OptionsandStrategiesto C-apture the Benefitsfor the Middle 'mt eASBSlXJDYFROM TUNISLA Economic and Social Commission for Western Asia (ESCWA) Distr. GENERAL EIESCWAlAGRJ200114 5 October 2001 0RIGIN.AL: ENGLISH ECONOMIC AND SOCIAL COMMISSION FOR WESTERN ASIA AGRICULTURAL TRADE AND THE NEW TRADE AGENDA: OPTIONS AND STRATEGIES TO CAPTURE THE BENEFITS FOR THE MIDDLE EAST CASE STUDY FROM TUNISIA United Nations New York, 2001 Copyright 2001, International Bank for Reconstruction and Developmenflhe World Bank and ESCWA. This material may be copied for research, educational or scholarly purposes only in member countries of the World Bank andlor ESCWA. All materials are subject to revision. The views and interpretations in this document are those of the individual author(s) andlor trainers and should not be attributed to the World Bank Institute, the World Bank or ESCWA. Preface This study is the result of fruitful cooperation between the United Nations Economic and Social Commission for Western Asia (ESCWA) and the World Bank. Mr. Nabil M. Chahrli served as consultant in the preparation of the study. The next round of negotiations of the World Trade Organization (WTO) will examine major issues related to the three pillars of the Uruguay Round Agreement on Agriculture (URAA) market access, domestic support and export subsidies. It will also focus on key issues under the new trade agenda that have important implications for developing countries, including intellectual property rights, trade-related environment and labour issues, food security, impact of trade policies on rural development, investment and competition policy. In preparation for the next WTO round, many developing countries in the Middle East have requested assistance in the following areas of concern: (i) quantitative economic analyses to evaluate the implications of the new trade agenda in trade and agricultural policy reform; (ii) preparing and formulating appropriate negotiating positions on market access, domestic support and export competition; (iii) evaluating options under second-generation issues, such as State trading, sanitary and phyto-sanitary measures, standards and intellectual property; (iv) enhancing human and institutional capacity to implement commitments under the URAA; and (v) strengthening analytical capacity to effectively participate in the next round of WTO negotiations. In response to that request, three country-specific case studies (on Egypt, the Syrian Arab Republic and Tunisia) were commissioned to address the issues identified above. In general, the preparatory analytical background contained in all three studies is intended to directly assist policy makers in the Middle East in evaluating their interests and assessing the effects of different negotiating modalities during the WTO negotiations. Moreover, the country-specific data compiled in each case will provide useful qualitative and quantitative analyses of the relevant issues. CONTENTS Page Preface................................................................................................................................................... Abbreviations ........................................................................................................................................ Introduction ........................................................................................................................................... Chupter I. OVERVIEW OF THE AGRICULTURE AND FOOD SECTOR ....................................... A. Tunisia's main agricultural products................................................................................... B. Role of agriculture in employment, trade and the overall economy ................................... I1. AGRICULTURAL AND TRADE POLICY .......................................................................... A. Agricultural policy .............................................................................................................. . . B. Prlclng policy ...................................................................................................................... C. Support and subsidization ................................................................................................... D. Trade policy ........................................................................................................................ E. Association Agreement between Tunisia and the European Union .................................... F. Estimates of progress in price and trade reforms .............................................................. I11. EXPERIENCES FROM THE IMPLEMENTATION OF THE URUGUAY ROUND ..... A. Market access...................................................................................................................... B. Domestic support ................................................................................................................ . . C. Export subs~d~es .................................................................................................................. D. Evolution of applied tariffs since Uruguay Round Agreement ........................................... IV . QUANTITATIVE EVALUATION OF POLICY OPTIONS. TRADE-OFFS AND STRATEGIES ........................................................................................................................... A. Welfare analysis of tariff formula reductions ..................................................................... B. Welfare analysis of complete import and export liberalization .......................................... C. Cross-country comparisons from the multi-market analysis of trade liberalization in Jordan, Morocco and Tunisia.......................................................................................... D. Domestic pricing reforms as a prerequisite for further trade liberalization ........................ E. Agriculture and industry linkages in a general equilibrium setting .................................... F. Policy implications and recommendations from the various quantitative assessments ...... V. POLICY ISSUES. OPTIONS. TRADE-OFFS AND STRATEGIES................................... Market access abroad .......................................................................................................... Domestic market access ...................................................................................................... Domestic support ................................................................................................................ Continuation of non-tariff protection measures .................................................................. Export credits...................................................................................................................... Continuation of the peace clause beyond 2004................................................................... Redefinition of State trading companies to conform to WTO regulations .......................... Introduction of the multi-functionality of agriculture ......................................................... Tariff rate quotas.................................................................................................................. CONTENTS (continued) Page SUMMARY AND CONCLUSIONS....................................................................................... 50 LISTOF TABLES Production of major products in volume. 1999 .......................................................................... Macroeconomic and agricultural indicators ............................................................................... Investment in agriculture and food sector................................................................................... Major Tunisian imports and their evolution ............................................................................... Major Tunisian exports and their evolution................................................................................ Geographical destination of exports and origin of imports ........................................................ Average protection coefficients of key products in 1984........................................................... Share of agricultural investment in total allocation .................................................................... Domestic and international price ratios for the major cereal crops in Tunisia ........................... Pricing structure for imported wheat in 1997............................................................................. Evolution of wheat prices and protection levels. 1995-1998...................................................... Exports of olive oil by type of operator...................................................................................... Olive oil exports by the private sector for the 1998-1999 season............................................... Producer and export prices for olive oil...................................................................................... into the European Community of . . . . . agricultural products orlglnatlng In Tun~s~a Arrangements applying.to selected imports................................................................................ Arrangements applying to selected imports into Tunisia of agricultural products . . orlglnating in the European Community..................................................................................... A comparison of nominal protection estimates for selected products ........................................ Post-Uruguay Round bound tariff lines by section..................................................................... Applied tariff expressed in ad valorem terms for selected agricultural and food products in 1998........................................................................................................................................ Applied tariff rates and share in imports, 1998 .......................................................................... Tariff rate quota indicators in Tunisia ........................................................................................ Total aggregate measurement of support commitments by Tunisia. 1995-1999........................ Product-specific aggregate measurements of support in 1995 ................................................... Product-specific aggregate measurements of support in 1999 ................................................... Domestic support in the base period as notified by Tunisia. 1995-1999.................................... Applied tariffs in 1995 and 1998................................................................................................ Data needs and markets analysed in welfare analysis of tariff reduction formulas .................... Key parameters used in the evaluation of the impact of TCFs on welfare ................................. Range of change in TCF parameters........................................................................................... Impact of TCFs on welfare gains................................................................................................ Partial equilibrium estimates of welfare changes from removing selected Tunisian agricultural tariffs based on MFN tariff values. 1998................................................................. CONTENTS (continued) Page 32. Partial equilibrium estimates of welfare changes from removing export quotas on selected Tunisian agricultural products.................................................................................... 33. Type of scenarios simulated in the multi-marketcross-country comparisons ......................... 34. Estimated welfare and income growth effects of price reforms............................................... 35. Welfare and income growth effects of price reforms by agro-ecological zone ....................... 36. lmpact of market liberalization on selected macroeconomic aggregates................................. 37. Results of agricultural reforms simulated with a CGE model for Tunisia............................... 38. Tunisia's main agricultural exports, 1990-1998....................................................................... LISTOF FIGURES I. Trends in cereal production...................................................................................................... 11. Evolution of GDP composition by major activity sector ......................................................... 111. Evolution of GDP composition by sector ................................................................................ IV. Evolution of exports by sector ................................................................................................. V. Evolution of imports by sector................................................................................................. V1. Share of agro-food products in total trade................................................................................ VII. Coverage rate of imports by exports for the agro-food sector ................................................. VIII. Composition of agricultural exports in Tunisia, 1979-1998.................................................... IX. Composition of agricultural imports in Tunisia, 1979-1998.................................................... X. Trends in agricultural support and sector contribution to GDP ............................................... XI. Evolution of tariff revenues from agricultural imports ............................................................ XI1. Comparison of bound and applied tariffs for Tunisia and an average for selected developing countries ................................................................................................................ XI11. Applied tariffs in 1998for selected agricultural and food items.............................................. XIV. Relation between applied tariff levels and share in major agricultural imports....................... XV. Ad valorem tariffs on staple foods ........................................................................................... XVI. Impact of tariff-cut formulas on welfare gains. reference case ................................................ XVII. Impact of tariff-cut formulas on welfare gains. case with high supply elasticities.................. XVIII. Impact of tariff-cut formulas on welfare gains. case with high supply and demand . . . elast~c~t~es ................................................................................................................................. XIX. Sensitivity of welfare gains to formula cuts and elasticity ranges ........................................... XX. Impact of removal of MFN tariffs............................................................................................ XXI. Share of ONH in world olive oil exports ................................................................................. I. Special authorization for the export of olive oil in Tunisia...................................................... I1. General structure of a multi-market model .............................................................................. . . Box.Main policy reforms in Tunlsla .................................................................................................... References............................................................................................................................................. vii ABBREVIATIONS AIPA Agricultural Investment Promotion Agency AMS Aggregate measurement of support ASAP Agricultural Structural Adjustment Programme BCT Central Bank of Tunisia BNA National Agricultural Bank CGC General Equalization Fund (Caisse Generale de Compensation) CGE Computable general equilibrium model CIHEAM International Center for Advanced Mediterranean Agronomic Studies D Tunisian Dinar EU European Union GATT General Agreement on Tariffs and Trade ha Hectare hl Hectolitre IATRC International Agricultural Trade Consortium IMF International Monetary Fund rNS Institute National de la Statistique IOOC International Olive Oil Council LDC Less developed country m3 Cubic metre MFN Most-favoured nation mm Millimetre MOA Ministry of Agriculture MOC Ministry of Commerce NFIC Net-food importing country NPC Nominal protection coefficient NTB Non-tariff barriers OECD Organization for Economic Cooperation and Development ONH Office National de 1'Huile QR Quantitative restrictions STE State trading enterprise TCF Tariff-cut formula TRQ Tariff rate quota LIRAA Uruguary Round Agreement on Agriculture USDOA United States Department of Agriculture USDOS United States Department of State WTO World Trade Organization INTRODUCTION There has been a broad trend towards liberalization of markets and trade in the Tunisian economy in recent decades. However, a lagging agricultural sector, chronic unemployment problems and slow structural reforms in some sectors have been keeping Tunisia from achieving higher growth rates. Concerns have been expressed that the country is not yet ready for participation in the globalized world economy and might turn away from open trade and gradual market liberalization, even though it has been combining that trend with the prudent macroeconomic policy it has been pursuing over the past fifteen years to generate economic growth and reduce poverty. In this context, Tunisia has strong interest in the new WTO round of multilateral trade negotiations covering all sectors, including agriculture. In a country with limited natural resources but with very close commercial ties to the European Union (EU), it is felt that the negotiations will have to be responsive to the needs of those that have not yet seen the fruits of the liberalization process to ensure that they can reap full benefits from integration into the world economy. The new "Millennium Round" of WTO negotiations would allow developing countries like Tunisia to converge their interests and policies with those of developed countries and thus enable them to join the free trade and open market-based world economy. There is now worldwide common interest in pushing forward the process of liberalization and in finding solutions for the increasingly complex economic, environmental, social and political issues associated with the requirements of a global economy and which are being addressed through the WTO and other bilateral agreements, such as the one Tunisia has with the EU in the context of the Euro-Mediterranean Partnership. It is, therefore, vital for a country like Tunisia to participate actively in the preparatory process for the new round of negotiations, especially with regard to agriculture. Despite its declining share in total gross domestic product (GDP), the agricultural sector still carries heavy weight illterms of employment and represents an important foundation in the economy of Tunisia. This report attempts to contribute to the preparatory process by reviewing some of the evidence on current trade distortion patterns and the progress made in market liberalization, drawing on Tunisia's experience with the Uruguay Round and implementation of the Agricultural Structural Adjustment Programme. That experience has shown that, despite notable efforts to liberalize agricultural trade and markets, domestic distortions remain through the use of various instruments (e.g., high applied "most- favoured-nation "(MFN) and bound tariffs, the existence of State trading enterprises (STEs) in key sectors and in handling key commodities, and important price interventions). The same could be said of Tunisia's main trading partners, as Tunisia's main exports still face substantial trade barriers. Tunisia. therefore, has a double stake in the negotiations. First, it should look at how far it can go in further opening its agriculture economy. Second, it should follow with close interest how the EU responds to alternative ways of settling items relative to the three main WTO pillars. In addition to a broad overview of the sector and a review of agricultural and trade policy and the status of Uruguay Round implementation, this report provides some quantitative analyses related to potential modifications in the rules affecting Tunisia's current trade flows. One issue that is examined is related to the performance of tariff-cut formulas on key imported commodities. One particular formula (deep harmonization) seems to be providing high welfare gains when applied in a medium range of domestic price reductions. As we move closer to a situation of zero tariffs or of keeping high levels of protection, tariff-cut formulas tend to fare the same. What seems to be as important as the type of formula is the propensity of Tunisian agriculture to rapidly respond to price signals, whether on the supply or demand side. The report also reviews a number of studies that have looked at various patterns of trade liberalization in agriculture using various methodologies-from multi-market analysis of trade and domestic pricing policies to computing general equilibrium (CGE) modelling of combined agriculture and industry tariff reductions and preferential access to the EU market. The results show that Tunisia should not shy away from deeper trade liberalization and that it should look for targeting schemes to help those segments of the population that are hurt by the process. For example, low rainfall areas could lose from reduced border protection, as those regions do not have any choice other than crop-livestock activities; hence, their markets should continue to be protected despite commitments taken to liberalize those strategic sectors. Overall, Tunisia would have little to gain from closely following the classical agenda backed by less developed countries (LDCs) or net-food importing countries (NFICs). It could potentially gain from embracing some features of the Cairns Group philosophy, but it should be extremely vigilant on how much the EU (the origin and destination of more than three quarters of its trade in agricultural and food products) would consent to in terms of reduced market distortion. Discussions on the future of STEs are to be followed closely, as Tunisia still maintains a system of para-statals and marketing boards with broad import and export responsibilities. Finally, Tunisia seems to be limited in its potential to gather and process relevant information concerning the items to be negotiated in the future round. This preparatory research, therefore, carries important weight in, first, bridging the gap between trade and agriculture policy makers and, second, in providing evidence on the potential implications of the discussions on the three URAA pillars as well as those related to elements of the new trade agenda. I. OVERVIEW OF THE AGRICULTUREAND FOOD SECTOR Agriculture plays an important role in the overall social and economic development of Tunisia and, as a result of agro-ecological factors, the sector is well diversified. There are three main producing regions in the country: (a) The northern region, which covers 19 per cent of the country and includesthe plains ofthe Medjerda river basin. This is the region with the highest rainfall, covering the 400-1,200 millimetres(mm); (b) The central region, which covers 23 per cent of the country. This is a semi-arid region, where barley, sheep and fruit trees constitutethe predominant production system; (c) The southern region, which covers the remaining 58 per cent of the country and is arid to semi- desert in character. Most agriculturalproduction in this region occurs in the Chott el Djerid oases and on the northern and eastern fringes. Rainfall is erratic, with an average annual precipitation of 100-200mm. Tunisia's agricultural lands cover an area of 9 million hectares, divided into 5 million in cropland, 3 million in range land and 1 million in forests. Water resources are a limiting factor. The irrigation potential is about 570,000 hectares, with 400,000 in full irrigation (FI) and 170,000 in supplemental irrigation (SI). However, only 340,000 hectares are currently under F1 and 50,000 under SI. Surface water has been estimated at 2.7 billion cubic metres (m3),of which 2.1 billion m3are accessible (Chouchane, 2000) A. TUNISIA'SMAIN AGRICULTURAL PRODUCTS The four most important agriculture activities are cereal crop, livestock, fruit trees and vegetables. Table 1 reports production, consumption and trade indicatorsfor the major agricultural and food items in Tunisia. Cereal production is one of the leading activities in the agricultural sector. Currently, 1.6million hectares are devoted to wheat and barley, representing around one third of the total cropping area. Half the farmers in Tunisia (approximately 240,000 people) cultivate grains. Out of 1.8million tons of cereals produced in 1999, 1.1 million tons were durum wheat. Figure I shows the contribution of each cereal crop to total grain production. Among the most notable trends, the relative increase in barley production after the 1995 drought hints at the potential changes in crop profitability and feed demand that have occurred in those past years. A strategy for the sector was put in place in 1997 to reach a cereal production of 2.5 million tons by the year 2006. One of Tunisia's most famous agricultural products worldwide is olive oil. Out of 2 million hectares planted in fruit trees, olive trees figure prominently in land usage throughout southern and central Tunisia. The remaining area is planted with almonds, dates and citrus. This sub-sector covers 60 per cent of Tunisia's agricultural exports and around 28 per cent of total agricultural GDP. Crop-livestock integration is an integral part of the agricultural systems prevailing in Tunisia's arid environments. With 654,000 heads of cattle, 7.5 million sheep and goats and 70,000 camels, livestock represents the bulk of activity for 335,000 flock owners. Seventy-threeper cent of farmers in Tunisia are in some way or another connected with rearing livestock. Representing 32 per cent oftotal agricultural production, this sector has generated 156,000tons of meat and 526,000 tons of milk annually over the past years. The vegetable sector in 1998covered around 140,000hectares, of which 16 per cent was planted with tomatoes, 11 per cent with potatoes, 12 per cent with peppers and 24 per cent with melons and watermelons. Vegetables contribute around 15 per cent to total agricultural production. The sector generates a surplus that is exported (BCT, 2001), as in the case of tomatoes (27,000 tons) and processed peppers (2,200 tons of hot sauce) generating 6.3 million Tunisian dinars. B. ROLEOF AGRlCULTURE PJEMPLOYMENT, TRADE AND THE OVERALL ECONOMY Table 2 gives a summary of major macroeconomic and agricultural indicators. Agriculture contributes between 11-16per cent to the GDP, depending principally on climatic factors, between 8-12 per cent to total exports and benefits from 15 per cent of the total investmenteffort.' It also accounts for 25 per cent of total employment in the economy. Over the period 1991-1997,Tunisia's rural areas lost around 4 to 5 percentage points in population. Currently, 37 per cent of the total population live in rural areas. The Tunisian economy has been achieving impressive growth since the mid-1990s,with the rate of growth averaging 5.2 per cent in real terms over the past five years (IMF, 2001). Against this background of strong performance, the agriculture sector has not manifested the same dynamism noticed in other sectors of the economy, despite increases in yield and productivity and the gradual liberalization efforts undertaken during the past decade. As shown in figures I1and 111,the share of agricultural and food production has been declining for more than thirty years. This diminished importance of agriculture is also reflected in Tunisia's patterns of trade, as evidenced by the trends in figures IV and V showingthe evolution of exports and imports over time. In the early 1970s' 25 per cent of the country's exports and 20 per cent of its imports were related in one way or another to the agro-food sector. Thirty years later, those shares have been reduced to an average of less than 10 per cent. Figure VI gives details on the most recent agro-food import and export shares in total trade, while figure VII shows the coverage rates of imports by exports. The average rate of coverage of exports by imports was around 65 per cent for the period 1993-2000. Tables 4 and 5 and figures VIII and IX provide details on the composition of agro-food trade for various time periods, while table 6 indicates the geographical features of the exchange of goods and merchandise. TABLE1. PRODUCTION OF MAJOR PRODUCTS IN VOLUME, 1999 Self-sufficiency Production Consumption Imports Exports ratios Drum wheat 1 144.0 1393.0 249.0 0.82 Soft wheat 253.0 1090.5 837.5 0.23 Barley 408.0 619.4 211.4 0.66 Milk 817.0 826.3 9.3 0.99 Beef 50.0 52.4 2.4 0.95 Lamb 46.0 46.0 0.0 1.00 Sugar 89.4 394.4 305.0 0.23 Potatoes 320.0 336.0 20.5 4.5 0.95 Tomatoes 930.0 937.1 7.1 0.99 Peppers 185.0 185.0 1.OO Watermelonsand melons 350.0 350.0 1.OO Other vegetables 526.0 526.0 1.OO Maize 0.0 681.4 681.4 0.00 Soybean cake 0.0 241.8 241.8 0.00 Vegetable oil 180.0 171.7 198.7 207.0 1.05 Cheese 11.2 12.7 1.5 0.88 Dates 1 18.8 95.9 22.9 1.24 Citrus 210.5 190.7 19.8 1.10 Raisins 76.0 76.0 1.OO Other fruits 448.7 461.2 12.5 0.97 Source: Ministry of Agriculture,Annuaire des Statistiques Agricoles: 1999 (2000) and author's computations. ' See table 3 for the evolution of investmentsin the sector. TABLE2. MACROECONOMIC AGRICULTURAL INDICATORS AND Unit 1991 1993 1995 1997 Population Total population Millions Urban population Millions Total labour force (TLF) Millions Agricultural population Millions Agricultural labour force (ALF) Millions ALFITLF Percentage Production GDP Millions US$ GDP per capita US$ Agricultural GDI' Millions US$ Agrjcultural GDPIGDP Percentage Agricultural GDPIALI US$ Exchange rate IJS$ITD 1.081 0.996 1.057 Factors of production Arable land and permanent crops (ALPC) Irrigated area Permanent pasture Forest and woodland Arable land and permanent cropslALI: Arable land and permanent croplpopulation Fertilizerslhectare ALPCItractor Source: Medagri, 2000 TABLE3. INVESTMENT IN AGRICULTURE AND FOOD SECTOR (in millions of Tunisian dinars) 1990 1991 1992 1993 1994 1995 1996 1997 1998' 1999' 2000" 2001" Agr~cult~~reand 14 8 14 7 12 3 11.9 I2 I 14.5 16.2 14.3 14.7 13 7 13 0 13 3 Food ~ndustrtes 2 8 3 3 2.7 2 7 2.7 3 2 3.2 34 4 0 3.7 3 2 3.2 Sub-total a g o - food 17 5 18 0 150 14.5 14 7 17.6 19.4 17 6 18.6 17.4 16.2 16.5 Total General 2 634.8 2 892 3 3 728.6 4 122 3 4 278.7 4 120.7 4 422.2 5 152.8 5 613 5 6 2689 7095 7750 Sotrrce: Budge1Economique,2000. * Semi-definitivedata. ** Estimates for 2000 and forecastsfor 2001 TABLE4. MAJORTUNISIANIMPORTS AND THEIR EVOLUTION (Percentage of total agricultural imports) 1993 1994 1995 1996 1997 1998 1999 2000 Durum wheat 0.9 1.7 14.4 7.2 5.6 10.1 5.0 7.8 Sott wheat 16.0 13.4 15.4 17.7 17.7 14.5 11.2 12.8 Barley 0.5 13.1 7.9 0.6 4.9 1.4 2.3 6.4 Maize 6.9 4.9 4.5 7.8 7.1 6.6 10.2 10.4 Potatoes 2.2 1.9 2.6 1.7 1.4 2.5 1.5 1.7 Coffee 1.1 2.3 2.6 2.2 2.3 2.0 2!. I 1.9 Tea 3.2 2.9 1 . 1 2.2 2.4 2.5 2.5 2.5 TABLE4 (Continued) 1993 1994 1995 1996 1997 1998 1999 2000 Meat 4.7 2.3 1.2 1.3 2.8 1.8 0.7 1.1 Milk and dairy 10.1 6.0 5.3 3.8 4.4 2.5 2.4 2.8 Vegetable oils 12.3 13.7 11.6 15.2 10.2 14.0 14.6 10.5 Sugar 11.6 14.0 7.2 11.8 9.5 10.8 9.2 7.6 Soybean cake 6.5 5.4 3.0 5.4 7.4 6.0 5.8 7.5 Tobacco 3.8 2.1 1.7 2.4 2.5 3.5 2.7 2.0 Total agro-food (Millionsofdinars) 533.8 669.9 956.5 742.3 975 1004.3 865.5 986.4 Total imports (Millionsofdinars) 6 172.1 6 647.3 7 464.1 7 498.8 8 793.5 9 489.5 10 070.5 1 172.8 Source: Institut National de la Statistique, 2001 TABLE MAJORTUNISIAN 5. EXPORTS AND THEIR EVOLUTION (Percentage of total agriculfural exports) 1993 1994 1995 1996 1997 1998 1999 2000 Olive oil 40.5 50.0 42.0 28.9 42.3 33.9 48.4 37.3 Fish and seafood 20.5 13.5 14.4 22.5 15.8 20.1 12.9 17.0 Dates 10.9 9.3 1 1 3 11.5 7.6 11.2 7.1 7.5 Citrus 2.2 1.3 2.1 2.2 I .O 1.5 1. I 1.4 Total agro-food (Millionsof dinars) 438.2 610.5 515.4 404.8 681.7 627 790.9 707.2 Total exports (Millionsof dinars) 3760 4 696.6 5 172.9 5372 6 147.9 6 518.3 6966.9 8 004.8 Source: Institut National de la Statistique, 2001. TABLE6. GEOGRAPHICAL DESTINATION OF EXPORTS AND ORIGIN OF IMPORTS 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Exports Africa 8 8 9.4 9.1 8.6 7.2 8 5 7.2 7.0 6.3 7.0 6.5 America 1.8 1.5 1 7 1 4 1 8 2.1 2.2 1.7 1 5 1.3 2.0 Asla 6.8 6.1 6.3 5 2 4.9 5 0 5 3 6.0 5.9 5.6 5 4 Europe 81.1 80.7 81 6 81 7 82 7 81.2 82.3 81.1 82.8 82.8 82.4 Other areas 1.5 2.2 1.3 3.0 3 3 3.1 3 0 4.3 3 6 3.3 3.7 Tolal 3087.4 3417.1 3 5 4 9 7 3 760 4 696.6 5 172.5 5 372 61 479 65 183 6966.9 8 004.8 Imports Afr~ca 49 4 8 5 5 4 4 5.7 6.8 6.7 59 4.6 5.1 6 6 Arner~ca 8.1 8.0 7.9 7.9 9.2 7 4 7.0 7.8 5.8 6 6 6.5 Asia 6.8 6.1 5 9 6.3 6 3 5.9 6.9 7.3 7.3 8.2 8.5 Europe 79.6 80.6 79.4 80 3 77 8 79 1 78 5 78.0 81.4 79.0 77.1 Other areas 0 7 0.5 1 2 I.O 0.9 0.7 0.8 0.9 0.9 I 0 1.2 Total 4 826.4 4 788 9 5 688.8 6 172 1 6 647 3 7 464.3 7 498.8 87935 94 895 10070.5 1 1728 Source: lnstitut National de la Statistique, 2001 Figure I. Trends in cereal production I3Barley , Soft Wheat 1ElDurum Wheat Source: Ministry of Economic Development, Budget Economique, 2000. Note: Data labels represent production values in 1,000tons. Figure 11. Evolution of GDP composition by major activity sector Source: Author's computations based on mS data. Figure 111. Evolution of GDP composition by sector u ~ ~ n c u l ~andr firhcncr u c .~ood ~ndurlrr n ~ u ~ l d nmalmalr, c r m l c r and glass n g D~echanxcal8: ele~tncal~ndurtrlrn m ~ h ~ mandwrubber ~ l ~ T e x u l mcloth~ngand leatkr .Mlsc ~ndurtnm n ~ ~ n ~ n g m~,dr-rbans .Encrgv n ~ a t r r m ~ m r t ~ a ! o n clvtl cng~mnng and .Transpan n~ounsm .~omrst~c tmde m ~ n r ca n ~ w r Source: Author's computations from database compiled from various issues of BCT Annual Reports, Budget Economique, Ministry of Economic Development, and Annuaire des Statistiques Agricoles, Ministry of Agriculture, 1972-1998. Figure IV. Evolution of exports by sector Agriculture and hshernes m ~ ~fmdo~ndustry r n ~ u ~ l d mater~alsand cerarnlcs ~ n a .Mechanical& electr~callndustner B ~ h e m l c aand rubber l .Texl~les leather 8r fmnvear nhjtsc Lndustraes .Mlncng .Hydrocarbons m~lectrlc~ry P ~ a l e r .~onstructnon & clvll engsneenng .Transpan .Tourlsrn - .~ornestlc - trade P ~ ~ sservlces c Autllor's computations based on INS data. Figure V. Evolution of imports by sector m ~ ~ r i c u l t uand fisheries r e .Ago-food industry O ~ u i l d i nmaterials and ceramics ~ OMechanical & electrical industries .chemical and rubber m~extiles.leather & footwear WMISC,Industries OMining m~~drocarbons ~ .~lectricit~ qwater ~ ~ o n s t r u c t i o&ncivil engineering 1 Transport .Tourism B~omestictrade WMisc. services Source: Author's computations based on INS data. Figure VI. Share of agro-food products in total trade +Imports Share +Exports Share 1993 1994 1995 19% 1997 1998 1999 2000 Source: Author's computations based on INS data. Figure VII. Coverage rate of imports by exports for the agro-food sector 1993 1994 1995 19% 1999 2000 Source: Author's computations based on INS data, 2001. Figure VIII. Composition of agricultural exports in Tunisia, 1979-1998 ~ L I anlmals V ~ OF IS^ & seafood n ~ a l r & eggs y o t h e r animal products aFlowers & plants Vegetables n ~ r u i t s .coffee, tea & splces .cereals DMilling industry products a011seeds m ~ a n n products ~ n ~ m ~ l f & other plants a vegetable 011s & fats m ~ e a&t fish preparatlons ..Sweets & sugar n ~ o c o & preparations a n ~ e r e a l bread preparatlons s nvegetables preparations m~iversefood preparations Beverages & alcohol~cdrlnks m ~ o o dindustry resldues -- - Source: Various volumes of INS Sfatistiquesdu Commerce Exfe'rieur,1979-1998 Figure IX. Composition of agricultural imports in Tunisia, 1979-1998 qBeverages & alcohol~cdr~r qDlverse food preparatlons qVegetables preparatlons OCerenls bread prcparatlons UCocoa & preparat~ons Sweets & sugar M e a t & fish preparatlons .vegetable 011R fats . ~ l f a & other plant5 ~ a n n products~ ~ n 0011seeds ~ , M I I ~nduhtnproduct> I I ~ ~ C e r e a l s C o f f e e tea B bplces n ~ r u ~ t s Vegetables OFloners & plants o t h e r anl~nalproducts qD a ~ q& eggs ~ F I &Sseafood ~ M e a t - 0- o 1979 1984 1989 1994 Live animals Various issues of INS Stntirtiq~fesn'trCommcr-ce ErtPriezlr. 1979-1998. 11. AGRICULTURAL AND TRADE POLICY Agricultural policy in Tunisia Passed.through three major phases that are linked to the overall . development objectives of the country.- The ~n~tialphase of the post-independence era was characterized by a "forced" collectivization policy (1962-1969), in which the agricultural sector was supposed to play a key role in the overall development of the country. To reach that goal, agrarian reforms were introduced with the establishment of a cooperative system of agricultural production, in addition to government schemes to modernize production systems. The abandonment of the collectivization experience in 1969 was the starting point of a liberalization policy in the economy, in which the agricultural sector played a limited role confined to providing cheap food to consumers and securing hard currencies through the export of surplus commodities. This second stage was marked by the modernization of the production system principally by the State and large farms and the establishment of large and medium hydraulic projects to mobilize water. To reach the objective of this second phase, the State provided input (e.g., feedstuff, fertilizers, machinery) and credit subsidies for producers and consumer subsidies on basic food items. At the same time, the Government fixed the producer prices of the major crop and livestock commodities for many years, including soft and durum wheat, barley, milk and beef at levels below world market prices, which often resulted in negative nominal protection (World Bank, 1986). Imports of food items and inputs were not viewed as competing with locally produced products but rather as a complementary scheme to fill the gap between a supply limited by climatic conditions and the fast-growing demand generated by the increase in population and urbanization. This gap, which has become structural since the mid-1970s. prompted the Government to institutionalize endowments from the State budget for the General Equalization Fund ("Caisse Generale de Compensation", commonly referred to as CGC). CGC is a public entity in charge of managing consumer subsidies and the farm support system. Prompted by mounting financial and sectoral imbalances, the beginning of the 1980s saw a reshuffling of agricultural policy in Tunisia, which culminated with the adoption and implementation of the Agricultural Structural Adjustment Programme (ASAP) in 1986. The Programme's main objective was to more fully realize the "~~ntapped"potential of Tunisian agriculture and enable the sector to contribute more substantially to the macroeconomic objectives of growth and balance of payments equilibrium. Its measures on the output side included changes in the incentive framework to reduce under-protection patterns for durum wheat and barley and shift the bias in favour of soft wheat (as shown in table 7). In livestock products, domestic production of milk and beef was set to benefit from a 15 per cent minimum tariff, in addition to progressive increases that would reduce the difference existing with world prices. On the input side, reductions were set for fertilizers, animal feed, herbicides and certified seeds. 011 the consumption side, the Programme also reduced subsidies on cereals and milk. The ASAP coincided with the beginning of Tunisia's 7th Development Plan, which set the following global development objectives: (a) Better resource allocation between the productive sectors and regions to lead to a more equitable income distribution; (b) Control of consumption growth in order to mobilize savings; (c) Improve~nentin the balance of payments to contain the national debt; (d) Reduction of ~~nemployment(Abaab and Elloumi, 1997). It was in this context of a better balance between sectors and regions that agriculture was supposed to benefit from higher investments and better price incentives for producers of agricultural and food-related commodities. Agricultural policy-making was aimed at improving the food balance, so that an adequate See Abaab and Elloumi (1997) for a review of agricultural policy in Tunisia up to 1995. Description of most of the pre- 1995 features of this policy draw Srom this source. supply of strategic commodities could be guaranteed. This could be considered as the genesis of the concept of food security in Tunisia and as an alternative to the food self-sufficiency targets set in previous development plans for the sector in particular and the economy in general. The overall development objectives mentioned above had important practical implications for the sector. The agriculture-specific development objectives were: (i) to modernize the production system and improve technical and economic efficiency; (ii) to favour the emergence of agricultural entrepreneurs in launching capital-intensive enterprises, and (iii) to increase investment in the sector. Agricultural policy in Tunisia is implemented through various instruments. The most important interventions are channeled through public investment, output and input incentive schemes, credit, marketing, and research and extension. Some of the main features of those policies, with particular relevance to one of the three WTO pillars (domestic support), are reviewed below. There has been an increase in the level of agricultural investment, from 12.9 per cent under the 5th Development Plan to 17.1 per cent under the 8th Plan. This upward trend does not seem to be sustained in the current five-year plan (1997-2001), as the share in total investment dropped to 13.3 per cent during the first half of the plan. Table 8 shows the share of agricultural investment in total investments. Credit policy, as related to the agriculture sector, has been driven by the aim to mobilize larger resources for farmers and rationalize credit distribution and use. In the phase preceding ASAP, the Government tried to set a system of incentives targeting agricultural entrepreneurs, organized either as individuals or as companies referred to as "SocietC de Mise en Valeur de Developpement Agricole". One of the measures implemented was to establish the Agricultural Investment Promotion Agency (AIPA) to help in the identification, promotion and follow-up of agricultural investment projects. This was followed by the establishment of a specialized bank, the National Bank for Agricultural Development (BNDA), to formally mobilize and distribute resources for the sector. BNDA was to complement the role of the National Bank of Tunisia (BNT) in managing the State budget devoted to agriculture. In the early 1980s, a new investment code was put in place for agriculture and fisheries, which offered specific advantages to the various categories of investors. One of the important features of credit policy was the passing of a law to help Tunisian companies in the management of State domains, which concerned long-term leases and stipulated that foreign participation in ownership was allowed as long as the majority stayed in the hands ofTunisian nationals. A second important credit policy move was the establishment of regional irrigation scheme development offices ("Office Regionaux de Mise en Valeur des Perimetres Irrigues") that distributed credit in kind to farmers. Those offices were instrumental in the development of greenhouses and increased fruit tree activity. Despite the large credit programmes put in place in the form of institutional credit, Government subsidized credit and projects with credit components, the policy failed to achieve the desired results in terms of resource mobilization, productivity gains and income generation in the sector. The complex credit system was marked by several endemic problems that were further aggravated by the financial crisis the country went through in the years preceding ASAP adoption and implementation. The system led to a degree of confusion between social subsidies and bank lending, which discouraged full repayment of debts and a higher degree of involvement by the banking sector. With ASAP, three main areas of intervention were targeted (World Bank, 1986): (i) encouraging the banking sector to play a more active and efficient role in credit allocation; (ii) iniproving the financial viability of agricultural credit operations by increasing interest spreads and harmonizing credit terms and conditions; and (iii) encouraging mobilization of rural savings through banking schemes. Credit policy in subsequent five-year plans had to obey a different philosophy, one that would give the private sector a more prominent role in credit operations. The reform process led to a number of important measures, the most important of which were the new 1988 Investment Code and the merger in 1989 of BNDA and BNT into the National Agricultural Bank (BNA). Results over the first half of the 1990s do not show evidence of significant increases in access to credit, since less than I5 per cent of the farmers benefited, and bank participation in the process remained limited, with only 9 per cent of credits going to the sector (Abaab and Elloumi, 1997). Another set of reforms implemented in 1994 brought about the removal of credit subsidies and a reduction in the wedge between market rates and agricultural lending rates. This led to a new investment code that provided consolidated incentives for all types of investment, which nevertheless took into account the specificities of the sector and tended to give agriculture a status. The new code is more consistent with the overall spirit of the ASAP being implemented nationwide. In financing agricultural investments, priority has been given to activities and farms with solid prospects in terms of profitability and long-term sustainability. It is now recognized that the State will continue to exercise its social role, but only if the criteria of economic performance is met. An illustration of this new credit policy philosophy is the removal of a plethora of interest rate bonuses covered from budgetary allocations. B. PRICING POLICY Since the implementation of ASAP, the State has tried to base its pricing policy on "real prices" by enabling market mechanisms to play their part (e.g., for fruits and vegetables, fish and meat). However, in the case of a number of so-called essential commodities, such as cereals, milk and olive oil, prices are still fixed. The prices set are intervention prices and producers are under no obligation to deliver set quantities produced at this price. Because of the importance of particular commodities in total production and trade, the pricing policy conducted in key agricultural markets is reviewed next, with focus on those aspects that are most relevant to trade implications. Cereal prices in Tunisia have been consistently higher than world prices in the past decade. Table 9 shows the most recent trends for the key cereals grown in Tunisia. Since 1980, guaranteed minimum production prices for cereals have been announced by the Government and are referred to as base prices ("prix de base"). Farm gate prices are computed as the base prices minus a special tax, the purpose of which is to cover expenses of the Cereals Marketing Board ("Office des Cereales" or O C ) ~and fund specific seed improvement programmes. As for imported cereals, prices are set according to a structure put in place in January 1991 and approved by the Finance and Economic Development Ministries as well as by OC (Lachaal, 1999). The final cost of imported wheat is computed by adding the tariffs to CIF prices, plus a margin paid to OC to cover its storage operations. Table 10 shows how the different components of the final cost of imported wheat were computed in 1997. Commercialization of wheat in Tunisia, and in particular sales to processors, is made in such a way that the wedge between the domestic and international price is filled in by a system of subsidization and compensation borne by the State. Support for producers and consumer subsidies are adjusted each year and sometimes several times a year (as occurred in 1989), according to fluctuations in the world market and the base prices announced by the ~overnment.~ Guaranteed prices for soft and durum wheat have been quite stable since their increase in 1996, as opposed to prices in the world market (as shown in table I I). Ben Cheikh (2000) reports nominal protection coefficients (NPCs) for the second half of the past decade and shows the differences in protection trends and magnitude existing between soft and durum wheat. In addition to providing producers with price support, OC manages cereal imports through public auctions. Among the major developments noted in recent years was the privatization in 1997 of wheat processing companies and bakeries under OC ownership and the liberalization of maize and soybean cake imports, a monopoly privilege long held by OC. In addition to OC, Tunisia has a number of STEs that are playing roles of various importance in the domestic and world market. Among those entities under direct tutelage of the Ministry of Agriculture (MOA) are: OC was created on 3 April 1962 under Decree No. 62-10. Its status was subsequently modified by Law No. 70-7 of 20 November 1970 and Law 86-67 of 16 July 1986. Lachaal (1999) gives a description of different public interventions in the wheat sector, from farm gate to processing. (a) Office National de I'Huile (ONH), charged with broad responsibilities in the collection and export of olive oil as well as the import of vegetable oils; (b) Societe de Dkveloppement Agricole et des Dattes (SODAD), a public company that emerged from the break-up of a bigger public enterprise, the Societe Tunisienne d71ndustrieLaitiere (STIL), and specializes in the commercialization of dates. This company has been losing its leadership in the sector, as the activity is being taken over more and more by the private sector. The set of STEs under the direct control of the Ministry of Commerce (MOC) includes: (a) Office du Commerce de Tunisie (OCT), a State trading enterprise with a monopoly in the import of "sensitive" food items such as coffee and tea; (b) Societe Ellouhoum: a public company in charge of cattle meat imports and domestic wholesale commercialization (catering mainly to public institutions such as hospitals, schools and so on). Out of this list of STEs dealing with trade, we focus next on another major player in the agricultural market, ONH, because of its far-reaching influence not only on the domestic market as a trading company but also on agricultural policy-making. In fact, the State utilizes ONH as an instrument of regulation, control and stabilization in the vegetable oil market far beyond its originally intended role as an olive oil exporter when it was first established.; Tunisia is one of the top players in the international olive oil market. Only second to the EU in terms of production, it recently climbed to the position of world leader in terms of exports. With 55 million trees, of which 1 1 million have not yet reached maturity (IOOC, 1998), the olive sector covers around 1,400,000 hectares, representing 3 1 per cent of total agricultural land. It is a major source of employment, as 1.5 million people live from activities related to various stages of its production and manufacturing (Maaroufi, 1999). Despite an inherent variability in production, Tunisia passed the 200,000-ton mark in the 199111992 season and the 300,000-ton mark in 199611997. Two phases characterize agricultural policy-making with respect to olive oil. Before the passing of Decree No. 93-2328 (27 October 1993), the marketing of olive oil and its by-products was under ONH monopoly. Despite changes in the responsibilities of ONH, the Office kept a tight grip on the sector through mandatory production restitution and price-fixing schemes. It was in the context of the ASAP initiated in 1986 and trade liberalization as committed to in the General Agreement on Tariffs and Trade (GATT) that Tunisia gradually started changing its vegetable oil policy. The Decree of October 1993 and Law No. 94-37 of 24 February 1994 marked major shifts in that policy by dismantling the public monopoly over the commercialization of olive oil and olive by-products. The marketing of olive oil is no longer under the strict monopoly of ONH and private traders have been allowed to collect olive oil either directly from producers or from ONH storage facilities. ONH remains an important player, as olive oil manufacturers have to abide by specific conditions6 if they retrocede their production directly to ONH in exchange for a premium. This bonus was fixed to 600 Dlton for the 1998-1999 season. Since 1994, the structure of olive oil exports has changed substantially, as private traders assumed a higher stake in the export market (between one quarter and one half of total exports in the seasons between 1996-1998 and almost 55 per cent in 1999) as shown in table 12. The volume of exports by traders is shown in table 13. One constraint private traders have to face is the requirement to export their oil to non-European markets, since the EU negotiated quota can only be filled with oil exported by ONH. A law to be passed in 2001 will allow private traders a small share of that quota.7 For the 1998-1999 agricultural season, most of ONH was established in 1962 by Decree No. 62-64 dated 20 August 1962, giving it the status of a State trading enterprise with autonomous financial management. ")live oil refineries wishing to deliver their oil to ONH at a premium have to be registered within the administrative structure sct by ONFI. The premium or price complement is determined at the end of the marketing season (BCT, 2001). There are currently 550 of these refineries, representing 38 per cent of the total number of refineries in the country (MOA, 1999). This is based on unofficial discussions in March 2001 with Commerce and Agriculture Ministry officials, hinting that a quota of 4,000 to 6,000 tons is soon to be made available for private traders wishing to export to the EU market within the 50,000-ton quota negotiated in the 2001 EU-FTA Amended Protocols. the exports from private traders were destined to the EU under a special trade channel called "Traffic de Perfectionnement Actif' or TPA (MOA, 1999), under which Tunisian olive oil destined to the EU enters the market as an input. This type of oil is later processed, conditioned and sold as a locally made product in one of the EU countries, with Italy taking the bulk of this trade. To be able to export olive oil, a private trader has to submit an application package8to ONH (MOA, 1993). As for the olive oil exported by ONH, 45,591 tons were channeled to the EU under the preferential quota system, 21,400 tons under the TPA system, 9,752 tons to the Libyan Arab Jamahiriya and the rest to various trading partners. The rationale for retaining the bulk of the EU quota for ONH is to use the export proceeds to cover two important ONH tasks in the domestic market. First, by providing a guaranteed price, ONH assumes a financial burden when prices are low. ONH, therefore, plays what the State perceives as a stabilization role in the domestic market as a result of volatile supply and demand conditions in the world market. Second, ONH covers from its own endowments an important health protection programme targeting millions of olive trees. Because of the presence of negative externalities, the failure of one farmer to take the right protective measures could have catastrophic implications for the surrounding tree growers. MOA and ONH look at this operation as a public good provision, the cost of which can only be borne by the State. There have been wide fluctuations in producer and export prices over the past decade, as shown in table 14. In addition to rainfall and climatic factors conditioning the level and quality of olive fruit and oil production, the export market is heavily dependent on demand and supply conditions, mainly in the EU market. Following the liberalization process initiated in 1994, important price fluctuations have taken place in the olive oil market between the beginning of the season and the end of the marketing year (Abaab and Elloumi, 1997), with prices moving up to 2,400 Dlton from a low of 1,500 Dlton. In addition to its intervention in the domestic and international olive oil markets through the setting of intervention prices and direct filling of the EU quota, ONH has a monopoly on the import of some vegetable oils. This last intervention will most probably be taken out of ONH mandate, since the sector is moving towards a deep restructuring that will allow the private sector to take over what the State has been handling since its independence. C. SUPPORT AND SUBSIDIZATION An important arm of Tunisian economic policy is the General Equalization Fund of the CGC, used to subsidize consumer goods and support farmers. Interventions of this Fund have been diminishing since 1997 as a result of concerns about achieving development plan targets. Disbursements went down from 333 million D in 1998to 206 million D in 1999 (BCT, 2001), as compared to an initial forecast of 310 million D budgeted in the Finance Law. The share of CGC disbursements in GDP dropped from 1.5 per cent in 1998 to 0.8 per cent in 1999(calculated in current prices). Lower subsidies and support are due to lower import prices, in particular those for cereals and vegetable oils, and to price adjustments for cereal by-products and the elimination of subsidies for triple super-phosphate and locally produced sugar. Despite these positive developments in terms of market liberalization and reduced State intervention in the economy, the weight of subsidies remains considerable in subsidizing cereal grains and by-products (127 million D) and vegetable oils (59 million D). Those two items represent more than 90 per cent of the total support channeled via CGC. The cost of funding CGC from the State budget rose to 237 million D in 1999versus 348 million D in 1998 (BCT, 2001). The decrease in the agricultural support mentioned above is reflective of a long-term downward pattern. Figure X shows the evolution of combined disbursements of CGC to the agriculture, fisheries and food industry sectors over the 1972-1998 period and compares those trends with the contribution of those sectors to GDP. While the contribution of agro-food has been consistently declining since 1972, support to the sector saw a dramatic increase until 1989. Since then, fluctuations in the level of support have tended to be negatively linked with the contribution of the sector to total GDP. Those trends provide evidence of the major shift that has occurred in agricultural policy-making, with ASAP support increasing only after periods of drought when it is really needed. ' Details of the application procedure are provided in annex I 16 Tunisia is a member of the WTO and has bilateral association agreements with Egypt, the EU, Jordan and Morocco. It also maintains preferential trade arrangements with Algeria and the Libyan Arab Jamahiriya and is a signatory of the Greater Arab Free Trade Agreement. While maintaining restrictions on designated "strategic" products, the Tunisian Government has been pursuing a programme of trade liberalization. Approximately 97 per cent of imported goods now do not need a license, compared to 23.6 per cent in 1986. The use of non-tariff barriers, however, remains extensive, with cases reported of shipments delayed or rejected (USDOS, 2001). Tunisia's import duties range from 10 per cent to 287 per cent according to the type of products. A consumption tax imposed on some luxury items or consumer products can reach a 500 per cent level. Tunisia 's value-added tax (VAT), consolidated into a single rate of 18 per cent, is computed on the base price of the goods plus any import duties, surcharges and consumption taxes. Agricultural products are generally assessed high import duties and in some cases face other import barriers, as will be seen in the next section. Tunisia still has non-tariff barriers, such as import licenses or quotas, on a number of products with the potential to compete against locally- produced equivalents, such as textiles. With respect to taxes imposed on foreign agricultural trade, Tunisia has a tariff system based on the nomenclature of the Organization for Economic Cooperation and Development (OECD): minimum, preferential and general tariffs (IMF, 2000). The preferential rate applies to imports from countries of the League of Arab States, provided they satisfy criteria of the rule of origin. The general tariff applies to goods from countries with no special commercial relations or to which the MFlV clause does not apply. Rates are ad valorem, generally in the 10to 43 per cent range, on CIF values. Some exemptions and deductions apply to a limited number of foodstuff imported under special temporary tax regimes. There is also a 3 per cent flat service fee on imports, regardless of their origin, applied on total import duties, including VAT consumption duty and anti-cyclical tax. The most recent development in foreign trade taxation with relevance to the agriculture sector since the passing of Law No. 95-6 of 23 January 1995 ratifying the URAA is Decree No. 2000-2445 of 20 October 2000, which reduced custom duties on a number of goods, among them: (a) Colza seeds for importers authorized by MOC for a quota not exceeding 4,400 tons; (b) Cheese to be used for transformation had the tariffs reduced to 27 per cent for a quota of 800 tons; (c) Animal feed ("luzerne") for a quota of 50,000 tons; (d) Barley feed, with no duties on a quota of 200,000 tons. Tariff rate and quota (TRQ) regulations are based on Decree No. 96-1 119 of 10 June 10, 1996.~TRQs are submitted for an authorization from MOC, called "authorization spCciale contingent tariffaire". The authorization remains valid for a period of two months and has to be returned to MOC if not used within the assigned period. A committee presided over by an MOC representative includes one representative each from the Finance, Agriculture and Industry Ministries, as well as from the Central Bank. MOC publishes a list of available TRQs, fixing the quantities, type of attribution and conditions for request submission. TRQs are allocated according to one or a combination of the following the three methods: (a) Method based on previous import patterns; (b) Method based on a chronological ordering of TRQ applications; (c) Method based on the proportion of quantities requested in the TRQ. If the committee finds none of those methods appropriate for a quota, it can decide to adopt any other method deemed relevant. E. ASSOCIATIONAGREEMENT BETWEEN TUNISIA AND THE EUROPEANUNION Trade volume and policy is largely shaped by an important bilateral trade agreement. Since March 1998, Tunisia has been bound by the Euro-Mediterranean Partnership Agreement establishing an association Republic of Tunisia, Official Journal, 2 1 June 1996. with the EU. The Agreement states that the EU and Tunisia are to gradually liberalize their reciprocal trade in agricultural and fishery products and that, from 1 January 2000, they need to assess the situation with a view to determining the liberalization measures to be applied by the two parties, effective 1 January 2001. Tunisia has obtained duty-free access to the EU market for commodities with a high share of agricultural exports, namely fish, seafood and dates. In contrast to the case of Morocco, the issue of Tunisia's increased agricultural access to the EU appears to be of limited importance, with the exception of olive oil. In fact, some of the limits set on a number of exports (e.g., citrus fruit and wine) have not been fully met (CIHEAM, 1999). In keeping with the spirit of the Euro-Mediterranean Agreement and the Barcelona process, the EU Commission was authorized to begin negotiations with Tunisia with a view to negotiating greater liberalization of agricultural trade. The issue of olive oil trade between the two partners was a key discussion item on the agenda. Steps were to be taken to avoid a break in the traditional trade in olive oil, as untreated olive oil originating in Tunisia entered the EU at a reduced rate of customs duty up to a maximum of 46,000 tons per year. A customs duty of Euros 7.811100 kg has been levied on imports into the EU during each marketing year of the period covered in the agreement. Following the most recent negotiations, Tunisia and the EU have agreed to amend the protocols related to their respective originating imports (Agriculture Protocols, Nos. 1 and 3). Tables 15 and /$show the arrangements for selected imports into the EU and Tunisia, to take effect starting 1 January 2001. F. ESTIMATES PROGRESS IN PRICE AND TRADE REFORMS OF Agricultural price interventions in Tunisia are implemented by using restrictions such as import quotas and export licenses, price guarantees and tariffs. It is difficult at present to find a detailed account of the level and extent of protection and export taxation or subsidization, owing to the use of a combination of policy instruments. Even when studies are carried out on the subject at the official level, there is limited public dissemination of the findings. What is needed in the case of Tunisia is a policy monitoring and evaluation project to tackle what seems to be a major gap in the analysis of agricultural trade and price policy. This project could replicate the experience of Latin American and Caribbean countries with surveillance projects (e.g., Valdes and Shaeffer, 1996). A quantitative evaluation of reforms undertaken by the Government should include the provision of policy indicators, such as nominal and effective rates of protection, producer and consumer subsidy equivalents and the effective rate of assistance. These rates could make the effects of agricultural and trade policies on agricultural incentives more transparent. Table 17 summarizes the information taken from various sources on those effects with regard to key agricultural commodities. It shows that cereal crop, livestock products and sugar had been heavily protected before 1995. With initiation of the liberalization process in agriculture, there seems to be a trend towards less protection for those commodities. That trend, however, would need to be verified by other measures of assessment. A number of trade and price policy reforms have occurred in Tunisia over the past twenty years (see box 1 for the major events that have shaped its policy landscape). Those reforms, whether domestic or external, had and will continue to have a major impact on protection trends. They will need continuous monitoring to enable policy makers to evaluate the direct and indirect income transfers that farmers and consumers are receiving. 1 Box. Main oolicv reforms in Tunisia 1 Date External measures Domestic measures 1976 Cooperation agreement with the EU. 1986 Stand-by agreement (IMF). ASAP implementation. Depreciation of the dinar. 1987 Revision of agreement with EU. Financial liberalization begins. Quota regulations on capital items lifted. Tariff reform. 1988 Lifting of quota regulations on consumer goods begins. ASAP implementation. 1990 Tunisia joins GATT. Income tax reform. 1991 Imposition of temporary complementary duties. Law on free prices and competition. Law on distribution. 1993 Investment code. 'OPending ratification by individual country parliaments. 18 1 1 Box (continued) Date External measures Domestic measures 1995 Ratification of URAA. Seek substantial recourse to credit because of problems with food subsidy programmes. 1996 Euro-Mediterranean Partnership Agreement with the EU Para-statals (Cereal and Oil Boards) begins to be implemented. seek substantial recourse to credit Unilateral dismantling of tariffs on some capital goods. because of problems with food subsidy Establishment of tariff rate quota regulations. programmes. 1997 9th Development Plan starts. 1998 Official agreement with the EU. Government waives all outstanding All import duties in list I eliminated. loans to farmers up to a 2,000 D ceiling Elimination of complementary duties on imports (rate was 10 per cent). Signatory of the Greater Arab Free Trade Agreement. Revised agreement with Morocco. 2000 Reduction of custom duties on animal feeds Loans to farmers up to a 2000 ceilings 200 1 New agreement with EU. Sources: IMF (2000), OECD (2000), BCT (2001). TABLE7. AVERAGEPROTECTION COEFFICIENTS OF KEY PRODUCTS IN 1984 (Percentage) Nominal Effective Durum wheat -20.0 -28.6 Soft wheat -07.3 10.2 Barley -18.7 -26.2 Citrus 27.5 37.2 Tomatoes -25.1 -30.4 Milk -0 1.4 15.8 Beef -03.3 -02.2 Lamb -12.3 -08.6 Source: World Bank. 1986. TABLE8. SHAREOF AGRICULTURAL INVESTMENT IN TOTAL ALLOCATION (Percentage) 5th 6th 7th Average Development Plan 1977-1981 1982-1986 1987-1991 1992-1996 1997-1999 Share in total investment 12.9 15.6 16.4 17.1 Sources: Abaab and Elloumi (1997) and Annuaire des Statistiques Agricoles (2000), Ministry of Agriculture TABLE9. DOMESTIC AND INTERNATIONAL PRICE RATIOS FOR THE MAJOR CEREAL CROPS IN TUNISIA Durum wheat Soft wheat Barley 1996 1.10 1.26 1.09 1997 1.14 1.40 1.10 1998 1.11 1.67 1.82 1999 1.54 1.86 1.58 Source: Ministry of Agriculture and Cereals Board. Domestic prices are farm gate, while border prices are average import prices. TABLE10. PRICING STRUCTURE FOR IMPORTED WHEAT R\I 1997 (D/ton) Durum wheat Soft wheat CIF price 246.16 180.82 Others charges including tariffs 46.16 34.33 OC margin 26.04 24.86 TotaI cost 320.9 1 240.01 Source: Lachaal. 2000. TABLE11. EVOLUTIONWHEAT PRICES AND PROTECTION LEVELS, 1995-1998 OF 1995 1996 1997 1998 Durum wheat guaranteed price (Dlton) 275 285 285 285 Soft wheat guaranteed price (Dlton) 240 250 250 250 Dumm wheat reference price (Dlton) 25 1 326 290 326 Durum wheat moving average price (~lton)' 218 251 273 297 Soft wheat reference price (Dlton) 190 246 222 196 Soft wheat moving average price (Dlton) 148 176 190 199 Durum wheat NPC (per cent) -9.5 - 12.5 -1.0 -12.5 Soft wheat NPC (per cent) 26 1.6 12.6 27.0 Source: Ben Cheikh (2000). a/ Moving averages are computed over a five-year period. TABLE12. EXPORTSOF OLIVE OIL BY TYPE OF OPERATOR (Tons) Season ONH Private Total exports Total production 1988 54 811 0 54 811 100 000 1989 42 680 0 42 680 54 000 1990 49 781 0 49 781 130 000 1991 156 358 0 156 358 165 000 1992 98 239 0 98 239 265 000 1993 110 518 0 110 518 135 000 1994 185 569 0 185 569 210 000 1995 104 195 2 000 106 195 70 000 1996 26 850 1 1 235 38 085 60 000 1997 57 927 53 000 110927 310 000 1998 88 500 28 500 117 000 90 000 1999 76 971 93 532 170 503 180 000 2000 115 000* 195 000 Source: ONH (1998) and Ministry of Agriculture ( 1 996, 1999,2000) Note: Two dots (..) indicate data not available. * Estimates as reported in BCT, 2001. TABLE 13. OLIVE OIL EXPORTS BY THE PRlVATE SECTOR FOR THE 1998- 1999 SEASON Number of exporters Range of volume (in tons) 3 1 1 000-16 000 Source: Ministry of Agriculture (1999). TABLE14. PRODUCERAND EXPORT PRICES FOR OLIVE OIL (D/ton) Producer price Producer price Premium to ONH Average price received Season ending year High grade Low grade deliveries by exporters (FOB) 1992 1 750 1 390 1 435 1993 1 750 1 390 1 445 1994 1315 1 055 1 582 1995 1315 1 055 300 2 402 1996 2 730 2 330 150 4 051 1997 1 800 1 420 2 289 1998 1 800 1 420 1714 1999 1 300 1 005 600 2 336 2000 2 300 2 000 Sources: Ministry of Agriculture (1999) and BCT (2001). Note: Two dots (..) indicate data not available. TABLE15. ARRANGEMENTS APPLYING TO SELECTED IMPORTS INTO THE EUROPEAN COMMUNITY OF AGRICULTURAL PRODUCTS ORIGINATING IN TUNISIA Agricultural protocols Amended protocols in new agreement (March 1998) 2001-2005 (b) (c> Tariff quotas Reference quantity (a) (Tons) (Tons) Cut flowers and flower buds 750 1 000 New potatoes 15 000 16 800 Almonds 1 000 1 120 Fresh oranges 31 360 35 123 Oranges, other than fresh 1 680 Apricots 2 000 2 240 Olive oil 46 000 50 000 Tomato concentrate 2 000 4 000 Apricot pulp 5 160 5 160 Mixtures of fruit 1 000 1 000 Wine of fresh grapes 179 200 179 200 hl Wine of fresh grapes with a designation of origin 5 600 56 000 hl Source: Chouchane, (2000) for column a and Commission of the European Communities (2000), annex I-A: Protocol No. 1 for columns b and c. TABLE16. ARRANGEMENTS APPLYING TO SELECTED IMPORTS INTO TUNISIA AGRICULTURAL OF PRODUCTS ORIGINATING IN THE EUROPEAN COMMUNITY Maximum customs duties-final customs duties Preferential tariff quotas (Percentage) (Tons) Live bovine animals 17 2000 Meat of bovine animals 27 8 OOO(**) Frozen poultry 43 400 Milk and cream 17 9 700 Butter 35 250" Processed cheese 27 450" Potatoes 43 16 500 Dumm wheat 17 17 000" Soft wheat 17-0' 230 000" TABLE16 (continued) Barley 17 12 000" Maize 20-0' 1 5 000 Rice 27-0' 4 000 Lucerne 29-0' 1 5 000 Vegetable oils 15-0' 100 000" Sugar (cane or beet) 15 72 000 Soybean (oil - cake) 20-0' 6 000 Source: Commission of the European Communities (2000), Annex I-B. Protocol No. 3. * The rate will be reduced to 0 per cent in five equal steps between 1 January 2001 and 1 January 2005. ** The quantities imported under the tariff quota opened by Tunisia within the WTO framework under the current access arrangements are deducted from the preferential tariff quota. TABLE17. A COMPARISON OF NOMINAL, PROTECTION ESTIMATES FOR SELECTED PRODUCTS Commodity NPC NPC Durum wheat 1.48" 0.88" Soft wheat 1.49" 1.27" Barley 1.37b1 1.70G1 Sugar 3.33h1 I .22G1 Milk 1.88h1 Meat 2 . 3 ~ ~ 0.95' Sources: a/ Ben Cheikh (2000), based on 1992 and 1998 protection levels. - b/ NabIi (1997), based on 1992 protection levels; C/ Author's computations based on prices for 1998 in Annuaire des SfatisfiquesAgricoles (2000). Ministry of Agriculture. Note: Two dots (..) indicate data not available. Figure X. Trends in agricultural support and sector contribution to GDP Sources: Author's computations from data compiled from various issues of BCT annual reports; Budget Economique, Ministry of Economic Development; and Annuaire des StafisfiquesAgricoles, Ministry of Agriculture, 1972-1999. Note: Agricultural GDP includes agriculture, fisheries and agro-food industries. By support, we intend the disbursements made by the General Equalization Fund (CGC) to these three sub-sectors. Figure XI. Evolution of tariff revenues from agricultural imports - - -- - -- -Ag Tar~ffsIAgImports T o t a l TarlffsTTotal Imports , - -- Source: Author's computations from data compiled from various issues of BCT annual reports; Budget Economique, Ministry of Economic Development; and Annuaire des Statistiques Agricoles. Ministry of Agriculture, 1972-1999. Note: Agricultural tariffs include those collected on imports of primary agricultural commodities, fish and agro-food industrial products. 111. EXPERIENCES FROM THE IMPLEMENTATION OF THE URUGUAY ROUND This section attempts to provide an account of the extent of implementation realized by Tunisia with respect to the "three pillars" of the Uruguay Round Agreement on Agriculture. A. MARKETACCESS For Tunisia, the market access issue should be looked at from two angles: access to Tunisia's home market by exporters from other countries, on the one hand, and access by Tunisian exporters to foreign markets, on the other hand. We will focus here on the former dimension of the problem, as the latter dimension has already been covered in the context of Tunisia's Euro-Mediterranean Partnership Agreement with its main trading partner, the EU. 1. Most-favoured-nationand the post- UruguayRound tarijTregime1' High protection for agricultural commodities in the form of tariffs continues to be one of the major factors restricting world trade. The global average tariff has been estimated at 62 per cent (USDA, 2001). Are Tunisian farmers benefiting from tariff protection or are they losing income because of lower prices resulting from those same tariffs? To help identify some elements of the answer, we report some tariff characteristics and indicators, including the scope and levels of bound rates, the gap between mean bound and applied rates, tariff line intervals, tariff dispersion and transparency. The levels of tariff protection described next refer to MFN bound tariffs. While one of the main goals of the Uruguay Round was to increase the scope of tariff bindings, Tunisia's bound lines in its tariff schedules was only 53 per cent, a number way below that of most developed countries and lower than the percentage of bound lines for developing nations, such as the Philippines, India, Malaysia or Thailand (between 63 per cent and 73 per cent). This is partly caused by the gap between product groups. As shown in table 20, the process of tariffication in the agricultural sector is almost complete, with bound tariffs of more than 95 per cent established. The extent of bindings is much lower in industry, where Tunisia has yet to tariffy about 53 per cent of its industrial lines. This contrasts with the difference of almost 76 per cent between the simple bound means in the two sectors. Tunisia is also a country that has bound its tariffs at levels that were on average 30 per cent higher than the applied MFN rates. The difference is very significant in the specific case of agriculture products (between 75 per cent and 95 per cent in the four agricultural and food commodity groups, as shown in table 18), with an average of 30 per cent across all lines. 2. Post-UruguayRound applied rates As shown previously, Tunisia maintains high bound tariffs in its WTO market access schedules. However, the actual applied tariffs that it imposes are lower than the tariff bindings. Figure 12 compares the final WTO bound tariffs with 1995 and 1998 applied tariffs for Tunisia and an average tariff indicator for a group of WTO member developing nations.I2 The final bound tariffs for Tunisia, as well as the average for the sample considered, are considerably higher than the respective mean applied tariffs in 1995 and 1998. Applied tariffs tend to be less than one third of the average bound tariffs. While the patterns for Tunisia fit well in the overall picture for a comparison of bound and applied tariffs, the consistently higher protection rates for Tunisia could constitute an area of concern for the country's trading partners. As is the case for all " This sectiondrawsfromthetariff characteristicsand indicators computed using variousdatabases, including AMAD,and reported in OECD (1999) and Gibson et al(2001). ''The data used for the computations in figure XI1 are drawn from table 5 in Gibson et a1 (2001). The selected sample of developing countries includes, in addition to Tunisia, the following WTO members: Argentina, Brazil, Colombia, CostaRica, Ecuador, El Salvador, Guatemala, India, Indonesia, Republic of Korea, Mexico, Morocco, Nicaragua, Pakistan, Panama, Paraguay, Philippines, Thailand, Trinidad and Tobago, Turkey, Uruguay and Venezuela. The average bound tariff represents the average tariff binding to be effective by the implementation period (1995-2004). countries having bound tariffs above currently applied rates, applied tariffs may be raised above published rates without necessarily informing or compensating the affected trading counterparts. The Tunisian Government has a great deal to manoeuvre if it wants to dampen the effects of world price fluctuations on its agricultural sector and food consumption. However, it should bear in mind that the replication of such behaviour by several other countries could introduce even more instability in the world market, the same instability that Tunisia wants to protect itself from. Against a relatively "high" average applied tariff rate of 35 per cent (for all lines) compared to a global world avera e of 17 per cent (Gibson, 2001), Tunisia's applied tariff profile for key imported agricultural commoditiej3 exhibits some interesting patterns, as shown in table 19 and figure 13. First, the unweighted average for the key commodities (referred to as "strategic" commodities in policy-maker jargon) is only 26.2 per cent. However, if tariffs are weighted by the respective volume in value, the average drops to 18.7 per cent. This 7 per cent difference shows that tariffs are lower for the most heavily imported commodities, which include soft wheat, durum wheat, soybean oil, sugar and corn. There exists, indeed, a strong correlation between the share in imports and the level of applied tariffs, as shown in figure 14. Second, the selected profile reported in table 19 and figure 13 shows limited variation in tariff levels by commodity, with a standard deviation of 10.6 per cent. There seem to be some broadly significant common features in the clustering of products into three distinct groups. The low-tariff group (less than 20 per cent) includes most of the cereal crops produced locally (wheat and barley), in addition to milk, soybean oil and sugar. The medium-tariff group (between 25 per cent and 35 per cent) has several livestock-related products (beef, cheese, butter and cows), while the high-tariff group (more than 40 per cent) includes sheep, corn oil and wheat flour. The same type of grouping analysis could be made by taking into consideration the weight in imports for each commodity, an approach that is not adequately emphasized when reporting average tariffs. Table 20 shows the average applied tariff and the share of three distinct groups of commodities. It clearly portrays the major thrust of Tunisia's tariff protection policy, with the most heavily traded commodities facing lower tariffs than the commodities with marginal importance in total imports. It is quite interesting to see that in Tunisia's applied tariff profile, the goods that tend to be more consumer- oriented and do not necessarily compete against locally-produced items have higher tariff rates than the goods with a competition potential, such as sugar, milk and grains. This could have important public finance implications in terms of the collection of tariff revenues if further cuts in tariffs are imposed on Tunisia, whether in a multilateral or bilateral setting. 3. TarzfSrate quotas The tariffication process negotiated at the Uruguay Round allowed countries to use, for certain agricultural products, tariff rate quotas (TRQs) consisting of an in-quota tariff (IQT) applied for imports within the quota and an over-quota tariff (OQT) applied on imports in excess of the quota volume. Tariffication provides also for two types of TRQs: minimum access and current access. The quantity allowed to be imported at the lower tariff or minimum access level was set at 3 per cent of the 1986-1988 base period consumption levels and to be increased to 5 per cent by 2005 for developing countries. If imports happen to exceed 5 per cent of domestic consumption in the base period, current access was to be granted to those products. In the agricultural and food chapter, Tunisia has 36 lines in its tariff schedules subject to an IQT and 36 other lines with an OQT. All lines were expressed on an ad valorem basis or had an ad valorem equivalent computed. As reported in table 18, the mean tariff in agriculture is around 117 per cent, with a mean IQT of 26 per cent and a mean OQT of 100 per cent. Patterns of protection with quota- based tariffs for specific agricultural commodities (figure 15) indicate the importance held by staple foods in Tunisia's trade policy and reflect the country's objective to protect those sensitive commodities. Compared to other countries that scheduled IQTs at very high levels (e.g., Morocco, Columbia, Malaysia, Switzerland and Indonesia, with IQTs between 65 per cent and 150 per cent, as reported in Gibson I ; The subset of commodities looked at includes those goods with more than 1 million D in import value. Taken together, they account for approximately 70 per cent of agricultural and food imports. Data are based on trade statistics from Statistiques du Commerce Exterieur, 1998, Institut National de la Statistique. et al, 2001), Tunisia's average of 26 per cent has, relatively speaking, followed the original aim of IQTs to allow market access for a limited quantity of imports at low tariffs. Table 21 provides some features in the use of TRQs by Tunisia, compared to patterns for an average based on a sample for 35 countries. It shows that while it is doing better than several other countries in terms of TRQ use, Tunisia's leveIs are still high compared to the numerous countries that are not using TRQs. This is an area where Tunisia could certainly improve its standing. 4. Tar@-dispersionin Tunisia's UruguqRound schedule To get a fair assessment of applied patterns of protection and the extent of the distortion effects of tariffs, mean tariff indicators have to be complemented with measures of dispersion. The degree of dispersion is either assessed with tariff peaks or by the standard deviation from the mean tariff. For the first indicator, a large number of peaks suggests a highly differentiated tariff schedule, as opposed to a more uniform tariff structure when a low level of tariff spikes is included (OECD, 1999). When using a national reference level for tariff rates (three times the national mean tariff), Tunisia's schedule shows a relatively "flat" structure, as the ratio of lines for which the tariff rates exceed the reference level to the total number of lines is 2 per cent. However, when an "international" reference level is used (15 per cent), the totality of the schedule exceeds this threshold. What is most relevant to this study is the fact that the largest number of peaks is in the agricultural sector, where all tariffs are above the "international reference" line. The other measure of dispersion, the standard deviation within products, is reported in table 20 for different tariff chapters and groupings. Tunisia has one of the highest dispersion levels, with an overall standard deviation measure of about 40 per cent -just behind India (42 per cent), Republic of Korea (44 per cent) and Norway (70 per cent) - a pattern mostly driven by the high standard deviation in live animals, fats and oiIs and dairy products. B. DOMESTIC SUPPORT Because of the recognition that domestic policies could distort trade, the Uruguay Round Agreement on Agriculture included rules on domestic support. Furthermore, the fact that the WTO has extensive rules governing domestic as well as border policies attests to the special status generally given to agriculture. Agricultural support in Tunisia (table 25), though relatively high compared to support in other sectors, does not regard the relation between the Uruguay Round rulings and their impact on domestic agricultural policies an area of primary concern. Tunisia is more concerned about how those rulings will affect the policies of its main trading partner, the EU. Nevertheless, domestic support as measured by the aggregate measure of support (AMS), aggregated across all commodities, should not exceed 59.3 million D in 2004 (Chouchane, 2000). Tunisia should not have difficulties maintaining support in a way consistent with its WTO commitments, as those same measures (reduction in output price support for grains and irrigation water subsidies) have been an integral part of the ASAP put in place in 1986. Table 22 reports trends in the support level reported to WTO. It shows that even though there is no downward trend in the pattern of support, Tunisia stayed in line with the commitments made. Support tends to be correlated with climatic conditions and trends in international prices. The bulk of the support has been in the main cereal crops (wheat and barley) and milk as was shown in the product specific reporting by Tunisia to WTO (table 23 for the 1995 measurement of support and table 24 for the 1999 figures). 'The low aggregate measure of support for 1999 is due to the reduced support in durum wheat and milk, as a result of an increase in the reference price used for the specific product measurement. C. EXPORTSUBSIDIES Tunisia does not provide direct subsidies for its agricultural and food exports. However, some subsidies are given in the form of reduction in marketing costs, mostly related to transportation (as allowed by article 9.4 of the Uruguay Round Agreement). Those subsidies for 1999, in the approximate amount of 6.7 million D (as reported by the Tunisian Centre for the Promotion of Exports), covered citrus products (5.5 tons), potatoes (3.9 tons), wine (560,000 bottles), dates (13.3 tons) and tomato paste (19.3 tons). This is an amount that could easily revert to Tunisia and be spent on promotional campaigns or on improving the access of information to exporters if those subsidies became restricted. D. EVOLUTION APPLIED TARIFFS SINCE THE URUGUAYROUNDAGREEMENT OF When comparing the applied tariffs in 1995 and 1998 (table 26), there does not seem to be any major change in the agricultural commodity schedule. There is a slight increase in the unweighted average, accompanied by a decrease in the dispersion of the tariffs. TABLE18. POST-URUGUAYROUNDBOUND TARIFF LINES BY SECTION Percentage of - Bound mean Difference in MFN Product group bound tariff lines tariffs bound tariffs* Standard deviation Agriculture 97.1 1 16.7 40.4 ~ sanimals and products e 56.4 116.7 Vegetable products 100.0 128.1 Fats and oils 92.7 120.5 Prepared food 93.4 113.4 Industry 46.3 41.2 All lines 52.8 59.0 -30.2 40.5 Source: OECD (1999). Two dots (..) indicate data not available * For the difference between MFN and post-Uruguay Round bound tariffs, a negative value indicates that MFN rates were lower than post-Uruguay Round bound rates. Figure XII. Comparison of bound and applied tariffs for Tunisia and an average for selected developing countries 0 Bound 1995 1998 ElTun~sia A v e r a g e Source: Gibson et a1(2001). Note: Averages for bound and applied tariffs are unweighted averages. Figure XIII. Applied tariffs in 1998 for selected agricultural and food items Corn oil Wheat flour Bananas Chickpeas Potatoes Live sheep Sesame Butter Live cows Peanuts Rice Cheese Beef meat Tobacco Palm oil Soft wheat Eggs Chicks Corn Barley 3urum wheat Sugar Soybean oil Tea Coffee bes for seeds Milk 0 5 10 15 20 25 30 35 40 45 50 Tariff Rate Source: AMAD (2000) TABLE19. APPLIEDTARIFF EXPRESSED IN AD VALOREM TERMS FOR SELECTED AGRICULTURAL AND FOOD PRODUCTS IN 1998 Product Weight in agro-food imports Bound tariff Applied tariff Live cattle 1.17 150 3 1 Live sheep Chicks Beef meat Milk Butter Cheese Eggs Potatoes for seeds Other potatoes Chickpeas Bananas Coffee TABLE1 9 (continued) Product Weieht in aero-food irn~orts Bound tariff Applied tariff Durum wheat 14.44 80 Soft wheat 20.83 100 Barley 1.97 75 Corn 9.41 17 Rice 0.58 60 Wheat flour 0.53 150 Peanuts 1.24 200 Sesame 0.74 150 Soybean oil 15.38 17 Palm oil 1.02 75 Corn oil 1.24 30 Sugar 14.57 100 Tobacco 1.08 75 Llnweighted average 113.3 Weighted average 80.1 - Sources: AMAD (2000) for applied and bound tariffs and INS Sfatisfiquesdu Commerce Exfericur (1998) for value of imports. Weights are author's computations. Note: Weights have been computed based on value of imports for these products. This partial list represents about 70 per cent of the total trade of agricultural and food items. Figure XIV. Relation between applied tariff levels and share in major agricultural imports 5 0% 100% 15 C?/o Weight in imports Source: AMAD (2000) plus INS import data (1998). TABLE20. APPLIEDTARIFFRATES AND SHARE IN IMPORTS, 1998 (Percentage) Commodity group Group average applied tariff Share in imports Group I 34 6 Butter, chicks, eggs, cheese, live sheep. wheat flour. potatoes, rice. sesame, chickpeas, bananas Group 11 23 20 Palm oil, tobacco, beef meat, live cows, corn oil. peanuts, milk, barley, coffee, potatoes for seeds, tea Group I11 17 74 Corn, durum wheat, sugar, soybean oil, soft wheat Source: Author's computations based on INS statistics for 1998 import levels and AMAD database for tariff rates. Figure XV. Ad valorem tariffs on staple foods (Computed using worldprices) 0 wheat barley sugar beef skim milk whole milk butter nce tobacco coffee Items Source: AMAD (2000). TABLE21. TARIFFRATE QUOTAINDICATORS IN TUNISIA TRQ indicator Tunisia World average* Average tariff (all lines) 110 49 Average in-quota tariff (TRQ lines) 26 63 Average over-quota tariff (TRQ lines) 109 128 Number of tariff lines w/o TRQs 286 Number of tariff lines with in-quota tariffs 14 Number of tariff lines with over-quota tariffs 43 TRQ lines as per cent of total 17 23 Source: ERS-USDA as reported by Gibson et al (2001). * Average over a sample of 35 countries, using TRQs in final URAA. Two dots (..) indicate data not available. TABLE22. TOTALAGGREGATE MEASUREMENT OF SUPPORT COMMITMENTS BY TLNISIA,1995-1999 Total AMS commitment level Current total AMS Share of commitment level used (a) (b) (Percentage) (Millions of D) (b)/(a) 1995 67.49 58.6 87 1996 66.58 5I .O 77 1997 65.67 53.3 81 1998 64.76 61.1 94 1999 63.85 29.3 46 Sources: WTO Committee on Agriculture, "Domestic support", background paper by the Secretariat, (GIAGMGISII, 13 April 2000) and "Notification by Tunisia" (GIAGMITUNII8, 12 January 2001). TABLE23. PRODUCT-SPECIFIC AGGREGATE MEASUREMENTS OF SUPPORT IN 1 9 9 5 Applied External Eligible Total market price Description administered price reference price production support of basic products Measure type (m) ( E M ) (Ep) (AAP-ERP) EP Dlton Dlton 1,000 ton Millions of D Durum wheat Producer buying price 275 20 1 349 25.8 Soft wheat Producer buying price 240 152 38 3.4 Barley Intervention price 150 104 15 0.7 Milk Intervention price 330 248 278 22.8 Olive oil Intervention price 1 185 2 401 4 -4.9 Sugar beet Intervention price 38 40 268 -0.6 Source: WTO Committee on Agriculture, "Notification by Tunisia" (GIAGMITUNI3, 15 November 1996) Notes: Tunisia used the external reference prices for the reference period and adjusted them to take account of inflation and the movement in the exchange rate. Associated feesllevies are included in the computations. For 1995, those fees were equal to zero. TABLE24. PRODUCT-SPECIFIC AGGREGATE MEASUREMENTS OF SUPPORT IN 1 9 9 9 Applied administered External reference Eligible Total market price Description price price production support of basic products Measure type (AAP) (ERP) (EP) (AAP-ERP)EP Dlton Dlton 1.000 tons Millions of D Durum wheat Producer buying price 285 261 345.7 8.2 Soft wheat Producer buying price 250 205 66.7 3.0 Barley Interventionprice 170 141 8.5 0.2 Milk Interventionprice 340 293 385 17.9 Olive oil Interventionprice 1,857 1,860 I I 0.0 Sugar beet Interventionprice 38 42 89 0.0 Source:WTO Committee on Agriculture,"Notification by Tunisia" (GIAGM/TUNI18, 12January 2001) Notes:Same as table 23 TABLE25. DOMESTICSUPPORT IN THE BASE PERIOD AS NOTIFIED BY TUNISIA, 1995- 1 9 9 9 (Units = Millions of D) Base period 1995 1996 1997 1998 1999 Durum wheat 23.7 25.8 30.8 18.9 26.2 8.2 Soft wheat 10.4 3.4 10.5 6.8 10.7 3.0 Barley 7.3 0.7 4.6 8.1 2.7 0.2 Milk 9.5 22.8 18.2 19.5 21.5 17.9 Olive oil -1 1 -4.9 -22.6 Sugar beet 3.6 -0.6 -0.6 Poultry meat 7.4 Beef 3.2 Tomatoes 0.5 Potatoes 1.6 1.5 1.7 9.0 Citrus fruits 1.2 Non-product-specific 9.9 8.3 7.1 7.2 8.3 Sources: WTO Committee on Agriculture, "Domestic support", background paper by the Secretariat (GIAGMGISII, 13 April 2000) and "Notification by Tunisia" (GIAGMITUNI18, 12 January 2001). TABLE26. APPLIEDTARIFFS IN 1995 AND 1998 HS class Applied tariffs (2 digits) Group of commodities 1995 1998 8 Fruits 43.00 43.OO 16 Meat preparation 43.00 43.00 20 Vegetable and h i t s preparation 43.00 43.00 43 Leather products 43.OO 43.00 11 Milling industry products 41.40 41.80 21 Various food preparation 40.40 4 1.20 7 Vegetables 42.05 40.84 22 Beverages and alcohol 39.88 40.24 2 Meat 38.80 38.90 19 Cereal preparation 37.29 38.86 9 Coffee and tea 37.65 37.65 18 Cocoa preparation 35.64 37.45 33 Essence oils 34.53 36.65 24 Tobacco 43.00 36.00 6 Plants and flowers 33.55 34.64 15 Oils and fats 32.43 33.93 1 Live animals 31.91 33.36 17 Sugar and sweets 29.70 31.40 10 Cereals 29.38 30.17 12 Oil seeds 25.75 27.94 5 Other animal products 26.16 27.84 4 Dairy 26.69 26.86 23 Food industry residues 24.58 26.58 14 Alfa and vegetables 22.42 25.08 4 1 Skins and leather 18.67 21.83 13 Tanning products 18.38 2 1.54 38 Various chemical products 20.00 2 1.50 35 Albumin products 17.75 20.00 29 Organic chemical products 17.00 20.00 50 Silk 17.00 20.00 53 Other textile fibres 17.00 20.00 52 Cotton 17.17 19.33 51 Wool 17.OO 19.00 Summary unweighted average 34.40 35.16 Statistics standard deviation 1 1.27 10.19 CV 0.33 0.29 Source: Author's computations based on AMAD Database, Release 1.O, September 2000. IV. QUANTITATIVE EVALUATION OF POLICY OPTIONS, TRADE-OFFS AND STRATEGIES To evaluate policy options, trade-offs and strategies in a quantitative setting, various methodologies related to agricultural pricing policy reforms and trade liberalization in Tunisia are used. The outcomes for Tunisia are also compared with two other WTO member countries in the Middle East and North Africa region (Morocco and Jordan) to be able to assess the direction and magnitude of the changes generated by policy reforms. First, we present a simple partial equilibrium evaluation of tariff-cut formulas for key agricultural markets in Tunisia. Because of intersectoral linkages between markets, the analysis is refined by using a multi-market framework to evaluate the impact of trade policy reforms for Tunisia, as well as for Morocco and Jordan. Domestic pricing reforms with implications on trade, for Tunisia only, are also looked at. We then report on the use of computable general equilibrium (CGE) modelling for the analysis of bilateral and multilateral trade liberalization and the potential implications for the agricultural sector in Tunisia. A. WELFAREANALYSIS OF TARIFF FORMULA REDUCTIONS This section applies a single-market framework to specific instances of price intervention at the border level to illustrate the range of effects arising from the application of various tariff formula reductions on the major food imports in ~unisia.'~With elasticity estimates and price and quantity data, we estimated the financial implications of a change in a set of commodity prices, the welfare transfers between producers and consumers, and the net gains and losses in economic efficiency. The approach is based on the conventional welfare analysis of price interventions in agriculture as described in Tsakok (1990) or Tweeten (1992). Table 27 shows the data needs and the markets analysed. Four tariff reduction formulas were considered for tariff reduction in earlier GATT rounds (USDA, 1998). The tariff-cut formulas (TCFs) are: (i) a straight linear cut; (ii) a linear harmonized, which is a linear reduction with an additional harmonization adjustment term; (iii) a deep harmonized, which is similar to the linear harmonized but can have deeper cuts in high tariff rates; and (iv) the Swiss formula, which applies an upper bound on all tariffs. Each tariff reduction formula implies a different tariff rate, but the objective is to find the formula that subjects higher tariff rates to larger cuts; i.e., harmonize the tariff structure. The formulas are defined as follows: (a) Linear cut (LC) t, = a t, (b) Linear with harmonization (LH) (c) Deep harmonization (DH) (d) Swiss formula (SW) Where th is the new tariff rate, this the initial tariff rate, and 0 5 a I 1 is a parameter, b is an additional harmonization term, c is a constant greater than the highest applied tariff rate, and e is an upper bound on all computed t, using the Swiss formula. Simulations were made by varying the values of the coefficients in the TCFs (see tables 28 and 29). One hundred different outcomes, in terms of tariff changes and their resulting impact on domestic prices, were used to compare the impact of TCFs on net welfare effect (N.W.E.) for the markets selected and were computed as: N.W.E. = gains to consumers- losses for producers - losses for government To be able to compare outcomes of the different TCFs and make appropriate ranking, a benchmark had to be used in terms of the resulting impact on the various tariffs. We chose to harmonize all tariff changes by using the resulting average price change for the commodity set selected. For example, LC (a=1/2), LH (a=1/3), DH(c=133) and SW (d=65) all lead to an average reduction of 20 per cent in the domestic price. Because of the different impact on the individual tariffs, the different TCFs lead to a l 4This work was specifically carried on in the context of the WorId BankIESCWA preparatory research and policy analyses on country-specific studies. different welfare gain outcome. We chose then to look for the highest welfare improvement that could be achieved for the same average price change. Figure XVI shows the relation between average price changes and net welfare gains at the national level, assuming a set of elasticity measures taken from previous studies (e.g., Lachaal et al, 1998). This set of simulations is referred to as "base" or "reference". The impact of elasticity values was also analysed by doubling the supply elasticities in a second set of simulations. We then combined that second set with a 100 per cent increase in demand elasticities to form a third set of simulations. Table 30 summarizes the results obtained for the four TCFs, under three sets of elasticity levels for three levels of price changes (15 per cent, 25 per cent and 35 per cent). Figures XVI-XVIII show the complete range of price changes and their impact on net gain outcomes for the various elasticity sets assumed. The key features and trends that emerge from this evaluation are the following: (a) As expected, the higher the price change as a result of deeper cuts, the higher the welfare gains; the change tends to be exponential rather than linear with an average increase of 175 per cent between the 15 per cent and 25 per cent price change levels, and 92 per cent between the 25 per cent and 35 per cent levels; (b) The higher the behavioural flexibility in supply and demand, the higher the welfare gains, though a change in supply elasticity tends to have a greater effect on welfare gains than changes in demand elasticities; roughly speaking, a doubling of the elasticity level leads to twice the welfare gains; (c) The higher the price change and the higher the elasticity of demand and supply, the higher the difference in performance between the DH formula and the other three formulas, with the lowest performance always obtained with the linear cut reduction. The LH and SW formulas tend to have a comparable performance in general. The deep harmonization cut, compared to the linear cut reduction, would provide from 1 to 3 million D at 15 per cent price change, 4 to 8 million D at 25 per cent price change and from 8 to 14 million D at 35 per cent price change. This advantage of the deep harmonization cut starts fading away once the average price change passes the 34 per cent mark to move towards the highest possible price reduction, which is the zero-tariff level situation and corresponds to the first-best full liberalization situation. This point of inflexion where the deep harmonization line changes curvature in figures XVI-XVIII is the level where the parameter of the formula is closest to the highest applied tariff in the schedule; (d) While these findings for the different TCF are well known, this evaluation tried to show the magnitude of the gains that could be realized if one formula is chosen for the set of commodities looked at. Three different areas for the different curves emerge. For changes in tariffs resulting in less than 20 per cent change in average price changes, there is no substantial difference in net welfare gains between the different TCFs. At most, DH would provide a slightly over 1.5 million D extra gain over the other TCFs. For price changes in the 20-45 per cent range, the DH formula provides a higher welfare gain with respect to the other three formulas. The gap in net welfare gain is highest when the constant in the DH formula is equal to the highest tariff rate in the schedule, which happens to be for meat. With an average price reduction of around 31 per cent, this gap in welfare gains is no less than 14 million D. Figure XIX shows on the same graph the lowest and highest performing formulas for the three sets of elasticities assumed. It shows that the higher the flexibility in supply and demand response, the greater the difference between the different TCFs for the intermediate range of price change. It also suggests that the farther we get from the zero-tariff situation or the high protection case, the more sense it makes to pick the right parameters and the right formula for cutting tariffs. B. WELFAREANALYSIS OF COMPLETE IMPORT AND EXPORT LIBERALIZATION In this section, we use partial equilibrium methods for calculating the welfare effects of trade policies'5 applied by Tunisia for its own imports and those policies principally followed by the EU as the major trading partner for the country's main exports. We next provide some estimates for the economic effects of Tunisia's l 5By ignoring the price and quantity changes for all but a few commodities, the methods of course have their shortcomings. However, given the limited data requirements and the difficulties associated with data collection by other methods, partial equiIibrium estimates have been used widely to calculate the distributional and net welfare effects of tariff and non-tariff barrier elimination. own trade policy reform, as well as estimates for reforms in EU market access for some key agricultural exports. The purpose is to provide an approximation of the magnitude of gains and losses arising from actions taken by the Tunisian Government and compare them with those of other trading partners. Table 3 1 shows the income transfer and efficiency costs associated with removing MFN tariffs on selected agricultural commodities in ~ u n i s i a . 'One feature to note from those estimates is the small size of ~ the net welfare effect compared to that of the income transfers, a common feature noted in many studies on the costs of protection (Bowen et al, 1998). On aggregate, the gain is slightly over 10 per cent of total consumer surplus, with the largest relative gain achieved in barley (18 per cent) and the lowest in sugar (4 per cent), though in absolute terms the highest gains are to be found in durum wheat and meat. Of interest also is the wide variation in tariff revenue losses across commodities as a percentage of the respective income transfers. The most important losses will arise in the soft wheat and vegetable oil markets (respectively, 28 per cent and 26 per cent of total losses), clearly indicating why those two markets have lagged behind in liberalization progress from the perspective of the Government. Finally, the second column in the table lists the fraction of the total change in consumer surplus arising in the import market and shows the different impact on importers that further trade liberalization could have. The impact on the various players affected by the reduced tariffs is better shown by looking at those same numbers expressed in relative terms to the total commodity basket, as summarized in figure XX. It clearly indicates that the stakes are different, depending on the product and the agent affected (consumer, producer or government). Those welfare gains should be compared to those that could arise from the EU removing some of the constraints faced by Tunisia's exporters on a set of major agricultural commodities. What is assumed is the equivalent of a voluntary export restraint imposed on Tunisia as a part of its commitment to a free-trade agreement with the EU. It is as if the importing country, in this case the EU, is asking from the exporting country to limit the supply of a number of commodities for which it can claim entry at a preferential tariff. Table 32 shows estimated net welfare gains from removing barriers to preferential entry into the EU on specific products exported by Tunisia. The estimation assumes a complete liberalization process leading to a complete elimination of ad valorem tariff equivalents. The results provide evidence on the importance of the restrictions faced by Tunisia when it comes to exports of olive oil. Most of the welfare gains are, in fact, captured in that market. Given the difficulties associated with the estimation of export supply elasticities and the extent of trade distortion arising from the imposition of quotas, the uncertainty about their 'true' value or level was dealt with by incorporating lower and upper values for the relevant estimates. With this type of reform in the policy of its main trading partner, Tunisia could therefore gain between 24 and 62 million D. When compared to the 26 million D Tunisia could gain by eliminating tariffs on its imports, there seems to be a rather strong argument for seeking better market access terms for its own exports, in particular those of olive oil. From the perspective of Tunisia, the Government should realize that a strategy based on requesting better market access for olive oil as a counterpart of eliminating tariffs on food imports could yield potentially higher payoffs for the country. If lower distortions are imposed on its exports, Tunisia should be looking at better ways to market its food exports, as more value could be added to processed olive oil, citrus- based products and other fruits. C. CROSS-COUNTRY COMPARISONS FROM THE MULTI-MARKET ANALYSIS OF TKADE LIBERALIZATION IN JORDAN, MOROCCOAND TUNISIA The previous analysis was based on the aggregation of individual welfare gains and gave limited attention to the potential interactions in existing markets on the supply side and commodities consumed on the demand side. This has its limitations, considering the important crop-livestock linkages existing in the Tunisian agricultural sector. For example, barley is one commodity with various linkages on the demand and supply side; it is used as a feed in the livestock sector and at the same time is a food item in the diet of many Tunisians. Straw from durum and soft wheat is also an important feed source for cattle and sheep. Therefore, reductions in tariffs and changes in domestic support schemes in one market tend to have important indirect effects on other markets through complementarity and substitution patterns. The partial equilibrium analysis of tariff reduction can be taken one step further by looking at pricing policy changes in a multi-market setting, allowing for linkages between various agricultural markets. Research carried out by l 6These estimates assume that imported and domestically produced goods are perfect substitutes various regional scientists involved in the ICARDA-IFPRI Policy and Property Rights project" has permitted analysis of the impact of agricultural policy reforms and trade liberalization on different agro- ecological zones, with particular emphasis on commodities produced in low rainfall areas (wheat, barley, olive oil, forage, sheep and cattle). Using a multi-market modelI8for Morocco, Jordan and Tunisia, Chaherli (2001) summarizes results from a cross-country analysis of various pricing policy reforms in agriculture and their impact on national welfare, production, consumption and government budget, as well as trade balance.19 Some of the most important results with relevance to the World BanWESCWA preparatory research project are presented next. Table 33 shows the type of scenarios simulated for the three countries. Tables 34 and 35 present some aggregates at the national and specific agro-ecological zone level. The most important findings are the following: (a) Welfare efects. When examined at the national level, all combinations of reductions in distortions tend to generate positive welfare effects, as it can be seen from the Hicksian Equivalent Variation (HEV) measures in table 34. Relatively large welfare gains in the three country cases arise, because consumers in the respective economies benefit from lower consumer prices as a result of lower tariffs on food-related commodities. Depending on the type of scenario, the welfare improvement ranges from .17 per cent to 1.36per cent of the original level of income for the three countries. However, a mild liberalization in output and input prices (scenarios 1 and 2) does not generate substantial welfare improvements in Morocco, as opposed to the cases with a more pronounced pricing policy change (scenarios 3 and 4). At least, there does not seem to be a worsening in the welfare of the three countries as a result of those reforms; (b) Income eflects. The positive welfare effects at the national level reported in table 34 are accompanied by a reduction in national agricultural income. The results are consistent across the three countries, but differ in their magnitude for income growth measures. The reduction in output protection and input subsidies for the selected commodities has a higher impact in Morocco, at a loss ranging from 1.63 per cent to 3.29 per cent. For the other two countries, the loss ranges from 0.76 per cent to 1.60 per cent in Tunisia and from 0.07 per cent to 0.23 per cent in Jordan. The different magnitudes tend to follow the relative importance of the agricultural sector in each economy; (c) TarifSreduction policy versus input policy. Input price increases tend to be welfare-reducing, whether with a 10 per cent or 20 per cent reduction in output prices. Reducing input subsidies by 10 per cent produces an additional welfare reduction (in terms of Hicksian equivalent variation) of around 0.1 per cent of income and a reduction of the same amount in terms of income growth in the three countries. However, in relative terms (by comparing with and without input price changes), the incremental impact of input pricing changes is higher in Morocco and Jordan than in Tunisia. This is partly due to the higher importance of feed and fertilizer subsidies in the former countries. In summary, we find that national welfare improves in the three countries, even though pricing policy changes lead to a drop in income. Overall, and as expected from the nature of pricing policy reforms, consumers are better off, while agricultural producers are worse off. The cross-country results presented above suggest that the process of output and input price policy reform improves the welfare of the whole nation, even with losses in income. The source of the positive welfare change is that welfare gains obtained by consumers are higher than the losses generated by reduced output prices and lower input subsidies. The HEV measures allow these two effects (consumption and production) to be put together and compared with the direct evaluation of changes in a money value of welfare change. " See Chaherli et a1(1999% 1999b)for a description ofthe project, as well as for a summaryof some related studies on agricultural policy reforms in Jordan, Morocco and Tunisia. '' A description of the model and associated methodology is provided in annex 11. See Braverman et al (1987a-c) and Sadoulet and de Janvry (1 995) for more details on the methodology and specific country studies using multi-market models. l 9Even though Morocco and Jordan are not currently covered by the World BankIESCWA preparatory research project, we believe that a comparison of Tunisia with those two countries is relevant to the project and could provide some insights that could be generaIized for the entire region, given the similarities existing in the types of agriculture studied. Economists have been exploring the trade-off between equity and growth by considering the proposition that economic growth may be accompanied by systematic changes in inequality. While it is unlikely that the removal of distortions in the economy could be challenged on efficiency grounds, the results obtained show that a welfare gain at the national level hides the existence of potential losers in the process of pricing policy reforms. Whether this comes as a result of reductions in tariffs, the lowering of output support or reduction in input subsidies, the low rainfall areas (LRA) seem to have difficulties benefiting from this type of reform in the crop and livestock markets. While basic trade theory and the philosophy behind structural adjustment programmes suggest that reduced tariffs and domestic market distortions will lead to reduced inequality, evidence from the real world suggests that openness has been accompanied by increasing inequality in a significant number of observed cases (Horton, Kabur and Mazumdar, 1995). The concern raised by the combined national-regional results in tables 34 and 35 is that even with negative growth, some tend to lose more than others, despite improvements in welfare at the national level. The outcome in terms of welfare changes illustrates the crucial importance of complementary reforms or transfer payments to the LRA to mitigate the potential losses from agricultural policy reform and trade liberalization. In the case of Tunisia, negotiating an option in future multilateral trade liberalization rounds that would mitigate the negative impact on rainfed agriculture seems to be a necessity, considering the high weight the Government of Tunisia has put on its income inequality reduction objective. While the high rainfall areas (HRA) could benefit from higher market access for fruit and vegetable exports, the LRA have very limited alternatives other than barley, durum wheat and sheep meat production. D. DOMESTIC PRICING REFORMS AS A PREREQUISITE FOR FURTHER TRADE LIBERALIZATION According to some studies that assess the Uruguay Round impact on the agricultural sector, the influence of the Round negotiations has been limited, with most of the change in agriculture occurring as a result of changes in Government policy (e.g., IATRC, 1997). Given the high average rates of protection that continue to exist in Tunisia, as shown in chapter I, would most of the potential gains that can be realized come, in fact, from further liberalization in the domestic Tunisian market or in the markets of Tunisia's main trade partners, whether on the export or import side? That question has been looked at by Lachaal et al (1998), using the same multi-market model presented in the previous section, by considering a number of scenarios in which domestic market distortions are reduced. The first scenario is one in which a multiple of price changes on the producer side is assumed. The idea in this scenario is to remove some of the distortions that give bias to the incentive structure in the production of agricultural products. '['he policy instruments used consist of the reduction of the guaranteed producer price of wheat, barley and beef by 10 per cent and an increase in the producer price of olive oil by 10 per cent. The second policy scenario is similar to the first one, except that it intends to remove some of the price distortions on the consumer side by bringing consumer prices closer to world levels. This policy regime consists of the reduction of the consumer domestic prices of barley and beef by 10 per cent and an increase in the consumer price of olive oil by 4.5 per cent to align it with the world price. The third scenario is a market liberalization scenario and is a combination of the first and second scenarios (S1+S2). The various policy scenarios analysed seem to have differential impacts on key welfare and income measures. The impact of selected regimes on agricultural budget revenue, balance of trade, agricultural incomes and the net welfare effect are summarized in table 36. Reduced support to producers would reduce incomes of the various areas by 17-23 per cent, but would improve the Government budget and the balance of trade. Overall, this would generate gains of around 44 million D for the nation. Gains from reduced consumer subsidies would still yield a net positive gain of 14 million D, despite the worsening of the Government budget, owing to increased wheat consumption (higher subsidies for bread). The market liberalization scenario (S3) yields the highest net welfare gains. By removing some of the distortions of current policies (i.e., bringing producer/consumer prices closer to world levels), this regime tends to improve domestic resource allocation and positively benefit the nation with a net increase of around 56 million D. Agricultural budget revenue seems to be positively affected by that reform. Producer and consumer support declines, and so would producer and consumer taxes. The net effect is a 6 per cent increase in agricultural budget revenue. Lachaal et al (1998) found also that the change in the structure of demand and supply would affect trade with world markets. The import bills of wheat and beef would greatly improve, while export revenues from olive oil would increase. The overall effect is an increase of the agricultural balance of trade by 4 1 per cent. E. AGRICULTURE AND INDUSTRY LINKAGES IN A GENERAL EQUILIBRIUM SETTING The analysis in this section takes us one step further in the range of tools used to evaluate options of trade liberalization. Chemingui and Dessus (1999) use a dynamic CGE model to evaluate bilateral and multilateral liberalization arrangements Tunisia could enter into. They look first at the implications of reducing agricultural support with limited EU concessions when it comes to Tunisian exports and find that it is not in the country's best interest to lower its barriers to imports. However, improved access to EU markets for exports, in which the country has a comparative advantage, would provide substantial welfare gains nationwide. Chemingui and Dessus then suggest that part of the benefits be redistributed to agricultural producers hurt by the process of reduced support to agriculture. Most importantly, they show that by granting the same preferences to other trading partners, Tunisia would maximize its growth potential and reduce losses in the sector. Table 37 provides some key aggregate indicators of the potential changes that could arise as a result of further openness in the Tunisian economy. In the scenario explored (consisting of the simulation of a process of multilateral reform in agriculture and industry combined with the Euro- Mediterranean Partnership Agreement and with tariff loss compensated by a rise in harmonized VAT), welfare gains were around 40 million D more than what could be achieved with the EU-Tunisia agreement alone. Rural welfare does not necessarily suffer from the process, as this type of liberalization offers opportunities for further exports that would compensate for reduced domestic support. Chemingui and Dessus claim, however, that the reform of Tunisian agriculture is viable only if it is accompanied by greater access to EU markets for its export products. F. POLICY IMPLICATIONS AND RECOMMENDATIONS FROM THE VARIOUS QUANTITATIVE ASSESSMENTS This analysis of tariff reduction patterns provides evidence of the importance of decisions regarding the choice of TCFs and the parameters associated with tariff reduction schemes. While it does not venture to provide highly reliable and definitive estimates on potential welfare gains, it nevertheless shows important trends and the magnitude of welfare gains for a reduced set of markets. Tunisia is committed to reduce its AMS level from 68.4 million D to 59.3 million D by 2004 (MOA, 1996). Its society could very well benefit from such an amount if a deep harmonization is used to reduce prices by around 25 per cent on average on cereals, milk and meat. Obviously, adding vegetable oils, sugar and other food items to the set of markets to be covered by tariff reduction would substantially increase the net welfare gains achieved via trade liberalization. Those gains will also very much depend on the flexibility of the agricultural system in responding to pricing signals. Improving the information channels for producers and consumers, providing diversification schemes in production and encouraging the consumption of fruits and vegetables while reducing intakes of cereal-based food could well lead to higher welfare gains from tariff reduction, regardless of the type of formula cuts. Tunisia, therefore, should conduct a thorough analysis, encompassing all major agricultural and food items affecting the balance of trade and their substitutes, of the net gains each type of formula could generate and its differential impact on consumers, producers and the Government budget before opting for a particular tariff reduction scheme. Depending on how far it wants to go with respect to domestic producer price reduction and how much political leverage the Government has to affect the most crucial agricultural markets, Tunisia should be able to design a tariffication system that would limit the adverse effects of lower tariffs on specific segments of the population without compromising the potential gains that could be achieved at the national level. The DH formula seems to be yielding the highest welfare gains for price reductions that maintain a degree of border protection for cereals and livestock products. When it comes to complete liberalization, results of the welfare analysis have shown that further concessions by the EU for Tunisian olive oil could be greater than the improvement in welfare arising from a complete dismantlement of agricultural tariffs on Tunisia's main imports. However, those results tend to rely on behavioural parameters related to Tunisia's import demand and export supply. With more flexibility in its agricultural production system, Tunisia could to a certain extent reap higher benefits than those presented earlier. The results of the multi-market and CGE analyses suggest that substantial gains in total welfare will accrue from the domestic market and from trade liberalization conducted within a continuing structural adjustment programme or multilateral trade agreement covering agriculture in a more substantial way. The results have also shown that while the net gains very much depend on differences in methodology and specification of economic models, the broad scale of magnitudes suggests that Tunisia's welfare gains are proportionate to how deep the reduction in market distortions is, whether domestically or for its main trading partner. However, by focusing on the national welfare gains, those types of studies suggest that there will be some potential losers whose fate will have to be considered. Whether it is the rural sector as a whole or in some segment of the rural population (e.g., low rainfall areas in the centre and south of the country), Tunisia should accompany the market and trade liberalization process with specific and targeted rural development programmes. This approach is not an easy one, given the size of the agricultural labour force, as it involves substantial budgetary outlays in a context where tariff revenues are reduced, though with savings realized on subsidization schemes. TABLE27. DATANEEDS AND MARKETS ANALYSED IN WELFARE ANALYSIS OF TARIFF REDUCTION FORMULAS s s e d -- Domestic prices Soft wheat Border prices Durum wheat Tariff rate Barley Elasticity of domestic supply Meat Elasticity of domestic demand Milk Value of domestic production at domesticprices Value of domestic production at border prices Value of domestic demand at domesticprices Value of domestic demand at border prices TABLE28. KEYPARAMETERS USED IN THE EVALUATION OF THE IMPACT OF TCFSOh WELFARE Soft wheat Durum wheat Barley Meat Milk Bound tariff 61 76 57 76 61 Supply elasticity 64 21 42 Demand elasticity 23 02 25 Value of production 56.1 294.9 71.0 305.0 187.1 Value of consumption 180.5 355.4 1 10.3 325.5 223.3 L5'ources:Author's calculations based on reductions of 24 per cent in bound tariffs by the end of URAA implementation. Elasticity estimates taken from Lachaal et a1 (1998). TABLE29. RANGEOF CHANGE IN TCF PARAMETERS Lower limit Upper limit A 0 1 B 10 10 C E 600 Note: E is an arbitrarity small vaIue TABLE30: IMPACTOF TCFS ON WELFARE GAINS (Millionsof D) Aggregate average price reduction 15% 25% 35% Base Linear 9 24 47 Linear harmonized 9 25 48 Deep harmonization 10 28 55 Swiss formula 10 26 48 Case with high supply elasticities Linear 14 40 78 Linear harmonized 15 42 80 Deep harmonization 17 47 90 Swiss formula 16 43 80 Case with high supply and demand elasticities Linear 17 49 96 Linear harmonized 19 50 97 Deep harmonization 20 57 110 Swiss formula 19 52 97 Source: Author's calculations Figure XVI: Impact of Tariff-cut formulas on welfare gains Reference case +LINEAR HARMONIZED 1 +HARMONIZED i (-SWISS FORMULA 1 10 15 20 25 30 35 40 Average reduction in prices (%) Figure XVII. Impact of tariff-cut formulas on welfare gains Case with high supply elasticities 15 20 25 30 35 40 Average price reduction (%) Figure XVIII: Impact of tariff-cut formulas on welfare gains Case with high supply and demand elasticities i+LINEAR HAR +HARMONIZED *SWISS FORMULA *LINEAR I5 20 25 30 35 40 Average price reduction (Oh) Figure XIX. Sensitivity of welfare gains to formula cuts and elasticity ranges 5 10 15 20 25 30 35 40 - - - Average price reduction(%) - - - L~near-Ref L~near-HI& ES L~near-H~gh ESIED Hsrrnon~zed-Ref tlarrnontzed-hgh ES Harmontzed-H~ghESIED TABLE34. ESTIMATED WELFARE AND INCOME GROWTH EFFECTS OF PRICE REFORMS Type of scenariolindicators S x p r e s s e d in percentage) Morocco Tunisia Jordan S1 HEV-"' Income g,rowthh" S2 HEV Income growth S3 HEV Income growth S4 HEV 0.17 1.36 0.62 Income growth -3.29 -1.60 -0.23 Source: Chaherli (2001). Notes: HEV is the Hicksian Equivalent Variation expressed as per cent of total national income for the base year; a positive number indicates an improvement in welfare. b/ The income growth indicator is at the national level Figure XX. Impact of removal of MFN tariffs Contributions to welfare measures by commodity - I vegetable 1 011 I Q sugar ~ M I I L OMeat O9arley ~ u r u mwheat m ~ o fwheat t 1 Consumer Surplus Producer Surplus Tariff Revenue Net Welfare Measure TABLE35. WELFAREAND INCOME GROWTH EFFECTS OF PRICE REFORMS RY AGRO-ECOLOGICAL ZONE Morocco Tunisia Jordan HRA LRA HRA LRA HRA LRA S1 HEV -2.80 3.1 0.1 1 -0.17 Agricultural income growth -5.28 -6.88 -9.1 -9.7 -0.13 -1.73 S2 HEV -2.90 Agricultural income growth -5.38 S3 HEV Agricultural income growth S4 HEV -5.60 -6.5 0.23 -3.11 Agricultural income growth -10.67 -13.97 -18.8 -20.4 -0.26 -6.21 Source: Chaherli (200 I). Notes: Two dots (..) indicate data not available. HRA and LRA refer respectively to high and low rainfall areas HEV is the Hicksian equivalent variation expressed as per cent of regional agricultural income for the base year; a positive (rcsp. negative) number indicates an improvement (resp, loss) in welfare. The income growth indicator is at the regional level No welfare indicator at the regional level could be computed for Tunisia because demand has been specified at the national level in thc Tunisian rnulti-market model. TWLE36. IMPACI' OF MARKET LIBERALIZATION ON SELECTED MACROECONOMIC AGGREGATES 1996Base year values (Itlillions of D) S 1 S2 S3 __ Partial agricultural budget revenue 18.77 92.23 -94.30 6.35 Partial agricultural balance of trade 109.25 33.71 7.98 4 1.69 Agricultural incomes Low rainfall area 223.96 -16.78 -1.03 -17.81 High rainfall area 249.76 - 17.49 -0.33 -17.82 Rest of the country 74.82 -22.45 -2.65 -25.10 Net welfare effect (Millionsof D) 44.36 14.09 56.43 Source: Lachaal et a1 (1998). Notes: All entries are percentage changes from base year levels, except for the net welfare effect computed in millions of Tunisian dinars. SI= reduction in the producer guaranteed price of wheat, barley and beef by 10 per cent and an increase in the producer price of olive oil by 10 per cent. S2 = reduction in the consumer domestic price of barley and beef by 10 per cent and an increase in the consumer price of olive oil by 4.5 per cent. S3 = S1 plus S2. TABLE37. RESULTSOF AGRICULTURAI. REFORMS SIMULATED WITH A CGE MODEL FOR TUNISIA (Billionsof D) MRAl scenario Real GDP 34.06 Total production Private consumption Exports Exports to EU Exports to rest of the world lmports lmports from EU lmports from rest if the world Variation in well-being (94) Rural households Urban households Total Solirces: Chelningui and Dessus (1999) and Dessus and Suwa (2000) Notes: All macroeconomic indicators are in billions of 1992 Tunisian dinars. MRAI scenario simulates a process of multilatcral rcform in agriculture and industry 1992 and 2010: values of the variables for these two years in the reference scenario (the Euro-Mediterranean Partnership Agreement. with tariff loss compensated by a rise in harmonized VAT). V. POLICY ISSUES, OPTIONS, TRADE-OFFS AND STRATEGIES Tunisia's experience in terms of URAA implementation shows that the country has broadly adhered to rules and reduction commitments for market access, domestic support and export subsidies commitments. Despite the high levels of bound tariffs, Tunisia's applied profile for key imported agricultural commodities does not reflect the existence of a highly distorted domestic market. With quasi insignificant export subsidies, a continuous effort to reduce domestic support in crop and livestock production and downward trends for applied tariffs, further reduction commitments in future trade negotiations do not seem to have the potential to negatively affect welfare and production levels in a way that would compromise the overall development objectives set by the Government of Tunisia. What it should be mainly concerned about is the extent to which its main trading partner, the European Union, would consent to further reductions in domestic support and increased access to its market. However, this does not imply that Tunisia should minimize its efforts in looking for improvements in current terms. There are a number of outstanding issues that Tunisia should be tackling up front and it should look for ways to further improve its trade position. Before looking at some elements in the new trade agenda, some outstanding issues of the "three pillars" are discussed below. A. MARKET ACCESS ABROAD Agricultural and food exports accounted for almost 10 per cent of total exports in 'Tunisia over the period 1990-1998. The bulk of the current trade occurs in a number of key commodities, as reported in table 38. The most important barriers restricting expansion of these exports could be combined under the following three main headings: (a) Terms of the Euro-Mediterranean Partnership Agreement. Under the recently negotiated agreement with the EU (2001-2004), Tunisia has a limiting quota for olive oil (50,000 tons). Complete removal of barriers to entry of this product could allow Tunisia to design a more aggressive policy in terms of production and exports. For the other key agricultural commodities, Tunisia has difficulties reaching the negotiated quota limit under the current production patterns. Yield improvements in citrus fruits and vegetables, as well as productivity gains in the food processing industry, could generate a higher potential in terms of export growth. Unlike Morocco, which has restrictive limits on most of its agricultural goods entering the EU, Tunisia's limited export potential is more related to other factors, including domestic supply conditions and limited export promotion programmes; (b) Preferential terms for Tunisia's competitors. Tunisia faces competition from other southern Mediterranean countries exporting the same type of products and getting preferential terms similar to those it receives for some of its own products. A more competitive world food market shaped by the competitiveness of products and not their origin could allow Tunisia to start exporting some items that are currently limited in that regard (e.g., flowers and plants, vegetables, processed food items); (c) Tariff escalation on processed and transformed commodities. Tunisia has a number of products with stalled export growth (e.g., olive oil of "low" grade, dates and oranges) and with a high share in total exports. To adapt to the new downward trends in prices, Tunisia will have to export those products at a higher level in the processing chain. Currently, however, tariffs are higher the more transformed those products are. With significant progress in the reduction of tariff escalation in multilateral trade negotiations, those products could have better export prospects. B. DOMESTIC MARKET ACCESS When it comes to reporting progress made in trade liberalization, Tunisia figures preeminently on top of the list of countries with high applied MFN and bound tariffs. A closer examination in chapter I1 at the tariff schedule for agro-food items reveals that unweighted averages have put Tunisia in a position that tends to underestimate efforts made to reduce protection on the items that count the most in its trade balance. In the set of items looked at, there are approximately 30 percentage points for bound tariffs and 8 percentage points for applied tariffs in the difference between unweighted and weighted means. This, however, does not necessarily mean that the Government should minimize its efforts to further reduce the overall tariff schedule, in particular for animal products, vegetable oils and cereals. As seen in the preliminary assessment of tariff-cut formulas, any step taken towards a zero tariff situation generates higher welfare gains. Tunisia, to a certain extent, should be looking into the position adopted by countries of the Cairns Group, as the net cost of lowering its own tariffs would seem to be much lower than the benefits it could get from lower tariffs applied to its agro-food exports. C. DOMESTIC SUPPORT With respect to reported AMS levels over the past five years, as seen in chapter 11,the bulk of reported support has been in durum and soft wheat and milk. The total amount tends to be "diluted" by the negative support to the olive oil sector, as a result of reference prices being higher than supported prices. However, since 1997, total market price support in olive oil has been reported as equal to zero. This raises an issue related to the method of calculating domestic support. Should Tunisia be pushing for global reduction or reduction by products? At the present time, a global reduction gives more leeway by adjusting at least one administered price to reductions in AMS consistent with yearly commitments. Reaching 59.3 million D by the end of the URAA implementation period will not prove to be a difficult task. If further reductions are to be sought, Tunisia ought to evaluate more strategically its support policy, considering the limited research and development efforts put into agriculture. Tunisia's planned investments in agricultural research over the 1997-2001 period are a mere 30 million D (Ministry of Agriculture, 1997) to cover a variety of programmes for five years. from strategic products, natural resources and animal health to biotechnology. If it wants to achieve a stronger growth path and increase its export potential, Tunisia should be "sacrificing" a portion of the current support programme for an export-oriented research and extension programme on commodities with a higher value-added than cereals and milk. D. CONTINUATIONOF NON-TARIFF PROTECTION MEASURES Tunisia is having difficulties dealing with constraints imposed by the EU on its olive oil exports as a result of non-tariff barrier (NTB) measures. Standards commonly agreed to by the International Olive Oil Council (100C) should be the benchmark used in the trade of olive oil worldwide. If one country or a group of countries has to impose tighter standards than those negotiated or commonly agreed upon within multilateral settings, those impediments to trade could hurt countries like Tunisia that have limited power to affect rules set by bigger countries. Future negotiations on non-tariff protection should clearly identify the cases under which NTB measures are allowed and rule out any other NTB intervention. E. EXPORTCREDITS Discipline in the area of export credits is an item of particular interest for Tunisia. The main purpose of any multilateral agreement would be to provide the institutional framework for the orderly use of officially supported export credits. Any arrangement should seek to encourage competition among exporters on the basis of the quality and cost of the goods and services exported, rather than on the basis of officially supported terms that would provide the most favourable price structure for customers. Tunisian exporters benefit from various export credits with official support. An arrangement that would place limitations on the terms and conditions of official support (e.g., minimum cash payments to be made at or before the starting point of credit, maximum repayment terms and minimum interest rates covered by official financing support) could seriously hurt a large number of exporters of agricultural and food commodities. At the same time, importers could face the same limitations. Tunisia needs, therefore, to assess the potential implications that changes in export credit mechanisms could have on its trade flows before deciding on whether to embark on this front. F. CONTINUATIONTHE PEACE CLAUSE BEYOND 2004 OF In view of its relatively strong outlook for the next decade and the potential for moving on to a higher growth path, Tunisia should be aiming at faster and deeper trade liberalization of its agricultural sector. While there are political interests in looking for a continuation of the peace clause beyond 2004 (Tunisia being a member of two different trading blocks, composed of the African and Arab countries), there is some evidence showing that Tunisia could indeed benefit from the extension. Tunisia has always tried to be at the forefront of countries in the same income group by looking for trade expansion opportunities, rather than seeking exemptions and gaining time by delaying further outward orientation. Tunisia has rushed, although at times hesitantly, towards openness in various multilateral and bilateral settings, while recognizing the potential negative effects reduced protection could have on its economy in general and agriculture in particular. It is this approach of steady opening and liberalization that has contributed to improving its socio- economic indicators, including those related to poverty and income distribution. While it is on the basis of those two indicators that extensions and exemptions have principally been sought by LDCs, Tunisia is adopting a different and opposite approach to improve the conditions of those hurt by the opening process. Social safety nets of the sort set up via the National Solidarity Fund are seen as the remedy to mitigate the harm done by the liberalization process, but without questioning the benefits of further trade openness. G. REDEFINITIONSTATETRADING COMPANIES TO CONFORM TO WTO REGULATIONS OF State trading enterprises (STEs) have legitimate status within the rules of GATTIWTO. While their operations are legally recognized, there are rules that constrain their ability to distort trade and markets. While there is often ajustified rationale for maintaining their activities from a national standpoint, there have been a number of concerns with respect to the impact they have on international trade. As pointed out by several studies on the subject (e.g., OECD, 2001 or Ingco and Ng, 1998), there are currently problems regarding the definition of an STE and the evaluation of its trade-distorting. Concerns have also been voiced with respect to the potential for circumventing the export subsidy commitments made in LIRAA and the unfair advantage it poses to the Government in terms of export credits, tax breaks and so on, which tend to reduce the export prices it offers.20Tunisia does have several potential cases at hand it should look at before entering future negotiations on the issue. As mentioned in chapter 11, OC and ONH are two of the most heavily involved STEs on the trade side. There will be concerns, for example, about the degree of market distortion likely to arise from ONH export activities in the olive oil sector and import activities for other vegetables oils (mainly those derived from corn, colza and soybeans). One specific concern that could arise is whether ONH is in a position to undermine market norms and distort international agriculture trade in the olive oil market. What has been ONH's influence in that market in recent years? The share of ONH in world exports - (CE,,) computed in the standard way, as described in OECD (2001)~'-has been declining since 1995, as shown in figure XXI. Would a share varying between 10-20 per cent be considered as a trade-distorting pattern? If it plans to maintain ONH's influence in world markets, Tunisia ought to look for ways tojustify ONH's role as a market-stabilizing factor rather than as a trade distorting structure. H. INTRODUCTIONOF THE MULTI-FUNCTIONALITY OF AGRICULTURE If adopted and implemented, the concept of multi-functionality of agriculture, coming as the initiative of the EU (with potential backing from LDCs or NFICs), could have important implications for Tunisian agriculture and trade prospects. Though the concept in itself could secure some degree of protection for its own agriculture, Tunisia should be concerned about the use of multi-functionality in Europe as a way to protect farmers in the northern Mediterranean shores from products originating on the southern shores. It is not necessarily in Tunisia's best interest to back this concept in order to potentially secure greater preferential access to the EU market, as multi-functionality could have some opposite effects. The more open EU agriculture is, the better the prospects are for increasing its market share in non-traditional products and maintaining its current share in products such as olive oil, dates and citrus fruits. I. TARIFFRATE QUOTAS Although TRQs were designed to guarantee market access for imports, their administration by governments has, in fact, led to controversy. Tunisia is not immune to this controversy, as TRQs have in practice become a new channel for discretionary government intervention, owing to the management of procedures to grant import licenses according to one of the methods mentioned in chapter 11. Rather than imposing stricter regulations on license-granting procedures or introducing new formulas such as license auctioning, Tunisia should back innovative ways to stabilize agricultural markets domestically without having to resort to TRQs if, indeed, the claimed objective is to maintain adequate food security levels. 20 See OECD (2001) for a complete review of those concerns. 2' CEwe= (CJW,)*CE, where CJW, is the national share in world exports and CE, the share of ONH in Tunisian exports VI. SUMMARY AND CONCLUSIONS The Uruguay Round agreements represent the most thorough reform of the world trading system since the establishment of the GATT in 1948. The agreements were critical in supporting the continued efforts to widen the negotiating agenda of the multilateral trading system and, through the years, a growing number of countries contributed in setting up the agenda. The agreements were a cornerstone for the subsequent application of multilateral rules to trade in agricultural products, an area that had remained largely outside of the GATT disciplines and was governed by special arrangements. Since the establishment of the WTO, Tunisia has been using its membership in the newly created structure, as well as its agreements with the EU and a number of Arab countries, to further liberalize trade. Tunisia seems to have recognized the real potential in opening up its economy by reducing government interventions and distortions in several agriculture markets. However, despite notable efforts in this arena, further improvements are being sought, in particular in the crop and livestock sectors and also in durum and soft wheat and milk. The Uruguay Round Agreement on Agriculture resulted in important new steps that set the stage for further tariff reductions. By its official entry into WTO, Tunisia agreed to a number of reductions concerning its bound tariffs, aggregate measure of support and export subsidies. The experience has shown that Tunisia maintains relatively high bound tariffs that are in the order of 80 per cent when the importance of commodities in total trade is taken into account. This rate is far above the average MFN tariff of 19 per cent. This is in sharp contrast to the same measures when an unweighted average is used, which puts the average bound and MFN tariffs at 113 per cent and 26 per cent, respectively. Tunisia's overall tariff schedule exhibits a wide variation, as measured by various dispersion indicators. The use of TRQs is also quite widespread. In terms of domestic support, Tunisia will be bound not to surpass 59 million D in total support to commodities such as wheat, barley, milk, sugar beet, olive oil and potatoes. The issue of export subsidies does not seem to be a major trade-distorting mechanism, as very few items are covered by the scheme. "How much will further cuts in domestic support and tariffs be all worth?" is a natural question to ask in the case of Tunisia, as it seeks to show the rest of the world that its economy aims at full integration within the globalized economy. This happens to be a difficult question to answer precisely without the aid of quantitative tools. Several analyses made reported estimates of gain that appear to be quite substantial in terms of income and welfare effects. With various formula cuts, Tunisia could improve its welfare by at least 60 million D on cereal crops and meat, an amount as big as the aggregate measure of support to be achieved by 2004. Based on various methodologies, ranging from multi-market analysis to CGE modelling, it is apparent that it is in Tunisia' s interest to reduce once and for all the market distortions existing in what it considers as strategic products. We have briefly reviewed a number of elements of concern in Tunisia's future negotiating strategy, considering its special relation to the EU as its main supplier and customer. We conclude that Tunisia should distance itself from the position where exemptions and extensions are sought and, with a better data management system in place, aggressively explore opportunities to increase its trade flows without compromising the delicate equilibrium of its balance of payments. To do so, Tunisia will have to invest more in the conduct of agricultural trade research. TABLE38. TUNISIA'S MAIN AGRICULTURAL EXPORTS, 1990-1998 Production Exports (a) (b) Share (In tons except for wines in hectolitres) @)/(a) Olive oil (tons) 159400 109 900 78.8% Wines 322 600 84 300 25.9% Dates 81 000 20 200 25.0% Oranges 109900 22 500 19.3% Fish and seafood 86 900 16000 18.4% Source:Ministry of Agriculture. Figure XXI. Share of ONH in world olive oil exports Source: Data compiled from ONH (1998) and Ministry of Agriculture (1996,1999,2000) for national exports and composition, and IOOC (2001) for world exports. ANNEXES Annex I SPECIALAUTHORIZATION THE EXPORT OF OLIVE OIL IN TUNISIA FOR This is a non-official summary translation (by the author of this report) of the different articles related to the Special Authorization for the Export of Olive Oil in Tunisia, called "Cahier des Charges". Article 2. To obtain authorization to exercise the profession of olive oil exporter, the commercial entity in question must submit an application to the Director General of ONH. Article 3. The requestor must submit a certificate of ownership or rental agreement regarding the storage facility used for handling the olive oil. Article 4. The application is approved if the following conditions are met: (a) The storage facility has been approved by a commission whose composition is set by ministerial decree; (b) Exports are either directed to any market outside the annual quota provided by the EU to Tunisia, or are in the form of conditioned olive oil in receptacles of less than 4.5 Kg and destined to any market, or are of superior quality; (c) The exporter must have a laboratory in which oil is tested according to physical, chemical and taste criteria. The Ministry of Commerce should approve the testing facility used. Article 6. The commission in charge of handling the applications is presided over by the MOA and includes the following members (or their representatives): the Director General of the Livestock Production Department at MOA, the Director General of the Food Production Department at MOA, the Chief Executive Director General of ONH, representatives of the Ministry of Economic Affairs, the Commerce and Industry Trade Union (UTICA), the Farmer's Union (UTAP), and the Chamber of Agriculture for Olive Oil. The commission takes decisions under majority rule, with the vote of the President being decisive in the case of equal votes. Article 7. Exports are made under an agreement between ONH and the private exporter, stipulating the quantity of oil involved, its quality, shipment and delivery conditions and mode of payment. Article 8. Contracts between the parties have to be approved by ONH. Article 9. Exporters have to abide by IOOC and CODEX norms. Source: Annex No. 2, p. 57, MOA (1 993). Annex I1 GENERAL STRUCTUW OF A MULTI-MARKET MODEL In this annex, the general structure of a multi-market model, as described in Sadoulet and de Janvry (1995) is presented. See Chaherli (2001) for the specific features of the multi-market models developed for three countries of the Middle East and North Africa region: Jordan, Morocco and Tunisia. Generic multi-market model equations ; i The producer side ( I ) Supply of commodity i: q l s= CkqlkS(P, r, zJ) (2) Demand of input h: X / = Ckx h k d(P, r) ! The factor supply side (3) Supply of input h: Xh" = CkXhkS(P, r) I 1 The regional income and final demand side (4) Regional income: Yk = nk (P, r) + R k (5) Demand of output i: d qrJ = x k Y , k ( P C , Ykj zd) Market equilibrium conditions I (6) Product markets: Ck qlk' = Ckqlk t NE, - Non-tradable: NE, = NE, ;( p ,, (7, endogenous) If tradable: p, = i,; (NE,, 4, endogenous) (7) Factor markets: J Ck XhkS = Ckqhk + NEh Non-tradable: NEh ah(rh, = ; Xh endogenous) ( 8 ) Balance of trade: BOT = C I N E , i E I Set of agricultural products h E H Set of variable factors k E K Set of regions Parameters Rk Non agricultural income Xhk Supply of fixed factors Variables nk Agricultural profits pci Consumer prices pi Producer prices qdi Final demand qsi Supply NEi Net exports of commodity i NEh Net exports of factor h rh Factor price qhk Factor demand Yk Total income The model incorporates four classes of agents: producers distinguished by region and commodity, consumers distinguished by region and social group (urban vs. rural), suppliers of factors (intermediate demand factor or inputs supplied by the rest of the economy) and the Government. On the production side, the model assumes that producers are profit maximizers. Their supply response function has output and input prices as arguments. On the factor side, each region supplies factor h as a function of product and factor prices (P, r). Total supply of factor h is obtained by summing over regional supply. Regional income (Yk) is given in equation (4) as the sum of profits made from agricultural income- generating activities and some other exogenous incomes, such as remittances and services. Final demand for each region is specified as a function of commodity prices at the consumer level, regional income and other exogenous shifters. Total demand for commodity is obtained in equation (5) by summing over regional demand. Equilibrium conditions on product (equation 6) and factor markets (equation 7) depend upon the tradability of each product and factor. For products/factors that are traded internationally, markets are assumed to clear through adjustments in net exports (NE = exports minus imports), which equilibrate supply and demand at exogenous border prices. On the other hand, for products/factors not traded internationally, markets clear through domestic prices moving to equate supply and demand. Finally, the balance of trade (equation 8) is a residual. It indicates the magnitude of the deficits or surpluses of the multi-market model equilibrium, with no feedback on the domestic price system. Following the general structure, as in Sadoulet and de Janvry (1995), equations are log-linearized, with the result that all equations become linear in the rates of change of the endogenous and exogenous variables, and the coefficients are transformed into shares or elasticities. A number of important simplifications occur in the log-linearization process, as the profit equation becomes a linear combination of changes in output and input prices weighed by the contribution of each product and input in terms of values in total profits. The simplification is a direct application of Hotelling's lemma, expressing product quantity as the derivative of the profit function with respect to product price and the input quantity as the negative of the derivative of the profit function with respect to input price. If we denote A as the matrix of coefficients associated with the endogenous variables (e.g., prices of the non-tradable commodities and imports of the tradable goods), B as the matrix of coefficients associated with the exogenous variables (e.g., prices for the tradable commodities and the policy instruments), E as the vector of rates of change in the endogenous variables and X as the vector of rates of change in the exogenous variables, the system can be simplified as: Solving the above system for the endogenous variables yields: Resolution of the model, however, relies on some theoretical and empirical restrictions regarding the choice of elasticities appearing in matrix A and B. These elasticities must satisfy: (i) the basic restrictions from supply and demand theory (e.g., symmetry of the Slutsky and Hessian matrices, convexity conditions derived from utility and profit maximization, etc.), (ii) empirical evidence from econometl-ic work and (iii) restrictions on parameters generated by specific conjectures about behavioural responses. Welfare measurement in the multi-market analysis is done with the aid of the Hicksian Equivalent Variation measure: HEV is a measure of welfare change resulting from a change in prices from po to p, and moving the economic agent from the old utility level Uo to a new utility level UI. This measure uses the expenditure function e(p,U) and provides an exact measure of compensation required to maintain a specified level of utility in the face of changes in the level of output and input prices as a result of policy reform. The HEV measure uses the pre-reform prices as the base and asks what income change at the pre-reform prices would be equivalent to the proposed change in terms of its impact on utility. REFERENCES* Abaab, A. and M. Elloumi. 1997. "Effet des politiques de prix, de subvention et de fiscalite sur les performances de I'agriculture tunisienne". Options Mediterraneennes, Ser. B, No. 11, 1997. Prix et Subventions: Effets sur les Agricultures Familiales Mediterraneennes. AMAD. 2000. Agricultural Market Access Database. User's Guide, Release 1.O. Ben Cheikh, 11. 2000. "Etude sur I'impact de la politique des prix et des subventions sur la rentabilite de la cerealiculture en Tunisie: 1975-1998". Unpublished Masters thesis, Ecole Nationale d'Administration. Tunis: December 2000. Bowen, Hany P., Abraham Hollander and Jean-Marie Viaene. 1998. Applied International TradeAnalysis. Ann Arbor: University of Michigan Press. Braverman, A,, J. S. Hammer and A. Gron. 1987b. "Multimarket analysis of agricultural price policies in an operational context: the case of Cyprus". The WorldBank Economic Review, Vol 1,No. 2. Braverman, A., J. S. Hammer and C. Y. Ahn. 1987a. "Multimarket analysis of agricultural pricing policies in Korea." In The Theov of Taxationfor Developing Countries. Eds. D. Newbery and N. Stem. Oxford University Press. Braverman, A., J. S. Hammer and J. J. Murdoch. 1987c."Wheat and maize price policies in Hungary: tradeoffs between foreign exchange and government revenue". Agricultural Economics. Central Bank of Tunisia (BCT). 2001. Annual Report, 1999. Accessed at . Chaherli, N. 2001. "A comparative study of agricultural pricing policy reforms in West Asia and North Africa: synthesis of multi-market analysis for Jordan, Morocco and Tunisia". Paper presented at the European Conference on Global Change and Regional Integration: the Redrawing of the Economic Boundaries in the Middle East and North Africa, held in London, 20-22 July 2001. Chaherli, N., L. Lachaal, B. Thabet, L. Mahfoudhi and A. Jabarin. 1999a. "Impact of market reforms on the low rainfall areas in West Asia and North Africa". In Institutional Reform and Development in the MENA Region. Ed. 1. Liman. Arab Planning Institute and Economic Research Forum for the Arab Countries, Iran and Turkey. Cairo. Chaherli, N., P. Hazell, T. Ngaido, T. Nordblom and P. Oram. 1999b. "Agricultural growth, sustainable resource management and poverty alleviation in the low rainfall areas of West Asia and North Africa". Proceedings of an international conference held in Amman, 2-6 September 1997. Proceedings published by Deutsche Stiftung fur internationale Entwicklung (German Foundation for International Development). Feldafing, Germany. Chemingui, M. A. and S. Dessus. 1999. "La liberalisation de I'agriculture Tunisienne et I'Union Europeenne: une vue prospective". Technical Paper No. 144. OECD Development Center. Paris. Chouchane, A. 2000. "Les petites et moyennes exploitations agricoles en Tunisie". Paper presented at the workshop of the Food and Agriculture Organization of the United Nations (FAO) on the future of agricultural small and medium enterprises in North Africa, held in Tunis, 18-20 October 2000. Commission of the European Communities. 2000. "Proposal for a Council decision on the conclusion of an agreement in the form of an exchange of letters between the European Community and the Republic of Tunisia concerning the reciprocal liberalization measures and amendment of the agricultural protocols of the ECITunisia Association Agreement". Document COM (2000) 825 final, 2000/0334 (ACC). Brussels, 12 December 2000. References, wherever possible, have been verijied. 56 Dessus, S. and A. Suwa (eds.). 2000. Regional Integration and Internal Reforms in the Mediterranean Area. Paris: OECD Development Center. Gibson, P., J. Wainio, D. Whitley and M. Bohman. 2001. "Profiles of tariffs in global agricultural markets". Agricultural Economic Report No. 796. Market and Trade Economics Division, Economic Research Service, United States Department of Agriculture. Horton, S., R. Kanbur and Mazumdar. 1995. "Equity and openness". Paper presented at the World Congress of the International Economics Association in Tunis, December 1996. Ingco, M. and F. Ng. 1998."Distortionary effects of State trading in agriculture: issues for the next round of multilateral trade negotiations". Policy Research Working Paper No. 1915. Washington, D.C.: World Bank. Institut Nationale de la Statistique. "Statistiques du Commerce Exterieur". Various issues 1979-1998. Tunis: Ministry of Economic Development. See . International Agricultural Trade Consortium (IATRC). 1997. "Implementation of the Uruguay Round Agreement on Agriculture and issues for the next round of agricultural negotiations". Commissioned Paper No. 12 on Bringing Agriculture into the GATT. Department of Agricultural and Applied Economics, University of Minnesota. . International Centre for Advanced Mediterranean Agronomic Studies (CIHEAM). 2001. Development and Agri-food Policies in the Mediterranean Region. CIHEAM 1999Annual Report. International Monetary Fund. 200 1. Country Report No. 0 1/36. . 2001. "Tunisia: statistical appendix". Country Report No. 01/37. . 2000. "Tunisia: recent economic developments". Country Report No. 00137. International Olive Oil Council (IOOC). 1998. Encyclopedie Mondiale de I'Olivier, ch. 12, "Legislation and Olive Oil Policies". Lachaal, L., B. Thabet, L. Mahfoudhi and N. Chaherli. "Impacts of market reforms on the low rainfall areas in Tunisia: a multi-market analysis". Paper presented at the Mashreq and Maghreb Pro-jectPolicy and Property Rights Conference, held in Hammamet, Tunisia, 26-29 November 1998. Lachaal, N. 1999. "Mise a niveau des entreprises et strategies sectorielles: cas du secteur minoterie- semoulerie en Tunisie". Unpublished thesis, Institut National Agronomique. Tunis, April 1999. Lasram, M. and A. Khaldi, 2000. "Tunisian agriculture: 1998 achievements and expectations for 1999". In Development and Agri-food Policies in the Mediterranean Region, the CIHEAM Annual Report for 1999. Maaroufi. H. 1999. "L'exportation de I'huile d'olive en Tunisie: enjeux et perspectives". Projet de Fin d'Etudes, Ecole SupCrieure d'Agriculture de Mograne. Medagri. 2000. Annuaires des Statistiques Agricoles. CIHEAM. Ministry of Agriculture (MOA). 2000. Annuaire des Statistiques Agricoles, 1999. . 1996. Annuaire des Statistiques Agricoles, 1995. . 1993. "Etude sur les possibilitks et contraintes likes a la libkralisation de la collecte de I'huile d'olive". Unpublished report prepared by COMETE Engineering. . 1992. Annuaire des Statistiques Agricoles, 1991 Nabli, Raja. 1997. "Conskquences possibles des accords recents en matibe d'echanges agricoles et alimentaires". Unpublished Masters thesis. Institut National Agronomique de Tunisie. Observatoire National de \'Agriculture (ONAGRI). 2000. "Performance des exportations des produits agricoles en Tunisie". Ministry of Agriculture, Tunisia. Accessed at . Organization for Economic Cooperation and Development (OECD). 2001. State Trading Enterprises in Agriculture. Agriculture and Food Series. OECD: Paris. . 1999. Post-UruguayRound Tarijp'Regimes:Achievements and Outlook. 0ECD: Paris. Sadoulet, E. and A. de Janvry. 1995. Quantitative Development Policy Analysis. Baltimore: The Johns Hopkins University Press. Tsakok, I. 1990. "Single market analysis: calculating the impact of price policy". In Agricultural Price Policy. Cornell University Press. Tweeten, L. 1992. "Measuring comparative advantage, protection, and the gains from trade liberalization". In Agricultural Trade: Principles and Policies. Westview Press. United States Department of Agriculture (USDOA). 1998. "Agriculture in the World Trade Organization". Accessed at . United States Department of State (USDOS). 2001. "FY 2000 Country Commercial Guide: Tunisia" Valdes, A. and B. Shaeffer. 1996. Surveillance of Agriculture Price and Trade Policies: A Handbook for Paraguay. Washington, D.C.: The World Bank. World Bank. 1986. Report No. P-4368-Tun, Vol. 1 World Trade Organization (WTO). "Notification by Tunisia". Committee on Agriculture. Various documents, Nos. G/AG/N/TLlN/I-2 1. Accessed at .