72393 A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R 0 2 WHAT IS THE COST OF A BOWL OF RICE? The Impact of Sri Lanka’s Current Trade and Price Policies on the Incentive Framework for Agriculture OCTOBER 2013 A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R 0 2 WHAT IS THE COST OF A BOWL OF RICE? The Impact of Sri Lanka’s Current Trade and Price Policies on the Incentive Framework for Agriculture © 2013 International Bank for Reconstruction and Development / International Development Association or The World Bank 1818 H Street NW Washington, DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org This work is a product of the staff of The World Bank with external contributions. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this work is subject to copyright. Because The World Bank encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given. Any queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2422; e-mail: pubrights@worldbank.org. Cover Photo: Sri Lankan red rice. Shutterstock, LLC. CONTENTS III CONTENTS Boxes, Figures, and Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v Acronyms and Abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi Executive Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii Chapter 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Chapter 2. Background: Sri Lanka Trade Policy and Agriculture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 2.1 Brief Introduction to the Complexity of Sri Lankan Import Policy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 2.2 Export Policy Framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 2.3 Brief Note on the Composition of Government Revenue and Expenditure . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 Chapter 3. Discussion and Interpretation of Agriculture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .11 3.1 Approach, Definitions, and Data Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .11 3.2 The Case of Agricultural Imports and Exports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 3.3 Measuring Support to Agriculture: NRP and ERP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 3.4 Additional Observations on the Treatment of Various Products . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .17 Chapter 4. Cost-effectiveness of Fertilizer Subsidies in Sri Lanka . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 4.1 Rationale for Subsidizing Fertilizers in Sri Lanka. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 4.2 Analytical Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 4.3 Fertilizer Demand and Supply Elasticities with Respect to Fertilizer Price . . . . . . . . . . . . . . . . . . . . . . . . . . 22 4.4 Policy Simulation to Estimate the Cost-effectiveness of the Fertilizer Subsidy . . . . . . . . . . . . . . . . . . . . . . . 23 Chapter 5. Income Distributional Implications of Agricultural Trade Policies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 5.1 General Characteristics of Paddy Farmers in Sri Lanka . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 5.2 Effects of Trade and Price Policy on Real Household Income of Paddy Farmers . . . . . . . . . . . . . . . . . . . . . . . 30 Chapter 6. What Does It All Mean? Concluding Comments and Recommendations. . . . . . . . . . . . . . . . . . . . . . . . . 33 6.1 In the Short Term, Who Wins and Who Loses Under Current Policies? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 6.2 Implications for World Bank Engagement in Sri Lanka . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 Appendix Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R B OX E S , F I G U R E S , A N D TA B L E S V BOXES, FIGURES, AND TABLES BOXES Box 2.1: Variation in Sri Lanka’s External Policy Framework, 2007–11 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4 Box 2.2: Impact of Value Added Tax on Agricultural Commodities in Sri Lanka . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6 FIGURES Figure 4.1: Farmers’ Gains from Subsidized Fertilizer. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 Figure 4.2: Cost-effectiveness of Fertilizer Subsidy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 TABLES Table 2.1: Primary Imported and Exported Agricultural Products, Selected Subsectors, 2009 . . . . . . . . . . . . . . . . . . . . . .3 Table 2.2: Government Expenditures and Revenues (Rs million) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 Table 3.1: Protective Taxes on Agricultural Importables, 2009 and 2011 (percentages) . . . . . . . . . . . . . . . . . . . . . . . . . 13 Table 3.2: Implicit Ad Valorem Import Tariff (percent) Based on Average Import Duty Collection Rate for Food and Nonfood Consumer Goods, 2008–11 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 Table 3.3: Hidden Cost of a Complex System: Unweighted Average Import and Other Taxes (percent) by Broad Commodity Groups, 2011 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 Table 3.4: Weighted Average Import and Other Taxes (percent) by Commodity Group, 2009 . . . . . . . . . . . . . . . . . . . . . 15 Table 3.5: NRPs and ERPs Calculated for 2009/10, Using the TPR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 Table 3.6: NRPs and ERPs Calculated for 2010/11, Using the TPR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 Table 4.1: Fertilizer Price With and Without Subsidy, 2009–10 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 Table 4.2: Implied Transfer Efficiency of Fertilizer Subsidy in Paddy Rice (2009–10), Without Binding Quota and Assuming That Farm Fertilizer Purchases Are Used in Production (Not Resold at Market Prices) . . . . . . . . . . . . . 24 Table 5.1: Food Balance Sheet, Sri Lanka, 2010 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 Table 5.2: Incomes and Expenditures of Paddy Farming Households by Sector (Rural or Urban) and Province . . . . . . . . . . 27 Table 5.3: Incomes and Expenditures of Paddy Farming Households by Farm Size and Poverty Status . . . . . . . . . . . . . . . 27 Table 5.4: Incomes and Expenditures of Paddy Farming Households by District . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 Table 5.5: Income Share from Rice and Expenditure Shares on Rice of Paddy Farming Households by Sector (Urban, Rural, Estate) and Province . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 Table 5.6: Income Share from Rice and Expenditure Shares on Rice of Paddy Farming Households by Poverty Status and Farm Size. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 Table 5.7: Income Share from Rice and Expenditure Shares on Rice of Paddy Farming Households by District . . . . . . . . . . 29 A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R VI B OX E S , F I G U R E S , A N D TA B L E S Table 5.8: Effect of Policy Changes on Paddy Farmer Net Income over Tradable Input Costs, 2010/11 . . . . . . . . . . . . . . . 32 Table 5.9: Simulated Impact of Price and Trade Policies in 2009/10 and 2010/11 on Rice Growers’ Net Household Income and Rice Expenditures, by Poverty Status and Farm Size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 Table 6.1: Simulation on the Impact of Price and Trade Policies on (a) Farm Income, (b) Real Household Income, and (c) Government Accounts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 Appendix Table 1: Unweighted Average Import and Other Taxes (percent) by Commodity Group . . . . . . . . . . . . . . . . . 39 Appendix Table 2: Estimated Fertilizer Consumption by Crop Sector and Imports of Fertilizer, 2009 (mt) . . . . . . . . . . . . . 40 Appendix Table 3: Cost Share of Tradable Inputs, 2009. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 Appendix Table 4: Import Taxes on Agricultural Inputs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 Appendix Table 5: Items Subject to Specific Commodity Levies (SCLs) and Their Tariff Equivalents, 2009 . . . . . . . . . . . . . 42 Appendix Table 6A: Trade Policy Measures on Rice, Potato, Onion, and Chilies, 2007–11 . . . . . . . . . . . . . . . . . . . . . . . 43 Appendix Table 6B: Trade Policy Measures on Coconuts and Related Products, Milk and Milk Products, Sugar, and Wheat, 2007–12. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 Appendix Table 7: Changes in Levels of Special Commodity Levy, 2008–12. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 W H AT I S T H E C O S T O F A B O W L O F R I C E ? T H E I M PA C T O F S R I L A N K A' S C U R R E N T T R A D E A N D P R I C E P O L I C I E S A C R O N Y M S A N D A B B R E V I AT I O N S VII ACRONYMS AND ABBREVIATIONS CD Customs duty CESS Commodity Export Subsidy Scheme CIF Cost, insurance, and freight DRC Domestic Resource Cost EPZ Export Processing Zones ERP Effective rate of protection g gram GDP Gross domestic product ha Hectare HIES Household Income and Expenditure Survey HS Harmonized Commodity Description and Coding System kg Kilogram m million mt metric ton NBT Nation Building Tax NRP Nominal Rate of Protection PAL Port and Airport Development Levy RIDL Regional Infrastructure Development Levy Rs Sri Lanka rupees SCL Special Commodity Levy SRL Social Responsibility Levy TPR Total protection rate WTO World Trade Organization VAT Value Added Tax A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R P R E FA C E IX PREFACE This report provides empirical evidence to inform the policy dialogue over the impact of current trade and price policies on the incentive framework for agriculture in Sri Lanka. This information is meant as an input to the policy dialogue; it complements a previous note (SASDA, September 2011) on the role of Sri Lanka’s agricultural sector as the country graduates to middle-income status. This analysis provides a quantitative assessments of: (1) the level of support to farmers producing import-competing products; (2) the degree to which final consumers are indirectly taxed by those policies; (3) the extent to which agricultural exports are taxed; (4) the contribution of trade policy to government revenue through tariffs on imports and taxes on exports; (5) an evaluation of the cost-effectiveness of the current fertilizer subsidy scheme; and (6) a better understanding of the web of income transfers between producers, consumers, and government accounts. The report is structured as follows. The introduction is followed by an overview in Section 2 of the current (2009–11) import and export policy framework and a brief analysis of government revenues and expenditures for agriculture. Section 3 reviews the data sources and methods used in calculating the degree of protection, followed by summary statistics on the degree of protection for 10 agricultural commodities. Section 4 looks closely at the cost-effectiveness of fertilizer subsidies, which represent a large share of the government agriculture budget. Section 5 focuses on how agricultural trade policies influence income distribution, particularly among different groups of rice producers. The last section recapitulates the main findings and highlights their implications, with an emphasis on the often implicit and unintended income transfers among producers, consumers, and the government. A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R ACKNOWLEDGMENTS XI ACKNOWLEDGMENTS The report has been prepared by a team lead by Norman Piccioni (Lead Rural Development Specialist, SASDA), under the supervision and with the assistance of Madhur Gautam (Lead Economist, SASDA). The note is based on a background report and findings by Alberto Valdes and Jeevika Weerahewa (consultants to SASDA). The study was carried out under the overall guidance of Simeon Ehui (Sector Manager, SASDA). The peer reviewers for the study were Willem G. Janssen (Lead Agriculturist, LCSAR), Mona Sur (Senior Agriculture Economist, EASNS), Grahame Dixie (Senior Agribusiness Specialist, ARD), and Will Martin (Research Manager, DECAR). The team would like to thank Susan Razzaz, Rajish Wijeweera, Dan Biller, and Seenithamby Manoharan for feedback and comments. The team would like to thank all of the government officials and stakeholders met during several visits to Sri Lanka. Overall support to the team was provided by Shane Ferdinandus and Priyantha Jayasuriya Arachchi (SASDO). A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R EXECUTIVE SUMMARY XIII EXECUTIVE SUMMARY In Sri Lanka over the next 10–20 years, agriculture’s contribution to the economy through export revenues, employment generation, and forward/backward linkages will remain significant, even as agriculture’s relative contribution to gross domestic product (GDP) continues to decrease. Poverty is overwhelmingly rural. While it has been steadily declining over the years, further reductions in rural poverty will require a comprehensive approach, such as shifting to high-value agriculture, promoting nonfarm economic activities in rural areas, and assisting people to move out of agriculture. In addition, natural resources will become increasingly scarce and complex to manage. With mounting environmental externalities, agricultural development will become even more closely intertwined with environmental protection. It will be vital to reduce agriculture’s large environmental footprint, make farming systems less vulnerable to climate change, and harness agriculture to deliver more environmental services. Innovative policy initiatives and strong political commitment will be needed to implement such strategies, which are likely to benefit sectors other than agriculture, including tourism. The Sri Lanka National Development Plan—the Mahinda Chintana—envisages agriculture as playing a primary role in the development of rural areas and the economy as a whole. In the Sri Lanka of the 21st century, agriculture can work in concert with other sectors to produce faster growth, reduce poverty, and sustain the environment. The Mahinda Chintana offers a coherent vision of a modern agricultural sector in an export-oriented, competitive economy, which is consistent with the experience in other transforming economies that rapidly rising rural-urban income disparities and continuing extreme rural poverty can be major sources of social and political tension. The important lessons emerging from these experiences indicate that these problems cannot be addressed sustainably through agricultural protection, which raises the price of food, because large numbers of poor people are net buyers of food. Nor can they be addressed through subsidies, which distort resource allocation and prevent a smooth transition to a high performing sector. As Sri Lanka consolidates its agricultural transformation in the next 10–15 years, a key area of interest will be whether the policy and insti- tutional framework is conducive to implementing the vision of the Mahinda Chintana, and doing so in a fiscally manageable and efficient manner. By examining current trade and price policies in relation to agriculture, this analysis contributes to a better understanding of the implications and trade-offs of policy-induced changes to the incentives facing agriculture. CURRENT TRADE AND PRICE POLICY FOR AGRICULTURE Sri Lanka took a noteworthy step when it became the first South Asian country to open its economy for international trade. Since 2004 poli- cies have become more complex and inward looking, encouraging import substitution, especially with respect to agricultural commodities. The high levels of protection to importables in Sri Lankan agriculture create an incentive framework that arguably lowers investments in producing exportables, potentially preventing the best use of resources in the agricultural sector. In particular, this analysis finds that they tend to reorient the mix of crops toward protected activities and discourage the development of new products and export diversifica- tion. Experiences from other countries show that a discretionary and complex trade regime can generate considerable uncertainty about expected returns. Uncertainty significantly deters private investment, both in primary agriculture as well as in agro-processing. The study shows that under the current policies producers of farm products that compete with imports gain at the expense of consumers, while export crop producers and taxpayers lose. The effective rate of protection, representing the trade regime’s overall income effect per unit of farm output, is positive and high for import-competing products such as rice, potatoes, and milk, meaning that Sri Lankan producers of those commodities receive artificially higher incomes from them. The trade regime’s overall income effect for exportable products is negative, implying that producers’ incomes are lower than they would be without the policy. For some exportable products, such as tea and rubber, the potentially lower income is offset by a generous fertilizer subsidy. A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R XIV EXECUTIVE SUMMARY A preliminary and partial analysis indicate that in 2011, buyers of rice were implicitly taxed an estimated 34 percent; milk and potato buyers, 45 percent; maize buyers, 39 percent; and chili buyers about 52 percent—based on the protective taxes (considering both import tariffs and para-tariffs charged on them—calculated for 2011 for broad commodity groups). While a more complete analysis with broader and more general economic effects on the impacts of the policies is warranted, these findings show that elevated prices can negatively affect the real incomes of lower- and middle-income households, especially in urban areas. Households that spend about 50 percent of their incomes on food and pay 40 percent or more on basic foods suffer a real income loss of 20 percent. It follows that there is an implicit and large income transfer from consumers to the import-competing farm sector. These simulations show significant differences across provinces and districts, as well as across income classes (poor and nonpoor) and farm-size classes. In absolute terms, the removal of border protection results in smaller losses for the poor than for the nonpoor, but rela- tive to their income the poor suffer larger losses on the income side and larger gains on the consumption side. With respect to farm size, households with larger land holdings suffer the largest absolute and relative income loss. The net real income loss on average for farms larger than 10 acres reaches 36 percent of revenues from paddy production and 20 percent of total household income. Small farms of up to 3 acres have losses relative to total income of approximately 5 percent. Substantial revenues generated from imports and exports through the border measures just mentioned are also experienced in the agri- cultural sector, mostly through farm input subsidies, mainly for fertilizer. The fertilizer subsidy comes at a high fiscal and opportunity cost, representing 68 percent of the combined agriculture/irrigation budget in 2009 and 59 percent in 2010. These funds are diverted from other activities, possibly funding for public goods in agriculture and other sectors. The increase in farmers’ net income is small relative to the fiscal cost of the fertilizer subsidy, with an efficiency ratio ranging between 1.37 and 2.38. In other words, on average, the government spends between 1.4 and 2.4 rupees per acre to increase farm income by only 1 rupee per acre. The distortions brought about by the current trade regime go beyond product markets, affecting the demand for purchased inputs, land, water, and labor. TAKE-AWAY MESSAGES The current incentive framework has to be interpreted in the context of the government’s multiple policy objectives, which are (1) the generation of fiscal revenue from trade taxes, (2) the promotion of a higher degree of food self-sufficiency, (3) the reduction of poverty in rural areas, and (4) support to the political base within the farming community. Border taxes on imports and exports of agricultural products and on tradable inputs have been the principal policy instruments to gener- ate fiscal revenue. These instruments support producers in the case of import-competing activities and tax producers of export-oriented products. In contrast to agricultural products, nonagricultural tradables have seen some decline in border protection since the trade reforms of the early 1990s. In the past, such protection acted as a significant indirect (implicit) tax on agricultural production. The combined impact of import duties plus other taxes at the border results in nominal protection rates for importables for most agricultural products on the order of 30–50 percent. The results indicate high protection of import-competing activities (farmers are shielded from external competition), a slight tax on export products like tea and rubber (and thus a slight disincentive to produce exportables), and a heavier tax on coconuts for export (note: export taxes are in the form of export cesses). Given that the Mahinda Chintana aims at improving Sri Lanka’s competitiveness and promoting exports, the de-facto disincentives created against export are most likely the unintentional outcome of a complex system in which trade policy initiatives have been added on an ad hoc basis over time. Nevertheless, this system is having a very significant impact on the agricultural sector and the economy as a whole. From a sector perspective, it would be useful to reassess the highly complex system of taxes, levies and other measures in light of a fuller analysis of its impact on producers and consumers, and determine whether changes are warranted to attain the overall policy goals. W H AT I S T H E C O S T O F A B O W L O F R I C E ? T H E I M PA C T O F S R I L A N K A' S C U R R E N T T R A D E A N D P R I C E P O L I C I E S EXECUTIVE SUMMARY XV From a public investment perspective, one area for consideration is the removal of the current distortions and their consequences on the sectoral incentive framework to maximize the social returns to public investments. Public investments focused on promoting public goods with positive externalities (for example, agricultural research, rural roads, sanitary and phyto-sanitary services, and the like) tend to yield much higher returns to the economy as a whole. From a strategic perspective, it would be important (but challenging) to link trade and price policy reforms (in the medium term) with complementary reforms (over the long term) in markets for production factors, such as land and irrigation to allow a smooth transition toward an economy as envisioned in the Mahinda Chintana. As agriculture is a highly tradable activity, trade reforms, if implemented, will affect—and increase—the average returns to land and water in agriculture. Together, such reforms would also probably cause land and water to be reallocated to higher-value crops. To capture the benefits of reforms more fully, land and water markets should allow enough flexibility to facilitate changes in the crop mix and tenure arrangements. A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R CHAPTER 1 — INTRODUC TION 1 Chapter 1 INTRODUCTION PURPOSE OF THE STUDY disproportionately high, having declined far more slowly than agri- The first country in South Asia to open its economy to interna- culture’s share in GDP. This difference suggests a growing wedge be- tional trade, Sri Lanka was regarded until recently as the most open tween relative incomes in agriculture and the rest of the economy. economy in South Asia. Since 2004, however, Sri Lanka has pursued The HIES 2009/10 reports that Sri Lanka’s rural poverty headcount is inward looking policies that have encouraged import substitution, 9.4 percent and the national poverty headcount is 8.9 percent. especially with respect to agricultural commodities. Agricultural and processed foods average 30 percent of total mer- Why are the results of this analysis of Sri Lankan trade and price poli- chandise exports (Central Bank of Sri Lanka 2008; Annual Report 2007; cies relevant? They are relevant because high levels of protection Anderson and Martin 2009) report that the value of exports of primary for importables and export taxes in agriculture often have a num- agricultural products as a share of the value of all primary production ber of implicit and unintended effects throughout the economy. averaged about 39 percent during 2000–04, the highest such share in High levels of protection on food products raise farm incomes for Asia in those years. Imports of primary agricultural products accounted producers of protected activities and also raise fiscal revenue, but for only a small share of apparent consumption (5 percent), and the they constitute an implicit tax on consumers, particularly lower- “self-sufficiency” ratio (the value of agricultural production relative to income families who spend a large share of the household budget apparent consumption) was 1.57, the highest in the region. on food. High levels of protection on food products also implicitly Sri Lanka’s agricultural trade regimes have been well documented create disincentives for exports. The mix of crops is not diversified over the years. The first study, by Surjit Bhalla for 1960–84, was part of but artificially reoriented to protected activities; the development of the Krueger, Schiff, and Valdes comparative study for the World Bank new products and export diversification are discouraged. Although (1992). Another World Bank study (1996), “Sri Lanka Nonplantation export taxes raise fiscal revenues, they reduce investment in export- Crop Sector Policy Alternatives,” covered the situation in the mid- oriented activities. In the longer term, policies that protect import- 1990s. A study by Bandara and Jayasuriya (2009) was also done for competing activities and tax exportables can reduce overall invest- the World Bank (as part of research led by Kym Anderson on distor- ment and growth in the agricultural sector. tions to agricultural incentives) and included data up to 2004. More Presently agriculture’s share of national GDP is about 12 percent, a recently, Pursell (2011) provided a useful description of the com- proportion which, as in most countries, has been declining in Sri plexities of trade policy in Sri Lanka, although not specific to agricul- Lanka as the economy grows. Agriculture employs about 32 percent ture. The World Trade Organization (WTO) Trade Policy Review on Sri of the workforce in a nation that remains overwhelmingly (80 per- Lanka (Central Bank of Sri Lanka 2009, Annual Report 2008) has also cent) rural; the Household Income and Expenditure Survey (HIES) described more recent developments in the nation’s trade regime. of the Department of Census and Statistics for 2009/10 records a All of these publications provide some historical perspective on total population of 20.3 million and a rural population of 16.3 mil- how Sri Lanka’s agricultural trade policy has evolved and how the lion. In other words, agriculture’s share of employment remains inherent trade-offs have changed over time. A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R 2 CHAPTER 1 — INTRODUC TION Using a Domestic Resource Cost (DRC) approach, some analysts through tariffs on imports and taxes on exports, and (6) the cost- have sought to understand whether Sri Lanka has a comparative ad- effectiveness of the current fertilizer subsidy scheme. This informa- vantage in rice production. For example, Rafeek and Samarathunga tion should improve our understanding of the complicated web of (2000), in a review through 1992, reported nominal rates of pro- income transfers among producers, consumers, and government tection (NRPs) and effective rates of protection (ERPs) for rice and accounts. demonstrated positive protection to rice farmers. In 2002, the No recent quantitative evaluation has been available for this ex- Department of Census and Statistics reported that more than 80 tremely complex system, which has been subject to continuous percent of paddy holdings were less than 1 hectare; fewer than revision. Experiences from other countries show that when a trade 5 percent of farmers had a holding greater than 2 hectares regime is so discretionary and complex, it can generate consider- (Department of Census and Statistics 2002). Several studies con- able uncertainty about expected returns. Such uncertainty acts cluded that the income earned from paddy was not sufficient to as a significant deterrent to private investment, both in primary meet the basic needs of a family, so poverty was prevalent among agriculture as well as in agro-processing, which in turn stifles the small-scale growers (Ranaweera et al. 1990; Gunawardena 2000; diversification of agricultural production and exports. Weerahewa et al. 2002). Shilipi (1995), using 1993 data in a DRC ap- proach, concluded that Sri Lanka had no comparative advantage From the longer-term perspective, reforms in trade and price policy in growing paddy rice. Similarly, Rafeek and Samarathunga (2000) can be linked with complementary reforms in factor markets (such concluded that US$ 1 of resources was used to produce only as markets for land and irrigation water). Certainly trade reforms US$ 0.56 worth of rice. Wijayaratna et al. (1996), however, showed would affect (and probably increase) the average returns to land that in 1993, when irrigation and land costs were excluded from and water in agriculture, because they would induce the realloca- the accounting—two peculiar assumptions—rice growing in Sri tion of those resources to higher-value uses. Improved markets for Lanka could be competitive. Kikuchi et al. (2000, 2001) concluded land and water would enable producers to benefit more fully from that although Sri Lanka’s comparative advantage in paddy farm- trade policy reform by giving producers greater flexibility to adopt ing had been declining over the years, the country was still com- more competitive crop mixes and tenure arrangements. Experience petitive in paddy cultivation under major irrigation schemes. Finally, with trade liberalization and deregulation in agriculture in other Weerahewa et al. (2002) and Thibbotuwawa and Weerahewa (2004) middle-income countries (for example, in Latin America during showed a comparative advantage in growing paddy on relatively the 1990s) shows that there were significant—and often unantici- larger farms. Weerahewa (2004), who assessed household-level pated—adjustments to output composition, often in response to impacts of rice trade liberalization, concluded that the poor would export market potential. In many cases, the product mix evolved reap the benefits of rice trade liberalization—a finding contrary to to include high-value products aimed at foreign markets. Examples the general expectation. include certain regions of Brazil, the central valley of Chile, parts of northern Mexico, and the coastal range of Peru. In light of the con- The main objectives of this analysis on the impact of current trade tinuing food security discussion in Sri Lanka, it is noteworthy that and price policies on agricultural incentives is to provide a quanti- those countries are net importers of several important food prod- tative assessment of (1) the level of support to farmers producing ucts. It is also notable (although not necessarily the most important import-competing products, (2) the degree to which final con- criterion for pursuing similar policies) that these countries show a sumers are taxed, (4) the extent to which agricultural exports are net agricultural trade surplus. taxed, (5) the contribution of trade policy to government revenue W H AT I S T H E C O S T O F A B O W L O F R I C E ? T H E I M PA C T O F S R I L A N K A' S C U R R E N T T R A D E A N D P R I C E P O L I C I E S C H A P T E R 2 — B A C K G R O U N D : S R I L A N K A T R A D E P O L I C Y A N D A G R I C U LT U R E 3 Chapter 2 BACKGROUND: SRI LANKA TRADE POLICY AND AGRICULTURE This section offers considerable detail on the evolution of trade 2.1 BRIEF INTRODUCTION TO THE COMPLEXITY OF policy and agriculture in Sri Lanka in recent years. It describes SRI LANKAN IMPORT POLICY the structure of trade in agricultural products and the prevailing In 1994, under the Uruguay Round WTO agreement, Sri Lanka import/export policy framework. It concludes with an examina- signaled the opening of its trade regime by committing to tion of overall trade policy in relation to government revenues “bound” tariffs on agricultural products that were considerably and expenditures, particularly with respect to agriculture. lower than those of its neighbors in South Asia. More recently, TABLE 2.1: Primary Imported and Exported Agricultural Products, Selected Subsectors, 2009 PRIMARY TOTAL PRODUCTION IMPORT VALUE EXPORT VALUE CONTRIBUTION TO COMMODITY AREA (HA) YIELD (KG/HA) (MT) (US$ 000S) (US$ 000S) AGRICULTURAL GDP (%) Import-competing activities Paddy 977,561 4,336 3,651,674 22,392 (rice) 2,903 16.75 Potato 4,139 14,910 61,705 22,973 11 8.54 Maize 50,857 2,550 129,769 6,352 0 (all other field crops) Chilies 13,554 3,420 46,414 42,728 727 Red onions 4,498 10,280 46,234 50,264 11 Green gram 8,569 1,080 9,258 13,364 61 Milk 1,136,860* 19,443,024** 140,226 48 7.45 (all livestock) Export products Tea 222,000 1,515# 291,000 22,038 661,825 10.51 Rubber 124,000 1,437 136,900 523 115,989 3.14 Coconuts 395,000 3,000 2,853 145 50,040 8.57 Purchased inputs Machinery – – – 60,088 – – Seed – – – 4,552 – – Fertilizer – – – 181,493 – – Fuel*** – – – 754,075 – – Chemicals – – – 32,261 – – Source: AgStat, Central Bank, Trade Map. *Stock of cattle. **Total milk production in liters. *** All users, not only agriculture. # Tea production is 291 m kg, and total tea area is 192,000 ha, according to the Central Bank of Sri Lanka. A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R 4 C H A P T E R 2 — B A C K G R O U N D : S R I L A N K A T R A D E P O L I C Y A N D A G R I C U LT U R E while maintaining its commitments within WTO (and perhaps BOX 2.1: VARIATION IN SRI LANKA’S EXTERNAL as a reaction to the global food crisis and subsequent export POLICY FRAMEWORK, 2007–11 restrictions by its trading partners), Sri Lanka introduced para- 2007: The five-band tariff structure of 0, 2.5, 6, 15, and 28 per- tariffs and other de facto protective measures on agricultural cent continued, but duty rates on consumer goods were sub- imports, along with import duty waivers on a few essential ject to frequent changes. Surcharges on customs duty (10 per- goods to contain inflationary pressures. cent of customs duty), Value Added Tax (VAT), Port and Airport Development Levy (PAL), Social Responsibility Levy (SRL), ex- Before describing import and export policies more fully, how- cise duty (on alcohol, tobacco products, and vehicles), and a ever, it is relevant to discuss the structure of Sri Lanka’s trade in cess on nonessential consumer items prevailed. Single specific customs duty rates were introduced on 10 food items, instead agricultural products in recent years. In U.S. dollar values, the na- of the surcharge, VAT, SRL, cess, and other charges. tion’s primary agricultural exports are tea, rubber, and, to a lesser 2008: Wide fluctuations in prices of essential consumer items extent, coconuts (table 2.1). Milk, red onions, and chilies are the on the world market necessitated temporary measures to re- main agricultural imports, followed by potatoes, rice, and tea duce the impact on domestic prices. Measures included duty (wheat is not produced in Sri Lanka and the entire requirement waivers on a few essential goods to contain inflationary pres- is imported). Tea is notable because Sri Lanka, a traditional tea sures. To curtail pressure on the trade deficit to expand, mea- sures were taken to adjust tariffs and impose margin deposit exporter, imports some varieties of tea for blending and subse- requirements on certain imports that were deemed nones- quent exportation. Additional agricultural imports include palm sential. Such measures were removed or reversed to recoup oil, sugar and sugar confectionary, onions, soybeans, and wheat. the government’s tax revenue. The five-band tariff structure The import bill for fertilizer and fuel is high. Table 2.1 also shows continued. The surcharge on customs duty was increased to 15 percent from 10 percent. To contain rising prices of essential the high share of rice in agricultural GDP. food items, the customs duty, VAT, PAL, SRL, and other charges The complexity of Sri Lanka’s system of trade taxes and tariffs re- applicable at the customs point for 11 essential food items were replaced by a lower, single Special Commodity Levy (SCL). flects not only the multiplicity of taxes but the year-to-year vari- ation and lack of uniformity in taxes and tariffs across products. 2009: The key policies implemented included introduction of a Simplified Value Added Tax scheme, reduction of VAT from Box 2.1 presents a year-by-year account of trade policy changes 15 percent to 12 percent, and the introduction of an Export relevant to agriculture. For example, in 2009, on customs duties Development Reward Scheme, with the objective of securing Sri Lanka maintained a five-band tariff structure: 0 percent for existing markets, penetrating new markets, and establishing imports of essential goods, 2.5 percent for basic raw materials, and promoting forward and backward linkages during times of crisis. Though the five-band tariff structure introduced in 6 percent for semiprocessed goods, 15 percent for intermedi- 2004 continued until 2009, several other trade-related taxes ate products, and 28 percent for other finished products. This were imposed during 2009 to generate revenue and protect structure was revised to four bands in June 2010 (0, 5, 15, and domestic industries, in addition to customs duty. They includ- 30 percent) (Central Bank of Sri Lanka 2010). ed a cess, SCL, Nation Building Tax (NBT), and VAT. The average import duty collection rate was 7.8 percent in 2009. With the Sri Lanka also applies several para-tariffs under an extremely in- escalation of prices of certain essential commodities, the gov- tricate system that influences the level, dispersion, and predict- ernment revised taxes on imports of some items. Accordingly, customs duty, PAL, NTB, SRL, VAT, and surcharges were replaced ability of the structure of prices in the economy. As a result, Sri by a lower SCL. Lanka’s import regime presently is cited as one of the most com- 2010: Policies in 2010 focused on promoting trade through plex and protectionist in the world (Pursell 2011), with a reputa- diversifying the export product base and markets. The govern- tion for ad hoc policy changes (as seen in the frequent changes ment took decisive steps to encourage exports of value-added in border charges). The WTO (2010) Trade Policy Review of Sri and finished goods and to minimize the adverse impact of Lanka recommended rationalizing the incentive regime to im- supply constraints and price increases in international markets prove resource allocation, improve overall economic efficiency, W H AT I S T H E C O S T O F A B O W L O F R I C E ? T H E I M PA C T O F S R I L A N K A' S C U R R E N T T R A D E A N D P R I C E P O L I C I E S C H A P T E R 2 — B A C K G R O U N D : S R I L A N K A T R A D E P O L I C Y A N D A G R I C U LT U R E 5 ƒ Excise (special provisions) duties: A tax on CIF value, plus on the domestic market. Tariff bands were reduced to a four- the CESS levy, plus the PAL. Alternatively excise tax can be a band customs duty structure of 0, 5, 15, and 30 percent. The specific tax on the quantity of imports. customs duty and SCL applicable on selected consumer and ƒ The Social Responsibility Levy (SRL) is a rate applied intermediate goods such as petrol, diesel, milk powder, maize, palm oil, wheat grain, malt extract, and PVC leather cloth were to the sum of the customs duty, plus surcharges, plus the reduced to stabilize price fluctuations in the domestic econo- excise (special provisions) duty. my. A cess was imposed on several exports of raw and semi- ƒ A Value Added Tax (VAT) is applied to the total value of the processed items. import after it passes through the previous system of charges 2011: The 2011 budget included far-reaching reforms to sim- and levies—that is, a rate of the CIF value, times the customs plify the tax structure while broadening the tax base to im- duty, plus the CESS, plus the PAL, plus the excise tax. prove revenue mobilization. Measures were taken to rational- ƒ The Nation Building Tax (NBT) is an additional rate mul- ize the income tax structure by reducing marginal tax rates for tiplied by the same value of the good used in determining personal income tax and raising the tax-free threshold, with a the VAT. goal of improving tax collection. The base for income taxation ƒ Finally, a Regional Infrastructure Development Levy was widened with the extension of the PAYE tax to public sec- (RIDL) was based on a rate that, in 2009, mercifully was zero, tor employees. To create a level playing field, all tax incentives and was repealed in 2011. were brought under the Inland Revenue Act, other than those that are identified as strategically important and approved by The cascading effect of this multiplicity of taxes and duties can be the Cabinet of Ministers. Meanwhile, the VAT system was sim- seen from the following: plified by replacing multiple tax rates with a single tax rate of 12 percent. Several other initiatives were also taken to improve ƒ Total customs duties = CIF value in dollars * duty rate. tax administration. A Tax Appeals Commission was set up to ƒ SUR = Surcharge = total customs duty * surcharge rate. expedite the appeal process. ƒ CESS levy = (CIF value + 10% CIF value) * Cess rate or CESS Source: Central Bank of Sri Lanka 2007, 2008, 2009, 2010. levy = quantity * unit rate of CESS (per ton). ƒ PAL = CIF value * PAL rate. ƒ Excise duty = (CIF value + 15% CIF value + customs duty + Cess under EDB + PAL) * rate of excise duty, or Excise duty = eliminate distortions, and diversify production and trade. The Quantity * Unit rate of excise duty. WTO report suggested reductions in import protection, the ƒ SRL = (customs duty + surcharge + excise duty) * SRL rate. rationalization of domestic support, and the implementation ƒ VAT = (CIF value + 10% CIF + customs duty + Cess under EDB of a more consistent trade policy to increase productivity and + PAL + excise duty) * rate of VAT. farmers’ incomes. ƒ NBT = (CIF value + 10% CIF value + customs duty + Cess under EDB + PAL + excise duty) * rate of NBT. At the base of the tax structure is the total Customs Duty (CD), ƒ RIDL was zero in 2009 and was repealed January 1, 2011, which is a percentage of the CIF (cost, insurance, and freight) with some remaining taxes on motor vehicles (ending 2012). value. In Sri Lanka (as in many other countries) these tariff ƒ Total taxes = NRP = customs duty + surcharge+ CESS levy + rates vary significantly across tariff lines. In addition, for some excise duty + VAT + NBT + PAL+SRL + RIDL. goods there are specific duties per unit of quantity imported. In years of domestic production shortfalls and/or higher than On top of that, a wide variety of other charges and levies exists: average world prices, the government has revised taxes on ƒ CESS levies (assessments under the Commodity imports of some consumer items such as rice, potatoes, red on- Export Subsidy Scheme) applied to both imports ions, big onions, garlic, sugar, and chilies. For example, in 2009 and exports, defined either ad valorem or as a specific charge per unit quantity. import taxes were lowered to make food items more affordable. ƒ Port and Airport Development Levy (PAL), which acts as A lower SCL tax was charged in place of customs duty, PAL, NBT, a simple tariff on the CIF value of the imports. and SRL. A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R 6 C H A P T E R 2 — B A C K G R O U N D : S R I L A N K A T R A D E P O L I C Y A N D A G R I C U LT U R E In this report, the total protection rate (TPR) on imports is de- VAT’s importance to the outcomes. Box 2.2 discusses the VAT in fined as: CD + PAL+ SUR + Cess + RIDL, expressed as percent- relation to agricultural products and inputs in Sir Lanka. ages of the CIF price. Excise Duty, RIDL, and SRL were excluded The SCL also requires special attention. It is applicable to select- because they do not apply to agricultural imports. VAT is a com- ed, “essential” food items, and when applied it replaces all other plex case, because in principle it applies to both domestic and border taxes on those specific goods. The original list of items international trade, but in practice it is seldom charged in the covered by the SCL included maldive fish and sprats, potatoes, domestic trade of farm products. In discussions of the prelimi- onions (red and Bombay onions and garlic), peas (chickpeas, nary results of the TPR analysis, questions were raised about the BOX 2.2: IMPACT OF VALUE ADDED TAX ON AGRICULTURAL COMMODITIES IN SRI LANKA The Value Added Tax (VAT) was introduced by Act No.14 of 2002 and has been in force since August 1, 2002. The VAT Act replaced the nearly identical Goods and Services Tax (GST) on the consumption of goods and services. Every person who carries out a taxable activ- ity should register for VAT if the value of his/her taxable supplies exceeds or is likely to exceed Rs 500,000 per any quarter of the year or Rs 1,800,000 per year. Certain products are exempt from VAT (including certain imports). The customs agency has identified the VAT- exempt goods by their HS code classification*, in consultation with Inland Revenue: unprocessed agricultural produce, unprocessed horticultural produce, unprocessed fishing products, bread, milk, and agricultural tractors. Under the VAT Act No. 14 of 2002, the sole power to change the VAT lies with the Ministry of Finance; the Director General of Customs is not permitted to make any changes, as per recommendations made by other ministries and authorities. Exports are within the official scope of the VAT system but are subject to VAT at zero-percent. As the table shows, trade in agricultural products and, importantly, their tradable inputs is, for the most part, exempt from VAT. VAT was not incorporated into the calculations of total protection rate in this note. VAT ON IMPORTED AND EXPORTED AGRICULTURAL PRODUCTS AND THEIR TRADED INPUTS INPUTS OUTPUT TAX SEED/PLANTING ITEM CATEGORY FERTILIZER MATERIAL MACHINERY PESTICIDES Imports Rice Exempt Exempt Exempt Exempt Paid Wheat Exempt n.a. n.a. n.a. n.a. Potatoes 12% Exempt Exempt Exempt Paid Onions 12% Exempt Exempt Exempt Paid Chilies 12% Exempt Exempt Exempt Paid Liquid milk, infant milk Exempt n.a. n.a. Exempt n.a. Powdered milk Exempt n.a. n.a. Exempt n.a. Exports Coconuts 0% Exempt Exempt Exempt Paid Tea 0% Exempt Exempt Exempt Paid Dried fish Exempt n.a. n.a. Exempt n.a. Rubber 0% Exempt n.a. Exempt n.a. Condiments (cardamom, 0% Exempt Exempt Exempt Paid cinnamon, cloves, nutmeg, pepper) Source: Central Bank of Sri Lanka. Note: n.a. refers to “not applicable”; the inputs are not used in production of the corresponding product. * HS code is the Harmonized Commodity Description and Coding System for tariff nomenclature. W H AT I S T H E C O S T O F A B O W L O F R I C E ? T H E I M PA C T O F S R I L A N K A' S C U R R E N T T R A D E A N D P R I C E P O L I C I E S C H A P T E R 2 — B A C K G R O U N D : S R I L A N K A T R A D E P O L I C Y A N D A G R I C U LT U R E 7 gram, dhal, and similar), some fruits (oranges, grapes and apples), to Sri Lanka. Even so, the main findings of this note suggest that chilies, condiments (coriander, cumin, and others), and tinned the current policy framework might in fact be working against the fish. Palm oil, coconut oil, sugar, and fish were added to the list government’s stated objective of greater food security. in a revision, and recently milk powder and rice were included. The SCL can and has been applied and removed on individual 2.2 EXPORT POLICY FRAMEWORK products, even during the same year. For example, in February Agricultural exports have been taxed in Sri Lanka in different 2009, the SCL on milk was increased from 35 to 55 Sri Lanka rupees periods in different forms. Broadly speaking, four main argu- (Rs) per kilogram, but then removed entirely on March 31 of 2009, ments are used to justify taxing agricultural exports. The first which meant that other duties and taxes were applicable begin- is to generate revenue for the government. The second argu- ning in April. Although the SCL might act as a discretionary vari- ment is made when the product is a major staple in the diet of able levy to stabilize the domestic price of a product, the evidence urban consumers, and the tax is intended to reduce the price of is unclear as to whether SCL revisions in fact have been made to food. The third applies to countries that are major exporters of absorb fluctuations in world market prices (at least based on an in- a specific product in world markets—that is, when the country spection of the correlation of SCL rates with CIF prices). Appendix is not a “price taker.” An economist’s response to this condition table 3.5 shows a brief history of apparently ad hoc changes in SCL is to adopt an “optimum export tax.” The fourth argument is to levels. The SCL is applied as a specific duty, which for the purpose lower the domestic price of raw materials used by higher-value of calculating the TPR is converted into an ad valorem equivalent. industries to encourage further processing of local materials. Historically, import and price policy related to agriculture has been The first argument most likely applies to the taxation of ag- associated with concerns about food security. This note does not ricultural products in Sri Lanka. The second is not likely to be attempt to address the multiple facets of the food security issue, relevant, because Sri Lanka’s agricultural exports are not staple which is much broader then the trade policy focus here. The pro- crops. If it applies, the third argument would apply only to tea, duction (supply) side and its many related issues would require a but this note does not examine its application to Sri Lanka to- more comprehensive analysis of factors, including factors affecting day. Determining the level of that export tax requires agreement productivity such as technology (research and extension services), on the import demand elasticities in foreign markets, which is markets and marketing, and land and labor issues. Food security seldom reached. The fourth argument could be considered a concerns are discussed by Vishwanath and Serajuddin (2010) and kind of “infant industry” argument to promote the development 1 Kelegama (2010). Kelegama is explicit is his conclusion (p. 247): of the agro-processing industry. The literature on applied inter- “Priority should be given to production of food for the people of national trade suggests that the effectiveness of this fourth mo- Sri Lanka over commercial export-oriented agriculture in allocation tive for export taxes is questionable. The industry-specific data of land and resources.” This notion is reflected in the government’s required to support or refute this argument for Sri Lanka are National Policy on Agriculture from 2007, which presents initiatives beyond the scope of this study. to increase domestic production and restrain the rising cost of food. Sri Lanka imposes border charges (export cesses) on certain Beyond attempting to stimulate staple production, one such initia- exports, and a few items are under quantitative export restric- tive, the Api Vavamu–Rata Nagamu plan, is a “cultivation drive,” with tions (export bans and export licenses). Sri Lankan legislation the objective of gaining self-sufficiency within three years in onion, allows export duties and cesses to be used to meet the following chili, soybean, kurakkan, grape, orange, and potato production. stated objectives: to assure the availability of raw materials for Political and policy concerns about food availability are not unique higher-value-added industries and to promote further process- ing of local materials; to finance export promotion activities; 1 In the volume edited by Ahmed and Jansen (2010). and to protect national security, archaeological items, and the A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R 8 C H A P T E R 2 — B A C K G R O U N D : S R I L A N K A T R A D E P O L I C Y A N D A G R I C U LT U R E environment. Legislation also allows export duties to be im- and the manufacture-in-bond scheme. The system is designed posed in response to increases in world market prices. Section for manufacturing and processing and is unlikely to be accessible 10(a) of the Customs Ordinance provides for the levying of ex- for the farm sector. Like many countries, Sri Lanka implements a port duties, and Section 10(c) provides for any royalty or cess duty rebate scheme, through which duties paid on imported raw leviable on the exportation of such goods under other laws. The materials and intermediate inputs used in the manufacture of Export Development Act No. 40 of 1979 provides details of the products for export may be partially or fully rebated or refunded export cesses charged for various items, and the Gazette issued once the final good is exported. Exporters who incur fiscal levies on May 31, 2010 provides detailed export cess rates. The rates are on imported inputs and use these items to manufacture products ad valorem for certain items and have specific duties for certain for export directly or indirectly are eligible. Exporters who intend others. A third category can have ad valorem or specific rates, to use the scheme must register with the Duty Rebate Unit of the depending on which is greater. Unweighted average export Sri Lanka Customs Export Division. Registration is required for each cess rates in percentage terms for the agricultural commodity export shipment and must be accompanied by an export invoice, categories were computed based on the ad valorem rates in the receipt or air waybill, and parties’ copy of the export shipment. first or third categories. Similarly, using the specific rates in the Under the duty rebate scheme, only customs duties and the second and third categories, unweighted average specific cess import surcharge are rebatable; the cess, excise duties, PAL, and rates were computed in relevant units. When an item was not other duties levied on imports are not included. Likewise, the listed in the gazette, the export cess rate was assumed to be zero, VAT does not qualify for payment of duty rebate. Locally pur- both ad valorem and specific. When a rate was not applicable, it chased imported items that have been used to manufacture was considered a missing value (Appendix table 8). export products are considered for duty rebate.2 As shown in Bandara and Jayasuriya (2009), export taxes for Although not specifically related to the agricultural sector, Sri rubber, coconuts, and tea were high from the 1960s to approxi- Lanka also has a scheme for Export Processing Zones (EPZs). mately the mid-1990s, when they fell significantly until 2004. In these 11 free-trade zones, administered by the Board of Although export taxes for these three traditional agricultural Investment, some 220 enterprises employ over 75,000 persons. products remain low, broadly speaking, they remain surprisingly Locating an enterprise in an EPZ entitles a company to tax holi- high for nontraditional exports. The ad valorem export tax rates days, duty-free imports, and concessionary land prices. are relatively high for such products as fruits, vegetables, milk, and chilies, with a mode of 30 percent; maize, cinnamon, and Exporters of nontraditional goods or services also enjoy a num- cloves have a mode of 10 percent. Sri Lanka traditionally has ber of tax concessions, such as a preferential income tax rate used export taxes as a source of government revenue, but as on profits from those exports and a full tax holiday for three countries modernize, they typically move toward other, more to seven years for new investments. These companies are re- broad-based tax sources. As noted, export taxes discourage in- quired to export at least 80 percent of their goods production vestment in export-oriented agriculture and reduce the diversity and 70 percent of services. An earlier program, which appears of exports. These taxes on exports contrast sharply with the fairly to have been discontinued, subsidized the cost of international high protection of the import-competing sector, which implies a strong disincentive against exports in agriculture. 2 Exporters must submit a claim to the commercial bank that received the remittance for the exported products, along with relevant export With respect to positive export incentives, a complex system documents. The rate of rebate, a percentage of the FOB value, is de- termined by the Director General of Customs, based on the cost state- comprising several export incentive programs is available in Sri ments submitted and approved by the Deputy Secretary to the Treasury. Lanka. Customs operates three schemes: a duty rebate (drawback) Products eligible for duty rebates approved by the Minister of Finance are gazetted periodically. Exporters dissatisfied with a decision may ap- scheme, a temporary importation for export processing scheme, peal within 30 days. The rebate can be claimed within six months from the date of export. W H AT I S T H E C O S T O F A B O W L O F R I C E ? T H E I M PA C T O F S R I L A N K A' S C U R R E N T T R A D E A N D P R I C E P O L I C I E S C H A P T E R 2 — B A C K G R O U N D : S R I L A N K A T R A D E P O L I C Y A N D A G R I C U LT U R E 9 transport/freight of goods such as fresh fruits and vegetables, • Increasing the export cess on bulk tea to Rs 10 per kilo- foliage, live plants and preparatory materials, gherkins, cut flow- gram from Rs 4 per kilogram. ers, floricultural products, and betel leaves.3 ƒ Coconut subsector • Allocating Rs 500 million for replanting, new planting, in- In seeking to understand the most recent agricultural export tercropping, and productivity improvement in the coconut regime, it is particularly instructive to examine the 2011 budget industry. • Allocating Rs 200 million to curtail the spread of crop proposal on the treatment of agricultural exportables. At least in disease in the Weligama area. terms of the number of policy initiatives, the 2011 government • Encouraging coconut cultivation in Northern and Eastern budget shows a desire to promote export agriculture, but it also Provinces. manifests a degree of complexity and discretion that could act as • Leasing of unused land by government authorities (unless an obstacle to farmers and agro-processers operating at a small such land is developed by June 2011). and medium scale. There also seems to be a mix of public goods ƒ Rubber industry and private subsidies. While fighting the spread of crop disease • Allocating Rs 750 million to the Department of Rubber Development to increase the replanting and new planting or promoting new technologies might be public goods, subsidies subsidy and to grant a 50 percent subsidy to popularize for replanting and credit are not. To what degree do these private the use of rain guards. subsidies compensate for the export tax? Is balancing export taxes • Increasing the export cess on raw rubber to Rs 8 per kilo- the object of some of these subsidies? The 2011 budget proposes: gram from Rs 4 per kilogram. ƒ Spice subsector ƒ Reducing the concessionary tax rate on exports of nontradi- • Implementing a five-year subsidy scheme for planting and tional products from 15 percent to 12 percent. replanting. ƒ Imposing a cess on most exports in raw and semiprocessed • Providing financial assistance for industrial spice proces- form (tea, coconuts, rubber, cashews, raw hides) to encour- sors to promote value addition. age the exportation of value-added products. • Funding a special program to develop hybrid planting ƒ Reducing duties and taxes on machinery and equipment materials. used in value-added activities. • Promoting infrastructure to support smallholders in the ƒ Reducing the income tax rate to 10 percent for industries spice subsector. with over 65 percent value addition and brand names with patent rights reserved in Sri Lanka. 2.3 BRIEF NOTE ON THE COMPOSITION OF ƒ Reducing the corporate tax rate to 28 percent for domestic GOVERNMENT REVENUE AND EXPENDITURE manufacturing industries. Sri Lanka’s trade policy has relied mostly on price-based rather ƒ Tea industry than nontariff-based measures and has been guided to a large • Increasing the subsidy for replanting and new planting to extent by revenue considerations, according to the Central Bank Rs 50,000 per hectare for smallholders producing tea. of Sri Lanka (2009). In 2009 tax revenue constituted 90 percent • Establishing a revolving fund through the banking system to provide related credit facilities to plantation companies of total government revenue (nontax revenue accounted for for replanting. the remaining 10 percent). The tax revenue came from VAT (26 • Distributing unused Regional Plantation Company land percent), income tax (21 percent), other taxes (20 percent), im- among small growers for replanting. port duties (18 percent), and excise taxes (15 percent). • Supporting the Tea Research Institute to promote technology and research to improve the quality of tea Table 2.2 reports government revenue and expenditures for 2009, through high-yielding varieties and reductions in 2010, and 2011. Excluding VAT, border-related taxes represent postharvest losses. between 30 percent and 35 percent of tax revenues, a signifi- cant level, consistent with the claim that border measures in 3 WTO document G/AG/N/LKA/1, May 20, 1999. Sri Lanka are in part designed to raise government funds and A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R 10 C H A P T E R 2 — B A C K G R O U N D : S R I L A N K A T R A D E P O L I C Y A N D A G R I C U LT U R E TABLE 2.2: Government Expenditures and Revenues (Rs millions) 2010 2011 CATEGORY SUBCATEGORY ITEM 2009 (PROVISIONAL) (APPROVED ESTIMATE) Import-competing activities Tax revenue* 618,933 725,671 861,943 Income taxes 139,558 135,623 154,883 VAT 171,510 220,168 238,390 Excise taxes 97,604 129,749 152,250 Border measures Import duties 79,560 64,369 92,803 PAL 36,286 49,650 63,000 Other border-related taxes* 83,013 109,713 153,595 Other taxes 11,402 16,400 7,022 Revenues from border measures as % of total tax revenue 32.1% 30.8% 35.9% Nontax revenue 80,711 92,549 101,377 Total revenue 669,644 818,220 963,320 Current expenditure 879,575 937,094 1,017,155 Agriculture and irrigation 43,963 44,081 51,678 Fertilizer subsidy 26,935 26,028 20,000 Capital expenditure 277,416 302,087 351,899 Lending repayment 44,936 41,025 50,609 Total expenditure and net 1,201,927 1,280,205 1,419,663 lending Source: Central Bank of Sri Lanka. * Note that the cess on tea and coconuts in 2010 was Rs 100,490 million, or 91.6% of other border taxes. ** Includes stamp duties, cess, SRL, RIDL, SCL, and others. not merely to provide protection to local production. In 2010, removals and waivers on border taxes on specific products, includ- export taxes on tea and coconuts represented half of all border ing the customs duty waiver on wheat grain, a reduction in taxes tax revenue; import duties represented 29 percent of revenues on milk powder, removal of the SCL on rice, and the removal of raised at the border. Evidently domestic agriculture in Sri Lanka the cess on selected imports. These new and temporary changes is a significant contributor to public revenues. 4 demonstrate the high variability in the level of border taxes ap- plied by Sri Lanka from year to year. In computing the actual bor- Import duties represented approximately 9 percent of total tax der taxes charged for the subset of activities studied in this note, revenues in 2010. It is relevant to note (based on data from the the authors adjusted the various categories of border taxes follow- Central Bank of Sri Lanka 2009) that the “other taxes” category ing the Central Bank’s reported changes each year. includes the cess, SCL, NBT, and VAT. As discussed in Sri Lanka’s On the expenditure side, it is relevant to mention that the Central Bank Annual Report for 2010, trade-related taxes included fertilizer subsidy (discussed in greater detail in Section 4) rep- import duties, excise taxes, and taxes under the category of “other” resented 68 percent of current expenditures on agriculture in (SCL, PAL, and others). The Central Bank report outlines temporary 2009, 59 percent in 2010, and 39 percent in 2011. Though still absorbing a very high proportion of the public sector budget 4 In 2010 tea had a cess rate of 2.5 percent and coconuts of 11.82 percent. for agriculture, the fertilizer subsidy appears to be declining. Is it possible that Sri Lanka is not a price taker in world markets for these two products, which might justify an “optimal export tax” policy? This is Fertilizer subsidies represented approximately 3 percent of total unlikely, but the authors leave this question for further examination. government expenditures in 2010. W H AT I S T H E C O S T O F A B O W L O F R I C E ? T H E I M PA C T O F S R I L A N K A' S C U R R E N T T R A D E A N D P R I C E P O L I C I E S C H A P T E R 3 — D I S C U S S I O N A N D I N T E R P R E TAT I O N O F A G R I C U LT U R E 11 Chapter 3 DISCUSSION AND INTERPRETATION OF AGRICULTURE This section reviews the data sources and methods used to measure from border taxes and expresses it as a percentage of the border how border taxes influence levels of protection and taxation and price. All three methods have been applied in this study, but the presents statistics summarizing the degree of protection for 10 main results reported are based on the first approach. agricultural commodities. The discussion highlights how support to These three methods allow the estimation of a “nominal rate of pro- agriculture varies by commodity. tection,” which is defined as a price wedge between the border and domestic price of a given product. The difference between these 3.1 APPROACH, DEFINITIONS, AND DATA SOURCES two prices, given the exchange rate, is adjusted for quality differ- Two complementary approaches can be used to measure the impact ences between the domestically produced and imported product. of border taxes on levels of protection and taxation. One approach The first method gives a more direct measure of the government is to estimate the “ad-valorem tariff equivalent” of the various import interventions affecting domestic prices. The second method, the and export taxes applied per ton of a given commodity at the border. direct price comparison approach, also requires an adjustment for This approach allows us to synthesize the considerable variety of bor- domestic marketing margins so that the border and the farm gate der taxes in Sri Lanka (described in the previous section) into a single price can be compared at the common point of competition. This measurement for each commodity in each year. Another approach adjustment captures the structure of marketing, which in some used in the literature is to compute the “ad-valorem tariff equivalent” markets could be highly competitive and in others could be subject following a direct price comparison between border and farm gate to monopsonist and monopolistic forces, which in themselves are prices. The resulting figure is often referred to as the “import parity not necessarily the result of government policies. A complexity in price” or “export parity price.” The second approach is used to capture calculating the price comparisons occurs with products that un- nontariff restrictions on imports or exports, such as quantitative re- dergo processing before reaching the final consumer. For example, strictions in the form of import quotas, licenses, export prohibitions, a farmer sells a primary product, such as paddy rice or fresh milk, but pre-import deposits, and administrative measures. The first approach the import equivalent of those products is processed (polished rice, does not capture border measures unrelated to price, although such powdered milk and other milk products). The analyst has to obtain measures restrict imports and exports and influence domestic prices. the various transformation coefficients to arrive at the price equiva- Many countries imposed quantitative restrictions in the 1970s and lent at a given point in the marketing channel. 1980s, but partly as a result of the WTO Uruguay Round Agreement The nominal rate of protection (NRP) is the relevant measurement on agriculture, the legal scope for applying quantitative restrictions to examine the impact of trade policy on the price paid by agro-pro- has been significantly reduced. Such restrictions are less significant cessors for primary products or by the final consumer at retail. The today, but they are sometimes applied in a subtle or disguised way. NRP is not always an appropriate indicator of incentives for agricul- In addition to these complementary approaches, a third approach tural producers, however, because it does not capture the impact of captures the implicit import duty or export tax from the revenue price and trade policy on the domestic price of intermediate inputs A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R 12 C H A P T E R 3 — D I S C U S S I O N A N D I N T E R P R E TAT I O N O F A G R I C U LT U R E such as agrochemicals, machinery, equipment, fuel, and seed. For appreciate from the subsequent discussion, makes it difficult to gain a more accurate measure of the impact of trade policy on farmer a good understanding of the impacts on incentives.5 incentives, we look at the value added at the farm level. The effec- As explained in Section 2, at the base of the tax structure is the total tive rate of protection (ERP) is defined as the ratio between value customs duty, which is simply the CIF value of the imported good added at domestic prices to value added at world reference prices, multiplied by the duty rate, a proportion of the CIF value. These tariff that is (Vj – Vj*)/ Vj*), where Vj* represents value added per unit of rates vary significantly across tariff lines. In addition, some goods are output at border prices for the output and tradable inputs, given subject to specific duties per unit of quantity imported. On top of this the exchange rate. otherwise simple tariff, a variety of other charges and levies is added. This note reports nominal and effective rates of protection. It does There are surcharges over the normal duty rate. Under the authority not compute produce subsidy equivalents. This common measure of the act establishing the Export Development Board, there are cess of agricultural support, used by the Organisation for Economic Co- levies (assessments under the Commodity Export Subsidy Scheme) operation and Development for developed countries, represents a either calculated ad valorem as 110 percent of total CIF value multi- measure of income support and is particularly relevant in countries plied by an assessment rate, or calculated as a charge per unit quan- with income transfers and decoupled payments to farmers. tity. The Port and Airport Development Levy (PAL) acts as a simple tariff on the CIF value of the imports. An excise duty is a complicated The analysis was performed using tariff and para-tariff rates for tax on CIF value adjusted upward for customs duties plus the cess 2009 and 2011 (no customs tariff guide was available for 2010, so levy plus the PAL. Alternatively, the excise tax can be a specific tax the import tax analysis for 2010 was not performed). An important on the quantity of imports multiplied by a simple rate per quantity data source was the Customs Tariff Guide, which includes changes unit. The Social Responsibility Levy (SRL) is a rate multiplied by the made within the year as announced by the gazette notifications sum of the customs duty plus surcharges plus the excise duty. Then and reported on the Sri Lanka Customs website. Tariff and para-tariff there is the more familiar Value Added Tax (VAT) on the total value data for 2009 were kindly provided by G. Pursell, who downloaded of the import after it passes through the previous system of charges the 2009 database from the website at one point in time in 2009. and levies (that is, a rate multiplied by 110 percent of the CIF value Tariff and para-tariff data for 2011 were obtained from the hard copy of the Customs Tariff Guide, which was published in the beginning 5 Interviewed in the Sunday Times of Sri Lanka in 2010, Saman Kelegama, of year. The export  cess rates were from the government notifica- Executive Director of the Institute of Policy Studies of Sri Lanka and Member of the Presidential Taxation Commission, said, “Basically, our tion of Sri Lanka Export Development Act No. 40 of 1979, issued in import duty structure has got complicated by various ‘add on’ taxes over May 2010. and above tariffs and VAT (and Excise where applicable). These ‘add on’ taxes are revenue generators to the Treasury and additional protection for some domestic manufacturers but they are basically nuisance taxes To estimate the NRP and its equivalent—referred to as the total for import traders and importers of manufacturing inputs…. To meet protection rate (TPR)—the total taxes were first calculated using the certain development expenditures, the following taxes were imposed and they are applicable at the border: (a) Nation Building Tax, (b) Port formula applied by the Sri Lanka Customs authority. As explained and Airport Development Levy, (c) Regional Infrastructure Development Levy, and (d) Social Responsibility Levy. To develop certain commodi- in Section 2, this formula imposes certain taxes on taxes, rendering ties the following taxes were imposed: (1) Commodity Export Subsidy the total taxes greater than the figure obtained from a simple sum Scheme (CESS), and (2) Special Commodity Levy. In 2009, import tariff revenue was close to 2 percent of GDP but the customs border revenue of taxes. As Pursell (2011) points out, the Sri Lankan tax system af- was 8 percent of GDP, thus nearly 6 percent GDP revenue has come from these additional taxes…. All these minor taxes are an irritant to the fecting imports is one of the most complex in the world. The WTO importer although they fulfill some revenue and protection objectives. Uruguay-Round agreement aims for simplicity, transparency, and These taxes have made the tax structure more complicated and less transparent. And above all, these taxes operate under different tax bases. predictability. The WTO does not favor specific import taxes, be- One reason for illegal imports and undervaluation of imports is due to these taxes. What Sri Lanka should aim at is a less complicated border cause they insulate domestic prices from border prices and make tariff structure which gives reasonable protection and revenue.” (See “Tax protectionist measures difficult for traders and analysts to discern. Issues: Current System Not Delivering Needed Revenue,” Sunday Times, November 21, 2010, http://sundaytimes.lk/101121/BusinessTimes/bt10. The complexity of Sri Lanka’s import tax regime, as the reader will html, accessed April 2012). W H AT I S T H E C O S T O F A B O W L O F R I C E ? T H E I M PA C T O F S R I L A N K A' S C U R R E N T T R A D E A N D P R I C E P O L I C I E S C H A P T E R 3 — D I S C U S S I O N A N D I N T E R P R E TAT I O N O F A G R I C U LT U R E 13 plus the customs duty, plus the cess, plus the PAL, plus the excise applicable to all imported goods, both for net import-competing tax). The Nation Building Tax (NBT) is an additional rate multiplied by activities and for net exporting activities. the same value of the good used in determining the VAT. A Regional Table 3.2 presents the implicit ad valorem import duty collection Infrastructure Development Levy (RIDL), based on a rate of zero in rate calculated for selected food and nonfood consumer goods. The 2009, was repealed in 2011. Additionally, certain unprocessed prod- data have several features that are relevant to this analysis. One is ucts are heavily taxed upon entering Sri Lanka, probably to prevent the high dispersion of tariffs across various products in a given year. their entry for sanitary reasons. One good example is green chilies ver- For example, in 2010 the rice tariff was 2.1 percent; wheat flour, 40.8 sus dried chilies: The tax on green chilies is over 90 percent, whereas it percent; and wheat grain, 12.4 percent. The average duty on nonfood is 50 percent for dried chilies. consumer goods was 13.3 percent. Note also the dramatic changes Since changes in tax rates occurred throughout the years studied in duty rates through time. For example, the rates for milk and milk here, year-end taxes were used in the calculations. This study com- products were 6.4 percent in 2010 and 29.2 percent in 2009; the rates putes the “total protective tax,” which is equivalent to the NRP and for sugar were 4 percent in 2010 and 32.8 percent in 2008. Rice is an follows the formula used by the Sri Lanka Customs authority. The interesting case, with a duty rate of 23.4 percent in 2006 and then formula treats as protective the normal customs duties, the PAL, the a dramatic decline in subsequent years associated with increasing surcharge, cess, and RIDL. In other words, protective taxes exclude world prices. The implicit ad valorem is obviously influenced by fluc- the remaining tax rates (namely, the excise duty, VAT, NBT, and SRL). tuations in world prices, but these sharp changes in rates also suggest To summarize: The total protection rate (TPR) on imports = CD + PAL a policy designed to lower or raise the domestic prices of these goods. + SUR + cess + RIDL, expressed as a percentages of the CIF price. The coverage of protective taxes described here is not universally ac- TABLE 3.1: Protective Taxes on Agricultural Importables, 2009 cepted. The issue in Sri Lanka is that transactions involving domestic and 2011 (percentages) products escape some taxes that by definition should apply to all 2009 PROTECTIVE 2011 PROTECTIVE transactions of domestically produced and imported goods, such COMMODITY TAXES (TPR) TAXES (TPR) as the VAT. This issue is particularly relevant in agriculture, a sector Chilies 42.97 51.50 Maize 40.82 39.42 with a high degree of informal transactions and lax tax enforcement. Rice 7.20 33.84 The result is an increase in protection levels at the border, because Wheat 28.05 24.31 imported goods are subject to the customs duties and associated Vegetables 51.97 52.30 para-tariff as well as to the combined value added, nation building, Milk 47.09 45.33 social responsibility, and excise taxes. Fruits 68.61 66.17 Potatoes 40.94 44.44 For this analysis, the commodity groups are defined at the six-digit Onions 29.55 53.86 Harmonized System (HS) level, as some commodity groups can be AVERAGE 39.69 44.34 both imported and exported. For example, tea includes six tariff Machinery 11.05 8.15 categories, some of which are for imports. Seed 7.56 5.00 Fertilizer 7.88 5.00 Fuel (all users) 8.88 15.66 3.2 THE CASE OF AGRICULTURAL IMPORTS Chemicals 11.13 12.5 AND EXPORTS UNWEIGHTED AVERAGE 9.30 9.26 Levels of protection arising from trade-related and other taxes are Source: Task team’s calculations. Note: The table shows the TPR at the border, which conceptually corresponds to the NRP. The reader presented for selected agricultural products and tradable inputs in will note that the TPR for some products at the border differs from the TPR at the farm level. The TPR considered here includes import duties, PAL, the surcharge on top of the customs duty, and table 3.1. These results do not include export taxes but simply tariffs the cess on imports. VAT is excluded. A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R 14 C H A P T E R 3 — D I S C U S S I O N A N D I N T E R P R E TAT I O N O F A G R I C U LT U R E TABLE 3.2: Implicit Ad Valorem Import Tariff (percent) Based on Average Import Duty Collection Rate for Food and Nonfood Consumer Goods, 2008–11 ITEM 2008 2009 2010 2011 Average import duty 4.6 7.8 5.0 3.8 collection rate* Consumer goods 12.6 19.4 12.6 11.0 Food and beverages 14.8 26.8 12.1 12.3 Rice 3.1 5.4 2.1 21.2 Wheat flour 22.6 22.6 40.8 8.1 Sugar 32.8 26.5 4.0 5.0 Wheat and meslin 6.7 32.2 12.4 12.2 Milk and milk products 2.0 29.2 6.4 1.2 Dried fish 9.4 10.9 12.2 12.1 Other fish products 8.7 18.3 25.0 21.3 Other 24.2 27.2 19.6 21.0 Nonfood consumer goods 8.4 6.7 13.3 9.7 Source: Sri Lanka Customs. * Actual import duty collection (including SCL wherever applicable) as a percentage of total imports (CIF values); 2010 estimates are provisional. TABLE 3.3: Hidden Cost of a Complex System: Unweighted Average Import and Other Taxes (%) by Broad Commodity Groups, 2011 COMMODITY CUSTOMS DUTY SURCHARGE PAL NBT CESS EXCISE DUTY VAT Import and other taxes on export-oriented products* Tea 30.00 – 5 2.95 3.30 – 5.25 Coconuts 23.00 – 5 2.94 12.00 – 14.77 Rubber 19.08 – 5 2.90 12.34 – 17.38 Cinnamon 27.50 – 5 3.00 7.33 – 11.74 Cloves 30.00 – 5 3.12 11.00 – – Average 25.92 – 5 2.98 9.19 – 9.83 Import and other taxes on import-competing products Chilies 30.00 – 5 2.93 16.50 – 17.58 Maize 22.5 – 5 2.99 11.92 – 15.63 Rice** 26.64 – 5 2.88 2.20 – – Wheat 13.13 – 5 2.69 6.19 – 10.41 Vegetables 24.94 – 5 3.17 22.36 – 19.03 Milk 25.21 – 5 3.11 15.13 – 13.69 Fruits 30.00 – 5 3.52 31.17 – 21.14 Potatoes 25.00 – 5 2.96 14.44 – 16.06 Onions 30.00 – 5 2.49 18.86 – 17.09 Average 24.44 – 5 2.98 14.91 – 14.30 W H AT I S T H E C O S T O F A B O W L O F R I C E ? T H E I M PA C T O F S R I L A N K A' S C U R R E N T T R A D E A N D P R I C E P O L I C I E S C H A P T E R 3 — D I S C U S S I O N A N D I N T E R P R E TAT I O N O F A G R I C U LT U R E 15 TABLE 3.3: Hidden Cost of a Complex System: Unweighted Average Import and Other Taxes (percent) by Broad Commodity Groups, 2011 COMMODITY CUSTOMS DUTY SURCHARGE PAL NBT CESS EXCISE DUTY VAT Agricultural tradable inputs Machinery 3.15 – 5 2.36 – – 5.92 Seed – – 5 2.30 – – 11.04 Fertilizer – – 5 0.10 – – – Fuel 5.71 – 5 1.50 – 4.95 7.96 Chemicals 7.50 – 5 2.45 – – 14.70 Average 3.27 – 5 1.74 – 0.99 7.92 * Using broad product categories, the table shows import tax rates on products for which Sri Lanka is traditionally a net exporter, such as tea, but still imports in small amounts (usually of a slightly different type or quality). ** A specific customs duty of Rs 20/kg was charged for rice as per the 2011 Customs Tariff Guide; this duty is equivalent to 26.64% when the average CIF price is concerned. TABLE 3.4: Weighted Average Import and Other Taxes (percent) by Commodity Group, 2009 COMMODITY TOTAL TOTAL TAX: END-OF-YEAR PROTECTIVE EXPORT VALUES IMPORT VALUES NET EXPORTS GROUP TAX SIMPLE SUM TOTAL TAX TAXES (TPR) (US$ 000S) (US$ 000S) (US$ 000S) Import and other taxes on export-oriented products Tea 42.41 39.86 38.14 32.60 1,209,889 35,778 1,174,111 Coconuts 37.79 33.08 35.22 18.48 124,928 35,629 89,299 Rubber 38.56 33.93 35.96 19.76 594,272 194,840 399,432 Cinnamon 44.28 39.41 41.58 25.07 78,861 882 77,979 Cloves 53.43 50.68 50.27 48.20 12,007 0 12,007 Average 43.29 39.39 40.23 28.82 403,991.4 53,425.8 350,565.6 Import and other taxes on import-competing products Chilies 59.80 52.71 32.29 15.78 728 42,765 –42,037 Maize 59.11 50.49 62.25 39.56 5 10,052 –10,047 Rice 8.46 8.01 6.16 5.01 5,322 22,566 –17,244 Wheat 34.00 30.24 14.76 5.00 39,915 269,317 –229,402 Vegetables 39.63 35.88 35.99 22.00 62,154 175,648 –113,511 Milk 39.29 37.28 11.88 32.78 26,368 183,577 –157,209 Fruits 90.87 77.27 87.38 64.25 11,597 19,197 –7,600 Potatoes 94.94 23.13 23.67 93.24 57 25,358 –25,301 Onions 22.41 20.13 20.11 5.14 243 50,476 –50,233 Average 49.83 37.24 32.72 31.42 16,265.4 88,772.9 –72,509.3 Agricultural tradable inputs Machinery 24.02 21.64 16.47 8.89 – 60,088 –60,088 Seed 9.13 8.59 6.83 5.11 – 4,552 –4,552 Fertilizer 11.45 10.92 9.10 7.88 – 181,493 –181,493 Fuel 7.45 7.13 6.46 5.33 – 754,075 –754,075 Chemicals 29.98 26.73 27.55 11.63 – 32,261 –32,261 Average 16.41 15.00 13.28 7.77 – 206,493.8 –206,493.8 Source: AgStat, Central Bank, Trade Map. Note: These broad product categories include products of various levels of processing. A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R 16 C H A P T E R 3 — D I S C U S S I O N A N D I N T E R P R E TAT I O N O F A G R I C U LT U R E Table 3.3, which presents import and other taxes for various export- 3.3 MEASURING SUPPORT TO AGRICULTURE: NRP oriented and import-competing commodities, highlights several AND ERP implications of the current import regime. One implication is that The NRP and the ERP on various agricultural products capture the producers of most of these products are highly protected, suggest- profile of incentives for each one. As a reminder, NRPs is the appro- ing that consumers pay a higher price for them. In stark contrast to priate indicator to assess the impact of policies on consumers, and an import tax system based on a uniform rate for all products that ERP measures the impact of policies on producers’ value added (farm is stable through time, table 3.3 shows that a variety of taxes under income). One expects farmers’ investments in and use of various in- the Sri Lankan regime all apply different rates to the same product. puts to respond more to changes in value added than to changes in This level of complexity implies high transaction costs, particularly product price, because changes in value added capture the impact for small and medium-sized businesses. Moreover, it raises admin- of policies on the net returns to farmers. Appendix table 3 reports the istrative costs for the government, is vulnerable to lobbies and cost structure for various products, with an emphasis on the composi- rent-seeking behavior, and makes transparency difficult, increasing tion and the share in cost of tradable inputs. Tables 3.5 and 3.6 report the risk for investors and probably reducing investment. It confirms NRPs and ERPs for major import-competing and export products the uncertainty of the trade regime cited in the WTO Trade Policy for 2009–10 and 2010–11. Positive values imply protection, which is Review (2010). associated with import-competing activities; negative values imply taxation, which is associated with export activities. The relative importance of these taxes measured at the border is shown in table 3.4, which reports the TPRs (equivalent to the NPR at These tables present some significant findings. One is the enormous the border) and the associated value of exports and imports of the difference between NRP and ERP rates for the same product in the products to which they apply. (The reader should note that table 3.4 same year, which shows the influence of incorporating the impact presents data for broad commodity groups and that, for instance, of policies on purchased (tradable) inputs in the analysis. Sri Lankan the value of exports of the “wheat” group does not represent stand- trade policies have had a significant impact, not only on the domes- ard grain.) tic price of primary agricultural products but also on the domestic TABLE 3.5: NRPs and ERPs Calculated for 2009/10, Using the TPR NRP (PERCENT) ERP (PERCENT) PRIMARY TRADABLE OF TRADABLE OF PRIMARY COMMODITY COMMODITY COMMODITY COMMODITY IMPORT-COMPETING Paddy Rice 6.75 36.49 Potatoes Potato 54.24 149.09 Maize Maize 102.18 232.58 Red onions Red onion 3.00 0.70 Green gram Green gram 17.27 18.71 Fresh milk Milk powder 74.93 101.18 EXPORTABLES Made tea Made tea –2.12 –3.97 Rubber latex Rubber latex –1.42 –1.08 Coconuts: Fresh nuts Desiccated coconut –30 –46.65 Source: Task team’s calculations. Note: NRPs and ERPs were calculated using the TPR at the farm level (except for tea, rubber, and coconuts, which are calculated using the TPR at the wholesale level). These TPRs refer to a particular HS code. For example, for milk the border price corresponds to powdered milk converted to its fresh milk equivalent. W H AT I S T H E C O S T O F A B O W L O F R I C E ? T H E I M PA C T O F S R I L A N K A' S C U R R E N T T R A D E A N D P R I C E P O L I C I E S C H A P T E R 3 — D I S C U S S I O N A N D I N T E R P R E TAT I O N O F A G R I C U LT U R E 17 TABLE 3.6: NRPs and ERPs Calculated for 2010/11, Using the TPR NRP (PERCENT) ERP (PERCENT) PRIMARY TRADABLE OF TRADABLE OF PRIMARY COMMODITY COMMODITY COMMODITY COMMODITY IMPORT-COMPETING Paddy Rice 61.75 111.71 Potatoes Potato 2.14 41.52 Maize Maize 58.70 – Red onions Red onion 5.00 11.49 Green gram Green gram 5.00 4.21 Fresh milk Milk powder 276.55 907.63 EXPORTABLES Made tea Made tea –2.02 –4.24 Rubber latex Rubber latex –1.06 –1.84 Coconuts: Fresh nuts Dry coconut –30 –31.41 Source: Task team’s calculations. Note: NRPs and ERPs were calculated using the TPR at the farm level (except for tea, rubber, and coconuts, which are calculated using the TPR at the wholesale level). These TPRs refer to a particular HS code. For example, for milk the border price corresponds to powdered milk converted to its fresh milk equivalent. price of tradable inputs, which are influential in the cost structure of is used in the calculation. For example, the NBT rate is 2 percent for the various agricultural activities, as reflected in the ERPs. This effect all the broad commodity groups, but the amount charged as NBT on the intermediate inputs is not captured by the NRPs. A second differs from one category to another, because the customs duties significant finding is the very high and positive ERPs for import- and cesses differ depending on the category. competing activities in contrast to the negative ERPs for exporta- bles. These results imply strong incentives against trade and exports, making the trade regime inward looking. A third finding is the wide 3.4 ADDITIONAL OBSERVATIONS ON THE TREATMENT OF VARIOUS PRODUCTS dispersion in NRPs and ERPs across the various agricultural activities, as though policy makers intended to guide resource allocation in A number of additional observations on the treatment of particular the sector in a highly selective way. Was this an intentional move by products under current trade and price policies emerge from the policy makers? analysis: ƒ Paddy/Rice: The fertilizer subsidy is very influential in the Certainly the trade regime contributed to the generation of fis- final ERP estimate for rice. cal revenues from trade taxes as well as a higher degree of self- ƒ Potatoes have been a politically sensitive product with sufficiency in some food products, and it probably contributed to no comparative advantage. The price of potatoes in India obtaining support from the political base in farming community. On is rather low, and a very high, specific SCL was imposed on the other hand, high NRPs on food products imply higher prices for potatoes in 2009. The resulting ERP is substantial. consumers, which, for lower-income families, represent a consider- ƒ Maize too has been politically sensitive. It is used mainly in able real income loss. animal feed and used to be imported from the world market. The government imposed heavy import taxes on maize to The numbers show the amounts of taxes charged under each cate- protect maize farmers from lower world market prices. gory, but as noted they differ from the respective rates, as a formula ƒ Green gram too is highly protected through an import tax. A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R 18 C H A P T E R 3 — D I S C U S S I O N A N D I N T E R P R E TAT I O N O F A G R I C U LT U R E ƒ Chilies are generally profitable. Import taxes on chilies ƒ Tea is a traditional export, and the effect of the export varied quite frequently. cess is cancelled by the fertilizer subsidy (given to only ƒ Red onion farmers are constrained by the lack of high-quali- smallholders). ty seed, and illegal seed imports from India are an issue. ƒ Rubber also has an export cess. Though fertilizer is sub- ƒ Achieving self-sufficiency in milk production appears to be sidized for smallholders, they use very little, so the ERP is a high priority of the government; taxes on powdered milk negative. and feed are high. ƒ Fresh coconut production is taxed heavily through taxes on desiccated coconut and fertilizers. W H AT I S T H E C O S T O F A B O W L O F R I C E ? T H E I M PA C T O F S R I L A N K A' S C U R R E N T T R A D E A N D P R I C E P O L I C I E S CHAPTER 4 — COST-EFFEC TIVENESS OF FERTILIZER SUBSIDIES IN SRI LANKA 19 Chapter 4 COST-EFFECTIVENESS OF FERTILIZER SUBSIDIES IN SRI LANKA This section explores the rationale for subsidizing fertilizer in Sri to producers. If, however, there is some remediable obstacle to the Lanka, along with the analytical approach used in this study to use of resources in domestic production of farm products (such as understand farmers’ gains from subsidized fertilizer and the cost- rice), then social costs may be less than private (farmer) costs. For effectiveness of the subsidy program. Data on fertilizer supply and example, there might be some credit market constraints, perhaps demand elasticities in relation to fertilizer prices are reviewed, and arising from insecure property rights or other constraints. A produc- a simulation of the cost-effectiveness of fertilizer subsidies clarifies tion subsidy delivered through fertilizer purchases might compen- some of the trade-offs involved in maintaining the subsidy program. sate for those constraints and stimulate greater levels of production than might be achieved in an imperfect market. Is that really the case, however? And are fertilizer subsidies the most cost-effective 4.1 RATIONALE FOR SUBSIDIZING FERTILIZERS IN SRI LANKA means of countering market imperfections? Sri Lanka has subsidized fertilizer for more than four decades, re- Beyond the standard economic calculation, governments may per- ducing the prices paid by farmers relative to the border price of ceive some potential noneconomic benefits associated with a pro- imported nitrogen fertilizers (most urea fertilizer is imported). As duction subsidy. For example, the government might desire a level mentioned, fertilizer subsidies presently constitute a major expen- of self-sufficiency in rice production, and a fertilizer subsidy could diture in the government’s agricultural budget: approximately 68 be a practical mechanism to stimulate higher yields and total pro- percent of public expenditures on agriculture in 2009, 59 percent duction.6 Alternatively, the government might desire to use the sub- in 2010, and 39 percent in 2011. The subsidy program is a source sidy as a form of income transfer to low-income farmers, although of concern, because it diverts resources from public goods such as a general subsidy to fertilizer would not target this group and be rural roads, education, health, and research, among others. What are less cost-effective than a simple direct income transfer. Economic the arguments for fertilizer subsidies today? Is the subsidy program analysis can inform the policy maker of the economic costs associ- a policy response to a perceived liquidity constraint on working ated with these governmental objectives (for example, in terms of capital? What are the program’s social costs and benefits? Does it the excess resource costs of subsidized production over the value of generate positive (or negative) externalities? the imports displaced). These broad questions lead us to consider the welfare economics of this policy. Excluding other considerations (such as externalities associated with the subsidy and noneconomic objectives), a priori, a 6 The term “food security” is used to describe a wide variety of policy ob- jectives. It is important to distinguish between the food supply level at subsidy would drive a wedge between marginal costs and marginal the national level and the affordability of food at the household level. The first kind of food security is related to the security of aggregate do- benefits, leading to a net welfare loss (in terms of the sum of pro- mestic supplies, whether produced domestically or imported. The sec- ducer and consumer surpluses) from the misdirection of resources. ond is a question of household income (poverty). According to Sri Lanka State of the Economy 2010, “the physical availability problem (of food) is Traditional welfare economics suggests that a better approach to less clear, but appears to be the least serious. In terms of nutrition secu- rity, utilization at the household level is a serious issue of concern” (IPS aid producers would be to make a lump sum transfer from taxpayers 2010:101). Sri Lanka is about 96 percent self-sufficient in rice. A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R 20 CHAPTER 4 — COST-EFFEC TIVENESS OF FERTILIZER SUBSIDIES IN SRI LANKA Before describing the analytical approach for evaluating the cost- production would have some small effect on reducing the rice price effectiveness of the subsidy program, some background is useful. and thus the returns farmers receive from using additional fertilizer. In the four decades since its inception, the fertilizer policy has been Ekanayake’s conclusion appears to contradict World Bank (2002), the subject of several studies, particularly in regard to rice produc- which stated (although without empirical analysis) that the removal tion; see World Bank (2002), Wickmasinghe et al. (2005), Ekanayake of the fertilizer subsidy in 1989 had a strong impact on fertilizer use, (2006), Wijetunga, Thiruchelvam, and Balamuralli (2008), Weerahewa, and subsequent reinstatement of the subsidy induced a significant Kodithuwakku, and Ariyawardana (2010), and Weerasooriya and increase in use in the 1990s. Ekanayake suggested that fluctuations Gunaratne (2010). Ekanayake (2006) has reported than an important in fertilizer use were associated primarily with changes in paddy initial motivation of the subsidy was to promote the cultivation of prices at the time rather than fertilizer price changes. Ekanayake’s high-yielding varieties, which were particularly responsive to addi- results do not imply that an increase in fertilizer use would have no tional fertilizer; Ekanayake, along with Weerahewa, Kodithuwakku, impact on paddy yields (for example, as documented in Wijetunga, and Ariyawardana (2010), has recommended alternative policies Thiruchelvam, and Balamuralli 2008), but they address the question that the government might pursue. of the conditions that led to the observed increase in fertilizer use Before 1997, the subsidy applied to four types of fertilizer, but the noted in the World Bank report. The underlying issue is the substi- scheme was revised to apply only to urea, which according to the tutability between fertilizers and other inputs. If the substitutabil- World Bank (2002) implies concentrating the subsidy on paddy ity is low (in the extreme, fixed proportions) with respect to other farmers (who account for 75 percent of urea consumption). At inputs that are not subsidized, then the demand elasticity would least until 2000, the major share of the subsidy went to wealthier be low as well. An increase in price would induce an increase in all rice farmers. About 51 percent of the rice area was cultivated by inputs, however. households in the top 40 percent of farm households in terms of Weerasooriya and Gunaratne (2010), using supply and area response expenditure. World Bank (2002) also cited evidence of an imbalance functions for 1983–2007, examined the link between changes in in fertilizer nutrients: the nitrogen level is “excessive” and levels of productivity and fertilizer use associated with the subsidy. They phosphorous and potassium are less than optimal, with the result concluded that the subsidy had indeed increased productivity in that phosphorous and potassium are mined from the soil, depleting geographic zones with low-input, rainfed production systems in fertility over the long term. comparison to irrigated systems, where the authors found a limited Based on econometric estimates of fertilizer demand elasticities for yield response. 1981–2004, Ekanayake (2006) concluded that changes in the fertil- A recent study quantifying the consequences in Kirirndi Oya of izer price did not significantly affect fertilizer use. He estimated a fer- different levels of water access for farm productivities and income tilizer demand elasticity of –0.15 for urea but found a stronger cor- (Weligamage 2009) estimated rice production functions. The basic relation between fertilizer use and paddy prices. Fertilizer demand input variables to explain production levels included water, rice was relatively inelastic with respect to fertilizer prices, but fertilizer seed, fertilizer applications, machinery costs, chemical use, and total demand shifted outward in response to higher output prices. labor use (the estimation controlled for seasonality and geographic This finding has important policy implications. It suggests that the differences). Other explanatory variables included household char- main result of the subsidy is to transfer income to paddy farmers acteristics, such as schooling, age of farmer, years of farming expe- rather than to stimulate fertilizer use (which is high by international rience, distance to irrigation canals, and a factor for the quality of standards). Sri Lanka is a slight net importer of rice, so any increase the water supply. Weligamage’s results indicate that the elasticity of in rice production arising from the subsidy should have, in principle, paddy yield with respect to fertilizer application is 0.15. Based on a no effect on rice prices. Price transmission from border to domestic simple Cobb-Douglas conceptual framework (discussed later), this markets, however, is not perfect, and it is likely that additional rice estimate implies that the elasticity of fertilizer demand with respect W H AT I S T H E C O S T O F A B O W L O F R I C E ? T H E I M PA C T O F S R I L A N K A' S C U R R E N T T R A D E A N D P R I C E P O L I C I E S CHAPTER 4 — COST-EFFEC TIVENESS OF FERTILIZER SUBSIDIES IN SRI LANKA 21 to fertilizer price is approximately –1.2 and that elasticity of crop use without the subsidy (X0 units), in which case the quota is bind- yield with respect to fertilizer price is –0.18. This implicit own-price ing and the farmer finds it profitable to purchase additional fertil- demand elasticity derived from the production-response function izer beyond the quota at the higher market price. In this case, the at the micro level is high compared to that found by Ekanayake, subsidy would not affect the quantity of fertilizer and thus not affect who used aggregate time-series data. Typically micro-level results yields, but it does transfer a lump sum in the form of a simple rebate from individual farm data demonstrate less input substitutability on a portion of fertilizer used. The third and more interesting pos- than aggregate time-series data, so these results are puzzling. sibility is that the quota is binding but additional units of fertilizer present lower marginal benefits (at point e) than the marginal cost, Although it may be possible that the main policy objective of the which is the market price. In this third case, the farmer gains from fertilizer subsidy at present is to induce an untargeted income trans- the cost savings on the pre-subsidy use, area abfg (= S · X0), plus the fer to paddy farmers in general, regardless of the potential impact area behg. The total subsidy cost is area achf, which is simply the on rice production, this study measures the fertilizer subsidy’s po- product of the quota and the subsidy (S · Quota). The efficiency of tential impact on rice supply and farmers’ income. This measure is the transfer is greater with the quota than without, measured by used because it addresses the fiscal cost of inducing an extra ton of the ratio of total subsidy costs relative to total farmer benefits: achf paddy production—the additional production associated with an ÷ (abgf+behg). increment of X rupees spent on the subsidy (in other words, the cost-effectiveness of the subsidy). This ratio can be then be com- With the ability to resell the fertilizer at the market price, the farmer’s pared with the social cost of importing an additional ton of rice opportunity cost is the market price less whatever transaction costs under different scenarios about the cost of importing the specific are involved in finding a buyer, and the farmer would now face a rice varieties consumed in Sri Lanka. higher marginal cost of using fertilizer on his or her own production. With resale, the farmer would buy as much fertilizer as possible with the subsidy, use the original level of X0 in his or her own production, 4.2 ANALYTICAL APPROACH and sell the difference. This outcome is efficient in terms of increas- Figures 4.1 and 4.2 illustrate the subsidy analysis. The curve MB rep- ing the income of the farmer with a given level of total subsidy. The resents the marginal benefit to the farmer of an additional unit of efficiency gain is given by the triangle bce, which otherwise would fertilizer (the marginal value of product). Without the subsidy, the have been lost in resource misallocation. If farmers do not resell, cost of an additional unit of fertilizer for the farmer equals the mar- then it pays to buy more: The quota is less than X0 units. ket price, so the farmer will use X0 units. With the subsidy and no restriction on use, the farmer faces a lower effective fertilizer price FIGURE 4.1: Farmers’ Gains from Subsidized Fertilizer and consumes X1 units. The additional income that farmers receive Rupees as a result of the subsidy is given by the rectangle abfg (= S X0 – the savings on the pre-subsidy use) plus the triangle bounded by the a b c d P market points bgi (the extra net value of additional production). The cost of the subsidy is the larger rectangle adif (= S·X1); the efficiency of the transfer is measured by the ratio of total subsidy costs relative to subsidy S e total farmer benefits—that is, adif ÷ (abgf + bgi). g h i P effective With a quota on subsidized fertilizer, three possibilities emerge. The f quota might fall to the right of the optimal amount of fertilizer use with the subsidy (X1 units), in which case the quota is ineffective. The quota might fall to the left of the optimal amount of fertilizer X0 Quota X1 Fertilizer tons A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R 22 CHAPTER 4 — COST-EFFEC TIVENESS OF FERTILIZER SUBSIDIES IN SRI LANKA FIGURE 4.2: Cost-effectiveness of Fertilizer Subsidy One practical problem in this instance is that we do not know what the functional form is, and we do not have the data and resources to estimate a function for this simulation exercise. Instead we begin P market with the two parameter estimates available to us: (1) fertilizer demand elasticity with respect to fertilizer price and (2) the observed share of fertilizer costs in total revenue. These estimates help to approximate subsidy S the elasticity of production with respect to the price of inputs. P effective In any production function, at the point of optimum input use the production elasticity is the revenue share of fertilizer costs. A simple MB Cobb–Douglas functional form illustrates this point: X0 Quota π = py(x) – wx (y = kxα) 4.3 FERTILIZER DEMAND AND SUPPLY ELASTICITIES py wx f.o.c.: α –w=0€α= WITH RESPECT TO FERTILIZER PRICE x py This note considers a simple production function in which the and the demand elasticity is –(1–α)–1. Therefore the demand elastic- elasticity of supply with respect to the price of an input (such as ity for an input with respect to its own price is greater than one fertilizer) is negative and less in absolute value than the elasticity of in absolute value. For example, if the fertilizer cost share is 0.4 (the demand for the input with respect to its own price. The production value of α above), then the own-price demand elasticity is –1.67; elasticity with respect to the input level is less than one if declining and the supply elasticity with respect to fertilizer price is less by marginal products are observed, which is what one would expect. 0.4 or –0.67. One is unlikely to observe a production elasticity greater than one It is unlikely, however, that a simple Cobb–Douglas production func- without some outside constraint on farmers’ decisions, such as a tion will yield a good approximation of the production response of limit on fertilizer expenditures. Otherwise the producer would con- fertilizers over all levels of the various inputs. Indeed, Ekanayake tinue to increase input use, gaining more in extra revenues at the (2006) provided empirical evidence that the elasticity of demand margin than paying in extra costs. for fertilizer with respect to its own price is small, on the order of With a production function y = f(x), output price, p, and input price, –0.1. A simple Cobb–Douglas does not fit well, because it would w, a profit-maximizing farmer would choose optimal levels of input otherwise imply that α (the constant revenue share) is negative. A use and output supply based on prevailing prices x* = x(p,w) and small elasticity of fertilizer demand with respect to fertilizer price is y* = y(p,w). The relationship between the supply response and the not, in and of itself, a bad or good thing, but it would cause the input price change is necessarily determined by the response of absolute value of the supply elasticity with respect to fertilizer price production to input use and the response of input use to a change to be even smaller. A reasonable approximation, given a demand in its own price: elasticity of –0.1, can be obtained by multiplying that elasticity by ∂y* = ∂y x ∂x* the share of fertilizer costs in total revenue. For example, if the share ∂w ∂x ∂w of all input costs (labor, water, land, and so forth) in total revenue is 0.05, then the supply (yield) elasticity with respect to fertilizer price Converting this relationship to elasticities, one finds would be approximately –0.005. Benefit-cost numbers can then be w . ∂y* = x* . ∂y x w . ∂x* y* ∂w y* ∂x x* ∂w calculated using –0.1 for fertilizer demand elasticity with respect to fertilizer price and using 0.1 for the cost share, which is higher than εyw = αx · εxw that observed under the subsidy. In other words, without the sub- Note that εyw < 0, εxw < 0, 0 < αx, which implies |εyw| < |εxw|. sidy, the fertilizer cost share would increase. W H AT I S T H E C O S T O F A B O W L O F R I C E ? T H E I M PA C T O F S R I L A N K A' S C U R R E N T T R A D E A N D P R I C E P O L I C I E S CHAPTER 4 — COST-EFFEC TIVENESS OF FERTILIZER SUBSIDIES IN SRI LANKA 23 We have discussed the Cobb–Douglas function at some length TABLE 4.1: Fertilizer Price with and Without Subsidy, merely to illustrate how it could be used to calculate the supply 2009–10 elasticity with respect to fertilizer based on information about SUBSIDY AS % OF PRICE WITH UNSUBSIDIZED the share of fertilizer costs in total revenue and an estimate of SUBSIDY (Rs/kg) PRICE the own-price elasticity of fertilizer demand. As noted, the results DISTRICT CROP UREA TSP MOP UREA TSP MOP presented here are not based on a Cobb–Douglas or any other Ampara East Paddy 8 8 8 87 90 89 functional form; they are based on estimates from Ekanayake All other districts Paddy 7 7 7 89 91 91 Anuradhapura Maize 64 74 74 0 4 3 (2006) and on the observed cost shares of fertilizers. The price Anuradhapura Chili* 63 77 76 0 0 0 differential with and without the subsidy is large, so the elas- Anuradhapura Brinjal 72 67 64 0 13 16 ticity of fertilizer use, given fertilizer’s low price with the sub- Source: Task team, based on data from Cost of Cultivation reports, Department of Agriculture, Sri Lanka sidy, is bound to yield a low demand elasticity, consistent with Note: TSP = triple super phosphate; MOP = muriate of potash. To obtain the 2009–10 subsidy rates for the Maha season (the rainy season from October 2009 to April 2010), market prices are Rs 63/kg Ekanayake’s estimate. for urea, Rs 77/kg for TSP, and Rs 76/kg for MOP. Estimates are based on Anuradhapura chili cultivation (rainfed), which is not subsidized. For the Yala season (2009), market prices were Rs 62/kg (urea), Rs 61/kg (TSP), and Rs 65/kg (MOP), based on Anuradhapura chili cultivation (irrigated), which is not subsidized. Note also that the chili mixture fertilizer is excluded. 4.4 POLICY SIMULATION TO ESTIMATE THE COST- Table 4.2 shows the implied transfer efficiency of the fertilizer sub- EFFECTIVENESS OF THE FERTILIZER SUBSIDY sidy in paddy rice during 2009–10. The elasticity of fertilizer use with Table 4.1 shows fertilizer prices for various districts and crops respect to its own price is assumed to be –0.1, which implies that an during 2009–10. These data come from cost of cultivation re- increase, for example, of 688 percent in the price of urea (from Rs 8 ports of the Department of Agriculture, which present farm to Rs 63) would cause fertilizer use to decline from 113 kg/acre to gate prices of fertilizer by district. Here the farm gate price of 35.3 kg/acre. The elasticity of yield with respect to fertilizer price is a highland crop is the price of fertilizer without subsidy. The assumed to be –0.01, which is consistent with the demand elastic- subsidy on the three fertilizers represents approximately 90 ity. The cost share of fertilizers is assumed to be roughly 10 percent percent of the unsubsidized price of fertilizer in rice. Because without the subsidy. In other words, the 688 percent increase in fer- rice farmers get their fertilizer practically for free, the cost share tilizer price associated with ending the subsidy would reduce yields of fertilizer is very low, on the order of 3–5 percent of revenues. on the order of 8–9 percent (note that agronomic research could Without some binding quota on subsidized fertilizer, it would help to refine this information). These estimates assume that farm- not be surprising that the marginal product of fertilizer would ers use all of their purchased fertilizer on their own production. The be very low, perhaps near zero, approaching the maximum in calculation of net benefits assumes that other input costs remain the physical response function. This situation would create a roughly constant. In effect, the output elasticity with respect to fer- large incentive for paddy farmers to buy subsidized fertilizer tilizer price is double the elasticity implied by Ekanayake. and sell to other users, in and out of agriculture, who have to pay unsubsidized market prices for fertilizers. If there is a quota 7 The estimates here are based on the cost share observed under the on purchases of subsidized fertilizer, then the yield response to present subsidy (from 5 percent to 10 percent). The hypothetical cost share without subsidy would be unchanged under a Cobb-Douglas ap- the subsidy would be zero, and changes due to a changing sub- proach, but it is difficult to believe that the cost share would remain sidy would be infra-marginal. The low share of fertilizer costs in steady over the entire range of possible quantities of fertilizer use. It is more likely that the cost share would rise as the increasing input price total revenues is primarily caused by the very low price per unit causes the input level to fall and as the marginal product declines. If one had a cost-share estimate without subsidy, a first-order approxima- of the amount of input being used, whether or not it is being tion would make use of a weighted average of cost shares, with and rationed. 7 The analysis that follows first examines a scenario in without the subsidy. With little or no subsidy, one would likely observe a higher marginal product and a higher cost share, and the farmer would which there is no quota or it is ineffective and then focuses on be earning higher additional revenues for an additional kilogram of fer- tilizer use. As the subsidy increases, the additional revenues due to an a scenario in which there is an effective quota. additional kilogram of fertilizer falls, likely moving toward the point of zero additional revenue. A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R 24 CHAPTER 4 — COST-EFFEC TIVENESS OF FERTILIZER SUBSIDIES IN SRI LANKA TABLE 4.2: Implied Transfer Efficiency of Fertilizer Subsidy in Paddy Rice (2009–10), Without Binding Quota and Assuming That Farm Fertilizer Purchases Are Used in Production (Not Resold at Market Prices) PRODUCTION VALUE, LESS FARMERS’ FERTILIZER COSTS (Rs/acre) INCREASE IN COST OF SUBSIDY TO FARMERS’ INCOME TAXPAYERS RATIO: TRANSFER SEASON DISTRICT WITH SUBSIDY WITHOUT SUBSIDY (Rs/acre) (Rs/acre) EFFICIENCY Maha Ampara East 56,931 50,726 6,205 11,152 1.80 Maha Anuradhapura 64,631 58,374 6,257 11,305 1.81 Maha Hambantota 68,512 61,902 6,610 11,250 1.70 Maha Polonnaruwa 68,174 61,619 6,555 11,277 1.72 Maha Gampaha 49,781 45,093 4,689 6,415 1.37 Maha Kandy 38,760 35,176 3,585 6,316 1.76 Maha Kurunegala 53,932 48,754 5,178 9,764 1.89 Maha Kurunegala 65,122 58,937 6,185 10,240 1.66 Maha Whole island 62,099 56,089 6,010 11,570 1.93 Maha Whole island 49,123 44,475 4,648 8,408 1.81 Yala Ampara East (irrigated) 68,126 61,477 6,650 11,142 1.68 Yala Anuradhapura (irrigated) 65,831 59,544 6,286 9,396 1.49 Yala Hambantota (irrigated) 74,271 67,297 6,974 9,636 1.38 Yala Polonnaruwa (irrigated) 58,470 52,695 5,775 10,060 1.74 Yala Gampaha (rainfed) 40,410 36,536 3,874 5,998 1.55 Yala Kandy (rainfed) 38,622 34,517 4,105 9,790 2.38 Yala Kurunegala (rainfed) 48,544 43,859 4,685 7,448 1.59 Yala Kurunegala (irrigated) 59,309 53,610 5,699 8,908 1.56 Yala Whole island (irrigated) 61,990 55,929 6,062 10,182 1.68 Yala Whole island (rainfed) 45,824 41,425 4,399 6,836 1.55 Source: Task team, based on elasticity coefficients discussed in text and data on fertilizer use and cost shares from Cost of Cultivation report. Note: The transfer efficiency shows the number of rupees required on average from taxpayers to transfer one rupee to a farmer. As shown in the final three columns of table 4.2, the increase in a To recapitulate, the simulation of the transfer efficiency of the fertiliz- farmer’s net income per acre is small relative to the subsidy cost per er subsidy was estimated based on parameters from various sources acre, an efficiency ratio ranging between 1.37 and 2.38. On average, in Sri Lanka, which report estimates of fertilizer response functions, then, the government spends between Rs 1.4 and Rs 2.4 per acre fertilizer demand elasticity with respect to fertilizer price, fertilizer to increase a farmer’s income by only Rs 1 per acre. This high inef- used in response to changes in rice prices, the share of fertilizer in rice ficiency ratio is partly the result of a small, postsubsidy cost share of production costs, and the prices paid by farmers with and without fertilizers. At such inefficiency levels, farmers have an incentive to the subsidies. The simulation reveals that the government spends resell fertilizer rather than waste it on obtaining a very small increase Rs 1.4 to 2.4 per acre to increase income by Rs 1 per acre. Relative in production. In terms of a farmer’s income, giving a farmer one to a direct income transfer, the current program seems highly cost- more unit of an already low-priced input adds less to output if farm- ineffective. Removing the subsidy would reduce rice yields by about ers must use that input only in rice production rather than some 8–9 percent. Finally, the artificially lower cost of fertilizer induces other activity. higher use, with possible negative environmental impacts. W H AT I S T H E C O S T O F A B O W L O F R I C E ? T H E I M PA C T O F S R I L A N K A' S C U R R E N T T R A D E A N D P R I C E P O L I C I E S C H A P T E R 5 — I N C O M E D I S T R I B U T I O N A L I M P L I C AT I O N S O F   A G R I C U LT U R A L T R A D E P O L I C I E S 25 Chapter 5 INCOME DISTRIBUTIONAL IMPLICATIONS OF AGRICULTURAL TRADE POLICIES It is widely accepted that policies and programs can be implement- incomes. Real household income depends both on income from ed cost-effectively if they are well targeted to their intended ben- paddy farming and household expenditures on food, of which rice is eficiaries (for example, to vulnerable segments of the population). an important component. An examination of differences in income Practical difficulties in targeting recipients have led many policy sources and expenditure patterns of paddy farming households by makers worldwide to adopt “one-size-fits-all” policies and programs, region and farm size is followed by a simulation of the effect on despite the large financial outlays. Experience in many countries in- household real income of a hypothetical removal of price and trade dicates that such programs often produce a number of unintended interventions. The analysis is based on data from the Household outcomes, and in certain situations they have undermined their Expenditure and Income Survey (HEIS) of 2009/10 (Department intended objectives. The incentives provided to the paddy subsec- of Census and Statistics) and the NPRs and ERPs estimated for this tor in Sri Lanka and their consequences are relevant in this regard. study, presented earlier. Successive Sri Lankan governments since independence have To put the simulation results in context, table 5.1 presents Sri Lanka’s regarded development of the paddy subsector as a key agricultural national food supply and per capita availability for 2010 by product policy objective. Policies were designed to generate higher incomes category. In 2010, Sri Lanka was almost self-sufficient in oil, meat, for paddy farmers and alleviate rural poverty. Most paddy farmers are and fish (although not tinned fish). Sugar was a major import (im- and have been small-scale, semisubsistence producers, and protec- ports supplied 94 percent of consumption), but overall, Sri Lanka tion of the paddy subsector was perceived as a means to raise their is not significantly dependent on food imports. Self-sufficiency incomes. The intended outcomes were to develop a rice subsector rates for fruits were 91 percent; cereals, 86 percent; vegetables, 85 that would provide the bulk of a staple food to consumers and to im- percent; milk and milk foods, 76 percent; root crops, 74 percent; prove the wellbeing of rural families, both poor farmers and laborers. dried fish, 49 percent; and pulses, 23 percent. Sri Lankans tend to prefer a cereal-based diet, although traditional root crops are still This section describes results of a simulation to assess how trade and an important source of energy for low-income families. price policies affecting the rice sector have influenced household A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R 26 C H A P T E R 5 — I N C O M E D I S T R I B U T I O N A L I M P L I C AT I O N S O F   A G R I C U LT U R A L T R A D E P O L I C I E S TABLE 5.1: Food Balance Sheet, Sri Lanka, 2010 000 mt PER CAPITA AVAILABILITY SELF- GROSS AVAILABLE FOOD CALORIES PER PROTEIN FAT SUFFICIENCY ITEMS PRODUCTION IMPORTS SUPPLY* FOOD NET** (g/day) DAY (g/day) (g/day) RATE All food items 8,375.87 1,945.75 10,211.67 7,558.40 1,002.16 2,688.36 67.05 46.00 81 Cereals 4,469.73 736.82 5,191.81 3,396.73 450.59 1,561.83 33.12 2.69 86 Roots, tubers 381.15 130.89 512.04 387.40 51.38 66.60 0.56 0.08 74 Sugar 33.38 539.2 571.53 545.57 72.37 289.37 0.00 0.00 6 Pulses and nuts 45.18 152.31 196.88 188.61 25.02 90.30 6.24 1.39 23 Vegetables 997.14 170.52 1,148.96 1,052.06 139.56 77.14 3.62 0.50 85 (including onions) Meat 134.63 1.55 133.87 133.87 17.77 23.26 4.30 0.68 99 Fish 384.67 13.63 384.93 137.81 18.28 24.33 3.57 1.00 97 Dried fish 46.57 48.69 95.26 95.26 12.64 30.97 6.41 0.51 49 Tinned fish 0.00 19.18 19.18 19.18 2.54 4.38 0.53 0.02 0 Milk and milk 231.37 72.53 304.87 220.41 28.75 71.33 3.51 3.95 76 foods Oil and fats 1,025.28 8.51 989.86 728.35 96.62 353.95 2.83 33.81 99 (including coconuts) Fruits 558.43 51.79 594.02 585.85 77.72 78.47 0.99 0.24 92 Source: Department of Census and Statistics. * (Production + Imports) – (Change in Stocks + Exports). ** Quantities set apart for Seed, Animal Feed, Waste, and Manufacturing are excluded. 5.1 GENERAL CHARACTERISTICS OF PADDY portion going to home consumption. Only purchased inputs are FARMERS IN SRI LANKA considered in estimating costs and net income. (Family labor is not The HEIS sample is nationally representative, consisting of 19,958 included in costs, and net income is therefore returns to owned land, households, of which 2,719 engage in paddy cultivation. As shown to whatever investments have been made, and to family.) Whenever in tables 5.2, 5.3, and 5.4, paddy farmers have an average per capita the estimate of costs is greater than cash income, the household monthly income of Rs 11,544, which is 27 percent higher than the is considered to have zero income. No regional price adjustments national average (Rs 9,112). On average, across all types of rice farm- were made for income, but consumption was adjusted to reflect ers and regions, the average share of household income from paddy regional differences in the price of rice (sometimes referred to as farming is 9.96 percent (table 5.5). The share of income from rice pro- “real” consumption). duction varies widely by region and farm size, however. Households Paddy farmers’ average per capita monthly expenditures on all with larger farms depend more on rice production as a portion of goods (including the imputed value of home-grown rice) total Rs household income relative to households with smaller farms (table 7,714, slightly lower than the national average (Rs 7,833). Of these 5.6). In some regions, such as Batticoloa, rice represents a large aver- expenditures, per capita monthly spending on food is Rs 3,331 (that age share of income across all farm sizes (table 5.7). is, food’s share of household expenditure is 43.18 percent). The aver- Income includes money received as well as an imputed value to age per capita monthly expenditure on rice is Rs 602 and includes housing and to food produced by the household for internal con- the imputed value of home-grown consumption, which averages sumption. Rice production is valued at market prices, including that Rs 368 or slightly more than 60 percent of spending on rice. In other W H AT I S T H E C O S T O F A B O W L O F R I C E ? T H E I M PA C T O F S R I L A N K A' S C U R R E N T T R A D E A N D P R I C E P O L I C I E S C H A P T E R 5 — I N C O M E D I S T R I B U T I O N A L I M P L I C AT I O N S O F   A G R I C U LT U R A L T R A D E P O L I C I E S 27 TABLE 5.2: Incomes and Expenditures of Paddy Farming Households by Sector (Rural or Urban) and Province PER CAPITA MONTHLY INCOME PER CAPITA MONTHLY EXPENDITURE PADDY IMPUTED HOME SECTOR AND PROVINCE COUNT PRODUCTION TOTAL TOTAL FOOD* RICE GROWN RICE Sector Urban 88 1,288 18,865 13,340 4,280 442 241 Rural 2,596 1,152 11,479 7,656 3,322 603 370 Estate 35 175 5,501 5,365 2,740 613 107 TOTAL 2,719 1,150 11,544 7,714 3,331 602 368 Province Western 211 414 22,738 11,812 4,142 517 224 Central 257 525 7,179 7,837 3,205 553 317 Southern 327 1,311 14,225 7,532 3,265 595 379 Northern 66 1,217 6,021 6,849 3,629 482 362 Eastern 202 1,459 5,692 6,091 3,192 658 384 North Western 489 938 14,323 7,619 3,434 572 363 North Central 613 2,342 9,457 7,959 3,295 637 484 Uva 367 866 7,084 6,194 3,012 707 365 Sabragamuwa 187 544 10,811 6,884 3,037 566 306 TOTAL 2,719 1,150 11,544 7,714 3,331 602 368 Source: Task team’s calculations, based on HEIS 2009/10. Note: “Food” excludes alcohol and tobacco. TABLE 5.3: Incomes and Expenditures of Paddy Farming Households by Farm Size and Poverty Status PER CAPITA MONTHLY INCOME PER CAPITA MONTHLY EXPENDITURES POVERTY STATUS PADDY IMPUTED HOME AND FARM SIZE COUNT PRODUCTION TOTAL TOTAL FOOD* RICE GROWN RICE Poverty status Nonpoor 2,545 1,198 12,174 8,099 3,453 610 378 Poor 174 495 3,025 2,506 1,686 487 226 TOTAL 2,719 1,150 11,544 7,714 3,331 602 368 Farm size ≤1 acre 1,525 507 8,808 7,443 3,244 583 315 >1 and ≤2 acres 569 1,317 12,157 7,433 3,338 622 442 >2 and ≤3 acres 323 2,256 14,880 7,847 3,466 648 446 >3 and ≤4 acres 113 2,754 12,378 7,847 3,439 654 468 >4 and ≤5 acres 79 3,413 51,207 13,355 4,360 528 456 >5 and ≤10 acres 69 5,419 20,736 10,739 3,837 663 532 >10 acres 15 11,009 20,079 10,569 3,535 562 450 TOTAL 2,693 1,157 11,573 7,722 3,334 601 368 Source: Task team’s calculations based on HEIS 2009/10. Note: “Food” excludes alcohol and tobacco. A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R 28 C H A P T E R 5 — I N C O M E D I S T R I B U T I O N A L I M P L I C AT I O N S O F   A G R I C U LT U R A L T R A D E P O L I C I E S TABLE 5.4: Incomes and Expenditures of Paddy Farming Households by District PER CAPITA MONTHLY INCOME PER CAPITA MONTHLY EXPENDITURES PADDY IMPUTED HOME DISTRICT COUNT PRODUCTION TOTAL TOTAL FOOD* RICE GROWN RICE Colombo 36 838 18,127 12,481 4,433 501 269 Gampaha 60 403 39,081 14,050 4,624 496 258 Kalutara 115 314 10,842 9,850 3,682 539 185 Kandy 97 346 7,129 8,411 3,273 564 275 Matale 137 717 7,470 7,366 3,193 549 392 Nuwara Eliya 23 475 5,891 7,222 2,896 518 149 Galle 86 489 25,769 7,988 3,580 502 191 Matara 115 1,126 8,227 7,155 3,048 569 392 Hambantota 126 2,262 10,635 7,550 3,226 709 534 Jaffna 12 557 3,874 6,456 3,726 471 384 Vavuniya 54 1,623 7,342 7,091 3,569 488 348 Batticoloa 23 2,928 6,106 6,052 3,527 641 172 Amparai 90 1,257 5,041 5,727 2,960 661 435 Trinco 89 1,508 6,671 6,691 3,504 655 341 Kurunegala 423 935 14,671 7,503 3,378 565 354 Puttalam 66 961 11,368 8,605 3,908 629 444 Anuradapura 353 2,224 10,387 7,944 3,208 629 467 Polonnaruwa 260 2,581 7,563 7,989 3,474 653 519 Badulla 149 832 8,125 6,596 2,936 685 322 Monaragala 218 902 6,005 5,776 3,089 730 409 Ratnapura 84 867 13,602 7,000 3,113 658 339 Kegalle 103 235 8,149 6,774 2,965 478 275 TOTAL 2,719 1,150 11,544 7,714 3,331 602 368 Source: Task team’s calculations, based on HEIS 2009/10. Note: “Food” excludes alcohol and tobacco. TABLE 5.5: Income Share from Rice and Expenditure Shares on Rice of Paddy Farming Households by Sector (Urban, Rural, Estate) and Province INCOME SHARE LESS HOUSEHOLD CATEGORY INCOME SHARE EXPENDITURE SHARE EXPENDITURE SHARE Sector Urban 6.83 3.31 3.52 Rural 10.03 7.88 2.15 Estate 3.17 11.42 –8.25 TOTAL 9.96 7.80 2.16 Province Western 1.82 4.38 –2.56 Central 7.32 7.06 0.26 Southern 9.21 7.90 1.31 Northern 20.20 7.03 13.17 Eastern 25.63 10.80 14.84 North Western 6.55 7.51 –0.96 North Central 24.76 8.00 16.76 Uva 12.23 11.42 0.81 Sabragamuwa 5.03 8.23 –3.20 TOTAL 9.96 7.80 2.16 Source: Task team’s calculations, based on HEIS 2009/10. W H AT I S T H E C O S T O F A B O W L O F R I C E ? T H E I M PA C T O F S R I L A N K A' S C U R R E N T T R A D E A N D P R I C E P O L I C I E S C H A P T E R 5 — I N C O M E D I S T R I B U T I O N A L I M P L I C AT I O N S O F   A G R I C U LT U R A L T R A D E P O L I C I E S 29 TABLE 5.6: Income Share from Rice and Expenditure TABLE 5.7: Income Share from Rice and Expenditure Shares on Rice of Paddy Farming Shares on Rice of Paddy Farming Households by Poverty Status and Farm Hhouseholds by District Size INCOME EXPENDITURE INCOME SHARE LESS INCOME DISTRICT SHARE SHARE EXPENDITURE SHARE SHARE LESS Colombo 4.63 4.02 0.61 HOUSEHOLD INCOME EXPENDITURE EXPENDITURE Gampaha 1.03 3.53 –2.50 CATEGORY SHARE SHARE SHARE Kalutara 2.89 5.47 –2.57 Poverty Nonpoor 9.84 7.53 2.31 status Kandy 4.86 6.70 –1.84 Poor 16.35 19.42 –3.07 Matale 9.60 7.45 2.15 TOTAL 9.96 7.80 2.16 Nuwara Eliya 8.06 7.17 0.88 Farm ≤1 acre 5.76 7.84 –2.08 Galle 1.90 6.29 –4.39 size Matara 13.69 7.96 5.73 >1 and ≤2 acres 10.84 8.37 2.46 Hambantota 21.26 9.39 11.88 >2 and ≤3 acres 15.16 8.25 6.91 Jaffna 14.36 7.29 7.07 >3 and ≤4 acres 22.25 8.33 13.92 Vavuniya 22.10 6.89 15.21 >4 and ≤5 acres 6.67 3.95 2.72 Batticoloa 47.96 10.59 37.37 >5 and ≤10 acres 26.13 6.17 19.96 Amparai 24.93 11.55 13.39 >10 acres 54.83 5.31 49.51 Trinco 22.60 9.79 12.81 TOTAL 10.00 7.78 2.22 Kurunegala 6.37 7.53 –1.16 Source: Task team’s calculations, based on HEIS 2009/10. Puttalam 8.45 7.32 1.14 Anuradapura 21.41 7.92 13.50 words, on average, paddy farming households depend primarily on Polonnaruwa 34.13 8.17 25.96 Badulla 10.24 10.39 –0.15 their own farms for the rice they consume, and rice accounts for 7.8 Monaragala 15.03 12.64 2.39 percent of their expenditure. Ratnapura 6.37 9.41 –3.03 Significant differences exist across various sectors, geographical Kegalle 2.89 7.06 –4.17 TOTAL 9.96 7.80 2.16 zones, income classes, and sizes of land holding. The small sample Source: Task team’s calculations, based on HEIS 2009/10. (35) of households residing in estates indicates that this group earns on average a small portion of income from paddy farming (only Rs 175). “Urban” farmers earn Rs 1,288 per capita, which represents a in Southern Province earn less than 2 percent of their income from small 3.26 percent of income. In contrast, rural households on aver- paddy farming; they obtain a large portion of income from other age earn 10 percent of their income from rice (Notably, expendi- activities. tures on rice are highest in the estate sector, which also consumes relatively higher levels of wheat flour). Nonpoor households (numbering 2,545 in the sample) earn more than twice as much from paddy farming as poor households North Central Province, comprising the administrative districts of (numbering 174). As a proportion of income, however, rice is more Anuradhapura and Polonnaruwa, is predominantly a paddy-grow- important for poor households (16.35 percent) than for nonpoor ing area; of the total national rice area (1.8 million acres), the two households (9.84 percent). Thus changes in income from paddy districts have 279,507 and 161,597 acres under paddy (Department farming will have greater relative impacts on poor farmers. In terms of Agriculture statistics 2011). The proportion of total income from of expenditures, nonpoor farmers spend more on rice (Rs 610), of paddy is higher in Northern, North Central, and Eastern Provinces. which 60 percent (Rs 378) is home grown. For the nonpoor, the In contrast, Western Province has the smallest share of income from proportion of total spending on rice is 7.53 percent, which is low. rice. In particular, farmers in Gampaha in Western Province and Galle Poor paddy farming households spend Rs 487 on rice, representing A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R 30 C H A P T E R 5 — I N C O M E D I S T R I B U T I O N A L I M P L I C AT I O N S O F   A G R I C U LT U R A L T R A D E P O L I C I E S 19.42 of their expenditures. In short, higher rice prices resulting from normal arbitrage in the market. Even if the tariff exemption period international trends or border protection would harm poor paddy were short, if it corresponded with the period during which farms farming households more than nonpoor households. were selling paddy, then in effect for farmers the tariff did not exist. Second, although taxes would have applied if imports would have Larger land holders earn a greater proportion of total income from been purchased, the domestic market was self-sufficient at an equi- paddy farming. Note that average total household income is high- librium price both below the border price plus tax but above the est among households with 4–5 acres and lower for households fob price. And so in that particular year, the tariff was redundant with holdings greater than 5 acres (although still greater than the and marginal changes would not have resulted in farm gate price average for farms of less than 4 acres). Households in the largest changes. Third, there might be lags in price transmission between farm-size class on average earn 56 percent of their income from border and domestic prices, so that the observed domestic price rice. The majority of paddy farmers have holdings of less than 1 acre is in transition to a higher price as arbitrage opportunities are rec- (1,525 farmers in the sample), and fully 78 percent of farms are less ognized, although slowly. Low price transmission might be due to than 2 acres. price controls, or some strong coordinated, monopsonistic behav- ior of buyers of paddy or rice. In terms of the urban consumer, the 5.2 EFFECTS OF TRADE AND PRICE POLICY ON REAL marginal price of rice could still be determined by the CIF price plus HOUSEHOLD INCOME OF PADDY FARMERS the tariff equivalent, so that some rents would be captured along In this section we present results for 2009–10 and 2010–11. Although the supply chain beyond the farm gate, perhaps by a government we expected to find significant changes in NRPs and ERPs between procurement agency. these two years due to the instability of the world price of rice, in fact the border price was not very different in the latter period: We considered three simulations to show the impact of policies on US$ 0.47/kg in 2009 and US$ 0.49/kg in 2010 (about 4 percent in- the net income of rice paddy producers expressed as value added crease), and in rupees from SLR 50.08 to SLR 53.60/kg (a 5.3 percent over tradable input costs. The three scenarios are: removal of the increase). The variable that did change significantly over these two fertilizer subsidy, but maintaining the border taxes; removal of all years was the import tax on rice expressed in TRP, which increased border taxes, but maintaining the fertilizer subsidy; and the removal from 7.2 percent in 2009 to 47 percent in 2010. In the short run at of both subsidy and border taxes. Table 5.8 shows the results for the least the observed farm gate price did not change notably, increas- two periods. For 2009/10 the simulations are based on the farm gate ing from 27.7 rupees in 2009/10 to 28.5 in 2010/11. In fact, applying equivalents without the interventions and the farm gate prices as the implicit margins observed in 2009/10 to the CIF price (which observed. For 2010/11, however, we take the long-term estimate of shows the price without taxes) in 2010/11, we find that the farm what the farm gate price would be if, in the absence of self-suffi- gate equivalent price should have been 28.4 rupees, very close to ciency, the opportunity cost of buying a ton of domestic rice were what was observed. Applying the border taxes of 47 percent and in fact the CIF price plus the official tax of 47 percent (without tax the previous year’s margins (in absolute terms) would have yielded exemptions). a farm gate price equivalent of 46 rupees, or 62 percent greater than For 2009/10, removing the fertilizer subsidy of 90 percent (but assum- the observed farm gate price. (The wholesale price equivalent with ing the same level of input use) would increase the price per kilo of the border taxes and implicit margins would have been 50 percent fertilizer from SLR 7 to SLR 72 (928 percent), which in turn would lead higher than the wholesale price observed.) to a decrease of 25 percent in the net income of producers. Removing There are three possible explanations (not mutually exclusive) the border taxes on all goods, output and tradable inputs, but keep- for these discrepancies. First, although official border taxes were ing the fertilizer subsidy, would cause a decline in the farmer’s output increased, there might have been exemptions during the import price of 6.8 percent and a decrease in per kilo costs of tradable inputs season, and the observed farm gate price is simply reflecting the of 7 percent, for a final decrease in net income of only 4 percent (SLR W H AT I S T H E C O S T O F A B O W L O F R I C E ? T H E I M PA C T O F S R I L A N K A' S C U R R E N T T R A D E A N D P R I C E P O L I C I E S C H A P T E R 5 — I N C O M E D I S T R I B U T I O N A L I M P L I C AT I O N S O F   A G R I C U LT U R A L T R A D E P O L I C I E S 31 21.6 to 20.8 per kilo). The case of removing border measures alone losses on the income side. Because the share of rice in poor house- yields an ERP of 4 percent due to the relatively low tariff in that year. hold expenditures is greater, the potential gains from tariff removal are greater on their consumption side. Finally, removing both border taxes and the fertilizer subsidy as exist- ed in 2009/10 leads to the 6.8 percent decrease in output price at the With respect to farm size, households that suffer the largest abso- farmgate and an increase of 40 percent in tradable input costs per lute and relative income loss have the largest land holdings. The kilo produced, because the fertilizer cost increased so much. Such net real income loss on average for farms of greater than 10 acres a large increase in input costs with the decline in the paddy price reaches 26 percent of paddy revenues from paddy production and reduces net income by 27 percent and gives the ERP of 36.5 percent. 15 percent of total household income. Small farms of up to 3 acres For the period of 2009/10, the fertilizer subsidy is driving the result. have losses relative to total income of less than 5 percent of total household income. The 2010/11 NRP estimate would lead to con- For 2010/11, removing the fertilizer subsidy, but maintaining tariffs, siderably larger household savings on food, particularly for the poor leads to an increase in the fertilizer price of 525 percent and a re- and for smaller-scale farmers. Nevertheless, across all farm sizes the duction in net income of 10 percent, for an ERP of 76.48 percent. potential household income loss from eliminating the high tariff Removing the tariffs, maintaining the fertilizer subsidy, leads to a fall rate for the 2010/11 period would be higher in absolute terms than of the output price of 38 percent and an increase in tradable input for 2009/10. costs per ton of 12 percent, for a notably large decline of 43 percent in farmer’s net income. Finally, a removal of tariffs and the subsidy These simulations are based on recently observed consumption and would lead to a fall in net income of 53 percent; therefore the exist- production patterns, so they do not consider substitution effects in ing policies prior to these hypothetical changes have an ERP of 112 consumption (such as moving from wheat to rice) or production percent. Note that these estimates are for the full-transmission case (such as switching to relatively more profitable crops). Nevertheless, for border prices. results of the simulation have implications for policy. On average, for the poor and for smaller farm households, the removal of protection The net effect on particular farm households will be determined by would have a small effect on real household income. The main ad- the reduction in income from paddy sales and the savings derived verse effect is felt by the medium- and larger-scale paddy farms. In from the lower price of purchased rice. The simulations show that other words, those who gain from the protectionist policies belong paddy farm households would experience a savings on rice ex- to relatively wealthier household categories. penditure that is significantly less that the reduction in per capita net income due to the fall in paddy prices. In 2009/10, the savings The simulation analysis described here is only for the paddy subsec- on purchased rice consumed by the average household would be tor, where benefits to producers are to be expected. It is important Rs 29, but the loss of farmer paddy income would average (over to remember, however, that a larger population is potentially af- both poor and nonpoor) Rs 307, yielding an average household fected by the current trade policy—namely, all consumers who are per capita net reduction in real income of Rs 279. This hypotheti- not paddy farmers, who would definitely benefit from a removal of cal income loss would represent 24 percent of paddy income, but protection to rice (a lower NRP). Lower-income, urban families in less than 3 percent of total household income. When the simula- particular would tend to spend relatively more of their budgets on tion is repeated with the estimated 2010/11 NRP, the savings effect rice. For example, if poor, non-farming families have a consumption is larger proportionally, but the loss of potential paddy income is expenditure share of 20 percent for rice, the impact of removing much larger in an absolute sense than in the 2009/10 year. trade taxes on rice in 2009/10 would lead to a rice-expenditure sav- ings of 5 percent, or about 1 percent of household expenditures. Note that poor farmers in absolute terms suffer smaller losses than But given the higher tariff of 2010/11, removing border taxes would nonpoor farmers when border protection is removed. Relative to lead to a 40 percent savings on rice expenditures, or about an 8 their total household income, however, the poor would suffer larger percent savings in total household expenditure. A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R 32 C H A P T E R 5 — I N C O M E D I S T R I B U T I O N A L I M P L I C AT I O N S O F   A G R I C U LT U R A L T R A D E P O L I C I E S It is a matter for Sri Lankan decision makers to determine the weight TABLE 5.8: Effect of Policy Changes on Paddy Farmer placed on various possible policy objectives (such as self-sufficien- Net Income over Tradable Input Costs, 2010/11 cy); this simulation has focused attention on the impact across LOSS OF INCOME FROM PADDY poverty and farm-size classes. Setting aside other possible objec- PRODUCTION tives of the present border protection policy and fertilizer subsidies, Simulation scenario 2009/10 2010/11 however, the current trade regime is regressive from the perspective Remove fertilizer subsidy, keep border taxes 25% 10% of poverty and small farms. Remove border protection, keep fertilizer 4% 43% subsidy Remove both 27% 53% TABLE 5.9: Simulated Impact of Price and Trade Policies in 2009/10 and 2010/11 on Rice Growers’ Net Household Income and Rice Expenditures, by Poverty Status and Farm Size OBSERVED FROM 2009 HEIS SIMULATED 2009/10 SIMULATED 2010/11 INCOME FROM EXPENDITURES INCOME EXP SHARE LOSS IN EXPENDITURE LOSS IN EXPENDITURE GROUP PADDY ON RICE SHARE % % INCOME SAVED GAP INCOME SAVED GAP Nonpoor 1,198 610 9.8 7.5 320 29 291 632 194 438 Poor 495 487 16.4 19.4 132 23 109 261 155 106 TOTAL 1,150 602 10 7.8 307 29 279 607 192 415 <=1 507 583 5.8 7.8 136 28 108 268 186 82 >1 & <=2 1,317 622 10.8 8.4 352 30 323 695 198 497 Size class in acres >2 & <=3 2,256 648 15.2 8.3 603 31 572 1,191 206 984 >3 & <=4 2,754 654 22.3 8.3 736 31 705 1,453 208 1,245 >4 & <=5 3,413 528 6.7 3.9 913 25 887 1,801 168 1,633 >5 & <=10 5,419 663 26.1 6.2 1,449 32 1,417 2,859 211 2,648 >10 11,009 562 54.8 5.3 2,943 27 2,916 5,809 179 5,630 TOTAL 1,157 601 10 7.8 309 29 281 610 191 419 W H AT I S T H E C O S T O F A B O W L O F R I C E ? T H E I M PA C T O F S R I L A N K A' S C U R R E N T T R A D E A N D P R I C E P O L I C I E S C H A P T E R 6 — W H AT D O E S I T A L L M E A N ? C O N C LU D I N G C O M M E N T S A N D R E C O M M E N D AT I O N S 33 Chapter 6 WHAT DOES IT ALL MEAN? CONCLUDING COMMENTS AND RECOMMENDATIONS This analysis has examined the period following the 30-year war; Part of the revenue earned through taxation is used in subsidizing during this period, the government has pursued multiple objec- fertilizer. To better understand the overall incentive framework, the tives: generating fiscal revenue from trade taxes, obtaining a higher team, in addition to looking at output prices, assessed the current degree of self-sufficiency in certain food products, reducing poverty policy on the fertilizer subsidy program, which represents a major in rural areas, and gaining political support in rural areas—primarily expenditure in the government´s agricultural budget (about 50 through the higher farm income associated with high protection of percent in 2009 and 38.7 percent in 2011). The subsidy raises some import-competing activities. In this endeavor, the government faces concern, to the extent that the transfer efficiency is low and that the difficult task of balancing the interests of farmers with those of the program diverts public funding from truly public goods (such consumers, given that higher protection to rice causes higher food as agricultural research or rural roads) to “private” goods (fertilizers). prices for consumers. An analysis of the transfer efficiency of the fertilizer subsidy in paddy farming in 2009–10 indicated that the gain in farmers’ net income is Border taxes on imports and exports of agricultural products and small relative to the fiscal cost of the subsidy: On average, the gov- tradable agricultural inputs have been the government’s principal ernment spends between Rs 1.4 and 2.4 per acre to increase farm policy instruments to generate fiscal revenue, support farmers income by Rs 1 per acre. engaged in import-competing activities, and taxing producers of export-oriented products. In contrast to agricultural products, non- The ERP calculations show the extent of incentives provided to the agricultural tradables have seen a decline in border protection since farming community. Of 10 sectors analyzed here, the contributions the trade reforms of the early 1990s, which removed a significant to GDP from paddy, tea, and coconut production are substantial. indirect (implicit) tax on agricultural production. According to the Paddy is highly protected (mainly through the fertilizer subsidy), tea estimates in this study for 2009 and 2011, the combined impact of is marginally protected, and coconut production is taxed (due to the import duties plus other taxes at the border (the TPR in the tables export cess on desiccated coconut and taxing of fertilizer imports). shown previously) yielded NPRs for importables for most agricul- Of the remaining commodities studied, maize, potatoes, and milk tural products on the order of 30–50 percent (ad valorem equiva- are highly protected; rubber is neutral. lent). For rice, the TPR changed from 7.2 percent during 2009 to 34 percent in 2011. 6.1 IN THE SHORT TERM, WHO WINS AND WHO LOSES UNDER CURRENT POLICIES? During 2010, through the application of cess taxes, exports of tea and rubber were taxed at low rates (2.5 percent or less), unlike cin- In the short term, who wins and who loses under the current trade namon (taxed at 8.3 percent), coconuts (11.8 percent), and cloves and price policy for agriculture? A simple partial equilibrium frame- (10 percent). Furthermore, it is claimed that revenues from export work that divides agricultural producers, buyers of farm products, cesses are not used fully for investments in the development of households as consumers, and taxpayers reveals that producers of these export activities. import-competing farm products gain at the expense of consumers, A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R 34 C H A P T E R 6 — W H AT D O E S I T A L L M E A N ? C O N C LU D I N G C O M M E N T S A N D R E C O M M E N D AT I O N S mainly due to border measures. The ERP measures the effect of As Vishwanath and Serajuddin (2010) note, a large proportion of trade policy on farm value added (returns to land, labor, owned rural households in South Asia are net buyers of food, so the impact capital, and profits) and indicates the overall income effect per unit of higher food prices on poverty could be quite large. In the future, of output of the trade regime on the farm production sector. an in-depth analysis of the impacts of Sri Lankan agricultural trade and price policy on real household income could take advantage For example, the ERP—the increase in farm-related income per ton of well-developed methods for analyzing the net impacts of food arising from policy—for paddy rice, an import-competing crop, was price changes (see, for example, Deaton 1997), although household 36 percent in 2009 and 112 for 2010/11. The ERP for potatoes was survey data would be required. 149 percent in 2009 and 41.52 percent in 2010/11; for milk it was 5.26 percent in 2009 and 11.25 percent in 2010/11. Among export- Based on the ERP estimates percentage change in value added of able crops, the ERP for coconuts was negative (–30 percent), imply- selected products during 2009/2010 and 2010/2011 were calcu- ing that incomes per ton were lower than they otherwise would lated. As expected, import competing activities show positive and be without the policy. For tea and rubber (also exportables), the high values and with significant differences across commodities ERPs are near zero, in part because of the fertilizer subsidy. In other within a given year and differences between the two years. In the words, income from paddy farming was 36 percent higher than case of rice 2010/2011 is explained by the higher border taxes not what it would be without government intervention, and income by the changes in border prices. This is different in the case of milk from milk production was 5.26 percent higher. The impact of policy products where the border prices went up considerably from 2009 on incomes from tea and rubber production was insignificant. The to 2011. These measures present first order effects where the level impact on coconut farming, however, was a 30 percent reduction in of output remains unchanged and it captures essentially an income income. The overall, aggregate effect on agricultural incomes in ab- transfer effect. The income of maize producers during 2009/2010 solute terms would depend on the levels of production of all crops was practically 70 percent above what they would have been in and any crop mix that might be fostered by the incentive structure. the absence of price and trade interventions (includes the fertilizer Clearly, incentives under the current policy regime are structured to subsidy). In contrast, income of coconut growers was 88 percent expand production of import-competing crops and discourage the lower than what could have been in the absence of interventions. production of exportables. No attempt was made in this study to assess the impacts of removal of price and trade interventions, but the results suggest a very clear Turning to buyers of commodities (agro-processors and house- direction of changes in outputs. The selected import competing holds), the NRP is the relevant indicator; it measures the percentage activities show a reduction in output while export competing activi- difference between observed prices and the prices that would pre- ties would experience an increase in production. The overall output vail in the absence of policy intervention. Current policies represent effect of removal of price and trade interventions on medium and an implicit tax of about 300 percent on maize and chili buyers; 96 longer term would probably reflect a significant change in output percent on milk buyers; 54 percent on potato buyers; and 12 per- mix of Sri Lankan agriculture toward a more diversified output and cent on rice buyers. Consumers of tea, rubber, and coconuts are export composition. subsidized, although only to a small degree. These elevated prices are likely to have a significant negative impact on the real incomes Table 6.1 (b) presents a simulation of the impact of removal of trade of lower- and middle-income households, especially in urban areas. and price interventions on household expenditures based on the Households that spend about 50 percent of their incomes on food NRP estimates at the border (not at farm level). The computations and spend 40 percent or more of that income on basic foods ef- were based on the expenditure shares of three products in the fectively suffer a real income loss of 20 percent. The final impact on consumer price index. The positive figures show the percentage households depends on their net food position: Farming households increase in household expenditures due to the price and trade in- gain somewhat because of their sales of protected farm products. terventions as captured by the NPR. In the case of coconut there W H AT I S T H E C O S T O F A B O W L O F R I C E ? T H E I M PA C T O F S R I L A N K A' S C U R R E N T T R A D E A N D P R I C E P O L I C I E S C H A P T E R 6 — W H AT D O E S I T A L L M E A N ? C O N C LU D I N G C O M M E N T S A N D R E C O M M E N D AT I O N S 35 TABLE 6.1: Simulation on the Impact of Price and Finally, table 6.1 (c) shows very simple computation impacts of the Trade Policies on (a) Farm Income, (b) Real price and trade interventions on the government side both on Household Income, and (c) Government expenditures and revenues. The revenue captures the border tax Accounts revenues on both imports and exports based on import duty col- (a) Impact on Farm Income (Change in Value Added) lection rates and export cess. On the expenditure side, it represents YEAR 2009/10 2010/11 the fiscal cost of fertilizer subsidy. PERCENTAGE CHANGE PERCENTAGE CHANGE ITEM IN VALUE ADDED IN VALUE ADDED Rice 26.73 52.77 6.2 IMPLICATIONS FOR WORLD BANK ENGAGEMENT Potato 59.85 29.34 IN SRI LANKA Maize 69.93 181.66 The high protection of importables in Sri Lankan agriculture creates Green gram 15.76 4.04 Milk 50.29 90.08 strong8 disincentives for crop and export diversification. Given the Tea –4.14 –4.43 objectives and statements in the Mahinda Chintana about promot- Rubber –1.09 –1.87 ing export agriculture, it appears that the extremely complex system Coconut –87.44 –45.80 of border measures and trade policy currently in place is slowing (b) Impact on Real Household Income (Change in rather than accelerating export growth, with no recent quantitative Expenditures for Food and Non-Food Items) evaluation available so far. YEAR 2009 PERCENTAGE 2011 PERCENTAGE From an investment perspective, in the absence of an adjustment of CHANGE IN TOTAL CHANGE IN TOTAL the incentive framework, it would be sensible to avoid investing in ITEM EXPENDITURE EXPENDITURE Rice 0.53 2.48 areas “contaminated” by significant distortions, where social returns Milk and milk products 1.56 1.50 on lending would be low (such as mechanization, fertilizer, and Coconut –0.71 –0.71 similar areas). Instead, lending should focus on promoting public (c) Impact on Government Revenues (2011) goods with positive externalities—such as agricultural research, ITEM REVENUES IN SLR MILLION rural roads, sanitary and phytosanitary services, and similar public Food imports 449.39 goods—that are less affected by a distorted incentive framework. Agricultural exports Tea 31.61 It is important to link the impact of a trade and price policy reform Rubber 2.89 (in the medium term) with complementary reforms (over the longer Coconut 79.80 term) in markets for production factors, such as land and irrigation. Subtotal 563.69 As agriculture is a highly tradable activity, trade reforms, if imple- Fertilizer subsidies (Expenditures) –248.35 mented, will affect (and probably increase) the average returns to NET EFFECT 315.34 land and water in agriculture. Together, such reforms would also probably cause land and water to be reallocated to higher-value crops. To capture the benefits of reforms more fully, land and water markets should allow enough flexibility to facilitate changes in the is a saving due to the export tax. Same calculations can be done crop mix and tenure arrangements. for the first quintile in the income distribution of households. It would clearly show a higher magnitude of the income effects at the household level considering the higher share of food in the total 8 Even if there were no export taxes nor cesses the high levels of protec- household expenditure at lower income levels. The figures in 6.1 (b) tion to import competing products represent an implicit high export tax equivalent due to the impact on relative incentives to remain producing represent the average share for the whole income distribution. importables instead of shifting some resources (land, water, labor, work- ing capital) to the production of agriculture exportables A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R 36 C H A P T E R 6 — W H AT D O E S I T A L L M E A N ? C O N C LU D I N G C O M M E N T S A N D R E C O M M E N D AT I O N S Experience from other countries shows that a trade regime as dis- An important result from this study is that trade policy in relation cretionary and complex as Sri Lanka’s can generate considerable un- to rice does not necessarily benefit the poor and small-scale paddy certainty about expected returns. This uncertainty can significantly farmer. The main benefits are likely captured by larger farms. If one deter private investment, both in primary agriculture as well as in considers the effect on those who do not grow rice, especially the agro-processing, and potentially reduce diversification in agricul- urban poor, the protection of paddy farming is regressive. tural production and exports. W H AT I S T H E C O S T O F A B O W L O F R I C E ? T H E I M PA C T O F S R I L A N K A' S C U R R E N T T R A D E A N D P R I C E P O L I C I E S REFERENCES 37 REFERENCES Abeyratne, F., E. Neville, W. G. Somaratne, and P. Wickramaarachchi. 1990. __________. 2001. “Comparative Advantage of Rice Production in Sri Lanka “Efficiency of Rice Production and Issues Relating to Protection.” with Special Reference to Irrigation Costs.” Paper presented at the Sri Lanka Journal of Agricultural Economics 1(1): 16–25. Workshop on Medium- and Long-Term Prospects of Rice Supply and Demand in the 21st Century, International Rice Research Institute Ahmed, S., and H. G. P. Jansen. 2010. Managing Food Price Inflation in South (IRRI), Los Baños, December 3–5, 2001. Asia. Dhaka: University Press Limited. Panagariya, A. 2011. “Avoiding Lopsided Spatial Transformation.” Chapter 6 in Anderson, K., and W. Martin, eds. 2009. Distortions to Agricultural Incentives in Reshaping Tomorrow: Is South Asia Ready for the Big Leap?, edited by E. Asia. Washington, DC: World Bank. Ghani. New Delhi: Oxford University Press India and the World Bank. Bandara, J., and S. Jayasuriya. 2009. “Sri Lanka.” Chapter 12 in Distortions to Pursell, G. 2011. “Sri Lanka’s Trade Policies: Back to Protectionism.” Economic Agricultural Incentives in Asia, edited by Kym Anderson and Will Martin, and Political Weekly (June 18), V. XLVI (25): 31–4. 409–40. Washington, DC: World Bank. Rafeek, M. I. M., and P. A. Samarathunga. 2000. “Trade Liberalization and Its Bhalla, Surjit S. 1991. “The Political Economy of Agricultural Pricing Policies Impact on the Rice Sector of Sri Lanka.” Sri Lankan Journal of Agricultural in Sri Lanka.” In The Political Economy of Agricultural Pricing Policy, edited Economics 3: 143–54. by Anne Krueger, Maurice Schiff, and Alberto Valdes. Vol. 2. Baltimore: Johns Hopkins. Ranaweera, N. F. C., P. A. Samarathunga, and A. A. B. Hafi. 1990. Economics of Paddy Production: Past, Present and Future. In Proceedings of the Central Bank of Sri Lanka. 2008. Annual Report 2007. Rice Congress, Department of Agriculture, edited by S. L. Amarasiri, Central Bank of Sri Lanka. 2009. Annual Report 2008. S. Nagarajah and B. M. K. Perera, 147–68. Peradeniya, Sri Lanka: Department of Agriculture. Central Bank of Sri Lanka. 2010. Annual Report 2009. Shilpi, F. 1995. “Policy Incentive, Diversification and Comparative Advantage Central Bank of Sri Lanka. 2011. Annual Report 2010. of Non-Plantation Crops in Sri Lanka.” Working Paper 2. Colombo: Deaton, A. 1997. The Analysis of Household Surveys: A Microeconometric World Bank. Approach to Development. Baltimore: World Bank and Johns Hopkins. Thibbotuwawa, R. M. M. I., and J. Weerahewa. 2004. “Policy Options for Department of Census and Statistics. 2002. Agriculture and Environment Sustainable Paddy Farming: Scope for Land Consolidation in Sri Lanka.” Statistics Division. Abstract published in the Proceedings of the Peradeniya University Research Sessions (PURSE). 7 pp. Ekanayake, H. K. J. 2006. “Impact of Fertilizer Subsidy in Paddy cultivation in Sri Lanka.” Staff Studies (Central Bank of Sri Lanka) 36 (1 and 2): 73–92. Vishwanath, T., and U. Serajuddin. 2010. ”Welfare Impacts of Rising Food Prices in South Asian Countries.” Chapter 2 in Managing Food Price Gunawardena, D. 2000. Consumption Poverty in Sri Lanka, 1985–1996: A pro- Inflation in South Asia, edited by A. Ahmed and H. G. P. Jansen, 45–84. file of poverty based on household survey data. Washington, D.C.: World Dhaka: University Press Limited. Bank. Weerahewa, J. 2004. “Impacts of Trade Liberalization and Market Reforms IPS (Institute of Policy Studies of Sri Lanka). 2010. Sri Lanka State of the on the Paddy/Rice Sector in Sri Lanka.” MTID Discussion Paper No. 70. Economy 2010: Growth and Stability in Post-conflict Economic Recovery. Washington, DC: International Food Policy Research Institute (IFPRI). Colombo. http://www.ifpri.org/publication/impacts-trade-liberalization-and- Kelegama, S. 2010. “Managing Food Price Inflation in Sri Lanka.” Chapter 8 in market-reforms-rice-sector-sri-lanka, accessed April 2012. Managing Food Price Inflation in South Asia, edited by A. Ahmed and H. Weerahewa, Jeevika, H. M. Gunatilake and Hansamali Pitigala (2002). Future G. P. Jansen. Dhaka: University Press Limited. Pp. 237–56. of Paddy Farming in Sri Lanka: Comparative Advantage, Scale and Rural Kikuchi, M., R. Barker, M. Samad, and P. Weligamage. 2000. “Comparative Poverty. Sri Lanka Economic Journal 3 (2): 104–144. Advantage of Rice Production in an Ex-rice Importing Country: The Case of Sri Lanka.” Paper presented at the 3rd Conference of the Asian Society of Agricultural Economists, Jaipur, October 18–20. A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D E PA R T M E N T D I S C U S S I O N PA P E R 38 REFERENCES Weerahewa, J. S., S. Kodithuwakku, and A. Ariyawardana. 2010. Case Study Wijayaratna, C. M., C. R. Panabokke, P. B. Aluwihare, S. H. Charles, and R. #7-11, “The Fertilizer Subsidy Program in Sri Lanka.” In Food Policy for Sakthivadivel. 1996. “Potential for Diversified Cropping in the Rice Developing Countries: The Role of Government in Global, National, and Lands of Sri Lanka.” IIMI Country Paper No. 14. Colombo: International Local Food Systems, edited by Per Pinstrup-Andersen and Fuzhi Cheng. Irrigation Management Institute (IIMI). http://cip.cornell.edu/dns.gfs/1289505412, accessed April 2012. Wijetunga, W. M. L. K., S. Thiruchelvam, and N. Balamurali. 2008. “Impact of Weerasooriya, S., and L. H. P. Gunaratne. 2009. “Impact of Fertilizer Subsidy ‘KETHATA ARUNA’ Fertilizer Subsidy Scheme on Paddy Production in on Productivity, Supply and Cultivated Extent of Paddy.” Abstracts of Minipe Scheme.” Abstracts of the Second Annual Research Forum of the Third Annual Research Forum of Sri Lanka Agricultural Economics Sri Lanka Agricultural Economics Association, October 3, University of Association, October 2, Hector Kobbakaduwa Training Institute, Peredeniya. http://www.slageconr.net/saea/arf/abstractsarf02.pdf, ac- Colombo. http://www.slageconr.net/saea/arf/abstractsarf03.pdf, ac- cessed April 2012. cessed April 2012. World Bank. 1996. “Sri Lanka, Nonplantation Crop Sector Policy Alternatives.” Weligamage, P., C. R. Shumway, K. A. Blatner, W. R. Butcher, and M. Giordano. Report No, 14564-CE. March 20. Washington, DC. 2009. “Production Function for Rice in Kirindi Oya Major Irrigation __________. 2002. “Sri Lanka: Poverty Assessment.” Report 22535-CE. System in Sri Lanka.” Presentation at the Third Annual Research Forum Washington, DC: Poverty Reduction and Economic Management of Sri Lanka Agricultural Economics Association, October 2, Hector Sector Unit, South Asia Region. Kobbakaduwa Training Institute, Colombo. WTO (World Trade Organization). 2010. “Trade Policy Review: Sri Lanka.” Wickramasinghe, W., G. Samarasingha, and S. Epasinghe. 2009. “Fertilizer Report by the Secretariat, WT/TPR/S/237, September 29. Geneva. policy on paddy farming. Evaluation of subsidy program 2005.” http://www.wto.org/english/tratop_e/tpr_e/tp337_e.htm, accessed Colombo, Sri Lanka: Hector Kobbekaduwa Agrarian Research and April 2012. Training Institute. W H AT I S T H E C O S T O F A B O W L O F R I C E ? T H E I M PA C T O F S R I L A N K A' S C U R R E N T T R A D E A N D P R I C E P O L I C I E S A P P E N D I X TA B L E S 39 APPENDIX TABLES APPENDIX TABLE 1: Unweighted Average Import and Other Taxes (percent) by Commodity Group 2009 2011 TOTAL TAX: PROTECTIVE TOTAL TAX: PROTECTIVE COMMODITY TOTAL TAX SIMPLE SUM TAXES TPR TOTAL TAX SIMPLE SUM TAXES TPR Import and other taxes on export-oriented products Tea 48.25 44.65 37.46 45.56 42.83 38.30 Coconuts 59.32 52.07 38.76 54.50 50.24 40.00 Rubber 58.40 50.59 36.88 55.15 49.10 36.42 Cinnamon 55.40 50.39 41.71 54.57 49.17 39.83 Cloves 53.43 50.68 48.20 49.12 47.00 46.00 Average 54.96 49.68 40.60 51.78 47.67 40.11 Import and other taxes on import-competing products Chilies 78.74 67.68 42.97 57.01 60.55 51.50 Maize 61.46 53.19 40.82 58.04 50.33 39.42 Rice 10.72 10.00 7.20 36.71 35.64 33.84 Wheat 45.91 40.72 28.05 37.41 33.2 24.31 Vegetables 77.71 66.57 53.17 70.78 62.16 52.30 Milk 66.37 57.69 47.09 62.13 54.46 45.33 Fruits 95.88 81.24 68.61 90.83 77.33 66.17 Potatoes 71.66 63.09 52.60 57.21 79.35 44.44 Onions 50.60 43.17 29.55 52.37 64.12 53.86 Average 60.63 52.62 39.69 58.05 57.46 44.34 Agricultural tradable inputs Machinery 23.19 21.01 11.05 16.43 15.18 8.15 Seed 23.69 21.26 7.56 18.34 16.60 5.00 Fertilizer 11.45 10.92 7.88 5.10 5.09 5.00 Fuel 21.67 19.50 8.88 25.13 22.00 15.66 Chemicals 29.40 26.23 11.13 29.65 26.50 12.50 Average 21.88 19.78 9.30 18.93 17.07 9.26 A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R 40 A P P E N D I X TA B L E S APPENDIX TABLE 2: Estimated Fertilizer Consumption by Crop Sector and Imports of Fertilizer, 2009 (mt) TOTAL CONSUMPTION FERTILIZER BY ALL CROP TYPE PADDY TEA RUBBER COCONUTS OFC EAC SECTORS IMPORTS Urea 250,741 112,363 2,063 3,478 2,757 689 377,406 401,318 Ammonium sulfate 2,754 10,345 320 2,345 8,190 1,040 27,965 24,573 Triple super 82,609 1,161 13 163 2,565 70 87,281 34,107 phosphate Muriate of potash 72,330 34,366 1,438 5,176 6,738 855 122,755 81,548 Total: All fertilizers 422,968 187,787 8,277 16,857 28,150 4,228 686,671 546,665 Source: AgStat, Department of Agriculture, 2010. APPENDIX TABLE 3: Cost Share of Tradable Inputs, 2009 COST SHARE OF TRADABLE PRIMARY COMMODITY OUTPUT TRADED TRADABLE INPUTS INPUTS* (%) Paddy Rice Fertilizers 2.08 Pesticides 1.65 Weedicides 4.31 Machinery 14.00 Potatoes Potatoes Fertilizers 14.39 Pesticides 2.97 Seeds 26.24 Machinery 3.26 Maize Maize Fertilizers 15.29 Weedicides 2.85 Seeds 5.41 Machinery 7.74 Chilies Dried chilies Fertilizers 12.76 Pesticides 2.39 Seeds 0.71 Red onions Red onions Fertilizers 10.35 Pesticides 0.68 Weedicides 0.44 Seeds 24.54 Machinery 4.43 Green gram Green gram Pesticides 3.35 Weedicide 3.32 Seeds 2.97 Machinery 6.92 (continued) W H AT I S T H E C O S T O F A B O W L O F R I C E ? T H E I M PA C T O F S R I L A N K A' S C U R R E N T T R A D E A N D P R I C E P O L I C I E S A P P E N D I X TA B L E S 41 APPENDIX TABLE 3: Cost Share of Tradable Inputs, 2009 (continued) COST SHARE OF TRADABLE PRIMARY COMMODITY OUTPUT TRADED TRADABLE INPUTS INPUTS* (%) Milk Powdered milk Feed 23.06 Made tea Made tea Urea 2.10 Muriate of potash 0.83 ZnO4 0.10 Glyphosate 0.36 Paraquat 0.18 Copper fungicide 0.45 Systemic fungicide 0.70 Diesel 5.33 Rubber (sheet) RSS Fertilizers 0.51 Fresh coconuts Desiccated coconut Fertilizers 34.64 * Share of each input as a proportion of farm-gate price of the produce. APPENDIX TABLE 4: Import Taxes on Agricultural Inputs TOTAL TOTAL TAX: TOTAL TAX: END-OF-YEAR PROTECTION HS-CODE DESCRIPTION FORMULA SIMPLE SUM TAX RATE (TPR) 8419.89 Equipment for pasteurization and sterilization of milk 8.45 9.50 6.15 5.00 8419.90 Parts 25.60 24.38 23.24 7.88 8422.30 Machinery for filling, closing, sealing, or labeling bottles, cans, boxes, bags, or other 25.60 24.09 10.42 5.72 containers; machinery for capsuling bottles, jars, tubes, and similar containers; machinery for aerating beverages 8433.51 Combine harvester-threshers 11.45 12.38 9.10 7.88 8434.10 Milking machines 8.45 9.50 6.15 5.00 8436.29 Other 8.45 9.50 6.15 5.00 8437.80 Other machinery 22.25 21.50 13.05 5.00 8438.20 Machinery for the manufacture of confectionery, cocoa, or chocolate 25.60 24.38 23.24 7.88 8438.80 Other machinery 42.68 37.67 16.17 12.02 2711.19 Other 30.29 28.40 20.08 8.45 1209.91 Vegetable seeds 8.45 9.50 6.15 5.00 3102.10 Urea, whether or not in aqueous solution 11.45 12.38 9.10 7.88 3103.10 Superphosphates 11.45 12.38 9.10 7.88 3104.20 Potassium chloride 11.45 12.38 9.10 7.88 3105.20 Mineral or chemical fertilizers containing the three fertilizing elements nitrogen, 11.45 12.38 9.10 7.88 phosphorus, and potassium 3808.91 Insecticides 36.99 34.15 34.43 17.65 3808.92 Fungicides 25.60 24.38 23.24 7.88 3808.93 Herbicides, anti-sprouting products, and plant growth regulators 25.60 24.38 23.24 7.88 A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R 42 A P P E N D I X TA B L E S APPENDIX TABLE 5: Items Subject to Specific Commodity Levies (SCLs) and Their Tariff Equivalents, 2009 DESCRIPTION (3) SCL (%) SCL (15) 0713.31.10 Green gram (moong) 17 Rs 15/kg up to 30/9/2009 0713.40.11 Massor dhal (red lentils), whole 13 Rs 15/kg up to 30/9/2009 0713.40.21 Yellow lentils, whole 13 Rs 15/kg up to 30/9/2009 1701.99.10 White crystalline cane sugar 30 Rs 14/kg up to 30/9/2009 1701.99.20 White crystalline beet sugar 30 Rs 14/kg up to 30/9/2009 0713.40.12 Massor dhal (red lentils), split 17 Rs 20/kg up to 30/9/2009 0713.40.22 Yellow lentils, split 17 Rs 20/kg up to 30/9/2009 0701.90 Potatoes, fresh or chilled 94 Rs 25/kg up to 30/9/2009 0703.10.20 B' onions 69 Rs 25/kg up to 30/9/2009 0305.59.20 Sprats, dried, not salted 19 Rs 30/kg up to 30/9/2009 0713.20 Chickpeas (garbanzos) 30 Rs 30/kg up to 30/9/2009 0713.10 Peas (Pisum sativum) 63 Rs 35/kg up to 30/9/2009 0904.20.11 Fruits of the genus Capsicum or of the genus Pimenta, dried or crushed or ground : Neither crushed 29 Rs 40/kg up to 30/9/2009 nor ground 0904.20.12 Fruits of the genus Capsicum or of the genus Pimenta, dried or crushed or ground : Crushed or ground 37 Rs 50/kg up to 30/9/2009 1604.11 Fish, whole or in pieces, but not minced : Salmon 25 Rs 85/kg up to 30/9/2009 1604.12 Fish, whole or in pieces, but not minced : Herrings 34 Rs 85/kg up to 30/9/2009 1604.13 Fish, whole or in pieces, but not minced : Sardines, sardinella, and brisling or sprats 30 Rs 85/kg up to 30/9/2009 1604.14 Fish, whole or in pieces, but not minced : Tunas, skipjack, and bonito (Sarda spp.) 31 Rs 85/kg up to 30/9/2009 1604.15 Fish, whole or in pieces, but not minced : Mackerel 33 Rs 85/kg up to 30/9/2009 1604.16 Fish, whole or in pieces, but not minced : Anchovies 13 Rs 85/kg up to 30/9/2009 1604.19 Fish, whole or in pieces, but not minced : Other 14 Rs 85/kg up to 30/9/2009 1604.20 Other prepared or preserved fish 12 Rs 85/kg up to 30/9/2009 0402.10 Milk and cream, concentrated or containing added sugar or other sweetening matter. In powder, granules, or 6 Rs 15/kg up to 21/01/2009 other solid forms, of a fat content, by weight, not exceeding 1.5% 0402.21 Milk and cream, concentrated or containing added sugar or other sweetening matter : Not containing added 5 Rs 15/kg up to 21/01/2009 sugar or other sweetening matter 0402.29 Milk and cream, concentrated or containing added sugar or other sweetening matter. Other 6 Rs 15/kg up to 21/01/2009 1006.10 Rice in the husk (paddy or rough) 2 Rs 1/kg December 2009 1006.20 Husked (brown) rice 1 Rs 1/kg December 2009 1006.30 Semimilled or wholly milled rice, whether or not polished or glazed 2 Rs 1/kg December 2009 1006.40 Broken rice 2 Rs 1/kg December 2009 0703.10.10 Red onions 28 Rs 10/kg November 2009 0703.20 Garlic 56 Rs 30/kg November 2009 W H AT I S T H E C O S T O F A B O W L O F R I C E ? T H E I M PA C T O F S R I L A N K A' S C U R R E N T T R A D E A N D P R I C E P O L I C I E S A P P E N D I X TA B L E S 43 APPENDIX TABLE 6A: Trade Policy Measures on Rice, Potato, Onion, and Chilies, 2007–11 PERIOD MONTH RICE POTATO ONION CHILIES Pre-crisis January 31, 2007 Full duty waiver applicable on Full duty waiver applicable on importation of B’ onions and red importation of chilies was extended onions was extended until February until February 15, 2007. 15, 2007. March 10, 2007 A duty waiver of Rs 10/kg on B’ Full duty waiver applicable onions was granted until April on importation of chilies was 30, 2007. extended. March 29, 2007 A duty waiver of Rs 5/kg was granted on importation of potatoes until April 30, 2007. June 1, 2007 Duty waivers of Rs 5/kg on potatoes Duty waiver of Rs 10/kg on B’ Full duty waiver applicable on was granted. onions was granted. importation of chilies was extended until June 30, 2007. Food crisis August 1, 2007 Single specific customs duty rates were introduced on 10 selected food items, instead of the surcharge on imports, Value Added Tax (VAT), Social Responsibility Levy (SRL), cess, Port and Airport Development Levy (PAL), and other charges applicable on them at Customs. August 15, 2007 A duty waiver of Rs 5/kg was granted on importation of B’ onions until December 31, 2007. August 20, 2007 A duty waiver of Rs 5/kg was granted on importation of potatoes until October 31, 2007. October 15, 2007 A full duty waiver was granted on importation of semimilled or wholly milled rice and broken rice until December 31, 2007. December 2007 Duty waivers, granted on essential food items, were removed except for edible oil. January 1, 2008 Surcharge on customs duty was increased to 15% from 10%. January 9, 2008 A single levy (SCL) was introduced on 10 essential goods in place of customs duty, VAT, PAL, SRL, and other charges. August 25, 2008 Customs duty on the importation of rice was suspended. September 16, 2008 SCL on the importation of potatoes was increased from Rs 15/kg to Rs 20/kg. Pre-GFC January 1, 2009 PAL rate was increased from 3% to 5%. February 4, 2009 SCL on importation of other potato SCL on importation of B’ onion was SCL on importation of the following was increased from Rs 20/kg to Rs. increased from Rs 20/kg to Rs 25/ food items was increased, valid for 25/kg, valid for 6 months. kg, valid for 6 months. 6 months: Chilies (neither crushed nor ground) from Rs 30/kg to Rs 40/ kg; chilies (crushed or ground) from Rs 40/kg to Rs 50/kg. Post GFC September 30, 2009 SCL on importation of other SCL on importation of B’ onion was SCL on importation of dried chilies potatoes was extended for another extended for another 3 months at was extended for another 3 months 3 months at previous rates. previous rates. at previous rates. November 10, 2009 A recoverable tax rate of Rs 5/kg SCL on importation of other potato SCL on importation of B’ onion was SCL on importation of the following was imposed instead of all taxes was reduced from Rs 25/kg to Rs reduced from Rs 25/kg to Rs 10/kg, food items was reduced, valid for and levies applied on importation 10/kg, valid for 3 months. valid for 3 months. SCL of Rs 10/kg 3 months: Chilies (neither crushed of rice. was imposed on importation of red nor ground) from Rs 40/kg to Rs 20/ onion, valid for 3 months, in lieu of kg; chilies (crushed or ground) from customs duty and all other taxes. Rs 50/kg to Rs 25/kg. December 21, 2009 SCL of Rs 1/kg on importation of rice was imposed, valid for 3 months, in lieu of customs duty and all other taxes. (Continued) A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R 44 A P P E N D I X TA B L E S APPENDIX TABLE 6A: Trade Policy Measures on Rice, Potato, Onion, and Chilies, 2007–11 (continued) PERIOD MONTH RICE POTATO ONION CHILIES Post GFC February 10, 2010 SCL on importation of potatoes was SCL on importation of Bombay extended for another 3 months at onion and red onion was extended previous rate. for another 3 months. March 20, 2010 SCL on importation of rice was removed and customs duty and other applicable taxes and levies reimposed. May 10, 2010 SCL on importation of potatoes was SCL on importation of Bombay SCL on importation of dried chilies extended for another 3 months at onion and red onion was extended was extended for another 3 months previous rate. for another 3 months at previous at previous rate. rates. August 10, 2010 SCL on importation of potatoes was SCL on importation of red onion SCL on importation of dried chilies extended for another 4 months at and Bombay onion was increased was extended for another 3 months previous rate. from Rs 10/kg to Rs 25/kg. at previous rate. September 10, 2010 SCL on importation of potatoes was increased from Rs10/kg to Rs 30/kg for another 3 months. October 30, 2010 SCL on importation of potatoes was SCL on importation of Bombay reduced to Rs 10/kg. onion was reduced to Rs 10/kg. December 10, 2010 SCL on importation of potatoes was SCL on the importation of Bombay extended for another 4 months at onion and red onion was extended previous rate. for another 4 months at previous rates. February 20, 2011 SCL on potatoes was increased from Rs 10/kg to Rs 20/kg for 3 months. April 10, 2011 SCL on the importation of red SCL on importation of chilies was onions and big onions was extended for another 4 months at extended for another 4 months. previous rate. May 3, 2011 SCL on the importation of potatoes was increased from Rs 20/kg to Rs 30/kg for 4 months. August 10, 2011 SCL on the importation of potatoes SCL on the importation of big SCL on importation of dried chilies was increased from Rs 30/kg to Rs onions was increased from Rs 10/kg was extended for another 4 months 35/kg for 4 months. to Rs 25/kg for 4 months. The SCL at the previous rate. on importation of red onions was extended for another 4 months. October 14, 2011 SCL on the importation of crushed or ground chilies was increased from Rs 25/kg to Rs 40/kg. November 22, 2011 SCL of Rs 40/kg for the importation of crushed or ground chilies. December 10, 2011 SCL on importation of potatoes was The SCL on the importation of The SCL on the importation reduced from Rs 35/kg to Rs 20/kg. red onions and big onions was of chilies (neither crushed nor extended for another 4 months. ground) was extended for another 4 months. Source: Central Bank of Sri Lanka. W H AT I S T H E C O S T O F A B O W L O F R I C E ? T H E I M PA C T O F S R I L A N K A' S C U R R E N T T R A D E A N D P R I C E P O L I C I E S A P P E N D I X TA B L E S 45 APPENDIX TABLE 6B: Trade Policy Measures on Coconuts and Related Products, Milk and Milk Products, Sugar, and Wheat, 2007–12 COCONUT AND PERIOD MONTH RELATED PRODUCTS MILK AND MILK PRODUCTS SUGAR WHEAT Pre-crisis June 1, 2007 Duty waivers of Rs 2/kg were granted on cane sugar, beet sugar, other sugar, and Sakkara; duty waivers of Rs 3/kg were granted on white crystalline cane sugar and white crystalline beet sugar. Food Crisis August 1, 2007 Single specific customs duty rates were introduced on 10 selected food items instead of the surcharge on imports, Value Added Tax (VAT), Social Responsibility Levy (SRL), cess, Port and Airport Development Levy (PAL) and other charges applicable on them at Customs. August 1, 2007 Duty on importation of copra was A duty waiver of Rs 1.50/kg was removed until February 29, 2008. granted on importation of beet sugar A 10% duty waiver was granted on and jaggery. importation of edible oil for a period of 1 month. December 2007 Duty waivers granted on essential food items were removed, except for edible oil. January 1, 2008 Surcharge on customs duty was increased to 15% from 10%. January 9, 2008 A single levy (SCL) was introduced on 10 essential goods in place of customs duty, VAT, PAL, SRL, and other charges. April 3, 2008 A duty waiver of 23% was granted on the importation of palm oil and coconut oil. May 1, 2008 SCL was introduced on the importation of milk powder. July 25, 2008 Duty waiver granted on the importation of palm oil and coconut oil was reduced from 23% to 13%. Pre GFC November 7, 2008 SCL on imported milk powder was SCL on the importation of sugar was Import duty on wheat grain was increased from Rs 5/kg to Rs 15/kg. increased from Rs 14/kg to Rs 16/ increased to 10% from 6%. A 5% cess kg. Customs duty on raw sugar was was imposed on the importation of increased from Rs 10/kg to Rs 12/kg. wheat flour. December 3, 2008 Customs duty on edible oils was increased. January 1, 2009 PAL rate was increased from 3% to 5%. January 21, 2009 SCL on the importation of milk powder was increased from Rs 15/kg to Rs 35/kg for 6 months. February 4, 2009 Customs duty on importation of wheat grain was revised from 15% to 15% or Rs 10/kg, whichever is higher. February 27, 2009 SCL on importation of milk powder was increased from Rs 35/kg to Rs 55/ kg, valid for 6 months. March 31, 2009 SCL of Rs 55/kg on importation Customs duty on importation of of milk powder was removed and wheat flour was increased from 15% customs duty was reimposed at 28% or Rs 12.50/kg to 15% or Rs 16/kg, or Rs 125/kg, whichever is higher. whichever is higher. Total recoverable tax/levies on milk powder imports was Rs 125/kg. Post GFC April 7, 2009 Customs duty on importation of some categories of edible oil in the form of crude was increased from 28% or Rs 35/kg to 28% or Rs 55/kg, whichever is higher. Customs duty on importation of virgin coconut oil and margarine was increased from 28% or Rs 40/kg to 28% or Rs 60/kg, whichever is higher. July 24, 2009 Total recoverable tax/levies on milk powder imports was increased from Rs 125/kg to Rs 145/kg. (Continued) A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R 46 A P P E N D I X TA B L E S APPENDIX TABLE 6B: Trade Policy Measures on Coconuts and Related Products (continued) COCONUT AND PERIOD MONTH RELATED PRODUCTS MILK AND MILK PRODUCTS SUGAR WHEAT Post GFC November 10, 2009 Total recoverable tax rate on SCL on importation of sugar was importation of milk powder was reduced from Rs 16/kg to Rs 6/kg, reduced from Rs 145/kg to Rs 100/kg. valid for 3 months. December 18, 2009 A full customs duty waiver was granted on importation of wheat grain until March 31, 2010. December 21, 2009 SCL on importation of sugar was reduced from Rs 6/kg to Rs 1/kg, valid for 3 months. February 10, 2010 SCL on importation of sugar was extended for another 3 months. February 16, 2010 Total recoverable tax rate on importation of milk powder was reduced from Rs 100/kg to Rs 50/kg. March 21, 2010 SCL of Rs 1/kg on importation of sugar was extended for another 3 months. May 1, 2010 A customs duty of Rs 2/kg was reimposed on importation of wheat grain. June 21, 2010 SCL on importation of sugar was Duty waiver of Rs 8/kg on importation increased from Rs 1/kg to Rs 5/kg for of wheat grain was removed, making another 3 months. the applicable customs duty 15% or Rs 10/kg, whichever is higher. September 21, 2010 SCL of Rs 5/kg on importation of sugar was extended for another 3 months. December 10, 2010 SCL on the importation of sugar was extended for another 4 months at previous rates. January 22, 2011 A full customs import duty waiver was granted on the importation of milk powder. Hence only the PAL and NBT will be applicable on the importation of milk powder. April 10, 2011 The SCL on the importation of sugar was extended for another 4 months. August 10, 2011 The SCL on importation of sugar was extended for another 4 months. December 10, 2011 The SCL on importation of sugar was extended for another 4 months. January 13, 2012 The SCL for sugar was increased from Rs 5/kg to Rs 10/kg. Source: Central Bank of Sri Lanka. W H AT I S T H E C O S T O F A B O W L O F R I C E ? T H E I M PA C T O F S R I L A N K A' S C U R R E N T T R A D E A N D P R I C E P O L I C I E S A P P E N D I X TA B L E S 47 APPENDIX TABLE 7: Changes in Levels of Special Commodity Levy, 2008–12 DATE SPECIAL COMMODITY LEVY (SCL) January 9, 2008 Instead of customs duty, Value Added Tax (VAT), Ports and Airport Development Levy (PAL), Social Responsibility Levy (SRL), and other charges applicable at Customs on importation of 10 essential food items, a single levy, an SCL, was introduced. SCL was introduced on importation of dried sprats and B’ onion (Rs 20/kg); potatoes, peas, chickpeas (Rs 15/kg); green gram (Rs 13/kg); masoor dhal (Rs 6/kg); chilies (Rs 30/ kg); fish (Rs 25/kg); sugar (Rs 14/kg). May 1, 2008 SCL was introduced on importation of milk powder in addition to the 10 food items announced earlier. August 25, 2008 SCL on importation of potatoes was increased from Rs 15/kg to Rs 20/kg. January 21, 2009 SCL on the importation of milk powder was increased from Rs 15/kg to Rs 35/kg. February 4, 2009 SCL on dried sprats, potatoes, and B’ onions was increased from Rs 20/kg to Rs 25/kg. SCL on chickpeas and peas was increased from Rs 15/kg to Rs 25/kg. SCL on green gram was increased from Rs 13/kg to Rs 15/kg. SCL on chilies (neither crushed nor ground) was increased from Rs 30/kg to Rs 40/kg. SCL on crushed or ground chilies was increased from Rs 40/kg to Rs 50/kg. SCL on some categories of prepared/preserved fish was increased from Rs 25/kg to Rs 40/kg. February 27, 2009 SCL on the importation of milk powder was increased from Rs 35/kg to Rs 55/kg. March 31, 2009 SCL on the importation of milk powder was removed. SCL on dried sprats was increased from Rs 25/kg to Rs 30/kg. SCL on peas was increased from Rs 25/kg to Rs 35/kg. SCL on chickpeas was increased from Rs 25/kg to Rs 30/kg. SCL on masoor dhal (whole) was increased from Rs 4/kg to Rs 15/kg; SCL on masoor dhal (spilt) was increased from Rs 6/kg to Rs 20/kg. SCL on some categories of prepared/preserved fish was increased from Rs 40/kg to Rs 85/kg. November 2009 SCLs reduced on some commodities including potatoes, red onions, B’ onions, garlic, chickpeas, white/red sugar, and chilies. December 2009 SCL was introduced on importation of rice, instead of all other taxes charge at Customs; applicable rate was Rs 1/kg. SCL on importation of sugar was reduced from Rs 6/kg to Rs 1/kg. February 10, 2010 SCL on importation of dried sprats, potatoes, Bombay onion, peas, chickpeas, green gram, dhal, canned fish, sugar, red onion, and garlic was extended for another 3 months. March 20, 2010 SCL on importation of rice was removed. March 21, 2010 SCL of Rs 1/kg on importation of sugar was extended for another 3 months. May 10, 2010 SCL on importation of dried sprats, potatoes, Bombay onion, red onion, peas, chickpeas, dhal, green gram, dried chilies, and canned fish was extended for another 3 months at previous rates. June 21, 2010 SCL on importation of sugar was increased from Rs 1/kg to Rs 5/kg for another 3 months. August 10, 2010 SCL on importation of red onion and Bombay onion was increased from Rs 10/kg to Rs 25/kg. SCL on importation of garlic was reduced from Rs 30/kg to Rs 10/kg. SCL on importation of whole peas and whole chickpeas was reduced from Rs 15/kg to Rs 10/kg. SCL on importation of whole dhal and split dhal was increased from Rs 1/kg to Rs 10/kg and Rs 2/kg to Rs 15/kg, respectively. SCL on importation of dried sprats, potatoes, split peas, split chickpeas, green gram, dried chilies and canned fish was extended for another 4 months as previous rates. September 10, 2010 SCL on importation of potatoes was increased from Rs 10/kg to Rs 30/kg for another 3 months. September 21, 2010 SCL of Rs 5/kg on importation of sugar was extended for another 3 months. October 30, 2010 SCL on importation of Bombay onion and potatoes was reduced to Rs 10/kg. December 10, 2010 SCL on the importation of dried sprats, potatoes, Bombay onion, peas, chickpeas, green gram, dhal, canned fish, sugar, red onion, and garlic was extended for another 4 months at previous rates. February 20, 2011 SCL on potatoes was increased from Rs 10/kg to Rs 20/kg for 3 months. April 10, 2011 SCL on the importation of dried sprats, red onions, big onions, garlic, peas, chickpeas, green gram, dhal, chilies, canned fish, and sugar was extended for another 4 months. May 3, 2011 SCL on the importation of potatoes was increased from Rs 20/kg to Rs 30/kg for 4 months. August 10, 2011 SCL was increased on the importation of potatoes (Rs 30/kg to Rs 35/kg), big onions (Rs 10/kg to Rs 25/kg), and garlic (Rs 10/kg to Rs 25/kg) for 4 months. The SCL on importation of dried sprats, red onions, peas, chickpeas, green gram, dhal, chilies, canned fish, and sugar was extended for another 4 months. October 14, 2011 SCL was increased on the importation of green gram (Rs 15/kg to Rs 30/kg) and crushed or ground chilies (Rs 25/kg to Rs 40/kg). SCL of Rs 90/kg was introduced on the importation of black gram (HS Code 0713.31.90), of Rs 75/kg on cowpeas (HS Code 0713.39.10), and of Rs 30/kg on millet (HS Code 1008.20). November 22, 2011 SCL was Rs 250/kg for the importation of maldive fish; Rs 100/kg for the importation of dried fish, black gram, cowpeas, and shelled groundnuts; Rs 50/kg for green gram, fennel seeds, mathe seeds, mustard seeds; Rs. 45/kg for neither crushed nor ground seeds of coriander; Rs 90/kg for crushed or ground seeds of coriander; Rs 150/kg for cumin; Rs 300/kg for turmeric; Rs 200/kg for neither crushed nor ground turmeric; Rs 60/kg for fresh orange; Rs 35/kg for fresh mandarin; Rs 120/kg for fresh grapes; and Rs 45/kg for fresh apples; Rs 40/kg for the importation of crushed or ground chilies; Rs 75/kg for millet; Rs 150/kg for kurakkan flour; Rs 200/kg for black gram flour. December 10, 2011 SCL on importation of dried sprats, B' onion, garlic, red onions, peas, chickpeas, dhal, chilies (neither crushed nor ground), canned fish, and sugar was extended for another 4 months. The SCL on importation of potato was reduced from Rs 35/kg to Rs 20/kg. January 13, 2012 SCL on the importation of garlic was increased from Rs 25/kg to Rs 40/kg for 4 months. SCL increased on the importation of whole peas (Rs 10/kg to Rs 15/kg) and split peas (Rs 15/kg to Rs 20/kg) for 4 months. SCL on the importation of dhal was increased to Rs 18, 22, 18, and 22/kg for whole red lentils, split red lentils, whole yellow lentils, and split yellow lentils, respectively. SCL was introduced on the importation of crude oil and palm stearin (Rs 65 per kg) and other edible oil (Rs 75/kg). SCL was increased for sugar from Rs 5/kg to Rs 10/kg. SCL was introduced for fish at 30% or Rs 35/kg, whichever is higher. A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R 48 A P P E N D I X TA B L E S APPENDIX TABLE 8: Export Cesses COMMODITY MEAN PERCENTAGE MEAN RS/KG MODE PERCENTAGE MODE RS/KG Export cesses on export-oriented products Tea 2.50 1.14 0.00 0.00 Coconut 11.82 12.62 0.00 0.00 Rubber 0.00 0.00 0.00 0.00 Cinnamon 8.33 0.00 10.00 0.00 Cloves 10.00 0.00 10.00 n.a Average 6.53 2.75 4.00 0.00 Export cesses on import-competing products Chilies 15.00 17.50 30.00 35.00 Maize 11.67 0.00 10.00 0.00 Rice 2.00 0.00 0.00 0.00 Wheat 5.63 0.00 0.00 0.00 Vegetables 20.26 26.87 30.00 35.00 Milk 16.96 55.43 30.00 100 Fruits 28.33 54.67 30.00 60.00 Potato 14.38 31.00 5.00 0.00 Onion 11.25 13.13 0.00 0.00 Average 13.94 22.07 15.00 25.56 W H AT I S T H E C O S T O F A B O W L O F R I C E ? T H E I M PA C T O F S R I L A N K A' S C U R R E N T T R A D E A N D P R I C E P O L I C I E S Agriculture and Rural Development (ARD) 1818 H Street, NW Washington, D.C. 20433 USA Telephone: 202-477-1000 Internet: www.worldbank.org/ard