Report No. 100 36795 SouthAsia: Agriculture and Rural Development Sector Human Development Sector Enhancing Food Security in Afghanistan: Private Markets and Public Policy Options August 2005 Discussion Paper Series ENHANCING FOOD SECURITY IN AFGHANISTAN: PRIVATE MARKETS AND PUBLIC POLICY OPTIONS Agriculture and Rural Development Sector and Human Development Sector South Asia Region The World Bank August 2005 Discussion Papers are published to communicate the results of The World Bank's work to the development community with the least possible delay. The typescript manuscript of this paper therefore has not been prepared in accordance with the procedures appropriate to formally edited texts. Some sources cited in the paper may be informal documents that are not readily available. The findings, interpretations, and conclusions expressed herein do not necessarily reflect the views of the International Bank for Reconstruction and Development / The World Bank and its affiliated organizations, or those of the Executive Directors of The World Bank 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. CONTENTS Acknowledgements..............................................................................................................i Abbreviations and Acronyms.............................................................................................ii Executive Summary ............................................................................................................iii I. Introduction ...............................................................................................................1 Food Policy Objectives and Scope........................................................................1 Key Issues .............................................................................................................2 Outline of the Report.............................................................................................2 II. Food Production .....................................................................................................3 Agro-ecological zones...........................................................................................3 Utilization of Agricultural Technologies ..............................................................6 Access to Land......................................................................................................6 Major Crops ..........................................................................................................7 Wheat ..........................................................................................................7 Other Grains ................................................................................................10 Non-Cereals.................................................................................................11 Opium..........................................................................................................11 Livestock.....................................................................................................13 Constraints on Agricultural Production ................................................................14 Irrigation......................................................................................................14 Agricultural Research and Extension..........................................................15 Rural Credit.................................................................................................15 III. International Trade, Domestic Markets and Availability of Food.............17 International Trade and Total Availability of Foodgrains ....................................17 Domestic Food Markets........................................................................................20 Wheat Milling .............................................................................................23 Production Shocks and Market Price Fluctuations ...............................................24 Food Aid ...............................................................................................................27 Policy Implications ...............................................................................................28 IV. Reducing Poverty and Increasing Household Food Security .....................31 Rural Poverty ........................................................................................................31 Urban Poverty and Food Security.........................................................................33 Infant Mortality and Malnutrition.........................................................................34 Reducing Poverty..................................................................................................34 Existing Social Protection Policies .......................................................................35 Food Aid Supported Programs..............................................................................36 Targeting of Food and Other Safety Net Interventions.........................................36 V. Conclusions and Policy Implications..................................................................41 Food Production....................................................................................................42 Domestic and International Markets .....................................................................43 Food Consumption, Poverty and Food Aid...........................................................44 Food Policy Coordination .....................................................................................45 Annexes Annex 1: Wheat Production Estimates in Afghanistan: NRVA and FAO data................49 Table A1.1: Afghanistan: Production of Wheat, 2002 and 2003 (FAO Estimates)........52 Table A1.2: Afghanistan: Production of Wheat, 2002 and 2003 (NRVA Estimates) ....53 Table A1.3: Net Imports under Alternative Estimates of Production and Demand 2003/4.............................................................................................................54 Annex 2: Summary Evaluation of the WFP's PRRO in Afghanistan ...............................55 Annex 3: Ensuring Food Security following Production Shortfalls ­ The Role of Food Aid ............................................................................................................57 Box A3.1: Where Markets are Weak: Going Carefully with Cash.................................58 Annex 4: Supplementary Annex Tables............................................................................61 Table A4.1: Regional Wheat and Rice Production, Trade and Consumption, 2001 (FAO data).........................................................................................................62 Table A4.2: Grain Production and Consumption in Afghanistan by Province, 2002 and 2003....................................................................................................64 Table A4.3: Grain Production and Consumption in Afghanistan by Region, 2002 and 2003....................................................................................................65 Table A4.4: Medium Sized Milling Facilities in Afghanistan, 2004..............................66 Table A4.5: Results of Anthropometric Surveys in Kabul (1995 ­ 2003)......................66 Table A4.6: Human and Social Capital ­ Policy Priorities and Programs......................67 Table A4.7: Afghanistan: LSP Significant Achievements in 1382 (FY2003/04)...........68 Table A4.8: Afghanistan: Food Aid Costs, 2002-03 ......................................................69 Table A4.9: Afghanistan: Food Aid Costs, 2002-03. (Alternative Estimates) ...............69 Box A4.1: National Emergency Employment Program .................................................70 VI. References ...................................................................................................................71 Tables, Figures and Boxes Table 1: Afghanistan: Wheat Availability, 2001-04 (thousand tons) ..............................iv Table 2.1: Afghanistan Land Cover by Agro-ecological Zone (thousand hectares)...........4 Table 2.2: Afghanistan: Land Holdings by Rural Household Groups................................7 Table 2.3: Wheat Production in Afghanistan by Region, 2002 and 2003...........................8 Table 2.4: Area, Production and Yield of Wheat in Afghanistan, 1999-2004....................9 Table 2.5: Poppy Cultivation in Afghanistan, 2003 (NRVA Survey Data)........................12 Table 2.6: Gross Revenues from Opium and Wheat Cultivation........................................13 Table 2.7: Livestock Population per Family.......................................................................14 Table 2.8: Afghanistan Livestock Ownership, 2003...........................................................13 Table 3.1: Afghanistan: Wheat Availability, 2000-2004....................................................17 Table 3.2: Afghanistan: Grain Imports, 1999/00 - 2003/04................................................19 Table 3.3: Wheat Production and Consumption in South and Central Asia, 2001.............20 Table 3.4: Regional Wheat Markets in Afghanistan...........................................................22 Table 3.5: Correlation Coefficients for Wheat Prices, January 2002 to March 2004.........25 Table 4.1: Distribution of Calories across Poverty Groups, Rural Afghanistan 2003........31 Table 4.2: Rural Poverty by Region....................................................................................32 Table 4.3: Afghanistan: Poverty and Food Consumption by Rural Household Groups, 2003 ..................................................................................................32 Table 4.4: Safety Net Program Participation, Rural Afghanistan 2003..............................37 Table 4.5: Program Targeting Indicators under Various Eligibility Criteria (%), Rural Afghanistan...........................................................................................38 Table 5.1: The Afghanistan Food System: Policy and Program Priorities ........................46 Figure 2.1: Afghanistan Wheat Production, 1999 to 2004...................................................9 Figure 3.1: Wheat and Wheat Flour Prices in Afghanistan, 2000-2004...............................25 Figure 4.1: Daily Caloric Distribution per Food Group, By Consumption Quintile:...........33 Box 3.1: Estimates of Total Availability of Wheat ..........................................................18 Box 3.2: Price Stabilization, Trade and Food Security: The Bangladesh Experience......26 Box 3.3: Food Aid and Incentives for Domestic Production............................................28 Box 4.1: Background to WFP PRRO in Afghanistan.......................................................37 Map Map 3.1: Afghanistan Regional Wheat Markets...............................................................22 ACKNOWLEDGEMENTS This report was written by a team led by Paul Dorosh, with major contributions from Abi Masefield, Christine Allison and Philippe Chabot. The discussion of constraints on agricultural production in Chapter 2 draws heavily from an Afghanistan Agricultural Sector Strategy document written by Deepak Alhuwalia. Surveys of wheat millers and traders conducted by Mercy Corps and a background paper on wheat markets by Philippe Chabot (Mercy Corps) underpin much of the analysis in Chapter 3. Chapter 4 is largely based on the World Bank's report Poverty, Vulnerability, and Social Protection in Afghanistan (2005). William Byrd, Dina Umali-Deininger and other members of the Afghanistan Country Team (Stephane Guimbert, Peter Middlebrook, To Nhu Dao and Renos Vakis) provided critical inputs and advice, which helped sharpen the report. Valuable comments were also received from peer reviewers Harold Alderman (AFTHD) and Arnaud Cauchois (European Commission). In Afghanistan, Wendy Johnecheck and Andrew Pinney in the Ministry of Rehabilitation and Rural Development were hugely helpful. The report was prepared under the overall supervision of Gajan Pathmanathan (Sector Manager, SASAR) and Alastair McKechnie (Country Director, Afghanistan). i ABBREVIATIONS AND ACRONYMS ACF Accion Contra Faim BCM Billion Cubic Meters CFW Cash for Work CSO Central Statistics Office FAO Food and Agricultural Organization FFE Food for Education FFT Food for Training FFW Food for Work IDP Internally Displaced People LSP PIP Livelihoods and Social Protection Public Investment Program MAAH Ministry of Agriculture and Animal Husbandry MDG Millennium Development Goal MISFA Micro-Finance Support Facility of Afghanistan MIWRE Ministry of Irrigation, Water and Rural Electricity MRRD Ministry of Rehabilitation and Rural Development NEEP National Emergency Employment Program NGO Non-governmental Organization NRVA National Rural Vulnerability Assessment NSP National Solidarity Program PRRO Protracted Relief and Recovery Operation SAF Securing Afghanistan's Future TISA Transitional Islamic State of Afghanistan UNICEF United Nations Children Fund UNODC United Nations Office for Drug Control WFP World Food Programme ii EXECUTIVE SUMMARY In the last two decades, Afghanistan has endured protracted conflict, political upheaval, and natural calamities (including droughts) that have threatened the food security and indeed the very lives of much of the population. Poverty and vulnerability are widespread, both in rural and urban Afghanistan. The country experienced major disruptions to productive investment, markets, public institutions and household incomes. However, a combination of private international trade, migration, entrepreneurial activities and other household coping strategies (including sales and slaughter of livestock) together with the work of humanitarian agencies have all helped to mitigate the extent of human suffering and loss of lives. Now, with growing peace and prosperity in the country, and an elected government, the conditions to ensure greater food security for the people of Afghanistan are in place. Moving forward, coherent public policies and programs are required to increase the country's self-reliance, mitigate the risks of future shocks to households, and improve the nutritional status and overall human development of its population. This report analyzes some key aspects of food security, namely production, trade, markets and food aid at the national level, and consumption at the household level. In doing so it aspires to make a contribution to the on-going work in Afghanistan regarding the attainment of the poverty and hunger Millennium Development Goal (MDG #1). The major findings of the report can be summarized as follows: · Food security (at the national level) does not necessarily require national self-sufficiency in wheat or other food staples, as long as the country has access to international markets. Rather, diversification into legal high-value crops and livestock products may be the most effective means of increasing food security, by generating foreign exchange and raising the incomes and purchasing power of the rural poor. · In spite of very difficult conditions, wheat markets in Afghanistan have performed fairly well and private sector international trade has helped to stabilize supply and prices. Therefore, further developing the infrastructure and institutions to support wheat markets and facilitating private sector trade is called for and will enhance food security. · At the household level, food insecurity in Afghanistan is largely caused by inadequate access to food resulting from low household incomes. For most of Afghanistan, where availability of food is not a constraint, increasing cash incomes is the more efficient means of enhancing food security of the poor. · Development of both private and public capacity for data collection and analysis is a high priority for effective formulation, assessment and implementation of food policies. Food Policy Objectives and Scope The ultimate objective of food policy in Afghanistan is to ensure food security for all Afghans, where food security is defined as access by all people at all times to the food needed for an active and healthy life (1996 World Food Summit). This objective is echoed in the first MDG, which calls for the eradication of extreme poverty and hunger, with a significant reduction in the number of people suffering from hunger by 2015. Achieving a reduction in hunger and increased food security in turn requires that the three major components are met: Availability of food (domestic food production, commercial imports, food aid and security stocks); Access to food (the ability to produce or purchase food at the household level, as determined by income, other iii resources, government transfer programs and market prices), and Nutrition (determined by actual food intake, health, sanitation and caring practices). The food policy framework proposed in this report focuses on policies related directly to food production, food markets, and consumption, with special attention to wheat, the major food staple of Afghanistan. This report is not intended as a final or complete food policy framework for Afghanistan, even for the narrower definition of food security delineated above. Instead, this report is designed to support a process of dialogue and debate by the Afghan government and civil society leading to the formulation and implementation of a national food policy and improved food security outcomes for the people of Afghanistan. Food Production Food production systems vary greatly across Afghanistan, and are influenced by both agro- ecological factors (e.g., elevation, soil type, rainfall, and availability of irrigation) and socio- economic factors (e.g., market proximity and infrastructure ­ roads, electricity, communications etc., ­ security, level of incomes). Due to mountainous terrain in much of the country and an arid to semi-arid climate, only about 14 percent of total area is cultivated with crops. Of this agricultural land, about 40 percent is irrigated. Wheat dominates crop production in Afghanistan: about 70 percent of crop area cultivated is planted with wheat (of which about 70 percent is irrigated). Other cereals (rice, barley and maize ­ the latter two crops produced largely for animal feed) account for another 15 percent of area cultivated. Poppy, an extremely labor intensive crop grown illegally on about 2 percent of cultivated land, is a major source of income for opium producers and traders, and some, though not necessarily all, farmers cultivating the crop. Availability of water is a limiting factor on agricultural production in most years, and wheat production has varied considerably in recent years because of both drought and deterioration in irrigation infrastructure. After severe droughts in 2000 and 2001, adequate rainfall and snow melt in late 2002 permitted a large expansion in area sown in November-December for the wheat crop harvested in April-May of 2003 and resulted in a record wheat harvest, estimated at 4.36 million tons.1 Table 1: Afghanistan: Wheat Availability, 2001-04 (thousand tons) Average 2002-04 2000 2001 2002 2003 2004 ('000 tons) (share) Availability 2357 2817 3546 4912 4005 4154 100% Production 1469 1597 2686 4362 2293 3114 75% Imports 888 1220 860 550 1712 1041 25% Commercial 650 908 500 300 1300 700 17% Food Aid 238 312 360 250 412 341 8% Note: Years indicate calendar year of main harvest (April - May). Data on imports and availability are for the following July/June fiscal year. For example, the figures for 2004 indicate production in April - May 2004 and availability for 2004/05. Data limitations mean that these estimates may carry significant margins of error. 2004 data are preliminary. Source: FAO/WFP Kabul. 1Chapter 2 and Annex 1 discuss data limitations and alternate estimates of wheat production in Afghanistan. iv Livestock (mainly sheep, goats, cattle and poultry) is also a major source of food and income for Afghan farmers, and is an integral part of most farming systems in the country. Range pasture land covers about 45 percent of total land area and has traditionally supported a large livestock population, with animals being the only source of income for some groups such as the Kuchi nomads. According to a 2003 census, the livestock population is about 34 million (including 12 million of poultry), about a two-thirds decline from pre-drought estimates (for 1995). Achieving substantial increases in agricultural and livestock production will require a number of measures, viz.: · Investments in irrigation rehabilitation and irrigation management reforms (establishment of a river basin approach and support for local management of water resources through water users associations). · Investments in agricultural research and extension, including reforms in extension services to make them more responsive to farmers. · Maintaining adequate incentives for efficient domestic production through policies and infrastructure investments that promote development of competitive agricultural markets and avoid price disincentive effects of subsidized imports (food aid or commercial sales. The optimal agricultural policy for Afghanistan most likely rests with diversification (away from poppy) into legal high-value crops and livestock products, and this is turn will promote food security by generating employment and raising incomes of the rural poor. Food security does not necessarily require national self-sufficiency in wheat or other food staples. (Nonetheless, in years of good wheat harvests, Afghanistan may approach wheat self-sufficiency.) As long as the country has sufficient purchasing power to acquire wheat and other food on international markets, and market flows are not blocked, commercial trade can ensure adequate supply of food at the national level. Similarly, as long as household incomes are sufficient to purchase available food in local markets, these households need not be self-sufficient in food either. Food Markets (International Trade and Domestic Markets) Food markets play a crucial role in food security, linking domestic producers and international suppliers to consumers, including urban, rural non-farm and farm households.2 Available evidence suggests that these markets have performed fairly well in Afghanistan under very difficult conditions of war, internal conflict and poor infrastructure. However, further development of efficient private markets is essential to translating food security at the national level (i.e., sufficient overall availability of food) into adequate availability of food at the local level. Because of the country's mountainous terrain, poor road infrastructure and tenuous security situation, internal marketing costs are very high. Available evidence suggests that regional markets for wheat are generally better integrated with markets in bordering countries than with other markets in Afghanistan. Only in years of good harvests are there significant flows of wheat between regions of the country, in particular from the northern region of the country to Kabul, and to other markets in the central and western regions. Wheat imports (both commercial imports and food aid) are a major source of wheat supply and total food availability in Afghanistan, averaging 1.04 million tons annually from 2002/03 to 2Only about one fourth of all rural households in Afghanistan are self-sufficient in wheat. v 2004/05 ­ that is 25 percent of total wheat supply over the period. Survey evidence and analysis of market prices suggest that since 2001 private sector imports (mainly originating from Pakistan and Kazakhstan) have involved substantial numbers of traders that respond quickly to shortages in domestic market supplies that would otherwise put upward pressure on prices. In this way, private sector imports have helped to stabilize both market supplies and market prices in major cities throughout Afghanistan. Food aid inflows also add to domestic supplies, but because their levels are determined by donor decisions and administrative procedures and not by market forces, they have the potential to reduce market prices. While this can benefit consumers in the short term, there is also a risk that it can depress producer prices below the levels that would prevail in markets in the absence of food aid. Thus farmer incomes and incentives for production may be affected, both directly (through reduced market prices) and indirectly (through reduced public and private incentives for investment in agricultural technology). Market price analysis and the significant scale of commercial imports suggest that food aid inflows to date have not had significant disincentive effects on market prices in Afghanistan. However, as discussed below, the high cost of food aid relative to commercial imports suggests that whatever possible a cash-based alternative aid program would be preferable. The experience of many countries indicates that efforts to stabilize food supply and prices through public sector procurement, storage and distribution of food can involve large fiscal subsidies while often having little sustainable impact on food security for the poor. Minimizing such government direct interventions in food markets, while at the same time investing in market development through security, rule of law and improved infrastructure will encourage private sector investment and further development of efficient private food markets. Specifically, government policies can promote efficient markets by: · maintaining incentives for private sector imports through a policy of free trade (no import licenses required apart from traders' registrations, and zero or minimal import tariffs and taxes), · maintaining free trade between and within provinces (no movement restrictions or taxes on domestic trade flows, and no limits on the location, quantity or prices of food purchases or sales), · promotion of private sector investment in storage and food processing (e.g., wheat milling) by improving the private sector investment climate and by avoiding restrictions on privately held stocks or the size of milling or other food processing operations, and · minimizing licensing requirements and avoiding imposition of price controls (e.g., administratively set price ceilings on bread). Food Consumption, Poverty and Food Aid Poverty and household food insecurity are widespread in Afghanistan. According to data from the 2003 National Rural Vulnerability Assessment (NRVA), almost half of the rural population had food expenditures less than what would be required to purchase adequate calories in vi 20033, and more than 20 percent of the rural sample reported actually consuming fewer than the poverty threshold of 2100 calories per day. Overall, wheat accounted for about 60 percent of calorie consumption for rural households, with poor households consuming a larger part of their calorie needs from wheat (over 70 percent). Comprehensive data on poverty or food consumption for urban areas are not yet available, though the limited evidence that exists suggests that a significant share of the urban population live in poverty, and experience acute seasonal malnutrition. Because private sector wheat markets have functioned well in most of Afghanistan in recent years, food insecurity at the household level is largely a function of income poverty. To address this wide-spread and chronic poverty, rapid economic growth is required, creating jobs and putting purchasing power into the pockets of the poor. In order to provide support to those unable to work (children, the elderly, disabled and infirmed) well-targeted transfer programs can be considered. These could provide support in the form of food if that were the most readily available resource, but given the availability of food in the market, cash-based programs would be preferable. Among other things, food-based transfers typically entail high costs, especially when the food is brought in from far afield.4 Thus, where food markets have the potential to function well if effective demand can be stimulated and availability of food is not a constraint, income-generating programs are generally a more efficient means of enhancing household food security than are food-based programs. Examples of such programs in Afghanistan include the National Emergency Employment Programme (NEEP) and the micro-finance programs. Where food is not readily available and the market is reluctant to respond to demand (thus potentially cause a price hike), provision of food can be most advantageous, especially for the poorest. In sum, to maximize the efficiency of resource use and minimize the risk of disincentives for domestic production and market development, food aid use should be restricted to situations where the comparative advantage over cash is clear, i.e., where: · availability of food in local markets is a constraint on food security; · the commodity distributed is an inferior good (i.e., a good for which quantity consumed falls as household incomes rise); · the food aid commodity can be fortified, so that it carries additional nutritional benefits that are not easily available to households in their normal diets; and / or · food aid imports are additional resources (i.e., in the absence of food aid, equivalent monetary resources would not be provided by donors). Increases in household access to food will not necessarily translate into major improvements in nutritional outcomes for poor individuals. This is a particular concern for the most vulnerable ­ pregnant and lactating women, and young children. Here, there is also an important role for 3The food expenditure cutoff used was Af. 4621/year ($94/year), equal to the cost of purchasing 2100 calories/day of the typical basket of food items consumed by the poorest 40 percent of households in rural Afghanistan. 4Average unit costs of cereals (mainly wheat) delivered in 2002-03 under the WFP Protracted Relief and Recovery Operations (PRRO) was $332/ton, compared to an average commercial retail price in Kabul, Faizabad, Herat, Jalalabad, Kandahar, and Mazar-e-Sharif of $207 per ton. This food aid cost figure includes estimates of all shipping and handling costs for cereals, but it does not include other (e.g., administrative) costs. vii nutrition and health education, improvements in sanitation and quality of water, primary health care and other measures that promote absorption of nutrients and general health. Much international experience shows that increased educational status of women contributes to improved nutritional status of their children due to a better understanding of feeding practices and the importance of hygiene. In this way, tough choices should be considered across programs, for example the relative benefits and costs of fortification of foods vis-à-vis an adult education program, assuming that resources are insufficient to do both. Food Policy Coordination The elements of a food policy framework outlined above can make a significant contribution to greater food security in Afghanistan. This framework needs to have in-built flexibility to adapt to changes in the food system, at both domestic and global levels. It also should be responsive to participatory consultations with key stakeholders. Moreover, effective formulation and implementation of specific food policies will require the establishment of a robust nationwide information system to forecast and monitor key indicators (e.g., market price stability, domestic food production, household purchasing power, commercial food imports, international trade flows, nutritional status, etc.) together with the development of both private and public analytical capacities. To this end, a high-level food policy framework committee may be formed consisting of representatives of the Ministries of Food and Agriculture, Commerce, Rural Rehabilitation and Development, and Social Affairs, together with responsible bodies for Disaster Management and National Statistics. The core economic ministries (Ministry of Economic Development and Ministry of Finance) might also be included. Such a committee can help Afghanistan continue to refine its broadly successful current policy framework in the context of the emerging National Development Strategy, and to avoid policy mistakes in responding to an evolving food security environment. viii I. INTRODUCTION In the last two decades, Afghanistan has endured protracted conflict, political upheaval, and natural calamities (including droughts) that have threatened the food security and indeed the very lives of much of the population. The country experience major disruptions to productive investment and activities, markets, public institutions and regular household incomes. However, a combination of international assistance, private international trade, migration, entrepreneurial activities and other household coping strategies (including sales and slaughter of livestock) helped to mitigate the extent of human suffering and loss of lives. Nevertheless, poverty and vulnerability remain widespread, both in rural and urban Afghanistan. Moving forward, coherent public policies and programs are required to increase the country's self-reliance, mitigate the risks of future shocks to households, and improve the nutritional status and overall human development of its population. Numerous challenges lie ahead. Production of food, particularly food grains, is characterized by low input use and low productivity. Moreover, food production is highly variable because of fluctuations in rainfall and snow melt. Private markets face high transaction costs due to damaged or non-existent infrastructure (roads, telecommunications, storage facilities and market yards) in much of the country. Furthermore, poverty and malnutrition remain widespread. According to the National Rural Vulnerability Assessment (NRVA), in 2003, about half of the rural population had food expenditures less than the food-based poverty line and about 20 percent of households consumed less than 2100 calories/person/day. Food Policy Objectives and Scope The ultimate objective of food policy in Afghanistan is to ensure food security for all Afghans, where food security is defined as access by all people at all times to the food needed for an active and healthy life (1996 World Food Summit). Achieving this food security in turn requires that the three major components of food security are met: Availability of food (domestic food production, commercial imports, food aid and security stocks); Access to food (the ability to produce or purchase food at the household level, as determined by income, other resources, government transfer programs and market prices), and Nutrition (determined by actual food intake, health, sanitation and caring practices). However, because food security, and in particular, access to food, is in part determined by incomes, a food policy could potentially encompass all policies and programs related to poverty reduction and economic development. Defining a food policy framework in this way would make it indistinguishable from the broader development strategy, and divert attention away from issues and policies more directly linked to food availability, access and utilization. Thus, while the eradication of extreme poverty, hunger and food insecurity remains the overarching objective of both food and overall development policy, this food policy framework focuses on policies related directly to food production, food markets (both international and domestic trade), and consumption, with special attention to wheat, the major food staple of Afghanistan. Such a policy framework can help to guarantee consistency across a range of sectoral policies that impact on national and household food security. An evolving food policy framework would also complement other activities to promote a comprehensive national strategy for development and the reduction of poverty and vulnerability (National Development Strategy). It is not the intention of this paper to seek early closure on food policy issues, but instead to provide analysis and to identify key policy issues and options in support of a food policy dialogue 1 Key Issues For Afghanistan, an effective food policy must address the key issues related to food security, including: · The distinction between food self-sufficiency and food-security · The role of private international trade in stabilizing food-grain prices ­ and what are the most efficient mechanisms · Policies measures to maintain incentives for domestic production of food · How best to ensure household food security, especially for the poor · The role of food aid in Afghanistan's food security. Outline of the Report This report is designed to provide data, analysis, and lessons from experience of other countries to help address these key food policy issues. The report begins with a description of the structure and development of the Afghanistan food economy, as background to an analysis of government food policy issues. Chapter 2 covers food production in Afghanistan in the context of the country's diverse agro-ecologies and overall agricultural system. Food processing and marketing, including both international and domestic trade, are addressed in Chapter 3. These two chapters together describe the determinants of availability of food at both the national level and in local markets. Chapter 4 examines household access to food and utilization of food, including consumption patterns and government food programs. Finally, Chapter 5 discusses government policy options in terms of the broad policy issues identified above and presents some concluding observations. 2 II. FOOD PRODUCTION5 Food production is central to food security in Afghanistan because most of the food consumed in Afghanistan derives from domestic production, and because production of both crops and livestock is a major source of rural incomes. As a result of sustained damage to markets, technologies and infrastructure, however, growth in Afghanistan's agricultural production fell dramatically from the period of pre-conflict (1961-1978) to the period of intense conflict (1978- 2001). Since 2002, however, there has been a strong recovery, coinciding with the end of a severe drought and improved security in many areas. This chapter begins with a description of agro-ecological zones in Afghanistan that greatly influence the agronomic potential of various crops and livestock. Because wheat dominates agricultural and food production systems, the major focus of the chapter is on wheat production systems, trends and constraints. Other agricultural products (rice and other food crops, livestock and poppy) that together account for about half of area cultivated and a large share of agricultural incomes are covered in less detail. Agro-ecological zones Afghanistan has an arid to semi-arid climate, with annual rainfall ranging from 100 to 400 mm. Thus, availability of water is a limiting factor on agricultural production. More than half the country receives less than 300 mm of rain or snow, with substantial annual and seasonal variation. For most of the country, about half of the annual precipitation typically takes place in the winter months between January and March. Temperatures range from about -10° C in winter to about 34° C in the summer. Food production systems vary tremendously across Afghanistan, according to both agro- ecological factors (e.g., elevation,6 soil type, rainfall, and availability of irrigation) and socio ­ economic factors (e.g., market proximity, infrastructure ­ roads, electricity, communications - security, level of incomes). Nonetheless, in spite of this diversity of systems and factors, it is useful to distinguish three broad regions: the northern plains (about one-fourth of national area), the central highlands (about half of national area), and the southwestern lowlands (the remaining one-fourth of national area).7 The northern plains region on the north side of the Hindu Kush range extends from Faryab province in the northwest to Takhar province in the northeast. Approximately one fourth of the land in this region is agricultural land (orchards, irrigated land and rainfed land), and the region accounts for 48 percent of national agricultural land (Table 2.1). About half of the agricultural land in the Turkestan plains (roughly one-half the total area of the region) is irrigated; crop production in the mountains in this region is mainly rainfed. 5The discussions of irrigation, agricultural technology and constraints to agricultural growth in this chapter draw extensively from World Bank (2004) Afghanistan Agricultural Policy Note. 6Altitudes vary from 150 to 8,000 meters; crops are cultivated at elevations as high as 4,000 meters. 7 FAO (1997) further divides the country into eleven agro-ecological zones. An approximate correspondence between the FAO zones and the three broad regions described above is given in Table 2.1. See also Appendix Table 1. 3 Table 2.1: Afghanistan Land Cover by Agro-ecological Zone (thousand hectares) % Irrig/ % Agric/ Orchards Irrigated Rainfed All Agric Total Agric Total % Natl Agric %Natl Area 1. Wakhan Corridor and Pamir Knot - 5 - 5 1,110 100.0 0.4 0.1 1.9 2. Badakshan 6 72 720 793 4,124 9.1 19.2 9.8 6.9 3. Central Mountains 10 479 859 1,338 11,452 35.8 11.7 16.5 19.1 4. Eastern Mountains 13 452 164 616 5,082 73.4 12.1 7.6 8.5 5. Southern Mountains and Foothills 4 523 234 756 6,439 69.1 11.7 9.3 10.7 Mountains subtotal 33 1,531 1,977 3,508 28,208 43.6 12.4 43.2 47.0 6. Northern Mountains and Foothills 8 301 2,266 2,568 9,030 11.7 28.4 31.6 15.1 7. Turkestan Plains 16 648 643 1,291 4,890 50.2 26.4 15.9 8.2 Northern 24 949 2,910 3,859 13,920 24.6 27.7 47.5 23.2 4 8. Herat-Farah Lowlands 1 210 1 211 4,887 99.6 4.3 2.6 8.1 9. Helmand Valley-Sistan Basin 22 388 22 411 4,841 94.5 8.5 5.1 8.1 10. Western Stony Deserts 0 77 0 77 3,687 99.9 2.1 0.9 6.1 11. Southwest Sand Deserts - 53 6 58 4,451 90.2 1.3 0.7 7.4 West/South 23 728 29 757 17,866 96.2 4.2 9.3 29.8 Total (excluding Herat) 80 3,208 4,916 8,124 59,993 39.5 13.5 100.0 100.0 Source: Adapted from Sloane (2001), p.4; authors' calculations. Original data from Afghanistan Agricultural Strategy, FAO, Rome 1997, Land Management Report, Table 3, p. 28. (Using data from 1972 land use by zone; data excludes Herat, for which data was unavailable at that time. Total area including Herat (4247.1 hectares) is 64,240 hectares.) Agriculture in the Hindu Kush mountain region is largely confined to valleys which cut between the mountains. Only about 12 percent of the land in the region is agricultural land (excluding pasture land). Thus, although the region accounts for 47 percent of national area, it accounts for only 23 percent on national agricultural land. Overall, about 40 percent of agricultural land in this region is irrigated. The river valleys of the Eastern Mountains (including areas surrounding Kabul and Jalalabad) are highly productive, with significant levels of double-cropping. In the other parts of the central highlands, however, productivity is generally low. Here, water is the limiting factor, and most irrigation is supplied by underground water. The west and south, nearly half of which is desert land, is the least densely populated region. Only four percent of the land is used for agriculture; nearly all of this land irrigated, since the area is generally too dry for rainfed agriculture. Overall, the region contains only about 15 percent of the country's agricultural land.8 Traditional surface systems, with intakes from rivers and streams, account for about 80 percent of the irrigated area. While these systems have generally survived, their community-based mechanisms for water management and maintenance have been adversely affected by local commanders who frequently have not respected water rights or the authority of the farmer elected and employed mirabs (water masters). This has had a significant impact on water distribution, both within some of these systems, and across irrigation systems along the same river (Sheladia Associates, 2004). Next in importance are larger modern irrigation systems, accounting for about 10 percent of irrigated area. Managed in the past through parastatal agencies, these have been greatly affected during the past twenty years as maintenance was neglected due to lack of staff, equipment and finances. Consequently only the upstream parts of the roughly 325,000 hectares of their command areas now receive proper irrigation. Karezes (a system of irrigation using underground tunnels), accounting for roughly five percent of irrigated area, have also suffered due to the war and some have been deliberately destroyed, sabotaged or mined to make rehabilitation difficult. The recent drought, coupled with a gradual disintegration of community responsibility, also led to indiscriminate groundwater pumping by some farmers leading to decline in groundwater levels in some areas and jeopardizing traditional springs and kareze fed supplies. Springs and wells constitute the remaining types of irrigation systems in the country. At the macro level, recent estimates indicate that the country has, on average, about 70 billion cubic meters (BCM) of annual water resources of which 55 BCM is surface water and 15 BCM is groundwater (Sheladia Associates, 2004). Annual water use, of which over 95 percent is for irrigation, is estimated about 20 BCM, which means that not more than 30 percent of the water resources are presently being used. Estimates of irrigated area vary. As per 1993 FAO satellite maps, which are reckoned to be the most objective and reliable data available, the upper limit scenario is about 1.5 million hectares of land that is `intensively irrigated' and about 1.6 million hectares of land that is `intermittently irrigated' depending on water availability. Actual area irrigated in recent years has been considerably lower ­ the 2003 estimate is 1.3 million hectares were intensively irrigated (SAF, Annex on Natural Resources Management). There is also considerable regional variation in water availability and use, with the north, northeast, west and southwest regions of the country accounting for about half the estimated surface water resources and three-fourths of the irrigated area in the country (World Bank, 2004a; Sheladia Associates, 2004). 8 Note that the figures in Table 2.1 do not include Herat, since data from this province was unavailable at the time of the last major agro-ecological zoning exercise (1972). 5 Utilization of Agricultural Technologies In spite of the frequent droughts and considerable economic and political uncertainties, Afghanistan's farmers have adopted new technologies on a large scale. According to a 2002/03 MAAH/MRRD/FAO/WFP survey covering 30 provinces, use of improved seeds, fertilizers and tractors is significant. Nearly 54 percent of wheat area in 2003 was sown with improved seed released in the last 10 years, while the rest was sown with local seeds or improved seeds introduced more than 10 years ago. There are regional variations, however, with less than 30 percent wheat area using improved seeds in the southwest and southern regions to over 75 percent in the east and northeast. Fertilizer use is also sizeable with a national average of about 180 kgs/ha in irrigated wheat, with regional variations ranging from an average of 345 kgs/ha in the central region to 80 kgs/ha in the southwest. Fertilizer is also commonly used for fruits, vegetables and other crops. Use of tractors is also significant and in 2002/03 about 46 percent of land was cultivated using a tractor compared to 39 percent in 2001/02. This increased use may have partly been prompted by the decrease in number of oxen due to the prolonged drought, although significant income for many farmers from the poppy sector (in Helmand particularly where more than 98 percent of land was cultivated using tractors) is also a factor influencing an increase in tractors. Access to Land In the Afghan context, ownership and access to land is complex.9 According to the 2003 National Rural Vulnerability Assessment survey, 85 percent of rural households cultivate land, 77 percent of households own land and 66 percent own irrigated land (Table 2.2).10 Pure tenants (those without any owned land) account for only eight percent of rural households. Sharecropping is not restricted to poorer households, though larger landholders may sharecrop out to pursue more profitable investments than offered by agricultural production (Grace and Pain, 2004). Nearly 70 percent of households not owning or cultivating land are poor (NRVA 2003), though landless households are also found among the non-poor: 18 and 15 percent of households in the top two consumption quintiles are landless (World Bank, 2005: Table 3.2). Very few Afghan women own resources such as land and livestock and as a result, access to income generating opportunities are relatively fewer than is the case with men. However, contrary to the commonly held assumption that Afghan women have very little involvement in income generating opportunities outside the home, women in rural Afghanistan frequently play a significant role in agricultural production. 9In this report, landed households include land owners who cultivate their own land and also share crop in or share crop out land. Sharecroppers owning no land are classified as landless. 10Average household size varies little by the land and farm characteristics described here, so shares of total rural population are approximately the same as shares of all rural households described above. 6 Table 2.2: Afghanistan: Land Holdings by Rural Household Groups Irrigated Irrigated Total Land Land Land Land Land Land HouseholdsPopulation Owned Owned OperatedOwned Owned Operated (mns) (%) (ha/hh) (ha/hh) (ha/hh) (%) (%) (%) Irrig Land Owners Large 0.14 6.7 30.92 5.56 6.42 31.6 47.4 28.6 Irrig Land Owners Small 1.29 60.7 5.32 0.67 1.16 50.3 52.6 47.8 Irrigated Land Tenants+ 0.11 4.9 0.96 0.00 1.28 0.8 0.0 4.6 Non-irrig Land Owners Lrg 0.07 3.3 21.45 0.00 4.38 10.9 0.0 9.7 Non-irrig Land Owners Sm 0.17 7.5 5.24 0.00 1.42 6.5 0.0 7.6 Non-irrig Land Tenants* 0.06 2.8 0.00 0.00 0.87 0.0 0.0 1.8 Total Farmers (Cultivators) 1.84 85.8 7.40 0.89 1.70 100.0 100.0 100.0 Total Rural Population 2.17 100.0 6.28 0.75 1.44 100.0 100.0 100.0 *Non-irrigated Land Owners (tenants) are farmers with no irrigated land. + Irrigated land (tenants) own no irrigated land but may own some rainfed land. Source: Calculated from NRVA 2003 data. Major Crops Wheat Wheat is the major food staple in Afghanistan, as well as the major crop. Area planted to wheat in 2002 was 2.3 million hectares or about 68 percent of total cultivated area, and 1.7 million hectares in 2003. Over the 2002 to 2004 period, irrigated wheat area averaged about 1.0 million hectares, approximately 10 percent more than rainfed wheat area. However, since yields on irrigated land are more than twice those on rainfed land (2.3 tons/ha versus 0.9 tons/hectare) of the same period, production on irrigated land accounts for about 70 percent of total national production. Average yields of irrigated wheat vary little across region; a similar pattern is observed for rainfed wheat (Table 2.3). About half of national wheat output is produced in the northern plains and foothills region, where both irrigated and rainfed crops are grown.11 This region is the major surplus region of Afghanistan, producing an estimated 283 kgs/capita of wheat (average 2002 and 2003), compared to a national average of 157 kgs/capita. In the mountain region of the country, over 85 percent of production is irrigated. Although the region accounts for 20 percent of national production, it is a deficit region (producing only 67 kgs/capita) due to the presence of large urban populations in Kabul and Jalalabad. The west ­ southwest region accounts for about 30 percent of wheat production, and produces approximately as much as it consumes (about 189 kgs/capita of production). Apart from some rainfed wheat cultivated in Herat and Badghis provinces, almost all of the region's wheat is irrigated. 11 According to NRVA 2003 data, nationally about 60 percent of rainfed wheat production is by farmers with no irrigated land; 86 percent of the country's rainfed wheat production is from the North. 7 Table 2.3: Wheat Production in Afghanistan by Region, 2002 and 2003 IRRIGATED Wheat RAINFED Wheat Total Wheat Area Yield Prod. Area Yield Prod. Area Yield Prod. Population Prod./cap Region (000'ha) (tons/ha) (000' tons) (000'ha) (tons/ha) (000' (000'ha) (tons/ha) (000' (mns) (kgs/person) tons) tons) 2002 Northern 464 1.9 901 449 0.8 351 913 1.4 1252 6.3 197.2 Mountains 225 2.1 477 38 0.8 31 263 1.9 508 10.6 47.7 West / SW 356 2.1 732 210 0.9 194 566 1.6 926 5.2 177.9 Total 1045 2.0 2110 697 0.8 576 1742 1.5 2686 22.2 121.0 2003 8 Northern 441 2.9 1279 984 1.1 1090 1425 1.7 2369 6.5 366.2 Mountains 308 3.0 911 45 0.5 23 353 2.6 934 10.8 86.1 West / SW 309 2.7 825 207 1.1 232 516 2.0 1057 5.3 199.3 Total 1058 2.8 3015 1236 1.1 1345 2294 1.9 4360 22.6 192.8 Growth Rate (2003 / 2002) Northern -5.0% 49.4% 42.0% 119.2% 41.7% 210.5% 56.1% 21.2% 89.2% 1.9% 85.7% Mountains 36.9% 39.5% 91.0% 18.4% -37.3% -25.8% 34.2% 37.0% 83.9% 1.9% 80.4% West / SW -13.2% 29.8% 12.7% -1.4% 21.3% 19.6% -8.8% 25.2% 14.1% 1.9% 12.0% Total 1.2% 41.1% 42.9% 77.3% 31.7% 133.5% 31.7% 23.3% 62.3% 1.9% 59.3% Note: Years shown in the table correspond to the calendar year of harvest. Most wheat is planted in November-December and harvested the following April-May. Source: MAAH/FAO National Crop Output Surveys, 2002 and 2003. Wheat production has varied considerably in recent years because of both drought and deterioration in irrigation infrastructure (Table 2.4 and Figure 2.1). After severe droughts in 2000 and 2001, adequate rainfall and snow melt in late 2002 permitted a large expansion in area sown in November-December for the wheat crop harvested in April-May of 2003. Total wheat production in 2003 was estimated at 4.36 million tons, a record harvest, 62 percent higher than in 2002. Slightly over half (54 percent) of the total 1.7 million ton gain came from increases in irrigated wheat production (1 percent gain in area and a 43 percent increase in yield). The remaining gain in production derived from a near doubling of upland area and a 32 percent gain in upland yields. Production estimates for 2004 were considerably less, however, only 2.29 million tons, a 47 percent decline relative to the 2003 estimate. Table 2.4: Area, Production and Yield of Wheat in Afghanistan, 1999-2004 1999 2000 2001 2002 2003 2004 Wheat Area (th. hectares) 2027 2029 1779 1742 2294 1766 Irrigated 1196 1189 1156 1045 1059 969 Rainfed 831 840 623 697 1235 797 Wheat Production (th. tons) 2500 1469 1597 2686 4361 2293 Irrigated 1988 1329 1514 2110 3017 1867 Rainfed 512 140 83 576 1345 426 Wheat Yields (tons/ha) 1.2 0.7 0.9 1.5 1.9 1.3 Irrigated 1.7 1.1 1.3 2.0 2.8 1.9 Rainfed 0.6 0.2 0.1 0.8 1.1 0.5 Note: Years shown in the table correspond to the calendar year of harvest. Most wheat is planted in November- December and harvested the following April-May. Source: FAO/WFP Crop and Food Supply Assessments 2003 and 2004. Figure 2.1: Afghanistan Wheat Production, 1999 to 2004 5.0 4.5 4.0 3.5 3.0 tons) 2.5 (mn2.0 1.5 1.0 0.5 0.0 1999 2000 2001 2002 2003 2004 Irrigated Rainfed 9 Evidence from the NRVA, however, suggests that the gains in wheat production between 2002 and 2003 as indicated in the FAO-WFP Crop Assessments may have been over-stated. Contrary to these official production estimates derived from crop surveys, NRVA data suggest that the total gain in production was only about 14 percent, with little change in yields between 2002 and 2003. Moreover, the large 62 percent increase in production is inconsistent with available price and trade data.12 The most likely scenario is that yield estimates for irrigated wheat were underestimated in 2002 (and earlier), and that the production increase between 2002 and 2003 was on the order of 25 percent (due mainly to an increase in non-irrigated production).13 The estimated irrigated wheat yields in Afghanistan averaged 1.8 tons/hectare in 2002 and 1.9 tons/hectare in 2004 (FAO/WFP Crop Assessments, Table 2.4), significantly less than averages of about 2.6 to 3.0 tons/hectare in neighboring Pakistan, Uzbekistan and Iran. Reported irrigated wheat yields approximated those of these neighboring countries in 2003, however (2.8 tons/hectare). Rainfed wheat yields in 2004 (0.53 tons/hectare), were also lower than yields in neighboring countries. Even in the major wheat producing provinces of the north and northeast regions, average irrigated yields in 2004 were only about 2.2 tons/ha (and only about 0.6 tons/ha for rainfed wheat). Relatively low use of improved wheat seed varieties is one reason for lower average yields. According to the agricultural survey of 2002, only about half (53 percent) of farmers used improved seed introduced in the last ten years. The remainder of the seed used was either indigenous seed or improved seed introduced more than ten years earlier, including the "zardana" variety widely used in Northern Afghanistan.14 There is substantial use of fertilizer, however, which is imported by the private sector. In 2002, an average of 180 kilograms of fertilizer per hectare, (mainly urea and di-ammonium phosphate) was used on irrigated wheat, totaling an estimated 187 thousand tons total on wheat alone. Thus, there would appear to be substantial scope for raising wheat yields and production in the country, provided security and economic incentives are favorable. Nonetheless, self-sufficiency in wheat is not necessary for food security, provided that the country has sufficient purchasing power to purchase wheat on international markets and market flows are not inhibited by excessively high import prices, export bans in neighboring countries, destruction of transport infrastructure (roads, bridges) or physical insecurity that effectively prevents trade.15 Moreover, cultivation of higher-value crops may result in higher farm incomes. Other grains Production of rice (291 thousand tons of milled rice equivalent), barley (410 thousand tons) and maize (310 thousand tons), together total only 525 thousand hectares and 1.01 million tons of 12The relatively small decrease in prices between 2002/03 and 2003/04 is unlikely to have led to only a 25 percent increase in per capita consumption implied under a no net trade scenario, suggesting that a 62 percent production increase would have resulted in a large exportable surplus. See Annex 1. 13The Swedish Committee for Afghanistan (SCA) agriculture survey conducted in the early 1990s also indicated that actual yield levels were higher than official estimates. See Maletta, H. and Favre, R. (2003). 14 Unfortunately, some of the seed provided in the emergency distribution programs in 2002 ad 2003 may have been of low quality. The amount of seed procured for distribution in autumn 2002 (23,000 tons) was nearly five times greater than what the contracted farmers were able to produce with a minimum quality control standard certified by FAO. 15Moreover, policies to provide additional incentives for wheat production (government procurement at high prices, import bans that raised domestic prices, subsidies explicitly targeted to wheat production), would involve pulling economic resources (land, labor) away from more productive uses. 10 grain, less than one quarter of both area and production of wheat in 2003. Production of paddy, the second most important cereal for food consumption, is concentrated in the northeast, (53 percent of Afghanistan's production in 2003). Maize is grown as a summer crop following winter wheat, particularly in eastern and central Afghanistan. Barley is mainly grown in higher altitude areas of central and northeast Afghanistan, including Badakshan. Most maize grain (about 60 percent) and barley (about 80 percent) are used as feed.16 These crops are also used as fodder for animals. Non-Cereals The important non-cereals are pulses, oilseeds, fruits, vegetables, fodder, and poppy. Area under poppy has increased dramatically in recent years and in 2003 was estimated to cover 80,000 hectares which is equivalent to about two percent of cultivated area or about one percent of arable area.17 Of the other non-cereal crops, fruits and vegetables are seen as particularly important, both in terms of past significance and also future cash crop prospects. The fruits sub-sector, which in the 1970s was a significant part of the Afghan economy contributing over 40 percent to total export earnings, at present covers about 2 percent of cultivated area with grapes, almond, apple, apricot and pomegranate leading the acreage. There has been a decline in area under fruit orchards in recent years ascribed mainly to conflict during the Taliban period and the drought. Most fruits are now for domestic consumption, with exports constrained due to lack of market access arising mainly from weak post-harvest management practices, lack of quality control, and lack of finance.18 Vegetables, increasingly preferred because of their short growing cycle in an uncertain socio-economic environment, also cover about 2 percent of cultivated area with the main ones being melons, watermelons, potatoes, onions and tomatoes. Fruits and vegetables are major cash crops, with marketed proportions varying from 50-70 percent of production (FAO, 2003). Cotton, which used to be significant in the 1970s, has declined to negligible levels, mainly because of the collapse of the domestic cotton industry. Opium 2.22. Opium poppy cultivation, although illegal, is widespread in Afghanistan and has been increasing in Afghanistan in recent years. Including processing and trade, poppy (and opium) account for an estimated one-third to one-half of Afghanistan's gross domestic product. According to the Afghanistan Opium Survey 2004 (UNODC, 2004), production may involve as many as 2.3 million people, 10 percent of the total population. NRVA (2003) data show only 5 percent of rural households cultivating poppy (8 percent of households in the northern region), (Table 2.5). 161998/99 FAO/WFP Crop and Food Supply Assessment. 17The growth of the poppy economy and its related complex issues has recently been analyzed in detail in World Bank, 2004a. 18Afghan dried fruits (mainly almonds and apricots) accounted for 60% of the world market in 1982; this declined to 16% by 1990 and presently is an insignificant 0.3%. See: ICARDA, 2003a; International Trade Center, Switzerland (www.intracen.org). 11 Table 2.5: Poppy Cultivation in Afghanistan, 2003 (NRVA Survey Data) Poppy No. of Poppy area Total land /total land Location and type farmers % of Poppy area cultivated cultivated cultivated Poor of farmer (`000) Households ('000 ha's) (ha/HH) (ha/HH) (%) HHs (%) All Afghanistan Poppy farmer 119 5 46.51 0.39 1.75 22 34 Non-poppy farmer 2054 95 - 0 0.88 0 52 Total 2173 100 46.51 0.30 0.92 2 51 Northern Poppy farmer 58 8 22.79 0.39 2.34 17 20 Non-poppy farmer 682 92 - 0 1.70 0 52 Total 741 100 22.79 0.38 1.75 2 49 West/South West Poppy farmer 12 3 2.94 0.24 0.55 44 23 Non-poppy farmer 457 97 - 0 0.55 0 52 Total 467 100 2.94 0.13 0.55 1 51 Mountains Poppy farmer 49 5 20.77 0.43 1.33 32 53 Non-poppy farmer 915 95 - 0 0.42 0 51 Total 964 100 20.77 0.29 0.47 5 52 Source: Calculated from NRVA (2003) data. Poppy cultivation has relatively little direct effect on total food crop cultivation at the national level. According to the UNODC, poppy area cultivated was about 80,000 hectares in 2002/03 and 131,000 hectares in 2003/04, only 3.5 ­5.0 percent that of wheat, and only about 1 percent of total arable land. Although opium is a very labor-intensive crop (350 days/hectare), particularly in harvesting, increased poppy cultivation does not appear to have significantly lessened labor availability for wheat. According to NRVA data (which may understate poppy area cultivated given the crop's illegal status), among poppy farmers, average poppy area is only 22 percent of total area cultivated. Prices of opium fluctuate enormously, however, and have fallen steeply in the last two years from $350/kg in 2001/02 to $283/kg to only $92/kg in 2003/04. Yields of poppy also fell by about 30 percent in 2004, so that gross value of output per hectare fell from $12,700/ha in 2003 to $2900 in 2004. Using these opium yield estimates, gross revenues of poppy were 7.5 times those of wheat in 2004 (and 27 times those of wheat in 2003). Assuming lower opium yields equal to about one- third those of UNODC, gross revenues of poppy were 8.7 times those of wheat in 2003 (Molla, 2003) and about 2.3 times those of wheat in 2004 (Table 2.6). 12 Table 2.6: Gross Revenues from Opium and Wheat Cultivation Opium Opium Opium Wheat Opium / Yield Price Revenue Revenue Wheat (kg/ha) ($/ton) ($/ha) ($/ha) Revenue 2003 UNODC Survey 45 283 12735 470 27.1 2004 UNODC Survey 32 92 2944 390 7.5 Low Opium Yield Est. 2003a 14 283 3962 456 8.7 Low Opium Yield Est. 2004b 10 92 916 390 2.3 aMolla (2003). bOpium yield calculated using percentage change from UNODC and Molla (2003) figure for 2003 opium yield. Because of the high value of poppy, it can be a major source of income for households that cultivate the crop, though some analysis suggests that after deducting for costs of hired labor, income earned from poppy for some farmers is not large (Mansfield, 2002). Nationally, one-third of poppy formers were poor according to NRVA (2003) data, though on average, consumption per capita of poppy farmers was 25 percent higher than that of other farmers. Nonetheless, cultivation of poppy involves substantial risks, including possible eradication of the crop by authorities. Moreover, involvement of drug traders in local society often results in threats, intimidation and violence against farmers and other community members. Livestock Livestock is a major source of food and income for Afghan farmers, traditionally being an integral part of most farming systems in the country. Range pasture land covering about 45 percent of total land area has traditionally supported a large livestock population, with animals being the only source of income for Table 2.7: Livestock Population per Family some groups such as the Kuchi 1995 1998 2003 nomads. Livestock population declined in the 1980s but had Cattle 3.7 2.5 1.2 recovered to pre-war levels by the Sheep 21.9 14.2 2.9 mid-1990s. However, the years of Goats 9.4 5.8 2.4 drought from 1999-2001 have had a Donkeys 1.1 0.7 0.5 serious adverse impact. As per a Camels 0.4 0.2 0.1 Poultry 11.6 6.8 5.0 2003 census the estimated livestock Source: FAO, National Livestock Census, Interim Report, 2003. population is about 34 million (including 12 million of poultry), which is a sizeable decline from estimates in the 1990s. By 2003 the number of cattle and poultry per family had fallen to one-third of the 1995 figure, with numbers of sheep and goats per family suffering an even steeper decline (Table 2.7). NRVA 2003 data presented in Table 2.8 show that livestock ownership is highest among large land owners, though these data apparently miss the nomadic Kuchi - about 10 percent of the total rural population. Average livestock ownership is highest in the Northern regions (8.4 sheep and 3.2 cattle per household) and lowest in the West/SW region (1.4 sheep and 0.7 cattle per household). 13 Table 2.8: Afghanistan Livestock Ownership, 2003 House- % Total Cattle Goats Sheep Camels Donkeys Poultry holds Populatio Owned Owned Owned Owned Owned Owned (mns) n per HH per HH per HH per HH per HH per HH Large Land Owners 0.21 10.0 6.96 6.08 20.57 0.12 2.54 14.35 Small Farmers/Tenants 1.64 75.8 1.43 1.95 2.31 0.01 0.69 6.38 Non-Crop Livestock Owners 0.22 9.5 0.53 2.44 3.59 0.11 0.65 5.42 Rural Non-Farm 0.11 4.6 0.00 0.00 0.00 0.00 0.00 7.46 Total Livestock Owners 2.17 100.0 1.80 2.30 4.08 0.03 0.83 7.10 Northern 0.74 32.7 3.23 3.49 8.40 0.05 1.39 8.68 West/SW 0.47 21.3 0.72 1.74 1.44 0.01 0.52 5.70 Mountains 0.96 46.1 1.22 1.67 2.03 0.03 0.54 6.53 Source: Calculated from NRVA 2003 data. Constraints on Agricultural Production19 Growth of agricultural production faces several major constraints, including lack of assured and timely irrigation, inadequate knowledge of improved production and post-harvest technologies, lack of access to credit, and insufficient access to markets. Irrigation Lack of assured and timely water supply at the farm level through the existing irrigation system is perhaps the most important constraint to agricultural growth. Irrigation infrastructure, as well as many supporting community based and government institutions, have seriously deteriorated or broken down due to years of war, strife and drought. Addressing irrigation related issues is high on the government's agenda, as is evident from current programs of rehabilitation of irrigation infrastructure by MIWRE using external assistance. Substantial scope exists for improving the quality (timely delivery) and expanding the quantity (area under assured irrigation) of irrigation, thereby reducing vulnerability to droughts and improving water supply in Afghanistan. Current best estimates of the irrigation potential are that 3.54 million hectares could be developed through rehabilitation and development of new infrastructure such as storage (SAF, 2004). Of this total 1.035 million hectares have never been irrigated but could be brought to intensive irrigation through development of new irrigation schemes. Of the remaining 2.4 million hectares, projects currently funded by external assistance (as of mid-2004) provide resources to rehabilitate infrastructure serving about 0.43 to 0.55 million hectares. Another 0.5 million hectares, mainly for areas with intermittent irrigation and land that has never been irrigated is planned to be developed by 2015 under the National Long Term Irrigation and Power Program. Further expansion will take place as absorptive capacity grows and management and technical skills develop. Recognizing the centrality of water, the Government of Afghanistan plans to invest about US$2.35 billion between 2004 and 2015 in four sub-programs related to water resources and irrigated agriculture. Three major institutional and policy reforms are also being implemented: adoption of a river basin approach to manage water resources; restructuring of MIWRE and privatization/ liquidation 19For further details, see World Bank 2004 (Afghanistan Agricultural Policy Note). 14 of parastatals involved in irrigation construction and maintenance; and adoption of a participatory approach involving water user groups in the rehabilitation, as well as O&M of all irrigation schemes in the future. More attention should also be paid to groundwater resources. Groundwater tables should be regularly monitored to curb indiscriminate exploitation of groundwater resources, especially in vulnerable areas. New regulations to curb groundwater extraction may not be enforceable in the prevailing uncertain security situation. Experience suggests that community participation and peer pressure can often be more effective in managing groundwater resources than policing by government staff. Efforts to integrate watershed development with irrigation improvement are also needed, requiring closer technical and financial analysis of watershed development options. Agricultural Research and Extension Productivity growth in agriculture is based largely on application of science, technology and information. Studies consistently show high returns to investments in agricultural research in developing countries, averaging over 40 percent.20 Research on comparative returns to public investments in R&D, irrigation, roads, education and electricity in China and India showed that public investment in India and the People's Republic of China showed that R&D contributed the most to promoting agricultural growth in both countries (Fan, Zhang and Zhang, 2002). These results suggest that while complementary investments in infrastructure are necessary, investment in agricultural research and extension is a key element in enhancing a country's competitive advantage by reducing production costs, improving product quality, and generally increasing efficiency along the commodity chain. Prior to the conflict, Afghanistan had a substantial agricultural research system which, at its peak, carried out its functions through 24 research stations and over 1000 staff, of which 25 percent were technical research staff. This system is now largely dysfunctional as a result of widespread infrastructure destruction and loss of skilled staff. The agricultural extension system, which used to operate through about 400 extension units spread across the nation, is in a similar state of disrepair. Capacity building efforts need to focus on updating technical skills but also developing new skills related to management, monitoring and evaluation, participatory approaches, and marketing. In order to revamp agricultural research and extension in Afghanistan, a national research strategy and a national extension strategy are needed. This new strategy should include a focus on adaptive research rather than basic research, greater participation by farmers in defining research priorities, and more emphasis on post-harvest issues. Private service delivery of extension services, which has been shown to be a more efficient way to serve clients in various programs in Chile, Bangladesh and India (World Bank, 2004b), may also be explored. Decentralization of extension services can facilitate user participation and a more responsive system. Sustainable financing mechanisms will also need to be developed. Rural Credit Rapid farmer adoption of improved technologies requires a functional rural financial system. In view of the ineffectiveness of directed lending to agriculture through publicly owned institutions, the government has decided to dissolve the Agricultural Development Bank. Currently, traditional sources (moneylenders, traders, family, friends, etc.) are the main source of 20Alston et. al., (1998). 15 agricultural credit, together with various NGO led microfinance initiatives. The Micro-Finance Support Facility of Afghanistan (MISFA) begun in 2003 is designed to be a primary apex facility for meeting rural credit needs. Administrative costs in the first year of operation were high, however, in part because of expenditures on establishing new facilities. Establishing a viable rural financial system in Afghanistan is a medium term rather than short- term process, with financing for agriculture recognized as part of the wider rural finance market. Possible steps to building a viable rural financial system include:21 · In the short term, establish an effective system of NGO and microfinance registration and monitoring, facilitate entry and expand outreach. · In the medium term, adopt new NGO legislation that encompasses microfinance activities, and other savings and credit associations; entrust larger, well managed institutions to expand their range of products, including deposit taking among members. · In the long run, consider transformation of some of the larger microfinance NGOs into banks or non-bank financial institutions. 21World Bank, 2004b. 16 III. INTERNATIONAL TRADE, DOMESTIC MARKETS AND AVAILABILITY OF FOOD Food markets play a crucial role in food security, linking domestic producers and international suppliers to consumers. Available evidence suggests that these markets have performed fairly well in Afghanistan under very difficult conditions of war, internal conflict and disruption of infrastructure. Nonetheless, transport and other marketing costs remain high, and improved efficiency of food markets can potentially benefit both producers and consumers. This chapter begins with a discussion of international trade, including both commercial trade and food aid imports, which together with domestic production determine total availability of food in Afghanistan. The chapter then covers the structure and performance of domestic markets, focusing on the size of the marketed surplus, regional flows, the wheat milling sector and marketing costs. Finally, specific policy options are analyzed, including food aid, public sector procurement and distribution, and emergency food stocks for price stabilization. As in the chapter on food production, much of the focus of this chapter is on wheat, the major food commodity. International Trade and Total Availability of Foodgrains Wheat imports (both commercial imports and food aid) are a major source of wheat supply and total food availability in Afghanistan, averaging 780 thousand tons annually from 1999/2000 to 2003/04, (24 percent of total wheat supply over the period) (Table 3.1). About one-third of these wheat imports (252 thousand tons per year) have been through food aid, mainly in the form of wheat flour. Following the record 2003 harvest, however, food aid flows were still about 250 thousand tons (down from 360 thousand tons in 2002/03), out of an estimated total imports of 550 thousand tons. Table 3.1: Afghanistan: Wheat Availability, 2000 - 2004 Average 2002-04 2000 2001 2002 2003 2004 ('000 tons) (share) Availability 2357 2817 3546 4912 4005 4154 100% Production 1469 1597 2686 4362 2293 3114 75% Imports 888 1220 860 550 1712 1041 25% Commercial 650 908 500 300 1300 700 17% Food Aid 239 312 360 250 412 341 8% Note: Years indicate calendar year of main harvest (April/May). Data on imports and availability are for the following July/June fiscal year. For example, the figures for 2004 indicate production in April/May 2004 and availability for 2004/05. Data limitations mean that these estimates may carry significant margins of error. 2004 data are preliminary. Source: FAO/WFP Kabul. There is considerable uncertainty surrounding these estimates of production and imports, however. Wheat production estimates for 2003 vary by 1.63 million tons, with NRVA estimate (2.73 million tons) 38 percent less than the FAO crop assessment figure (4.36 million tons), (see Box 3.1). Uncertainties regarding the volume of import trade are smaller in absolute terms, but there is no reliable independent customs data on Afghanistan's wheat imports. 17 According to FAO data (based on exports of other countries to Afghanistan), private commercial imports of wheat in 2002/03 equaled 500 thousand tons, and total commercial imports of cereals, (including 151 thousand tons of rice from Pakistan), were 651 thousand tons.22 By comparison, total food aid imports were only 394 thousand tons, (360 thousand tons of wheat). Thus food aid accounted for only 42 percent of total wheat imports and 38 percent of total grain reports. Box 3.1: Estimates of Total Availability of Wheat There is substantial uncertainty regarding the levels of total wheat supply (domestic production plus imports) and domestic consumption per capita, in large part because of uncertainties in the underlying data on production and population.1 Using the official FAO production estimates for 2003 of 4.36 million tons and CSO population estimates, per capita availability of wheat was 205 kilograms/person. By comparison, per capita consumption is 131 kgs/person in Pakistan and 170 kgs/person in Iran. Moreover, NRVA one-week consumption recall data extrapolated over a one year period would imply annual consumption of 163 kilograms/capita for rural households (which account for about 70 percent of the population). Since the survey was conducted in the months just following the record harvest of mid-2003, it is likely that annual rural wheat consumption is in fact lower. However, NRVA production estimates of 2.73 million tons and CSO population estimates imply a national (both urban and rural) per capita consumption of only 140 kgs/person. In addition, NRVA- based production estimates for 2002/03 of 2.44 million tons (only 12 percent less than in 2003/04), would imply an increase in per capita consumption of only 3 percent between the two years, even assuming that population increased only by only 2 percent. Such a small increase in per capita consumption would appear to be inconsistent with an 11 percent increase in incomes (as measured by total GDP), a positive income elasticity of demand and a decline in real prices in major markets. 2 Thus, production in 2002 and 2003 likely lies in between the NRVA and FAO estimates. Using an average of the estimates for 2003 production (3.55 million tons) and assuming that total population increased by 1.0 million between 2002/03 and 2003/04 (consistent with estimates that Kabul's population alone grew by 1.0 million refugees between the start of 2002 and mid-2004, World Bank 2005), results in a per capita availability of 168 kilograms/person/year. Assuming a 25 percent increase in production between 2002 and 2003, (slightly more than the estimated 23 percent increase in production implied if one assumes that irrigated wheat yields did not change between these two years ­ see Annex 1), per capita consumption increased by about 9 percent between 2002/03 and 2003/04. 1Trade figures are also uncertain since accurate Afghanistan customs data does not exist. Import estimates through 2002/03 are based largely on export data of neighboring countries. 2Given an estimated 10 percent decline in real prices, and assuming, conservatively, an income elasticity of demand of 0.5 for wheat products and an own-price elasticity of demand of ­0.5, per capita demand for wheat would rise by 10.6 percent. Likewise, over the 2000/01 ­ 2002/03 periods, food aid accounted for only 31 percent of total wheat imports and 26 percent of total grain imports (Table 3.2). Although in 2003/04, food aid wheat declined somewhat to 250 thousand tons, commercial imports also declined to 300 thousand tons, and the share of food aid in total imports rose was 45 percent, essentially the same as in 2002/03. 22 Earlier WFP estimates placed commercial wheat imports in 2002/03 at 911 thousand tons, (18.3% of estimated total availability). (Neun and Fitzherbert (2003);, vol 1 p. 52 table). 18 Table 3.2: Afghanistan: Grain Imports, 1999/00 - 2003/04 Average 1999/00 2000/01 2001/02 2002/03 2003/04 2000/01-03 Commercial Imports Wheat and wheat floura 290 650 908 500 300 686 Kazakhstan 283 333 173 191 n.a. 233 Pakistan 0 300 716 204 n.a. 407 Other 6 17 19 105 n.a. 47 Rice 206 297 247 151 n.a. 232 Pakistan 184 297 179 151 n.a. 209 Other 22 0 68 0 n.a. 23 Other Cereals 62 19 14 0 n.a. 11 Total 558 966 1169 651 n.a. 929 Food Aid Food aid wheat and floura 102 239 312 360 250 304 Food aid total grain 109 246 337 394 n.a. 326 Food aid/Total Imports Wheat and wheat floura 26% 27% 26% 42% 45% 31% Total grain 16% 20% 22% 38% n.a. 26% Note: Wheat flour quantities are converted to wheat equivalents. n.a. denotes data not available. Source: FAO and World Food Programme /Kabul. Pakistan and Kazakhstan are the major source of commercially imported wheat, mainly in the form of wheat flour. In recent years, Pakistan has typically exported around 600 thousand tons of wheat flour to Afghanistan through formal and informal channels. These exports have been officially encouraged in years of good harvests. In spite of movement restrictions on wheat in Pakistan's Punjab province to facilitate government wheat procurement, wheat flour exports continued in 2004. Kazakhstan has the largest wheat surplus in the region (Table 3.3), and exports to Russia, Iran and the world market, as well as to Afghanistan. Wheat from Kazakhstan, which has no common border with Afghanistan transits through Uzbekistan, Tajikistan or even Iran. Iran is the largest wheat importer in the region, with imports of 6 million tons, almost ten times the magnitude of Afghanistan's wheat imports (Annex Table A4.1). 19 Table 3.3: Wheat Production and Consumption in South and Central Asia, 2001 Population Per capita Production Net Imports Per capita Income* Cons. (mns) (US$) (mn tons) (mn tons) (kgs) Afghanistan* 22.6 300 3.5 0.70 172.1 Pakistan 145.0 440 19.0 -0.69 131.1 Iran 71.4 1680 9.5 6.36 169.9 Kazakhstan 16.1 1260 12.7 -3.21 125.8 Kyrgyzstan 5.0 270 1.2 0.07 210.0 Tajikistan 6.1 180 0.4 0.35 121.3 Turkmenistan 4.8 750 1.8 0.28 211.4 Uzbekistan 25.3 360 3.8 0.53 155.1 Regional Ave 273.7 801 48.3 3.7 145.8 Notes: Rice in milled equivalents; Afghanistan per capita income is 2002 GDP/capita including opium. Official GDP/capita (without opium) for 2002 is $186/person. Other country per capita income figures are Gross National Income/capita for 2000; Afghanistan production data is for the 2003 harvest using best estimates (see Annex 1.) Afghanistan calories per capita from NRVA (2003) rural average; note that wheat consumption in NRVA is 162.6 kgs/person/year. Sources: FAO Food Balance Sheets (FAOSTAT) and NRVA (2003); World Bank (2004a) and World Bank World Development Indicators. Smaller quantities of rice (150 to 300 thousand tons per year in recent years) are also imported, almost all from Pakistan. With rice production equal to 291 thousand tons (milled rice equivalent) in 2003 (FAO data), rice imports accounted for one third to one half of total rice supply, and 4-7 percent of total rice and wheat availability (assuming wheat availability of about 3.6 million tons). Domestic Food Markets NRVA data indicate that total marketed wheat in 2003 was 683 thousand tons, equal to 25 percent of production (2.73 million tons) and 22 percent of total consumption (3.1 million tons). The northern region of the country accounts for about 90 percent of total marketed grain, (620 thousand tons) in 2003. Only minimal quantities of domestic wheat production were sold in the West/Southwest (48 thousand tons) and Mountain regions (17 thousand tons) of the country.23 Comparisons of wheat production estimates with an estimated consumption standard of 150 kgs/ person/ year suggest broadly similar patterns of marketable surplus by region. Only four 23The NRVA data are consistent with survey findings reported in Grace and Paine (2004) showing that most sample households had no or little marketed surpluses of wheat. 20 provinces, all located in northeast Afghanistan, have sizeable surpluses of production over consumption: Kunduz (542 thousand tons), Takhar (179 thousand tons), Baghlan (192 thousand tons) and Badakshan (72 thousand tons), (Annex Tables A4.2 and A4.3).24 Regional differences in agro-ecological conditions, economic incentives for production, concentration of land ownership, and population densities largely account for the above differences in estimated marketed surplus. In addition, mountainous terrain, poor road infrastructure and the tenuous security situation, all contribute to high marketing costs and generally weak market links across various regions of Afghanistan. In fact, several regional markets for wheat appear to be better integrated with markets in bordering countries than with other markets in Afghanistan. In broad terms, available evidence suggests that there are five wheat marketing regions of Afghanistan (Table 3.4 and Map 3.1). The northern region, with its major market in Mazar-e- Sharif, is surplus in wheat, supplying wheat to markets in the central region (in particular, Kabul) and the West Region (Herat). Significant amounts of wheat are imported from Kazakhstan via Uzbekistan and Mazar-e-Sharif (and in past years from Kazakhstan via Tajikistan and Kunduz25) into the northern region. The central/east region, including Kabul and Jalalabad, is deficit in wheat, and is supplied both with imported wheat flour from Pakistan (via Peshawar) and the northern region of Afghanistan (via the Salang tunnel). The southern region is also a deficit region, with strong international links to the Pakistan wheat flour market via Quetta, Pakistan and the Spin Boldak/Chaman border to Kandahar. Given the high risks of transport between the Kabul and Kandahar markets, however, there has been little wheat trade between the southern and central/eastern markets. Relatively little solid empirical information is available about markets in the western region, which includes Herat. Anecdotal evidence suggests there are links with domestic markets in the north (Mazar-e-Sharif) and the south (Kandahar). Imports from Iran (perhaps originating from Kazakhstan) may also supply this market. Finally, markets in the Central-West Highlands are perhaps the most isolated because of low population densities, and poor physical and institutional market infrastructure. Because of the high marketing costs and weak links to international markets, supply of wheat in this region is likely even more vulnerable to shocks than supply in other regions of Afghanistan. 24These estimates of total marketable surplus total to almost 1 million tons, about 300 thousand tons greater than the NRVA data on marketed quantities. This suggests that average consumption in the surplus districts may have exceeded 150 kgs/ person/ year and/or that NRVA data under-reports total marketed quantities. 25Neun and Fitzherbert (2003) and personal communications with Daniel Molla (WFP/Kabul) and Anthony Fitzherbert. 21 Table 3.4: Regional Wheat Markets in Afghanistan Domestic International Regional Market Major Cities Surplus/Deficit Market Links Market Links From Uzbekistan North Mazar-e Sharif Surplus To West and Central through Mazar From North and From Iran and West Heart Deficit South Pakistan From Pakistan Central/East Kabul, Jalalabad Deficit From North (Peshawar) Central-West Highland None Deficit Weak links Little or no trade South Kandahar Deficit To West From Pakistan (Quetta) Map 3.1: Afghanistan: Regional Wheat Markets 22 Wheat Milling Afghanistan has a severe shortage of wheat milling capacity. As of mid-2004, there were only eight medium- to large-scale industrial mills in the country (Annex Table A4.4). Five of these mills are publicly owned and located in the major cities of Kabul, Mazar-e-Sharif, Kandahar, Herat, and Pol-e-Khumri (Baghlan province). All five of these mills were built by the Soviet Union in the 1980s, but have suffered severe damage since then and only two (in Kabul and Mazar-e-Sharif) are operational in any form. The Kabul mill processes an average of 45 tons of second or third class wheat per day, and all of the output is sold to the Afghan national army. Baking facilities at the mill are leased out to a private firm periodically that manufactures sweetened breads and cake products. The Mazar facility provides services for the storage of grain, and periodically provides milling services. However, quantities milled are limited, and the mill is operated infrequently. In addition to the five publicly owned mills, there are three private medium-sized flour mills located in Kabul and Mazar-e-Sharif. These mills have significantly less storage capacity (2 to 6 thousand tons) than the public mills did when built (40 to 60 thousand tons). Kabul Flour Mills has been in operation since early 2003. It procures grain in the north of the country through its buying agents in Kunduz, Mazar-e-sharif and other cities), but is operating at only about 30-50 percent of milling capacity.26 The two privately owned mills in Mazar have a capacity of only 45 and 60 tons per day that are generally producing at capacity. All three private mills are owned by firms that also have operations in a number of other products. Apart from these three private and five public mills, all other milling capacity in the country consists of small scale, zero extraction rate `chakki' style mills. According to the July 2004 Mercy Corps survey of 61 mills in Kabul, Kunduz and Kandahar (Chabot, 2004), average mill capacity is between 1.6 and 2.4 tons per day. Mills in Kunduz (in the surplus northeast of the country) milled an average 13.7 tons/month (about 40 percent of capacity); mills in Kabul and Kandahar milled only about 8.0 tons/month (about 15 to 20 percent of capacity). With only three medium-sized mills operational on a regular basis, industrial milling capacity in Afghanistan is only a small fraction of total wheat flour processing needs. Assuming all three private modern mills are operated 300 days per year and produce 305 tons of flour per day, total capacity is 91,500 tons per year, about 3 percent of national consumption. The capacity of Kabul Flour Mills (60,000 tons) represents approximately 13 percent of estimated consumption for the city of Kabul.27 Actual quantity milled is less than half the capacity, about 5-6 percent of estimated consumption. Including the 137 registered chakki mills in Kabul and assuming 7.9 tons of grain milled per month (as reported in the Mercy Corps survey), total milling capacity in Kabul is about 73,000 tons, about 16 percent of estimated total annual consumption.28 Thus, there appears to be substantial scope for investment in medium sized industrial milling facilities in Kabul, as well as the rest of the country. 26The mill is processing between 50-75 tons per day during the months of May through July, and 80-100 tons per day during the winter months of October-February. 27Annual wheat flour consumption for Kabul city is estimated at 450,000 tons -- 150 kgs/person x 3.0 million people. 28These estimates are consistent with results from the Mercy Corps trader survey showing 80 to 90 percent of wheat flour sold in Kabul is imported. 23 Production Shocks and Market Price Fluctuations In spite of large fluctuations in domestic production caused by periodic droughts, domestic grain prices in Afghanistan have been remarkably stable. A major reason for this relative stability in prices is imported grain flows that add to total availability of grain. Although food aid has been a significant source of import supply, private sector imports appear to have been the major stabilizing factor in Afghanistan's markets. The limited survey evidence on the private sector import trade suggests that it is highly competitive and very responsive to market forces in Afghanistan wheat markets. Informal interviews indicate that private traders/importers typically hold small stocks, and react quickly to perspective demand for wheat flour by increasing imports.29 Comparisons of retail prices in Kabul with wholesale prices in Peshawar (the nearest major wheat market in Pakistan) 30 and Lahore (a major wholesale wheat market) also suggests that the price of private sector imports (the import parity price) was a major determinant of the price of wheat in Kabul from 2001 through mid-2003 (Figure 3.1). For example, in 2002, the average import parity price of wheat from Lahore was $158/ton ($138/ton plus an estimated $20/ton transport and marketing margins), compared with an average retail price in Kabul $152/ton.31 The correlation coefficient between this estimated import parity price and the Kabul retail price for January-December 2002 was 0.93,32 suggesting that market prices in Kabul were effectively determined by the cost of imported wheat and wheat flour from Pakistan. Over a longer period (January 2002 through March 2004), the correlation coefficient of the ex: Lahore import parity price with the Kabul retail wheat price is only 0.47, however. Correlation coefficients between Kandahar and Lahore and between Jalalabad and Lahore33 are somewhat higher, 0.88 and 0.66, respectively. Correlation coefficients of average monthly prices of wheat (expressed in US$/ton) across most cities in Afghanistan suggest even lower degrees of market integration (Table 3.5). The only pair of domestic markets with a correlation coefficient greater than 0.6 is Mazar and Faisabad in the wheat surplus northern regions of Afghanistan.34 29 Neun and Fitzherbert (2003), p. 58. 30Transport time by road between Kabul and Peshawar is approximately 8-10 hours. Though it is generally unsafe for cars to travel this road after dark, truck traffic continues throughout the night. 31A better measure of integration across markets would be comparisons of wholesale prices; unfortunately, no data on wholesale prices in Kabul exists. The marketing margin used here approximates transport costs, differences in quality between Pakistan wheat/flour and the average quality in Kabul markets, and the margin between wholesale and retail prices. 32By comparison, correlation coefficients in the well-integrated North American wheat markets for US Soft Red Winter, US Hard Red Winter and Canadian wheat prices over the January 2002 through March 2004 ranged from 0.88 (US Soft Red Winter and Canadian Wheat) to 0.94 (U.S. Hard Red Winter and Canadian Wheat). 33Prices for Lahore are import parity prices calculated as the price in Lahore plus a fixed dollar marketing and transport margin to Kabul. 34Preliminary regressions of prices of wheat in various provincial markets and the price of wheat in Kabul of the form: Pprovincial = 0 + 1 PKabul , produced statistically insignificant coefficients for 1, however. 24 Figure 3.1: Wheat and Wheat Flour Prices in Afghanistan, 2000-2004 400 350 300 ) 250 ton e($/c 200 Pri 150 100 50 0 00-n 00l- 01-n 10l- 02-n 20l- 03-n 30l- 04-n Ja Apr-00 Ju Oct-00 Ja Apr-01 Ju Oct-01 Ja Apr-02 Ju Oct-02 Ja Apr-03 Ju Oct-03 Ja Apr-04 Kabul wheat flour Import Par ex: Lahore Kabul retail Table 3.5: Correlation Coefficients for Wheat Prices, January 2002 to March 2004 Kabul Kandahar Jalalabad Herat Mazar Faisabad Lahorea Kabul 1.00 Kandahar 0.42 1.00 Jalalabad 0.30 0.49 1.00 Heart 0.34 0.29 -0.05 1.00 Mazar 0.24 0.05 -0.30 0.51 1.00 Faisabad 0.22 -0.13 -0.57 0.54 0.79 1.00 Lahorea 0.47 0.88 0.66 0.34 -0.02 -0.24 1.00 a Data for Lahore, Pakistan are import parity prices calculated as the price in Lahore plus a marketing and transport margin to Kabul, estimated at $20/ton. Source: Calculated from price data from WFP/Kabul and the World Bank. Several factors may account for these rather low correlations between wheat prices in Kabul, Kandahar and Jalabad with estimated import parity prices in Pakistan. Because most of the trade in wheat is in the form of flour, not grain, changes in the markups between grain and flour in both Afghanistan and neighboring countries affect the correlation in wheat grain prices. Moreover, available data on wheat flour prices do not specify qualities of wheat flour; comparisons of imported and domestic flour prices are likewise problematic.35 Finally, the margins between prices of imported and domestic flour are not constant, but change according to relative availability of the two types of flour in the market. The very low correlation coefficients between 35In Kabul markets in mid-2004, prices of Pakistan flour commanded about a 10 percent premium over Uzbek 1st class flour; in Mazar markets, domestic 1st class flour was priced at about a 10 percent premium over domestic 3rd class flour. 25 many pairs of domestic markets, however, most likely reflect lack of trade between these markets. Nonetheless, the relative stability of domestic wheat prices (at prices approximating import parity prices) in the face of extremely large fluctuations in production suggests that private trade has played an important role in stabilizing prices and enhancing food security in Afghanistan, as it has in other countries (see Box 3.2). Box 3.2: Price Stabilization, Trade and Food Security: The Bangladesh Experience In an effort to stabilize domestic market prices and the availability of food, many countries have intervened heavily in food markets, buying and selling grain, and accumulating large national stocks as emergency reserves. For example, for most of the last forty years, India has relied on procurement of grain at fixed support prices, movement restrictions, limits on private stocks, large national food stocks, and subsidized sales of food through ration shops to stabilize prices. A policy of market development and trade liberalization is an alternative, potentially less- costly and more effective option, however, as evidenced by the experience of Bangladesh following massive floods in 1998. In mid-1998, as major floods spread across much of Bangladesh and rice prices rose to import parity levels (the export price of rice in the exporting country plus transport and normal marketing costs), inducing massive inflows of rice by hundreds of small traders. Private sector rice imports, 2.42 million tons in this period according to official estimates, were 6.1 times larger than government rice distribution. During the 1998 calendar year alone, private sector imports, mainly from India, reached 2.26 million tons. Had the government of Bangladesh imported this grain itself, the average additional cost of the imported rice delivered to local delivery points would have been approximately 50 to 100 million dollars. And, if the government had subsidized this rice by selling it at the Open Market Sales price used for very limited government sales in urban centers, the total fiscal cost would have been 160 to 210 million dollars The liberalized trade policy also helped the Government of Bangladesh to achieve price stabilization without large government stocks. In 1988/89, at the time of another major flood that caused an 18.1 percent reduction in the monsoon season rice harvest, (relative to trend), public stocks averaged 1.167 million tons from August to November, (10.9 kilograms/person). These stocks combined with public sector imports enabled the government to use public distribution channels and stabilize markets and reach flood-affected households. Average public stocks ten years later in August-November 1998, however, were only half the per capita levels of ten years earlier (5.5 kilograms/person). Nonetheless, private sector imports rather than public distribution provided the main source of additional market supplies to compensate for the 2.04 million ton crop loss. Other factors were also important in enabling Bangladesh to stabilize food supplies. The large expansion in the winter season boro rice and wheat harvests over more than two decades helped reduce the relative importance of the monsoon (aman) rice crop in total domestic production, and minimize the length of time between major domestic grain harvests. Moreover, large-scale import flows would not have been possible without development of private domestic markets through investment in market infrastructure (particularly, roads and bridges) and pro- market policies including no limits on private stocks or movement restrictions, sufficient access to foreign exchange, made possible in part through expansion of Bangladesh exports, and availability of grain in world markets. Source: Dorosh (2001, 2004). 26 Food Aid Food aid inflows also add to domestic supplies, but because their levels are determined by administrative procedures and not by market forces, they have the potential to reduce market prices below international border price levels. In this case, food aid reduces producer prices below the levels that would prevail in markets in the absence of food aid, reducing farmer incomes and incentives for production, both directly (through reduced market prices) and indirectly (through reduced public and private incentives for investment in agricultural technology). However, there is no evidence of major disincentive effects of food aid on wheat markets up to the bumper wheat crop in 2003, and the continuance of private sector imports during that year suggests that marketing or milling constraints, rather than food aid were most responsible for localized decline in producer prices. During 2001/02 and 2002/03, (major drought years), substantial private sector imports took place along with food aid distribution, and retail market prices for wheat in Kabul mirrored import parity prices. In spite of large amounts of food aid, there was no evidence of major price disincentive effects of food aid 2001/02 and 2002/03. Prices of wheat and wheat flour were at or near import parity levels in this period, and substantial private sector imports took place along with food aid distribution. After the record wheat harvest in mid-2003, however, domestic wheat grain prices may have fallen below import parity levels; wheat prices in Kabul fell substantially below import parity. Retail wheat prices in Kabul averaged about $165/ton from September 2003 to May 2004, while Pakistan import parity prices rose to about $195/ton (18 percent higher) in the same period.36 Afghanistan's imports of wheat flour from Pakistan continued (albeit at a slower pace than in earlier years), in part because higher income urban households demand imported wheat flour, which is of higher quality than domestic wheat flour. (The shortage of milling capacity in Afghanistan to produce higher quality wheat flour is also a factor.) With domestic prices below import parity prices, food aid added approximately 300 thousand tons to wheat availability (equivalent to about 8 percent of total wheat consumption in 2003/04 using the "best estimate"). Depending on the price responsiveness of consumer wheat demand, this increase in supply may have reduced market prices by perhaps 4 to 8 percent, in comparison with a program of cash transfers of the same value to food aid recipients.37 Under the NRVA- and FAO-based estimates of availability, the effects of food aid inflows range from 3 to 9 percent. While these adverse effects on food aid on market prices are relatively small, food aid inflows may also reduce public sector incentives to make investments in irrigation, research and extension that can raise domestic production and rural incomes (see Box 3.3). 36Pakistan's wheat prices in 2003/04 were higher than in 2002/03 largely because the Pakistan wheat harvest in mid-2003 (19.25 million tons) was affected by a drought and was 1.83 million tons less than the record 2000 harvest. After a failure to import wheat through public tenders, government stocks fell to extremely low levels prior to the 2004 wheat harvest, and the Government of Punjab placed movement restrictions on wheat after the harvest in 2004 to keep market prices low and to facilitate its own procurement. 37This calculation assumes an own-price elasticity of demand for wheat of between ­0.5 and ­1.0. Alderman (1988) estimated own-price elasticities of demand for wheat in Pakistan in the mid-1980s of ­ 0.96 for rural and ­0.34 in urban areas. 27 Box 3.3: Food Aid and Incentives for Domestic Production Maintaining food security at the national and household level is a major priority for most developing countries, both for the welfare of the poor as well as for political stability. In order to help assure food security, developing country governments have adopted various strategies including efforts to increase production (often with an explicit goal of food self-sufficiency), government intervention in markets, public distribution of food and maintenance of national food security stocks. Food aid, both for short-term emergency relief and program food aid that helps address medium-term food "deficits", is often a major component of these food security strategies. However, if not well managed, food aid can reduce domestic prices and producer incomes, as well the incentives for public and private investment in food production. Country experiences regarding food aid and domestic production have varied considerably. In South Asia, both India and Bangladesh have dramatically increased food production and reduced the role of food aid in national food supply and in programs to reduce food insecurity at the household level. In contrast, food aid has been a major share of total food supplies in Ethiopia for more than thirty years, and in spite of some gains in grain production since the mid-1990s, emergency food aid appeals are an almost annual occurrence. Following drought-induced production shortfalls and a temporary cutoff of U.S. food aid due to a political dispute in the mid-1960s, India adopted a national food self-sufficiency policy to increase domestic food production and eliminate the need for food aid inflows. Beginning in the late 1960s India rapidly increased its domestic food production through substantial public investments in irrigation and agricultural technology, and maintenance of producer incentives through large-scale government interventions in grain markets. By the mid-1970s, the country had achieved its wheat and rice self- sufficiency objectives. Only in the mid-1990s did major market liberalizations in agriculture occur, however, and the government still plays a dominant role in food markets in major producing states (World Bank, 2004). Bangladesh also rapidly increased its food production, using similar technologies to those adopted in India (expansion of irrigation and increased use of improved seeds and fertilizer), even as large-scale food aid inflows continued. Food aid inflows generally did not exceed the amount of wheat that would have been demanded in the absence of food aid and so producer prices were not adversely affected (Dorosh et. al., 2004). Bangladesh also adopted a more market-oriented approach than India with less government intervention, reaping its largest gains in production and food security after liberalizations of internal and external markets for food grain and agricultural inputs beginning in the late 1980s (Ahmed, Haggblade and Chowdhury, 2001; Ahmed, 2004). Ethiopia's gains in food production have been much more modest than those of India or Bangladesh, in part because the country has a much drier climate, with relatively far fewer opportunities for irrigation and multiple-cropping. War and insufficient investment in roads and market infrastructure have also limited the development of private markets, lessening market incentives for domestic production. Since 1994, economic reforms and increased adoption of improved seeds and fertilizer have enabled grain production to keep pace with population growth rate, but only after a dramatic decline in per capita grain production over the previous two decades (del Ninno, Dorosh and Subbarao, 2005). Policy Implications Because of the country's mountainous terrain, poor road infrastructure and tenuous security situation, internal marketing costs are very high. As a result, regional markets for wheat are generally better integrated with markets in bordering countries than with other markets in Afghanistan. Only in years of good harvests are there significant flows of wheat between regions 28 of the country, in particular from the northern region of the country to markets in the central region (in particular, Kabul) and the West Region (Herat). Development of efficient private markets is thus essential to translating food security at the national level (i.e., sufficient overall availability of food) to adequate availability at the local level. However, because of the country's mountainous terrain, poor road infrastructure and tenuous security situation, internal marketing costs are very high. Reducing these costs and increasing the efficiency of private markets will require both investments in domestic market infrastructure and a favorable environment for private sector trade and investment. Well-functioning markets facilitate the flow of food commodities from surplus to deficit areas, potentially benefiting both producers and consumers. This trade can be facilitated by a policy of free trade between and within provinces (no movement restrictions or taxes on domestic trade flows, and no limits on the location, quantity or prices of food purchases or sales). Private sector investment in storage and food processing (e.g., wheat milling) can be encouraged by placing no limits on privately held stocks or the size of milling or other food processing operations and minimizing licensing requirements and avoiding imposition of price controls (e.g., administratively set price ceilings on bread). The experience of many countries indicates that public sector procurement, storage and distribution of food can involve large fiscal subsidies. Minimizing direct government interventions in food markets will encourage private sector and investment. If the government does intervene in food markets, it is important to maintain private sector incentives for trade and storage through a transparent and consistent set of rules for government purchases and sales, and a sufficient margin between public sector procurement and sales prices that is in excess of normal private sector storage and processing costs. 29 30 IV. REDUCING POVERTY AND INCREASING HOUSEHOLD FOOD SECURITY38 The ability to access food by different households is determined by a number of factors, including purchasing power, the local availability of food, and for individuals within the household, by socio-cultural relations. Own-production of food, wages (either in cash or in kind) from employment, or profits from own-enterprises are major sources of income. Household access to food can also be enhanced by government and NGO programs that supplement cash or food resources, as well as through private gifts, informal social support networks or borrowing (formal or informal). But even where household access to food is sufficient, utilization of food and nutritional outcomes may be inadequate, either because actual food intake (quantity, quality and diversity) is insufficient or because various health factors (water-borne diseases) inhibit the bodies' ability to utilize food. As discussed in this paper, total availability of food, particularly cereals, is not the only, or necessarily the most significant constraint on food consumption for the country as a whole. The NRVA data corroborate this: average per capita availability is about 2800 calories/day (Table 4.1), significantly above the 2100 calories per person estimated minimum daily requirement. Thus, increasing the ability of households to purchase more food rather than increasing availability of food appears to be the main pathway for reducing chronic food insecurity in Afghanistan. Table 4.1: Distribution of Calories across Poverty Groups, Rural Afghanistan 2003 Calories per Cost of Distribution of Person per Distribution of purchasing food Day Calories (%) calories Afs/p.a. expenditure (%) Quintiles Q1- Poorest 20% 1708 12 2292 9 Q2 2270 16 3580 13 Q3 2650 19 4762 18 Q4 3106 22 6237 23 Q5 - Top 20% 4263 31 10151 38 Overall 2800 100 5403 100 Source: NRVA 2003, World Bank 2005. Rural Poverty In Afghanistan, widespread poverty translates into chronic household food insecurity for millions of people who struggle to acquire sufficient food on a regular basis. Data from the NRVA suggest that 48 percent of the rural population in the sample of 11,200 rural households had food consumption expenditure (in the summer of 2003) that fell below the consumption poverty line, that is the cost of purchasing 2100 calories/day of the typical (to rural Afghanistan) basket of food items among the poorest households.39 Because of variation in diets and sharing of meals, around 38This chapter is largely based on World Bank (2005) Afghanistan Poverty, Vulnerability, and Social Protection: An Initial Assessment. 39An individual is classified as poor if his annual total food-consumption per capita is below the poverty line, defined as the annual cost of consuming 2100 calories/day of the typical basket of food items among the poorest. For further explanation of the methodology used to calculate food-consumption poverty, see World Bank (2005), Appendix 2. 31 22 percent of the population are found to actually be consuming less than 2100 calories/day. Along both dimensions there appear to be variation within the country (Table 4.2). Table 4.2: Rural Poverty by Region Percent of Percent of people falling Poverty Percent of people sample below the severity consuming less population per capita food expenditure (P2)* than 2100 cal/day threshold Rural Afghanistan 100 48 7 22 By Region North-east 11 33 5 13 South-east 12 39 4 15 Central 15 43 5 24 South 19 42 4 24 East 14 47 4 16 West 10 62 11 25 North 16 66 12 26 West-central 3 71 17 37 * A larger number suggest more "deep" poverty. Source: NRVA 2003 When asked about their ability to meet their food needs over the last 12 months prior to the survey (summer of 2003), about a third of the rural sample of households claimed they "usually" had problems meeting those needs, while another third "sometimes" had food security problems. As to be expected, poor households were more likely to have problems meeting food requirements compared to the non-poor or better off households. The factors underlying rural poverty in Afghanistan are complex, including insufficient assets (physical and financial), insecurity, recurrent droughts, poor infrastructure, low levels of education and skills, and inadequate public services. Access to land, especially irrigated land, is one important determinant of incomes and food consumption. Rural non-farm households on average consumed only 2,300 calories/person/day, 17 percent less than the national average (Table 4.3). Table 4.3: Afghanistan: Poverty and Food Consumption by Rural Household Groups, 2003 Households % Total Poor HHs Poor HHs Poor HHs Annual food Average Per capita (mns) population (mns) (as % of HH (% of expends/HH calories wheat cons. group) rural poor) (000 Afgh) per capita (kg) Farmers (cultivators) 1.84 85.8 0.87 47.4% 79.5 41.2 2,921 164.5 Large Land Owners 0.21 10.0 0.07 31.6% 6.0 54.3 3,481 175.0 Small Land Owners 1.46 68.1 0.69 47.3% 62.8 40.7 2,876 164.4 Tenants 0.18 7.7 0.12 66.8% 10.7 29.6 2,586 151.4 Landless w/ Livestock 0.22 9.5 0.14 63.1% 12.6 31.0 2,555 153.2 Rural Non-Farm 0.11 4.6 0.09 80.0% 7.9 24.2 2,311 147.1 Total 2.17 100 1.10 48 100 39.3 2,800 162.6 Source: NRVA 2003, World Bank 2005. Authors' calculations. 32 Households in rural Afghanistan are heavily dependent on wheat as a source of calories. Typically 60 percent of calories derive from wheat, with another 10 percent from other cereals and tubers (NRVA sample). When the data are analyzed by group, there is greater dietary diversity among the upper consumption quintiles, while poor households are less likely to meet needs for dietary diversity. In particular, more than 75 percent of the daily calories for the lowest quintile are derived from cereals and roots compared with about half for the highest quintile (Figure 4.1). Figure 4.1: Daily Caloric Distribution per Food Group, By Consumption Quintile 100% 80% 60% 40% 20% 0% Q1 Q2 Q3 Q4 Q5 Wheat Other cereal and roots Oil fat Pulses and nuts Sugar/honey Dairy Vegetables Meat, fish and eggs Fruits Source: NRVA 2003. Urban Poverty and Food Security The current urban population of Afghanistan is estimated to be around 6.4 million, which constitutes about 30 percent of the total population40. Owing to the influx of returning refugees in 2002 and thereafter, the population of Kabul alone has increased from 1.7 million in 2000 to over 2.5 million (possibly over 3 million41) in 2003.42 Little quantitative data is currently available on urban consumption patterns and poverty. In households surveyed recently in Kabul (ACF, 2004), for some 53 percent food clearly continues to be a primary concern (ranking third after income insecurity and tenure insecurity). When asked to give the main reasons for taking loans, food security was cited as the primary purpose, (followed by health expenditure). Household interviews confirmed that decreasing consumption of food (particularly eggs and meat) is a common phenomenon, as is the purchase of food on credit. Kitchen gardens provide an important component of food availability for many households, but tend to be less prevalent for poorer groups for whom overcrowded residential 40TISA (2004) Securing Afghanistan's Future. 41Central Statistical Office currently estimates 2,799,300. 42TISA (2004) Technical Annex Urban Development. 33 areas do not allow for such opportunities.43 An assessment of households in poorer districts of Kabul found that 13 percent owned livestock to supplement their diet. However, for many households in these areas a typical meal was reported to consist only of bread and tea. Eighty- five percent of women reported eating fruit less than once a week, 49 percent less than once a month. Ninety-seven percent respondents consume dairy products less than once a month. The daily expenditure on food for 62 percent families is between $1­2 /day, and for 36 percent families, less than $1 /day.44 Infant Mortality and Malnutrition Afghanistan has some of the worst indicators for child health and development anywhere in the world. The infant mortality rate in Afghanistan is 165 per 1000 live births, with one in four children dying before they reach their fifth birthday45 (around 300,000 annually). Around half of Afghan children suffer from chronic malnutrition (stunting). One of the key contributing factors to this crisis is the lack of access to safe drinking water for 87 percent of the population and the absence of adequate sanitation facilities for 99 percent of the population. The unavailability of detailed nationwide anthropometric information hinders an in depth and systematic analysis of malnutrition. Most malnutrition occurs during pregnancy and the first 2 years of life. Typically, there are a number of factors associated with malnutrition: (i) household level access to food; (ii) access to health services and facilities; and (iii) home care. All three factors are present in rural Afghanistan. Moreover, imbalance in the diet, particularly among poor households with inadequate caloric intake, often implies inadequate intake of essential micro-nutrients as well, which can lead to serious household problems. Results of the anthropometric surveys undertaken in Kabul from 1995 ­ 2003 (Annex Table A4.5) show high levels of chronic malnutrition (stunting) for children under five and significant seasonal variations in acute malnutrition (wasting). The results point to a marked improvement in nutritional status since 2002, although the absolute figures remain high. The fact that acute malnutrition peaks occurred during the summer of 2002 may be correlated to the marked increase in the incidence of diarrheal disease at this time of year. Surveys in Badghis Province46 suggests that the incidence of chronic malnutrition is similar to that in Kabul, while the incidence of acute malnutrition may be higher in Kabul than that among rural households. The nationwide 2003 Multiple Indicator Cluster Survey47 data show that, as a result of over-crowded unsanitary conditions, infant mortality rates are frequently higher among sections of the urban poor than for their rural counterparts.48 Reducing Poverty Sustained, equitable economic growth is the essential path for poverty reduction in Afghanistan. In the absence of growth, and the conditions necessary to facilitate it, the scope for poverty reduction is severely constrained. Sustained over some years, economic growth can increase opportunities for most Afghan people. Its impact will depend on how well equipped poor people are to participate in growth and to what extent growth helps the poor to build their assets ­ 43ACF (2004). 44CARITAS Germany (2003). 45Source, UNICEF, 2003. 46UNICEF / MoPH (2003.) Global chronic malnutrition 57.5% and global acute malnutrition 6.5%. 47UNICEF / CSO (2003). 48World Resources Institute (1997) The Urban Environment. 34 physical, financial, human and social. Having the ability to accumulate physical capital such as (irrigated) land, livestock and small businesses, having access to (micro) finance and business support services, as well as finding employment opportunities will be fundamental to the success of any poverty reduction program. Likewise, improvements in the welfare of poor people will require human capital accumulation ­ education, skills and good health ­ and the provision of infrastructure. Existing Social Protection Policies Social Protection is at the heart of poverty reduction, both providing people with the means to escape chronic poverty as well as cushion against income loss when shocks ­ such as drought ­ occur. Government interventions to improve food security are one component of its Livelihoods and Social Protection Program. This program has been developed to address the needs of poor and vulnerable people and ensure both their direct (self-reliance) and indirect (targeted public transfers) inclusion in national development. In the 2003/04 (1382) development budget, the Government sought to initiate a strategic shift beyond purely humanitarian relief to a more integrated and efficient programmatic approach able to provide sustained support to poor people so that they are better able to protect their assets and rebuild their livelihoods (see Annex Table A4.6). Four key ministries share cross-cutting mandates especially related to livelihood promotion and social protection. This is to be achieved both through sharpening the pro-poor focus of all sector programs, as well as through investment in additional targeted programs for vulnerable groups. The total investment budget for the Livelihoods and Social Protection Public Investment Program in 2003/04 (SY 1382) was US$233.5 million, about 13 percent of the total National Development Budget of around US$1.8 billion. In addition to the LSP PIP, a food aid program (300,000 tons of food) accounted for about US$150 million. The LSP consists of six sub-programs, each with its own set of priority interventions, including many not directly implemented by the Government (Annex Table A4.7).49 Two of these programs, the National Emergency Employment Program (NEEP) and the National Solidarity Program (NSP) are particularly important sources of public employment and resource transfer. NEEP is a labor-based approach to investments in rural infrastructure, which in FY2003/04 (Afghan year 1382), generated 8 million labour days of employment for the construction of rural infrastructure, including 5000 km of rural roads using mainly cash payments (see Annex Box A4.1.) The NSP provides block grants to villages, empowering communities to manage their own resources (see Annex Box A4.2.). The other programs are: · the Microfinance Support Facility of Afghanistan which promotes the establishment of sustainable financial service delivery to the rural poor; · the Business Development Services Program, which provides sustainable support to enterprises; · the National Water Supply and Sanitation Program, designed to enhance the health of rural communities; and 49This reflects the Government's commitment to focus its existing resources on policy development rather than involvement in direct service delivery, through entering into purchaser ­ provider relationships with relevant agencies. 35 · the National Disaster Risk Management Program, emergency response and interventions and institutional strengthening for risk management and disaster mitigation. Food Aid Supported Programs All of the programs in the Livelihoods and Social Protection area are designed to raise real incomes and improve the welfare of the poor. As such, they can potentially contribute to household food security by increasing access to food for poor households. None of these programs involve direct transfers of food, however. Alongside these programs is an array of programs utilizing food aid resources. In the absence of a functioning public administration and with frequents droughts and protracted political instability for several decades, food aid interventions in Afghanistan have played a crucial role in saving lives, protecting welfare and preserving capital assets. In recent years, however, with the re-emergence of a state presence, development of national and sub-national markets, improvement of climatic conditions following a three-year drought, and improved security in some parts of the country, the need for humanitarian food aid appears to be diminishing. Nonetheless, given weather-related fluctuations in production, and chronic food insecurity at the household related to inadequate access to food, food aid remains a potentially important instrument if properly targeted and utilized. Food programming within Afghanistan currently takes place within the framework of the Protracted Relief and Recovery Operation, implemented on behalf of the Government by WFP. The stated objective of the two-year PRRO that started on April 1, 2003, is to contribute to the protection and re-establishment of livelihoods and household food security by providing assistance to over 9.2 million beneficiaries over a two-year period. The total food requirements are 618,989 tons of commodities, valued at US$337.5 million. The main focus of the PRRO is on recovery activities such as food for work (FFW), food for education (FFE) and food for training (FFT). Relief activities constitute around one third of assistance and include support to returning refugees, internally displaced persons, the acutely malnourished, vulnerable households and social institutions. A mid-term evaluation of existing food aid programs undertaken during the second quarter of 2004 suggested cutting back several programs in accordance with medium-term government strategy to reduce levels of food aid. Small programs involving Institutional Feeding and Orphanages, TB Patient Feeding, and Urban Bakeries may be significantly reduced or eliminated, with an increased focus placed on programs with a greater development or training component, and where the benefits of food aid justify the costs of food aid handling and administration (see Annex 2). Targeting of Food and Other Safety Net Interventions Given the scarcity of financial and administrative resources in Afghanistan, it is important that policies and programs involving direct transfers are efficiently and effectively targeted. Data from the NRVA suggest that food for work (FFW) and cash for work (CFW) programs designed with the dual objectives of creating rural infrastructure and providing short-term employment do not efficiently target poor households. Relief food aid, administered mainly by the World Food Program, is relatively more efficiently targeted, but has much smaller coverage. 36 Box 4.1: Background to WFP PRRO in Afghanistan · With costs estimated at US$370 million50 at the time of launching, the two-year program (launched March 2003) was WFP's largest single country PRRO. This operation was designed to assist over nine million vulnerable people in rural and urban areas including returning refugees and Internally Displaced People, largely through FFW, FFT, school feeding programmes and returnee packages. · The overall goal of the operation is to contribute to re-establish and stabilize livelihoods and household food security in the context of Afghanistan's National Development Framework. · The current PRRO only began implementation in April 2003, but some of these activities have implemented under the EMOPs since in January 2000, or since the EMOPs were introduced that same year. · The PRRO is implemented by close to 180 non governmental implementing partners, of which 140 are reported to be national NGOs. FFW is the most widespread Table 4.4: Safety Net Program Participation, Rural Afghanistan 2003 program, accounting for over one-third of food aid. More Food for Cash for Relief food Welfare work work Neutral than a third of rural distribution households in the NRVA The extreme poor sample participated in the (poorest 20 percent) 24 22 33 20 program, compared to almost The moderately 20 percent and 13 percent poor (middle 60 participation in CFW and percent) 60 61 53 60 Relief Food Assistance, The better off respectively. Three percent (top 20 percent) 16 17 14 20 of rural households participated in other public Total 100 100 100 100 and private programs, not Concentration identified in the survey. coefficient -0.08 -0.05 -0.18 0 Sources: NRVA 2003; and World Bank 2005. Relief Food Aid is better targeted to the poorest 20 percent of households than are the other two programs: one-third of relief food aid recipients are among the poorest rural households. The shares of participants in FFW and CFW, however, are essentially the same for all consumption quintiles, with more than 15 percent of participants in each program coming from the richest 20 percent of households.51 The concentration coefficient, which captures whether program participation is progressive or regressive, indicates that FFW and CFW are welfare neutral; only the RFA programs are significantly progressive.52 50Food commodity value estimated to be around 46% of total costs (WFP, 2003). 51While it is possible that they were less well off at the time of participating in the program (the data refer to the year preceding the survey) and as such were more eligible beneficiaries, this seems quite unlikely. 52The concentration coefficient ranges from +1 to ­1, with positive values indicating a regressive distribution, negative values a progressive distribution, and a value of zero a perfectly equitable distribution. The formula for calculating the concentration coefficient is: 2 N - 1 + ix N i i = 1 N 1 37 All of these programs have limited coverage. The exclusion error (the percentage of the eligible population that is not participating in the program) is large by all targeting eligibility criteria considered here: (i) food consumption poverty; (ii) food intake gap (food insecurity); (iii) whether the households had unemployed household members; and (iv) if the household suffered employment shocks. For example, for FFW, only 37 percent of poor households participate in the program; 63 percent of the poor are excluded. There are significant errors of inclusion (non eligible households that participate in the programs), as well. In FFW programs, 45 percent of participants are non-poor households. Similarly, using "food intake gap" (e.g., households who consume less than 2100 calories per capita), as the criteria for program eligibility, 85 percent of the households that are food insecure do not participate in the RFA program, while more than 75 percent of the households that would not be eligible as they are not "food insecure" do receive food aid. Table 4:5: Program Targeting Indicators under Various Eligibility Criteria (%), Rural Afghanistan Assumed targeting criteria: Inclusion error: Exclusion error: % of non-eligible group % of eligible group receivinga not receivingb Poverty Food for work 45 63 Cash for work 47 79 Relief food 38 83 Food intake gap Food for work 81 70 Cash for work 79 85 Relief food 77 85 Unemployed in household Food for work 88 67 Cash for work 89 75 Relief food 85 83 Employment shock in household Food for work 94 60 Cash for work 95 80 Relief food 92 80 aPercentage of non-eligible group participating in program as a share of total program participants. bPercentage of eligible group not participating in program as a share of the total eligible population. Source: NRVA 2003; World Bank (2005). Further evaluation53 is imperative in order to improve some of the targeting and selection criteria that may exist in these programs, and to help restructure and redesign some of the objectives and goals of the LSP PIP. Nonetheless, this initial analysis suggests that, given the high cost of where N is the total number of groupings of the welfare variable used (e.g., 5 food consumption quintiles) and xi is the coverage rates among group i. 53To fully evaluate these programs, additional information not collected in the NRVA is needed, especially data on the level of transfers to households. The role of social support networks in the highly heterogeneous socio-cultural contexts of Afghanistan should also be investigated. 38 public sector food aid programs (average unit cost of cereals of $332/ton54), cash employment programs or income transfers are potentially more efficient. Households that receive these wages or cash transfers can then purchase their food in local markets. Thus, in order to maximize the efficiency of resource use and minimize the risk of disincentives for domestic production and market development, food aid use should be restricted to situations where the comparative advantage over cash transfers is clear, i.e., where: · availability of food in local markets is a constraint on food security; · the commodity distributed is an inferior good (i.e., a good for which quantity consumed falls as household incomes rise); and / or · the food aid commodity can be fortified, so that it carries additional nutritional benefits that are not easily available to target households in their normal diets. This applies to relief programs, as well, though it is crucial to distinguish areas where markets are likely to supply demand for food spurred by cash-based programs and areas where markets are poorly developed. Only in the latter case are food-in-kind programs likely to be more efficient than cash-based programs (See Annex 3). In general, these conditions are not met for wheat and wheat flour under current food aid programs. Thus, where food markets are functioning well, cash-based programs are a more efficient means of enhancing food security of the poor than are food-based programs. In particular, the National Emergency Employment Programme (NEEP), the National Solidarity Programme (NSP) and other national cash-based interventions are expected to provide employment to nearly four million people in 2004/05. Other programs such as the National Vulnerability Programme have the potential to increase household food security through raising incomes of the poor, and micro-finance programs offer promise for consumption smoothing and risk management. 54This figure includes estimates of all shipping and handling costs for cereals; it does not include personnel or material costs of implementing programs. The figure cited above for cost of cereals delivered ($332/ton) assumes that shipping and handling costs are proportional to commodity values; assuming shipping and handling costs are proportional to commodity weights raises the estimated cost of cereals delivered to $421/ton (see Annex Tables A4.8 and A4.9). Note that much of the food aid is delivered to rural areas, which involves higher costs than grain delivered to urban markets. 39 40 V. CONCLUSIONS AND POLICY IMPLICATIONS After decades of war, political upheaval, and natural calamities (including droughts), Afghanistan faces numerous challenges in reducing poverty and increasing food security. Productivity of agriculture, particularly food grains remains low, production is highly variable and susceptible to drought, and private markets face high transaction costs due to damaged or non-existent market infrastructure. Moreover, poverty and malnutrition remain widespread. According to the National Rural Vulnerability Assessment (NRVA) in 2003, about half of the rural population had food expenditures less than the amount required to purchase adequate calories55 and about 20 percent of households consumed less than 2100 calories/person/day. Achieving food security in Afghanistan requires policies and programs that address the three major aspects of food security, · Availability of food (from domestic production or imports) at the national and local levels, · Access to food by households (which is largely determined by their purchasing power and by public and private transfers), and · Nutritional aspects, (including consumer choice in consuming food, diversity of diets, and health factors that influence absorption of food by the body). Inclusion of all factors influencing food availability, access and nutrition, would make the scope of food policy as wide as the scope of overall development strategy. A food policy framework is nonetheless needed to help guarantee consistency across a range of sectoral policies that impact on national and household food security. This food policy note focuses on policies related directly to production, processing and marketing, and consumption of food, especially of wheat, the major food staple of Afghanistan. The major findings of the report can be summarized as follows: · Food security does not necessarily require national self-sufficiency in wheat or other food staples, as long as the country has access to international markets. Rather, diversification into legal high-value crops and livestock products may be the most effective means of increasing food security, through raising the incomes and purchasing power of the rural poor. · In spite of very difficult conditions, wheat markets in Afghanistan have performed fairly well and private sector international trade has helped to stabilize supply and prices. Therefore further developing wheat markets and facilitating private sector trade is called for and will enhance food security. · At the household level, food insecurity in Afghanistan is largely caused by inadequate access to food resulting from low household incomes. For most of Afghanistan, where availability of food is not a constraint, policies to promote growth in rural incomes are the most efficient means of enhancing food security of the poor. · In the short run, food aid should be targeted to poor households in parts of the country where markets are failing; even here, however, close attention should be paid to producer and trader incentives. 55The food expenditure cutoff used was Af 4621/year ($94/year), equal to the cost of purchasing 2100 calories/day of the typical basket of food items consumed by the poorest 40 percent of rural households in Afghanistan. 41 · Development of both private and public capacity for data collection and analysis is a high priority for effective formulation and implementation of food policies. Food Production Food production is central to food security in Afghanistan because most of the food consumed in Afghanistan derives from domestic production, and because production of both crops and livestock is a major source of rural incomes. Production systems vary tremendously across Afghanistan, and are influenced by both agro- ecological factors (e.g., elevation, soil type, rainfall, and availability of irrigation and socio- economic factors (e.g., market proximity and infrastructure ­ roads, electricity, communications etc., security, level of incomes). Due to mountainous terrain in much of the country and an arid to semi-arid climate, only about 14 percent of total area is cultivated with crops. Of this agricultural land, about 40 percent is irrigated. Availability of water is a limiting factor on agricultural production in most years. Wheat production has varied considerably in recent years because of both drought and deterioration in irrigation infrastructure. After severe droughts in 2000 and 2001, adequate rainfall and snow melt in late 2002 permitted a large expansion in area sown in November-December for the wheat crop harvested in April-May of 2003, officially estimated at 4.36 million tons (MAAH/FAO). There remains considerable uncertainty regarding wheat production estimates, however, and estimated production in 2004 is only 2.29 million tons. Similarly, livestock population and production (mainly sheep and goats, cattle and poultry) has been heavily influenced by drought, as well as war. Achieving substantial increases in agricultural and livestock production would require: · Investments in irrigation rehabilitation and irrigation management reforms (establishment of a river basin approach and support for local management of water resources through water users associations) can help bring more area under cultivation, reduce the effects of drought, and enhance yields. · Investments in agricultural research and extension, including adaptation of existing plan varieties for local conditions, reforms in extension services to make them more responsive to farmers, and legalization of contract farming and the private sector seed industry can lead to significant increases in agricultural productivity. · Maintaining adequate incentives for efficient domestic production through policies and infrastructure investments that promote development of competitive agricultural markets and avoid price disincentive effects of subsidized imports (food aid or commercial sales). Food security does not necessarily require national self-sufficiency in wheat or other food staples, however. As long as the country has sufficient purchasing power to acquire wheat and other food on international markets, and market flows are not blocked, commercial trade can ensure adequate supply of food at the national level. Similarly, as long as total household incomes are sufficient to purchase available food in local markets, these households need not be self-sufficient in food, either. Indeed, diversification into legal high-value crops and livestock products is a promising strategy for raising employment and incomes of the rural poor, increasing household food security, and reducing poppy cultivation in Afghanistan. 42 Domestic and International Markets Food markets play a crucial role in food security, linking domestic producers and international suppliers to consumers, including both rural non-farm and farm households (only about one fourth of all rural households are self-sufficient in wheat). Available evidence suggests that these markets have performed fairly well in Afghanistan under very difficult conditions of war, internal conflict and disruption of infrastructure. Nonetheless, transport and other marketing costs remain high, and further development of efficient private markets is thus essential to translating food security at the national level (i.e., sufficient overall availability of food) to adequate availability at the local level. Private sector imports and food aid can add to food security through increasing overall supply (availability) of food, especially in the winter months. These imports can also improve access to food for households by containing market price increases. Survey evidence and analysis of market prices suggest that since 2001 private sector imports (mainly originating from Pakistan and Kazakhstan) have involved substantial numbers of traders that respond quickly to shortages in domestic market supplies that would otherwise put upwards pressure on price. In this way, private sector imports have helped to stabilize both market supplies and market prices in major cities in Afghanistan. Food aid inflows also add to domestic supplies, but because their levels are determined by administrative procedures and not by market forces, they have the potential to reduce market prices below international border price levels. Thus, food aid can reduce producer prices below the levels that would prevail in markets in the absence of food aid, reducing farmer incomes and incentives for production, but potentially benefiting consumers. Market price analysis and the significant scale of commercial wheat and wheat from inputs suggest that food aid inflows to date have not had significant disincentive effects on market prices in Afghanistan. Future flows, however, could lead to disincentive effects if not carefully managed, and as discussed below, the high costs of food aid suggest that where the option to switch to cash-based interventions exist, this should be the preferred direction. The experience of many countries indicates that efforts to stabilize food supply and prices through public sector procurement, storage and distribution of food can involve large fiscal subsidies. Minimizing government direct interventions in food markets will encourage private sector investment, and further development of efficient private food markets. More specifically, government policies can promote efficient markets by: · Maintaining incentives for private sector imports through a policy of free trade (no import licenses required apart from traders' registrations and zero or minimal import tariffs and taxes). · Maintaining free trade between and within provinces (no movement restrictions or taxes on domestic trade flows, and no limits on the location, quantity or prices of food purchases or sales) · Promotion of private sector investment in storage and food processing (e.g., wheat milling) by improving the private sector investment climate and by avoiding restrictions on privately held stocks or the size of milling or other food processing operations. · Minimizing licensing requirements and avoiding imposition of price controls (e.g., administratively set price ceilings on bread). 43 Given scarce public sector resources and the overall effectiveness of private international trade flows to supply markets and help stabilize prices, there appears to be little case for costly, large-scale national stocks at this time. Before any large-scale investments in public storage are made, detailed studies of the likelihood of production and trade disruptions, and the benefits and costs of public stocks should be undertaken. Food Consumption, Poverty and Food Aid Because private sector wheat markets have functioned well in most of Afghanistan in recent years, food insecurity at the household level is largely a function of income poverty. To address this wide-spread and chronic poverty, rapid economic growth is required, combined with well- targeted programs to assist the most vulnerable households in the short run. Targeted programs do not necessarily need to involve food transfers or food aid, however, which typically entail substantially higher costs. (Average unit costs of cereals - mainly wheat - delivered under the WFP Protracted Relief and Recovery Operations was $332/ton56, compared with an average retail market price of wheat flour of $207/ton in six major Afghan cities57 in 2003). Alternatives to food-based safety net programs include the National Emergency Employment Programme (NEEP), the National Solidarity Programme (NSP) and other national cash-based interventions that are expected to provide support to nearly four million people in 2004/05. Wherever possible, these should be preferred to food aid. Where food aid is the only possible intervention, and food aid imports are additional resources for the Afghanistan government (i.e., in the absence of food aid, equivalent monetary resources would not be provided by donors), the use of food aid should be restricted to situations where the comparative advantage over cash is clear. This will help to maximize the efficiency of resource use and minimize the risk of disincentives for domestic production and market development. These situations occur where: · availability of food in local markets is a constraint on food security; · the commodity distributed is an inferior good (i.e., a good for which quantity consumed falls as household incomes rise); · the food aid commodity can be fortified, so that it carries additional nutritional benefits that are not easily available to target households in their normal diets; and · food aid can help create physical assets and support human capital accumulation. Increases in access to food will not necessarily translate into major improvements in nutritional outcomes for poor individuals, particularly for the most vulnerable ­ pregnant and lactating women and young children. Here, there is also an important role for nutrition and health education, improvements in sanitation and quality of water, primary health care and other measures that promote absorption of nutrients and general health. Much international experience shows that increased educational status of women contributes to improved nutritional status of their children. Benefits and costs of programs involving fortification of foods should be assessed 56This figure includes estimates of all shipping and handling costs for cereals; it does not include personnel or material costs of implementing programs, and assumes shipping and handling costs are proportional to the value of the commodity (See Annex Tables A4.8 and A4.9). Note that much of the food aid is delivered to rural areas, which involves higher costs than grain delivered to urban markets. 57Kabul, Faizabad, Herat, Jalalabad, Kandahar, and Mazar-e-Sharif; WFP price data. 44 in comparison with benefits and costs of measures addressing the major factors affecting nutritional outcomes listed above. Food Policy Coordination The broad food policy framework outlined above, and summarized in Table 5.1, can provide a basic framework for policies and programs to enhance food security in Afghanistan. This framework must have in-built flexibility to adapt to changes in the food system, at both domestic and global levels and in response to ongoing participatory consultations with key stakeholders. Moreover, effectively formulation and implementation of specific food policies will require the establishment of a robust and nationwide information system to forecast and monitor key indicators (e.g., market price stability, domestic food production, household purchasing power, commercial food imports, international trade flows, nutritional status, etc.) together with the development of both private and public analytical capacities. This analytical capacity and an information system to forecast and measure domestic food production, and to monitor food market prices and international trade flows are crucial to developing early warning capabilities and effective targeting procedures. Finally, a high-level food policy framework committee may be formed consisting of representatives of the Ministries of Food and Agriculture, Commerce, Rural Rehabilitation and Development, and Social Affairs, together with responsible bodies for Disaster Management and National Statistics. The core economic ministries (Ministry of Economic Development and Ministry of Finance) might also be included. Such a committee can help Afghanistan to continue and refine its broadly successful current policy framework and to avoid policy mistakes in responding to an evolving food security environment. 45 Table 5.1: The Afghanistan Food System: Policy and Program Priorities Production Markets (Local, national, regional and global) Consumption /Social Protection Lead Ministry of Food and Agriculture Ministry of Commerce (MoC) Ministry for Rural Rehabilitation and Development ministry (MRRD) Key Natural Resource Management Public Trade and Investment and Private Sector Public Livelihoods and Social Protection PIP Investment Investment Program (NRM PIP) Investment Program (TIPS PIP) Health and Nutrition PIP; Education PIP Programs Priority Programs: Priority Programs: Priority Programs: National Agriculture Program (NAP) National Private Sector Program (NPSP) National Vulnerability Program (NVP) National Irrigation and Power Program National Skills Development and Market Linkage National Solidarity Program (NSP) (NIPP) Program (NSDMLP) National Emergency Employment Program (NEEP) Food Promote efficient and competitive domestic Promote efficient rural, urban and cross border Enhance both access to food and consumption / utilization System production of food (for domestic and global market development for through providing (greater quantity and quality) through social protection Objective markets) for broad based growth and poverty adequate price incentives for producers, sufficient interventions targeted to the most vulnerable, minimizing reduction through investments in margins for processors and traders, and affordable adverse impacts on other food system stakeholders infrastructure and agricultural technology. food at local levels for consumers. (producers and traders). Present Domestic wheat and rice production Regional grain markets appear to be more closely About half the population lives in poverty; in 2002-03, Situation accounted for about xx to yy percent of linked with markets in neighboring countries than around 20% rural people had inadequate calorie intakes. annual supply from 2001 to 2004. with markets within Afghanistan. The majority of Afghans have virtually no reliable access to 46 Livestock (mainly sheep, goats and cattle) Private sector imports (about twice as large as food safe water sources or health care facilities. populations declined by an estimated xx aid inflows) help stabilize national grain supplies percent from 1995 to 200; Livestock incomes and prices. Almost half of all rural households benefited from relief were xx percent of crop incomes in 2004. food / food for work or cash for work during 2002-3. There is very little modern grain milling capacity. Workfare programs appear largely welfare neutral. Illicit opium poppy production and processing accounts for between one-third to The domestic trade environment is relatively Economic constraints and livelihood strategies vary across one half of GDP. undistorted by formal public interventions such as regions and socio-economic groups, suggesting the need for high import tariffs, quantitative restrictions, or a multi-dimensional strategy for poverty reduction. Lack of assured and timely irrigation water export subsidies. supply at the farm level is a major important constraint to agricultural growth. Recurring Development of food markets and private droughts cause major production shortfalls. investment in food processing are constrained by inadequate transport, communications and power The agricultural research and extension infrastructure, weak financial institutions, system, roads and the financial network were insecurity and corruption. also extensively damaged due to prolonged conflict and insecurity. Existing laws, regulations, and procedures inhibit private sector investment instead of encouraging it. Table 5.1: The Afghanistan Food System: Policy and Program Priorities (cont.) Production Markets (Local, national, regional and global) Consumption/Social Protection Policy and Establish objective of food self-reliance Continued investment in market and business Promote sustained and broad based economic growth -- a Program (having sufficient incomes at national and infrastructure such as roads, power, water, and precondition for widespread and lasting poverty reduction. Priorities household levels to acquire necessary food) telecommunications. rather than food self-sufficiency (own- Ensure that poor people are equipped to participate and production of all food utilized) Promote competitive markets and avoid market benefit from growth through accelerated investments in impediments such as (formal or informal) education and health. Promote the efficient and competitive movement restrictions, stock limits, taxes on domestic production of food commodities domestic trade flows between provinces, or Promote productive social protection instruments such as (including wheat) through: government interventions that reduce private community-based funds, labor intensive public works and incentives for trade. micro-finance. (i) Expansion of the scope and effectiveness of irrigation; Maintain incentives for private sector imports in Given the high cost of direct food transfers, limit use of food (ii) Accelerated generation, dissemination response to market demand. aid and direct food transfers to situations where: and promotion of appropriate technologies · Food aid imports are additional resources for the for both agricultural production and Review alternative mechanisms for ensuring timely Afghanistan government (i.e. in the absence of processing by both women. response to rapid onset disasters including a) food aid, equivalent monetary resources would not (iii) Programs designed to encourage strengthening early warning and price monitoring be provided by donors) diversification through improved quality of systems; b) promotion of private sector trade flows; · Availability of food in local markets is a constraint 47 horticultural products: c) coordination with international agencies. (and injection of cash transfers is unlikely to (iv) Investments in sustainable systems for induce market commodity flows from other affordable animal health and nutrition related Promote diversification of production and trade regions) services patterns including enhanced integration into the (v) Strategically targeted use of global trading system based on both free and fair and/or there are substantial targeting or nutritional benefits government subsidies; trade (removal of trade barriers and WTO to food transfers in-kind because: accession) · The commodity distributed is an inferior good that is self-targeting to poorer Strengthen institutional capacity in terms of If food aid is utilized, monitor potential consumers physical infrastructure, staff skills, and disincentive effective of food aid on domestic · The food commodity can be fortified, so that knowledge of modern and competitive wheat markets. it carries additional nutritional benefits that agricultural sector. are not easily available to targeted Minimize government direct interventions in food households in their normal diets Improve regulatory framework. Remove markets such as public procurement, storage and illegal taxation systems imposed on distribution of food commodities. Develop commercially viable programs of flour fortification agricultural producers. and nutritional education. Establish an effective legal regulatory framework to reduce market transaction costs. Implement additional measures as outlined in the National Vulnerability Program targeted to the poorest and most vulnerable people who are not able to participate in more immediately productive programs. 48 Annex 1 Wheat Production Estimates in Afghanistan: Comparing NRVA and FAO Crop Assessment Data The National Rural Vulnerability Assessment (NRVA) survey data on changes in wheat production between 2002 and 2003 are not consistent with the official (FAO) estimates. The FAO estimates indicate a 59 percent in production, from 2.686 million tons in 2002 to 4.360 million tons in 2003. NRVA data show a much smaller increase ­ only 12 percent. Note that the NRVA derived estimate for total production for 2002 is 2.435 million tons (only 9 percent below the FAO estimate for that year). The NRVA-based production estimate for 2003 (2.729 million tons), however, is 38 percent below the FAO estimate for that year. This NRVA figure is derived using weights that assume that village level data are representative of villages in a given agro-ecological zone in the corresponding district, using the share of population in an agro-ecological zone by district from the NRVA district level questionnaire.58 Changing the weights used does not change the main results: using weights that do not utilize information on agro-ecological zones and that imply that the villages sampled are representative of their district, the percentage increase in production is 12.1 percent. Using only the village level weights (with no adjustment for the size of the village in district or province population), the average percentage increase in production per farmer is only 11.0 percent. Moreover, according to the NRVA data, only seven provinces had production increases greater than 30 percent: Zabul (171 percent), Badghis (120 percent), Hilmand (50 percent), Herat (41 percent), Kapisa (47 percent), Kabul (38 percent) and Balkh (35 percent). Under-reporting of the number of large farms would not explain the low percentage increase in wheat production, either. According to NRVA data on increases in wheat production by farm size, the largest farmers (who cultivated more than 4 hectares)59 had the smallest percentage increase in production (4.1 percent). The largest percentage increases in production were for farmers with between 1 and 2 hectares of wheat area cultivated ­ 20.4 percent. Thus, the discrepancy between percentage increases in wheat production derived from the NRVA survey and the percentage increase reported in the FAO crop assessments is apparently not due to the sampling design of the NRVA survey. Two major (non-exclusive) explanations are possible: · The NRVA survey understates the percentage increase in wheat production in 2003 because of inaccurate data (e.g., under-reporting of 2003 production by farmers). · The FAO estimates for 2002 and 2003 overstate the percent increase in wheat production in 2003.60 58In the NRVA data, 3804 households (34%) did not respond to questions on wheat production; 92 households (0.8%) reported zero production in 2002 or 2003. The weights described here include an adjustment for non-responding households. 59 These farmers accounted for 1.4 percent of the farms nationally (1.8% of households who reported wheat production). 60Note that the large increase in national wheat production reported by FAO is largely due to reported increases in rainfed wheat production. According to NRVA data, five provinces had extraordinarily large 49 Evidence from Price and Trade Data Price and trade data for Afghanistan suggest that a very large percentage increase in wheat production is not plausible. Using estimated consumption for 2002/03 as a base, and adjusting consumption for the effects of a 1.9 percent population increase, an estimated 10 percent in incomes (equivalent to the percentage change in poppy-inclusive GDP), and a real price decline of 10 percent, results in an increase in total food demand by up to 400 thousand tons. However, FAO estimates for 2002 and 2003 suggest that net production increased by 1.5 million tons (1.68 million tons less a 10 percent adjustment for seed, feed and wastage), implying that net imports declined by at least 1.1 million tons. Given that estimated net imports in 2002/03 were only 500 thousand tons, the FAO 2003/04 reported production increase would imply net exports of at least 0.6 million tons. In contrast, traders consistently report significant quantities of private sector imports of wheat and wheat flour (Table 3). In order for estimated net imports to be in the range of 0.3 to 0.6 million tons, the change in production between 2002 and 2003 is estimated as about 300 to 600 thousand tons (10 to 25 percent) -- a figure fairly close to the NRVA estimate of 0.3 million tons (12 percent).61 Conclusions There remains significant uncertainty regarding wheat production data in Afghanistan. Neither the levels nor the percentage changes in area, yields and production are consistent across data sources. Evidence from the analysis of demand suggests that the changes in production implied in the official FAO estimates for 2003 (relative to 2002) are not plausible, and the change in production between the two years was probably on the order of 10 to 25 percent. Given the greater availability of data in 2003, it is plausible that this data is more accurate, which would suggest higher availability of wheat in 2002 than earlier indicated, as well as a significant under- reporting of wheat production in the NRVA data. To the extent that consumption of wheat in the NRVA is internally consistent with NRVA production, consumption levels may be under-stated and poverty levels overstated. Moreover, if the higher production figures are used, private sector imports are only about 10 percent of supply.62 (Even with the lower production figures, wheat imports are at most only about 15 percent of supply). Thus, Afghanistan may be nearing the point of market-determined self-sufficiency in wheat. Indeed, it is possible that when more modern wheat milling capacity is installed in Afghanistan, this product will replace imported flour to a large extent, such that even with current production levels, total wheat imports may decline. increases in rainfed wheat production: Logar (736%), Kabul (569%), Paktika (191%), Badghis (131 percent), Herat (70%). 61One possible explanation for the discrepancy in the percentage change in production (2003 vs. 2002) in the two data sources is that yields in 2002 for irrigated land were grossly understated. Using 2003 irrigated wheat yields (2.85 tons/ha) in place of the original 2002 FAO figure (2.02 tons/ha) results in a total production of 3.554 million tons, (22.7 percent higher than the original FAO 2002 figure). The percentage change in production between 2002 and 2003 is then only 22.7 percent instead of the original figure of 62.3 percent. Moreover, the 22.7 percent change is more in line with the NRVA estimated change (12.1 percent). However, this upward adjustment in yields makes FAO average yields on irrigated land (2.85 tons/ha) almost 60 percent higher than the corresponding NRVA yields for both 2002 and 2003. See Annex Table 1. 62Food aid accounts for another 6-7 percent of supply. 50 Finally, further work is needed to improve the methodology for data collection on food availability in the future, so as to help avoid errors of either over-estimating or under-estimating food assistance needs. 51 Table A1.1: Afghanistan: Production of Wheat, 2002 and 2003 (FAO Estimates) IRRIGATED Wheat RAINFED Wheat Total Wheat Area Yield Prod. Area Yield Prod. Area Yield Prod. Population Prod./cap Region (000'ha) (tons/ha) (000' tons) (000'ha) (tons/ha) (000' tons) (000'ha) (tons/ha) (000' tons) (mns) (kgs/person) 2003 Northern 441 2.9 1279 984 1.1 1090 1425 1.7 2369 6.5 366.2 Mountains 308 3.0 911 45 0.5 23 353 2.6 934 10.8 86.1 West / SW 309 2.7 825 207 1.1 232 516 2.0 1057 5.3 199.3 52 Total 1058 2.8 3015 1236 1.1 1345 2294 1.9 4360 22.6 192.8 2002 Northern 464 1.9 901 449 0.8 351 913 1.4 1252 6.3 197.2 Mountains 225 2.1 477 38 0.8 31 263 1.9 508 10.6 47.7 West / SW 356 2.1 732 210 0.9 194 566 1.6 926 5.2 177.9 Total 1045 2.0 2110 697 0.8 576 1742 1.5 2686 22.2 121.0 Growth Rate (2003 / 2002) Northern -5.0% 49.4% 42.0% 119.2% 41.7% 210.5% 56.1% 21.2% 89.2% 1.9% 85.7% Mountains 36.9% 39.5% 91.0% 18.4% -37.3% -25.8% 34.2% 37.0% 83.9% 1.9% 80.4% West / SW -13.2% 29.8% 12.7% -1.4% 21.3% 19.6% -8.8% 25.2% 14.1% 1.9% 12.0% Total 1.2% 41.1% 42.9% 77.3% 31.7% 133.5% 31.7% 23.3% 62.3% 1.9% 59.3% Note: Northern is defined as north and north-east regions. Source: FAO/WFP 2003; authors' calculations. Table A1.2: Afghanistan: Production of Wheat, 2002 and 2003 (NRVA estimates) IRRIGATED Wheat RAINFED Wheat Total Wheat Area Yield Prod. Area Yield Prod. Area Yield Prod. Population Prod./cap Region (000'ha) (tons/ha) (000' tons) (000'ha) (tons/ha) (000' tons) (000'ha) (tons/ha) (000' tons) (mns) (kgs/person) 2003 Northern 605 1.96 1188 610 1.03 628 1215 1.49 1816 6.47 280.7 Mountains 305 1.69 515 128 0.29 38 433 1.28 552 10.84 50.9 West / SW 209 1.45 303 55 1.06 58 264 1.37 361 5.30 68.0 Total 1119 1.79 2006 793 0.91 723 1912 1.43 2729 22.62 120.7 2002 53 Northern 537 2.02 1085 572 1.00 570 1109 1.49 1656 6.35 260.8 Mountains 287 1.67 480 113 0.45 51 400 1.33 531 10.64 49.9 West / SW 168 1.31 220 42 0.69 29 210 1.19 249 5.20 47.9 Total 992 1.80 1785 727 0.89 650 1719 1.42 2435 22.19 109.7 Growth Rate (2003 / 2002) Northern 12.7% -2.8% 9.5% 6.6% 3.2% 10.1% 9.6% 0.1% 9.7% 1.9% 7.6% Mountains 6.1% 1.2% 7.3% 13.2% -34.7% -26.1% 8.1% -3.7% 4.1% 1.9% 2.1% West / SW 24.4% 10.5% 37.5% 31.3% 52.4% 100.1% 25.8% 15.1% 44.8% 1.9% 42.0% Total 12.7% -0.3% 12.4% 9.1% 2.0% 11.2% 11.2% 0.8% 12.1% 1.9% 9.9% Note: Northern is defined as north and north-east regions. Source: NRVA 2003; authors' calculations. Table A1.3: Net Imports under Alternative Estimates of Production and Demand, 2003/04 Per capita demand No Change Low Elasticity High Elasticity Change in Per Capita Demand 0.0% 4.1% 10.6% Low: 12% Increase (NRVA) 243 374 581 7% 10% 15% Medium: 20% Increase 52 183 390 1% 5% 10% High: 62% Increase (FAO) -973 -842 -635 -27% -22% -16% Notes: Net imports as a percentage of total utilization shown below import figures. Low Elasticity: Income elasticity of demand = 0.2; own-price elasticity of demand = -0.2. High Elasticity: Income elasticity of demand = 0.5; own-price elasticity of demand = -0.5. Calculations use an approximate real price decline of 10 percent and per capita income increase of 10 percent (11 percent total GDP growth and 1.8 percent population growth). Dollar price of wheat rose 6.3% (Kabul) with 3.6% nominal exchange rate appreciation. = 2.6% increase in Afghani price deflated by 10.5% inflation = 7.1 percent real price decline. Sources: Author's calculations; FAO and WFP Kabul data; World Bank (2004a), Table 1.1 54 WORLD FOOD PROGRAMME EVALUATION REPORTS SUMMARY EVALUATION REPORT OF PRRO 10233 IN AFGHANISTAN EXECUTIVE SUMMARY Annex 2 Mid-term Evaluation of Food Aid Programs in Afghanistan, 2004 This report synthesizes the findings of an independent evaluation of the Afghanistan protracted relief and recovery operation 10233 approved for April 2003 to April 2005. The evaluation was commissioned by the Office of Evaluation and fielded in May 2004. The evaluation team found that WFP Afghanistan is doing a commendable job in implementing WFP's mandate despite the security situation, which has worsened since the PRRO was designed at the end of 2002. The operation has a number of strengths such as strong synergy between project design and government priorities, effective logistics and pipeline management, and creative and innovative project design, especially in activities related to nutrition. Interviews with partners and beneficiaries reflected general satisfaction with WFP assistance in Afghanistan. In line with WFP's shift to results-based management, the evaluation team looked at results at the output and outcome levels. However, because of a weak monitoring system it was difficult to get reliable data on beneficiary numbers or any results at the outcome level. The evaluation team based its findings on triangulation between country office monitoring data, interviews with partners, studies and interviews with beneficiaries. Support for internally displaced people in camps was very effective, but otherwise the evaluation found that the relief objectives were not fully met because (i) free food distributions had been reduced in reaction to government and donor criticism of food aid and (ii) food for work was spread so thinly that the average household received only enough to support a family for 45 days, regardless of the food gap. This was exacerbated by a "one-size-fits-all" approach to food for work that paid insufficient attention to differences between districts with acute/very high food insecurity and those with moderate food insecurity. In many cases, the food assistance was insufficient to have a significant impact on the livelihoods of those who received it. Recovery objectives were partially met through a combination of food for education, food for training and food for work; the effectiveness of school feeding was reduced, however, by an unreliable biscuit supply. Performance in asset creation was highly satisfactory in Badakhshan, where WFP has supported food for work for over a decade and established long-standing relationships with implementing partners who mobilized resources for non-food inputs. In provinces that did not have previous experience with asset-oriented food for work, the quality and sustainability of the work are less satisfactory. In its second year of implementation, the PRRO should concentrate on ensuring greater effectiveness at the outcome level of (i) food for work, (ii) food for training, (iii) food for education, (iv) assistance to the rural vulnerable and (v) food fortification by working more closely with partners that can provide the non-food inputs necessary to achieve the outcomes and 55 by ensuring a better match between assessed needs and food rations. Assistance to the urban vulnerable needs to be reconsidered, because the changed urban economic environment has made the women's bakeries a less effective instrument for addressing urban vulnerability. Free food for internally displaced people in camps should continue. The country office should improve the linkage between assessment, programming and monitoring ­ especially of outcomes ­ in order to improve programme effectiveness. 56 Annex 3 Ensuring Food Security following Production Shortfalls ­ The Role of Food Aid Periodic wheat production shortfalls in Afghanistan can threaten food security in two major ways: by directly reducing availability of food and by reducing access to food through loss of incomes and higher food prices. Food aid, one of the major policy instruments utilized to address food insecurity, averaged 304 thousand tons per year from 2000/01 to 2002/03 period, and accounted for 11 percent of total availability of wheat. (Private commercial imports accounted for an estimated 24 percent of total availability in the same period.) Whether food aid is the most appropriate instrument for enhancing food security in any given context, however, depends crucially on whether household food insecurity is due to problems of availability, access or both. Afghanistan's food production is highly variable, due mainly to large fluctuations in the availability of water from snow melt or rainfall. Severe droughts, combined with a hazardous security environment, led to poor harvests in 1998, 1999 and 2002, and especially severe reductions in wheat and other agricultural production in 2000 and 2001. Moreover, food markets appear to be only weakly linked within much of Afghanistan because of the mountainous terrain in much of the country, inadequate road infrastructure, physical insecurity, and relatively small size of markets in isolated parts of the country that limit opportunities for economies of scale in trade. Thus, in most years, the evidence suggests that little grain flows across various regions of the country, e.g., from the northeast surplus provinces of Kunduz, Takhar, and Baghlan to major deficit regions such as Kandahar in the south or Herat in the West. Nonetheless, Afghanistan's domestic markets are strongly linked to markets in neighboring countries and imported grain flows from these countries have added to domestic availability of grain and stabilized market prices in recent years. According to available survey evidence, almost all of this commercially imported wheat in 2004 has been in the form of flour. Imported wheat flour flows from Peshawar, Pakistan to Jalalabad and Kabul in eastern Afghanistan; Pakistan wheat flour also moves through Quetta to supplement food supplies in Kandahar in southern Afghanistan. Wheat flour from Kazakhstan, Uzbekistan and Turkmenistan adds to food supplies in the north in drought years and also flows through to Kabul and other parts of Afghanistan. Finally, although Iran is a wheat deficit country, some reports indicate that Iranian flour or flour originating from Kazakhstan, supplies Herat and other markets in the western region of Afghanistan. Because of these private sector import flows, wheat and wheat flour prices in major Afghanistan cities have generally been close to import parity prices (the cost of imported wheat grain or flour from the lowest cost international supply source) despite the drought and physical insecurity conditions in recent years. Private sector imports, combined with food aid flows, have to a large extent offset production shortfalls, ensuring adequate food availability at a national level (and in various major markets). These private market inflows do not necessarily ensure adequate food availability in more isolated markets, such as those in the mountainous areas of west-central Afghanistan ­ nor do they guarantee food security at the household level, since poor households may lack the resources to purchase sufficient food. Thus, widespread chronic food insecurity persists even in years with 57 relatively good harvests, as evidenced by the findings of the NRVA for 2003. For these households, other interventions are needed to increase their access to food. One of the standard interventions is food aid ­ this is discussed below. However, interventions do not necessarily need to be food transfers. Where market availability of food is not a constraint, increasing household incomes is the best way to enhance their access to food, either by enabling purchases of existing local food supplies, or by stimulating traders to respond to enhanced effective demand. This is the first best solution. In situations where it is not possible to increase household incomes (in the short run) and/or where market availability of food is a constraint, food aid can be an acceptable instrument for welfare enhancement. Guidelines to Ensure the Appropriate Use of Food Aid In order to maximize the efficiency of resource use and minimize the risk of disincentives for domestic production and market development, food aid use should be restricted to situations where the comparative advantage over other interventions, for example cash transfers, is clear. Food aid and food transfers in-kind will tend to be most appropriate where: · Lack of availability of food in local markets (most often signaled by high prices) is a constraint on food security, as is the case for rural areas in mountainous regions (particularly in central and western Afghanistan) with poor road and marketing infrastructure. These isolated markets tend also to be those where purchasing power is weakest. · Where weak market demand, rather than poor infrastructure, is the binding constraint on volume of market flows, cash-based programs may induce inflow of additional market supplies. · In some cases, the provision of food can be more effective than other interventions in targeting benefits to the poor. This is particularly true where the commodity distributed is an inferior good (i.e., a good for which quantity consumed falls as household incomes rise). This is likely not the case for wheat flour in Afghanistan, which is consumed in large quantities by all income groups. · Food in kind can be fortified, so that certain interventions (such as supplementary or school based feeding programs) may carry additional nutritional benefits. · The food aid imports provide an additionality to the national budget, i.e., in the absence of food aid, Afghanistan could not obtain the equivalent value of resources in cash to resource other programs for the poor. Box A3.1: Where Markets are Weak: Going Carefully with Cash Where market analysis highlights weak integration regarding the ability of traders to effectively respond to increased purchasing power (e.g., as a result of cash based safety net resource transfers), the following questions can be used to frame assessment and monitoring: · Under what circumstances will traders respond to an increase in demand? · What level of purchasing power is necessary and at what distance from supply to ensure an inflow of food or other items? · How do prices behave following an injection of cash? · At what level of cash inflow does inflation become inevitable? · How do beneficiaries (gender/economic status) in varying circumstances (emergency/non- emergency) spend cash? · How can investment in livelihoods be encouraged while allowing beneficiaries the flexibility to satisfy their consumption needs? Peppiatt, D., J. Mitchell and P. Holzmann. 2000. Buying Power: The Use of Cash Transfers in Emergencies. British Red Cross. 58 Conclusions Apart from its important role in addressing rapid onset emergencies, food aid can be a valuable resource for increasing availability of food in the short term in areas in where markets are not functioning well. However, food aid has the potential to lead to reduced incentives for domestic production and market development. In order to avoid these adverse effects, care should be taken to limit the use of food aid to programs and areas where food aid is a more efficient mechanism and where other interventions are not feasible in the short-run. Increased analytical capacity and reliable data are needed to monitor prices and availability and to conduct the detailed analysis needed to assess food aid needs in the short and medium term. 59 60 Annex 4 Supplementary Annex Tables 61 Table A4.1: Regional Wheat and Rice Production, Trade and Consumption, 2001 (FAO data) Net Imps/ Total Grain Grain Population Production Prod/cap Consumption Cons/cap Net Imports Cons Calories/cap Calories/cap Cal Share (mns) (mn tons) (kgs) (mn tons) (kgs) (mn tons) (percent) (calories/day) (calories) (percent) Afghanistan* Wheat 22.6 3.5 156.8 3.9 172.1 0.70 18.0% 2859 1367 47.8% Rice 22.6 0.3 12.9 0.4 18.3 0.15 36.6% 2859 187 6.5% Pakistan Wheat 145.0 19.0 131.2 19.0 131.1 -0.69 -3.6% 2457 1042 42.4% Rice 145.0 3.9 26.8 1.8 12.4 -2.40 -133.2% 2457 127 5.2% Iran Wheat 71.4 9.5 132.5 12.1 169.9 6.36 52.5% 2931 1459 49.8% Rice 71.4 1.3 18.6 1.7 23.8 0.65 38.1% 2931 248 8.5% 62 Kazakhstan Wheat 16.1 12.7 789.3 2.0 125.8 -3.21 -158.5% 2477 954 38.5% Rice 16.1 0.1 8.2 0.1 6.1 -0.01 -5.2% 2477 59 2.4% Kyrgyzstan Wheat 5.0 1.2 238.7 1.0 210.0 0.07 7.1% 2882 1633 56.7% Rice 5.0 0.0 2.2 0.0 2.3 0.00 6.9% 2882 22 0.8% Tajikistan Wheat 6.1 0.4 63.3 0.7 121.3 0.35 46.4% 1662 943 56.7% Rice 6.1 0.0 4.3 0.0 6.5 0.00 0.8% 1662 64 3.9% Turkmenistan Wheat 4.8 1.8 364.1 1.0 211.4 0.28 27.4% 2738 1644 60.0% Rice 4.8 0.0 5.4 0.0 5.3 0.00 6.2% 2738 52 1.9% Uzbekistan Wheat 25.3 3.8 149.9 3.9 155.1 0.53 13.4% 2197 1207 54.9% Rice 25.3 0.0 1.8 0.2 6.3 0.10 64.7% 2197 61 2.8% Regional Ave Wheat 273.7 48.3 176.5 39.9 145.8 3.7 9.3% 2553 1180 46.2% (w/o Afghan) Rice 273.7 5.5 19.9 3.8 14.0 -1.7 -43.1% 2553 144 5.6% * Afghanistan data is for 2003/04 using best estimates; calories per capita in Afghanistan is NRVA rural average. Table A4.1: Regional Wheat and Rice Production, Trade and Consumption, 2001 (FAO data, cont.) Net Imps/ Total Grain Grain Population Production Prod/cap Consumption Cons/cap Net Imports Cons Calories/cap Calories/cap Cal Share (mns) (mn tons) (kgs) (mn tons) (kgs) (mn tons) (percent) (calories/day) (calories) (percent) Afghanistan* Wheat 22.6 3.5 156.8 3.9 172.1 0.70 18.0% 2859 1367 47.8% Rice 22.6 0.3 12.9 0.4 18.3 0.15 36.6% 2859 187 6.5% Pakistan Wheat 145.0 19.0 131.2 19.0 131.1 -0.69 -3.6% 2457 1042 42.4% Rice 145.0 3.9 26.8 1.8 12.4 -2.40 -133.2% 2457 127 5.2% Bangladesh Wheat 140.4 1.7 11.9 3.3 23.7 1.67 50.1% 2187 202 9.2% Rice 140.4 24.2 172.4 21.8 155.3 0.15 0.7% 2187 1547 70.7% 63 India Wheat 1024.9 69.7 68.0 59.2 57.8 -3.01 -5.1% 2487 493 19.8% Rice 1024.9 93.2 90.9 78.2 76.3 -2.19 -2.8% 2487 756 30.4% Nepal Wheat 23.6 1.2 49.1 0.9 38.4 0.00 -0.4% 2459 351 14.3% Rice 23.6 2.8 117.7 2.4 99.9 0.04 1.8% 2459 922 37.5% Sri Lanka Wheat 19.1 0.0 0.0 0.8 41.9 0.73 91.0% 2274 293 12.9% Rice 19.1 1.8 94.0 1.8 92.3 0.05 2.8% 2274 894 39.3% * Afghanistan data is for 2003/04 using best estimates. Rice in milled equivalents; calories per capita in Afghanistan is NRVA rural average; Note that wheat cons in NRVA is 162.6 kgs/person/year. Source: Calculated from FAO Food Balance Sheets (FAOSTAT). Table A4.2: Grain Production and Consumption in Afghanistan by Province, 2002 and 2003 Population Estimated (1,000s) Consumption Production Balance Production Balance Province (150 kg/cap) 2002 2002 2003 2003 North 2,984.5 447,675 281,312 (166,363) 335,822 (111,853) Faryab 794.1 119,115 92,557 (26,558) 100,447 (18,668) Jawazjan 447.5 67,125 37,888 (29,237) 47,435 (19,690) Sar-i-Pul 474.8 71,220 45,522 (25,698) 50,068 (21,152) Balkh 949.6 142,440 78,151 (64,289) 105,668 (36,772) Samangan 318.5 47,775 27,194 (20,581) 32,204 (15,571) Northeast 3,046.9 457,035 1,328,174 871,139 1,442,198 985,163 Baghlan 726.6 108,990 265,749 156,759 300,971 191,981 Kunduz 833.2 124,980 624,967 499,987 666,966 541,986 Takhar 761.4 114,210 278,717 164,507 293,175 178,965 Badakhshan 725.7 108,855 158,741 49,886 181,086 72,231 West 1,857.0 278,550 133,636 (144,914) 200,653 (77,897) Heart 1208 181,200 88,965 (92,235) 125,741 (55,459) Farah 343.4 51,510 24,158 (27,352) 29,852 (21,658) Badghis 305.6 45,840 20,513 (25,327) 45,060 (780) West Central 884.1 132,615 70,717 (61,898) 56,264 (76,351) Ghor 492.4 73,860 53,182 (20,678) 35,167 (38,693) Bamyan 391.7 58,755 17,535 (41,220) 21,097 (37,658) Central 5311.2 796,680 119,112 (677,568) 144,056 (652,624) Kabul 3445 516,750 22,729 (494,021) 31,306 (485,444) Parwan 737.2 110,580 30,909 (79,671) 31,256 (79,324) Kapisa 364.9 54,735 28,317 (26,418) 41,696 (13,039) Logar 315.4 47,310 11,264 (36,046) 14,380 (32,930) Wardak 448.7 67,305 25,893 (41,412) 25,418 (41,887) Source: National Risk and Vulnerability Assessment (NRVA). 64 Table A4.2: Grain Production and Consumption in Afghanistan by Province, 2002 and 2003 (cont.) South 1,977.9 296,685 94,126 (202,559) 102,878 (193,807) Paktia 401.3 60,195 19,723 (40,472) 22,721 (37,474) Paktika 357.2 53,580 16,669 (36,911) 20,505 (33,075) Khost 304.6 45,690 22,826 (22,864) 25,189 (20,501) Ghazni 914.8 137,220 34,908 (102,312) 34,463 (102,757) East 1,922.9 288,435 248,164 (40,271) 256,309 (32,126) Nangahar 1105.7 165,855 144,420 (21,435) 150,462 (15,393) Laghman 378.1 56,715 60,523 3,808 60,561 3,846 Kunar 328.1 49,215 35,897 (13,318) 38,792 (10,423) Nuristan 111 16,650 7,324 (9,326) 6,494 (10,156) Southwest 2706.9 292575 114200 -291835 157827 -248208 Kandahar 913.9 137,085 15,384 (121,701) 19,065 (118,020) Helmand 756.4 113,460 67,138 (46,322) 100,825 (12,635) Zabul 249.1 37,365 1,661 (35,704) 4,505 (32,860) Nimroz 151.5 22,725 2,101 (20,624) 2,421 (20,304) Uruzgan 636 95,400 27,916 (67,484) 31,011 (64,389) TOTAL 20,691 3,103,710 2,389,441 (714,269) 2,696,007 (407,703) Source: Calculated from National Risk and Vulnerability Assessment (NRVA) data. Table A4.3: Grain Production and Consumption in Afghanistan by Region, 2002 and 2003 Estimated Pop. Consumption Production Balance Production Balance Province (1,000s) (150 kg/cap) 2002 2002 2003 2003 North 2,984.5 447,675 281,312 (166,363) 335,822 (111,853) Northeast 3,046.9 457,035 1,328,174 871,139 1,442,198 985,163 West 1,857.0 278,550 133,636 (144,914) 200,653 (77,897) West Central 884.1 132,615 70,717 (61,898) 56,264 (76,351) Central 5,311.2 796,680 119,112 (677,568) 144,056 (652,624) South 1,977.9 296,685 94,126 (202,559) 102,878 (193,807) Southwest 2,706.9 406,035 114,200 (291,835) 157,827 (248,208) TOTAL 20,691.4 3,103,710 2,389,441 (714,269) 2,696,007 (407,703) Source: Calculated from National Risk and Vulnerability Assessment (NRVA) data. 65 Table A4.4: Medium Sized Milling Facilities in Afghanistan, 2004 Storage Milling Capacity Is it functional # of employees Capacity MT MT/day at all? Public mills Kabul 60,000 150 Yes 430 Kandahar 40,000 0 No 0 Mazar-e-Sharif 40,000 60 Yes 209 Pol-e-Khumri 40,000 0 No 0 Herat 40,000 - - - Private mills Kabul Kabul Flour Mills 6,000 200 Yes 120 Mazar-e-Sharif Sayed Kemal 3,000 60 Yes 31 Sayed Jamal 2,000 45 Yes 120 Source: Chabot (2004). Table A4.5: Results of Anthropometric Surveys in Kabul (1995 ­ 2003)63 Data of Survey Acute Malnutrition Chronic Malnutrition (percent) (percent) November 1995 6.2 -- May 1996 6.7 -- December 1996 5.1 55.2 June 1997 6.8 52.1 December 1997 7.5 63.0 February 1999 8.7 61.3 February 2000 2.8 53.7 October 2000 8.0 57.5 March 2001 3.8 62.6 August 2002 11.7 47.9 August 2003 8.1 34.4 November 2003 4.2 44.1 Acute malnutrition (wasting) is measured as percentage of children with weight for height Z-scores below 2 standard deviations of reference population. Chronic malnutrition (stunting) is measured as percentage of children with height for age Z-scores below 2 standard deviations of reference population. Source: UNICEF (2003). 63Ibid 66 Table A4.6: Human and Social Capital ­ Policy Priorities and Programs Policy Priority Associated Programs 1382 1383 (Public Investment Budget Budget Programs and their Sub- (million) (million) components) Prioritization of investment in human capital Education PIP 109.9 277.0 (health, nutrition and education) for immediate and longer term impact on poverty through promotion Public Health and Nutrition 153.5 289.4 of universal access to services PIP Build a more open democratic society through Culture Media and Sports 19.0 34.3 freedom of expression and non-discrimination PIP Mainstream challenges of ensuring stability, Refugees and Internally 31.5 82.6 security and economic prosperity with regard to Displaced PIP (+handled as sustainable re-integration of returnees and IDPs cross cutting issue in LSP PIP) Special attention to access and inclusion of All PIPs; Livelihood and different groups of women and men to benefits Social Protection PIP derived from economic growth 233.5 216.1 Focus of targeted investments in livelihood Micro-finance and Business promotion to foster self-reliance and accelerate Development Services Sub- poverty reduction Programs of LSPPIP Provision of specific targeted interventions, National Vulnerability subsidized support and direct social transfers for Program (National Priority extreme vulnerability (poor disabled; poor female Program within LSP PIP) headed households, destitute households, women and children at risk, etc.) Ensure public capacities and responsibilities to National Disaster Risk provide emergency response Management Sub-Program of LSP Community action, involvement and ownership (National Solidarity 32.6 168.8 key to sustainability of reconstruction efforts Program, National through transfer of resources and decision rights to Emergency Employment the local level Program, and National Drinking Water Project ­ Sub Programs of LSP Mainstreaming a labor-based approach to National Emergency 68.2 72.1 investments in rural infrastructure Employment Program ­ Sub Program of LSP 67 Table A4.7: Afghanistan: LSP Significant Achievements in 1382 (FY2003/04) National Emergency The NEEP generated 8 million labour days through establishment of a Employment Program nationwide employment based safety net for the construction of rural (NEEP) infrastructure (including 5000 km of rural roads) using mainly cash payments National Solidarity Program The NSP, Afghanistan's flagship intervention for community (NSP) empowerment, provided resources through block grants for community managed small-scale reconstruction and development in rural areas, and promoted more participatory and inclusive decision making and governance at the village level. The target is to cover the country's approximately 20,000 villages over a four year period. As of May 2004, 4,600 village level community development councils had been elected, 3,400 village development councils completed and 2,350 sub-projects submitted to approval. Comprehensive Disabled The delivery of services to 30,000 disabled people in Herat, Farah, Afghan's Program (CDAP) Kandahar, Ghazni, Mazar and Takhar. Community Based Rural The construction of over 12,000 water points across the country with Water Supply and prioritisation given to areas where pressure on host communities resulting Sanitation Program from returning refugees and IDPs has resulted in particularly acute needs (CBRWSS) for safe water and sanitation facilities. National Area Based The formulation and implementation of immediate recovery and Development Program development projects in areas where the local economy has been (NABDP) particularly damaged, based on local resources and existing / historical skills and productive enterprises. 317 projects valued at over $17.9 million across 34 provinces. Micro-Finance Support The establishment of MISFA following a National Micro-Finance Facility in Afghanistan Workshop in February to provides funds for the competitive contracting of (MISFA) specialist professional agencies to sustainably deliver financial services to poor Afghans. Over 20,000 families received loans in 1382. As of May 2004 30,000 loans have been disbursed to clients (90% of whom are women) and the repayment rate is 98%. 68 Table A4.8: Afghanistan: Food Aid Costs, 2002-03a Planned Total Total Quantity Price Plan Cost Shipping* Cost Cost Total Cost (tons) ($/ton) (mn $) (mn $) (mn $) (%) ($/ton) Cereals 496,551 167 82.9 82.1 165.04 26.3% 332 Mixed / blended foods 44,042 539 23.7 23.5 47.25 7.5% 1073 Miscellaneous 3,757 152 0.6 0.6 1.14 0.2% 303 Oils and fats 61,698 761 47.0 46.5 93.45 14.9% 1515 Pulses, vegetables 12,941 338 4.4 4.3 8.71 1.4% 673 Total Commodities 618,989 256 158.6 157.0 315.57 50.2% 510 External transport 40.2 12.7% LTSH 78.5 24.9% ODOC 6.8 2.2% DSC 31.4 10.0% Total overhead 157.0 49.8% Total budget 618.989 510 315.6 100.0% No. of beneficiaries 9,243,000 aWFP Protracted Relief and Recovery Operations (PRRO) budget for April 2002 to December 2003. Notes: Shipping costs include all "overhead" items (external transport, LTSH, ODOC, DSC). Shipping costs are allocated across commodity using each commodity's share in total commodity pre-shipping cost (plan cost of $158.6 million). Source: World Food Programme, Kabul; Neun and Fitzherbert (2003), p. 66. Table A4.9: Afghanistan: Food Aid Costs, 2002-03. (Alternative Estimates)a Planned Price Plan Cost Shipping* Total Cost Total Cost Total Cost Qty (tons) ($/ton) (mn $) (mn $) (mn $) (%) ($/ton) Estimate 1: Cereals 496,551 167 83 82 165 52% 332 Non-cereals 122,438 618 76 75 151 48% 1229 Total 618,989 256 158.6 157 316 100% 510 Estimate 2: Cereals 496,551 167 83 126 209 66% 421 Non-cereals 122,438 618 76 31 107 34% 871 Total 618,989 256 159 157 316 100% 510 aWFP Protracted Relief and Recovery Operations (PRRO) budget for April 2002 to December 2003. Estimate 1: Shipping costs allocated by value shares of commodities. Estimate 2: Shipping costs allocated by weight shares of commodities. Source: Authors' calculations; World Food Programme, Kabul; Neun and Fitzherbert (2003), p. 66. 69 Box A4.1: National Emergency Employment Program Afghanistan's National Emergency Employment Program was launched as a National Priority Program in 2002. While initially consisting as one component of the World Bank-funded Community Empowerment and Labour Intensive Public Works Project, the framework of NEEP has expanded to encompass three ministries and a diversified portfolio of coordinated interventions establishing public works / infrastructure projects integrating a labor-based approach. To date the achievements of NEEP are significant and include rapid start up investment in infrastructure while providing employment opportunities with cash wages, signaling the Government's determination to break out of a largely humanitarian programming framework. Since the start of 1382, 7.2 million labor days have been created and over seven thousand km of roads constructed and rehabilitated since the inception of the program as well as other infrastructure outputs (e.g., 51 bridges, 188 wells, 43 schools, 574 canals and 870 other structures). In 1382 the total level of investment was $31 million and this rose to $61 million in 1383. The experience with NEEP to date highlights critical challenges facing `state building' and `reconstruction' agendas. These include: (i) rationalizing ministerial mandates around rural infrastructure investments; (ii) addressing private sector capacity constraints; (iii) accelerating efforts for institutional strengthening at sub-national level; (iv) ensuring resource transfers are increasingly channeled through provincial structures; and (v) the promotion of pro-poor participatory planning processes in support of community driven development principles. Originally conceived as a vehicle for the provision of short-term employment opportunities to poor Afghans, it is clear that NEEP needs to adapt to the changing landscape of the development agenda in Afghanistan. However, given the increased significance of the private sector in generating employment opportunities through out Afghanistan, coupled with the expansion of the National Solidarity Program and infrastructure and investment programs64, future directions for NEEP to evolve as an integrated framework for investment in rural and urban infrastructure are currently under scrutiny. The potential role of NEEP in supporting national processes of policy development is significant, including the establishment and promotion of optimum standards and norms for planning and investment in rural infrastructure and integrating a labor-based approach as a core pro-poor commitment and as part of an integrated social policy. There may, therefore, be a strong case to shift from NEEP to a National Rural / Urban Infrastructure Support Program. 64These include (i) Transport Public Investment Program - Over the next seven years the required funding commitments for transport total $7,519 billion (SAF, 2004) (ii) National Solidarity Program ­ particularly the introduction of block grant allocation for inter-village infrastructure priorities through district level committees, (iii) Afghanistan Stabilisation Program ­ provincial stabilization funds for rural infrastructure, establishing basic systems for sub-national government and coherent resource allocation mechanisms (iv) Private Sector Development and Skills Development National Priority Programs ­ to ensure that the growth of the private sector is sustainable complementary to investment priorities (v) National Vulnerability Program ­ to promote initiatives to ensure that the extremely poor and most at risk are able to participate in growth and access the benefits. 70 VI. REFERENCES Action Contre la Faim. (ACF). 2004. Kabul Vulnerability Mapping. Afghanistan Government. 2003. National Risk and Vulnerability Assessment. Afghanistan Government. 2002. National Development Framework. Afghanistan Government. 2004. Securing Afghanistan's Future: Accomplishments and the Strategic Path. Ahmed, Raisuddin. 2004. "Rice Economy of Bangladesh: Progress and Prospects", Economic and Political Weekly. 39(36): 4043-4052. Ahmed, Raisuddin, Steven Haggblade, and Tawfiq-e-Elahi Chowdhury (eds.) 2000. Out of the shadow of famine: Evolving food markets and food policy in Bangladesh. Baltimore, Md., U.S.A.: Johns Hopkins University Press for the International Food Policy Research Institute. Alderman, Harold. 1988. Estimates of consumer price response in Pakistan using market prices as data. Pakistan Development Review. 27 (Summer): 89-107. Alston, J.M., M.C. Marra, P.G. Pardey and T.J. Wyatt. 1998. Research Returns Redux: A Meta- Analysis of the Returns to Agricultural R&D. EPTD Discussion Paper No. 38. Washington, D.C.: International Food Policy Research Institute (IFPRI). CARITAS Germany. 2003. Women in Kabul: A Needs Assessment. Chabot, Philippe. 2004. A Preliminary Review of the Structure of Wheat and Flour Markets in Afghanistan. Report to the World Bank South Asia Agriculture and Rural Development Unit (processed). del Ninno, Carlo, Paul A. Dorosh, and K. Subbarao. 2005. Food Aid and Food Security in the Short- and Long Run: Country Experiences from Asia and sub-Saharan Africa. Dorosh, Paul A. 2001. "Trade Liberalization and National Food Security: Rice Trade between Bangladesh and India", World Development, 29(4): 673-689. ______ . 2004. Trade, Food Aid and Food Security in Bangladesh: Evolving Rice and Wheat Markets. Economic and Political Weekly. 39(36): 4033-4042. Dorosh, Paul A., Quazi. Shahabuddin, M.A. Aziz and Naser Farid. 2004. "Food Aid and Producer Price Incentives: Implications of Increases in Domestic Foodgrain Production", in Dorosh, P., C. del Ninno and Q. Shahabuddin, (2004). The 1998 Floods and Beyond: Towards Comprehensive Food Security in Bangladesh. Dhaka, Bangladesh: International Food Policy Research Institute and University Press, Ltd. Fan, S., L. Zhang, and X. Zhang. 2002. "Growth, Inequality and Poverty in Rural China: The Role of Public Investments". Research Report 125. Washington, D.C.: International Food Policy Research Institute. 71 Food and Agriculture Organization (FAO). 1997. Afghanistan Agricultural Strategy. Land Management Report. Rome. FAO Food Balance Sheets (FAOSTAT). www.fao.org. FAO/WFP. Afghanistan 2003 Crop and Food Supply Assessment. Rome. Grace, Jo and Adam Pain. 2004. Rethinking Rural Livelihoods in Afghanistan. Afghanistan Research and Evaluation Unit. Kabul, Afghanistan. International Center for Agricultural Research in Dry Areas (ICARDA). 2003. Needs Assessment on Horticulture in Afghanistan. January. Maletta, Hector. 2003. Winters of Content: Seasonal Changes in Wages and Food Prices in Afghanistan, 2002-2003. FAO and Transitional Government of Afghanistan. Ministry of Agriculture and Animal Husbandry. November. Kabul. Manuscript. Maletta, Hector and Raphy Favre. 2003. "Agriculture and Food Production in Post-War Afghanistan: A Report on the Winter Agricultural Survey 2002-03", Ministry of Agriculture and Animal Husbandry. August. Kabul. Manuscript. Mansfield, David. 2002. "The Economic Superiority of Illicit Drug Production: Myth and Reality. Opium Poppy Cultivation in Afghanistan." Paper prepared for the International Conference on Alternative Development in drug control and cooperation, Feldafing, January 7- 12, 2002. Molla, D. 2003. "Food Aid, Wheat Prices and Poppy Cultivation in Afghanistan: Is there a link?" November. World Food Programme, Kabul. Manuscript. Neun, Hansjorg and Anthony Fitzherbert. 2003. Assistance to the Transitional Authority of the Islamic State of Afghanistan to Elaborate Policy Guidance Regarding Food Aid and utilization in Afghanistan. Final Report to European Commission, Kabul. October 2003. Peppiatt, D., J. Mitchell and P. Holzmann. 2000. Buying Power: The Use of Cash Transfers in Emergencies. British Red Cross. London, U.K. Sheladia Associates. 2004. Final Report for Framework of Water Resources Management & Draft Report on Proposed Water Resources Development Strategy, Afghanistan. February. Sloane, Peter. 2001. Food Security Strategy for Afghanistan. Report prepared for the Working Team on Food Security. Islamabad, Pakistan. UNICEF and Central Statistics Office (CSO). 2003. Moving Beyond 2 Decades of War: Progress of Provinces, A Multiple Indicator Cluster Survey (MICS). United Nations Office on Drugs and Crime (UNODC). 2004. Afghanistan: Opium Survey 2004. November. World Bank. 2004a. Afghanistan: State Building, Sustaining Growth, and Reducing Poverty. Report No. 29551-AF. Washington, D.C. 72 ________. 2004b. Agriculture Investment Sourcebook. Washington, D.C.: Agriculture and Rural Development Department. ________. 2004c. India: Re-energizing the Agricultural Sector to Sustain Growth and Reduce Poverty. Washington, D.C.: Agriculture and Rural Development Unit. South Asia Region. ________. 2005. Afghanistan: Poverty, Vulnerability, and Social Protection: An Initial Assessment. Report No. 29694-AF. Washington, D.C. World Resources Institute. 1997. The Urban Environment. Washington, D.C. 73