, with a higherproportion of net food importing countries,gain only if they liberalize themselves (underthis comparative static analysis, with noaccount being taken of short-term adjustmentcosts). However, as in developing countries,producers gain more from the more ambitiousscenario. Table 9 shows for all scenarios thatthe least-developed countries have the highestpercentage increase in export revenues,although from a low base.

For the Group of 20 developingcountries that formed a coalition andsupported a counter-proposal to the EC-United States joint proposal and for the CairnsGroup, the qualitative results are similar to theresults for the group of developing countries.Producers gain slightly more than consumerslose and overall they gain more from moreambitious scenarios.

(d) Wider Implications of Welfare Estimates

The merit of each proposal thusdepends on whether policymakers emphasizethe gains to producers, exporters, consumersor taxpayers. The estimated welfare impacts arethe sum of these three effects. Transfersbetween these groups are equally weighted, so

the net welfare effect is positive if the estimatedbenefits to consumers and taxpayers exceed thelosses to producers, as is often the casefollowing the reduction of tariffs. A problemwith such welfare estimates is the presumptionof equal weights. The existence of policiesfavouring producers (in developed countries)or consumers (as used to be the case in somedeveloping countries) is evidence thatpolicymakers favour one group over another.From the observed policies in the currentround of negotiations it seems thatgovernments tend to attach greater weight totheir producers rather than consumers. Thereare often internal political reasons for this. Indeveloping countries, the argument goes thatin order to achieve development needs, thepoor rural population, where a large sharedepends on the agricultural sector, must besupported.21 The negotiation strategy of mostdeveloping countries, namely to demandimproved access to developed countries’markets and the elimination of trade-distortingsubsidies and to protect their own markets, isa strategy aimed at maximising the producersurplus in these countries, at least in the shortterm. However, it has negative impacts onconsumers as a result of higher domestic foodprices.

Cancún Harbinson Conservative Ambitious

US$ m % US$ m % US$ m % US$ m %

Developed - 938 - 1 1 189 1 185 0 10 475 11Developing 12 272 13 16 557 17 9 815 10 31 106 32Least Developed 904 22 1 254 30 859 21 2 109 51World 12 237 6 19 001 10 10 859 5 43 690 22Group of 20 7 861 15 10 951 21 6 489 12 20 594 40Cairns 6 415 8 8 297 10 4 967 6 15 805 20

Source: ATPSM simulations.

Table 9Export revenue impacts from Cancún scenarios

21 Support for agriculture runs counter to the import-substitution industrialization policies, which favourindustry over agriculture as a development strategy.

29

If producers happen to be the poorestmembers of society and the targets ofgovernment support, it may be that reformsthat reverse these policies are detrimental topoverty reduction programmes. Assessment ofthe impact of alternative proposals should betaken in the light of such considerations. Theeconomic analysis merely points out thepotential impacts of reforms. Social, politicaland environmental considerations should beassessed. The modelling has little to say onpoverty reduction for example, because it is notclear a priori whether it is producers orconsumers who constitute these groups andthat is a judgement for the governmentconcerned. Furthermore, in many developingcountries a large proportion of the populationare subsistence farmers and this adds to thecomplexity. Rising world prices tend to benefitproducers, and hence many of the rural poor,although this is not the case if tariffs arereduced by more than the rise in world prices.Price rises are greatest for the temperateproducts, such as livestock and grains and,since developing countries tend to be importersof these products, such increases are to the

detriment of consumers. The analysis indicatesthe likely impacts on consumers and producers,but the desirability of these impacts needed tobe assessed for each country depending on itsindividual circumstances and objectives. Thus,rather than looking at total welfare, policymakers and negotiators may prefer to look athow consumers and producers are affected, andbring this in relation with the poverty structureand possible distribution effects in their owncountry.

(e) Special and Differential Treatment

Developing countries, and especiallythe Group of 20, have been seeking highercommitments for developed countries andlower commitments for themselves. Mostdeveloped countries do not support a highdegree of special and differential treatment fordeveloping countries, although there isprovision for some differentiation specified inthe Doha Declaration. Both interests can bebetter understood by looking at changes inproducer surpluses following the liberalizationof various magnitudes. Figures 7 and 8 show

Figure 7Special and differential treatment: Impact on developing countries

Figure 8Special and differential treatment: Impact on developed countries

-60

-40

-20

0

20

40

60

1 2 3 4

Increasing level of ambition in developing countries

Producer Surplus

Consumer Surplus

Welfare

-10

-5

0

5

10

15

1 2 3 4

Increasing level of ambition in developing countries

Producer Surplus

Consumer Surplus

Welfare

30

the results when the reductions in thedeveloped countries are fixed at the UruguayRound level and the reductions in thedeveloping countries are increased up to thesame values.22 The producer surplus indeveloping countries (here including least-developed-countries) decreases when the levelof ambition is increasing in developingcountries (Figure 7), which can be interpretedas a decreasing degree of special and differentialtreatment because the lower reductioncommitment for developing countries isprogressively eliminated so that in scenario 4both country groups have the some cuts.Producers of protected products are worse offas liberalization erodes domestic prices. Theopposite holds for the producer surplus indeveloped countries (Figure 8) and also for theconsumer surplus in developing countries(Figure 7).

Figures 9 and 10 show the results whenboth groups of countries liberalise andtherefore bring the market access and allocativeefficiency effects together. For this, the fourscenarios defined in Table 3 were used.Interestingly, the producer surplus indeveloping countries is inversely U-shaped. Asthe level of ambition increases, the negativedomestic prices effect from reducing tariffsstarts to dominate the positive world priceeffect of developed countries reducing theirtariffs. This lag occurs because of the gapbetween bound and applied tariffs indeveloping countries. However, as domesticprices fall in developing countries consumersare better off and overall welfare is increasing.In developed countries, all three measures arestrictly monotonically increasing or decreasingbecause there is no gap between bound andapplied rates.

22 Uruguay Round approach means a reduction in bound tariff rates by 36 per cent in developed and 24 percent in developing countries, a reduction in export subsidies by 36 (24) per cent and a reduction in domesticsupport by 21 (14) per cent.

Figure 9Impact of liberalization on developing countries

Figure 10Impact of liberalization on developed countries

-30000

-20000

-10000

0

10000

20000

30000

1 2 3 4

Conservative, Cancun, Harbinson, Ambitious

Producer Surplus

Consumer Surplus

Welfare

-40000

-20000

0

20000

40000

60000

1 2 3 4

Conservative, Cancun, Harbinson, Ambitious

Producer Surplus

Consumer Surplus

Welfare

31

(f ) Declining Ambitions

As the negotiations have progressed theproposals appear to be converging. The annualestimated global welfare gains are about US$7billion in the Cancún scenario and US$13billion in the Harbinson scenario comparedwith US$5 billion and US$22 billion in theConservative and the Ambitious scenariosrespectively. However, the latest Cancúnproposal is near the bottom of this range. Thus,the flexibil ity given to developed anddeveloping countries by the Cancún marketaccess formula waters down the welfare gains.Even though the formula contains one tariff-harmonizing Swiss formula component withrather ambitious coefficients of 25 fordeveloped and 50 for developing countries,implementing the Uruguay Round numbersfor the linear-cut portion of the formula givesoverall welfare effects that are not much higherthan a continuation of the Uruguay Roundapproach, along the lines of the initial EUproposal. Assuming the same smaller exportsubsidy and domestic support reductions in theCancún scenario as in the Harbinson scenario(that is, reducing export subsides by 45 per centand domestic support by 55 per cent) furtherreduces the global welfare gains. However,since developing countries have in generalhigher bound tariff rates and since their tariffsdo not in general vary as much as developedcountries’ tariffs, the Swiss formula part in theCancún formula would require relativelyhigher reductions in developing countries. Thisdepends on the coefficients that would be

chosen. The Group of 20 proposal and thedraft Cancún text, first revision, proposed fordeveloping countries the opportunity to applythe Uruguay Round reductions to all tariffs.This increases the degree of special anddifferential treatment with the above shownconsequences, namely higher producer surplusand a lower consumer surplus and welfare indeveloping countries.23

(g) Sensitivity Analysis

The declining ambition of the Cancunscenario in comparison to the Harbinsonscenario depends on the assumed numbers forthe blended formula. In order to verify theresults a further simulation with a moreambitious blended formula wasimplemented.24 With these revised numbersboth developed and, above all developingcountries liberalize their own markets morethan under the “standard” Cancún scenario.Least-developed countries do not liberalize andchanges in domestic support and exportsubsidies are as in the standard Cancúnscenario. The results presented in table 10should be compared with table 8.

Table 10 shows the results from theambitious Cancún scenario. The overall levelof ambition lays between the standard Cancúnand the Harbinson scenario. However, sincedeveloping countries have to makeconsiderable tariff reductions consumer surpluslosses are lower than in both the standardCancún and the Harbinson scenarios. The

23 For simulation results see Peters and Vanzetti (2003).

24 For developed countries the average cut in the Uruguay Round formula component is 50 per cent andthe minimum 25 per cent as it was over all bands in the Harbinson proposal of March 2003. For developing countriesthe average is 32.5 per cent and the minimum 15 per cent for non-Special Products, as in the Harbinson draft.Special Products are reduced by 10 per cent, which is the average reduction for these products proposed by Harbinson.The coefficients in the Swiss formula part remain at 25 and 50, respectively. The distribution of tariff lines amongthe three components in the blended formula is a third each in developed countries. In the first third the 33 per centmost sensitive products are reduced by the minimum reduction of 25 per cent. In developing countries the distributionis: 8 per cent of tariff lines Sensitive Products, 31 per cent Uruguay Round formula, where within this category 27per cent are reduced by the minimum of 15 per cent, 31 per cent Swiss formula and 31 per cent reduced to 5 percent.

32

balance of higher world prices increases andlower domestic tariffs leaves producers indeveloping countries no worse off than thestandard Cancún scenario. Competitiveagricultural producers, such as in the UnitedStates and the Cairns Group countries, arebetter off in the ambitious Cancún scenariothan in both the Harbinson and the standardCancún scenarios. Producers in least-developedcountries are also better off from higher worldprices but consumers in these countries are

worse off. The global welfare gains of US$10billion are between the standard Cancún andthe Harbinson scenarios. This is also true fordeveloped countries, whereas developingcountries are better off and least-developedcountries are worse off, with the ambitiousCancún scenario compared to the standard andthe Harbinson scenarios. In summary, theanalysis indicates that the impacts are notparticularly sensitive to reasonable values ofassumed tariff reductions.

Table 10Results from an ambitious Cancún scenario

Consumer Producersurplus surplus Welfare

US$ m US$ m US$ m

Developed 24 379 -17 289 8 985Developing -10 758 17 388 1 722Least Developed -2 627 2 370 - 231World 10 993 2 469 10 476Group of 20 -8 462 11 144 1 360Cairns -6 849 9 488 2 164European Union 21 497 -20 741 4 430United States -5 646 6 488 831

Source: ATPSM simulation.

33

The road to Cancún has proved longand winding, and in the negotiations onagriculture WTO Members were not able toagree on the road ahead. Therefore, there isno agreement on modalities for furthercommitments (at the time of writing –November 2003). For the 5th Ministerialmeeting at Cancún, the General Councilprepared a draft framework for establishingmodalities in agriculture. Since the draft didnot contain specific targets, it is difficult tojudge the text and to assess its economic effects.In this paper we assume possible numbersdrawn from previous proposals including onefrom the Chair of the Committee onAgriculture. In an effort to reach a negotiatedoutcome, proposals have been modified byincreasing the flexibility in the formulae.

The missing numbers in the EC-United States proposal make it difficult toassess the level of ambition but it seems morein line with the limited liberalisation goal ofthe EU. The analysis shows that if the EU wereto put a sufficiently high weight on conservingproducer surplus, they would tend to favour amore conservative approach. In the UnitedStates, producers gain more from an ambitiousapproach and, since agricultural tariffs arealready rather low in the United States, thiscountry has little to lose and much to gain fromimposing this formula on highly protectedcountries. However, for two reasons it isunderstandable that the EU and the US agreedon a flexible approach. First, the fallbackposition is either the status quo or, given theDoha Declaration, a limited liberalizationscenario such as a continuation of the flexibleUruguay Round type of approach. Thisincreases the EU’s bargaining power. Second,both countries are better off with a multilateral

agreement than going it alone. Thus, these twocountries want an agreement, not only onagriculture, but also in the other negotiatingareas of the Doha Round Single Undertakingnot analysed here, including non-agriculturalmarket access and services. The resultingcompromise between the EU and the UnitedStates is at the expense of net-food exportingcountries, producers in developing countriesand global welfare gains. A result that wasshown previously for other liberalizationscenarios (see for example Vanzetti and Peters,2003) and also using general equilibriummodels is that developing countries gain – as agroup – more from a more ambitiousliberalisation scenario because positiveallocative efficiency effects start to outweighthe negative terms of trade effects. However,least-developed countries and net-food-importing developing countries suffer a welfareloss if tariffs and subsidies are ambitiouslyreduced. Tariff cuts reduce quota rents tocountries with preferential access and,reductions in exports subsidies raise prices oftemperate products that are imported ratherthan exported by many of these countries.

Since there are efficiency gains fromliberalization, finding an agreement may be aquestion of compensating the losers. This istrue within a country and among countrygroups. Therefore, one possible solution to thecurrent impasse is for developed countries toswitch support from output-related to direct-income support. By decoupling domesticsupport from production levels, the EU CAPreform is a step in the right direction.However, it is unlikely that this is a big enoughstep to encourage other countries to follow. Ifproduction levels were to be reducedsufficiently for export subsidies to be

VI. CONCLUSIONS

34

eliminated, a successful agreement would bemuch more likely. On a multilateral level, theWorld Bank and the IMF could play a role incompensating losing developing countries,although this is beyond the bounds of theWTO system. The EU could also compensateACP countries that lose from the erosion ofpreferences, just as it compensates its ownproducers for reductions in output relatedsupport.

Limitations to the analysis should bekept in mind. The conclusions are based onthe simulation of several proposals for anagreement on agriculture. However, not allelements of the proposals could be capturedadequately. To simulate the draft CancúnMinisterial text, assumptions about thebracketed numbers had to be made. Thus, theresults have to be interpreted with care.

There are further limitations of theanalysis. One is the lack of knowledge of thedistribution of quota rents. A second is theassumption in the model that domestic pricesare determined by the higher out-quota tariff,in spite of the number of observed unfilledimport quotas. This may lead to anoverestimation of the impacts. Limitations thatapply to all models of this nature are thatestimated annual gains are static rather thandynamic and, that adjustment costs are nottaken into account. A more ambitious scenariocauses higher adjustment costs, which are likely

to be higher in developing countries than indeveloped countries. Finally, data quality is anissue, especially when considering the resultsfor a particular country or sector.

In spite of these limitations, the resultsprovide a useful indication of the possibleimpacts of an agreement on a framework likethe draft Cancún Ministerial text. At this stage,it seems likely that the modalities will not bevery ambitious, and potentially importantwelfare gains will be forgone. However, ifdeveloping countries push for a moreambitious round, then they would likely comeunder pressures to undertake moreliberalization than they seem willing or ableto consider at the present stage of theirdevelopment, despite potential longer termgains. Least-developed countries and net-food-importing countries should be aware of thepossible negative impacts that they may faceas a result of rising food prices, although thismay be advantageous to their producers, whichinclude some of the poorest sections of society.Finally, developing countries should note theflexibility that a Uruguay Round type approachgives developed countries. Some flexibility isnecessary to forge an agreement, but too muchmeans there is little progress in promoting tradeand development. In pointing out theseconflicts, this analysis may make a smallcontribution to getting the negotiations backon the road.

35

AMAD database http://www.amad.org.

Braun, J. von, P. Wobst and U. Grote (2002), “ ‘Development Box’ and Special and DifferentialTreatment for Food Security of Developing Countries: Potentials, Limitations andImplementation Issues”, ZEF – Discussion Papers on Development Policy, Bonn.

de Gorter, H. (1999), “Market access, export subsidies and domestic support measures: issues andsuggestions for new rules”, contributed paper at The Conference on Agriculture and theNew Trade Agenda in the WTO 2000 Negotiations, sponsored by the World Bank, October1-2.

European Commission (2002), “WTO and agriculture: European Commission proposes moremarket opening, less trade-distorting support and a radically better deal for developingcountries”, Brussels. FAOSTAT database http://apps.fao.org.

European Commission and United States (2003), “Joint EC-US Paper, Agriculture”, WTOdocument JOB(03)/157, Geneva.

OECD (2001a), Agricultural policies in OECD: Estimates of support for agriculture,http://www.oecd.org/xls/M00022000/M00022536.xls.

Peters, R. and D. Vanzetti (2003), ‘Searching for a Compromise, A Game Theoretic Approach tothe WTO Negotiation’, European Trade Study Group Conference Paper, Madrid, 15-17September, http://www.etsg.org/etsg2003/ etsg2003_programme_mk1.htm.

Roberts, I. and F. Jotzo (2001), “2002 US Farm Bill: Support and Agricultural Trade” ABAREResearch Report 01.13, Canberra.

Ruffer, T. and P. Vergano (2003), An Agricultural Safeguard Mechanism for Developing Countries,Oxford Policy Management.

Skully, D. (2001), “Liberalising tariff-rate quotas”, chp. 3 in ERS/USDA, Agricultural Policy Reform– The Road Ahead, pp. 59-67, Washington DC.

UNCTAD TRAINS database http://www.unctad.org/trains/index.htm.

US Department of Agriculture (2002), U.S. Proposal for Global Agricultural Trade Reform, ActualLanguage of U.S. WTO Agriculture, Proposal Presented in Geneva, Washington, http://www.fas.usda.gov/itp/wto/actual.htm.

REFERENCES

36

Vanzetti, D. and R. Peters (2003), The Good, the Bad and the Ugly, Three Proposals for AgriculturalPolicy Reform. In: S. Inama and N. Xuto, eds. Processing Towards the Doha DevelopmentAgenda, ADB and UNCTAD.

Vanzetti, D. and R. Sharma (2002), “Impact of Agricultural Trade Liberalisation on DevelopingCountries: Results of the ATPSM Partial Equilibrium Model”, invited paper at theInternational Agricultural Trade Research Consortium summer symposium on “The DevelopingCountries, Agricultural Trade and the WTO”, Whistler Valley, British Columbia, Canada,16-17 June.

World Bank (2001), “Global Economic Prospects and the Developing Countries 2001”, WashingtonD.C. http://www.worldbank.org/prospects/gep2001/.

WTO (2000), Tariff and other Quotas. G/AG/NG/S/7, Geneva.

WTO Committee on Agriculture, Special Session (2003). Negotiations on Agriculture, first draftof Modalities for the further Commitments, Revision, Geneva.

WTO General Council (2003), Preparations for the Fifth Session of the Ministerial Conference,Draft Cancún Ministerial Text, Second Revision. JOB(03)/150/Rev.2, Geneva.

37

Developed Developing Developing (cont.) Least developedAustralia Albania Latvia AfghanistanBrunei Algeria Lebanon AngolaCanada Argentina Libya BangladeshChina, Hong Kong Armenia Lithuania BeninChina, Taiwan Azerbaijan Macedonia Burkina FasoCyprus Bahamas Madagascar BurundiEuropean Union Barbados Malawi Central African Rep.French Polynesia Belarus Malaysia CambodiaIceland Belize Malta Cape VerdeIsrael Bolivia Mauritius ComorosJapan Bosnia Herzegovina Mexico CongoKuwait Botswana Moldova Congo Dem. Rep.Macao Brazil Mongolia DjiboutiNeth. Antilles Bulgaria Morocco EritreaNew Zealand Cameroon Namibia EthiopiaNorway Chad Nicaragua GambiaSlovenia Chile Nigeria GuineaSwitzerland China Pakistan Guinea BissauU. A. Emirates Colombia Panama HaitiUnited States Costa Rica Papua New Guinea Lao PDR

Croatia Paraguay LesothoCuba Peru LiberiaCzech Rep. Philippines MaldivesDominica Poland MaliDominican Rep. Romania MauritaniaEcuador Russia MozambiqueEgypt Saudi Arabia MyanmarEl Salvador Seychelles NepalEstonia Slovakia NigerFiji South Africa RwandaGabon Sri Lanka Sao TomeGeorgia St. Lucia SenegalGhana St. Vincent Sierra LeoneGrenada Suriname Solomon IslandsGuatemala Swaziland SomaliaGuyana Syria TanzaniaHonduras Tajikistan TogoHungary Thailand UgandaIndia Trinidad Tobago VanuatuIndonesia Tunisia YemenIran Turkey ZambiaIraq TurkmenistanIvory Coast UkraineJamaica UruguayJordan UzbekistanKazakhstan VenezuelaKenya Viet NamKorea DPR YugoslaviaKorea Rep. ZimbabweKyrgyzstan

Appendix

Table A1Country coverage in ATPSM

Note: Among the 49 least-developed countries, Bhutan, Chad, Equatorial Guinea, Kiribati, Madagascar, Malawi,Samoa, Somalia, Sudan, Togo and Tuvalu are not included in the model.

38

The ATPSM modelling frameworkATPSM is a deterministic, comparative static, partial equilibrium model. This means that thereare no stochastic shocks or other uncertainties, and there is no specific time dimension to theimplementation of the policy measures or to the maturing of their economic effects. The comparativestatic nature of the model doesn’t imply that the policies take effect instantaneously. Rather, it iscomparing two states at a similar point in time, one with the policy change, the other without.Finally, whereas the model aims at estimating far-reaching details of the agricultural economy, itdoes not deal with the repercussions of barrier reductions on other parts of the national economy.Thus, neither effects on the government budget (except for tariff revenues and subsidies to exportsand domestic production) nor on the industrial and service parts of the economy or the labourmarket are subject to analysis. Simplifying the model in these respects allows for a detailedspecifications of policies in a large number of countries for numerous commodities.

Equation systemAfter a trade policy change, like a change in tariffs, export subsidies and/or domestic support, isspecified, the model calculates the new equilibrium. The equation system for all countries has fourequations:

, , , , ,1

ˆ ˆ ˆ1) ;J

i r i i r C i j r Ci jjj i

D P Pη η=≠

= +∑

Meat Vegetables01100 Bovine meat 05420 Pulses01210 Sheepmeat 05480 Roots, tubers01220 Pigmeat 05440 Tomatoes01230 Poultry FruitDairy products 05700 Apples & pears02212 Milk, fresh 05710 Citrus fruits02222 Milk, conc. 05730 Bananas02300 Butter 05790 Other tropical fruits02400 Cheese BeveragesCereals 07110 Coffee green bags04100 Wheat 07120 Coffee roasted04400 Maize 07131 Coffee extracts04530 Sorghum 07210 Cocoa beans04300 Barley 07240 Cocoa butter04200 Rice 07220 Cocoa powderSugar 07300 Chocolate06100 Sugar 07410 TeaOils Tobacco and cotton22100 Oil seeds 12100 Tobacco leaves42000 Vegetable oils 12210 Cigars

12220 Cigarettes12230 Other tobacco - mfr.26300 Cotton linters

Table A2Commodities in ATPSM

39

, , , , ,1

ˆ ˆ ˆ2) ;J

i r i i r P i j r Pi jjj i

S P Pε ε=≠

= +∑

, , , , , ,ˆˆ3) ;i r i r i r i r i r i rX M D D S S∆ = ∆ − +

, , , ,

,

ˆ4) , where ;1 1 1

new init new m di r i r i r i r y

new init new d m y i r

A A A PM D D D AA A A P

σ

αα

∆ = − − = + + +

where: D, S, X, and M denote demand, supply, exports and imports, respectively;^ denotes relative changes and ∆ absolute changes;PC denotes consumer price, PP producer price, Pd price for domestic supply, Pm price for imports (see below);ε denotes supply elasticity, η denotes demand elasticity;i and j are commodities indexes, r is a country index;y=init indicates initial values and y=new indicates values after the policy changes;σ denotes the Armington elasticity between imports and domestically produced goods.

Equations 1 and 2 specify that the new demand and supply are determined by the pricechanges, trade policy changes and the corresponding elasticities and cross-price elasticities. Equation4 ensures that the relation of imports and domestic supply are determined by the price ratio ofdomestic supply and imports.

m d

d m

PMD M P

σαα

= −

Equation 3 clears the market, so that production plus imports equals domestic consumptionand exports.

These equations can be transformed into matrix notation and the equation system solvedarithmetically for world prices by matrix inversion. A market equilibrium requires that, globally,the sum of the change in exports equals the total change in imports for each commodity:

5);0)(

1=∆−∆∑

=

N

nnn MX

PricesDomestic prices are all functions of the world market price and border protection or special domesticsupport measures. Thus, domestic price data is not required and transaction costs (such as wholesaleand retail margins) are not taken into account. All protection measures are expressed in tariff rateequivalents.

40

The relationship between world and domestic prices is complicated by the existence oftwo-way trade of the one (aggregated) good. To accommodate heterogeneous goods with oneprice, the approach taken here is to estimate a composite price and a composite tariff for determiningthe domestic consumption and production price, respectively. To derive a composite price, productsare divided into three groups: imports, exports and, production supplied to the domestic market(Sd).

First, a domestic market price wedge (td) is computed as the weighted average of two tariffs,the export tariff (tx) and import tariff (tm), where the weights are exports (X) and imports (M): td =(X tx + M tm)/(M + X).

The price for domestic supply is Pd=Pw(1+td), where Pw is the world price, and the price for

imports is Pm=Pw(1+tm) . Then, a composite consumer price is computed as PC =

( )1 1/1 1m m d dP P

ρσ σ σ σα α−− −+ . The producer price wedge is computed as the weighted average of the

export tariff (tx) and the domestic market price wedge (td), where the weights are exports (X) anddomestic supply (Sd) plus the domestic support tariff (tp): ts = (X tx + Sd td) / S + tp. The producerprice is Ps=Pw(1+t

s). The calculations of consumer and producer prices are applied both to the

baseline and the final tariffs.

A feature of this structure is that if there are no exports, domestic producer prices aredetermined by the tariff plus the domestic support. If there are no imports the export subsidyeffectively determines the producer price. Finally, if there is two-way trade the share of totalproduction or consumption influences the importance of each tariff.

The need for a composite price such as this is the requirement for one price with essentiallytwo goods. The heterogeneous nature of imports and exports also requires a means of specifyingthe volume of either imports or exports. In this model imports are specified so that the relation ofimports and domestic supply are determined by the price ratio of domestic supply and imports(equation 4). This is the so-called Armington specification. Exports are determined as the residualof production, consumption and imports.

Trade revenueOnce changes in world prices and hence domestic prices are determined from the model solution,volume changes can be derived from equations 1-4. Given the volume responses DX, DM, DS,and DD, the trade revenue and welfare effects can be computed. The trade revenue effect of thepolicy changes is computed for each country and each commodity from:

( ) ( ) ( )[ ] ( )MXPMMXXPPR www −−∆+−∆+∆+=∆ 1

Secondly, there is a change in quota rents DU, which generates a further trade revenueeffect (in each country and for each commodity):

( )[ ] UXXXUUR −∆+∆+=∆ 2.

The total trade revenue effect is the sum of these components: DR = DR1 + DR2 .

41

WelfareThe welfare change has three components. The first two are the changes in producer surplus (DPS)and consumer surplus (DCS). These changes depend on the domestic market price changes andtheir own price domestic demand and supply volume responses. The change in producer surplus isalso dependant on the change in quota rent. For each country and commodity:

( ) ( )20.5 ; 0.5 ;p d c dPS P S S R CS P D D∆ = ∆ + ∆ + ∆ ∆ = −∆ + ∆

The third part is the change in net government revenue (DNGR), consisting of change intariff revenue, change in export subsidy expenditure and, change in domestic support expenditure.For each country and commodity:

( )( ) ( ) ( )[ ] ( )

( )[ ] ( )( )[ ]4444 34444 214444 34444 21

44444444 344444444 214444 34444 21

enditureportdomesticinChange

ddd

endituresubsidyortinChange

xxx

revenuequotaofoutinChange

oo

revenuequotawithininChange

www

StSSttXtXXtt

QMtQQMMttQtQQtt

DSESTRNGR

expsupexpexp

0

)(

()

−∆+∆+−−∆+∆+−

−−∆+−∆+∆++−∆+∆+=

∆−∆−∆=∆

−−−

The sum is the total welfare effect: DW = DPS + DCS + DNGR..

APTSM is able to estimate the economic effects of changes in within-quota and out-of-quota tariffs, import, export and production quotas, export subsidies and, domestic support onproduction, consumption, prices, trade flows, trade revenues, quota rents, producer surplus andwelfare. The assumptions of filled quotas made here imply that changes in within-quota tariffs andimport quotas will not have price and quantity effects, as these instruments are not binding.(However, they do change the distribution of rents.)

DataVolume data are from 2000 and are compiled from FAO supply utilization accounts.25 The year2000 represents the base year for the model. Most of the price data is also from FAOSTAT.Parameters on elasticities and feedshares are from FAO’s World Food Model. These are based on atrawling of the literature and are not econometrically estimated specifically for the model. Some ofthe elasticities were modified by the authors to adjust them to the special ATPSM features. Within-quota tariffs, out-of-quota tariffs and global quotas, notified to the WTO, are obtained from theAMAD database where available or directly from the WTO and aggregated to the ATPSMcommodity level.26 Export subsidy and setaside data are notified to the WTO. Bilateral trade flowdata relate to 1995 and are from the UNCTAD Comtrade database. These are used to allocateglobal quotas to individual countries. The UNCTAD TRAINS database is the source of informationon applied tariffs.

25 This is a revision of data used in previous applications of the model (Vanzetti and Sharma, 2002) andresults in a substantial downward revision of welfare estimates.

26 AMAD database http://www.amad.org

42

43

UNCTAD Study Series on

POLICY ISSUES IN INTERNATIONAL TRADEAND COMMODITIES

No. 1 Erich Supper, Is there effectively a level playing field for developing countryexports?, 2001, 138 p. Sales No. E.00.II.D.22.

No. 2 Arvind Panagariya, E-commerce, WTO and developing countries, 2000, 24 p.Sales No. E.00.II.D.23.

No. 3 Joseph Francois, Assessing the results of general equilibrium studies of multi-lateral trade negotiations, 2000, 26 p. Sales No. E.00.II.D.24.

No. 4 John Whalley, What can the developing countries infer from the UruguayRound models for future negotiations?, 2000, 29 p. Sales No. E.00.II.D.25.

No. 5 Susan Teltscher, Tariffs, taxes and electronic commerce: Revenue implicationsfor developing countries, 2000, 57 p. Sales No. E.00.II.D.36.

No. 6 Bijit Bora, Peter J. Lloyd, Mari Pangestu, Industrial policy and the WTO, 2000,47 p. Sales No. E.00.II.D.26.

No. 7 Emilio J. Medina-Smith, Is the export-led growth hypothesis valid for develop-ing countries? A case study of Costa Rica, 2001, 49 p. Sales No. E.01.II.D.8.

No. 8 Christopher Findlay, Service sector reform and development strategies: Issuesand research priorities, 2001, 24 p. Sales No. E.01.II.D.7.

No. 9 Inge Nora Neufeld, Anti-dumping and countervailing procedures – Use orabuse? Implications for developing countries, 2001, 33 p. Sales No. E.01.II.D.6.

No. 10 Robert Scollay, Regional trade agreements and developing countries: The caseof the Pacific Islands’ proposed free trade agreement, 2001, 45 p. Sales No.E.01.II.D.16.

No. 11 Robert Scollay and John Gilbert, An integrated approach to agricultural tradeand development issues: Exploring the welfare and distribution issues, 2001,43 p. Sales No. E.01.II.D.15.

No. 12 Marc Bacchetta and Bijit Bora, Post-Uruguay round market access barriersfor industrial products, 2001, 50 p. Sales No. E.01.II.D.23.

No. 13 Bijit Bora and Inge Nora Neufeld, Tariffs and the East Asian financial crisis,2001, 30 p. Sales No. E.01.II.D.27.

44

No. 14 Bijit Bora, Lucian Cernat, Alessandro Turrini, Duty and Quota-Free Access forLDCs: Further Evidence from CGE Modelling, 2002, 130 p. Sales No.E.01.II.D.22.

No. 15 Bijit Bora, John Gilbert, Robert Scollay, Assessing regional trading arrange-ments in the Asia-Pacific, 2001, 29 p. Sales No. E.01.II.D.21.

No. 16 Lucian Cernat, Assessing regional trade arrangements: Are South-South RTAsmore trade diverting?, 2001, 24 p. Sales No. E.01.II.D.32.

No. 17 Bijit Bora, Trade related investment measures and the WTO: 1995-2001, 2002.

No. 18 Bijit Bora, Aki Kuwahara, Sam Laird, Quantification of non-tariff measures,2002, 42 p. Sales No. E.02.II.D.8.

No. 19 Greg McGuire, Trade in services – Market access opportunities and the ben-efits of liberalization for developing economies, 2002, 45 p. Sales No. E.02.II.D.9.

No. 20 Alessandro Turrini, International trade and labour market performance: majorfindings and open questions, 2002, 30 p. Sales No. E.02.II.D.10.

No. 21 Lucian Cernat, Assessing south-south regional integration: same issues, manymetrics, 2003, 32 p. Sales No. E.02.II.D.11.

No. 22 Kym Anderson, Agriculture, trade reform and poverty reduction: implicationsfor Sub-Saharan Africa, 2004, 30 p. Sales No. E.04.II.D.5.

No. 23 Ralf Peters and David Vanzetti, Shifting sands: searching for a compromise inthe WTO negotiations on agriculture, 2004, 46 p. Sales No. E.04.II.D.4.

45

United Nations publications may be obtained from bookstores and distributors throughout the world.Please consult your bookstore or write to:

United Nations Publications

All orders from North America, Latin America, the Caribbean and Asia and the Pacific shouldbe sent to:

United Nations PublicationsRoom DC2-853, 2 UN PlazaNew York, NY 10017, USATelephone: (212) 963-8302, Toll Free 1-800-253-9646 (North America only)Fax: (212) 963-3489E-mail: publications@un.org

Customers in Europe, Africa and the Middle East should send their orders to:Section des Ventes et CommercialisationBureau E-4, CH-1211Geneva 10, SwitzerlandTelephone: 41 (22) 917-2613/2614Fax: 41 (22) 917-0027E-mail: unpubli@unog.ch

46

QUESTIONNAIRE

UNCTAD Study Series onPOLICY ISSUES IN INTERNATIONAL TRADE

AND COMMODITIES(Study series no. 23: Shifting Sands: Searching for a Compromise

in the WTO Negotiations on Agriculture)

Readership SurveySince 1999, the Trade Analysis Branch of the Division on International Trade in Goods and

Services, and Commodities of UNCTAD has been carrying out policy-oriented analytical workaimed at improving the understanding of current and emerging issues in international trade ofconcern to developing countries. In order to improve the quality of the work of the Branch, itwould be useful to receive the views of readers on this and other similar publications. It wouldtherefore be greatly appreciated if you could complete the following questionnaire and return to:

Jenifer Tacardon-MercadoTAB/DITC, Rm. E-8076

United Nations Conference on Trade and DevelopmentPalais des Nations

CH-1211 Geneva 10, Switzerland

1. Name and address of respondent (optional):

2. Which of the following describes your area of work?

Government Public enterprisePrivate enterprise institution Academic or researchInternational organization MediaNot-for-profit organization Other (specify) _________________

3. In which country do you work? _________________________________________

4. Did you find this publication Very useful Of some use Little useto your work?

5. What is your assessment of the contents of this publication? Excellent Good Adequate Poor

6. Other comments: