THE WORLD BANK 46858 Water Management in Agriculture Ten Years of World Bank Assistance, 1994­2004 THE WORLD BANK GROUP WORKING FOR A WORLD FREE OF POVERTY The World Bank Group consists of five institutions--the International Bank for Reconstruction and Development (IBRD), the International Finance Corporation (IFC), the International Development Association (IDA), the Multilateral Investment Guarantee Agency (MIGA), and the International Centre for the Settlement of Investment Disputes (ICSID). Its mission is to fight poverty for lasting results and to help people help themselves and their envi- ronment by providing resources, sharing knowledge, building capacity, and forging partnerships in the public and private sectors. THE INDEPENDENT EVALUATION GROUP ENHANCING DEVELOPMENT EFFECTIVENESS THROUGH EXCELLENCE AND INDEPENDENCE IN EVALUATION The Independent Evaluation Group (IEG) is an independent, three-part unit within the World Bank Group. IEG-World Bank is charged with evaluating the activities of the IBRD (The World Bank) and IDA, IEG-IFC focuses on assessment of IFC's work toward private sector development, and IEG-MIGA evaluates the contributions of MIGA guarantee projects and services. IEG reports directly to the Bank's Board of Directors through the Director-General, Evaluation. The goals of evaluation are to learn from experience, to provide an objective basis for assessing the results of the Bank Group's work, and to provide accountability in the achievement of its objectives. It also improves Bank Group work by identifying and disseminating the lessons learned from experience and by framing recommendations drawn from evaluation findings. W O R L D B A N K I N D E P E N D E N T E V A L U A T I O N G R O U P Water Management in Agriculture Ten Years of World Bank Assistance, 1994­2004 2006 The World Bank Washington, D.C. http://www.worldbank.org/ieg ©2006 The International Bank for Reconstruction and Development / The World Bank 1818 H Street NW Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org E-mail: feedback@worldbank.org All rights reserved 1 2 3 4 5 09 08 07 06 This volume is a product of the staff of the International Bank for Reconstruction and Development / The World Bank. The findings, interpretations, and conclusions expressed in this volume do not necessarily reflect the views 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 judgement on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The International Bank for Reconstruction and Development / The World Bank encourages dissemination of its work and will normally grant permission to reproduce portions of the work promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center Inc., 222 Rosewood Drive, Danvers, MA 01923, USA; telephone: 978-750-8400; fax: 978-750- 4470; Internet: www.copyright.com. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2422; e-mail: pubrights@worldbank.org. Photo credit: Cover photo of seedlings sprouting. ©Photowood Inc./CORBIS. ISBN-10: 0-8213-6904-0 ISBN-13: 978-0-8213-6904-3 e-ISBN: 0-8213-6905-9 DOI: 10.1596/978-0-8213-6904-3 World Bank InfoShop Independent Evaluation Group E-mail: pic@worldbank.org Knowledge Programs and Evaluation Capacity Telephone: 202-458-5454 Development (IEGKE) Facsimile: 202-522-1500 E-mail: eline@worldbank.org Telephone: 202-458-4497 Facsimile: 202-522-3125 Printed on Recycled Paper Contents vii Acronyms and Abbreviations ix Acknowledgments xi Foreword xiii Executive Summary 1 1 Introduction 2 World Bank Strategies for AWM 3 Integrating the Strategy with Rural Development 4 Renewing the Focus of the Rural Strategy 4 Objectives of the Study 7 2 AWM Relevance to the Bank Declined 7 Overview 8 AWM Investment Shrank 9 Aligning Overall Bank Lending with Its Poverty Strategy 9 The Compact Had Unintended Consequences 11 Agriculture's Shrinking Contribution to Growth and Employment 12 Borrowers Are Becoming Focused on Urban Challenges 13 Demand Is Still High for Assistance to Agriculture 13 AWM Is Now Less Focused on New Investment 14 Emphasis on Comprehensive Rural Development Is Growing 14 Policy Evolution Is Reflected in the CASs 15 Regional Differences Are Captured in the CASs 16 Mixed AWM Responses to New Bank Policies 19 Development Objectives Have Changed 21 Less New Construction and Lower Costs 23 Policy Reform Has Languished 24 Implications for Management 27 3 Global Relevance Remains High 27 Overview 27 Benefits of Good AWM Are Substantial 28 Irrigation Impact on Production, Income, and Poverty WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 29 Impact on Employment and Wages 30 Impact on the Nonfarm Economy 31 Impact on Food Prices and Diet Quality 31 Impact on Empowerment of the Poor 32 Demand for Food and Better AWM 36 Current Evaluation Captures Only Partial Impacts 37 4 Outcomes Need Improved Reporting 37 Overview 37 Outcome and Performance Ratings 38 Performance by Bank Region 38 Less than Satisfactory Performance Ratings 39 What Benefits Were Expected for Bank AWM? 41 Benefits Produced Were Less Than Expected 42 Economic Efficiency and Competitiveness Are Declining 45 Social Impacts of AWM Projects 45 Monitoring and Evaluation 47 Results of Poor M&E Design 51 5 Project Design and Impact 51 Overview 51 Poverty Indicators Need More Attention 53 Assessments of Water User Associations Are Too Optimistic 53 Institutional Support for Water User Associations Is Neglected 54 Unrealistic Incentives for Cost Recovery 56 Incentives to Boost Water Use Efficiency Are Frequently Neglected 56 Irrigation Is Only Part of the Solution 57 Bank Staffing and Training Are Weak 59 6 Findings and Conclusions 59 Changing Global and Bank Priorities 61 AWM Remains Relevant 62 Increasing Relevance and Performance 65 Appendixes 67 A: Study Methodology 73 B: Background Information and Trends, 1994­2004 87 C: Detailed Project Design Analysis 95 D: Analysis of Monitoring and Evaluation 105 E: Analysis of Country Assistance Strategies 121 F: Analysis of Project Completion Reports 129 Endnotes 135 Bibliography Boxes 17 2.1 Bank Strategy for Sub-Saharan Africa: The Importance of Agriculture 24 2.2 Armenia: Creating Synergistic Packaging for AWM Lending 32 3.1 Effects of New Irrigation in India's Andrah Pradesh Area, 2005­06 33 3.2 Yemen: Evolution of an Integrated Water Resources Management Strategy i v C O N T E N T S 40 4.1 Main Project Lessons 55 5.1 Madagascar: Unrealistic Strategy for Transferring Responsibility for O&M 55 5.2 Tanzania: The Extent of Cost Recovery Figures 2 1.1 Bank Lending for AWM 4 1.2 Agriculture Sector and AWM Lending in Total Bank Lending 8 2.1 Most Bank AWM Investment Went to Asia in 1994­2004 10 2.2 Bank Resources to Prepare Projects Declined in the Late 1990s 10 2.3 AWM Projects Are among the Most Costly to Prepare 13 2.4 Borrower Demand for Agricultural Investment Remains High 14 2.5 CASs Reflected Changes in Bank Policy 15 2.6 Discussions of Institutions for Water Management Became More Prominent in CASs 16 2.7 Poorest Countries Received Most Bank AWM Investment, 1994­2004 19 2.8 Relative Importance of Development Objectives for All AWM Projects, 1994­2004 20 2.9a Importance of Physical Development Objectives Either Stabilized or Declined in Importance 20 2.9b Improved Attention to Social and Institutional Issues 21 2.10 Comparison of Dedicated and Nondedicated Project Objectives Reveals Strengths and Weaknesses 22 2.11a Reduced New Construction and Improvement in Dedicated Projects 22 2.11b Reduced Improvement and Rehabilitation in Nondedicated Projects 22 2.12 Share of Nondedicated Lending Increased 23 2.13 Overall Commitment Declined 23 2.14 Other Agricultural Investment Declined Steeply among Large Borrowers 34 3.1a Declining Agricultural Value Added Is Faltering in Some Countries and Regions 34 3.1b But Is Growing Apace in Others 37 4.1 Outcomes: Satisfactory Ratings 38 4.2 Sustainability: Likely and Highly Likely Ratings 38 4.3 Institutional Development: Substantial Ratings 39 4.4 Performance Comparisons, by Region 41 4.5a Area of Completed Projects Is Falling 41 4.5b And Future Projects May Be More Expensive 42 4.6 Regional Variations in Infrastructure Costs Are Large 42 4.7 Economic Evaluations: Rural Sector Is Better Than Other Bank Sectors 43 4.8 Relatively Low Economic Impact of Rural Projects, 2001­04 43 4.9 Economic Efficiency is Declining 45 4.10 Neglect of Social Impacts in Completion Reports 46 4.11 Improved Designs of M&E Systems 46 4.12 Logical Framework Greatly Improved M&E 47 4.13 But Attention to Outcome Indicators Needs Improvement 47 4.14 Unsatisfactory M&E of Outcome Indicators, Particularly for Poverty v WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 47 4.15 Rigorous Evaluation Tools and Attribution of Benefits Needed 48 4.16 Poor M&E Planning Leads to Delayed Baseline Creation 48 4.17 Good M&E Needs Vigilant Supervision 49 4.18 Early Baselines Lead to High-Quality Surveys 52 5.1a Poverty Coverage Is Improving 52 5.1b Quality of Appraisal of Poverty Aspects Is Improving 53 5.2a Less Appraisal of Social Impacts in Dedicated Projects 53 5.2b But All Projects Have Improved the Quality of Analysis 54 5.3a Limited Focus on Good Water Management Essentials in All AWM projects 54 5.3b Focus of Essentials for Nondedicated Projects Are Particularly Limited 57 5.4a Coverage of Credit and Marketing Links Has Improved 57 5.4b But Only in Nondedicated Projects Tables 11 2.1 Cost of Safeguard Policies Is High, 1994­2005 11 2.2 Changes in the Bank's Skills Mix, 1997­2005 14 2.3 Developing Countries: Annual Investment for Water Services 18 2.4 Largest Borrowers Took a Smaller Share of Investment 18 2.5 Reform in India Improved Spending on O&M and Revenue Collection 28 3.1 Improving Agricultural Water Management Significantly Contributes to the Millennium Development Goals 44 4.1 Irrigation Benefits in China's Hunan Province Are Technology Dependent v i ACRONYMS AND ABBREVIATIONS ADB Asian Development Bank AFD African Development Bank AFR Africa ARD Agriculture and Rural Development Department ASAL Agriculture sector adjustment loan AWM Agricultural water management CAS Country Assistance Strategy CDD Community-driven development EAP East Asia and Pacific ECA Europe and Central Asia ERR Economic rate of return FAO Food and Agricultural Organization of the United Nations FY Financial year of the World Bank (July 1­June 30) GDP Gross domestic product HIPC Heavily Indebted Poor Countries (Initiative) I&D Irrigation and drainage IBRD International Bank for Reconstruction and Development ICR Implementation Completion Report IDA International Development Association IEG Independent Evaluation Group (formerly OED) IFPRI International Food Policy Research Institute LAC Latin America and the Caribbean M&E Monitoring and evaluation MNA Middle East and North Africa O&M Operations and maintenance OED Operations Evaluation Department (changed to IEG) PAD Project appraisal document PPAR Project Performance Assessment Reports SAR South Asia SSA Sub-Saharan Africa WUA Water user association v i i Acknowledgments This report was written by George T. Keith The draft report was finalized taking into Pitman, supported by Ridley Nelson, Marcello account comments from the Agriculture and Basani, and Anita Yang (consultants). Mark Rural Development Department's agricultural Rosegrant and Rowena A.Valmonte-Santos of water sector team, led by Salah Darghouth; the International Food Policy Research Institute additional information was provided by Sanjiva undertook the literature review of the Cooke and Julienne Roux. William Hurlbut economic, social, and poverty impacts of provided editorial assistance and administrative agricultural water management that forms the assistance was provided by Soon-Won Pak. basis of chapter 3. Christopher Gerrard and Helen Chin edited the report for publication. Christopher Willoughby provided peer review. Director-General, Evaluation: Vinod Thomas Director, Independent Evaluation Group­World Bank: Ajay Chhibber Manager, Sector, Thematic, and Global Evaluation Division: Alain Barbu Task Manager: George T. Keith Pitman i x Foreword The purpose of this study is to update the water portfolio for the period 1994­2004, and Review of World Bank Experience in Irriga- updated lending trends to include 2005. tion (IEG 1994) and to broaden the scope of Emerging issues are thus identified and lessons evaluation to include all water lending for are drawn from the World Bank's rich experi- agricultural development. Since that first ence. This study also updates and elaborates the study, the proportion of World Bank lending for more general findings presented in Bridging agricultural water management continued to Troubled Waters (IEG 2002a). decline, a trend that started in the late 1970s when the subsector received 11 percent of the The study is primarily based on the analysis of a lending, falling to less than 2 percent in 2001­03. wide range of World Bank data and reports, It has since staged a strong recovery and including 131 project appraisal documents, 129 reached over 4 percent in 2005. Commitments country assistance strategies, and 71 implemen- for agricultural water management account for a tation completion reports, covering the Bank's quarter of all lending for agriculture and rural experience in 56 countries. In addition, it draws development--more recently, this amount upon the detailed findings from IEG's project increased to one-half. performance reports, country assistance evalua- tions, and several special sector and thematic This study undertook a detailed assessment of evaluations. The methodology used in the the design and performance of the agricultural evaluation is described in detail in appendix A. Vinod Thomas Director-General, Evaluation x i Executive Summary While no country has been able to decrease poverty through agricultural development alone, at the same time, no country ... has solved its problem of poverty without creating a dynamic agricultural sector (Timmer 2003). A t the turn of the millennium, irrigated land made up about a fifth of the arable area in developing countries, having doubled to about 200 million hectares since the early 1960s. Yet, this relatively small proportion of arable investments for infrastructure maintenance, will land produced 40 percent of all crops and close require better regulation and management. to 60 percent of cereal crops--but it also accounted for about 80 percent of all water use The World Bank has long been the largest source in developing countries. As the world's popula- of assistance for agricultural and rural develop- tion expands from its present 6.5 billion to a ment. This has included a range of structural and projected 8.2 billion in 2030, demand for agricul- nonstructural measures to harness, control, and tural production and, thus, water will increase. manage surface and ground water to improve agricultural production. These measures have Increased attention to efficient water manage- involved widely variable combinations of irriga- ment will be essential to meeting that demand. tion, drainage and flood control, water conser- Globally, water is an increasingly scarce vation and storage, on-farm water management, commodity--in the next 25 years more than a and more recently, institutional support to quarter of the developing world population will improve sustainability, user operation and face severe water scarcity. The rapidly growing management, and cost recovery. Collectively, numbers of urban and industrial consumers will these interventions are called agricultural be prepared to pay more for water than the water management (AWM). agricultural sector at present. The use of this limited resource will, therefore, require efficiency The Bank's engagement with AWM has evolved improvements and tradeoffs. Groundwater, the considerably since its first comprehensive sector main source of water associated with most strategy in the 1993 Water Resources Manage- private sector agricultural investment in South ment: A World Bank Policy Paper. Even so, the and East Asia, the Middle East, and North Africa irrigation and drainage and the natural resource regions, is already overexploited. Thus, more management subsectors were identified as high competition for water and the degradation of risk in the Bank's 1997 rural development existing supplies, owing to pollution and reduced strategy, Vision to Action (World Bank 1997a). x i i i WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 That strategy emphasized a supportive policy Since 1994 the average loan amount per AWM framework for projects, an enabling environ- project has fallen from $59 million to a low of ment for private sector development, and a $15 million in 2001. This was because of the participatory, decentralized approach to the increasing share of projects in the Europe and design and implementation of projects. It Central Asia Region; a move away from shifted the objectives and the design of agricul- freestanding irrigation and drainage projects to tural water projects from a narrow agricultural more general rural development or social fund- focus to a broader rural development approach. type projects, in which AWM was a minor Subsequently, the Bank's 2001 Water Resources component; and a marked reduction in new Sector Strategy highlighted that provision of construction in favor of rehabilitation of existing water infrastructure was an important infrastructure. component of growth. However, the details of water development strategy and business plans During the period 1994­2004 agricultural water for efficient and sustainable service provision projects directly benefited up to 12 million and management were delegated to the main households and more than 60 million people. water-using sectors, which for agriculture is the The average project served slightly less than Bank's Agriculture and Rural Development 115,000 farm households, mostly defined as Department (ARD). Subsequently, ARD issued small family farms, and covered an area of an updated rural development strategy in 2003, 134,000 hectares. The investments were Reaching the Rural Poor (World Bank 2003a), economically sound and averaged an economic which sought to focus the Bank's rural return of 22 percent. However, more recently, lending--including that for water--on extend- returns have declined to about 17 percent ing Bank endeavors to reach the rural poor. because of lower commodity prices, smaller incremental benefits, and overly optimistic Despite these strategies and the increasing appraisal of institutional constraints. demand for food and water, the proportion of total Bank lending to agriculture fell from 31 A recent impact evaluation by IEG in India's percent in the late 1970s to less than 10 percent Andrah Pradesh area has reaffirmed irrigation's in the early 2000s. Similarly, the share of Bank role in reducing rural poverty (IEG 2006a). lending for agricultural water management, after Irrigation increased net farm income by just over peaking at 11 percent of all Bank commitments 60 percent, about half of which came from in the 1970s, fell to less than 2 percent by 2000. increased cropping intensity and most of the Recently, following a strong drive from ARD it remainder from higher yields, with only a small has grown to 4 percent. part attributable to changes in the crop mix. However, irrigation had a very modest impact The total amount of Bank lending between 1994 on income distribution. The top quartile and 2004, for the 161 projects that included benefited the most in absolute terms, and the quantifiable agricultural water management second quartile benefited the most in relative components, was $13.2 billion, which went to 56 terms, experiencing income growth of 30 countries. Within this total commitment less percent. The poorest quartile experienced a low than half--42 percent or $5.6 billion--was benefit, but their already low income meant that specifically for agricultural water management they also experienced income growth of 20 components. Almost two-thirds went to South percent, compared with 19 percent for the top and East Asia, and half to China, India, Indone- two quartiles. Dynamic effects also had an sia, and Pakistan. Mexico, with only two impact on income distribution. Households operations, is the third largest borrower. This subject to repeated negative shocks became regional distribution follows the pattern of Bank heavily indebted and depleted their assets, lending established for agricultural management constraining their ability to undertake produc- since the 1970s. tive investments. Reducing the negative impact x i v E X E C U T I V E S U M M A R Y of bad years by irrigation thus aided asset education, health, and other social services accumulation and helped households grow out peaked at 31 percent ($5.8 billion) of total of poverty. lending in 2003 (compared with 12 percent in 1990), before falling back to 18 percent in 2005. Irrigation investments in Andrah Pradesh increased the demand for labor, particularly for Also contributing to the reduction in agricultural women. This increase in the demand for labor lending during the 1990s was the secular decline led to an increase in average wages of 5 percent in agricultural prices (owing to the success of for men and 10 percent for women because of the Green Revolution) and reductions in the high demand for female labor for weeding government involvement in agriculture (such as and harvesting paddies. In an IEG survey, input and credit subsidies). Dissatisfied with women accounted for 63 percent of agricultural previous public sector­led approaches to employment in 2005 and 64 percent in 2006. agricultural development, the Bank began to There is a considerable body of evidence that experiment with a more diversified menu of women's incomes have a larger impact on child subsector strategies, depending on the charac- welfare (health and education) than do men's teristics of each subsector and the level of each incomes. country's development. Central and state governments continue to fund research, Given the relative importance of AWM and the extension, and livestock services (because of the fact that the last comprehensive independent strong public-goods elements), while private evaluation of this subsector was in 1994, this sector investment tends to be associated with study set out to answer three questions: land markets, agricultural marketing, and rural finance. Local governments are a key to improv- · Why has Bank investment in agricultural water ing rural infrastructure, and local communities management declined so precipitously? to improving the management of renewable · Are agricultural water projects relevant to the natural resources, such as pastures, forests, and Bank's renewed focus on poverty alleviation fisheries, provided that incentives are in place and institutional and policy reform? such as harvest or property rights. · What should be done to improve perform- ance and relevance? Finally, internal Bank factors also played a role. Overall Bank budget constraints may also have Reasons for Reduced AWM Lending in contributed to reduced lending for agriculture the 1990s and AWM, because the Bank's administrative Following its 1990 World Development Report budget for lending preparation declined from on Poverty, the Bank adopted a strategy that $150 million in 1993 to $122 million in 2000; and targeted efficient, labor-intensive growth and economic and sector work declined from 13 greater attention to social concerns, including percent to 7 percent of the Bank's budget education and health care. With the stronger during the same period. (In real terms, the Bank focus on reducing poverty, lending to the social budget for lending preparation declined by 44 sectors increased while lending for infrastruc- percent between 1993 and 2001.) While the ture, agriculture, and the environment fell after administrative budget for lending preparation 1993. International Development Association has recovered since 2001, it was still only three- (IDA) replenishment agreements (IDA10­12) quarters of its 1993 value in 2005. In this context also required increases in the share of invest- of increased competition for scarce Bank budget ment lending in the social sectors, and the resources, the rural sector, and the AWM subsec- Heavily Indebted Poor Countries (HIPC) initia- tor in particular, were at a disadvantage because tive required beneficiary countries to allocate of the relatively high costs of preparation and funds freed up from debt service to public supervision, fiduciary, and safeguard concerns. expenditure on the social sectors. Lending to (Agricultural projects were a quarter more x v WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 expensive to prepare than the average Bank shortages is an increasing concern--particularly project, and AWM projects that required in China, India, Pakistan, Yemen, much of the additional preparation costs to meet the Bank's Middle East, and in a number of Central Asian safeguard policies were almost twice as countries. expensive.) A reduction in the number of techni- cal staff may also have contributed to reduced Irrigation boosts growth and reduces poverty sector lending. The introduction of new lines of directly and indirectly, benefiting the poor in business under the 1997 Strategic Compact several ways. Poor farmers directly benefit from (World Bank 1997b) brought about a change in increases in their production, which enables the skills mix of Bank staff--all of the targeted them to increase their own consumption or sectors, except rural and private sector develop- provide a surplus of marketed products, thereby ment, increased their staff. Within the rural increasing their farm income. Small farmers and sector there was a reduction in the number of landless laborers benefit from agricultural agricultural and irrigation specialists. (By 2002 employment opportunities and higher wages, there were only 16 irrigation and drainage and a wide range of rural and urban poor benefit specialists left in the ARD sector.) from related growth in the rural and urban nonfarm economy. Larger crop harvests from AWM Remains Relevant irrigated areas lead to strengthened staple and No country has successfully tackled rural nonstaple food output, which lowers prices and poverty without developing a dynamic agricul- benefits all consumers, particularly the poor. ture sector. In most of the Bank's client Even so, it is the "package" that matters for countries this is dependent on efficient water effective poverty alleviation and not just the management, good drainage, and flood protec- supply of irrigation water. Investments in agricul- tion. Sound water management increases tural water management may not reduce poverty agricultural productivity and this has substantial directly in any significant way unless accompa- positive impacts on rural employment and the nied by other complementary interventions. rural nonfarm economy. Although agriculture's contribution to growth and employment contin- The Importance of Agricultural ues to shrink as economies make the transition Development and Sound Water from agriculture and subsistence production to Management Is Increasingly more reliance on industry, processing, and Recognized in CASs services, IDA borrowers have consistently Evaluation of the Country Assistance Strategies placed the highest priority on infrastructure and (CASs) and projects approved during the period agricultural development in the Bank's client 1994­2004 shows a change toward a more surveys. In a recent IEG assessment in Madagas- comprehensive approach to rural development, car, for example, farmers reported that they were with a growing emphasis on building social able to send their children to primary school only capital. Project objectives encompassing after irrigation and road access to markets had community support and participation, income improved their incomes--the uptake of invest- and employment, and support for capacity ments in education was strongly conditioned on building and institutional development the impact of infrastructure investment. increased. Conversely, objectives that are central to the new policies--addressing poverty Many borrowers are seeking external support to reduction, agricultural development and produc- improve the productivity of agriculture through tion, and environment and natural resources private sector growth, agribusiness, better management--declined in importance. One communications, marketing and trade, and reason for these changes is that development improved input efficiency, particularly for water. objectives have become more practical and In some countries, managing agricultural water achievable by focusing on measurable outcomes x v i E X E C U T I V E S U M M A R Y rather than global targets. For example, portfolio. There is a marked difference in the increased attention to income and employment type of infrastructure components financed by almost offsets the decrease in poverty-reduction dedicated and nondedicated projects even objectives. though most contain a mix of physical interven- tions ranging from some new construction, Attention to the technical and social issues of redesign and upgrading, to repair of damage agricultural water management has become caused by deferred maintenance, referred to as more polarized. The more general projects, in rehabilitation. Among the dedicated projects, which water-related activities are in the minority, rehabilitation or improvement of large irrigation are building water infrastructure with less systems now account for more than 80 percent attention to issues of technical efficiency and of Bank commitments. Nondedicated projects, sustainability. This may not be an issue where after an initial focus on rehabilitation in the mid- agricultural water management projects are part 1990s, now support the construction of new of a broader package of rural development that systems that are small scale, community owned, deals with social, human, and economic and integrated in social development programs. development. But these findings indicate the Because rehabilitation projects averaged $2,900 importance of integrating agricultural water per hectare, while new construction projects management projects within country rural averaged $6,600 per hectare, there was a strategies and ensuring that they are adequately substantial fall in the cost of projects. As a result, supported either by parallel operations that the average loan amount per project fell from $59 address critical omissions, or by improving the million in 1994 to a low of $15 million in 2001. skills mix of appraisal teams preparing agricul- tural water management components of Dedicated irrigation and drainage projects with nonwater projects. policy content--large or small--only give broader policy issues modest attention. For While most CASs discussed the importance of about 20 percent of the cases reviewed, policy agriculture policy, less than half discussed it in was not addressed at all, either because it was the context of economic growth; greater no longer relevant or because it was being prominence was given in the CASs to addressed outside the project. Many of the community-driven development, general rural appraisal documents implicitly assume either development, and reform of agricultural institu- that policy reform is largely complete, or that it tions. In part this is the result of economic is beyond the project's scope--particularly in evaluation that neglects the analysis of growth cases where irrigation and drainage was only impacts and poverty-alleviation effects of invest- one of many components, or where the size of ment in AWM. the investment was small in relation to the norm for earlier periods. Yet, in many cases, important Projects Are Smaller and Cheaper policy issues remain. For example, in Brazil, the but Broader Policy Issues Have Been Bank's analysis shows that there is a need to Neglected increase the security and enforceability of water Low-cost approaches are increasingly important. rights, to introduce water charges that reflect The average Bank commitment to agricultural the economic value of water, and to clarify the water management projects declined for two roles and responsibilities of institutions. These reasons: a change in the type of infrastructure recommendations are valid for many of the financed and the greater emphasis on nonstruc- Bank's clients, particularly in countries such as tural and capacity-building components. India, Jordan, and Mexico. This clearly indicates Freestanding projects dedicated to water that a comprehensive approach--as opposed to management now comprise only about 40 an irrigation-led one--is required for agricultural percent of the agricultural water management development in developing countries. x v i i WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 Better Private Sector Participation Is later--the primary reason for the improvement Needed in the design of monitoring and evaluation was While the principle of user participation to the introduction of logical frameworks in the improve management of public sector irrigation late 1990s and their mandatory use in project projects remains valid and is still widely appraisal documents. While the overall quality supported, farmers often lack the skills needed of indicators improved, only a fifth of sampled to manage the larger irrigation systems, and the projects had good poverty output indicators. need for continuing government support has been underestimated. Projects have tended to Only 11 percent of projects were designed to give more emphasis to strengthening water user have the tools that would allow rigorous impact associations than to strengthening the broader assessment: this includes well-defined output authorization and institutional framework in and outcome indicators, good baselines, and which they must function. Cost-recovery targets independent control groups unaffected by have been wildly ambitious and unrealistic project interventions that would allow the because of inadequate social assessment. And counterfactual to be determined. Another 41 frequently, essential credit, inputs, extension, percent would allow determination of what and marketing linkages have been neglected. happened before and after project implementa- Projects also have not planned for the gradual tion, but not a robust attribution of observed phasing out of support as the user groups changes. Slightly fewer than half the projects did mature. not have any means of verifying project impacts--no surveys or baselines--even though Simultaneous attention to community operation more than two-thirds of them included outcome and management, and physical modernization or impact indicators. of water distribution networks has not been very common, reducing the efficacy of both interven- Key Indicators Are Infrequently Reported tions. Where this has been done, the results can Outcomes from 71 projects in the portfolio that be outstanding, as shown in China's Tarim Basin have been completed reveal that while all of and in Armenia. Where the potential synergy has them provide qualitative accounts of policy or not been captured, the outcomes have been institutional outcomes, less than half can define disappointing. quantifiable outcomes and impacts. There are three reasons for this. First, almost a third of the The complementarity among irrigation invest- projects (20) could either define benefits only ments and extension, marketing, and credit very generally (for example, the community- services can be improved, particularly for driven development projects) or very narrowly, dedicated projects. While there was a big such as the six output-oriented emergency- increase in the share of irrigation projects that disaster recovery projects. Second, planning and addressed credit and marketing constraints after setting up of monitoring and evaluation is poor. 1998, most of this increase derives from nonded- And third, very often there is a lack of relevant icated projects. indicators because the results chain linking inputs to outputs and impacts is either weakly Monitoring and Evaluation Was Poor developed or missing. but Is Improving Slowly Throughout the study period there was system- In comparison with the rest of the Bank, the atic improvement in the overall quality of rural sector is more assiduous in carrying out monitoring and evaluation systems. The overall economic evaluations, and more projects-- annual average rating increased from slightly about half--are reevaluated at completion. In above modest in 1994 to substantial 10 years the agricultural water subsector, there is a partic- x v i i i E X E C U T I V E S U M M A R Y ular need for more attention to broaden use and should be added to the benefit stream. projects' economic analysis to demonstrate Capturing these effects is clearly very difficult growth and poverty impacts, thereby increasing but options for some assessment could be project relevance to Bank country directors and developed from the analysis of earlier experi- ministers of finance. And in most projects, the ences and the literature. Better economic impacts of capacity building and institutional analysis would also help the selection of the most reform are not factored into the benefit stream. relevant project objectives and components, and This is particularly important because the help simplify project design, thus avoiding economic efficiency of rural projects (based on projects overloaded with too many objectives. more easily measured impacts such as incremental crop production) has been less Give more attention to enhancing water use than most other sectors in the Bank, and it has efficiency. Agriculture uses 70 percent or more been declining. of all water resources in the Bank's client countries. Increasing water use efficiency will Reports on how many people benefit, their become increasingly important as inter-sectoral social status, and what benefit they realize are competition for scarce water increases. More not very common. Slightly under half of the attention to irrigation system modernization is projects report how many farmers benefit, but needed to ensure that systems designed for top- less than a fifth report how many people this down supply management are redesigned with affects or the social distribution of benefits. the provision of appropriate volumetric measurement for demand management by There Is Potential for Increasing water users. Simultaneously, more care should Relevance and Performance be given to developing supporting institutions Demonstrate growth impact though better and incentive structures for water user groups. economic analysis. The relevance of agricul- Greater social assessment and financial capacity- tural water management operations to borrow- building is generally required. ers and to Bank country directors could be increased through better analysis of links to Match sector staffing to needs. There is ambiva- economic growth and more attention to lence about technical staffing in the agricultural demonstrating social impact and poverty sector, and for irrigation and drainage, brought reduction. More explicit and thorough results about by the Bank's preference for less techni- chains are needed in project design. Current cally demanding operations and more fungible project economic analysis is typically limited to generalists. If better economic analysis leads to input and production impacts with almost no increased demand for AWM projects, then attention to modeling employment, poverty attention will have to be given to recruiting staff alleviation, growth, environmental, and institu- that can deliver technically sound and relevant tion-building impacts because it is often projects--a pressing issue because about half of regarded as impractical to quantify their the senior technical staff in AWM will retire in benefits. Consequently, these potential benefits the next two to five years. are omitted. Provide sufficient resources. Increasing the Currently, the value of water saved through contribution of agricultural water investments more efficient agricultural use is neglected if that to economic growth and poverty reduction may water is not used to expand agricultural produc- require additional Bank resources for front-line tion. Yet, increased urban demand puts a much development. If it is demonstrated that higher value on water. This is clearly a benefit inadequate or inefficient agricultural infrastruc- derived from more efficient agricultural water ture hinders economic growth and slows x i x WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 poverty reduction, then the Bank may have to prepare. The distorted incentive structure that consider increasing resources for project this creates when budgets are tight should be preparation and economic, sector, and advisory recognized and rectified. This is particularly work. The Bank should not continue to allow important because the agricultural sector economically justified lending to agriculture to provides most of the employment for the rural languish just because projects are expensive to poor. x x 1 Introduction S ince its inception the World Bank has been the world's largest source of assistance for agricultural and rural development. This has included a whole range of structural and nonstructural measures to harness, con- trol, and manage surface and ground water to improve agricultural production. Typically these measures include widely variable to grow high-yield seed varieties and to apply combinations of irrigation, drainage and flood adequate plant nutrition as well as pest control control, water conservation and storage, on- and other inputs, thus increasing yields. In farm water management, and more recently, successful irrigation projects, for example, institutional support to improve sustainability, conversion from rain-fed to irrigated crops user operation and management, and cost- typically increases crop yields twofold or more; recovery. Collectively, these interventions are and, using either reservoir or groundwater termed agricultural water management (AWM). storage, it may be possible to extend the normal Bank lending for AWM peaked in the 1970s growing season and produce two or more (figure 1.1). irrigated crops each year in warm climates. Effective irrigation and drainage contribute to In 1998, irrigated land made up about one-fifth food production, generate rural employment, of the total arable area in developing countries, and raise the incomes of farmers. Successful having doubled in size since the early 1960s to agricultural water management minimizes reach about 200 million hectares. Despite the production risks (drought, poor drainage, and relatively small proportion of total arable land flooding), boosts output, and provides irrigated, this produces 40 percent of all crops incentives for farmers to invest in other inputs and close to 60 percent of cereals--it also and agronomic improvements. In high-risk accounts for about 80 percent of all water use in environments, farmers tend to rely on robust developing countries.1 The Food and Agricul- but low-yielding varieties of crops and are tural Organization of the United Nations (FAO generally unwilling to invest in fertilizer and 2003) estimates that since the early 1960s more other inputs because of the risk of failure and than 70 percent of agricultural production loss. Conversely, good agricultural management increases in developing countries have been the makes it economically and financially attractive result of yield increases, much of it associated 1 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 water irrigation from canals and tanks and Figure 1.1. Bank Lending for AWM drainage projects have been built primarily through public investment. Groundwater abstraction, initially developed through public 12 expenditure on large well fields, is now financed 10 mainly through private investment because of its small scale and manageability (Barker and (percent) 8 Molle 2004). Because an integrated approach to water resources development is needed to lending 6 minimize harmful externalities and resource Bank degradation, government has a major regulatory and planning role. total 4 of 2 Expansion of private investment has also Share increased the pool of resources available for 0 agricultural water management. However, 1950s 1960s 1970s 1980s 1990s 2000s private sector irrigation development has been Source: World Bank data. limited mainly to groundwater and, to a lesser extent, smaller commercial surface water with irrigation and improved on-farm water systems growing high-value crops. Two-thirds of management. In the Philippines, for example, groundwater irrigation in India and Mexico is yields of irrigated rice are twice those of rain-fed privately managed.5 In India and elsewhere in varieties. South Asia, and in Latin America and the Middle East and North Africa, public investments have Poor drainage is as much an impediment to crop facilitated private irrigation investment. In South production as insufficient water and about half Asia, private tubewell irrigation systems have the world's irrigated land suffers from drainage grown most rapidly in areas with reasonably problems.2 Twenty-five million hectares of prime good roads, research and extension systems, agricultural land have become unproductive access to credit, and electric or diesel energy. As due to irrigation-induced waterlogging and a result, these have been concentrated in and salinity.3 Two hundred and fifty million hectares around the command areas of large, publicly of rain-fed cropland need improved drainage. developed surface irrigation systems. Improved drainage can also produce substantial benefits to health, reduction of damage to roads World Bank Strategies for AWM and buildings, and flood control. In Egypt, The first Bank water strategy, presented in subsurface drainage increased the annual net "Water Resources Management: A World Bank income of the traditional farm up to $375 per Policy Paper" (1993), evolved in response to hectare.4 In the Mardan project area of Pakistan, growing unease within the Bank that water crop yields increased between 27 percent and operations were failing to deliver sustainable 150 percent. In Mexico, economic rates of return development, and growing international of the drainage subprojects, based only on the concern about the mismanagement of global changes in agricultural yields, ranged from 15 water resources and poor service levels, particu- percent to 22 percent. larly for the poor. When the strategy was issued, water-related projects were among the poorer Investment in agricultural water management performers in the Bank's portfolio. This was requires substantive support from the govern- emphasized by the influential Wapenhans ment and the private sector in order to attain its Report (World Bank 1992), which was highly full efficiency. Large- and medium-scale surface critical of the quality of the Bank's water lending 2 I N T R O D U C T I O N based on the findings of sector reviews of water evaluation of this strategy found that the urban supply and sanitation, irrigation and drainage, bias often shown by governments, and and trends in project outcome ratings.6, 7 sometimes echoed by the Bank's country directors, continued to impede selectivity and The strategy paper recognized that improving strategic mainstreaming.9 Despite increasing performance in meeting water needs requires support for sector investment programs, borrowing countries to reform their water progress with sector and sector-level policy and management institutions, policies, and planning institutional reforms remained slow, restricting systems. It also acknowledged that this would rural growth in many countries. While the require changes to the Bank's internal strategy helped to improve the design and processes, training, skills mix, and resources implementation of rural projects, it was not assigned to water and water-related operations. adequately incorporated into CASs and there The main recommendation was that a new remained room for improvement in knowledge approach--recognizing that water is a scarce management, training, and monitoring. natural resource, subject to many inter-depend- encies in conveyance and use--be adopted by Agricultural water management received minor the Bank and its member countries. Specifically, attention in the Vision to Action strategy. Major the aim of the strategy was to maximize the actions proposed included improved inter- contribution of water to countries' economic, sectoral coordination and planning of all water- social, and environmental development while using sectors, resolution of water allocation ensuring that resource and water services are disputes among sectors and countries, support managed sustainably. This was to be achieved for efforts to decentralize irrigation manage- through the establishment of comprehensive ment based upon water user associations and analytical frameworks to foster informed and greater attention to sustainability of infrastruc- transparent decision making, with an emphasis ture, including full recovery of operation and on demand management, promotion of maintenance costs. decentralized implementation processes, and use of market forces to guide the appropriate Independent evaluation of the Bank's 1993 mix of public and private sector provision of Water Resources Management Strategy water services. conducted during the period 1999­2000 did not assess the efficacy of Vision to Action on agricul- Integrating the Strategy with tural water management.10 Generally, agricul- Rural Development tural water management performed poorly on The Bank's 1997 rural development strategy, compliance with the main elements of the water Vision to Action (World Bank 1997a), attempted strategy compared with the other water-using to shift the Bank from a narrow agricultural sectors--energy, environment, and water supply focus to a broader rural development approach, and sanitation. Even so, agricultural water integrated with country assistance strategies projects were superior on social assessment and (CASs), with particular focus on 18 countries.8 participation but inferior on institutional and Four subsectors--agricultural research and financial issues, and their supervision was more extension, forestry, irrigation and drainage, and problematic. Among the recommendations was natural resource management--were identified that more attention was needed on loss of skills as high risk. To address the risks the strategy through net loss by retirement of experienced emphasized a supportive policy framework for water sector staff; better guidelines were also projects, an enabling environment for private needed on best practice. sector development, and a participatory, decentralized approach to the design and Subsequently, the Bank's new Water Resources implementation of projects. Independent Sector Strategy, finalized in 2001, confined itself 3 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 to the broader policy issues and new directions development, which is the main engine for for water resources management, deferring broad-based rural development. While detailing and delegating the detailed strategy and priorities for Bank assistance--including business plans for efficient and sustainable tackling the well-known systemic institutional service provision and management to the main problems of irrigation and drainage--Reaching water-using sectors, in this case, agriculture.11 the Rural Poor promotes several new initiatives. The 2001 strategy identified the falling invest- New approaches to irrigation that yield benefits ment in hydraulic infrastructure as a critical disproportionately to the poor are promoted, as failing of the Bank's lending in the 1990s. Thus, is greater attention to monitoring and evalua- it proposed institutional support to facilitate tion to ensure accountability of monopoly development of "high risk­high reward" service providers in irrigation, and linking irriga- infrastructure, in recognition of the controversy tion reform to broader issues of development surrounding large-scale investment to harness and political economy. Overall, the approach is water resources. to encourage coherent and integrated rural development that is more closely aligned with Renewing the Focus of the Rural Strategy Poverty Reduction Strategies Papers and CASs. The renewed rural strategy, Reaching the Rural But Reaching the Rural Poor's prescriptions for Poor, was launched in 2002 to realign the Bank's improving agricultural water management are declining rural lending with the incidence of not based on systematic review of the Bank's poverty (World Bank 2002b). While three- experience or a detailed evaluation of implemen- quarters of the world's poor reside in rural areas, tation experience. the proportion of total Bank lending to agricul- ture fell from 31 percent in the late 1970s to less Reaching the Rural Poor succeeded in refocus- than 8 percent in the early 2000s but has since ing the Bank on improving its approach to shown a strong resurgence (figure 1.2). The new agriculture and agricultural water management strategy focused on creating an investment and better equipping of its staff to do the job. climate conducive to rural growth and empower- This has not only rejuvenated lending but also ing the poor to share in the benefits of growth. produced two comprehensive internal Bank evaluations of the subsector and its perform- Reaching the Rural Poor stressed the centrality ance--Agricultural Growth for the Poor: An of sound water management to agricultural Agenda for Development (2005c) and Reengag- ing in Agricultural Water Management: Figure 1.2. Agriculture Sector and AWM Lending in Challenges and Options (2006). In addition, Total Bank Lending Bank staff were supported in implementing these new directions through development of 14 % Agricultural sector practical guidelines Shaping the Future for Water in Agriculture: A Sourcebook for Invest- 12 % Agricultural water ment in Agricultural Water Management management 10 (2005d). 8 Objectives of the Study Percent 6 With less than three years of implementation experience, it is too early to evaluate the efficacy 4 of the rural development strategy laid out in 2 Reaching the Rural Poor or of the subsequent detailed guidance to Bank staff. However, more 0 1992 1994 1996 1998 2000 2002 2004 2006 than a third of the Bank's rural investments Source: World Bank data. between 1994 and 2004 dealt with agricultural 4 I N T R O D U C T I O N water management. Given the relative · Why did Bank investment in agricultural water importance and specialized nature of the management decline so precipitously? subsector, and that the last comprehensive, · Are agricultural water projects relevant to the independent evaluation of this subsector was in Bank's renewed focus on poverty alleviation, 1994, it is appropriate to evaluate the perform- and on institutional and policy reform? ance of agricultural water management since · What should be done to improve perform- then. The main study questions are: ance and relevance? 5 2 AWM Relevance to the Bank Declined T his chapter demonstrates that the commitment to agricultural water man- agement declined in the 1990s as the Bank's development agenda fo- cused more on social, human development and environmental concerns, and good governance, and because it was seen in the Bank as less relevant to the needs of borrowers. Overview waned. Since 2002, budgets and staffing have Although borrowers have become more modestly improved and, refocused by the new concerned with issues of urbanization and social rural and water sector strategies, lending for development, particularly as food security is no rural development and AWM has resurged. longer a concern for most countries, infrastruc- ture and agriculture remain at the top of their Within agricultural water management there has developmental priorities. Within the Bank, the been a strong trend toward more general 1997 action plan, Strategic Compact: Renewing agricultural development projects that are more the Bank's Effectiveness to Fight Poverty closely aligned to the Bank's objectives and to (1997b), significantly reduced budgets for agricultural sector strategies, that focus more on project preparation, a trend accelerated by a poverty alleviation, human and social develop- substantial shift toward development policy ment, and capacity building. At the same time, lending during the mid- to late 1990s. At the attention to environmental issues has steadily same time, the skills mix of Bank staff was fallen and, within the fewer and more special- realigned to the Strategic Compact, resulting in ized AWM projects, the approach also has a loss of technical staff and their replacement become more technically focused. More general with staff having more fungible skills. Enhanced AWM projects--many of which use community- fiduciary and safeguard provisions increased the driven development--are building water costs of project preparation such that AWM infrastructure with less attention to issues of projects became among the most expensive to technical efficiency and sustainability. These prepare. Squeezed by budget pressures, high findings indicate the importance of integrating costs, muted advocacy, and new development AWM projects within country rural strategies and initiatives, country directors' interest in AWM ensuring that they are adequately supported 7 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 either by parallel operations that address critical Figure 2.1. Most Bank AWM Investment omissions, or by improving the skills mix of Went to Asia in 1994­2004 appraisal teams. It is important to realize that there is no "ideal" AWM package--country EAP needs and preferences should drive project 28% design, be it entirely a water project or part of a more general or sequenced approach to rural development. ECA 11% AWM Investment Shrank The proportion of total Bank lending to agricul- ture fell from 31 percent in the late 1970s to less LAC than 10 percent in the early 2000s. Similarly, the 10% SAR share of Bank lending for AWM, after peaking at 36% 11 percent of all Bank commitments in the 1970s, was less than 2 percent in the early 2000s. MNA 9% Recently it has grown to 4 percent. The size and AFR composition of the AWM portfolio also changed 6% in response to borrowers' preferences and Source: World Bank data. evolving Bank strategies for poverty alleviation and agriculture. thirds went to the South Asia and East Asia Regions (figure 2.1), with 51 percent to China, Globally, the international financial institutions India, Indonesia, and Pakistan. After India and are small players in agricultural water manage- China, Mexico, with only two operations, is the ment. In the 1990s, the irrigated area in develop- third-largest borrower, accounting for almost 8 ing countries expanded at an average rate of 2.5 percent of the total amount committed. This million hectares a year and increased the stock regional distribution follows the pattern of Bank of irrigated land to 207 million hectares by 2000. lending established for irrigation and drainage This represents an annual investment of about since the 1970s (IEG 1995). Within this total $36 billion in current prices, including the costs commitment, less than half (42 percent or $5.6 of operations and maintenance (O&M) for past billion) was specifically for agricultural water irrigation investments. In the same period, management components. international financial institution investment averaged $1.6 billion a year, 4 percent of global In practice, the Bank allocated more to AWM, investment, half of which came from the World but this is difficult to quantify. Although a textual Bank.1 The balance of global investment came search of the Bank's data for 1994­2004 found mostly from the public sector, from private that 371 projects included some discussion of investment in small-scale, mainly groundwater- AWM, only 161 could be quantified--these form based irrigation, and small contributions by the portfolio used for this evaluation. For bilateral development partners--there are, consistency, the lending for agricultural water however, no reliable global estimates of their components was estimated from the cost tables relative contribution.2, 3 in project appraisal documents (PADs). However, the subsequent random sampling of The total amount lent by the World Bank PADs and more detailed cost allocation found between 1994 and 2004 for the 161 projects that that in several projects the agricultural water included quantifiable AWM components was component was almost 20 percent more costly $13.2 billion. This represents almost 6 percent than those identified using only the sector code. of the total Bank commitments during the The remaining 210 projects that mention AWM period and it went to 56 countries. Almost two- were not analyzed further for two reasons. First, 8 AW M R E L E VA N C E T O T H E B A N K D E C L I N E D 153 of the projects were either social funds or public/private sector, and rural development. community-driven development (CDD) types of Subsequently, the Bank's 2000/2001 World intervention that did not make an a priori Development Report: Attacking Poverty, allocation of the loan amount because it was not advocated an increased focus on public sector known what interventions the beneficiaries governance and institutional reforms that would would choose.4 Second, the remainder had only empower and foster participation of the poor in very small, if any, AWM components that would the conduct of public institutions and delivery not be cost effective to assess. of public services (IEG 2004a). During the period 1994­2004, the average loan With the renewed focus on poverty, lending to amount per AWM project consistently fell from the social sectors increased while that for $59 million in 1994 to a low of $15 million in infrastructure, agriculture, and environment fell 2001. Since then it has recovered (figure 1.2). after 1993. International Development Associa- While there is high variability from year-to-year tion (IDA) replenishment agreements in the number of project approvals, averaging 15 (IDA10­12) required increases in the share of a year, they exhibit no significant trend over investment lending in the social sectors, and the time. The amount of lending declined until Heavily Indebted Poor Countries (HIPC) Initia- 2003, when it recovered as a result of two large tive required beneficiary countries to allocate projects in Mexico and Vietnam and several large funds released from debt service to public projects in India. Five factors explain the falling expenditure on the social sectors. Lending to lending for AWM in the portfolio: (i) realignment education, health, and other social services of lending with the Bank's strategy for assisting peaked at 31 percent ($5.8 billion) of total poverty, (ii) a shift in borrowers' priorities, (iii) a lending in 2003 (compared with 12 percent in growing emphasis on integrated rural develop- 1990), before falling back to 18 percent in 2005. ment, (iv) changing development objectives in AWM, and (v) increased use of low-cost The Compact Had Unintended approaches. Consequences The Compact had an adverse impact on Aligning Overall Bank Lending with overall Bank lending, and on lending for all Its Poverty Strategy infrastructure--the rural sector and agriculture Following the 1990 World Development Report in particular. on Poverty, the Bank adopted a two-pronged strategy that targeted efficient, labor-intensive Infrastructure lending declined from a peak of growth and greater attention to social concerns, $10.3 billion (44 percent of the Bank's lending including education and health care. This was portfolio) in 1993 to as low as $5.1 billion (26 put into effect through the Bank's 1997 plan, percent) in 2002. The largest declines were in Strategic Compact: Renewing the Bank's electric power and water supply. In the mid- Effectiveness to Fight Poverty, as part of the 1990s, before the Asian financial crisis, rapid systemic reforms introduced by President James growth in the volume of private sector invest- Wolfensohn. The Compact focused on ments was widely expected to continue implementing four priority-change programs: (i) unabated. Coupled with serious concern about refueling current business activity to improve the environmental and social impact of several client services, (ii) refocusing the development large-scale projects--particularly those for agenda, (iii) retooling the knowledge base, and water--this caused the Bank to reduce its (iv) revamping institutional capabilities to support for public investments in civil works. support a more agile, creative, and client- With the introduction of the Infrastructure orientated work environment. Sectors and Action Plan in 2003, however, infrastructure themes highlighted for attention were environ- lending is now recovering, reaching $6.9 billion ment and social, finance, human resources, (32 percent) in 2005. 9 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 lending, harmonization and streamlining of Figure 2.2. Bank Resources to Prepare Projects information and knowledge management Declined in the Late 1990s systems, and improved client services5--these were at the expense of Bank budgets for lending 50 Lending preparation Project supervision 41 preparation, which declined from $150 million (%) Other client services in 1993 to $122 million in 2000, and for economic and sector work, which fell from 13 percent to 7 percent of Bank budget in the same budget 26 20 period. While the decline in budget for lending Bank 17 preparation bottomed out in 2001, in 2005 it of 13 14 8 was still below the level of 1993 (figure 2.2). In 8 Share real terms the Bank's budget for lending preparation declined by 44 percent between 0 1993 and 2001.6 Even so, the total number of FY93 FY96 FY00 FY05 Bank loans fell by only 13 percent in the same Source: World Bank data. period as a result of efficiency improvements and a preference for smaller, less risky projects Budget Constraints Led to Reduced Lending in sectors other than infrastructure and rural The additional financing to implement the development.7 Compact--$250 million--was agreed to on the basis that efficiency savings would enable the Overall budgets for lending preparation, Bank to be more effective in 2001 but on the however, remain tight. While the budget for same overall administrative budget as 1997. lending preparation recovered by 2005, it was While there were major achievements--the still only three-quarters of its 1993 value. And decentralization of many directors to country while use of trust funds supplemented project offices, an increase in the number of satisfactory preparation costs, these have declined Bank- projects, reduced processing time for Bank wide from a third of overall cost in the period 1999­2001 to 16 percent in 2005. Generally, the agriculture and rural sector has attracted more Figure 2.3. AWM Projects Are among the Most Costly trust funds, but their contribution to overall to Prepare preparation costs fell from about 40 percent to about 25 percent between 1999 and 2005. 600 Competition for an increasingly scarce Bank budget reduced country directors' interest in the rural sector and the AWM subsector because thousands) 400 of the relatively high costs of preparation and (US$ supervision, fiduciary and safeguard concerns, cost and performance issues. In that sense the opportunity cost of continued investment in a traditional line of Bank business was high 200 preparation because new development initiatives promoted AWM under the Compact were much cheaper to verageA Development policy lending prepare, were seen as less risky, and delivered Other investment results more quickly.8 And preparation costs of 0 AWM projects are among the highest in the 1994 1996 1998 2000 2002 2004 Bank--even compared with development policy Source: World Bank data. lending --and in a time of tight budgets this led 1 0 AW M R E L E VA N C E T O T H E B A N K D E C L I N E D to reduced lending to the subsector (figure Table 2.1. Cost of Safeguard Policies Is High, 2.3).9 For example, in 2001 the average new 1994­2005 (US$ thousands) lending per sector staff was $22.5 million for public sector governance, $15.5 million for Average Plus finance, and $6.2 million for rural development. Preparation Safeguards Only education and health had similar cost-to- Sector Board Cost Partial Full lending ratios. Clearly, if only lending volume Rural Sector 447 445 655 was a priority, rural, education, and health sector Water Supply and Sanitation 437 432 492 projects would not be selected unless they could Energy and Mining 401 403 553 be made much larger. While the moral impera- Urban Development 372 379 588 tive and international pressure to lend for Environment 349 355 483 education and health is strong, this is not the Transport 348 360 381 case for agriculture--even though most of the Private Sector Development 329 347 320 poor live in rural areas. Yet leaving countries Source: World Bank data. prone to droughts, floods, and famines because the projects are too expensive to prepare is not a defensible option for the Bank as the lender of Investment lending was also reduced by the rise last resort. in the importance of adjustment lending during the 1990s.13 From about a quarter of total Bank The primary reason for increased preparation lending in 1990, adjustment lending--now costs of rural, AWM, and infrastructure is called development policy lending--steadily because their size and spatial impact may increased to more than half of all Bank lending adversely affect the environment, human settle- by 1999 and stayed around that level until 2002. ments, and employment. Normally they include Although it declined subsequently it is expected substantial contracting and procurement. to increase with the growing preference within Consequently, when the Bank's fiduciary and the international financial institutions for safeguard policies are invoked, preparation general budget support operations.14 costs rise (table 2.1). Agriculture's Shrinking Contribution The Compact's introduction of new lines of to Growth and Employment business brought about a change in the skills The Bank's approach to attacking poverty laid mix of Bank staff (table 2.2).10 All of the targeted out in its 2001 Strategic Directions paper was the sectors, except the rural and private sector, rationale for strategy articulated in Reaching the significantly increased their staff. Within the Rural Poor (2002b). Underpinning the focus on rural sector there was a reduction in the number of agricultural and irrigation staff because they did not fit the new requirements for less techni- Table 2.2. Changes in the Bank's Skills Mix, cally specialized and more fungible skills--in 1997­2005 (number of professional staff) 2002 there were only 16 irrigation and drainage Sector 1997 2000 2005 staff in the Agriculture and Rural Development Department (ARD).11 The primary means of Human Development 342 516 542 downsizing was early retirement without Rural 289 314 309 replacement. From 425 staff in 1985­86, Private Sector 158 164 151 numbers in the rural sector declined to 330 in Environment 139 199 207 1996 and 309 in 2005. A survey of rural sector Finance 115 186 147 managers indicated that 330 staff was the critical Social 53 129 145 threshold they needed to discharge their Source: World Bank, Human Resources Department data. Professional staff includes headquarters sectoral responsibilities.12 and country office staff at grade GE and above. 1 1 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 poverty was growing evidence that accelerating were mostly assuaged. Declining prices of economic growth was the fastest way to raise staples--particularly the irrigated rice that people out of poverty--a 1 percent increase in accounts for the majority of food grains per capita income reduces the proportion of consumed--helped by improved nonwater people living on less than a dollar per day by an inputs, markets, and trade, increased the food average of 2 percent because economic growth access of the poor. Among the Bank's borrow- per se does not systematically affect the distribu- ers, food security remains an issue in a number tion of income (Revallion 2001, 2004, IEG 2005). of Sub-Saharan countries (for example, Ethiopia, While most of the poor reside in rural areas, Kenya, Madagascar, Malawi, Mali, Mozambique, however, sectors other than agriculture and Zambia), in postconflict Afghanistan, and in frequently offer more rapid economic growth. Indonesia as a consequence of natural and economic shocks. Thus, in most developing Agriculture's contribution to growth and countries agricultural production has met employment continues to shrink as economies performance expectations. Even so, many are make the transition from agriculture/subsis- seeking external support to improve the tence to more reliance on industry, processing, productivity of agriculture through private and services. This shift of economic focus has sector growth, agribusiness, better communica- lowered policy makers' attention to agricultural tions, marketing, and trade. policy and water management. Between 1980 and 2000 the share of agriculture in the world's Dramatic growth in the urban populations of gross domestic product (GDP) fell from 7.9 to 5 developing countries is posing severe percent, but the decline was much greater for economic, political, and social challenges that the regions where most of the AWM infrastruc- have displaced the attention given earlier to ture is located. It fell by a quarter to 25.1 percent rural development. Rural populations are antici- in South Asia and halved in East Asia and Pacific pated to decline slightly from 3.3 billion in 2003 to only 12.6 percent. The only region where the to 3.2 billion in 2030,17 and agricultural employ- share of agriculture in GDP remained ment will continue to contract--as discussed unchanged was Sub-Saharan Africa (17 percent), earlier, agriculture employment has declined in and in the Middle East and North Africa where it the Philippines, Egypt, and Brazil in the past two actually increased--from 10.3 to 14.3 percent. decades. Therefore, while 48 percent of the In comparison, among all low- and middle- world's population lived in urban areas in 2003, income countries, agriculture's value added to this is projected to rise to 61 percent by 2030. Of growth in the 1990s shrank to about half of that the two billion new urban dwellers, almost all added by the industry and service sectors and will be in developing countries. about a quarter of that added by exports of goods and services.15 Between 2005 and 2050, eight developing countries are expected to account for half of the Agricultural employment has become less world's projected population increase: India, important. In the Philippines, for example, Pakistan, Nigeria, the Democratic Republic of between 1980 and 2000, agricultural employ- Congo, Bangladesh, Uganda, Ethiopia, and ment fell from 60 to 47 percent.16 In the same China (in order of the size of their contribution period, Egypt saw a decline from 45 to 29 to population growth). Four of the eight percent while in Brazil it fell from 34 to 26 (Bangladesh, China, India, and Pakistan) are percent. among the world's leading irrigation states. Increased urbanization will require more water Borrowers Are Becoming Focused supplies and it is expected that this water will on Urban Challenges come from the increased efficiency of agricul- Food security concerns that were the focus of tural water use in conjunction with less agricultural development in the 1960s to 1980s pollution that reduces water resources. 1 2 AW M R E L E VA N C E T O T H E B A N K D E C L I N E D Demand Is Still High for Assistance political realities, it is too bureaucratic, and it to Agriculture does not explore alternative policy options. In global polls covering 55 countries during the While most of the sectors targeted by the Strate- period 2002­04 Bank clients gave the highest gic Compact align with priorities identified by priority to agriculture where they perceived the clients, it is clear that agricultural development Bank to be moderately effective (figure 2.4), is the exception in terms of investment and though International Bank for Reconstruction Bank resources allocated to this sector. and Development (IBRD) countries, as expected, gave this a much lower priority than did IDA AWM Is Now Less Focused on countries. The next highest priority was for New Investment infrastructure, in which the Bank received the Escalating construction costs for dams and highest scores for effectiveness, followed by related irrigation infrastructure, together with environment, financial systems, economic concerns over environmental impacts and growth, and poverty reduction. One interpreta- involuntary resettlement, reduced demand for tion of this sequence is that IDA governments new AWM infrastructure.18,19 Irrigated environ- put the highest priority on public service ments also faced increased competition for infrastructure, agriculture, and sound financial water and reduced water supplies owing to management to enable growth. The Bank also system degradation and reduced infrastructure needs to improve its effectiveness in many areas, development. particularly poverty reduction and assistance to improve public sector performance and good Poor performance and maintenance problems governance. The Bank's greatest weaknesses--a of public sector irrigation induced disillusion- consistent trend seen in all the client surveys ment among governments and policy makers, throughout the period 2002­05--is that it thus quelling interest. Much of this was because imposes technical solutions without regard to the rapid expansion of irrigation infrastructure Figure 2.4. Borrower Demand for Agricultural Investment Remains High 3.58 Infrastructure effective 3.48 Financial system More 3.38 Environment 3.28 Education Private sector Economic growth Regulatory framework Corruption Agriculture effectiveness 3.18 Health sector Gender Natural resources management Coordinated strategy bank 3.08 Transparency in governance Reduce poverty Social protection Digital Mean Divide 2.98 Public sector 2.88 effective Judicial system Less 2.78 3.76 3.86 3.96 4.06 4.16 4.26 4.36 4.46 4.56 Lower priority Mean importance Higher priority Source: World Bank Client Survey, data aggregated from more than four years, more than 10,000 decision makers, and 55 countries. The survey indicates how clients perceive the relative importance and effectiveness of the Bank's development interventions, on a scale from 1 (negligible) to 5 (high). 1 3 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 preference will be given to public investments in Table 2.3. Developing Countries: Annual Investment basic water supply, sanitation, and environmen- for Water Services tal needs because agriculture is regarded as Annual Cost of Investment primarily a private sector activity. And financially (US$ billions) it is projected that demands for water supply Investment sector 2000 2002­25 and sanitation will far outweigh the investment Municipal Wastewater Treatment 14 70 needed for continued growth of AWM (table Agriculture 32 40 2.3). Diminishing attention to agriculture among Industrial Effluent 7 30 the Bank's IBRD borrowers was, therefore, Sanitation and Hygiene 1 17 consistent with increased attention to social issues, urbanization, and growth. Drinking Water 13 >13 Environmental Protection 7.5 10 Emphasis on Comprehensive Rural Total 75 180 Development Is Growing Source: Camdessus Panel 2003. Borrowers' preferences and Bank policies and strategies are translated into country-specific, outpaced effective public management and local three-year rolling action plans through CASs institutions.20 Therefore, by the early 1990s developed in collaboration with governments institutional development and reform, benefici- and civil society. An evaluation of the CASs in the ary management, and upgrading and rehabilitat- period 1994­2004 shows a change toward a ing existing irrigation systems became more more comprehensive approach to rural important than area expansion.21 development, with a growing emphasis on building social capital--a result of the new rural With competition for scarce public sector strategy. While most CASs discussed the financial resources, it is likely that political importance of agriculture policy, less than half discussed it in the context of economic growth, giving greater prominence to community-driven Figure 2.5. CASs Reflected Changes in Bank Policy development, general rural development, and reform of agricultural institutions (figure 2.5).22 Agricultural water issues were less frequently 100 1994­97 discussed. 1998­01 2002­04 Coverage of privatization, trade, and markets, (%) 80 also declined overall, perhaps because less CASs attention was needed as they became self- in 60 sustaining in many countries. Indeed, a prevail- ing view among the Bank's country directors discussion who were interviewed by IEG was that the of 40 Bank's major contribution was helping to get the country's policy environment right, not assisting agriculture, which is seen as primarily a private Frequency 20 sector activity. Policy Evolution Is Reflected in the CASs 0 Attention to irrigation reform to reduce Agricultural Agricultural- Rural Rural subsidies and improve sector efficiency was development led economic development development growth as poverty fairly constant during the 1990s but grew after alleviation 2001, featuring in almost half the CASs (figure Source: IEG review of 129 CASs. 2.6). Within this trend was growing recognition 1 4 AW M R E L E VA N C E T O T H E B A N K D E C L I N E D Figure 2.6. Discussion of Institutions for Water Management Became More Prominent in CASs 50 1994­97 40 1998­01 (%) 2002­04 CASs in 30 discussion of 20 Frequency 10 0 Irrigation Irrigation & Irrigation & Cost Water Water users Water reform & rural agriculture recovery management charges development development Source: IEG review of 129 CASs. that institutional reform will depend on greater Regional Differences Are Captured capacity building in the water sector. Contrarily, in the CASs discussion of agricultural water charges almost In the MNA Region water resources manage- disappeared, being internalized in participatory ment concerns now top the agenda because of irrigation management, while discussion of cost rapid urbanization and growth of agriculture, recovery modestly increased. Discussion of which now uses 90 percent of all water cross-cutting water resource management resources.23 Even so, except in critically water- issues generally declined, except in the Middle overdrawn countries like Yemen, discussion of East and North Africa (MNA) Region. Reduced agricultural development and water issues in the interest in water tariffs is of some concern CAS dropped by half and irrigation to zero. In because it avoids the increasingly important part this was caused by the very uneven progress issue of the opportunity cost of water in agricul- on removing agricultural subsidies on inputs ture as urban and environmental competition and protective tariffs on outputs. In Jordan, for scarce water increases. The general soften- protective tariffs were removed via a structural ing of the approach to AWM and tariff issues is reform program, yet high implicit subsidies to explained in part by the comparative doubling agricultural water use and the government's of discussions on irrigation in the context of unwillingness to reform the sector distorted comprehensive rural development--this is also crop production, making it regionally noncom- a reflection of the changing skills mix of Bank petitive. Consequently, the Bank is investing in staff caused by retirement of AWM specialists sectors that are willing to reform, such as and their replacement with generalists. telecommunications and urban water supply, 1 5 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 and where social needs are greatest, such as IBRD/IDA blends accounted for another 9 education. Morocco has a similar history except percent of projects (mostly in China, Indonesia, that most protective tariffs remain in place. and Pakistan). Eight loans to China and one to Indonesia account for the increased share going At the other end of the spectrum, attention to to lower-middle-income countries in the period rural development in Sub-Saharan Africa CASs 1998­2001. Argentina, Lebanon, Mexico, and increased fourfold in the past 10 years, and a Turkey received most of the lending to upper- slight drop of interest in agricultural develop- middle-income countries. There is some ment has since revived in the past two years (box evidence of targeting the poorest countries 2.1). Sub-Saharan Africa is the only region where following the 1997 Vision to Action strategy. As population is growing faster than agricultural the volume of lending sharply contracted in the GDP and absolute poverty is increasing. While period after 1999, an increasingly larger there are some notable agricultural successes portion--reaching more than 95 percent in based on irrigation (for example, in Mali and 2002--went to the lower-income group. Niger), some countries also have problems. In Senegal, for example, irrigation-related bilharzia Performance issues rather than poverty consid- is a major public health issue. In Madagascar, the erations were the main drivers of reduced second largest irrigator in the region, irrigation lending among the biggest borrowers. Although has made little difference to agricultural produc- China, India, Indonesia, and Pakistan remained tivity because of low skills, input use, technol- the Bank's most consistent and prominent ogy, and access to markets. borrowers, the amount they borrowed for AWM fell significantly in the past 11 years (table 2.4).24 Mixed AWM Responses to New This falloff in borrowing reflects greater Bank Bank Policies attention to performance, institutional issues, Bank lending for AWM was primarily to the and governance, as well as country agreements to poorest countries (figure 2.7). And except for 4 better use existing agricultural water infrastruc- of the 11 years, more than half of all lending was ture and to reform irrigation institutions. The to those in the lower-income group. Overall, IDA shrinkage of lending is generally in accord with credits accounted for 55 percent of all AWM the CASs agreed with these countries. projects and 52 percent of commitments, and In India during the 1990s the Bank's approach to irrigation development became far more focused on assisting states, such as Andrah Figure 2.7. Poorest Countries Received Most Bank Pradesh, that were willing to reform institutional AWM Investment, 1994­2004 shortcomings and organizational inefficiencies 25 and better manage irrigation expenditures. In 1994­97 Andrah Pradesh, greater spending on O&M 20 1998­01 improved reliability of water deliveries and 2002­04 (%) farmers' willingness to pay water fees (table 2.5). 15 Even so, the subsector remains extensively total of subsidized. This trend away from new construc- 10 tion was accelerated because water resources Share and, by association, irrigation projects were 5 perceived to be risky for the Bank's reputation, especially in the light of the controversies 0 Lower- Lower- Upper- surrounding the Bank's engagement with the income middle- middle- Sardar Sarovar Project, resettlement, and the income income ensuing Morse Report that cooled interest in the Source: World Bank data. sector. 1 6 AW M R E L E VA N C E T O T H E B A N K D E C L I N E D Box 2.1. Bank Strategy for Sub-Saharan Africa: The Importance of Agriculture Agricultural production remains an important component Figure A. GDP Depends Much on Agriculture of GNP in most Sub-Saharan countries and the adverse 12 effects of periodic droughts has severe impact on eco- (%) 10 Agriculture growth GDP growth nomic growth, sometimes for several years, as happened rate 8 6 in the early 1990s.a With a high reliance on rainfall and lack growth 4 of irrigation, most farmers are unwilling or unable to risk 2 investment to improve agricultural productivity. Instead, 0 Annual -2 they have increased the area cultivated using traditional -4 methods because land is the least constrained resource. -6 Even so, yield increases have stagnated as soil nutrients 1970 1975 1980 1985 1990 1995 2000 become depleted--Sub-Saharan Africa (SSA) fertilizer Source: FAO 2004. use is only 5 kilograms per hectare; in East Asia it is 194 Figure B. Agricultural Area Increased but Yields Have Stagnated, 1980­2003 kilograms per hectare. Only 2 percent of water resources (percent change) 47 are harnessed compared with 36 percent in SAR and 53 50 42 42 42 37 percent in MNA.b The FAO projects that to meet basic 40 human needs, irrigation in Sub-Saharan Africa will have 30 to expand by 2 to 4 million hectares in the next three Percent 20 15 8 decades. 10 3 -1 -1 Working together with the New Partnership for Africa's 0 MNA SAR LAC EAP SSA Development, African agricultural ministers, the African -10 Change in cultivated area Change in yield Development Bank, the United Nations Economic Com- Source: FAO 2004. mission for Africa, FAO, and the World Bank developed the Figure C. Sub-Saharan Africa Is the Least Irrigated Region, 1980­2001 2003 Comprehensive African Agriculture Development Pro- (percentage of land irrigated) gram. The primary purpose of the program is to raise farm 50 production, increase economic growth, and reduce hunger 1980 1990 2001 40 through concentration on four pillars: 30 · Extending land area under sustainable land manage- Percent 20 ment and reliable water control systems, · Increasing food supply and reducing hunger, 10 · Improving rural infrastructure and capacities for mar- 0 ket access, and SAR MNA EAP LAC SSA Source: FAO 2004. · Promoting, over the long term, agricultural research and technology dissemination. Figure D. More CASs Are Emphasizing Agricultural Productivity and Water Management, 1994­2004 (percentage of all Sub-Saharan Africa CASs) Annual investment needed under this plan averages 70 1994­97 1998­2001 2002­04 $17.9 billion, of which three-quarters will be for land, water, 60 and rural infrastructure investment and their operations 50 40 and maintenance. The sum is large; but is less than the Percent 30 amount spent on food imports. 20 ­­­­­­­­­­­­­­­­­­­­­­­­­­ 10 a. Regression of agricultural growth on GNP during the period 1990­2001 yields an R2=0.5104 and a regression coefficient of +1.1898. 0 Irrigation & water Irrigation & rural Irrigation & General irrigation b. Bruinsma 2003. resource development agricultural strategy management productivity Source: IEG analysis of 33 CASs. 1 7 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 Table 2.4. Largest Borrowers Took a Smaller Share of Investment FY94­FY98 FY99­FY04 Borrower Investment in Investment in typology (and agriculture water only agriculture water only number of (US$ No. of (US$ No. of countries) millions) (Percentage) Projects millions) (Percentage) Projects Consistent (4) 1,884 62 24 956 38 18 Regular (9) 336 11 15 646 26 29 Periodic (10) 255 8 13 384 15 17 Occasional (12) 444 15 13 370 15 11 One-off (21) 131 4 8 157 6 13 Total 3,050 100 73 2,512 100 88 Source: World Bank data and IEG analysis. An earlier independent review of India's water productivity of the rain-fed agriculture sector sector (IEG 2002b) found that half of the irriga- that accounts for much of the residual rural tion projects completed in the 1990s had poverty. Allied with an early 2002 Planning satisfactory outcomes, a third were judged to Commission embargo of new works and the have substantial institutional development transfer of an increased portion of the budget impact, and fewer than 20 percent were responsibility to the individual states, these evaluated as having likely sustainability. A imperatives led to a reduced level of lending for country assistance evaluation found that the irrigation infrastructure rehabilitation and selective and more relevant focus on reforming modernization. states had a larger impact on rural development institutions than the disparate and enclave The Bank's approach in China before 1990 was projects of earlier years. There was also a need to finance new and conventional irrigation to move away from state-monopolized, mono- projects prepared by the line agencies with little crop irrigation to more diversified agriculture, input by the Bank. During the 1990s this and to give greater attention to increasing the changed as Bank staff introduced innovative institutional and participatory management components, improved procurement practices, Table 2.5. Reform in India Improved Spending on and increased the Bank's role in providing O&M and Revenue Collection knowledge about modernizing water manage- ment and using water more efficiently. The Proportion of O&M costs Bank's decrease in lending reflected a move (percent) from its more conventional banking role to a Spent on Realized from focus on broader development goals and policy, establishment water a decrease in lending that was accelerated by the State and staff charges switch from concessional IDA credits to more Reforming: commercial IBRD loans.25 Andrah Pradesh 38 74 Nonreforming: Decreased lending to Pakistan for agricultural Assam 99 0.1 water management and infrastructure was the Gujarat 50 28 result of its low overall portfolio performance-- Haryana 85 27 in the late-1990s Pakistan was ranked among the Source: India Planning Commission 2002. 25 worst performers Bank-wide.26 Consequently, 1 8 AW M R E L E VA N C E T O T H E B A N K D E C L I N E D an aggressive portfolio improvement strategy, nonagricultural employment in those countries implemented in mid-FY98, closed 32 projects, caused many people to revert to the land for reducing the portfolio to 16, and whittled the subsistence production. Therefore, the Bank's pipeline of potential projects from 30 to 12. small interventions not only assisted poverty Many large and important projects and alleviation in the medium term but also provided programs--notably the Left Bank Outfall Drain, an entry point for policy discussions aimed at the National Drainage Program, the Chashma rationalizing the region's aging and oversized Right Bank Canal Project, and the Social Action infrastructure, which was frequently environ- Program--did not have the desired impact mentally damaging and uneconomic to operate. because of poor design, poor feasibility, poor coordination, delays in implementation, or Development Objectives Have Changed corruption. The main risk factors in the National Considering the whole portfolio, the primary Drainage Project were very uneven provincial development objective of projects was agricul- buy-ins to institutional reform and inadequate tural production, followed by community attention to reform and capacity building. And support and participation, and measures to because the water sector only slowly adapted to increase the efficiency and sustainability of the government's development agenda, which irrigation and drainage (figure 2.8)--develop- was based on decentralization, participation, ment priorities that accord with the Bank's 1993 and management transfer, investment for most Water Resources Management Policy and 1997 agricultural water activities contracted. Vision to Action. Like India and Pakistan, Indonesia's top-down Figure 2.8. Relative Importance of Development public sector focused on building new Objectives for All AWM Projects, 1994­2004 infrastructure--while deferring maintenance of existing irrigation and drainage works--and Dedicated projects Nondedicated projects (% of objective) became increasingly challenged during the mid- 1990s. The Water Resources Sector Adjustment Agriculture development & production 46% Loan (1999) brought institutional reform to the Community support & participation 71% fore. Thereafter, the reduction in infrastructure components, along with the financial crises, I&D efficiency & sustainability 16% decentralization, and increased requirements for provincial financing, caused lending for Income & employment 66% agricultural water management to shrink. Water resources management 30% A new set of smaller country clients for AWM Institutional development & support 50% emerged as commitments to the biggest borrowers declined. Before 1999 Europe and Poverty reduction 68% Central Asia (ECA) accounted for only 11 Policy & institutions percent of projects in the portfolio; afterward, 57% with 29 loans, it accounted for a third. The Environment & natural res. mgnt. 58% relatively small size of most ECA agricultural water projects is more a reflection of the size of Risk management 53% the economies in the region and the limited IDA envelope, than the Bank's identification of Agriculture support services 78% needs. The transition to market economies and Private sector participation 53% loss of the former Soviet Union market for agricultural outputs has put most of the region's 0 20 40 60 80 investments in agricultural water management Number of occurrences, 1994­2004 in jeopardy. Yet, simultaneously, the collapse of Source: IEG analysis of 161 PADs. 1 9 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 Because development objectives and Figure 2.9a. Importance of Physical components differ according to the relative Development Objectives Either importance of AWM in projects, the portfolio Stabilized or Declined in Importance was further disaggregated into 67 dedicated 30 projects, where more than half of the Bank's 1994­97 commitments were for AWM, and 94 nondedi- 25 1998­01 in cated projects. All of the dedicated projects are 2002­04 20 managed by ARD plus one by the Environment objective Department. While dedicated projects tend to of (percentage) 15 address risk, agricultural, and infrastructure period 10 issues, nondedicated projects focus on social Occurrence time concerns and agricultural support services. 5 0 While most development objectives were Agriculture I&D efficiency Water Environment development & sustainability resources and natural aligned with Bank and sector policies during the & production management res. mgnt. period 1994­2004 there were some notable Source: IEG analysis of 161 PADs. exceptions. Objectives central to the new strate- Figure 2.9b. Improved Attention to gies--addressing poverty reduction, agricultural Social and Institutional Issues development and production, and environment and natural resources management--declined 14 1994­97 in importance (figure 2.9a). Conversely, 12 1998­01 objectives encompassing community support in 2002­04 10 and participation, income and employment, and support for capacity building and institutional 8 objective of (percentage) development increased (figure 2.9b). One 6 reason for these changes is that development period 4 time objectives have become more practical and Occurrence 2 achievable by focusing on measurable outcomes 0 rather than global targets. For example, Community Income and Poverty Institutional increased attention to income and employment support & employment reduction development participation & support almost balances the decrease in poverty- Source: IEG analysis of 161 PADs. reduction objectives. dedicated and nondedicated projects would ARD's dedicated projects clearly show increas- improve to meet both technical and policy ing interest in making the physical aspects of imperatives. AWM work and declining interest in policy, institutions, and the environment. In contrast, The few non­ARD projects distributed across nondedicated ARD projects, in which AWM seven departments preclude statistical analysis. professionals have a minor role, emphasize However, the effect of this diversified responsi- development priorities more aligned with bility on the outcomes and impacts of the current Bank policies. The net effect of nonded- agricultural water components outside ARD is icated AWM on the development objectives of generally difficult to discern in the Bank's the more traditional, technically oriented implementation completion reports. This is approach to AWM shown by dedicated projects because outcomes, if reported at all, are is evident in figure 2.10. The main lesson from confined to the few key objectives that reflect these findings is that the skills mix in the teams the rationale for a project as articulated by the designing AWM projects need broadening to initiating Sector Board. Minor agricultural water avoid design capture by professional enclaves components are, therefore, most frequently within the Bank. That way the quality-at-entry of accounted for in terms of inputs and outputs 2 0 AW M R E L E VA N C E T O T H E B A N K D E C L I N E D Figure 2.10. Comparison of Dedicated and Nondedicated Project Objectives Reveals Strengths and Weaknesses Community support & participation Income and employment Strengths of Agriculture support services nondedicated projects Poverty reduction Policy & institutions Environment & natural resource management Agriculture development & production Institutional development & support Weaknesses of nondedicated Risk management projects Water resources managment AWM effciency & sustainability -20 -15 -10 -5 0 5 10 15 20 Change in occurrence of objectives, 1994­2004 (percentage) Source: IEG analysis of 161 PADs. This analysis is based upon a comparison of all dedicated projects with all nondedicated projects designed in the period FY94­FY04. The occurrence of each objective in each cohort was expressed as a percentage of all objectives and the differences between the cohorts is plotted above. Thus, for example, AWM efficiency and sustainability was 4.6 percent for nondedicated projects and 20.2 percent for dedicated projects, and the difference was ­16 percent. such as the length of canal or drainage works, or of large irrigation systems or improvement of number of wells installed. There is invariably no them, and by FY02­FY04 these accounted for description of what these facilities achieved in more than 80 percent of Bank commitments in terms of area irrigated or drained and the impact this group of projects (figure 2.11a). Examples on crop production, rural incomes, or the include the 1996 Punjab Private Sector Ground- environment. water Development Project that focused on transferring after rehabilitation some 4,200 Less New Construction and Lower Costs public tubewells to farmer organizations, As overall lending for agriculture fell, the Bank's decreasing the leakage from all types of canals, commitment to freestanding AWM projects and improving the efficiency of water distribu- dedicated to water management decreased. tion on communal and on-farm watercourses. In Since 1994, nondedicated projects have Mexico the 2002 On-Farm and Minor Irrigation comprised 58 percent of the AWM portfolio. Networks Improvement Project covered improvements to irrigation canals and drains, The type of infrastructure components financed land leveling, and installation of new high-tech by dedicated and nondedicated projects are sprinkler, drip, and microsprinkler irrigation markedly different, even though most projects systems. contain a mix of physical interventions ranging from some new-builds, redesign and upgrading, Nondedicated projects, while initially focusing and repair of damage caused by deferred on rehabilitation in the mid-1990s (figure 2.11b), maintenance referred to as rehabilitation. now focus primarily on building new systems Dedicated projects now prioritize rehabilitation that are small scale and community owned. For 2 1 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 repair 24 irrigation systems and to build a Figure 2.11a. Reduced New Construction and replacement system. Improvement in Dedicated Projects 100 The average Bank commitment to AWM projects 1994­97 1998­01 2002­04 declined for two reasons: a change in the type of 80 infrastructure financed, and the greater with (%) emphasis on nonstructural and capacity- 60 building components discussed above. This projects trend reflects not only the needs of deferred component 40 O&M of infrastructure in the SAR and East Asia this and Pacific (EAP) Region but also the growing Dedicated 20 importance of clients in the emerging economies of the ECA Region with similar 0 problems. Because dedicated AWM projects New Significant Rehabilitation construction improvement became more focused on rehabilitation, the and/or expansion costs of these projects fell: rehabilitated projects Source: IEG analysis of 161 PADs. averaged $2,900 per hectare while new construction averaged $6,600 per hectare. The increased focus on new construction in Figure 2.11b. Reduced Improvement and dedicated projects increased their share of Rehabilitation in Nondedicated Projects commitments for AWM (figure 2.12). Even so, the overall commitment to dedicated projects 100 fell much faster than the commitment to 1994­97 1998­01 2002­04 nondedicated projects grew and overall commit- 80 ments to AWM declined (figure 2.13). with (%) 60 In the 1980s and early 1990s investment in AWM projects was also strongly supported by parallel invest- component 40 ment in agriculture. This has now changed this among the biggest borrowers (figure 2.14). Nondedicated 20 0 Figure 2.12. Share of Nondedicated New Significant Rehabilitation construction improvement Lending Increased and/or expansion 50 Source: IEG analysis of 161 PADs. (%) 40 example, the 1995 Brazil Northeast Rural Poverty Alleviation Project--Ceara provided, via a CDD lending 30 approach, $107 million for infrastructure. Of AWM this, 77 percent was for basic infrastructure of 20 (water supply, drains, sewage disposal) and 22 percent was for "productive" investments, Share 10 mainly irrigation that included, for example, irrigation kits to cover 0.8 hectare plots and 0 sprinkler irrigation systems to irrigate 10 1990 1995 2000 2005 hectares areas of new lands. In Bolivia, the 1998 Fiscal year El Niño Emergency Assistance Project helped to Source: IEG analysis of 161 PADs. 2 2 AW M R E L E VA N C E T O T H E B A N K D E C L I N E D Many of the appraisal documents implicitly Figure 2.13. Overall Commitment assume either that policy reform is largely Declined complete, or that it is beyond the project's 800 scope--particularly where irrigation and drainage Dedicated was only one of many components, or where the Nondedicated size of the investment was small relative to the norm for earlier periods. Yet, in many cases, 600 important policy issues remain. For example, in millions) Brazil the Bank's analysis shows that there is a (US$ need to increase the security and enforceability 400 of water rights, introduce water charges that reflect the economic value of water, and clarify amount the roles and responsibilities of institutions.29 200 AWM These recommendations are valid for many of the Bank's clients, particularly in countries such as India, Jordan, and Mexico. In Madagascar, for 0 example, the extremely poor performance of 1990 1995 2000 2005 irrigation, despite decades of Bank support, Fiscal year caused a withdrawal from the sector in 2001. Yet, Source: IEG analysis of 161 PADs. subsequent analysis has shown that most of the problems are the result of a distorted incentive Policy Reform Has Languished structure for seeds and fertilizer, inadequate The Bank's water strategy gives considerable attention to agricultural extension and emphasis to the enabling environment. severance of its linkage to agricultural research, However, project appraisal documents make and inadequate transport infrastructure that only modest proposals for policy reform and precludes efficient markets (World Bank 2005a completion reports usually conclude that and Minten 2006). There is a danger in many reform expectations at appraisal were unrealis- countries that the Bank will scale back lending tic, particularly for cost recovery.27 Does the lack for irrigation before the policy reforms needed of policy reform content reflect a lack of need, a to get the balance right between public and lack of leverage, or a pragmatic decision to private intervention are complete. tackle this business outside the project, using other Bank vehicles? All three of these consider- ations played a role. Independent evaluations of Figure 2.14. Other Agricultural Investment Declined attempts to reform agricultural policy have Steeply among Large Borrowers found that passing laws and developing regula- 2,500 tions is easier than implementing them--and Other agriculture that attention to the detail of implementation is I&D 2,000 often slighted, typically because the time span of a single project is not long enough to come to 1,500 grips with the toughest issues.28 amount millions) US$ 1,000 Dedicated irrigation and drainage projects with policy content--large or small--only give it 2002 Commitment 500 modest attention. For about 20 percent of the (in cases reviewed, policy was not addressed at all, 0 either because it was no longer relevant or 1990­92 1993­95 1996­98 1999­01 2002­04 because it was being addressed outside the Fiscal years project. Source: World Bank data and IEG analysis. 2 3 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 Policy reforms need integration across the some of the Bank's recent economic and sector whole agricultural sector. The success of irriga- work for Morocco.31 tion can adversely affect the welfare of rain-fed farms in developing countries.30 Since the early Implications for Management 1960s, yield improvement was by far the largest Within the more specialized AWM projects that factor, not just in the developed world but also have significant components devoted to irriga- in the developing countries, where it accounted tion, drainage, and flood control, the approach for 70 percent of increased production. Thus, also has become more technically focused. This while irrigated farms increased output the most, may not be an issue where AWM projects are many rain-fed and marginal farms also benefited. part of a broader package of rural development Unlike the irrigated farms in the developing projects that deal with social, human, and world, however, rain-fed farms are generally not economic development, as is the case, for subsidized. Hence, while irrigated farms can example, in Armenia (box 2.2). increase (subsidized) production to maintain their incomes, rain-fed farms do not have that However, the more general AWM projects-- option but are subject to the price structure. many of which embody CDD-type approaches-- Consequently, they have little option but to are building water infrastructure with less accept falling grain prices. attention to issues of technical efficiency and sustainability. Independent evaluation of CDD This paradox clearly indicates that a comprehen- approaches shows that this is a major problem-- sive approach--rather than an irrigation-led more than three-quarters of Bank task managers one--is required for agricultural development for these projects raised sustainability of project in developing countries. It also shows that if infrastructure as a major issue.32 This is the market-distorting subsidies for irrigation were classic decentralization dilemma: how to give removed, then irrigated farms would move out power to local stakeholders while at the same of competition with rain-fed farms and grow time providing them with sufficient technical higher-value crops. Further inducements to rain- support. fed farms would also arise from the elimination of the high subsidies in developed countries. A These findings indicate the importance of good example of the analysis needed to get integrating AWM projects within country rural irrigation and subsidies right is provided by strategies and ensuring that they are adequately Box 2.2. Armenia: Creating Synergistic Packaging for AWM Lending The first Irrigation Rehabilitation Project (FY95) in Armenia targeted velopment Project, aimed at improving system efficiency, re- decayed public sector infrastructure and focused on engineer- ducing operational costs, and enhancing the effectiveness of ing and institutions. This was complemented by the FY06 Social water user associations, and the Natural Resources Management Investment Fund that included modest support for small-scale ir- and Poverty Reduction Project--targeted marginal farmers and rigation. Subsequently, in FY99, the Dam Safety Project (and its those in mountain and hill areas. The FY04 Agricultural Reform Pro- second phase in FY04) secured bulk irrigation supplies, the Title ject Supplemental Credit provided critical support for commer- Registration Project secured farmers' rights to land and thus cial fruit growers. The FY06 Rural Enterprise and Small-Scale collateral, while the Agricultural Reform Support Project aimed Commercial Agriculture Development Project supported the de- to improve rural finance, agro-processing, and agricultural in- velopment of Armenia's small- and medium-scale rural busi- stitutions, including research and extension. The FY00 Judicial Re- nesses and the ability of farmers and rural entrepreneurs to form Project aimed to improve governance and thus private access markets and by stimulating market-oriented private and commercial transactions. Two projects in FY02--the Irrigation De- public investments in rural areas. 2 4 AW M R E L E VA N C E T O T H E B A N K D E C L I N E D supported by parallel operations that address ing the skills mix of appraisal teams preparing critical omissions. Where the country lending nondedicated AWM is also essential: irrigation window is small it is essential that projects engineers should be included on appraisal address the most critical issues and include teams where water development or its manage- enabling actions that may lie outside the water ment is part of a broader social package. or agricultural sector. Comprehensive appraisal Similarly, dedicated projects need social that includes upstream and downstream scientists, anthropologists, and rural financial linkages thus becomes vitally important, as does specialists to address issues related to capacity economic modeling and analysis. And improv- building, incentives, and social sustainability. 2 5 3 Global Relevance Remains High A gricultural water management to increase food production contributes toward meeting the Millennium Development Goals, provides basic food security needs for growing populations, and has important catalytic growth impacts on the rural economy. It is also vital to the conservation of in- creasingly stressed water resources. Overview ment opportunities and higher wages within This chapter summarizes the evidence and the adopting regions, and thereby increase shows that AWM remains relevant to the Bank's migration opportunities for the poor to other development agenda. agricultural regions. · It can benefit a wide range of rural and urban Benefits of Good AWM Are Substantial poor through growth in the rural and urban Sound AWM has the potential to boost growth nonfarm economy. and reduce poverty to benefit the poor in · It can lead to lower food prices for all con- several ways. It is directly relevant to the Millen- sumers, rural or urban. Crop harvest from ir- nium Development Goals (table 3.1). Hazell and rigated areas leads to strengthened staple or Haddad (2000) classify the direct and indirect nonstaple food output. This abundant supply ways that irrigation and other new water lowers the prices of staples and other food management technology can increase growth and thereby cuts the food bill of small farmers and reduce poverty: and the poor from the rural areas as well as that of the urban poor (Lipton, Litchfield, and Fau- · It can benefit poor farmers directly through an res 2003). Irrigated land must, therefore, be increase in their level of own-farm produc- supported in order to sustain low food prices tion. This may involve production of more (Carruthers, Rosegrant, and Seckler 1999). food and nutrients for their own consump- · It can empower the poor by increasing their ac- tion, and increasing the output of marketed cess to decision-making processes, increasing products for increased farm income. their capacity for collective action, and reduc- · It can benefit small farmers and landless la- ing their vulnerability to shocks via asset ac- borers through greater agricultural employ- cumulation. 2 7 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 Table 3.1. Improving Agricultural Water Management Significantly Contributes to the Millennium Development Goals How water management contributes to achieving goals Millennium goals Directly contributes Indirectly contributes Poverty: To halve by 2015 Assured water and its good management raise · Secondary employment in agro-processing the proportion of rural incomes and off-farm activities the world's people · Reduced vulnerability to water-related whose income is Water as a factor of production in agriculture, hazards reduces risks in investments and less than $1/day industry, and other types of economic activity production · Reduced ecosystems degradation boosts Investments in water infrastructure and services local-level sustainable development act as a catalyst for local and regional · Improved health from better quality water development increases productive capacities · Ensured ecosystems integrity to maintain water flows to food production · Reduced urban hunger by cheaper food grains from more-reliable water supplies Hunger: To halve by 2015 Improves food security by mitigating drought, · Improved school attendance from improved the proportion of flood, and drainage risks health and reduced water-carrying burdens, the world's people especially for girls who suffer from Water as direct input into irrigation, including · Community-based organizations for water hunger supplementary irrigation, for expanded food management improve social capital of production women · Reduced time and health burdens from Reliable water for subsistence agriculture, home improved water services lead to more gardens, livestock, tree crops balanced gender roles Source: Soussan 2002. In the remainder of this section, the evidence percent. Analysis at both the macro and sectoral on irrigation impacts on production, impact, levels and at the microirrigation or farm levels and poverty is reviewed. In undertaking the showed that irrigation has significant impacts on review, IEG acknowledges the contribution of crop production, farm income, and poverty. At previous reviews, including Bhattarai (2004) and the macro level, Fan, Hazell, and Thorat (1999); Heyd (2004). Fan and Hazell (2000); and Fan, Zhang, and Zhang (2002) found that irrigation investment Irrigation Impact on Production, Income, has significant effects in increasing crop produc- and Poverty tion and reducing poverty, but that the impact of Rapid agricultural and economic growth, irrigation is lower than that of rural roads and together with direct social spending and agricultural research. appropriate social policies, is essential for poverty reduction in developing countries. Other studies have shown much stronger According to De Haan and Lipton (1998) impacts of irrigation on crop productivity and economic growth explains between one-third poverty alleviation. Datt and Ravallion (1997) and one-half of poverty reduction in Asia. Their found that differences in the growth rate of report, based on a summary of various studies average agricultural output per unit of crop area on developing Asia, found that a 1 percent were important in explaining cross-state differ- growth in per capita GDP is associated with a ences in rural poverty reduction between 1958 decline in the incidence of poverty of 0.82 and 1994. The initial endowments of physical 2 8 G L O B A L R E L E VA N C E R E M A I N S H I G H infrastructure and human resources played a income (Dhawan 1988, 1999; Vaidynathan et al. major role in explaining the trends in rural 1994). As revenue increases, the purchasing poverty reduction. Higher initial irrigation power of rural farmers is positively affected, intensity, higher literacy rates, and lower initial enabling them to procure a wide variety of food. infant mortality all contributed to higher long- In effect, a balanced food diet with adequate term rates of poverty reduction in rural areas. intake of micronutrients is assured for the Bhattarai (2004), in an econometric analysis farmers' families (Lipton, Litchfield, and Faures using state-level data for India, 1970­94, showed 2003). In Bangladesh and India, farmers that rural literacy, followed by irrigation, had cultivate three rice varieties (boro, aman, and higher impacts on agricultural productivity than aus) per year in irrigated areas, in contrast to other input factors, including fertilizer, modern only one rice crop for rain-fed agriculture. This crop varieties, and road density. The impact of practice of crop diversification offers substan- irrigation on reducing poverty was even higher tially higher crop productivity and helps than that of rural literacy and significantly higher increase returns to farmers' land and labor than roads, fertilizer, and modern varieties. The resources at the household level (Dahawan and marginal impact of groundwater sources of Datta 1992; Hussain and Hanjra 2004a, 2004b). irrigation on spatial and temporal reduction of the rural poverty level in India was, in general, Impact on Employment and Wages higher than that of the canal irrigation. Improved access to irrigation generates both direct and indirect employment in the rural Barker et al. (2004) investigated the different economy. While direct employment is generated investment variables that contribute to output by the construction and maintenance of irriga- growth in Vietnam. Based on their results, tion projects, more importantly, indirect investment in irrigation ranked the highest in employment is generated through enhanced explained share of total agricultural output land-use intensity, cropping intensity and growth, at 28 percent from 1991 to 1999, productivity, as well as its multiplier effects in followed by research and development at 27 allied activities like livestock and rural nonfarm percent, and road improvement at 11 percent. sectors (Patel 1981; Dhawan 1988; Saleth 1997). Pardey et al. (1992) studied the socioeconomic Patel (1981) found that the average additional effects of past investments in rice and soybeans employment generated per hectare per annum in Indonesia. They argued that investment in of farming in those 10 irrigated-system-level technology alone did not affect yield and output studies in Gujarat was 50 person-days per growth, but substantial increases in fertilizer hectare per year. Another analysis showed that application, pesticide use, and irrigation services in Gujarat, while the nonirrigated farms used have contributed significantly to gains in total only 484 hours of total labor per gross cropped production. Rosegrant, Kasryno, and Perez area, canal-irrigated farms used 729 hours, well- (1998) found that irrigation investment irrigated farms used 835 hours, and conjunctive explained 29 percent of rice production growth use of groundwater- and canal-irrigated farms in 1969­90, as well as 11 percent of maize used 1,456 hours (Bhattarai 2004). production growth, 35 percent of cassava production growth, and 7 percent of soybean The most effective way to tackle poverty may be production growth. to marry irrigation with other complementary investments. A modeling exercise for Ethiopia At the micro level, a number of studies have found that the impact on poverty of combined established that irrigation increases crop yield investment in irrigation and roads was substan- per hectare per season, land-use intensity, and tially greater than investment either in irrigation cropping intensity, leading to increased land or in roads alone.2 In Bangladesh and Nepal it productivity (gross crop outputs per unit of land has been found that benefits from investments per year) that in turn helps improve farm in roads or in new inputs are higher when land 2 9 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 is irrigated (Brabben et al. 2004). Detailed within the surrounding nonfarm economy. The surveys in Madagascar found agricultural yields multipliers are particularly large in Asia-- and access to essential inputs to be lower and between 1.5 and 2.0. That is, each dollar increase the incidence of poverty higher the farther a in agricultural value generated by irrigation or village is located from a good, public, all-weather other investments or new technology leads to road (Minten 2006). In summary, it is the an additional $0.5 to $1.0 of additional income "package" that matters for effective poverty created in the local nonfarm economy (Hazell alleviation and not the mere supply of irrigation and Ramasamy 1988; Haggblade and Hazell water; and "investments in irrigation sectors 1989). The multipliers are about half as large in may not reduce poverty directly in any signifi- Africa and Latin America (Haggblade and Hazell cant way unless accompanied by other comple- 1989; Dorosh, Haggblade, and Hazell 1998). mentary [public] interventions" (Hassain and Lower multipliers in Africa are attributable to low Hanjra 2004). There are, however, many per capita incomes, poor infrastructure, and complementary private investments that will be farming technologies that require few stimulated through markets after the initial purchased inputs (Dorosh, Haggblade, and irrigation and other public investments have Hazell 1998). In a comprehensive review of been made. The skill of good economic and growth linkages, Haggblade and Hazell (1989) sector work feeding into project appraisal is found that the strength of growth linkages is determining what the minimal public invest- higher in labor-abundant regions, and increases ment should be and avoiding projects that try to with regional development and per capita deliver everything and fail. incomes. In particular, irrigated regions dominated by medium-sized farms and modern Impact on the Nonfarm Economy input-intensive farming systems generate the Agricultural growth--and the irrigation invest- largest multipliers. The multipliers are smaller in ments that contribute to it--has significant rain-fed farming systems and in regions impacts on the rural nonfarm economy as well. dominated by very small farms or by large Agriculture can influence nonfarm activity in at estates. least three ways: through production, consump- tion, and labor market linkages. On the produc- Relatively few studies have estimated the tion side, a growing agriculture sector requires multiplier effects of investment in irrigation on inputs (fertilizer, seeds, pesticides, pumps, the nonfarm economy, but those that have done sprayers, machinery repair services) that are so have confirmed such multiplier effects. In a either produced or distributed by nonfarm comprehensive study of the Muda Dam in firms. Moreover, increased agricultural output Malaysia, Bell, Hazell, and Slade (1982) found stimulates forward production linkages by that for every one dollar of value-added providing raw materials that require milling, generated directly by the dam, another 83 cents processing, and distribution by nonfarm firms. were generated in the form of indirect or Consumption linkages arise when growing farm downstream effects, resulting in a multiplier of incomes boost demand for a range of consumer 1.83. Bhattarai, Narayanmoorthy, and Barker goods and services. Demand increases as rising (2002), analyzing a panel of state-level Indian per capita incomes induce diversification of data, found that the direct and indirect income consumption into nonfood goods and services, benefits of irrigation investment exceeded the many of which are provided by local firms direct benefits to the farming community by (Rosegrant and Hazell 2000). more than 3 to 1. An ex-post analysis (Olsen 1996) of the Grand Coulee Dam and the Numerous empirical studies consistently show Columbia Basin Project showed that invest- that agricultural growth does, in fact, generate ments made in the basic sectors generated important income and employment multipliers between 1.5 to 1.7 dollars of total income within 3 0 G L O B A L R E L E VA N C E R E M A I N S H I G H the local area for each dollar produced by the the supply is, relative to demand (Alston, basic sectors. Norton, and Pardey 1995). Lower food prices are of benefit to rural and urban poor alike, and Irrigation also has substantial indirect impacts because food typically accounts for a very large on nonfarm employment. The employment share of their total expenditure, the poor gain growth multipliers are predominantly driven by proportionally more than the nonpoor from a increased rural household demands for decline in food prices (Pinstrup-Andersen and consumer goods and services as farm incomes Hazell 1985). These price effects may be muted rise, many of which are supplied by small, in open economies with low transport costs, and informal, and labor-intensive rural nonfarm more countries now fall into this category than firms. This leads to high nonfarm employment before because of recent rounds of market elasticity within rural regions; each 1 percent liberalization policies. But many poor countries increase in agricultural output is often associ- still face high transport costs because of poor ated with a 1 percent increase in rural nonfarm infrastructure, remoteness from world markets, employment (Gibb 1974; Hazell and Haggblade or inefficient marketing institutions, and may 1991). Mellor (2001) indicates that the additions still face considerable domestic price determina- to employment in the rural nonfarm sector tion even after market liberalization. In many stimulated by agricultural growth can be as landlocked African countries, for example, much as twice that for the agriculture sector. In domestic prices still fall sharply when domestic a study of rural nonfarm employment in 114 food production increases suddenly. Many Indian villages in different states, it was found traditional food crops are also not traded in that the nonfarm employment is determined by world markets, and hence their prices also the extent of irrigation. Similarly, using village- continue to be determined within the countries level census data of Tamil Nadu, Jayaraj (1992) that grow them. showed that irrigation is a major variable in explaining the magnitude of rural nonfarm Technological changes that smooth seasonal employment. In a survey of canal-irrigated and food supplies (such as irrigation and short- rain-fed farming villages in Tamil Nadu, India, season rice varieties) can also help smooth Saleth (1995) estimated that nonfarm employ- seasonal price variations and this can be of ment is about 46 person-days per hectare higher considerable benefit to the poor. The rural poor in canal-irrigated villages than in rain-fed villages. may also obtain enhanced food security from increased production within their region if it Impact on Food Prices and Diet Quality displaces food purchases from outside the Technological change, including the spread of region that previously had to be priced to cover irrigation, contributes to increases in the high transport costs. aggregate output of affected commodities and often to lower unit costs. This has proved to be Impact on Empowerment of the Poor one of the most important ways through which Hazell and Haddad (2000) have pointed out that poor people have benefited from technological "the assets which individuals, households, and changes in agriculture (Scobie and Posada 1978; communities control are critical for the capacity Rosegrant and Hazell 2000; Fan, Hazell, and to cope with vulnerability and to establish Thorat 1999). secure livelihoods. In many developing countries, the poor are highly dependent on If the national demand for these products is natural resources within their local environ- downward sloping (that is, export opportunities ments; hence poverty can be exacerbated by are constrained by trade policy or by high not having access to those resources. The cycle transport costs), then the output price will fall. can be perpetuated by not having access to The price decline will be greater the more elastic technologies and inputs in order to be better 3 1 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 able to use those resources, or by not being able Demand for Food and Better AWM to participate in the design and evaluation of The FAO predicts that a projected world popula- those technologies." Irrigation is one of the key tion of 8.2 billion in 2030 will drive a new round assets that empowers farmers (see box 3.1). of investment in irrigation and drainage, albeit at Box 3.1. Effects of New Irrigation in India's Andrah Pradesh Area, 2005­06 The income benefits from irrigation accrue in the form of higher project area, many more households had zero or negative income net income for cultivating households and higher wage income in 2005 than in 2006 when irrigation became available. As a result, and increased employment. The evaluation found that irrigation the bottom 10 percent of households in 2005 got 20 percent of the increased net farm income by just over 60 percent, about half of farm income in 2006. So, while the impact of irrigation in a good or which comes from increased cropping intensity and most of the normal year has a modest impact on the income distribution, in remainder from higher yields, with only a small part attributable drought years there is an extreme worsening in income inequality to changes in the crop mix (figures A and B). as many farmers experience negative income. The presence of ir- However, irrigation has a very modest positive impact on income rigation can mute this effect. distribution. The top quartile benefits most in absolute terms, the sec- With the good rains improving agricultural performance in 2006, ond quartile benefits most in relative terms, experiencing income combined with new irrigation, there was greater labor demand, par- growth of 30 percent. The poorest quartile has a low benefit, but their ticularly for women. Each household supplied, on average, 223 already low income means that they also experience income growth days of agricultural labor in 2006 compared with 155 in 2005. The of 20 percent, compared with 19 percent for the top two quartiles. increase in demand for labor led to an increase in average wages Dynamic effects also have an impact in distribution. House- of 5 percent for men, and 10 percent for women because of the high holds subject to repeated negative shocks become heavily in- demand for female labor to weed and harvest paddies. In the IEG debted and deplete their assets, constraining their ability to survey, women accounted for 63 percent of agricultural employment undertake productive investments. Reducing the impact of bad in 2005 and 64 percent in 2006. There is a considerable body of ev- years thus aids asset accumulation and helps households grow out idence that women's incomes have a larger impact on child wel- of poverty. Because of prolonged drought in the Andrah Pradesh fare (health and education) than do men's incomes. Source: IEG 2006a. Figure A. Distribution of Incremental Irrigation Figure B. Irrigation Reduces Poverty Headcount but Not Benefits Is Skewed toward the Better Off the Poverty Gap Rupees per hectare Poverty (thousands) measure 4,000 0.5 Incremental farm income Without project With project 0.4 3,000 Incremental wages Incremental employment 0.3 2,000 0.2 1,000 0.1 0 0 1 2 3 4 Upper Lower Upper Lower (Low) Wealth quartile (High) poverty line poverty line poverty line poverty line Poverty headcount Poverty gap Source: IEG data. 3 2 G L O B A L R E L E VA N C E R E M A I N S H I G H half the average rate of the preceding four subject to multiple cropping, is likely to increase decades. If the historic supply response will hold by a third by 2030, or 83 million hectares, in the future as it has in the past, the FAO provided water is available. projects that 80 percent of future increases in food production will come from intensification Better Management of Limited Water through higher yields, increased multiple Resources cropping, and shorter fallow periods--and Achieving these projections for expanded irriga- much of this is possible in rain-fed areas tion will be difficult. Irrigated environments face without irrigation. The balance of new arable increased competition for water but also land will come from developing countries reduced water supplies owing to system (especially those in Sub-Saharan Africa degradation and reduced investments in and Latin America), which have the potential to infrastructure development to store and distrib- add about 120 million hectares of new arable ute water. Globally, water is becoming an land. increasingly scarce commodity--one-quarter of the developing world's population will face The expansion of irrigation--40 million severe water scarcity in the next 25 years hectares, or about 1.3 million hectares a year (Seckler, Molden, and Barker 1998). The net until 2030--is projected to be strongest in South effect will be increasing real costs of water at the Asia, East Asia, and the Middle East and North farm level and declining profitability of irrigated Africa regions, where almost all arable land agriculture, taking into account all maintenance potential is used. Harvested irrigated area, and environmental costs. Box 3.2. Yemen: Evolution of an Integrated Water Resources Management Strategy Groundwater is Yemen's only significant year-round source of · Where groundwater is a state's major resource its sustain- water. Until recently, despite a large water-sector lending pro- able management should be the highest priority. gram, the Bank paid minimal attention to water resources man- · Technological, managerial, and institutional improvements agement in Yemen. The lacuna in recognizing the growing water to increase water conservation of all water-using sectors crisis over some 30 years of Bank lending is linked to a virtual ab- are essential. sence of strategic analysis and sector work in Yemen's water sec- · To manage costs and sustain supplies, urban water projects tor until the 2000s. Urban water supply and rural irrigation projects need to give water resource management a higher priority. proceeded without linkages between them, and out of 28 water · Link urban and rural water use through integration of plan- projects approved since lending to the water sector began in FY73, ning and management. It is inevitable that rural-urban water only one project--the Land and Water Conservation Project transfer will be needed in the future as urban populations (FY92)--focused on water conservation. Because withdrawal is quickly grow. greater than replenishment from rainfall, groundwater levels · Equitable water markets need to be developed. continue to decline--in heavily populated areas by as much as 6 meters per year. And mining of water from the deep aquifers, Fortunately, the problems have been increasingly recognized by where water is not renewable, is increasingly practiced. De- the Bank. A sector strategy has been prepared, and several re- spite this, groundwater management was neglected. Instead, cent projects are starting to tackle water depletion head-on, for supply-side infrastructure projects continued, without consid- example, the Sana'a Basin Water Management Project (FY03) and eration of ways to better manage the groundwater resource. the Groundwater and Soil Conservation Project (FY04). The following lessons can be drawn from the experience: Source: IEG 2006b. PPARs of one rural and two urban water projects in Yemen. In less detail, the review also considered the overall water-sector program in Yemen. IEG's draft Yemen CAE has also discussed the water crisis. 3 3 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 In the Middle East and North Africa, for example, groundwater resources, a situation encouraged a large proportion of irrigation is based on by the huge subsidies for electricity supplied to groundwater resources and most of these are all agriculture. Yet, 13 million hectares within fully exploited. Additional dams to provide India's 388 irrigation projects do not yet receive storage for the unreliable rains is likely to be water because of incomplete infrastructure.4 controversial and expensive, and better agricul- tural water demand management is clearly a Uneven Policy Reform Retards Agricultural solution (box 3.2). Productivity A comparison of the growth in value added by In China there is a water crisis in the eastern agriculture indicates that the Bank's biggest third of the country: groundwater levels are borrowers for AWM appear to have run out of declining (at half a meter a year) and surface and steam and that agricultural growth is slowing. In ground water are being allocated by both contrast, other borrowers have grown consis- deliberative policies and unregulated appropria- tently and more quickly (figures 3.1a and 3.1b.). tion--the implicit values in different uses being The economies where growth slowed shared far removed from economic realities. The value several common characteristics: high levels of of water for municipal and domestic use is a public intervention, highly regulated or ineffi- multiple of 10 to 40 times its value to agricul- cient agricultural markets, few incentives for ture.3 private sector development, poorly targeted provision of public goods and subsidies, and In India, almost all readily available surface water inefficient inputs including water use. is allocated and almost three-quarters of its groundwater is fully utilized. And four of the Agricultural reforms in China spurred productiv- biggest irrigation states--Haryana, Punjab, ity, rural incomes rose 15 percent a year between Rajasthan, and Uttar Pradesh--are mining 1978 and 1984, and the share of the population Figure 3.1a. Declining Agricultural Value Added Is Faltering in Some Countries and Figure 3.1b. But Is Growing Apace in Others Regions 1,800 1,800 China Vietnam Mexico 1,600 Pakistan India Average SSA 1,600 1,400 1,400 1,200 1,200 US$ US$ ha ha 1,000 1,000 per per added 800 800 added alueV 600 alueV 600 400 400 Albania Bangladesh Peru 200 200 Iran Yemen 0 0 1990 1992 1994 1996 1998 2000 2002 1990 1992 1994 1996 1998 2000 2002 Source: World Bank, Statistical Information Management and Source: World Bank, Statistical Information Management and Analysis data. Analysis data. 3 4 G L O B A L R E L E VA N C E R E M A I N S H I G H in rural poverty declined from a third to 11 Similarly, despite the big push for agricultural percent (IFPRI 2005). This was the result of the sector reform in Mexico in the 1980s and early rural nonfarm sector that developed to meet the 1990s, sustained growth in agricultural value growing demand from better-off farmers and to added seems elusive. Between 1979­81 and provide employment for surplus agricultural 2000­02, value added per worker grew by 22 labor.5 In turn, the rural nonfarm economy percent compared with 59 percent for Latin stimulated growth in the urban sector, which America as a whole. A primary reason is that the has been the main engine up to the present. rural workforce is comprised not of market- Even though agricultural diversification grew oriented producers, but of semi-subsistence substantially, the slowdown in agricultural value producers, combining income from production added since the late 1980s has been largely the for their own consumption with casual wage result of state dominance in procurement of earnings, remittances from migrant relatives, food grains, which remains the principal crop and government transfers.7 These problems sector. There is still too high a reliance on collec- reflect the weakness in the incentive environ- tively owned township and village enterprises, ment and the absence of institutions needed to inadequate finance for the private sector and generate and disseminate improved farm small farmers, and restrictions on farm inputs technologies, including AWM. which slow farmers' response to market forces (IEG 2005). Among the key interventions identi- The countries where agricultural growth has fied were a move away from production-based occurred have a number of characteristics in projects toward projects that would raise returns common. All have adopted substantial reforms to labor, increase resource use efficiency, and that have empowered farmers and water user help solve market failures in rural China. Specifi- groups, made determined efforts to increase cally, these include the provision of public goods agricultural water use efficiency, get farmers in relation to water management, environmental involved in paying for agricultural inputs-- sustainability, and agricultural research. particularly for O&M--and have liberalized agricultural policy, and improved markets and In India, minimum support prices and input access to credit. At the same time, these subsidies for agriculture, intended to support countries continued to develop and support all technological innovation and growth, instead other sectors of the economy--in contrast to became inefficient income-support interven- the partial approach in the poorer performing tions (IFPRI 2005). Agribusiness and private countries. The main lesson, therefore, is that sector initiatives are hampered by outmoded reform has to be evenly balanced across the regulations on trade in essential commodities, economy--adopting a stop-start approach does lack of rural electricity that adversely affects not provide the right signals for farmers to agro-processing and related industry, and perform and for the private sector to become policies aimed at protecting rural employment. involved. Most of the sustained growth in Consequently, incentives to increase crop Bangladesh, for example, has been the result of diversification and output are weak. This, allied private tubewell development and market with inefficient water management, groundwater liberalization assisted by better information, a overextraction, and slow diffusion of research, partial result of the Grameen Bank's microcredit has slowed the growth of agriculture. The Indian for village cellular phones. government has recently realized that its neglect of the agriculture sector is affecting aggregate There are clearly huge institutional and policy GDP growth and has, since 2005, embarked on a issues related to rural development, agricultural new agricultural strategy that involves substan- water use, and groundwater that remain tial public investment, particularly in irrigation, unresolved. Not only do they threaten agricul- improved agricultural water management, and a tural productivity but they also have profound focus on rain-fed agriculture.6 implications for potable water supplies and for 3 5 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 equitable sharing of scarce water supplies Current Evaluation Captures Only between rural and urban areas and for environ- Partial Impacts mental management. The Bank has a strong Agricultural water management produces comparative advantage in this multifaceted impacts that are highly relevant to the Bank's arena--natural resources management, agricul- development objectives. Despite this, reporting ture, power, and trade--and thus more support of the results tends to focus primarily on the for policy, sector, and analytical work by the physical achievements and infrequently Bank is highly relevant. describes higher-level impacts on poverty, employment, welfare, and to the rural economy as a whole. The next two chapters discuss why this has occurred. 3 6 4 Outcomes Need Improved Reporting T he previous chapters showed that lending for water infrastructure and institutions has become more closely aligned with the Bank's rural strat- egy and poverty alleviation objectives. Therefore, the relevance of AWM for rural development is high. Overview Outcome and Performance Ratings Evaluation of outcomes discussed in this Between 1997 and 2000 annual average IEG chapter show that there is insufficient attention outcome ratings for AWM projects were margin- to monitoring and evaluation (M&E) of ally better than the Bank as a whole but dropped outcomes and impacts. Robust results relevant by over 30 percent in 2002­03 (as lending to the to the Bank's mission are lacking. The present subsector reached its all-time low), before level of M&E fails to provide adequate informa- recovering in 2004 (figure 4.1). Overall, 74 tion to inform Bank management of progress percent of AWM projects had satisfactory toward strategic objectives--particularly outcomes. Given that most agricultural water poverty alleviation and the Millennium Develop- ment Goals--and needs to be overhauled. Figure 4.1. Outcomes: Satisfactory Ratings Economic appraisal needs to go beyond the project and indicate value added to the sector and economy as a whole--not only the direct 100 impact of infrastructure improvements but also 90 of institutional reform and capacity building. 80 Much greater attention to indicators and evalua- 70 tive frameworks is needed to unambiguously 60 50 determine and attribute the development Percent40 impacts of Bank lending. Better demonstration Weighted by disbursement 30 of the positive impacts on growth and poverty By project 20 alleviation of rural development and AWM 10 would also strengthen the sector's case for an 0 increased budget for economic and sector work 1994 1996 1998 2000 2002 2004 and lending preparation. Source: IEG project database. 3 7 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 Institutional Development Figure 4.2. Sustainability: Likely and Highly Likely Although the overall AWM average of 43 percent Ratings is slightly higher than the Bank average of 41 percent, it is still lower than Reaching the Rural 100 90 Poor's target of 50 percent. As with sustainabil- 80 ity, exit ratings declined dramatically after FY00 70 (see figure 4.3), which suggests the factors 60 responsible for the decline in performance were 50 sown by the lack of attention to institutional Percent 40 30 Weighted by disbursement development and support in the early to mid- By project 20 1990s and its narrow focus.1 Although this 10 project design problem was rectified after 1997, 0 these projects have not yet been completed. 1994 1996 1998 2000 2002 2004 Source: IEG project database. Performance by Bank Region The ECA and Latin America and the Caribbean Figure 4.3. Institutional Development: Substantial (LAC) Regions have performed best with Ratings respect to outcomes, sustainability, and institu- 100 tional development, and the SAR and Africa Weighted by disbursement 90 By project (AFR) Regions have performed the worst (figure 80 4.4). Since 1994, MNA has had the greatest 70 improvement in outcome ratings. ECA exhibited 60 a strong outcome rating cohort in the 50 Percent 40 FY94­FY98 exits, perhaps reflecting the spurt of 30 political and land ownership changes that took 20 place in the region in the early 1990s. In marked 10 contrast, project performance in AFR has contin- 0 1994 1996 1998 2000 2002 2004 ually declined. Source: IEG database. Less than Satisfactory Performance Ratings projects have a life of seven years, these The majority of projects that perform poorly outcomes reflect project design between 1987 primarily do so because of institutional and 1997 and the effect of subsequent supervi- problems. These may affect the government's sion. In comparison with the Bank's Strategic ability to efficiently implement projects or Compact target of 75 percent, satisfactory establish viable mechanisms to sustain them. outcomes were achieved in half of the years. Most instances indicate the need for more thorough appraisal and less institutionally Sustainability ambitious projects given the long time needed The overall average percentage for the period to change attitudes and bring about reform. A FY94­04 had 56 percent rated "likely" or better, review of 75 IEG evaluation summaries since 1998 which is still quite low with respect to the Bank shows that the following factors accounted for average of 60 percent. However, the average most cases of modest institutional development: hides a significant upward trend from among the lowest levels in the Bank to a level of 90 percent in · Poor coordination among implementing agen- FY04 (see figure 4.2). This is higher than the cies; rural strategy target of 60 percent and better than · Lack of capacity within the main implementing the Bank as a whole. agency; 3 8 O U T C O M E S N E E D I M P R O V E D R E P O R T I N G · Insufficient buy-in to sector reform and reor- ganization; Figure 4.4. Performance Comparisons, by Region · Neglect of complementary agricultural ser- vices, credit, marketing, and diversification; 90 and 80 · Weak commitment to cost recovery and/or 70 user participation in system management and 60 operation. 50 Percent 40 30 The Colombia Small-Scale Irrigation Project, for 20 example, attempted to scale up the experience 10 of an earlier project before the implementing 0 agency's capacity was strengthened. As a result, ECA LAC MNA EAP SAR AFR the agency was ill equipped for agricultural and Outcomes Sustainability Institutional development impact community organization and there was little Source: IEG project database. extension support for crop diversification and Note: Percentage of regional project outcomes rated satisfactory, sustainability rated as likely, and essential marketing. Conversely, Mali's Irrigation institutional development impact rated as substantial. Promotion Project failed because civil servants saw the introduction of the private sector as too The average dedicated project was designed to risky and were unwilling to delegate the task to serve slightly more than 150,000 farm more knowledgeable nongovernmental organi- households, mostly defined as small family zations. In Uruguay's Natural Resource Manage- farms. Only a few projects also targeted medium ment and Irrigation Development Project the or large family farms.3 Projects ranged from big main implementing agency could not projects that aimed to serve over 400,000 harmonize the approach of several different families, such as Pakistan's Punjab Private Sector government agencies primarily because the Groundwater Development Project (1997) and scope of the project was not matched to agency Egypt's Third Pumping Stations Rehabilitation capacity. Similar problems affected Nepal's Project (1999), to very small projects that aimed Second Mahakhali Irrigation Project and Indone- to serve less than 10,000 farm families, such as sia's Provincial Irrigated Agricultural Develop- those in Yemen's Groundwater and Soil Conser- ment Project. In view of these lessons the vation Project (2004) or the Dominican current trend to smaller projects is highly Republic's Irrigated Land and Watershed Project appropriate. Box 4.1 summarizes many of the (1995). Nondedicated projects, which are key lessons from completed projects. generally assumed to be quite small, were each expected to benefit, on average, nearly 120,000 What Benefits Were Expected for households--only a small proportion of which Bank AWM? would benefit from agricultural water develop- The 161 agricultural water projects were ment or management. designed to improve national and local institu- tions, build capacity, and directly benefit up to The average area that was planned to be 12 million households and more than 60 million irrigated was around 400,000 hectares per people during the period 1994­2004. About 54 project.4 Omitting the four huge projects million people, comprising 9 million designed to benefit an area larger than one households, are expected to benefit directly million hectares (such as Pakistan's National from the dedicated projects and a further 7 to 18 Drainage Project) drops the average to about million people in nondedicated projects. Thus, 190,000 hectares, with quite a bit of variation the total number of direct beneficiaries could be across and within the time periods. Consistent between 61 and 72 million.2 with findings related to the project's financial 3 9 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 Box 4.1. Main Project Lessons Project Design irrigation and drainage works--this was a problem in the 1996 In- · Institutional reform and policy sequencing are important; do not donesia Village Infrastructure for Java Project. overstretch weak institutions. The 1995 Niger Pilot Private Irrigation Project could not handle the reforms for this reason, while Institutional Issues overly optimistic assumptions about reforming deep-seated · Keep institutional issues high on the agenda and do not allow them institutional problems undermined the effectiveness of the to be crowded out by investment in civil works as occurred in the 1999 Armenia Third Sector Adjustment Credit. Peru Sierra Natural Resources Management and Poverty Project. · Design poverty-targeting mechanisms that are simple, verifiable, · Develop an institutional development road map and detail steps that transparent, and minimize political interference. are important in order to define relationships and the input of · Look for commercial opportunity, agro-processing links, product stakeholder institutions (a lesson from the Indonesia Java Ir- quality, and certification. Although the 1994 Uruguay Natural rigation Project). Clearly spell out institutional responsibilities Resources Management and Irrigation Project supported links (as in the Mexico On-farm & Minor Irrigation Networks Im- to product markets, marketing in low-value products became provement Project). an issue. · Get the balance between public and private focus right and, if nec- · Where appropriate, adopt a basin approach that addresses water essary, improve both. Improved performance of community or- allocation, and integrate higher-level government reform with ganizations is frequently constrained by mistrust of government farmer-level institutional reform. The 1994 Indonesia Java Ir- agencies. Government unwillingness to accept enabling private rigation Improvement Project showed strong performance on sector development frequently slows down market-driven de- basin planning but still had weaknesses in overall integration. mand--this was insufficiently addressed in Albania. · Agree on a plan for supporting services such as extension, credit, and · Make allowances in design for the managerial weakness of com- marketing. The 1997 Nepal Irrigation Sector Project showed munity organizations (the 1997 Mali Pilot Private Irrigation Pro- the risk of relying on a separate initiative to deliver improved motion Project had problems in this area). Social mobilization services; while in the 1995 Vietnam Irrigation Rehabilitation Pro- and "after-care" is important for sustainability of water users ject, the support activities started too late. associations, as demonstrated in the 1997 Pakistan Punjab Private Sector Groundwater Development Project. Implementation · Adapt to cultural constraints and work with local leaders on gender · Public awareness programs are important before major system issues. Do not forget to give this continuous attention during su- changes. In several Pakistan projects, this was a case of "too pervision. little too late." In the Sri Lanka Mahaweli Project a series of · The rationale for water charges needs to be transparent to farm- awareness workshops for stakeholders contributed to the re- ers--clearly identify the incentives. When it is not transparent, form program. they frequently refuse to cooperate. · Train stakeholders in new participation concepts. There was weak · Increase water delivery efficiency before increasing water tariffs. This understanding of what "microcatchment" management and avoids passing on the costs of agency organizational ineffi- "participatory planning" meant to stakeholders of the 1997 ciency to farmers. Peru Sierra Natural Resources Management and Poverty Al- leviation Project. Technology, Knowledge, and Capacity · Operational manuals should be completed before loan approval so · Technology supporting services need to accompany water devel- that all staff understand what is expected and to uncover any opment. The 1997 Mali Pilot Private Irrigation Promotion Project flaws in design. The absence of operational guidelines for the limited dissemination of technology despite strong farmer in- 1996 Indonesia Nusu Tenggara Agricultural Area Development terest. project resulted in poor first-half performance. · Ensure information transparency. Village notice boards were em- · Give more attention to modernization and the construction quality of powering in the Indonesia Village Infrastructure Project. 4 0 O U T C O M E S N E E D I M P R O V E D R E P O R T I N G size, all the projects covering a very large area were designed before 1999. The change over Figure 4.5a. Area of Completed Projects Is Falling time was substantial. After 1998 the average project area fell by two-thirds. Projects designed 200,000 to cover large areas are now rare. 171,593 Benefits Produced Were Less 150,000 (ha) 133,859 Than Expected area 113,648 Outcomes from 71 projects in the portfolio that 100,000 116,890 92,075 have been completed reveal that while all irrigated provide qualitative accounts of policy or institu- tional outcomes, less than half can define verageA 50,000 60,338 quantifiable outcomes and impacts. There are 122 three reasons for this. First, almost a third of the 0 projects (20) could either define benefits only 1992 1994 1996 1998 2000 2002 very generally (for example, the CDD projects) Approval FY or very narrowly, such as the six output-oriented Source: IEG database. emergency-disaster recovery projects.5 Second, planning and setting up of M&E is poor. Third, more recent projects are proportionately more very often there is a lack of relevant indicators expensive (figure 4.5b). because the results chain linking inputs to outputs and impacts is either weakly developed There is also high regional variation in unit area or missing (chapter 5). costs, average project costs per hectare in MNA and AFR being about three times more Of the 43 projects with data, nearly all report the expensive than those in ECA (figure 4.6). When irrigated area, primarily because this is easily average project costs are weighted by area, AFR measured and closely linked to disbursement becomes the most expensive region for irriga- through well-defined contract arrangements. tion investment followed by MNA--and as also Overall, the total area that benefited was 5.45 demonstrated by Kikuchi and Inocencio million hectares (82 percent of expectations) (forthcoming), irrigation investment in AFR and and this averaged 146,000 hectares per project MNA are two to three times more expensive but with a substantial decline throughout the period (figure 4.5a). Unless drainage was a Figure 4.5b. And Future Projects May Be More project objective, such as in the 1996 Estonia Expensive Agricultural Development Project or the 1994 China Songliao Plain Agricultural Development 5,000 Project, it was generally subsumed under the irrigated area and was only reported for 14 4,000 -0.3705 Cost per ha = 3,089 Area percent of projects--even though 40 percent of (US$) 3,000 projects included some drainage. ha per 2,000 For the projects reporting area and cost (38), Cost the overall average per project was $2,123 per 1,000 hectare. When three outliers are eliminated, the average project per hectare cost reduces to 0 $1,293.6 There are significant diseconomies of 0 100 200 300 400 500 scale as project areas get smaller and, therefore, Project area (thousands ha) Source: IEG analysis of 43 completion reports. 4 1 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 attempt was made to estimate an economic rate Figure 4.6. Regional Variations in Infrastructure of return (ERR). In general, project economic Costs Are Large analysis seems to have been allowed to slide in 2,500 the past 20 years, either being avoided or being Average per project Average per ha done peremptorily. In comparison with the rest 2,000 of the Bank, however, the rural sector is more assiduous in carrying out economic evaluations, (US$) 1,500 and more projects are reevaluated at comple- cost tion (figure 4.7).8 In the agricultural water 1,000 subsector, there is a particular need for more Actual attention to project economic analysis to 500 demonstrate growth impact and impacts of institutional reform and efficiency improve- 0 ments. This is particularly important because its MNA AFR SAR LAC EAP ECA economic efficiency is less than most other Bank Region sectors in the Bank (figure 4.8) and is declining. Source: IEG analysis of 43 completion reports. Of the 71 completed projects, slightly under half than the other three regions. Kikuki and Inocen- (32) had implemented AWM and estimated the cio qualify their finding because they found that ERRs at both entry and exit to the portfolio. The AFR projects with acceptable economic remaining projects had either ERRs derived performance (economic rates of return above from non­AWM activities or did not estimate an 10 percent) were no more expensive than those ERR because they were either sectorwide in in other regions--they infer that higher unit cost scope, dispersed CDD-type projects, or were due to the relatively high proportion of emergency projects that focused primarily on unsuccessful projects compared with other short-term reconstruction.9 The overall average regions.7 The six completed Bank projects in ERR estimated at exit was 22 percent, similar to Africa designed after FY93 are too few and varied the overall average at appraisal and close to the to draw any such conclusion. rural sector average of about 20 percent. When Compared with earlier investment, the current Figure 4.7. Economic Evaluations: Bank portfolio of AWM projects is more cost Rural Sector Is Better Than Other effective. Kikuchi and Inocencio show that the Bank Sectors average cost of AWM in the period 1967­2003 80 was $5,021 per hectare, ranging from $6,590 for FY94­97 new construction to $2,882 for rehabilitation. As FY98­01 the current AWM portfolio comprises about 90 60 percent rehabilitation projects by area, the FY02­04 balance being new construction, the average (%) cost should be about $3,253 per hectare. In fact, ERR 40 at $2,123 per hectare it is about a third less expensive. Does this mean that AWM manage- verageA ment projects are now more economic? The 20 next section examines this issue. Economic Efficiency and Competitiveness Are Declining 0 Rural Rural Bank Bank Economic efficiency has been generally satisfac- ENTRY EXIT ENTRY EXIT tory in those agricultural projects where an Source: World Bank data. 4 2 O U T C O M E S N E E D I M P R O V E D R E P O R T I N G weighted for project area, the overall average Figure 4.8. Relatively Low Economic Impact of Rural increased to 23 percent but there is no statisti- Projects, 2001­04 cally significant increase of ERR with project size. Almost 60 percent of the ERRs estimated at Energy and mining appraisal were not achieved and, on average, Health, nutrition, and population appraisal overestimated economic efficiency by 8 percent.10 Transport Information technology ERRs Are Declining Social protection Even so, the average annual ERRs decrease over Environment time, whether plotted against the date of entry Urban development to the portfolio or exit from it, falling from 25 Water supply and sanitation percent in 2002 to 17 percent in 2006 (figure Social development 4.9). Clearly the estimates of project economic Rural sector efficiency provide few incentives for the Bank to Private sector development invest more in AWM. IEG's review found that Financial sector there was no change in ERR methodology Education during the period of review even though there is now general acknowledgment that many 0 10 20 30 40 50 positive and negative externalities are missing Average sector ERR (%) from the analysis. A review of the economic Source: World Bank data. analysis of projects in China, for example, reveals that most appraisals are too simplistic China's Hunan Province, for example, the because they avoid econometric analysis of conversion of rain-fed grain to irrigated early nonproduction impacts that would justify rice increased crop value sevenfold compared investment. In particular the impact of public with the impact of modernization and upgrad- goods and institutions was underestimated (IEG ing of existing irrigation (table 4.1), much larger 2005.) The reasons for the downward trend in than the relative incremental costs.12 ERRs are not clear without detailed econometric analysis and case studies, but the Implementa- tion Completion Reports (ICRs) do reveal some Figure 4.9. Economic Efficiency is Declining of the causes. 40 Lower Commodity Prices. Globally, commodity Exit year prices for most food grains have fallen because 35 Entry year of the overall higher productivity of agriculture 30 and freer trade. As a result, for Indian projects, ERR the parity price for the dominant paddy crop in 25 2004 was 18 percent lower than what was 20 projected in 1997 at appraisal. Similar changes average occurred in East Asia; and in Central Asia, the 15 world price of the principal cash crop of Bank- Annual financed irrigation projects, cotton, fell by more 10 than half between 1994 and 2001.11 5 Smaller Benefits. In addition, the increased 0 emphasis on rehabilitation over new construc- 1990 1995 2000 2005 2010 tion, while reducing unit area costs, also Fiscal year produces much smaller incremental benefits. In Source: World Bank data. 4 3 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 Peru's irrigation-subsector project delay of two Table 4.1. Irrigation Benefits in China's Hunan years was caused by exogenous events: budget Province Are Technology Dependent caps as a result of International Monetary Fund Production value of main strictures, El Niño floods, and the withdrawal of crop per ha (Yuan `000) cofinanciers. With overly optimistic appraisals Sector Before After Change this led to underachievement of physical targets Modernization (early rice) Irrigated Irrigated and insufficient funding for a matching-grants 6.49 7.32 0.83 program to induce farmers to adopt new irriga- New Irrigation Rain-fed Irrigated tion technologies; even so, an ERR of 24 percent was achieved (against 39 percent anticipated). 1.73 7.56 5.83 Source: Yangtze Basin Water Resources Project. Consequences of Inappropriate Lending Delayed Benefits. This is of concern because Instruments delays increase costs and reduce benefit In Indonesia, the $350 million Water Sector streams. Most delays are a symptom of institu- Adjustment Loan was delayed by four years tional problems, lack of local capacity, and overly because project objectives were unrealistic in optimistic appraisal; but some are a result of the time available. Although the first two exogenous events. Many projects are extended tranches were disbursed on time against agreed to take care of procedural problems. In policy actions and institutional reform, the third Kazakhstan, for example, the irrigation and was cancelled because agreed actions were drainage project was extended for a year never completed, even though there were four because of the government's restructuring of one-year extensions--a realistic recognition of the project implementation unit. This delayed the time needed to undertake the ambitious the awarding of contracts, but it did not increase water sector reforms--and acknowledgment engineering costs or reduce benefits--the ex- that an adjustment loan to mitigate Indonesia's post ERR was 32 percent compared with 27 financial crisis was an inappropriate instrument percent ex ante. to implement water sector reform. Although a new water sector law was issued in early 2004 Delayed Benefits Affect Economic Efficiency and regulatory guidelines drafted, these were Completion of the Nepal irrigation sector not issued. Consequently, procedures for water project was delayed by two years and its allocation, water rights, and regulations for estimated ex-post ERR was 10 percent compared farmer-managed irrigation systems remain with 15 percent ex ante. The main reasons for temporary, undermining farmers' incentives to the delay were capacity limitations in the manage systems and pay operational and implementing agency, which slowed procure- maintenance costs. ment and reduced civil works construction; underestimation of how long would be needed Keeping Procurement and Implementation to build farmers' capacity to take over operation Together and management of irrigation systems; and the Having procurement managed independently of need to rebuild substandard works in the hills the implementing agency does not generally and Terai. India's Haryana and Tamil Nadu water work well, as happened in Lebanon and Iran. In resources consolidation project were both Lebanon's Irrigation Rehabilitation and Modern- extended for similar reasons and both cost less ization Project, procurement by the national than expected. Even so, a year's extension in Council for Development and Reconstruction Haryana reduced benefits and lowered the ERR was not well coordinated with the Ministry of from 18 to 14 percent: in Tamil Nadu a 30-month Agriculture. As a result, and compounded by extension lowered its ERR from 15 to 11 percent. problems over farmers' wishes to have agreed 4 4 O U T C O M E S N E E D I M P R O V E D R E P O R T I N G works redesigned and resulting difficulties with Figure 4.10. Neglect of Social Impacts in variation orders, the project was delayed by 30 Completion Reports months. In Iran similar problems arose between the Ministry of Budget and the Ministry of Jihad and Agriculture over the Irrigation Rehabilita- Reports the number 37 tion Project, which led to cancellation of some of farmers benefiting 49 project components. Subcontracting Procurement Overcomes 23 Reports which groups benefit Capacity Constraints 17 When the Dominican Republic's National Institute for Water Resources subcontracted the Appraisal private sector to implement components of the Reports the number 21 Completion Irrigated Land and Watershed Management of people benefiting 17 Project, results were satisfactory. When the institute undertook its own design and procure- 0 25 50 75 ment, however, the results were disastrous, Proportion of projects (percent) (n = 43) leading to a year's delay. Delegation of Source: IEG analysis of 71 ICRs. implementation of Niger's Pilot Private Irriga- tion Project to a specialized nongovernmental South Asia, the same data provides revenue lists organization worked very well, particularly in for recovery of O&M charges, water costs, and overcoming government's skepticism over the land taxes. These data typically contain nothing role of the private sector. Even so, the govern- on the social attributes of the farmers unless ment support for the project was negligible until there are social surveys. It is notable from figure farmers became enthusiastic about the results 4.10 that while the number of projects reporting achieved by the nongovernmental organiza- farmer numbers increased between appraisal tions, after which the government requested an and completion, the reverse is true of which extension of 18 months to build the conditions particular social groups and which people (for necessary for sustainability. example, women) benefit. While farmer numbers are associated with use of project Social Impacts of AWM Projects inputs (training, credit, extension services, and Reports on how many people benefit, their cooperative or user group formation), very few social status, and what benefit they realize are social groups or individuals can be directly not very common. Slightly under half of the linked with such easily measured project inputs. projects report how many farmers benefit but The falloff in information on social impact is less than a fifth report how many people benefit primarily because the special efforts to track the or the social distribution of benefits (figure 4.10). impact of the project interventions are not made because of poor M&E. Reporting of the number of beneficiaries improved significantly after 1998 when the new- Monitoring and Evaluation format PADs required this information in the All AWM projects that have civil works project description.13 In most AWM projects, components have quite good M&E systems to beneficiaries are defined as the farmers within track inputs and related outputs, but the quality the project area and the number is primarily of the system declines as evaluation focuses on from general administrative surveys used to outcomes and impacts.14 Therefore, of the 32 establish the basis for water user groups or projects that calculated ERRs only 2 created associations. In many instances, particularly in "without-project" controls prior to implementa- 4 5 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 attributed solely to the Bank's project-level Figure 4.11. Improved Designs of M&E Systems interventions. The findings from a systematic evaluation of the design of M&E systems in the random selection of 80 AWM projects are 4 discussed below. 15,16 Throughout the study period there was system- atic improvement in the overall quality of M&E score 3 systems. The overall annual average rating increased from slightly above "modest" in FY94 to "substantial" 10 years later (figure 4.11). The average primary reason for the improvement in the Annual 2 design of M&E was the introduction of logical frameworks in the late 1990s and their mandatory use in PADs (figure 4.12). This forced Bank task managers to develop results chains linking inputs to outputs, and in doing so, clarify 1 and simplify development objectives. As a result, 1990 1995 2000 2005 the focus on output indicators, and their overall Source: IEG analysis of 80 PADs, by year of approval. quality, increased significantly--although there Note: 4 = high; 3 = substantial; 2 = modest; 1 = negligible. remains room for improvement. tion and this rose to 3 by project completion. A closer look at scores for individual questions This clearly raises serious question about the reveals, however, that increased attention to robustness of the conclusions drawn by most monitoring the outcomes and impacts only projects that assert improvements in observed occurred in the most recent two to three production, and that farmers' incomes can be years--most attention was given to monitoring indicators of project implementation to provide feedback for better management (figure 4.13). Figure 4.12. Logical Framework Greatly Improved At the same time, project documents showed M&E increased support to build local M&E capacity, particularly as the quality of existing M&E among 4 the new ECA borrowers caused the rating for this question to dip between 1997 and 2002. 3 While the overall quality of indicators improved, only a fifth of sampled projects have good rating poverty output indicators (figure 4.14). 2 Objectives Omitting those projects that do not have direct Quality social impacts--such as the ones that improve Focus on outcome physical efficiency or build only infrastructure, 1 indicators or adjustment operations, the proportion of Logical framework projects with substantial or better poverty indicators increases to almost 40 percent. 0 1990 1995 2000 2005 Viewed statistically, dedicated and nondedicated Approval FY projects show no significant difference in the 12 Source: IEG analysis of 80 PADs, by year of approval. quality ratings, except for a much clearer defini- Note: 4 = high; 3 = substantial; 2 = modest; 1 = negligible. tion of desired outputs.17 This is not unexpected 4 6 O U T C O M E S N E E D I M P R O V E D R E P O R T I N G Figure 4.13. But Attention to Outcome Figure 4.14. Unsatisfactory M&E of Outcome Indicators Needs Improvement Indicators, Particularly for Poverty Feedback loops for Support for M&E Poverty indicators management capacity-building High Overall outcome indicators M&E system in Evaluative baseline place ex ante specified 4 Substantial indicators Modest of 3 Negligible Quality rating None Quality 0 10 20 30 40 50 2 Share of projects in each class (%) Source: IEG analysis of 80 PADs. after project implementation, but not a robust 1 attribution of observed changes. Slightly fewer 1995 2000 2005 than half of the projects did not have any means Approval FY of verifying project impacts--no surveys Source: IEG analysis of 80 PADs, by year of approval. or baselines--even though more than two- Note: 4 = high; 3 = substantial; 2 = modest; 1 = negligible. thirds of them included outcome or impact indicators. These results compare well with given that dedicated irrigation and drainage those of the World Bank's review of impact projects almost always define physical works, evaluation plans.18 crop yield improvements, or capacity-building targets such as the number of water user associ- Results of Poor M&E Design ations (WUAs)--and many of the outputs of Only a third of 32 completed projects in the nondedicated projects are difficult to define random sample had a baseline before the because of their CDD nature. Figure 4.15. Rigorous Evaluation Tools and A rigorous evaluative framework is often missing Attribution of Benefits Needed in AWM projects and robust attribution of benefits is difficult. When projects in the sample Outcome and/or impact indicators and are classified according to attributes that would 11 baseline and control groups allow exogenous and confounding factors to be eliminated, the results are less strong. Only 11 Output and/or impact indicators and 41 baseline percent of projects were designed to have the tools that would allow rigorous impact assess- Output indicators as well as outcome or ment (figure 4.15). Specifically this includes 34 impact indicators, but no baseline well-defined output and outcome indicators, good baselines, and independent control Output indicators, but no outcome or impact 14 indicators or baseline groups unaffected by project interventions that would allow the counterfactual to be 0 25 50 determined. Another 41 percent would allow Proportion of projects with these attributes (%) determination of what happened before and Source: IEG analysis of 80 PADs. 4 7 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 had rectified this by the project start, five by the Figure 4.16. Poor M&E Planning Leads to Delayed middle, and six at the completion. This may not Baseline Creation be a fatal flaw if higher-level evaluation tech- niques are applied, such as random sampling 12 of the project and control area and use of propensity-score matching--Brazil's Ceara 10 project and Algeria's social safety net project 8 Number are examples of this approach. of 6 · Poorly designed and planned M&E almost al- projects 4 ways guarantees late attention to establishing Baseline 2 a baseline (figure 4.16). The Madagascar Irri- Middle & end gation Rehabilitation Project paid almost no at- 0 tention to the design of M&E systems and, Modest Start Substantial consequently, no baseline was ever established. High The Peru Irrigation Subsector Project's mod- Quality of M&E planning est attention to M&E--which was assigned to Source: IEG analysis of 32 randomly selected PADs with ICRs. the project coordination unit--only produced project started and less than half attempted to a logical framework and indicators during the establish a baseline during the project. Other ex- final phase of the project. As a result, there was post surveys only added another 3 percent. no baseline survey and economic benefits were Slightly more than 20 percent never established based on a nonrandom ex-post survey of ben- a baseline. The primary reasons can be traced to efiting households. Conversely, Pakistan's project design and inadequate attention to M&E Groundwater Privatization Project's excellent during supervision: M&E design led to the early establishment of a baseline and controls, and excellent ex-post · When no evaluative baseline is specified it is surveys and case studies. likely to be established very late. Thirteen proj- · While good planning for M&E may lead to ects had no evaluative baseline specified; two early baselines, without vigilant supervision or good local capacity it may not be imple- mented (figure 4.17). The Albania Irrigation Re- habilitation Project had substantial planning for Figure 4.17. Good M&E Needs Vigilant Supervision M&E but it was only implemented at the end of the project. In Sri Lanka's Mahaweli Re- structuring and Rehabilitation Project the in- 40 ability of the project coordination unit to use consultants effectively negated the Bank's at- tempts to get a good M&E system established. Share 30 of Even so, an ERR was calculated using update projects appraisal data but it neglected to take account (%) 20 of the spillover benefits of an Asian Develop- Quality of ment Bank (ADB) project that overlapped the planning Bank's project area. 10 · Early baselines are most often associated with High high-quality ex-post surveys (figure 4.18). De- 0 Substantial tailed impact assessment of the Jordan Agri- Start Modest cultural Sector Adjustment Loan (ASAL) was Middle End undertaken in partnership with GTZ Baseline (Gesellschaft für Technische Zusammenar- Source: IEG analysis of 32 randomly selected PADs with ICRs. beit). The good monitoring data of the Mexico 4 8 O U T C O M E S N E E D I M P R O V E D R E P O R T I N G On-Farm and Minor Irrigation Networks Im- Figure 4.18. Early Baselines Lead to High-Quality provement Project enabled discrimination of Surveys incremental project benefits from the parallel Bank-financed Irrigation and Drainage Sector Project. Pakistan's Baluchistan Community Ir- 40 rigation and Agricultural Project is probably the best practice, particularly in its candid and 30 transparent description of how project im- Share pacts were estimated. of 20 projects Ex-post An econometric analysis of all the variables (%) survey 10 quality affecting the quality of M&E indicates that the major determinants of early baseline creation Substantial 0 are (i) a clear definition of the outputs expected Modest Start and (ii) discussion of the baseline and its use in Middle Negligible End the PAD.19 Conversely, the lack of a baseline is None strongly linked to the size of the project loan and Baseline Source: IEG analysis of 32 randomly selected PADs with ICRs. the quality of the logframe analysis. The latter may seem surprising, but great attention to the logframe at appraisal may have led to Thus, staff training in M&E needs to be subsequent complacency, particularly if a new improved at all levels. More importantly, greater task manager takes over soon after approval. efforts should be made to build and improve This is quite often the case because senior staff M&E capacity in borrowing countries to ensure in Bank headquarters tend to manage projects continuity and sustainability of this essential until Executive Board approval, after which function so as to ensure accountability and good supervision is delegated to country-based staff. governance. 4 9 5 Project Design and Impact W hile the potential of agricultural water management projects to make substantial contributions to growth and poverty are high, most projects lack a results chain that links their interventions to outcomes and impacts. Overview included an explicit poverty-alleviation element. This chapter shows that the design of Bank AWM This is because a well-focused poverty-targeted projects for poverty impact is weak: beneficiar- project design is possible even when there is not ies need to be more clearly characterized, an explicit poverty statement. Until FY01 gender analysis need more attention, targeting dedicated projects were far less effective than needs improvement, and land- and water-asset nondedicated projects at appraising the poverty ownership require more focus. There is room aspects of projects. Since then the gap has for closing the gap between the quite rich litera- closed, a notable achievement (figure 5.1a), and ture and the less rich project design through the quality of the analysis has improved signifi- better training for staff and transfer of experi- cantly for all projects--although dedicated ence between countries. projects lag nondedicated projects (figure 5.1b). Poverty Indicators Need More Attention Gender analysis is now undertaken for three- Only about 34 percent of the 80 projects quarters of all AWM projects, up from the 50 reviewed had an explicit poverty objective percent prevalent during the 1990s. The quality (sometimes, surprisingly, even when the CAS of analysis has also improved following the same had an explicit poverty statement that referred pattern as that for poverty aspects (figure 5.1b). specifically to the rural sector). While the percentage among the nondedicated projects is Analysis of distributional aspects has improved about 45 percent, among the dedicated projects in both coverage and scope. A comparison of the analyzed, only 24 percent mentioned poverty as pooled averages in FY94­FY04 for five social an explicit objective. appraisal indicators (figure 5.2a) shows that more nondedicated projects identified benefici- The extent to which the project design itself was aries and targeted them, identified who and judged to have incorporated a poverty focus was which groups benefit (distributional benefits), rated independently of whether the objectives and projected likely impacts on employment 5 1 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 markedly for dedicated projects (figure 5.2b). Figure 5.1a. Poverty Coverage Is Improving And more attention is being given to distribu- tional aspects by all projects. 100 Dedicated Non-dedicated These findings on employment and income benefits contrast with the now quite widespread 80 use of social assessments. Overall, 53 percent of (%) the projects in the sample indicate that a social 60 assessment was carried out with no statistically sample significant difference between dedicated and of 40 nondedicated projects. However, there is a clear Share improvement: from only 22 percent of projects in FY94­FY99, the number with social assess- 20 ments rose to 79 percent in the past five years. 0 Weak poverty indicators are attributable partly FY94­97 FY98­01 FY02­04 to the lack of poverty objectives or poverty Source: IEG analysis of 80 PADs by year of approval. design. Another explanation may be that many staff appear to feel that measuring poverty change is simply too difficult. While measuring Figure 5.1b. Quality of Appraisal of Poverty Aspects poverty is clearly not easy, the literature offers a Is Improving number of indicators.1 Examples of somewhat better cases of monitoring impact include the 4 1997 Peru Natural Resources Management Dedicated Non-dedicated Project and the 2004 Nepal Poverty Alleviation Fund, both, not unexpectedly, nondedicated projects. The 2004 India Madhya Pradesh Water 3 Sector Restructuring Project, a dedicated appraisal project, has a good analysis of the percentages of of the poor in the selected areas and it projects how many will be lifted above the poverty line Quality 2 by the project, and how much employment will be generated. It outlines some actions to reach the poor, including support for fish ponds and the use of tribal development plans. 1 FY94­97 FY98­01 FY02­04 The contrast between the previous finding that Source: IEG analysis of 80 PADs by year of approval. analysis of poverty is weak and the increased Note: 4 = high; 3 = substantial; 2 = modest; and 1 = negligible. application of social assessment is puzzling. One explanation might be simply that the social and incomes. In contrast, there are slightly more assessments are still not adequately analyzing dedicated projects with social assessment. Even who the poor are and why. A second explanation so, high-quality analysis of the specific benefits is might be that more analysis is being done than is done for less than half of all AWM projects, and being revealed in the PAD. If so, this presents a impacts on employment and incomes are problem for managers and reviewers. Another projected for less than a third of projects. explanation might be that social assessments often come too late in preparation. Observation As with the overall poverty focus, the quality of suggests that all three of these reasons may be appraisal for benefit targeting has risen involved. 5 2 P R O J E C T D E S I G N A N D I M PA C T Assessments of Water User Associations Figure 5.2a. Less Appraisal of Social Impacts in Are Too Optimistic Dedicated Projects There is considerable disillusionment with the performance of water user groups or associa- 100 tions and a common view that, for some years, Dedicated Nondedicated the Bank has had unrealistic expectations.2 80 While there is widespread support for the princi- ple of user participation to improve water (%) management and maintain irrigation systems-- 60 and there is much evidence to show that user sample participation is effective--staff also recognize of 40 that farmers often lack the skills needed to manage the larger irrigation systems and that Share the need for continuing government support 20 has been underestimated. Also, Bank staff have tended to treat the strengthening of user associ- 0 ations as an end in itself, rather than as part of a Benefit Analysis of Social Impact on broader strategy of raising the efficiency of water targeting distributional assessment employment and benefits incomes use--this is a reflection of the increasing Source: IEG analysis of 80 PADs by year of approval. proportion of nonwater staff designing AWM projects. Staff expressed a concern that, in some cases, the Bank has bypassed long- Figure 5.2b. But All Projects Have Improved standing public sector frameworks for water the Quality of Analysis management, failing to appreciate the potential of existing institutions and overestimating the 4 scope for reform. FY94­97 FY98­01 FY02­04 Institutional Support for Water User Associations Is Neglected 3 Projects have tended to give more emphasis to appraisal strengthening WUAs than to strengthening the of broader authorization and institutional framework in which they must function. 2 Quality Strengthening of user or community groups was substantial--a rating that holds for the nondedicated as well as the dedicated 1 projects. By comparison, institutional strength- Dedicated Nondedicated Dedicated Nondedicated ening at higher government levels was only Benefit targeting Distributional analysis modest. In the dedicated projects the point of Source: IEG analysis of 80 PADs by year of approval. reference was usually a WUA, whereas for nondedicated projects it was typically a broader Note: 4 = high; 3 = substantial; 2 = modest; and 1 = negligible. community group only partly focused on managing water.3 that enable community groups to be effective managers of water. The appraisal documents reviewed suggest that there tends to be too narrow a focus on increas- The Bank has found that "User participation ing participation by user or community groups. changes, but does not eliminate, the role of As illustrated in figures 5.3a and 5.3b, there has government agencies in irrigation development. been insufficient attention to the critical issues Building support from policy makers and agency 5 3 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 Some projects seem to get the balance right Figure 5.3a. Limited Focus on Good Water between strengthening user associations and Management Essentials in All AWM Projects strengthening counterpart institutions. The 2000 Kyrgyz On-Farm Irrigation Project 4 100 established and provided training to govern- Quality of analysis Projects with this attribute (%) ment support units for user associations. These 80 support units, located in the Department of (%) 3 Water Resources at the oblast (regional) and 60 appraisal rayon (district) levels, are expected to become a sample of of permanent part of the organizational structure.6 40 The 1995 Egypt Irrigation Improvement Project Quality Share 2 invests $10 million in broad-based institutional support, and will spend a further $5 million to 20 introduce an Irrigation Advisory Service that will provide information and technical support 1 0 aimed at strengthening user groups, as well as a Strengthening Support to Support for Focus on Addressing Peoples' community- WUAs on autonomy water use shortfalls water campaign to raise awareness of environmental based water of WUAs efficiency in cost rights issues. However, these are the exceptions. Few organizations control recovery appraisal documents indicate how support to Source: IEG analysis of 80 PADs, by year of approval. water-user groups will be backed up with Note: 4 = high; 3 = substantial; 2 = modest; and 1 = negligible. support to the broader institutional framework of which they are a part or how support will be maintained when the project ends. The Figure 5.3b. Focus of Essentials for Nondedicated documents also do not contain a plan for the Projects Are Particularly Limited gradual phasing out of support as the user groups mature. 100 Dedicated Nondedicated In Central Asia the institutional environment for 80 WUAs is particularly weak. In Uzbekistan, (%) Kazakhstan, and Kyrgyzstan, a number of 60 problems have been reported: corrupt officials sample who interfere with the operation of user groups, of 40 illegal water withdrawals by the politically well Share 20 connected, poorly managed government systems of credit and input supply, controlled 0 commodity markets, and cases of extortion at Strengthening Support to Support for Focus on Addressing Peoples' highway checkpoints and in bazaars. This is community- WUAs on autonomy water use shortfalls water compounded by severe inequalities in the distri- based water of WUAs efficiency in cost rights organizations control recovery bution of land and assets and a backlog of irriga- tion maintenance that dates back to Soviet Source: IEG analysis of 80 PADs, by year of approval. times.7 All these issues hamper attempts to make WUAs effective. staff as well as farmers and other water users is essential for successful participatory projects Unrealistic Incentives for Cost Recovery and involves paying close attention to the There are also unrealistic expectations about incentives relevant to each group."4 Bank cost recovery following the handover to user strategy argues that irrigation reforms should be groups. Many of the appraisal documents supported by broader institutional reforms if reviewed contain ambitious timetables for WUAs are to function effectively.5 recovering operation and maintenance costs 5 4 P R O J E C T D E S I G N A N D I M PA C T Box 5.1. Madagascar: Unrealistic Strategy for Transferring Responsibility for O&M The appraisal report for the 1995 Madagascar Second Irrigation ing to 90 percent by the final year. During implementation the in- Rehabilitation Project said about the first project (which had centive for groups to take O&M appeared to have been under- been completed) that "the level of financial contributions to O&M mined by continued government support and no exit strategy. In does not yet reflect a sufficient level of commitment." However, response to this slow project progress, instead of reviewing it is not clear in the appraisal report for the second project how strategy, the Bank pushed ahead with investments in large the incentive for users to pay up would be increased, other than schemes--apparently in order to restore the level of disburse- a mention that "new regulations on handover are being finalized" ments. Yet, as the completion report shows, it was precisely in and that O&M responsibilities "will be defined." The completion these larger schemes that the social cohesion needed to achieve report for the same project showed that cost recovery remained satisfactory handover was lacking, and the challenge of cost low, never exceeding 15 percent--according to the appraisal re- recovery was largest. port, it had been projected to reach 15 percent in year 1, and ris- from users (boxes 5.1 and 5.2). For dedicated financial incentives and the considerable projects the projected rate of increase in the uncertainties in such a "blind" handover is a recovery of operation and maintenance costs problem that project design usually overlooks. averaged 65 percent.8 Several projects aimed Assessments of the impact of WUAs are few. both to establish user associations and to Notable exceptions are in the Philippines and recover 100 percent of costs--within the span of Indonesia.10 In Indonesia a 2002­03 study by the a single project typically lasting less than seven University of Gajah Mada compared the quality years. The appraisal documents typically contain of irrigation improvements, with and without no fallback strategy if the targets prove to be WUA involvement. They found conventional unrealizable--particularly as the civil works rehabilitation without WUA strengthening had required to provide timely and adequate water ERRs in the range of 10­18 percent; in those supplies take most of the project period to with enhanced WUA capacity this increased to complete. the range of 30­40 percent. It is perhaps unfair to expect users to commit to Morardet et al. 2005 found that, in eastern and handover if it is not clear to them how much southern Africa, irrigation management transfer they will be expected to pay for O&M following was incomplete in all the countries they studied, transfer.9 While tariff projections may be difficult with particularly marked lags in Nigeria for public agencies to make given the variability and Madagascar. They note that the classic of systems and handover changes, the lack of remedy has been to combine increased public Box 5.2. Tanzania: The Extent of Cost Recovery The appraisal report for the River Basin Management and Small- basin, which is impossible to interpret given the absence either holder Irrigation Improvement Project states that "a system of of the target or of some estimate of what 100 percent recovery water charges . . . would be put in place to ensure that the O&M would amount to in absolute terms. Another annex table shows costs . . . would be covered." The report goes on to say that full what percentage of users paid up but, again, fails to specify recovery will be achieved in three years. The completion report the target set at appraisal. There is a vague statement that for this project fails to make clear to what extent this target was "funding . . . raised from water user fees . . . has increased over realized. The report notes that "Each scheme has devised its own time . . ." Therefore, there is no clear indication of how close to mechanism of attending to O&M . . ." The annex on performance recovering costs the project came--a failing shared by most of indicators quotes a dollar figure for the amount collected by the completion reports reviewed for this study. 5 5 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 spending on O&M with higher water charges; supply approach, such as the South Asia but, typically, this has not led to better mainte- warabundi system. It is rare for appraisal nance. documents or country water strategies to detail the various steps that must be followed if real Incentives to Boost Water Use Efficiency water use efficiency is to be achieved--"real" in Are Frequently Neglected the sense of allowing for the reuse of a percent- There is a disparity between the richness of the age of "lost" water. This is perhaps surprising literature on water pricing and efficiency and the given the extensive literature on the issue (such limited coverage of these issues in the appraisal as Perry 2001).14 Most appraisal documents documents reviewed for this study. It is often simply outline a strategy to fix water delivery difficult to discern from these documents what infrastructure, to raise cost recovery, and to system of water charging has been used in the increase participation through handover. past or is proposed for the future.11 The China Water Conservation Project is one of The pricing of water to achieve efficiency is the few examples of an appraisal document extremely complex; and in many borrowing offering clear guidance on how to achieve countries, it is unrealistic to apply marginal water system and technical water efficiency. Drawing pricing through volumetric outlets and charges. from the literature, it clearly articulates the issue There are exceptions to the general trend-- of "real" water savings, noting that although South Africa, for example, is moving toward systems higher in a basin may be inefficient, the volumetric pricing, and in Armenia and Iran the basin as a whole may be efficient because it principle is accepted. Even where volumetric allows for recovery of losses. The appraisal pricing is, in principle, feasible, it may be document succinctly demonstrates how the thwarted: in Jordan, the rate of meter tampering design of the system and the operating process was as high as 20 percent in some areas.12 influence the level of unrecoverable losses. The project includes activities specifically designed Alternatives to water pricing may be more practi- to address efficiency--physical improvements cal. One interesting variant is to charge by area to the system, agronomic measures, and according to crop type, applying average water management reforms.15 consumption for that particular crop, but allowing farmers who think they can do better Even where there has been agreement to pay for by using a volumetric charge, and providing the water, the means to measure volumes delivered meters at their own cost. Various studies have are not built into most Bank projects, except for shown that some alternatives to metering give the few concerned with groundwater. This is almost as good water use efficiency at lower because most large surface water projects were collection costs. Perry (2001) notes that many of designed for top-down supply management by the assumed advantages of water pricing at the public agencies, where equitable rotational margin can be achieved through physical distribution was more important than worrying rationing, which is easier and more transparent about what individual distributaries or farmers to farmers.13 Furthermore, from an equity received. As a result, the equipment and capabil- standpoint, it may not make much of a differ- ity to regulate supplies according to demand is ence what pricing approach is used; farm size is missing. What is needed is not only rehabilita- far more important (Tsur et al. 2004). tion but modernization of hydraulic control structures to allow volumetric measurement. Few appraisal documents or country water strategies offer a clearly articulated strategy for Irrigation Is Only Part of the Solution water use efficiency, whether it is through some There is scope for increasing the complemen- form of volumetric pricing, an area/crop-based tarity among irrigation investments and approach, a lower-cost proxy, or a restricted- extension, marketing, and credit services, partic- 5 6 P R O J E C T D E S I G N A N D I M PA C T ularly for dedicated projects.16 While there was a policy reform and social issues relevant to irriga- big increase in the share of irrigation projects tion, but also a continuing loss of staff with that addressed credit and marketing constraints technical and engineering skills (mainly through after FY98, most of this increase was for nonded- retirement). Some managers have been hiring icated projects (figures 5.4a and 5.4b). Comple- (or recruiting through cross-support from other mentarities among extension, marketing, and regions) staff who know how to work through credit services find their strongest expression in the details of policy and institutional reform, the India Sodic Lands Project, the Azerbaijan aiming to use these experts to convince Irrigation Distribution Systems and Manage- wavering borrowers in their region to embrace ment Improvement Project, and the Niger the reform agenda. Northeast­Private Irrigation Promotion Project. Conversely, the 1994 Mexico On-farm and Minor Managers expressed no strong views about Irrigation Networks Improvement Project--a training and most had no training plan, follow- $569 million project--apparently failed to marry ing Bank policy that it was up to staff to take the extension and credit services effectively, compro- initiative. Some staff reported that training on mising the objective of raising productivity. institutional reform and water use efficiency had almost ground to a halt. The irrigation study Bank Staffing and Training Are Weak tours are no more--even though these were Most sector managers interviewed reported a well received by staff in the past. Informal continual shortage of staff with the right irriga- lunchtime seminars are still common but are tion and drainage expertise--even though often sparsely attended, reflecting the pressure about half had either recently hired persons on staff to focus on the day-to-day details of with these skills or had some recruitment project processing. Staff were moderately pending. In Africa, recruitment of a lead water enthusiastic about "Water Week" but those management specialist is at the top of the specializing in agricultural water use tended to regional priority list. The overall impression was report that the event had been captured by the of a slight increase in staff equipped to address urban water and sanitation contingent. Figure 5.4a. Coverage of Credit and Figure 5.4b. But Only in Nondedicated Marketing Links Has Improved Projects 100 100 FY94­98 FY99­04 Dedicated Nondedicated 80 80 60 60 Percent Percent 40 40 20 20 0 0 Extension Credit Marketing Extension Credit Marketing Source: IEG analysis of 80 PADs. Source: IEG analysis of 80 PADs. 5 7 6 Findings and Conclusions T his study provides answers to three questions: Why has Bank invest- ment in agricultural water management declined so precipitously? Are agricultural water projects relevant to the Bank's renewed focus on poverty alleviation and institutional and policy reform? What should be done to improve performance and relevance? Changing Global and Bank Priorities countries, agriculture's value added to growth in Four factors account for most of the decline in the 1990s shrank to about half of that added by lending for agricultural water management: the industry and service sectors, and about a (i) a shift in borrowers' priorities, (ii) realign- quarter of that added by exports of goods and ment of lending with the Bank's strategy for services. Food security concerns that were the poverty reduction, (iii) changing development focus of agricultural development in the 1960s to objectives for agricultural water management, 1980s were mostly assuaged. Declining prices of and (iv) increased use of low-cost approaches. staples--particularly irrigated rice, helped by improved nonwater inputs, markets, and trade-- Borrowers' Priorities Have Changed increased the food access of the poor. In most Agriculture's contribution to growth and developing countries agricultural production employment continues to shrink, lowering has, therefore, met performance expectations. policy makers' attention to agricultural policy and water management. Between 1980 and 2000 Dramatic growth of urban populations in the share of agriculture in global GDP fell from 8 developing countries poses severe economic, percent to 5 percent, but the regions with the political, and social challenges that have most agricultural water management infrastruc- displaced the attention given earlier to rural ture saw an even greater decline. The share fell development. The rural population is antici- by a quarter, to 25 percent, in South Asia and pated to decline slightly from 3.3 billion in 2003 halved, to only 13 percent, in East Asia and to 3.2 billion in 2030. Forty-eight percent of the Pacific. The only region where agricultural GDP world's population lived in urban areas in 2003, remained unchanged was Sub-Saharan Africa and this is projected to rise to 61 percent by (17 percent); and in the Middle East and North 2030. With competition for both water and Africa it actually increased, from 10 percent to 14 scarce financial resources, political preference percent. Among all low- and middle-income has been given to the provision of basic water 5 9 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 supply, sanitation, and environmental needs investments in civil engineering works. Even so, because agriculture is regarded as primarily a Bank lending for agricultural water management private sector activity. Disillusionment among was primarily to the poorest countries. As the governments and policy makers about poor volume of lending sharply contracted in the performance and maintenance problems of period after 1999, an increasingly larger portion public sector irrigation also quelled interest in went to the lower-income group and this the subsector. Much of this was because the reached more than 95 percent in 2002 (though expansion of irrigation infrastructure outpaced investments in Africa also declined). public management capacity and local institu- tions. Therefore, diminished attention to A new set of smaller clients also emerged as agriculture among the Bank's borrowers was commitments for agricultural water manage- consistent with increased attention to social ment to the biggest borrowers declined. Before issues, urbanization, and growth. 1999, ECA accounted for only 11 percent of projects in the portfolio; afterward, with 29 Realignment in Lending loans, it accounted for a third. The Bank's Lending was realigned with the Bank's strategy smaller interventions in ECA not only assisted for poverty reduction for two reasons. First, poverty alleviation in the medium term but also Bank policy changes increased the share of provided an entry point for policy discussions lending allocated to the social sectors. Second, aimed at rationalizing the region's aging and agriculture was not on the fastest growth path. oversized infrastructure, which was frequently Following its 1990 World Development Report environmentally damaging and uneconomic to on Poverty, the Bank adopted a two-pronged operate. strategy that targeted efficient, labor-intensive growth and greater attention to social concerns, Budget Constraints Squeezed AWM including education and health care, a strategy Budget constraints within the Bank and new later reemphasized in the 2000/2001 World initiatives squeezed out AWM projects. The 1997 Development Report: Attacking Poverty. With Strategic Compact: Renewing the Bank's the renewed focus on poverty, lending to the Effectiveness to Fight Poverty significantly social sectors increased while lending for reduced budgets for project preparation, a trend infrastructure, agriculture, and the environment accelerated by a substantial shift toward fell after 1993. IDA replenishment agreements development policy lending during the mid- to (IDA10­12) also required increases in the share late 1990s. At the same time, the skills mix of of investment lending in the social sectors, and Bank staff was realigned with the Compact, by a the HIPC initiative required beneficiary loss of technical staff and replacement with staff countries to allocate funds released from debt having more fungible skills. Enhanced fiduciary service to public expenditures on the social and safeguard provisions increased the costs of sectors. As social sector investment increased, project preparation such that AWM projects are lending for infrastructure declined. among the most expensive to prepare. Squeezed by budget pressures, high costs, Underpinning the focus on poverty was growing muted advocacy, and new development initia- evidence that accelerating economic growth was tives, country directors' interest in AWM waned. the fastest way to raise people out of poverty. Since 2002, budgets and staffing have modestly Yet, though most of the poor live in rural areas, improved and, refocused by the new rural and development of agriculture has not been the water sector strategies, lending for rural path to the most rapid economic growth. This development and AWM has shown a resurgence. fact, coupled with serious concern about the The main lesson is that vital investment in rural environmental and social impact of several large- areas and AWM will not take place unless scale projects--particularly those for water-- directors' incentive structures are reformed and caused the Bank to reduce its support for public budgets that are commensurate with the 6 0 F I N D I N G S A N D C O N C L U S I O N S challenge are provided to enable staff to be reasons: a change in the type of infrastructure effective. financed and the greater emphasis on nonstruc- tural and capacity-building components. Changes in Development Objectives for AWM Freestanding projects dedicated to water Evaluation of the country assistance strategies and management now comprise only about 40 projects approved during the period 1994­2004 percent of the agricultural water management shows a change toward a more integrated portfolio. There is a marked difference in the approach to rural development, with a growing type of infrastructure components financed by emphasis on building social capital. Project dedicated and nondedicated projects, even objectives encompassing community support though most contain a mix of physical interven- and participation, income and employment, and tions ranging from some new-builds, redesign support for capacity building and institutional and upgrading, and repair of damage caused by development increased. Conversely, objectives deferred maintenance, referred to as rehabilita- that are central to the new policies--addressing tion. Among the dedicated projects, rehabilita- poverty reduction, agricultural development and tion or improvement of large irrigation systems production, and environment and natural now account for more than 80 percent of Bank resources management--declined in importance. commitments. Nondedicated projects, with an One reason for these changes is that develop- initial focus on rehabilitation in the mid-1990s, ment objectives have become more practical now build new systems that are small scaled, and achievable by focusing on measurable community owned, and integrated in social outcomes rather than global targets. For development programs. Because rehabilitated example, increased attention to income and projects averaged $2,900 per hectare, while new employment almost balances the decrease in construction averaged $6,600 per hectare, there poverty-reduction objectives. was a substantial fall in the cost of projects. As a result, the average loan amount per project fell Attention to the technical and social issues of from $59 million in 1994 to a low of $15 million agricultural water management has become in 2001. more polarized. This may not be an issue where agricultural water management projects are part AWM Remains Relevant of a broader package of rural development Demand for increased global agricultural endeavors that deal with social, human, and production will require better management of economic development. But the more general increasingly scarce water resources. As the projects, in which water-related activities are in world's population grows from its present 6.5 the minority, are building water infrastructure billion to 8.2 billion in 2030, the FAO projects with less attention to issues of technical that a new round of investment in irrigation and efficiency and sustainability. These findings drainage will follow, albeit at half the average indicate the importance of integrating agricul- rate of the preceding four decades. The balance tural water management projects within country of new arable land will come from developing rural strategies and ensuring that they are countries that have the potential to add about adequately supported either by parallel 120 million hectares of new arable land. The operations that address critical omissions, or by expansion will be strongest in South Asia, East improving the skills mix of appraisal teams Asia, the Middle East, and North Africa regions preparing agricultural water management where almost all arable land potential is utilized. components of nonwater projects. Harvested irrigated area, subject to multiple cropping, is likely to increase by a third, or 83 Low-Cost Approaches Are Increasingly million hectares by 2030. Important The average Bank commitment to agricultural Better regulation and management will be water management projects declined for two required because of more competition for water 6 1 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 and degradation of supplies owing to pollution Even so, it is the "package" that matters for and reduced investments for infrastructure effective poverty alleviation and not just the maintenance. Globally, water is becoming an supply of irrigation water. Investments in increasingly scarce commodity--more than a agricultural water management may not reduce quarter of the developing world population will poverty directly in any significant way unless face severe water scarcity in the next 25 years. accompanied by other complementary interven- And groundwater, the mainstay of most private tions. sector investment in South and East Asia, the Middle East, and North Africa regions, already is Increasing Relevance and Performance being extensively overexploited and mined. The The relevance of agricultural water management net effect will be increasing real costs of water at operations to borrowers and to Bank country the farm level and declining social profitability of directors can be increased through better irrigated agriculture. analysis of links to economic growth, more attention to demonstrating social impact and Irrigation boosts growth and reduces poverty poverty reduction, and better management. directly and indirectly, benefiting the poor in While most CASs discussed the importance of several ways. Poor farmers directly benefit from agriculture policy, less than half discussed it in increases in their production, which may the context of economic growth; greater increase their own consumption and provide a prominence was given to community-driven surplus of marketed products for increased farm development, general rural development, and income. Small farmers and landless laborers reform of agricultural institutions. benefit from agricultural employment opportu- nities and higher wages, and a wide range of Demonstrate Growth Impact rural and urban poor benefit from related This is particularly important as the economic growth in the rural and urban nonfarm efficiency of all rural sector projects is less than economy. Crop harvest from irrigated areas most other sectors in the Bank, and is declin- leads to strengthened staple or nonstaple food ing--it is ninth among the 13 sectors reporting output, which lowers prices and benefits all measures of economic efficiency. Investment in consumers, particularly the poor. agricultural water management is economically efficient but is becoming less competitive. The Agricultural growth generates important income annual average economic rate of return for and employment multipliers within the completed agricultural water projects steadily surrounding nonfarm economy. The multipliers declined from 25 percent in FY00 to 17 percent are particularly large in Asia, between 1.5 and 2.0 in FY06. The primary reasons for this are of the incremental agricultural benefits diseconomies of scale, as average projects generated, but they are only half as large in became smaller in area, global commodity Africa and Latin America. Multipliers are higher prices declined, and benefits were smaller and in labor-abundant regions, and increase with delayed. regional development and per capita incomes. Specifically, irrigated regions dominated by Measure Social and Financial Impact medium-sized farms and modern input- Reports on how many people benefit, their intensive farming systems generate the largest social status, and how they benefit are not very multipliers. Multipliers are smaller in rain-fed common despite a substantial increase in the farming systems and in regions dominated by use of social assessment. Slightly less than one- very small farms or large estates. This poses a half of the projects report how many farmers dilemma for decision makers: a poverty-targeted benefit but less than a fifth report how many intervention aimed at small farmers may not be people benefit or the social distribution of the most efficient way of increasing agriculture's benefits. While the number of projects report- contribution to economic growth. ing farmer numbers increased between 6 2 F I N D I N G S A N D C O N C L U S I O N S appraisal and completion, reports on outcomes Increase Focus on Policy and Institutional for particular social groups and people (particu- Reform larly women) declined. While farmer numbers PADs make only modest proposals for policy are associated with the use of project inputs reform and completion reports usually conclude (training, credit, extension services, and cooper- that reform expectations at appraisal were ative or water user group formation), very few unrealistic, particularly for cost recovery. social groups or individuals can be directly Dedicated irrigation and drainage projects with linked with such easily measured project inputs. policy content--large or small--only give it The falloff in information on social impact is modest attention. Many appraisal documents primarily because the results chains linking implicitly assume either that policy reform is inputs to critical outcome indicators is largely complete, or that it is beyond the frequently missing, a problem exacerbated by project's scope--particularly where irrigation poor M&E. and drainage was only one of many components, or where the size of the invest- Improve Monitoring and Evaluation ment was small. Yet, in many cases, important Current M&E does not provide adequate policy issues remain to be tackled. Therefore, information to inform Bank management of the Bank frequently scaled back lending for progress toward strategic objectives--particu- irrigation before the policy reforms needed to larly poverty alleviation and the Millennium get the balance right between public and private Development Goals--and needs an overhaul. intervention were completed; examples are Overall quality of M&E design improved in the Morocco, Nepal, and the Philippines. Yet there late 1990s with the introduction of logical have been notable successes, particularly in the frameworks and their mandatory use in PADs. ECA Region and Egypt. Even so, the quality of the M&E systems declined as evaluation has increasingly focused Build Support for Water User Groups on outcomes and impacts. Only a third of There is considerable disillusionment with the completed projects had a baseline before the performance of water user groups or associa- project started and less than half attempted to tions and a widespread view that, for some establish a baseline during the project. Slightly years, the Bank has had unrealistic expectations more than 20 percent never established a for them. While the principle of user participa- baseline. And only 9 percent of projects that tion is still widely supported, farmers often lack calculated ERRs created "without-project" the skills needed to manage the larger irrigation controls. This raises questions about the robust- systems and the need for continuing govern- ness of the conclusions drawn by most projects ment support has been underestimated. that assert improvements in observed produc- Projects have tended to give more emphasis to tion and farmers' incomes and that attribute it to strengthening WUAs than to strengthening the the Bank's project-level interventions. Even broader authorization and institutional when there was good M&E design, inadequate framework in which they must function. They supervision--possibly because of the also did not contain a plan for the gradual widespread practice of delegating supervision phasing out of support as the user groups to country staff--sometimes reduced effective mature. implementation of M&E. More training of all staff is indicated. Current ICR guidelines would Move Beyond Simple Cost Recovery benefit from a mandatory section on who the Expectations about cost recovery following beneficiaries are and how they benefit. Much handover to user groups are frequently unrealis- greater attention is needed to establish indica- tic and too ambitious. Most appraisal documents tors and evaluative frameworks to unambigu- simply outline a strategy to fix water delivery ously determine and attribute the development infrastructure, to raise cost recovery, and to impacts of Bank lending. increase participation through handover. Few 6 3 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 offer a clearly articulated methodology for 1998--most of this increase derives from improving water use efficiency, whether it be nondedicated projects. through some form of volumetric pricing, an area/crop-based approach, or lower-cost proxy, Correct Staff Mix Is Important or a restricted-supply approach, such as the Markedly different strengths and weakness South Asia warabundi system, and too few link between dedicated and nondedicated water this to the redesign of water supply systems. projects are related to the skills base of task Simultaneous attention to community managers. Nondedicated projects scored highly operation, management, and physical modern- on social and institutional factors but poorly on ization of water distribution networks is not very attention to the quality and sustainability of common, reducing the efficacy of both interven- their (minor) water-engineering components. tions. Where this is done, the results can be Conversely, dedicated projects were good on outstanding, as shown in China's Tarim Basin the engineering but tended to neglect institu- and in Armenia. Where the potential synergy is tional issues, social concerns, and incentives for not captured, the outcomes have been farmers and organizations to improve their disappointing. efficiency. In the past, these omissions were taken care of by parallel operations--but with Embed AWM in Sectorwide Approaches the shrinkage of rural sector lending this is a The complementarity among irrigation invest- problem. Sector managers expressed no strong ments and extension, marketing, and credit views about training and most had no training services can be improved, particularly for plans, following Bank policy that it was up to dedicated projects. While there was a big staff to take the initiative--perhaps it is time to increase in the share of irrigation projects that introduce training plans to mitigate revealed addressed credit and marketing constraints after weaknesses. 6 4 APPENDIXES APPENDIX A: STUDY METHODOLOGY Overview and the FAO's 2004­06 agricultural statistical The report is based on a desk study that databases. reviewed a wide range of instruments, products, · Informal review of economic and sector and databases: work. Forty-four documents from economic and social work, from 1999 to 2004, on agri- · Portfolio analysis. One hundred and sixty- cultural water were used to update the data- one PADs approved from 1994 to 2004, cov- base used for the 2002 "Bridging Troubled ering all 6 regions and 56 countries were Waters" report, to identify common themes analyzed. A random selection of 80 projects and emerging issues. were subjected to detailed evaluations of proj- ect design, including development objectives, 1. Portfolio Analysis results chain linkages, and M&E. The trend analysis of the lending patterns in · Analysis of country assistance strategies. time and by region drew primarily upon sources One hundred and thirty CASs, covering 54 within the World Bank and IEG, and particularly countries during the past 10 years were ana- upon the Business Warehouse database, lyzed in relation to five major areas of con- through December 2004. Most lending cern: irrigation, water resource management, operations were recoded by the Bank back to agriculture, rural development, and poverty FY1990 because that year was chosen as the base alleviation. year for measuring progress toward the Millen- · Meta-evaluation of IEG's country assis- nium Development Goals. tance evaluations. Thirty-two Country As- sistance Evaluations were analyzed to support The new coding system allows for up to five the meta-evaluation of the Bank's perform- themes and five sectors per activity. Themes are ance in irrigation, water resources manage- clearly separated from sectors, with themes ment, and agriculture; and to gather evaluative corresponding to the goals/objectives of Bank comments on effectiveness in identifying prob- activities, and sectors indicating the parts of the lems and strategies in those sectors. economy that receive the Bank's support. Every · Project completion and performance as- operation is coded along both the sectoral and sessment reports. ICRs and PPARs from 92 thematic dimensions. However, the system completed agricultural water projects were ex- omits projects where the amount committed to amined for common lessons from the past irrigation and drainage is too small to be decade and to observe emerging trends. included as a sectoral component, which · Interviews with Bank managers and sec- requires some caution in interpreting the figures tor specialists. (IEG's detailed sampling of appraisal documents · Literature review. found that in several dedicated projects the cost · Time series analysis. Various time series of agricultural water components was almost 20 data was gathered from the Bank's Statistical percent more than those identified by using only Information Management and Analysis database the sector code). This is especially the case for 6 7 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 CDD projects,1 such as many Social Funds Sampling Procedure. The population was strati- projects, where there is not a detailed ex ante fied twice according to the dedicated/nondedi- allocation of project costs. cated definition and according to the two time cohorts, FY94­FY98 and FY99­FY04. In this way Following this procedure a total of 161 projects it was possible to obtain four groups of projects were identified. Their main characteristics are (table A2). For each group, the number of detailed in appendix B. projects obtained was sampled following a simple random-sample procedure. Therefore, Detailed Review of Project Documents the overall sampling procedure can be defined The analysis focused on 80 projects, randomly as a stratified, proportionate, random-sample drawn from the larger population of 161 agricul- procedure. tural water projects. To facilitate the analysis, the projects were divided into two groups, The final list of projects is quite representative dedicated and nondedicated. The former group of different project types, regions, and identifies all those projects where the amount countries. In addition to the 80 selected committed on irrigation and drainage is greater projects, 18 dedicated projects were reviewed, than 50 percent of the total IDA/IBRD amount to include the entire population of the 60 committed for the whole project. Nondedicated dedicated projects. In order not to introduce a projects have less than half of Bank financing selection bias, statistical data from these 18 devoted to agricultural water management projects were not used. activities. Each sampled project was rated against 34 criteria developed from issues raised Each of the 161 projects was assigned a contigu- by Bank poverty, rural and water strategy ous ascending number, and then tables of statements, and key documents in the literature random numbers were used, blindly selecting a related to poverty, institutions, and policy as starting point in the table. Whenever the regards water (table A1). ICRs and PPARs selected project happened to be a supplement, covering the closed projects in the sample were the original project was scored instead, but only also reviewed where available, which provided a if it belonged to the same time cohort. When clearer picture of the design of the projects in this did not happen, another project was relationship to the outcomes and the effective- selected following the random-selection ness of M&E. procedure described above. Records of the procedure have been kept. Sampling stopped Randomly sampled project documents were when 80 projects had been selected. The scrutinized and scored. The rating scale used projects selected and IEG's ratings of them are was from 1 to 4. A blank was defined as no signif- given in appendix C. icant evidence of the feature, 1 was defined as occurrence of the feature to a small extent, 2 was Assessment of Monitoring and Evaluation defined as occurrence of the feature to a The 80 random projects were independently moderate extent, 3 was defined as occurrence of reviewed to determine how well M&E had been the feature to a substantial extent, and 4 was designed. A follow-up analysis using ICR output defined as occurrence of the feature to a very and outcome data from these completed high level. There was also a Not Applicable (NA) projects was used to determine how well M&E and an Addressed Outside the project (AO) had been implemented. This was based on the rating for some criteria, mainly policy. As is application of 17 evaluation questions (table A3) evident, the mid-point on the range where there and their overall categorization, to determine was any occurrence at all lay between 2 and 3. In the overall quality of M&E from an evaluation a few cases, there was simply a Yes/No rating. In and impact assessment perspective (table A4). a few cases, percentages were requested, for Results are presented in appendix D. example, percentage cost recovery. 6 8 A P P E N D I X A : S T U D Y M E T H O D O L O G Y Table A1. The 34 Criteria Used for Describing/Rating the Project Design Poverty mentioned as explicit objective? (Y/N) Extent to which project design incorporates a poverty focus (even if no poverty objective). (1 to 4 or NA) Extent of direct targeting of benefits towards poorer, e.g. selection of location, communities, households. (1 to 4 or NA) Given both the objectives and design, does this warrant classification as a poverty-focused intervention? (Y/N) Quality of analysis distributional aspects (who, why poor, what to do, power relationships) incl. location (e.g., head/tail, elite). (1 to 4, NA) Was a social assessment carried out? (Y/N or NA) Is project employment impact or wages analyzed or substantively discussed? (Y/N or NA) Extent of analysis of, or substantive discussion of, water rights of beneficiaries. (1 to 4 or NA/AO) Extent of policy content including legislation, pricing, rights, but excluding institutional. (1 to 4 or NA/AO) Extent of institutional reform, e.g., public/private shift, new organization (excludes pricing covered under policy). (1 to 4 or NA/AO) To what extent does project aim to reform or significantly strengthen public institutions at central level? (1 to 4 or NA/AO) To what extent does project aim to reform or significantly strengthen public inst. below central, e.g., region/district? (1 to 4 or NA/AO) Extent of gender focus in project design. (1 to 4 or NA) Total number of beneficiaries in household. (give no. or NG, for not given) Percent of poor in beneficiary total if known. (% or NG) Hectares of land benefiting. (ha or NA) Predominant hectares benefiting large farms (L) or small (S) as defined by PAD. (or NA) Quality of M&E indicators of poverty performance. (1 to 4 or NA) Quality of M&E design. (1 to 4 or NA) If not classified as poverty-focused, list in logical sequence any clearly plausible project poverty logic. If not classified as poverty-focused, what might have been the poverty impact of the "without project" scenario? (qualitative) If not classified as poverty-focused, to what extent do the previous two criteria suggest that the project is indirectly pro-poor? (1 to 4 or NA) To what extent does project aim to strengthen community orgs. or participation, e.g., through WUAs? (1 to 4 or NA/AO) To what extent does PAD propose autonomy for WUAs (in fee collection, retention, expenditure, and water management)? (1 to 4 or NA) To what extent does PAD indicate support to WUAs on water management? (1 to 4 or NA) To what extent is a broader sector strategy of which this project is a part clearly outlined in the PAD? (1 to 4 or NA) Estimate approximate planned average percent cost recovery for capital investment. (give % or NA) Estimate approximate preproject average percent cost recovery for O&M. (give % or NA) Estimate approximate planned average percent cost recovery for O&M. (give % or NA) To what extent does project address water efficiency, e.g., through water charges, regulations, or technical design? (1 to 4 or NA/AO) To what extent does PAD indicate how shortfalls in cost recovery are to be handled, e.g., public subsidy, cross-subsidy? (1 to 4 or NA) To what extent does PAD address environmental water issues such as quality, groundwater depletion, etc. (1 to 4 or NA) Which of following are supported or linked to significant extent in project design? Marketing (M), Extension (E), Credit (C) (or NA/AO) To what extent does the PAD propose collaboration with other water agencies, e.g., environment, fisheries, agriculture? (0 to 4 or NA) 6 9 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 Table A2. Distribution of Randomly Sampled Projects, primary analysis was textural, aided by the by Project Type commercial software package Atlas Ti.2 The analysis interrogated the whole text of each CAS Projects to determine the presence of 21 key phrases sampled Dedicated Nondedicated Total (table A5). When a phrase was encountered, the FY94­FY 98 20 17 37 paragraph in which that phrase occurred was FY99­FY 04 22 21 43 extracted and stored. Following scrutiny to Total 42 38 80 determine the relevance of each hit, the relevant hits were added to the summary database. In 2. Analysis of Country Assistance this way, 544 AWM references in 124 CASs, Strategies containing phrases relevant to this study's CASs covered 54 countries, and during the evaluation questions could be related to time, period FY94­FY04 there were 130, each of 51 countries, and region. Detailed results are which was entered into a special database. The presented in appendix E. Table A3. Evaluation of M&E Design and Implementation Evaluation Questions Evaluative score Objectives: level of clarity (1= negligible, 4 = high) 1 to 4 Logical framework (0 = absent, 1 = negligible, 4 = high) 0 to 4 M&E system in place ex ante 0 to 4 M&E system specified by project 0 to 4 Quality of M&E plan 0 to 4 Desired outputs clearly defined 0 to 4 Desired outcomes clearly defined 0 to 4 Indicators well structured 0 to 4 Focus on outcome indicators 0 to 4 M responsibility assigned 0 to 4 M coordination assigned 0 to 4 Support for M&E capacity building 0 to 4 Feedback loops for management? 0 to 4 Evaluative baseline specified 0 to 4 M&E as condition of lending? 0 or 1 ICR (0= active project, 1 = completed project) 0 or 1 Baseline status (0 = none, 3 = start, 2= middle, 1=end) 0 to 3 Ex-post survey(s) 0 to 4 Transparency of evaluation (0 = none, 1 = negligible, 4 = high 0 to 4 Table A4. Typology for Classification of M&E Findings from Project Design Output indicators, but no outcome or impact indicators Output indicators as well as outcome or impact indicators, but no baseline Output and/or impact indicators and baseline Output and/or impact indicators and control groups but no baseline Outcome and/or impact indicators and baseline and control groups 7 0 A P P E N D I X A : S T U D Y M E T H O D O L O G Y Table A5. Topics Researched in Analysis of planned and actual outputs 130 CASs and outcomes reported by each PAD and ICR in the portfolio. In addition to the qualita- Irrigation tive evaluation of the objectives and Institutional Development/Reform (Irrigation & Water) components of each PAD and ICR, the study also Participation in Irrigation analyzed the quantitative output data (table A6). Rural Development Results are summarized in appendix F. Irrigation & Rural Development Rural Poverty Alleviation Categorization of achievements, issues, Role of Rural Development in Poverty Alleviation problems, and lessons learned. Sixty-three Institutional Development in Rural Development ICRS and eight PPARs of projects approved from Participation in Rural Development FY94 were used to create a database and the Rural Development & Agricultural Growth lessons learned. After categorization, 408 Agricultural Development lessons were classified into 11 types (table A7) Agricultural Growth and Irrigation for further analysis and review (detailed in Agricultural Development & Economic Growth appendix F). Institutional Reform/Development in Agriculture Participation in Agriculture 5. Interviews of Bank Managers and Staff Water Resource Management A total of 17 managers were interviewed in the Irrigation & Water period April­May 2005. All of the interviews Cost Recovery were conducted under a guarantee of confiden- Water Users tiality. The list of questions, presented below, Water Tariffs was designed as a guide, rather than as a formal Technical Assistance questionnaire, to allow the discussion to follow leads as they occurred. Despite the variety of 3. Meta-evaluation of IEG's Country respondents contacted (belonging to different Assistance Evaluations departments, gender, etc.), the sample was not The methodology for CAEs was similar to that selected randomly, and for this reason no statis- used for the evaluation of CASs. tical inferences were drawn from this exercise. The main goal of the interview was to collect 4. Project Completion and Performance additional and competent views to inform the Assessment Reports analysis conducted on the operational Two types of analysis were undertaken: (i) a documents, and to have additional internal qualitative review of achievements, issues, insights on reasons, causes, and impressions problems, and a categorization of lessons with respect to the role and the trends of the learned; and (ii) a more detailed analysis of the irrigation and drainage subsector. The main outputs and outcomes reported by each ICR. questions were as follows: Table A6. Quantitative Comparison of Predicated and Actual Outputs from Projects Irrigation--area planned and actually achieved (ha). Drainage--area planned and actually achieved (ha). Farmers/Farm families--number of household units or number of individual farmers targeted and reported as being reached at the end of the project. People--number of individuals targeted and reached by the project. Engineering costs--planned and actual costs. Institutional development and capacity building costs--planned and actual costs. 7 1 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 Table A7. Typology Used to Classify 3. What has been and what could be the role of Frequency of Lessons Drawn from I&D in rural poverty alleviation? ICRs and PPARs 4. What are your views on Bank staffing, training, skills, and management in the subsector? Project Design Implementation & Procedural 6. Literature Review Community-Driven Development A systematic review of the global state of irriga- Cost Recovery tion and drainage and its impact on economic Knowledge/Information/Skills Mobilization growth and poverty was conducted by the Technology International Food Policy Research Institute Economics (IFPRI), a policy research organization known Institutional worldwide for its studies on the role of irrigation Bank Processes in agricultural development. The main objective Targeting of this general review was to identify issues and Disaster-Related challenges in the agricultural water sector in the 1. What are the main reasons Bank-wide for the period 1994­2004. The results of this review decline/change in irrigation and drainage (I&D) form a separate supporting paper. lending (reasons prioritized)? 2. What does the I&D subsector need to do to be more relevant to evolving Bank priorities? 7 2 APPENDIX B: BACKGROUND INFORMATION AND TRENDS, 1994­2004 Lending Analysis A broader search among all the Bank projects From FY94 to FY04, the World Bank approved approved in the period under review added at 161 agricultural water projects that designated least 210 more projects that were likely to "irrigation and drainage" (I&D) as a sector of include some sort of intervention in agricultural intervention. The details of these projects are water, 153 of which presented characteristics of shown in table B6 at the end of this appendix. CDD projects. This high number suggests the Because the system provides up to five sectors importance given to the participatory of intervention for each project, some of the approaches of many irrigation projects, and it is selected projects show the I&D sector as the not surprising to see almost half of the 161 main sector of intervention; for others it is projects present at least some characteristics of shown as the third, or even fifth, sector. For each CDD projects. project the system provides the amount specifi- cally spent on I&D. Unfortunately, the system Out of the 161 selected projects, 60 "dedicated" omits projects where the amount committed to projects were identified (table B1) with more I&D is too small or is considered too negligible than 50 percent of the total IDA/IBRD amount to be included as a sectoral component. This is committed for the whole project designated for especially the case for CDD projects, for I&D. These projects were subjected to particu- example, in many Social Funds projects, where larly close scrutiny in the study. Note that the 50 the allocation of project costs is not detailed ex percent cutoff was based on the planned alloca- ante. tion, not actual allocation. Out of the 161 projects, 119 show Rural Table B1. Number of Dedicated and Development as the sector board, which means Nondedicated Projects per Year that 42 projects are not classified as Rural Development projects. These were classified as FY # dedicated # nondedicated shown in table B2. 1994 5 6 1995 8 11 All but one of the dedicated projects have Rural 1996 3 6 Development as the sector board (the one 1997 5 11 exception lists the Environmental Department). 1998 7 11 Most (155 out of the 161 projects) have "invest- 1999 8 18 ment" as the lending instrument type, while six 2000 5 6 projects are identified as "adjustment." None of 2001 4 9 the six is listed among the dedicated projects. 2002 4 6 2003 2 9 The 161 projects are in a total of 56 countries. 2004 9 8 The total amount committed by the Bank in Total 60 101 these 161 projects is US$13.2 billion (in 2002 Source: World Bank's Business Warehouse database. US$), roughly the 5.6 percent of the total Bank 7 3 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 Table B2. Sector Boards of 161 Figure B1. Distribution of Projects by I&D Projects Region, per Time Cohort Sector board No. ECA Rural sector 119 Urban development 11 EAP Social protection 10 SAR Environment 4 FY94­98 Water supply 4 MNA Social development 3 FY99­04 AFR Transport 3 Financial sector 2 LAC Private sector development 2 0 5 10 15 20 25 30 35 Public sector governance 2 Percent Health nutrition and population 1 Source: World Bank's Business Warehouse database. Source: World Bank's Business Warehouse database. temporal cohorts (FY94­FY98 and FY99­FY04) lending committed for the period. The total shows the changes over time. amount committed specifically on I&D (with all the caveats about the data) is US$5.56 billion (in The most striking figure seems to be the 2002 US$), which represents about 2.36 percent increase in the number of operations in the ECA of the total Bank lending committed for the region.1 However, looking at the cumulative period. lending amounts in projects containing I&D per region and at the amount specifically committed Regional Lending Patterns and Number of on I&D from FY94 to FY04 (in 2002 dollars) the Operations picture changes (see figure B2). On the one Table B3 and figure B1 show the geographical hand, the Asia regions dominated irrigation distribution of the operations. The ECA region lending (thus confirming the figures of the leads with respect to the number of operations. 1980s; see IEG 1995).2 On the other hand, the Dividing the number of projects into two ECA region had many small projects (on average, US$16 million per project committed specifically to I&D; figure B3). Table B3. Geographical Distribution of I&D Projects per Year Figure B2. Commitment Amount Fiscal in Irrigation Projects per Region, Year ECA EAP SAR MNA AFR LAC TOT FY94­FY04 1994 0 3 3 2 0 3 11 5,000 4,679 1995 3 3 2 4 3 4 19 4,500 1996 2 2 2 2 1 0 9 million) Total commitment amount for 4,000 3,691 projects containing I&D 1997 1 3 3 4 3 2 16 US$ 3,500 I&D commitment amount within 1998 3 5 4 2 1 3 18 2002 3,000 these projects 1999 10 8 2 1 3 2 26 (in 2,500 1,908 2000 5 0 2 2 2 0 11 2,000 1,579 amount 1,517 2001 4 5 1 1 1 1 13 1,500 1,286 1,175 2002 2 0 5 1 1 1 10 1,000 832 617 703 551 2003 4 2 0 4 1 0 11 500 206 2004 4 2 6 1 2 2 17 Committment 0 SAR EAP ECA LAC MNA AFR TOTAL 38 33 30 24 18 18 161 Source: World Bank's Business Warehouse database. Source: World Bank's Business Warehouse database. 7 4 A P P E N D I X B : B A C K G R O U N D I N F O R M AT I O N A N D T R E N D S , 1 9 9 4 ­ 2 0 0 4 India alone has accounted for almost 23 percent Figure B3. Average I&D Commitment of the total amount committed to I&D, China Amount per Project and Number of for 14 percent, Pakistan for 7 percent, and Projects per Region Indonesia for 6 percent. These countries are consistently the largest borrowers; they account 70 40 Value for the highest number of operations in the 38 33 35 million) 60 Number period and for the greatest consistency (at least 30 US$ 30 six years with at least one operation in the whole 50 24 period). Mexico, with only two operations, is the 2002 25 (in 40 projects third largest borrower, accounting for almost 8 18 of 20 percent of the total amount committed to I&D. 30 amount 18 15 Number 20 Concerning the number of operations per 10 country, China, with 13 projects, accounts for 39 10 committment 5 percent of the total number of operations in the I&D 0 0 EAP region (and 8 percent of the overall total); ECA EAP SAR MNA AFR LAC India, with 12 projects, for 40 percent of the total Source: World Bank's Business Warehouse database. number of operations in the SAR region (7 percent of the total); Pakistan, with 9 projects, for 30 percent of the total number of operations 1999­2003, but rose sharply in 2004. However, in the SAR region (almost 6 percent of the total); 2004 had two big projects: the Mexico Irrigation Indonesia, with 8 projects, for 23 percent of the total number of operations in the EAP region (5 percent of the total). Figure B4. Specific I&D Amount Committed, per Region Figure B4 shows that EAP has eclipsed SAR in its share of total I&D lending, with SAR experienc- EAP ing the greatest drop in I&D commitment SAR amount. ECA exhibits the greatest jump, but the ECA share is still quite low. MNA FY99­04 FY94­98 Figure B5 shows the amount committed by the LAC Bank specifically to I&D, as a ratio of the total AFR Bank committed amount.3 The percentage fell 0 5 10 15 20 25 30 35 40 45 from about 3 percent in the period 1990­94, to Regional amount as % per time cohort an average of less than 1.7 percent in the period Source: World Bank's Business Warehouse database and calculations. Table B4. Number of Projects and I&D Commitment Amount per Region Projects ECA EAP SAR MNA AFR LAC TOT Number of I&D Projects 38 33 30 24 18 18 161 Total I&D Projects Commitment Amount (in 2002 US$ million) 1,517 3,691 4,679 1,175 832 1,286 13,181 I&D Commitment Amount (in 2002 US$ million) 617 1,579 1,908 551 206 703 5,563 Average I&D Commitment Amount per Project (in 2002 US$ million) 16 48 64 23 11 39 35 Source: World Bank's Business Warehouse database. 7 5 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 Lending and Time Periods Figure B5. I&D Commitment Amount as Percentage of Total Bank Committed Figure B6 shows the average amount committed Amount specifically on I&D per project per year (in 2002 US$). The average amount committed on I&D 4.5 per project fell from above $50 million in the 4.0 3.8 period 1990­94 to an average of less than $30 3.5 million in the last four years. The reason that the 3.2 3.1 3.1 3.0 2.9 average project size got smaller is because there 2.5 were fewer large dedicated irrigation projects 2.5 Percent 2.0 and the size of the irrigation and drainage 1.7 1.7 1.5 components in nondedicated projects shrank. 1.1 1.2 1.0 Figure B7 shows, in percentage terms, the trend 0.5 in large projects (defined as projects where an 0 1990 1992 1994 1996 1998 2000 2002 2004 amount larger than $80 million was committed Source: World Bank's Business Warehouse database and calculations. specifically to I&D) versus the trend in small projects (defined as projects where an amount smaller than $30 million was committed to I&D). and Drainage Modernization Project and the For the period 1990­2004 there is a decreasing Vietnam Water Resources Assistance Project. trend for large projects and an increasing trend Totaling over $370 million overall (in 2002 US$), for small projects. (from 40 percent in 1990 to these two projects made up half of the lending an average above 60 percent after 1995). to the subsector. Again, the figures are to be treated with caution (as stressed above, IEG's Figure B8 validates the previous analysis by detailed sampling of appraisal documents found showing, in percentage terms, the trend for the that in several dedicated projects the cost of number of dedicated projects out of the total agricultural water components was almost 20 number of I&D projects approved per year; and percent more than those identified when using the trend for the I&D amounts committed to only the sector code). However, this additional dedicated projects as a percentage of the total information does not change the conclusions amount committed by the Bank to I&D. about the overall trend. Figure B7. Percentage of Projects with Figure B6. Average I&D Commitment Large versus Small I&D Committed Amount per Project Amounts 80 90 70 67 80 59 70 60 60 50 48 49 million 43 50 40 Small I&D projects US$ 33 Large I&D projects 29 Percent 40 30 2002 30 In 20 20 20 10 15 10 0 0 1990 1992 1994 1996 1998 2000 2002 2004 1990 1992 1994 1996 1998 2000 2002 2004 Source: World Bank's Business Warehouse database and calculations. Source: World Bank's Business Warehouse database and calculations. 7 6 A P P E N D I X B : B A C K G R O U N D I N F O R M AT I O N A N D T R E N D S , 1 9 9 4 ­ 2 0 0 4 Figure B8. Lending and Number of Figure B9. Agriculture Sector and I&D Dedicated Projects Decreased Lending as Percentage of Total Bank Lending 100 I&D lending in 16 90 dedicated Agriculture sector 80 projects as % of 14 all I&D lending (%) I&D 70 12 60 lending 10 50 Percent Dedicated I&D Bank 8 40 projects as % 30 of total I&D total 6 projects of 20 4 10 Share 2 0 1990 1992 1994 1996 1998 2000 2002 2004 0 1990 1992 1994 1996 1998 2000 2002 2004 Source: World Bank's Business Warehouse database and calculations. Source: World Bank's Business Warehouse database and calculations. The same is true for the agriculture sector as a whole (figure B9). In fact, the decrease in the Figure B10. I&D Commitment Amount agriculture sector as a whole is greater than the per Group of Borrowers one experienced by the I&D subsector alone. 70 As shown in table B5 and figure B10, the volume 60 FY94­98 of lending to the four most consistent and large FY99­04 I&D borrowers of the last 15 years (India, China, 50 amount cohort) Pakistan, and Indonesia) has declined. 40 time per 30 The typology of borrowers was based on the committment 20 number of operations per borrowers. However, I&D (percent a high correlation among the number of 10 operations, consistency of operations per 0 borrower (number of years with at least one Consistent Regular Periodic Occasional One-off operation, i.e., frequency of borrowing across Source: World Bank's Business Warehouse database and calculations. Table B5. Lending Pattern and Number of Projects per Borrower Type Borrower FY94­FY98 FY99­FY04 typology I&D comm. amount I&D comm. amount (and number of projects) US$ (millions) Percent Projects US$ (millions) Percent Projects Consistent (4) 1,884 62 24 956 38 18 Regular (9) 336 11 15 646 26 29 Periodic (10) 255 8 13 384 15 17 Occasional (12) 444 15 13 370 15 11 One-off (21) 131 4 8 157 6 13 Total 3,050.5 100 73 2,512 100 88 Source: World Bank's Business Warehouse database. 7 7 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 borrowers with respect to number of operations Figure B11. Consistently Large (Indonesia, Pakistan, India, and China) are also Borrowers: Commitment Amount in among the five more consistent borrower (along Other Agriculture versus I&D with Yemen) and among the five receiving the Commitment Amount highest amount in the period (with Mexico). 2,500 Mexico, one of the largest I&D borrowers, Other agriculture million) I&D accounts only for a small number of operations 2,000 US$ containing I&D, and for this reason does not appear in the group of the consistently large 2002 1,500 (in I&D borrowers. 1,000 amount The decrease in "other agriculture" (all agricul- 500 ture but I&D) for the consistently large borrower (figure B11) suggests a steady drop in Commitment 0 agriculture-supportive interventions. The FY90­92 FY93­95 FY96­98 FY99­01 FY02­04 decrease in I&D commitment amount was, Source: World Bank's Business Warehouse database and calculations. therefore, not matched by an increase in the the years), and amount committed to I&D per amount committed elsewhere in the agriculture borrower was found. In fact, the first four sector after 1999. 7 8 A P P E N D I X B : B A C K G R O U N D I N F O R M AT I O N A N D T R E N D S , 1 9 9 4 ­ 2 0 0 4 page) M M M NEG SU M M SU M M SU SU SU M SU next on 1 Institutional Development continues Evaluation UL UL UL UL L UL L L UL UL L L L L HL able(T IEG Sustainability MU MS MS U S U MS S U MU S S MS MS S Outcome 3 9 6 4 36 36 18 24 48 24 12 18 15 18 34 months Extension -03 -01 -97 -04 Date Closing Dec-04 Dec-04 Sep-00 Mar Dec-05 Dec-01 Dec-01 Jun-02 Dec-06 Dec-05 Dec-02 Sep-03 Jun-05 Jun-03 Jun-07 Mar Jun-06 Apr Dec-00 Apr Dec-05 4 7 7 Bank millions 41 21 30 60 27 55 58 50 80 80 57 60 120 205 150 210 166 100 102 120 Lending US$ (*) 5 8 7 otalT Cost millions 50 26 60 41 63 85 70 80 57 242 202 382 259 552 304 136 148 182 211 107 FY94­FY04 asterisk US$ an by FY Information 1998 1996 1995 1997 1999 1995 1995 1994 1996 1995 1994 1996 1994 1997 1995 1999 1996 1995 1995 1995 1994 1994 1997 Entry Portfolio, indicated Project are Invest M . ater R ater ADP W W Dev Inv W Services Pep I Sewerage Irrigation Sector Corridor & Project Sec analysis II Promotion Plain Rehabilitation Basin Imp Delta Improvement unisT Management Rehabilitation Sector Hexi Private Irr enggaraT Resources Safety Agriculture Safety For ater detailed Name Agricultural Social Irrigation Irrigation Agric Pilot Agricultural Songliao Gansu angtzeY ater Java Nusa Dam W Irrigation Mekong Social Irrigation ASAL AT Irrigation Agricultural Greater W for Faso selected Country Burkina Ethiopia Madagascar Mali Mozambique Niger Zambia China China China Indonesia Indonesia Indonesia Philippines ietnamV ietnamV Algeria Egypt Jordan Jordan Lebanon unisiaT unisiaT Agricultural randomly B6. * ID ableT Projects Bank P000296 P000771 P001522 P001738 P001799 P001994 P003218 P003593 P003594 P003596* P003954* P004008* P004010 P004613* P004834* P004845* P004978* P005173* P005310* P005321* P005344* P005721* P005731 7 9 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 page) SU M M SU M SU SU H SU SU M M N M M SU M next on 1 Institutional Development continues Evaluation L NE UL L UL L L L L L UL L NE UL L L L able(T IEG Sustainability S MS U S MU S S HS S S MS MS MS U MS S S Outcome 6 0 6 12 36 24 12 18 18 12 21 30 12 months Extension -10 -04 Date Closing Jun-04 Dec-05 Dec-06 Jun-98 Dec-04 Dec-08 Dec-02 Jun-01 May-01 May-06 Mar Jun-02 Dec-04 Jun-04 Jun-02 Jun-10 Dec-01 Dec-95 Jun-00 Mar Sep-04 Jun-02 Dec-04 Bank millions 27 25 75 20 28 85 41 10 43 42 35 15 80 20 66 60 15 27 258 250 276 283 285 Lending US$ otalT Cost millions 51 46 28 43 74 45 57 47 39 31 59 74 75 18 39 100 172 118 483 250 421 491 785 US$ FY Information 1997 1998 1998 1994 1995 1997 1994 1995 1995 2000 2003 1996 1996 1998 1995 2003 1994 1994 1994 1995 1995 1996 1998 Entry Project Project Mgmt Mgmt etlands Irrig W II Development atersh Res Rehab Mgmt & Program W & Adjust NRM Sup Infrastructure WRCP Comm & Subsector Natl at Resource Sys Irrig WRCP W Privatization Farmer City Drainage Irrigation Subsector Sector Nadu Land Rehabilitation Rehabilitation Dist & Of Drainage Name Natural AGRI Small Belize Irrig Irrigation Irrigation Irrig Irrig Irrig/Drainage Irrig Agriculture Irrig Priv Cotton Drainage, Haryana Public Balochistan Madras amilT Balochistan Natl Country unisiaT emenY Argentina Belize Dominica Peru Uruguay Albania Armenia Azerbaijan Azerbaijan Estonia Kazakhstan urkeyT Uzbekistan Uzbekistan India Pakistan Pakistan India India Pakistan Pakistan (continued) B6 ID ableT Bank P005736 P005902* P006041 P006104 P007020* P008037* P008173 P008270 P008277* P008284* P008286* P008403* P008510 P009072* P009122 P009127* P009964* P010447 P010453 P010461 P010476 P010482* P010500* 8 0 A P P E N D I X B : B A C K G R O U N D I N F O R M AT I O N A N D T R E N D S , 1 9 9 4 ­ 2 0 0 4 page) M M M SU M SU SU SU M SU SU SU SU M next on 1 Institutional Development continues Evaluation NE UL UL L L L L L L L L L HL L able(T IEG Sustainability MS MS U HS S S S S S HS S HS HS S Outcome 0 6 0 6 6 7 9 24 18 18 12 18 months Extension -06 -08 -04 Date Closing Dec-01 Sep-04 Jun-04 Dec-06 Sep-98 Jul-04 Jun-06 Jun-01 Dec-03 Mar Jun-04 Dec-99 Jun-01 Jun-01 Jun-06 Jul-00 Dec-03 Mar Aug-06 Jun-07 Mar Bank millions 56 80 57 73 50 13 26 51 70 36 63 15 51 291 325 103 105 140 140 120 115 Lending US$ otalT Cost millions 74 84 33 26 84 60 87 29 51 105 346 103 477 258 175 106 117 140 180 253 139 US$ FY Information 1997 1996 1998 1998 1995 1997 2000 1995 1997 1998 1997 1995 1995 1995 1998 1997 1997 2002 1999 2000 1997 Entry Project Ii Smll Comm & III Investment 3 Groundwater Devt Emerg.Recons Restructuri (Bahia). Ceara Reform Mgmt WRCP .- .-Sergipe Station WRCP Infrastructu Sector Infrastructure Pov Pov Pov Investment Infrastructure NRM Sector Sector Rehab Privatization Irrigation Basin Irr Name Pvt Orissa Irrig Mahaweli illageV ater AP W Rural Agrarian Irrig Riv Hldr Dz-Mascara Rural Rural Rural illageV Farm Rajasthan Pumping Rehabilitation Rural Sierra Lanka Country Pakistan India Nepal Sri Indonesia India unisiaT Brazil Philippines Macedonia anzaniaT Algeria Brazil Brazil Bolivia Indonesia Azerbaijan India Egypt Mali Peru (continued) B6 ID ableT Bank P010501* P010529* P010530* P034212* P034891* P035158* P035707* P035717 P037079* P038399* P038570 P038695 P038884* P038885* P040085 P040521 P040544 P040610 P041410* P041723* P042442* 8 1 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 page) H NEG SU SU SU SU next on 1 Institutional Development continues Evaluation UL HL L L L L able(T IEG Sustainability U HS S U HS MS Outcome 0 0 6 18 17 36 months Extension -11 -06 -07 -05 Date Closing Dec-01 Mar Jun-05 Jun-07 Dec-05 Jul-06 Feb-07 Dec-04 Jan-06 Dec-01 Jun-05 Jul-03 Mar Dec-06 Jun-05 Nov-05 Mar Dec-04 Dec-02 Mar Sep-07 Oct-07 Jun-06 Bank millions 10 80 38 50 35 36 65 30 35 20 20 90 42 25 23 150 200 150 543 120 300 194 149 Lending US$ otalT Cost millions 11 46 47 43 86 35 46 24 29 90 69 25 30 103 278 273 364 150 830 240 849 287 174 US$ FY Information 1997 2004 2000 2000 1998 2002 2001 1998 1998 1997 1997 1997 1998 1999 1998 1999 2000 1999 1998 1999 1999 2002 2001 Entry Project Pilot Ii Ar Rehab Support Fund Ii Ag.Dev Supply II Poor. Flood Ii Cyclone Poverty Rehabilitation Agric Drainage Rehabilitation Control Irrigation Lands ater Support Restructuring Dev Alleviation Basin Infrastructure alleyV Fund W Privatization 2 Rural Ent Darya Emerg. Econ Sodic WRCP Name aizT Irrigation Irrigated National Irrigation Rur Syr arimT Forest. Emergency Urban AP AP Anning Iail-2 Farm On-Farm Poverty ASIL Social UP UP Piaui Republic Republic Country emenY Romania Mauritania Egypt Kyrgyz Uzbekistan Kazakhstan China China emenY Madagascar India India China China ajikistanT Kyrgyz Pakistan unisiaT Cambodia India India Brazil (continued) B6 ID ableT Bank P043367 P043881* P044711 P045499 P046042* P046043* P046045* P046563 P046952 P048522* P048697 P049301 P049385* P049665* P049700 P049718* P049723* P049791 P050418 P050601 P050646* P050647* P050881 8 2 A P P E N D I X B : B A C K G R O U N D I N F O R M AT I O N A N D T R E N D S , 1 9 9 4 ­ 2 0 0 4 page) SU SU SU M M SU M SU SU M M M M next on 1 Institutional Development continues Evaluation L L L NE UL HL L L L L NE UL L able(T IEG Sustainability MS MS S S S HS S S S MS U S S Outcome 0 3 0 4 0 0 0 12 12 12 18 26 15 months Extension -03 -07 -08 -01 -02 -03 -06 Date Closing Dec-04 Jun-01 Mar Dec-02 Jun-06 Jun-07 Mar Apr Dec-09 Mar Jun-05 Dec-00 Jun-06 Dec-01 Mar Jul-00 Jul-02 Dec-05 Jun-05 Mar Jun-04 Sep-03 Mar 5 9 9 5 5 Bank millions 65 10 89 25 27 12 25 28 74 40 76 50 20 27 20 100 150 369 Lending US$ 5 6 otalT Cost millions 65 17 15 31 33 14 28 41 40 10 76 50 29 32 14 24 200 106 150 186 685 US$ FY Information 1999 1999 1999 1998 1999 2001 2002 2001 2003 1998 1999 1998 2001 1999 1999 1999 2001 2000 2000 2000 1999 1999 2000 Entry Project Support AT Agric Rehab orks Fund Ii W Irrigation Com Project Irrigated Services illageV Recovery Rehab Rehab Irrigation Social Emergency Emergency Emergency 3 Development Scale Devt oñ Plateau Earthquake Conservation oñ oñ I 2 Flood Infra 2 Ni ater Ni Ni Agriculture Name Private SAC Community Szopad Guanzhong Rural Irrig Irr/Drain Small El Loess Hebei W El El ASAL ASAL SIF North-East Agricultural Northeast Emgy Rural Faso Lanka Herzegovina Country Burkina Armenia Albania Philippines China Madagascar Armenia Georgia Bosnia- Bolivia China China China Kenya Guyana Bulgaria Bulgaria Armenia Sri Bangladesh Cambodia urkeyT ajikistanT (continued) B6 ID ableT Bank P050886 P051171 P051309 P051386 P051888 P051922* P055022* P055068* P055434 P055974* P056216 P056491 P056516 P056595 P057271 P057925 P057926 P057952 P058070* P058468 P058841* P058877 P058898* 8 3 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 page) NEG SU SU M SU next on 1 Institutional Development continues Evaluation UL HL HL L L able(T IEG Sustainability U S HS S S Outcome 6 0 30 17 35 months Extension -04 Date Closing Dec-01 Sep-09 Dec-09 Mar Jun-07 Dec-07 Jun-02 Jun-03 Dec-09 Nov-04 Sep-07 Jun-09 Jun-10 Dec-11 Jun-07 Dec-09 Jan-06 Jun-12 Jan-09 Jun-06 5 6 Bank millions 40 70 10 21 80 27 24 17 27 20 99 21 103 111 100 300 100 158 Lending US$ 6 7 otalT Cost millions 68 14 26 30 30 18 41 38 32 116 123 133 125 125 300 154 176 125 US$ FY 1999 2003 2003 1999 2001 2001 1999 1999 2004 1999 1999 2003 2004 2004 2001 1999 2000 2003 2002 2004 2002 2001 2001 Information Entry Infra. Project Irrigation Mgmnt & Rural Recovery Agric. Assistance Asst ater Mgmt Supplmt Emergency W Support Flood Asst Cleanup Based Development Improvement Georges atershed Safety ankT Flood II orks Agriculture W Supp. Resources Basin Integrated Management Resources W Emg Flood River Emergency OFWM Dam Priv ater Name Emg Nura W Flood Irrigation Community angtzeY TSALA Modern. ater Hurricane Fadama W Irrig Sana'a Jiangxi W Agricultural Comm Suplmt Rainfed Disaster Anatolia Karnataka NWFP Farm & Republic incentV Country ajikistanT Kazakhstan Indonesia Kyrgyz emenY ietnamV China Dominica Nigeria Indonesia Armenia emenY China ietnamV Laos Albania ajikistanT Grenadines Morocco St. urkeyT India Pakistan ajikistanT (continued) B6 ID ableT Bank P059055 P059803* P059931 P062682* P062714* P062748* P063123 P063201 P063622 P064118 P064879* P064981 P065463 P065898 P065973 P066335* P068786 P069124 P069923 P070950 P071033* P071092* P072760* 8 4 A P P E N D I X B : B A C K G R O U N D I N F O R M AT I O N A N D T R E N D S , 1 9 9 4 ­ 2 0 0 4 SU 1 Institutional Development Evaluation NE IEG Sustainability S Outcome 0 months Extension -07 -11 -09 Date Closing Dec-07 Jun-05 Jun-08 Oct-07 Feb-05 Sep-08 Oct-09 Jun-09 Sep-09 Dec-04 Sep-07 Dec-08 Dec-08 Feb-09 Jun-09 Jul-08 Mar Mar Mar 5 7 Bank millions 39 35 12 20 40 95 34 42 40 61 15 15 65 443 250 238 95 Lending US$ 8 otalT 48 40 15 40 25 53 45 42 40 85 17 40 81 Cost millions 504 143 377 368 95 US$ FY 2002 2001 2001 2002 2002 2003 2004 2004 2003 2004 2003 2002 2004 2004 2003 2004 2004 2004 2004 2004 2004 Information Entry II SIL Mgmt Project FUND 2 Agency Reconstruct SIL 2 Solidarity Fund Development Resource Pos Soil Project Alleviation Promotion SOCIAL & & Employment Safety Peoples Community Rehab Mgmt National Dam Credit Earthquake Matruh Rural Alleviation Poverty Emergency Development SupplementalI Irrig OFWM Kecamatan Res II Irrigation Services Conservation (Prodepine2) Empowerment Project ater Name Priv SUPPL Flood Social Gujarat Second Agr Groundwater Second Indigenous ESRDF Emergency Emer Sindh Third Poverty W Second Emergency NEIAP Irrigation Lanka Country Niger Cambodia Cambodia Morocco India Egypt Chad emenY Algeria Ecuador Ethiopia Afghanistan Afghanistan Pakistan Indonesia Nepal Albania Pakistan Afghanistan Sri Armenia (concluded) B6 ID ableT Bank P072996* P073310* P073394 P073531* P074018 P074075 P074266* P074413* P076784 P077257* P077457 P077533 P078936* P078997* P079156* P081968* P082128* P082977 P084329* P086747* P088499 8 5 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 Explanation of Evaluation Terms Institutional Development Impact: The for Table B6 extent to which a project improves the ability of Outcome: The extent to which the project's a country or region to make more efficient, major relevant objectives were achieved, or are equitable, and sustainable use of its human, expected to be achieved efficiently. Possible financial, and natural resources through: (i) ratings: Highly Satisfactory (HS), Satisfactory better definition, stability, transparency, enforce- (S), Moderately Satisfactory (MS), Moderately ability, and predictability of institutional arrange- Unsatisfactory (MU), Unsatisfactory (U), Highly ments and/or (ii) better alignment of the mission Unsatisfactory (HU). and capacity of an organization with its mandate, which derives from these institutional arrange- Sustainability: The resilience to risk of net ments. Institutional Development Impact benefit flows over time. Possible ratings: Highly includes both intended and unintended effects Likely (HL), Likely (L), Unlikely (UL), Highly of a project. Possible ratings: High (H), Substan- Unlikely (HU), Not Evaluable (NE). tial (S), Modest (M), Negligible (N). 8 6 APPENDIX C: DETAILED PROJECT DESIGN ANALYSIS Detailed Review of Project Documents design of the projects in relationship to the The analysis focused on a total of 80 projects, outcomes and the effectiveness of M&E. randomly drawn from the larger population of 161 agricultural water projects. To facilitate Randomly sampled project documents were analysis, the projects were divided into two scrutinized and scored using a rating scale from groups: dedicated and nondedicated. The 1 to 4. A blank was defined as no significant former includes all projects where the amount evidence of the feature, 1 was defined as committed to I&D is larger than 50 percent of occurrence of the feature to a small extent, 2 was the total IDA/IBRD amount committed for the defined as occurrence of the feature to a whole project. Nondedicated projects have less moderate extent, 3 was defined as occurrence of than half of Bank financing devoted to agricul- the feature to a substantial extent, and 4 was tural water management activities. Each defined as occurrence of the feature to a very sampled project was rated against 34 criteria, high extent. There was also a Not Applicable developed from issues raised by Bank poverty, (NA) and an Addressed Outside the project (AO) rural, and water-strategy statements, as well as rating for some criteria, mainly policy. As is key documents in the literature related to evident, the mid-point on the range where there poverty, institutions, and policy related to water was any occurrence at all lay between 2 and 3. In (table A1). ICRs and PPARs covering the closed a few cases, there was simply a Yes/No rating. projects in the sample were also reviewed where Percentages were asked for, such as percentage available, which provided a clearer picture of the cost recovery, in some cases. 8 7 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 Table C.1. Questionnaire Bank Project ID P****** D 1,522 Dedicated 1 Q1 Proportion of poor in beneficiary total if known Is the project sensitive to the Bank's Poverty alleviation agenda? Q2 Is poverty alleviation mentioned as explicit objective? 1 Q3 Extent to which project design incorporates a poverty focus (even if no poverty objective) Q4 Extent of direct targeting of benefits towards poorer, e.g., selection of location, communities, households Q5 Given both the objectives and design, does this warrant classification as a poverty-focused intervention? 1 Q6 Quality of analysis distributional aspects Q7 Was a social assessment carried out? Q8 Is project employment impact or wages analyzed or substantively discussed? Q9 Extent of gender focus in project design 1 Q10 Predominant hectares benefiting large farms (L) or small (S) as defined by PAD 1 Q11 Extent of analysis of, or substantive discussion of, water rights of beneficiaries Institutional aspects Q12 To what extent is a broader sector strategy of which this project is a part clearly outlined in the PAD? 2 Q13 Extent policy content including legislation, pricing, rights, but excluding institutional reform 3 Q14 Extent institutional reform, e.g., public/private shift, new organization (excludes pricing covered under policy) Q15 To what extent does project aim to reform or significantly strengthen public institutions at central level? 4 Q16 To what extent does project aim to reform or significantly strengthen public institutions below central, e.g., region/district? 4 Community based organizations Q17 To what extent does project aim to strengthen community orgs. or participation, e.g., thru WUAs? 4 Q18 To what extent does PAD propose autonomy for WUAs (in fee collection, retention,expenditure, and water management)? 1 Q19 To what extent does PAD indicate support to WUAs on water management? 3 Cost Recovery Q20 Planned average cost recovery for capital investment (%) Q21 Preproject average cost recovery for O&M (%) 40 Q22 Planned average cost recovery for O&M (%) 90 Q23 To what extent does project address water efficiency, e.g., through water charges, regulations or technical design? 1 Q24 To what extent does PAD indicate how shortfalls in cost recovery are to be handled, e.g., public subsidy, cross-subsidy? Cross-sectoral linkages Q25 To what extent does PAD address environmental water issues such as quality, groundwater depletion, etc.? 3 Q26 To what extent does the PAD propose collaboration with other water agencies? 1 Q27 Which of following are supported or linked to significant extent in project design? Marketing (M), Extension (E), Credit (C) E 8 8 A P P E N D I X C : D E TA I L E D P R O J E C T D E S I G N A N A LY S I S Table C.2. Project Analysis Results ID# P00 P00 P00 P00 P00 P00 P00 P00 P00 P00 P00 P00 P00 P00 P00 P00 P00 P00 3596 3954 4008 4613 4834 4845 4978 5173 5310 5321 5344 5721 5902 7020 7701 8037 8277 8284 Dedi- cated 1 1 1 1 1 1 1 1 1 1 1 1 Q1 >40 53 HIGH 70 >75 80 50 Q2 1 1 1 1 1 1 1 Q3 1 3 2 1 2 4 1 1 4 Q4 1 4 1 1 4 2 1 2 4 2 Q5 1 1 1 Q6 2 2 2 1 3 Q7 1 1 1 1 Q8 1 1 1 1 1 Q9 3 3 3 1 1 3 3 Q10 1 1 1 1 1 1 1 1 1 1 3 1 1 Q11 2 3 2 Q12 2 4 3 3 1 4 4 2 4 1 2 2 3 2 4 3 1 3 Q13 2 4 1 3 2 Q14 2 3 2 3 1 4 1 3 2 3 Q15 3 4 2 1 1 3 2 2 3 1 3 3 Q16 3 3 3 2 2 3 2 2 3 3 Q17 4 3 3 4 1 2 1 4 2 2 3 3 3 4 Q18 3 3 4 2 1 4 2 3 2 3 4 4 Q19 3 2 1 2 2 3 3 2 3 3 3 3 2 3 Q20 20 4 10 0 50 0 50 100 Q21 66 60 60 30 73 0 30 78 20 <20 Q22 100 80 100 100 85 100 100 100 100 100 100 100 Q23 3 3 1 2 2 4 2 3 2 2 4 4 1 2 Q24 4 1 2 2 2 3 1 4 1 Q25 3 4 1 4 1 4 4 3 4 3 2 3 2 4 2 2 3 Q26 2 1 2 3 3 3 2 2 1 1 1 Q27 E EMC E NA EC E E E EC E EC NA (Table continues on next page) 8 9 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 Table C.2 (continued) P00 P00 P00 P00 P00 P01 P01 P01 P01 P01 P03 P03 P03 P03 P03 P03 P03 P03 ID# 8286 8403 9072 9127 9964 0482 0500 0501 0529 0530 4212 4891 5158 5707 7079 8399 8884 8885 Dedi- cated 1 1 1 1 1 1 1 1 1 1 1 Q1 43 >80 55 >70 40 66.2 45 Q2 1 1 1 1 1 1 1 Q3 1 1 2 1 2 1 2 4 3 1 2 4 4 Q4 1 1 2 1 1 2 4 3 2 4 4 Q5 1 1 1 1 1 Q6 2 2 4 2 3 3 2 2 1 2 2 2 2 Q7 1 1 1 1 1 1 1 Q8 1 1 1 1 1 1 1 Q9 2 1 2 2 1 1 1 1 2 3 3 3 Q10 1 3 1 1 2 4 1 1 1 1 2 1 1 Q11 1 4 2 3 3 3 Q12 4 2 1 3 4 2 4 3 4 3 2 4 3 4 2 3 3 3 Q13 3 4 3 3 4 3 3 4 3 Q14 2 3 4 4 4 4 3 3 1 4 3 4 1 1 Q15 3 3 2 2 4 3 4 2 4 3 3 1 3 3 3 3 1 1 Q16 1 2 4 3 1 2 3 3 3 3 2 2 Q17 4 3 2 3 3 2 2 4 4 2 3 2 4 3 4 3 3 3 Q18 2 3 3 2 4 2 3 4 3 2 2 2 3 2 2 Q19 3 3 2 2 3 2 2 2 3 3 2 3 3 3 Q20 20 70 95 25 55 25 10 7 0 10 0 10 10 Q21 100 20 <20 0 16 40 <20 0 0 >90 10 0 0 0 Q22 100 100 10 100 100 100 100 100 100 100 100 100 100 100 100 75 75 Q23 2 2 1 4 1 1 4 2 2 3 3 Q24 1 2 1 2 1 1 1 Q25 3 4 2 4 3 1 4 2 2 1 3 4 2 Q26 2 3 2 1 1 2 1 3 2 2 3 Q27 MEC E NA E E E MC ME E MEC EM AO (Table continues on next page) 9 0 A P P E N D I X C : D E TA I L E D P R O J E C T D E S I G N A N A LY S I S Table C.2 (continued) P04 P04 P04 P04 P04 P04 P04 P04 P04 P04 P04 P04 P05 P05 P05 P05 P05 P05 ID# 1410 1723 2442 3881 6042 6043 6045 8522 9385 9665 9718 9723 0646 0647 1922 5022 5068 5974 Dedi- cated 1 1 1 1 1 1 1 1 Q1 HIGH >80 48 HIGH >40 74 75 47 >70 70 Q2 1 1 1 Q3 2 4 4 2 1 3 2 4 2 Q4 3 4 1 2 3 1 4 3 4 2 Q5 1 1 1 1 1 1 Q6 3 3 1 1 2 1 2 4 3 2 Q7 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Q8 1 1 1 3 1 Q9 3 2 2 4 3 3 Q10 1 1 1 2 1 4 1 1 1 1 1 1 1 1 1 Q11 1 Q12 2 4 3 4 2 4 2 4 2 4 3 2 3 4 4 1 3 Q13 4 3 3 3 1 4 3 4 Q14 2 3 2 4 3 1 3 4 3 4 2 4 Q15 3 3 2 4 3 1 3 3 1 2 3 4 4 3 3 Q16 4 2 3 3 1 3 2 1 3 3 4 4 3 3 Q17 3 3 4 1 3 2 4 4 3 4 4 3 Q18 4 3 2 2 1 3 4 2 3 4 Q19 2 2 1 2 1 2 3 3 3 2 3 3 Q20 30 >20 7 50 0 15 25 50 >15 30 Q21 0 15 <20 0 0 <10 0 0 <20 Q22 100 100 100 <20 100 0 100 100 100 100 100 100 Q23 3 4 2 2 2 3 1 3 1 3 4 2 Q24 1 1 1 3 1 1 Q25 4 2 2 3 2 4 2 2 1 3 3 2 1 2 2 2 Q26 2 2 3 2 2 3 1 3 3 1 Q27 EC E E ECM E M E ECM E EMC (Table continues on next page) 9 1 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 Table C.2 (continued) P05 P05 P05 P05 P05 P06 P06 P06 P06 P06 P06 P07 P07 P07 P07 P07 P07 P07 ID# 7271 8070 8841 8898 9803 2682 2714 2748 4879 5463 6335 1033 1092 2996 3531 4266 4413 7257 Dedi- cated 1 1 1 1 1 1 Q1 64 28 >40 HIGH 37 HIGH 86 Q2 1 1 1 1 1 1 Q3 2 3 4 2 4 4 3 4 2 2 4 2 1 4 Q4 3 3 4 1 4 4 2 4 2 2 4 1 1 4 Q5 1 1 1 1 1 1 1 1 1 1 Q6 3 4 3 4 3 3 1 2 1 4 Q7 1 1 1 1 1 1 1 1 1 1 1 1 1 Q8 1 1 1 Q9 3 3 3 4 4 1 3 3 2 4 Q10 1 1 1 1 1 1 1 1 1 1 1 4 1 Q11 2 3 1 2 Q12 3 3 2 4 4 4 3 4 3 3 3 2 2 3 2 4 Q13 3 3 3 1 1 4 2 2 1 Q14 1 1 4 2 1 3 1 3 3 1 3 4 2 2 1 Q15 2 2 2 3 1 3 1 3 1 1 2 2 3 Q16 3 3 3 3 3 3 2 2 3 2 2 2 3 Q17 2 3 3 4 2 4 2 4 4 3 2 3 1 3 Q18 3 4 2 4 2 1 1 Q19 2 1 3 3 2 3 3 3 2 3 3 Q20 10 20 30 5 5 10 12 22 20 50 27 Q21 0 0 0 0 0 0 30 0 0 Q22 100 100 100 50 100 100 100 100 100 Q23 3 3 2 3 2 4 Q24 1 2 3 2 2 2 1 Q25 2 1 3 4 3 2 3 1 4 1 2 1 4 2 Q26 1 2 2 4 1 1 4 3 1 Q27 CME MC E C-AO E M E M C- E AO MEC CM EM M (Table continues on next page) 9 2 A P P E N D I X C : D E TA I L E D P R O J E C T D E S I G N A N A LY S I S Table C.2 (concluded) P07 P07 P07 P08 P08 P08 P08 ID# 8936 8997 9156 1968 2128 4329 6747 Dedi- cated 1 1 1 1 Q1 60 HIGH Q2 1 1 1 Q3 3 3 2 4 1 3 3 Q4 2 3 2 4 3 3 Q5 1 1 1 1 1 1 Q6 1 3 4 1 3 Q7 1 1 1 1 Q8 1 Q9 3 4 3 1 2 2 Q10 1 1 1 1 1 1 1 Q11 2 1 Q12 2 3 4 3 2 4 3 Q13 2 1 Q14 2 2 1 Q15 1 1 3 2 2 Q16 2 2 2 3 2 Q17 2 4 3 2 4 4 4 Q18 3 3 4 3 3 2 Q19 2 4 1 3 1 2 Q20 30 10 10 0 10 Q21 50 0 100 Q22 100 100 100 100 100 Q23 3 3 1 Q24 Q25 3 3 1 2 1 Q26 2 3 3 2 Q27 E C CM E E 9 3 APPENDIX D: ANALYSIS OF MONITORING AND EVALUATION The 80 random projects were independently application of 17 evaluation questions (table D1) reviewed to determine how well M&E had been and their overall categorization, to determine designed. A follow-up analysis, employing ICR the overall quality of M&E from an evaluation output and outcome data from these completed and impact assessment perspective (table D2). projects, was used to determine how well M&E Findings are presented in table D3; evaluation had been implemented. This was based on the ratings are given in table D4. Table D1. Evaluation of M&E Design and Implementation Evaluation Questions Evaluative score Objectives: level of clarity (1= negligible, 4 = high) 1 to 4 Logical framework (0 = absent, 1 = negligible, 4 = high) 0 to 4 M&E system in place ex ante 0 to 4 M&E system specified by project 0 to 4 Quality of M&E plan 0 to 4 Desired outputs clearly defined 0 to 4 Desired outcomes clearly defined 0 to 4 Indicators well structured 0 to 4 Focus on outcome indicators 0 to 4 M responsibility assigned 0 to 4 M coordination assigned 0 to 4 Support for M&E capacity building 0 to 4 Feedback loops for management 0 to 4 Evaluative baseline specified 0 to 4 M&E as condition of lending 0 or 1 ICR (0= active project, 1 = completed project) 0 or 1 Baseline status (0 = none, 3 = start, 2= middle, 1=end) 0 to 3 Ex post survey(s) 0 to 4 Transparency of evaluation (0 = none; 1 = negligible; 4 = high 0 to 4 Table D2. Typology for Classification of M&E Findings from Project Design Output indicators, but no outcome or impact indicators Output indicators as well as outcome or impact indicators, but no baseline Output and/or impact indicators and baseline Output and/or impact indicators and control groups but no baseline Outcome and/or impact indicators and baseline and control groups 9 5 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 ADP x x x x x x x x x 12 38% Reworked x x x 3 9% Ex-post Survey Without Px P x x x x x x P x x P x x P x x x x Px Px P x x P x 27 84% Ex-post Survey With x x x x 4 13% Survey During Project control. x x x x x x x x x x x x x x xx 15 47% Ex-ante Survey Without independent na as Projects x x 2 6% or Control Without Project project the x x x x x x x x x x x x x x x x x x Completed 18 after 56% or in Baseline Without Project before either x x x 3 Late 9% Baseline means Implemented "without" III Resources Irrigation Investment Project occurrence; Assistance ater Irrigation Pilot Project Irrigation Arrangements ADP W Improvement Development Irrigation Groundwater Resources projects partial Irrigation Sector Management Minor Drainage of WRCP Irrigation Supply = II Plain Rehabilitation Restructuring all Basin Res & Subsector Rehabilitation Rehabilitation & P WRCP Sector Reform Smallholder of M&E Private Irriagtion enggaraT Sector WRCP Pradesh Emergency Nadu and Natural ater W 2 full; Niño Irrigation in of Project Irrigation Pilot Songliao angtzeY Java Nusa Irrigation Irrigation Agricultural Natural On-Farm Irrigation Irrigation Irrigation Agriculture Irrigation Privatization Haryana amilT Balochistan Private Orissa Irrigation Mahaweli Andrah Agrarian RBM Sierra aizT ASIL El NE otalT Percent measure this of Summary Lanka Lanka Country Madagascar Niger China China Indonesia Indonesia ietnamV Lebanon unisiaT unisiaT Mexico Peru Albania Armenia Estonia Kazakhstan urkeyT India India Pakistan Pakistan India Nepal Sri India Philippines anzaniaT Peru emenY unisiaT Guyana Sri D3. ID occurrence = X ableT Project P001522 P001994 P003593 P003596 P003954 P004008 P004834 P005344 P005721 P005736 P007701 P008037 P008270 P008277 P008403 P008510 P009072 P009964 P010476 P010482 P010501 P010529 P010530 P034212 P035158 P037079 P038570 P042442 P043367 P050418 P057271 P058070 Key: 9 6 A P P E N D I X D : A N A LY S I S O F M O N I T O R I N G A N D E VA L U AT I O N 7020 1995 2.5 ? 2 2 2 3 2 ? 2 4 4 2 3 0 3 2005 2.5 3.5 1 page) next on y 5321 1995 2.5 2 2 2.5 2.5 3 2.5 3 2.5 3 3 3 3 2 11 2 3 3 1 2001 continued n 4 0 3 3 5310 1995 2.5 3 4 3 3 4 4 3 4 3 1 1 4 3 1 1997 able(T y 4 0 3 5173 1995 2.5 3 4 3 3 2.5 2.5 2 3 4 3 1 y 4834 1995 2.5 0 0 2 2 3 1 0 0 4 4 2 3 0 1 3 2 2 1 2004 n 9964 1994 2.5 0 2 3.5 3.5 3 2 3 2 4 3 3 4 3 1 2 2 3 1 2002 y 2 2 1 1 2 3 2 2 1 4 4 2 3 0 1 0 2 3 1 8277 1994 2002 3 0 2 4 3 4 4 4 3 4 4 1 4 3 1 1 3 4 1 7701 1994 2003 Projects 2 0 2 2 2 3 5721 1994 2.5 2.5 2 4 2.5 3 3 2.5 1 1 2001 2.5 2 1 Sampled yny 3 0 2 3 2 3 2 2 2 3 3 2 3 0 1 0 2 3 1 5344 1994 2004 3 0 2 0 1 4 3 1 2 4 2 1 4 0 1 0 3 4 1 3954 1994 2003 Randomly 80 of ny 3 0 0 2 2 4 2 2 0 3 4 3 3 2 1 3 2 1 3596 1994 2003 M&E y 3 0 2 2 1 1 1 1 0 4 2 2 3 0 11 0 2 2 1 for 1522 1994 2002 and and outcome baseline Ratings project ante defined building outcome as no groups by defined indicators ex ypologyT no but indicators indicators baseline well P###### clearly indicators management specified no control clearly assigned capacity but as ID: place plan structured assigned for evaluation but impact impact impact and Questions in specified of clear M&E M&E well loops baseline indicators indicators, Evaluation and/or framework of outputs outcomes outcome for status survey(s) Evaluation groups Project system system on indicators, indicators and/or and/or baseline D4. Approved responsibility coordination Completed impact impact Bank FY Evaluation Objectives Logical M&E M&E Quality Desired Desired Indicators Focus M M Support Feedback Evaluative FY RCI post Baseline Ex ransparencyT or or baseline control and Impact Output Output Output Output Outcome ableT 1 2 3 4 5 6 8 9 10 5 6 7 11 12 13 14 15 1 2 3 4 5 9 7 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 y 4 2 3 3 3 3 2 2 2 4 4 2 4 2 1 0 3 2 1 10529 1996 2005 page) next on y 0 3 4 4 3 3 3 4 4 4 1 1 1 4 4 1 10501 1996 2.8 3.5 3.8 2002 continues n 0 2 3 2 2 3 2 2 0 1 1 4 4 1 10482 1996 3.5 2.5 2.5 2.5 3.5 2003 able(T n 4 3 3 8403 1996 2.5 2.5 3 2 2 2 4 2.5 1 3 2 1 1 2 3 1 2003 y 3 0 2 1 2 4 2 3 2 4 3 2 3 0 1 3 3 4 1 8037 1996 2005 n 3 0 2 4978 1996 3.5 3.5 3.5 2.5 3 3 4 4 3 4 0 1 1 3 2001 2.5 1 n 3 2 4 4 4 3 3 4 4 4 4 4 4 1 1 3 3 1 38885 1995 3.5 2002 n 0 4 4 3 2 2 2 4 4 3 4 3 1 2 4 4 1 38884 1995 2.5 2.5 2002 n 0 1 2 2 3 2 1 4 2 2 3 0 1 0 3 1 34891 1995 3.5 1.5 1999 2.5 0 2 3 3 4 2 4 3 3 4 2 3 0 1 3 3 3 1 43178 1995 3.5 2006 yy 7020 1995 2.5 ? 2 2 2 3 2 ? 2 4 4 2 3 0 1 3 2005 2.5 3.5 1 and groups baseline baseline control and and and outcome baseline project ante defined building outcome as no by defined indicators ex ypologyT no but indicators indicators well P###### clearly indicators management specified clearly assigned capacity but as ID: place plan structured assigned for evaluation impact impact impact Questions in specified of clear M&E M&E well indicators indicators, outcome loops baseline and/or (continued) framework of outputs outcomes for status survey(s) Evaluation baseline groups Project system system on indicators, indicators and/or and/or no D4 Approved responsibility coordination Completed post impact impact Bank FY Evaluation Objectives Logical M&E M&E Quality Desired Desired Indicators Focus M M Support Feedback Evaluative FY ICR Baseline Ex ransparencyT or or but control Impact Output Output Output Output Outcome ableT 1 2 3 4 5 6 8 9 10 5 6 7 11 12 13 14 15 1 2 3 4 5 9 8 A P P E N D I X D : A N A LY S I S O F M O N I T O R I N G A N D E VA L U AT I O N y 3 3 3 3 3 4 3 4 3 4 4 3 4 4 1 3 2 3 1 35158 1998 2005 page) next on y 4 4 3 2 2 4 4 4 4 4 4 1 3 0 1 3 1 2 1 34212 1998 2004 continues y 3 0 3 3 3 3 3 2 2 4 1 1 3 2 10530 1998 2.5 2.5 3.5 2005 2.5 able(T y 3 3 3 3 3 4 4 3 4 1 10500 1998 2.5 2.5 2.5 3.5 2.5 y 3 0 2 2 2 9072 1998 2.5 2.5 2 2 3 2.5 2 3.5 2 1 2 1 4 11 2006 n 4 4 1 5902 1998 2.5 3 4 4 4 4 4 2.5 2 3 4 1 n 4 2 2 2 1 4 3 2 0 0 4 0 3 3 1 1 3 2 1 48522 1997 2002 n 0 1 2 3 2 3 2 4 2 3 1 2 1 42442 1997 2.5 2.5 3.5 1.5 2005 2.5 2.5 n 3 0 4 3 3 3 2 2 2 4 3 3 4 2 1 2 4 3 1 37079 1997 2004 4613 1997 2.8 0 3 3.5 3.5 3 4 3.5 4 3 2.5 2 4 4 1 ny 4008 1997 3.5 2 2.5 3 3 3.5 3 3 3 3 2.5 3 4 2.5 1 2 3 3 1 2004 but indicators, and indicators groups impact impact or baseline baseline control or and and and outcome project ante defined building outcome as by defined indicators ex ypologyT no indicators indicators well P###### clearly indicators management specified clearly assigned capacity but as ID: place plan structured assigned for evaluation impact impact impact Questions in specified of clear M&E M&E well outcome loops baseline and/or (continued) framework of outputs outcomes for status survey(s) Evaluation baseline groups Project system system on indicators, indicators and/or and/or no D4 Approved baseline responsibility coordination Completed post Bank FY Evaluation Objectives Logical M&E M&E Quality Desired Desired Indicators Focus M M Support Feedback Evaluative FY ICR Baseline Ex ransparencyT but no control Impact Output Output Output Output Outcome ableT 1 2 3 4 5 6 8 9 10 5 6 7 11 12 13 14 15 1 2 3 4 5 9 9 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 n 4 2 4 3 2 4 1 1 3 4 1 62682 1999 2.5 2.5 2.5 2.5 3.5 2.5 2.5 2005 page) next on n 3 1 3 2 4 3 3 0 1 58841 1999 2.8 2.5 2.5 2.5 2.5 3.5 continues y 4 3 3 4 3 3 3 2 4 4 1 50646 1999 2.5 2.5 2.5 2.5 able(T n 2 3 3 3 4 4 3 4 3 1 49718 1999 3.5 2.5 2.5 2.5 2.5 n 4 3 4 2 2 2 2 4 0 1 49665 1999 2.5 2.5 2.5 2.5 2.5 y 3 3 4 4 4 3 0 1 41410 1999 2.5 2.5 2.5 2.5 3.5 3.5 2.5 y 3 3 2 2 2 4 3 3 3 3 2 2 3 0 1 4845 1999 n 3 2 3 2 2 2 1 2 2 2 1 3 0 11 0 2 2 1 55974 1998 2.5 2001 n 3 2 3 3 3 2 2 3 2 2 4 2 1 49385 1998 2.5 2.5 4 3 4 4 3 3 4 4 0 1 46042 1998 2.5 2.5 2.5 2.5 3.5 yy 3 0 2 2 4 4 3 3 4 4 1 38399 1998 2.5 2.8 2.5 2.5 but indicators, and indicators groups impact impact or baseline baseline control or and and and outcome project ante defined building outcome as by defined indicators ex ypologyT no indicators indicators well P###### clearly indicators management specified clearly assigned capacity but as ID: place plan structured assigned for evaluation impact impact impact Questions in specified of clear M&E M&E well outcome loops baseline and/or (continued) framework of outputs outcomes for status survey(s) Evaluation baseline groups Project system system on indicators, indicators and/or and/or no D4 Approved baseline responsibility coordination Completed post Bank FY Evaluation Objectives Logical M&E M&E Quality Desired Desired Indicators Focus M M Support Feedback Evaluative FY ICR Baseline Ex ransparencyT but no control Impact Output Output Output Output Outcome ableT 1 2 3 4 5 6 8 9 10 5 6 7 11 12 13 14 15 1 2 3 4 5 1 0 0 A P P E N D I X D : A N A LY S I S O F M O N I T O R I N G A N D E VA L U AT I O N n 4 2 2 2 3 2 3 2 4 4 2 4 0 1 50647 2001 2.5 page) next on y 3 2 4 3 3 3 3 4 3 4 4 0 1 46045 2001 3.5 2.5 continues y 4 3 2 2 2 4 3 3 3 4 4 3 4 0 58898 2000 able(T n 3 3 3 2 3 4 3 4 4 3 1 58070 2000 2.8 2.5 2.8 2.8 y 4 3 4 4 3 4 4 2 4 3 1 49723 2000 3.5 2.5 3.5 2.8 n 3 3 4 4 4 2 4 1 1 41723 2000 3.5 2.5 3.5 3.5 3.5 3.5 y 3 4 3 2 2 3 2 2 3 2 3 2 3 0 11 35707 2000 y 4 4 3 8284 2000 3.5 3.8 4 4 4 4 4 4 3 4 3 1 n 3 3 3 3 3 4 2 4 4 1 73310 1999 3.5 3.5 3.5 3.5 3.5 3 3 2 4 3 4 0 1 66335 1999 2.5 3.5 2.5 2.5 3.5 2.5 3.5 yn 4 3 4 2 4 4 1 64879 1999 3.5 2.5 2.5 2.5 2.5 2.8 2.5 2.5 but indicators, and indicators groups impact impact or baseline baseline control or and and and outcome project ante defined building outcome as by defined indicators ex ypologyT no indicators indicators well P###### clearly indicators management specified clearly assigned capacity but as ID: place plan structured assigned for evaluation impact impact impact Questions in specified of clear M&E M&E well outcome loops baseline and/or (continued) framework of outputs outcomes for status survey(s) Evaluation baseline groups Project system system on indicators, indicators and/or and/or no D4 Approved baseline responsibility coordination Completed post Bank FY Evaluation Objectives Logical M&E M&E Quality Desired Desired Indicators Focus M M Support Feedback Evaluative FY ICR Baseline Ex ransparencyT but no control Impact Output Output Output Output Outcome ableT 1 2 3 4 5 6 8 9 10 5 6 7 11 12 13 14 15 1 2 3 4 5 1 0 1 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 y 3 4 3 3 3 4 3 3 4 4 3 3 4 3 1 8286 2003 page) next on n 3 1 3 2 2 4 3 3 4 0 1 73531 2002 2.5 3.5 3.5 2.5 continues y 4 2 2 3 4 4 4 4 2 2 3 3 2 1 72996 2002 3.5 able(T y 3 3 2 2 2 3 3 2 2 2 2 3 3 2 1 71033 2002 n 3 3 4 3 3 4 2 4 3 1 46043 2002 3.5 2.5 2.5 2.5 2.5 y 3 3 2 2 3 4 4 3 4 2 1 71092 2001 2.5 3.5 2.5 2.5 n 3 1 3 3 3 3 3 3 2 4 0 1 62748 2001 2.5 2.5 2.5 n 3 2 2 1 1 2 2 2 2 2 1 3 2 1 62714 2001 2.5 y 4 3 1 2 2 3 2 3 2 2 2 2 3 0 1 55068 2001 3 3 4 3 4 3 3 4 1 55022 2001 3.5 3.5 3.5 3.5 3.5 2.5 ny 2 3 3 3 4 3 4 4 1 51922 2001 3.5 1.5 3.5 3.5 3.5 3.5 but indicators, and indicators groups impact impact or baseline baseline control or and and and outcome project ante defined building outcome as by defined indicators ex ypologyT no indicators indicators well P###### clearly indicators management specified clearly assigned capacity but as ID: place plan structured assigned for evaluation impact impact impact Questions in specified of clear M&E M&E well outcome loops baseline and/or (continued) framework of outputs outcomes for status survey(s) Evaluation baseline groups Project system system on indicators, indicators and/or and/or no D4 Approved baseline responsibility coordination Completed post Bank FY Evaluation Objectives Logical M&E M&E Quality Desired Desired Indicators Focus M M Support Feedback Evaluative FY ICR Baseline Ex ransparencyT but no control Impact Output Output Output Output Outcome ableT 1 2 3 4 5 6 8 9 10 5 6 7 11 12 13 14 15 1 2 3 4 5 1 0 2 A P P E N D I X D : A N A LY S I S O F M O N I T O R I N G A N D E VA L U AT I O N y 4 3 4 4 3 3 4 4 3 4 4 1 82128 2004 3.5 3.5 3.5 page) next on n 3 2 3 3 3 3 3 4 3 1 81968 2004 2.5 2.5 2.5 2.5 2.5 continues y 3 4 3 4 3 4 3 4 3 4 4 3 4 4 1 78997 2004 able(T n 2 4 4 4 4 3 4 4 4 4 4 1 77257 2004 3.5 3.5 3.5 y 3 3 4 4 3 4 0 1 74413 2004 3.5 2.5 2.5 3.5 2.5 2.5 3.5 n 3 3 3 3 3 3 3 1 74266 2004 2.5 3.5 3.5 2.5 2.5 3.5 3.5 y 2 3 3 2 2 2 4 3 1 43881 2004 3.5 2.5 2.5 3.5 2.8 2.5 3 3 3 4 3 4 3 3 4 3 2 4 3 1 79156 2003 3.5 3 2 0 3 4 4 3 3 3 4 4 3 4 3 1 78936 2003 nyn 3 3 3 3 2 4 0 1 59803 2003 2.5 2.5 2.5 2.5 2.5 2.5 2.5 y 4 4 2 2 2 4 3 3 3 3 3 2 4 0 1 9127 2003 but indicators, and indicators groups impact impact or baseline baseline control or and and and outcome project ante defined building outcome as by defined indicators ex ypologyT no indicators indicators well P###### clearly indicators management specified clearly assigned capacity but as ID: place plan structured assigned for evaluation impact impact impact Questions in specified of clear M&E M&E well outcome loops baseline and/or (continued) framework of outputs outcomes for status survey(s) Evaluation baseline groups Project system system on indicators, indicators and/or and/or no D4 Approved baseline responsibility coordination Completed post Bank FY Evaluation Objectives Logical M&E M&E Quality Desired Desired Indicators Focus M M Support Feedback Evaluative FY ICR Baseline Ex ransparencyT but no control Impact Output Output Output Output Outcome ableT 1 2 3 4 5 6 8 9 10 5 6 7 11 12 13 14 15 1 2 3 4 5 1 0 3 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 14% 34% 41% 0% 11% Percent otalT 11 27 33 9 3 4 3 4 4 3 3 3 4 4 2 3 3 1 86747 2004 3.5 ny 4 0 1 3 3 2 4 4 4 0 1 84329 2004 3.5 3.5 2.5 3.5 but indicators, and indicators groups impact impact or baseline baseline control or and and and outcome project ante defined building outcome as by defined indicators ex ypologyT no indicators indicators well P###### clearly indicators management specified clearly assigned capacity but as ID: place plan structured assigned for evaluation impact impact impact Questions in specified of clear M&E M&E well outcome loops baseline and/or (concluded) framework of outputs outcomes for status survey(s) Evaluation baseline groups Project system system on indicators, indicators and/or and/or no D4 Approved baseline responsibility coordination Completed post Bank FY Evaluation Objectives Logical M&E M&E Quality Desired Desired Indicators Focus M M Support Feedback Evaluative FY ICR Baseline Ex ransparencyT but no control Impact Output Output Output Output Outcome ableT 1 2 3 4 5 6 8 9 10 5 6 7 11 12 13 14 15 1 2 3 4 5 1 0 4 APPENDIX E: ANALYSIS OF COUNTRY ASSISTANCE STRATEGIES One hundred and thirty CASs were reviewed, Table E1. Topics Researched in covering 54 countries during the period FY94­ 130 CASs FY04, each of which was entered into a database. The primary analysis was text-based, Irrigation aided by the commercial software package Atlas Institutional Development/Reform (Irrigation & Water) Ti.1 The analysis interrogated the entire text of Participation in Irrigation each CAS to determine the presence of 21 key Rural Development phrases (table E1). When a phrase was encoun- Irrigation & Rural Development tered, the paragraph in which that phrase Rural Poverty Alleviation occurred was extracted and stored. Following Role of Rural Development in Poverty Alleviation scrutiny to determine the relevance of each hit, Institutional Development in Rural Development the relevant hits were added to the summary Participation in Rural Development database. By this method, 544 AWM references, Rural Development & Agricultural Growth in 124 CASs containing phrases relevant to this Agricultural Development study's evaluation questions could be related to Agricultural Growth and Irrigation time, 51 countries and regions. Detailed tables Agricultural Development & Economic Growth of the results follow. Institutional Reform/Development in Agriculture Participation in Agriculture Water Resource Management Irrigation & Water Cost Recovery Water Users Water Tariffs Technical Assistance 1 0 5 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 page) 3 2 3 6 2 1 1 1 next 1995 19 8 50 8,574 on Cambodia continues 1 2 1 2 1 1 2 1 1995 12 9 14 Brazil 32,163 able(T 2 21 3 4 1 4 1 1995 17 7 30 11,153 Bangladesh 2 1 1 4 3 6 1995 19,705 Argentina 1 1 1 2 1 1 1 1 1994 10 9 27 ietnamV 4,990 1 1 1 1 2 1 11 1 1 1994 10 9 7 Peru 1,647 1 1 2 5 2 1 1 1994 13 7 38 Niger 375 1 2 1 1 1 1 1 8 7 5 1994 3,836 Mexico Faso 1 5 2 1 9 4 1994 35 1,640 Burkina ater W Alleviation Growth Agriculture Results of Development Growth in Poverty Rural in Economic 1997) in & Irrigation GDP Analysis Development Agricultural of 1996) & and Development % Irrigation Rural Agriculture Management a CAS Development/Reform in Alleviation Development ater Rural Development in Development Growth Development Reform/Development in as W mentions topics Resources E1. & & Assistance earY Rural of of Development Poverty of Development Resource Users ariffsT added meters/capita Recovery ableT ater ater ater CAS Country Irrigation Institutional Participation Rural Irrigation Rural Role Institutional Participation Rural Agricultural Agricultural Agricultural Institutional Participation W Irrigation Cost W W echnicalT Number Number Agriculture (value Freshwater (cubic 1 0 6 A P P E N D I X E : A N A LY S I S O F C O U N T R Y A S S I S TA N C E S T R AT E G I E S page) 1 1 1 1 3 2 1 2 2 2 16 01 next 26 1995 702 on Uzbekistan continues 4 1 4 1 1 4 1 1 2 1 20 01 26 1995 1,858 Pakistan able(T 1 1 1 1 1 1 6 6 5 1995 3,836 Mexico 2 4 1 1 8 4 11 1995 Macedonia 1 4 2 7 3 5 1995 158 Jordan 1 1 1 5 1 1 10 6 16 1995 12,839 Indonesia 2 3 1 2 1 2 1 2 1 15 9 27 1995 India 1,957 1 1 6 1 3 1 1 14 56 1995 1,889 Ethiopia 3 1 1 5 37 20 1995 China 2,304 ater W Alleviation Growth Agriculture of Development Growth in Poverty Rural in Economic 1997) in & Irrigation GDP Development Agricultural of 1996) & and Development % Irrigation Rural Agriculture Management a (continued) Development/Reform in Alleviation Development ater Rural Development in Development Growth Development Reform/Development in as W mentions topics Resources E1 & & Assistance earY Rural of of Development Poverty of Development Resource Users ariffsT added meters/capita Recovery ableT ater ater ater CAS Country Irrigation Institutional Participation Rural Irrigation Rural Role Institutional Participation Rural Agricultural Agricultural Agricultural Institutional Participation W Irrigation Cost W W echnicalT Number Number Agriculture (value Freshwater (cubic 1 0 7 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 page) 2 1 3 2 next 5 1996 3,836 on Mexico continues 2 2 1 1 7 5 1996 52 Laos 9,840 able(T 1 1 4 3 21 1 1 2 2 1996 18 10 12 Ecuador 26,842 1 1 2 4 3 1996 39 Chad 2,269 Faso 5 1 11 1 8 4 1996 35 1,640 Burkina 3 3 1 1996 10 2,154 Bulgaria 5 1 3 9 3 1996 22 1,068 Azerbaijan 3 1 4 2 1996 12 483 Algeria 1 4 1 1 1 1 1 1 1 1 1995 13 10 27 ietnamV 4,990 ater W Alleviation Growth Agriculture of Development Growth in Poverty Rural in Economic 1997) in & Irrigation GDP Development Agricultural of 1996) & and Development % Irrigation Rural Agriculture Management a (continued) Development/Reform in Alleviation Development ater Rural Development in Development Growth Development Reform/Development in as W mentions topics Resources E1 & & Assistance earY Rural of of Development Poverty of Development Resource Users ariffsT added meters/capita Recovery ableT ater ater ater CAS Country Irrigation Institutional Participation Rural Irrigation Rural Role Institutional Participation Rural Agricultural Agricultural Agricultural Institutional Participation W Irrigation Cost W W echnicalT Number Number Agriculture (value Freshwater (cubic 1 0 8 A P P E N D I X E : A N A LY S I S O F C O U N T R Y A S S I S TA N C E S T R AT E G I E S page) 1 1 1 next 1997 on Bosnia continues 1 2 3 2 1997 44 2,411 Armenia able(T 1 1 1 6 1997 19,705 Argentina 5 3 8 2 1996 16 8,703 Zambia 1 1 2 1 3 8 5 1996 emenY 18 260 2 1 3 2 1996 14 unisiaT 385 1 2 1 1 5 4 1996 ajikistanT 11,186 1996 Lanka 2 1 1 3 1 8 5 22 2,361 Sri 2 4 4 3 2 2 1 1996 18 7 20 4,492 Philippines ater W Alleviation Growth Agriculture of Development Growth in Poverty Rural in Economic 1997) in & Irrigation GDP Development Agricultural of 1996) & and Development % Irrigation Rural Agriculture Management a (continued) Development/Reform in Alleviation Development ater Rural Development in Development Growth Development Reform/Development in as W mentions topics Resources E1 & & Assistance earY Rural of of Development Poverty of Development Resource Users ariffsT added meters/capita Recovery ableT ater ater ater CAS Country Irrigation Institutional Participation Rural Irrigation Rural Role Institutional Participation Rural Agricultural Agricultural Agricultural Institutional Participation W Irrigation Cost W W echnicalT Number Number Agriculture (value Freshwater (cubic 1 0 9 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 page) 6 1 1 3 1 next 1997 13 6 16 12,839 on Indonesia continues 1 2 6 2 1 1 21 1 5 1 2 2 2 1 2 1997 31 15 27 India 1,957 able(T 0 0 1997 35 Georgia 10,737 1 2 1 2 1 8 6 1997 56 1,889 Ethiopia 2 2 1 1 2 3 4 11 1 1997 17 9 16 47 Egypt 2 1 4 1 1 1 2 3 1 1997 16 9 16 28,571 Colombia 1 2 2 5 3 1 1 1 2 1 1997 19 10 20 China 2,304 2 1 6 3 1 1 1997 14 6 50 8,574 Cambodia 1 2 3 1 1 1 6 1997 15 7 14 Brazil 32,163 ater W Alleviation Growth Agriculture of Development Growth in Poverty Rural in Economic 1997) in & Irrigation GDP Development Agricultural of 1996) & and Development % Irrigation Rural Agriculture Management a (continued) Development/Reform in Alleviation Development ater Rural Development in Development Growth Development Reform/Development in as W mentions topics Resources E1 & & Assistance earY Rural of of Development Poverty of Development Resource Users ariffsT added meters/capita Recovery ableT ater ater ater CAS Country Irrigation Institutional Participation Rural Irrigation Rural Role Institutional Participation Rural Agricultural Agricultural Agricultural Institutional Participation W Irrigation Cost W W echnicalT Number Number Agriculture (value Freshwater (cubic 1 1 0 A P P E N D I X E : A N A LY S I S O F C O U N T R Y A S S I S TA N C E S T R AT E G I E S page) 1 2 5 2 1 next 1998 12 6 54 Albania 12,758 on continues 1 4 1 1 11 1 1 3 1998 13 8 31 5,476 able(T Uzbekistan 1 1 1 1 4 4 1997 urkeyT 17 3,126 3 2 5 2 1997 48 anzaniaT 2,623 2 1 3 2 1997 21 1,637 Romania 1 1 1 1 1 2 1 1 9 8 7 1997 Peru 1,647 1 1 1 1 2 1 7 6 1997 38 Niger 375 1 13 4 2 1 3 1 1 2 1 1997 29 10 20 1,110 Morocco 2 1 3 2 1 6 1 1 1997 17 8 32 24,590 Madagascar ater W Alleviation Growth Agriculture of Development Growth in Poverty Rural in Economic 1997) in & Irrigation GDP Development Agricultural of 1996) & and Development % Irrigation Rural Agriculture Management a (continued) Development/Reform in Alleviation Development ater Rural Development in Development Growth Development Reform/Development in as W mentions topics Resources E1 & & Assistance earY Rural of of Development Poverty of Development Resource Users ariffsT added meters/capita Recovery ableT ater ater ater CAS Country Irrigation Institutional Participation Rural Irrigation Rural Role Institutional Participation Rural Agricultural Agricultural Agricultural Institutional Participation W Irrigation Cost W W echnicalT Number Number Agriculture (value Freshwater (cubic 1 1 1 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 page) 0 0 next 1998 27 Kenya 1,031 on continues 2 2 1 1998 48 anzaniaT 2,770 able(T 4 4 1 6 1998 ajikistanT 13,017 1 1 2 2 1998 11 3,483 Macedonia 5 2 7 2 1998 18 24,663 Bulgaria 1 1 2 2 1998 16 Bolivia 38,625 2 1 1 4 3 1998 26 ietnamV 11,647 1 2 4 2 1 1 1998 11 6 41 Nepal 9,199 1 6 1 2 1 2 1998 13 6 21 9,636 Bangladesh ater W Alleviation Growth Agriculture of Development Growth in Poverty Rural in Economic 1997) in & Irrigation GDP Development Agricultural of 1996) & and Development % Irrigation Rural Agriculture Management a (continued) Development/Reform in Alleviation Development ater Rural Development in Development Growth Development Reform/Development in as W mentions topics Resources E1 & & Assistance earY Rural of of Development Poverty of Development Resource Users ariffsT added meters/capita Recovery ableT ater ater ater CAS Country Irrigation Institutional Participation Rural Irrigation Rural Role Institutional Participation Rural Agricultural Agricultural Agricultural Institutional Participation W Irrigation Cost W W echnicalT Number Number Agriculture (value Freshwater (cubic 1 1 2 A P P E N D I X E : A N A LY S I S O F C O U N T R Y A S S I S TA N C E S T R AT E G I E S page) Faso 6 1 2 1 1 2 1 1 next 2000 15 8 32 1,671 on Burkina continues 1 1 1 2 1999 198 Jordan able(T 4 1 3 1 9 4 1999 emenY 17 254 1 1 3 1 6 4 1999 11 2,467 Dominica 1 1 1 2 1 6 5 1999 17 Zambia 12,001 2 1 1 1 2 7 5 1999 53 Laos 56,638 2 1 8 3 1 1 1999 16 6 19 3,831 Azerbaijan 2 1 1 1 1 1 1 8 7 5 1999 4,779 Mexico 4 1 1 2 1 1 1 1 1 1999 13 9 17 4,393 Philippines ater W Alleviation Growth Agriculture of Development Growth in Poverty Rural in Economic 1997) in & Irrigation GDP Development Agricultural of 1996) & and Development % Irrigation Rural Agriculture Management a (continued) Development/Reform in Alleviation Development ater Rural Development in Development Growth Development Reform/Development in as W mentions topics Resources E1 & & Assistance earY Rural of of Development Poverty of Development Resource Users ariffsT added meters/capita Recovery ableT ater ater ater CAS Country Irrigation Institutional Participation Rural Irrigation Rural Role Institutional Participation Rural Agricultural Agricultural Agricultural Institutional Participation W Irrigation Cost W W echnicalT Number Number Agriculture (value Freshwater (cubic 1 1 3 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 page) 7 7 1 next 2000 urkeyT 18 3,213 on continues 4 4 1 2000 48 anzaniaT 2,770 able(T 1 1 1 2000 Belize 3 4 2 6 2000 27,865 Argentina 3 1 11 5 3 9 2000 Uruguay 37,971 1 3 2 3 9 4 9 2000 Brazil 42,459 1 1 1 1 1 1 6 6 2000 13 unisiaT 439 6 3 1 1 7 2 1 2000 21 7 51 41,407 Cambodia 1 4 1 1 4 2 1 1 2000 15 8 32 12,746 Mozambique ater W Alleviation Growth Agriculture of Development Growth in Poverty Rural in Economic 1997) in & Irrigation GDP Development Agricultural of 1996) & and Development % Irrigation Rural Agriculture Management a (continued) Development/Reform in Alleviation Development ater Rural Development in Development Growth Development Reform/Development in as W mentions topics Resources E1 & & Assistance earY Rural of of Development Poverty of Development Resource Users ariffsT added meters/capita Recovery ableT ater ater ater CAS Country Irrigation Institutional Participation Rural Irrigation Rural Role Institutional Participation Rural Agricultural Agricultural Agricultural Institutional Participation W Irrigation Cost W W echnicalT Number Number Agriculture (value Freshwater (cubic 1 1 4 A P P E N D I X E : A N A LY S I S O F C O U N T R Y A S S I S TA N C E S T R AT E G I E S page) 1 1 1 1 next 4 4 2001 49 1,795 on Ethiopia continues 1 1 1 2 1 6 5 2001 33 2,767 Armenia able(T 1 3 1 1 4 1 1 2001 12 7 17 1,080 Morocco 6 1 1 3 2 1 1 2001 15 7 38 Chad 5,904 2 2 1 1 2 1 3 1 1 1 2001 15 10 20 12,625 Indonesia 3 3 3 1 1 3 2 1 4 1 1 2001 23 11 17 Egypt 949 2 4 2 2 2 5 1 1 3 1 1 2001 24 11 21 9,636 Bangladesh 2 1 1 8 1 3 4 3 3 8 2 1 1 1 1 2001 40 15 28 India 1,947 0 0 2000 Bosnia ater W Alleviation Growth Agriculture of Development Growth in Poverty Rural in Economic 1997) in & Irrigation GDP Development Agricultural of 1996) & and Development % Irrigation Rural Agriculture Management a (continued) Development/Reform in Alleviation Development ater Rural Development in Development Growth Development Reform/Development in as W mentions topics Resources E1 & & Assistance earY Rural of of Development Poverty of Development Resource Users ariffsT added meters/capita Recovery ableT ater ater ater CAS Country Irrigation Institutional Participation Rural Irrigation Rural Role Institutional Participation Rural Agricultural Agricultural Agricultural Institutional Participation W Irrigation Cost W W echnicalT Number Number Agriculture (value Freshwater (cubic 1 1 5 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 page) 1 1 5 1 1 1 next 2002 13 8 4 4,779 on Mexico continues 2 1 12 1 1 6 5 2002 19 4,508 able(T Mauritania 2 1 1 21 7 5 2 2002 198 Jordan 2 1 11 4 3 2002 Guyana 0 0 9 2002 Chile 32,007 1 1 1 3 1 1 8 6 2002 12 24,663 Bulgaria 1 2 2 1 1 1 1 9 7 2002 25 Albania 12,758 2 2 1 2001 41 2,318 Nigeria 1 1 2 2 2001 16 9,222 Romania ater W Alleviation Growth Agriculture of Development Growth in Poverty Rural in Economic 1997) in & Irrigation GDP Development Agricultural of 1996) & and Development % Irrigation Rural Agriculture Management a (continued) Development/Reform in Alleviation Development ater Rural Development in Development Growth Development Reform/Development in as W mentions topics Resources E1 & & Assistance earY Rural of of Development Poverty of Development Resource Users ariffsT added meters/capita Recovery ableT ater ater ater CAS Country Irrigation Institutional Participation Rural Irrigation Rural Role Institutional Participation Rural Agricultural Agricultural Agricultural Institutional Participation W Irrigation Cost W W echnicalT Number Number Agriculture (value Freshwater (cubic 1 1 6 A P P E N D I X E : A N A LY S I S O F C O U N T R Y A S S I S TA N C E S T R AT E G I E S page) 2 1 2 2 1 1 next 2003 10 7 6 Brazil 42,659 on continues 1 1 2 2 2003 16 3,831 able(T Azerbaijan 11 1 3 1 1 5 1 2 1 2 2003 18 10 11 485 Algeria 2 1 1 1 1 1 7 6 2002 emenY 15 254 1 3 1 5 3 2002 23 ietnamV 11,647 2 2 1 2 1 8 5 2002 35 5,476 Uzbekistan 1 1 5 2 3 3 6 7 2 2 2002 32 10 14 4,393 Philippines 1 1 1 3 3 8 2002 Peru 1,641 1 1 3 1 2 3 1 1 1 1 1 2002 16 11 23 1,938 Pakistan ater W Alleviation Growth Agriculture of Development Growth in Poverty Rural in Economic 1997) in & Irrigation GDP Development Agricultural of 1996) & and Development % Irrigation Rural Agriculture Management a (continued) Development/Reform in Alleviation Development ater Rural Development in Development Growth Development Reform/Development in as W mentions topics Resources E1 & & Assistance earY Rural of of Development Poverty of Development Resource Users ariffsT added meters/capita Recovery ableT ater ater ater CAS Country Irrigation Institutional Participation Rural Irrigation Rural Role Institutional Participation Rural Agricultural Agricultural Agricultural Institutional Participation W Irrigation Cost W W echnicalT Number Number Agriculture (value Freshwater (cubic 1 1 7 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 page) 2 2 1 1 1 4 4 1 1 2 2 1 1 next 2003 23 13 36 Mali 9,438 on continues 1 1 2 2 2 1 9 6 2003 29 23,094 able(T Madagascar 1 1 2 2 2003 12 3,483 Macedonia 1 1 1 3 2 2003 11 7 17 12,625 Indonesia 11 1 2 1 2 3 3 1 1 2003 15 9 21 Georgia 11,632 12 1 2 1 1 7 1 3 1 1 1 2003 20 11 42 1,795 Ethiopia 0 0 9 2003 Ecuador 26,305 3 3 2 1 1 2003 10 5 15 China 2,285 3 1 2 2 2 1 1 2003 12 7 38 Chad 5,904 ater W Alleviation Growth Agriculture of Development Growth in Poverty Rural in Economic 1997) in & Irrigation GDP Development Agricultural of 1996) & and Development % Irrigation Rural Agriculture Management a (continued) Development/Reform in Alleviation Development ater Rural Development in Development Growth Development Reform/Development in as W mentions topics Resources E1 & & Assistance earY Rural of of Development Poverty of Development Resource Users ariffsT added meters/capita Recovery ableT ater ater ater CAS Country Irrigation Institutional Participation Rural Irrigation Rural Role Institutional Participation Rural Agricultural Agricultural Agricultural Institutional Participation W Irrigation Cost W W echnicalT Number Number Agriculture (value Freshwater (cubic 1 1 8 A P P E N D I X E : A N A LY S I S O F C O U N T R Y A S S I S TA N C E S T R AT E G I E S page) 0 0 next 2004 Bolivia 38,625 on continues 1 1 1 1 1 1 2 1 9 8 2004 24 2,676 Armenia able(T 3 2 1 3 9 4 2004 11 27,865 Argentina 1 1 1 1 4 4 2003 urkeyT 13 3,213 1 1 1 3 3 2003 23 ajikistanT 13,017 2003 Lanka 1 2 2 2 2 1 1 1 1 13 9 20 2,329 Sri 1 2 1 2 1 1 1 1 1 1 2003 12 10 40 Niger 3,204 2 1 2 3 1 9 5 2003 40 Nepal 9,199 2 3 3 14 1 8 1 2003 32 7 23 12,746 Mozambique ater W Alleviation Growth Agriculture of Development Growth in Poverty Rural in Economic 1997) in & Irrigation GDP Development Agricultural of 1996) & and Development % Irrigation Rural Agriculture Management a (continued) Development/Reform in Alleviation Development ater Rural Development in Development Growth Development Reform/Development in as W mentions topics Resources E1 & & Assistance earY Rural of of Development Poverty of Development Resource Users ariffsT added meters/capita Recovery ableT ater ater ater CAS Country Irrigation Institutional Participation Rural Irrigation Rural Role Institutional Participation Rural Agricultural Agricultural Agricultural Institutional Participation W Irrigation Cost W W echnicalT Number Number Agriculture (value Freshwater (cubic 1 1 9 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 1 1 2 1 3 2 3 2 2 1 2 2004 20 11 13 unisiaT 439 3 1 1 2 2 2 2 9 1 1 6 6 2 1 2004 39 14 23 India 1,947 1 2 1 1 5 1 2 2004 13 7 17 Bosnia ater W Alleviation Growth Agriculture of Development Growth in Poverty Rural in Economic 1997) in & Irrigation GDP Development Agricultural of 1996) & and Development % Irrigation Rural Agriculture Management a (concluded) Development/Reform in Alleviation Development ater Rural Development in Development Growth Development Reform/Development in as W mentions topics Resources E1 & & Assistance earY Rural of of Development Poverty of Development Resource Users ariffsT added meters/capita Recovery ableT ater ater ater CAS Country Irrigation Institutional Participation Rural Irrigation Rural Role Institutional Participation Rural Agricultural Agricultural Agricultural Institutional Participation W Irrigation Cost W W echnicalT Number Number Agriculture (value Freshwater (cubic 1 2 0 APPENDIX F: ANALYSIS OF PROJECT COMPLETION REPORTS Two types of analysis were undertaken: tion of the objectives and components of each PAD and ICR, the study also analyzed the quanti- (i) A more detailed analysis of the outputs and tative output data (table F1). outcomes reported by each ICR, and (ii) A qualitative review of the achievements, is- Categorization of achievements, issues, sues, and problems, and a categorization of problems and lessons learned. Sixty-three ICRs lessons learned. and 8 PPARs of projects approved from FY94 were used to create a database, and the lessons Analysis of planned and actual outputs and were extracted. After classification, 408 lessons outcomes reported by each PAD and ICR in the were classified into 11 types (table F2) for portfolio. In addition to the qualitative evalua- further analysis and review. Table F1. Quantitative Comparison of Predicated and Actual Outputs from Projects Irrigation--area planned and actually achieved (ha). Drainage--area planned and actually achieved (ha). Farmers/Farm families--number of household units or number of individual farmers targeted and reported as being reached at the end of the project. People--number of individuals targeted and reached by the project. Engineering costs--planned and actual costs. Institutional development and capacity-building costs--planned and actual costs. Table F2. Typology for Classifying Frequency of Lessons Drawn from ICRs and PPARs Project Design Implementation & Procedural Community-Driven Development Cost Recovery Knowledge/Information/Skills Mobilization Technology Economics Institutional Bank Processes Targeting Disaster-Related 1 2 1 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 Table F3. Irrigated Area Planned and Achieved Project Description Irrigation Area, ha Project ID Country Project Name Approval FY PAD Planned ICR Achieved P008270 Albania Irrigation Rehabilitation 1995 73,486 100,149 P004978 Algeria Social Safety I 1996 115,000 154,000 P008277 Armenia Irrigation Rehabilitation 1995 148,000 146,800 P058468 Bangladesh Agricultural Serv. Innovation & Reform 2000 100 122 P038884 Brazil Rural Poverty--Ceara 1995 12,000 11,700 P050886 Burkina Faso Pilot Private Irrigation 270 500 P003593 China Songliao Plain ADP 1994 35,500 36,590 P003596 China Yangtze Basin Water Resources Project 1995 457,800 465,600 P063123 China Yangtze Flood Emergency Rehabilitation 1999 44,200 44,000 P007020 Dominica Irrigated Land & Watershed Management 1995 7,259 397 P063201 Dominica Hurricane Georges Emergency Recovery 1999 15,000 15,000 P008403 Estonia Agriculture 1996 60,000 76,819 P057271 Guyana El Niño Emergency Assistance Project 1999 72,430 275,000 P004964 India Water Resources Consolidation Haryana 1994 280,000 280,000 P035158 India Andrah Pradesh Irrigation Iii 1997 318,000 318,000 P010529 India Water Resources Consolidation Orissa 1996 265,358 314,953 P010476 India Water Resources Consolidation Tamil Nadu 1995 508,108 435,529 P003954 Indonesia Java Irrigation Improvement & WRM 1994 218,000 285,377 P005321 Jordan TA For Agriculture 1995 17,500 17,500 P008510 Kazakhstan Irrigation & Drainage 1996 30,000 32,200 P062682 Kyrgyz Flood Emergency 1999 39,000 34,300 P005344 Lebanon Irrigation 1994 27,000 24,300 P001522 Madagascar Irrigation II 1995 20,000 25,295 P001738 Mali Irrigation Promotion 1997 1,000 20 P007701 Mexico On-Farm & Minor Irrigation 1994 400,000 394,000 P010530 Nepal Irrigation Sector Develeopment 1998 59,600 46,250 P072996 Niger Pilot Private Irrigation 1995 2,000 1,735 P063622 Nigeria Fadama II 2004 50,000 35,000 P010453 Pakistan Balochistan Community Irrigation And Agriculture 1996 4,089 6,478 P010501 Pakistan Private Sector Groundwater 1997 1,400,000 900,000 P042442 Peru Sierra Natural Resources Management 1997 12,350 26,843 P008037 Peru Irrigation Subsector Project 1997 200,000 435,000 P03079 Philippines Agrarian Reform 1997 9,750 10,019 P034212 Sri Lanka Mahaweli Restructuring 1998 31,500 28,790 P038570 Tanzania River Basin Mgmt & Small Holder Irrigation 1997 7,000 5,059 P005721 Tunisia Agricultural Sectoe Investment 1994 9,292 9,292 P050418 Tunisia ASIL 2 1998 4,250 6,151 P009072 Turkey Privatization Of Irrigation 1998 1,497,900 220,000 P004834 Vietnam Irrigation Rehabilitation Project 1995 67,700 133,889 P043367 Yemen Taiz Water Supply Pilot 1997 596 596 Total 6,521,038 5,353,253 Average (ha/project) 163,026 133,831 1 2 2 A P P E N D I X F : A N A LY S I S O F P R O J E C T C O M P L E T I O N R E P O R T S Table F4. Actual and Planned Costs: Engineering and Infrastructure, Institutional Development and Capacity-Building Total Engineering Total Institutional & Infrastructure Development & Capacity- Costs US$ millions building Costs US$ millions Percent of Percent Project ID Country PAD Planned ICR Actual Planned PAD Planned ICR Actual of Planned P008270 Albania 31.50 35.84 114% 4.30 3.87 90% P004978 Algeria 18.70 17.00 91% na na ­ P008277 Armenia 37.43 39.67 106% 8.78 12.18 139% P058468 Bangladesh 1.80 1.56 87% na na ­ P038884 Brazil 28.44 23.32 82% 4.60 3.90 85% P050886 Burkina Faso 0.95 1.82 192% 3.01 2.59 86% P003593 China 29.40 40.10 136% 4.00 5.90 148% P003596 China 166.31 187.45 113% na 26.78 ­ P063123 China 7.47 7.76 104% na na ­ P007020 Dominica 23.18 8.15 35% 5.00 5.29 106% P063201 Dominica 13.38 11.78 88% na na ­ P008403 Estonia 10.90 9.07 83% 13.30 9.33 70% P057271 Guyana 4.14 4.95 120% P004964 India 449.60 461.57 103% 33.80 12.80 38% P035158 India 366.00 390.00 107% 2.47 0.97 39% P010529 India 223.40 257.96 115% 39.30 14.09 36% P010476 India 240.90 218.10 91% 17.70 21.80 123% P003954 Indonesia na na ­ na na ­ P005321 Jordan 5.15 na ­ 1.04 na ­ P008510 Kazakhstan 108.14 90.19 83% 5.93 4.46 75% P062682 Kyrgyz 13.40 12.09 90% 0.70 0.60 86% P005344 Lebanon 61.00 67.02 110% 16.18 11.37 70% P001522 Madagascar 13.80 10.96 79% 7.40 9.15 124% P001738 Mali 0.57 0.22 39% 2.34 0.86 37% P007701 Mexico 568.80 396.70 70% 39.60 21.20 54% P010530 Nepal 72.42 73.86 102% 16.02 16.49 103% P072996 Niger 2.60 2.24 86% 2.90 3.51 121% P063622 Nigeria 74.80 64.20 86% 16.80 40.00 238% P010453 Pakistan 17.10 19.50 114% 17.00 14.70 86% P010501 Pakistan 97.50 25.31 26% 7.30 4.37 60% P042442 Peru 67.20 85.10 127% 31.00 5.10 16% P008037 Peru 140.50 79.90 57% 6.00 13.00 217% P03079 Philippines 18.60 19.00 102% 7.50 6.10 81% P034212 Sri Lanka 79.20 75.80 96% 28.90 24.70 85% P038570 Tanzania 15.87 14.49 91% 11.57 10.32 89% P005721 Tunisia 105.20 110.40 105% 3.60 1.50 42% P050418 Tunisia 22.50 24.30 108% 9.10 9.60 105% P009072 Turkey 50.05 40.37 81% 6.50 4.16 64% P004834 Vietnam 108.10 110.90 103% 4.60 7.00 152% P043367 Yemen 0.40 0.75 188% 0.30 0.30 100% Total 3,296.40 3,039.40 378.54 327.99 Average per project 84.52 79.98 95% 9.96 8.41 84% Note: na = not available. 1 2 3 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 Table F5. Engineering and Infrastructure: Average Costs Per Hectare Total Engineering Costs Total Engineering (Excluding expensive projects) Costs US$/ha US$/ha Percent of Percent Project ID Country PAD Planned ICR Actual Planned PAD Planned ICR Actual of Planned P008270 Albania 429 358 83% 429 358 83% P004978 Algeria 163 110 68% 163 110 68% P008277 Armenia 253 270 107% 253 270 107% P058468 Bangladesh 18,000 12,787 71% P038884 Brazil 2,370 1,993 84% 2,370 1,993 84% P050886 Burkina Faso 3,519 3,640 103% 3,519 3,640 103% P003593 China 828 1,096 132% 828 1,096 132% P003596 China 363 403 111% 363 403 111% P063123 China 169 176 104% 169 176 104% P007020 Dominica 3,193 3,193 P063201 Dominica 892 785 88% 892 785 88% P008403 Estonia 182 118 65% 182 118 65% P057271 Guyana 57 18 32% 57 18 32% P004964 India 1,606 1,648 103% 1,606 1,648 103% P035158 India 1,151 1,226 107% 1,151 1,226 107% P010529 India 842 819 97% 842 819 97% P010476 India 474 501 106% 474 501 106% P003954 Indonesia P005321 Jordan 294 -- 0% 294 -- P008510 Kazakhstan 3,605 2,801 78% 3,605 2,801 78% P062682 Kyrgyz 344 352 103% 344 352 103% P005344 Lebanon 2,259 2,758 122% 2,259 2,758 122% P001522 Madagascar 690 433 63% 690 433 63% P001738 Mali 570 10,774 1890% P007701 Mexico 1,422 1,007 71% 1,422 1,007 71% P010530 Nepal 1,215 1,597 131% 1,215 1,597 131% P072996 Niger 1,300 1,291 99% 1,300 1,291 99% P063622 Nigeria 1,496 1,834 123% 1,496 1,834 123% P010453 Pakistan 4,182 3,010 72% 4,182 3,010 72% P010501 Pakistan 70 28 40% 70 28 40% P042442 Peru 5,441 3,170 58% 5,441 3,170 58% P008037 Peru 703 184 26% 703 184 26% P03079 Philippines 1,908 1,896 99% 1,908 1,896 99% P034212 Sri Lanka 2,514 2,633 105% 2,514 2,633 105% P038570 Tanzania 2,267 2,864 126% 2,267 2,864 126% P005721 Tunisia 11,322 11,881 105% P050418 Tunisia 5,294 3,951 75% 5,294 3,951 75% P009072 Turkey 33 184 549% 33 184 549% P004834 Vietnam 1,597 828 52% 1,597 828 52% P043367 Yemen 671 1,258 188% 671 1,258 188% Average per project 2,146 2,123 99% 1,494 1,293 87% Note: na = not available. 1 2 4 A P P E N D I X F : A N A LY S I S O F P R O J E C T C O M P L E T I O N R E P O R T S Table F6. Institutional Development and Capacity-Building: Average Costs per Hectare Total Institutional Development & Total Institutional Development & Capacity-building Costs Capacity-buildingCosts US$/ha Excluding expensive projects) US$/ha Percent of Percent Project ID Country PAD Planned ICR Actual Planned PAD Planned ICR Actual of Planned P008270 Albania 59 39 66% 59 39 66% P004978 Algeria -- -- 0% -- -- 0% P008277 Armenia 59 83 140% 59 83 140% P058468 Bangladesh -- -- 0% -- -- -- P038884 Brazil 383 333 87% 383 333 87% P050886 Burkina Faso 11,148 5,180 46% 11,148 5,180 46% P003593 China 113 161 143% 113 161 143% P003596 China -- 58 0% -- 58 -- P063123 China -- -- 0% -- -- -- P007020 Dominica 689 13,325 1935% -- -- -- P063201 Dominica -- -- 0% -- -- P008403 Estonia 222 121 55% 222 121 55% P057271 Guyana -- -- 0% -- -- 0% P004964 India 121 46 38% 121 46 38% P035158 India 8 3 39% 8 3 39% P010529 India 148 45 30% 148 45 30% P010476 India 35 50 144% 35 50 144% P003954 Indonesia -- -- 0% -- -- 0% P005321 Jordan -- -- 0% -- -- 0% P008510 Kazakhstan 198 139 70% 198 139 70% P062682 Kyrgyz 18 17 97% 18 17 97% P005344 Lebanon 599 468 78% 599 468 78% P001522 Madagascar 370 362 98% 370 362 98% P001738 Mali 2,340 42,116 1800% -- -- -- P007701 Mexico 99 54 54% 99 54 54% P010530 Nepal 269 357 133% 269 357 133% P072996 Niger 1,450 2,023 140% 1,450 2,023 140% P063622 Nigeria 336 1,143 340% 336 1,143 340% P010453 Pakistan 4,157 2,269 55% 4,157 2,269 55% P010501 Pakistan 5 5 93% 5 5 93% P042442 Peru 2,510 190 8% 2,510 190 8% P008037 Peru 30 30 100% 30 30 100% P03079 Philippines 769 609 79% 769 609 79% P034212 Sri Lanka 917 858 94% 917 858 94% P038570 Tanzania 1,653 2,040 123% 1,653 2,040 123% P005721 Tunisia 387 161 42% 387 161 42% P050418 Tunisia 2,141 1,561 73% 2,141 1,561 73% P009072 Turkey 4 19 436% 4 19 436% P004834 Vietnam 68 52 77% 68 52 77% P043367 Yemen 503 503 100% 503 503 100% Average per project 795 1,860 234% 720 487 68% Note: na = not available. 1 2 5 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 Table F7. Economic Rates of Return Project ID Country Project Approval FY Exit FY PAD Predicted ICR Actual P008270 Albania Irrigation Rehabilitation 1995 2000 17.0 38.0 P008277 Armenia Irrigation Rehabilitation 1995 2001 50.0 29.0 P003593 China Songliao Plain 1994 2003 26.0 40.0 P003596 China Yangtze Basin Water Resources 1995 2006 24.0 26.0 P008403 Estonia Agriculture 1996 2002 16.0 14.6 P057271 Guyana El Niño Emergency Assistance 1999 2002 19.0 16.0 P009964 India Haryana WRCP 1994 2002 18.0 14.0 P010476 India Tamil Nadu WRCP 1995 2005 13.3 11.9 P010529 India Orissa WRCP 1996 2006 16.7 13.9 P035158 India Andrah Pradesh Irrigation III 1997 2005 24.0 14.7 P003954 Indonesia Java Irrigation Improvement 1994 2003 16.6 15.5 P004008 Indonesia Nusa Tenggara Develeopment 1996 2004 17.0 14.5 P008510 Kazakhstan Irrigation & Drainage 1996 2005 27.0 32.0 P005344 Lebanon Irrigation 1994 2004 19.0 23.0 P001522 Madagascar Irrigation II 1995 2005 27.0 12.0 P007701 Mexico On-farm & Minor Irrigation 1994 2003 19.0 28.0 P010530 Nepal Irrigation Sector Project 1998 2004 15.3 9.7 P001994 Niger Pilot Private Irrigation 1995 2002 na 67.0 P010482 Pakistan Balochistan Community Irrigation 1996 2002 14.0 11.4 P010501 Pakistan Private Sector Groundwater 1997 2002 23.9 29.0 P008037 Peru Irrigation Subsector Project 1997 2005 38.7 24.1 P042442 Peru Sierra Natural Resources 1997 2004 12.0 12.3 P037079 Philippines Agrarian Reform 1997 2004 22.0 26.0 P034212 Sri Lanka Mahaweli Restructuring 1998 2005 14.0 15.0 P058070 Sri Lanka NE Irrigation 1999 2006 14.0 10.5 P038570 Tanzania RBM & Smallholder Irrigation 1997 2004 12.1 10.4 P005721 Tunisia Agricultural Sector Investment 1994 2001 15.0 19.0 P0075736 Tunisia Natural Resources Management 1997 2005 13.7 18.5 P050418 Tunisia ASIL 2 1998 2003 18.2 11.4 P009072 Turkey Privatization of Irrigation 1998 2004 72.0 54.0 P004834 Vietnam Irrigation Rehabilitation Project 1995 2003 17.0 20.5 P043367 Yemen Taiz Water Supply Pilot 1997 2002 na 26.0 Simple Average 21.7% 22.1% Weighted Average by Area 24.3% 22.9% Weighted Average by Engineering Costs 21.10% 19.50% Note: na = not available. 1 2 6 A P P E N D I X F : A N A LY S I S O F P R O J E C T C O M P L E T I O N R E P O R T S Table F8. Number of People Benefiting from AWM Projects Project ID Country Project Description PAD Projected ICR Actual P062682 India Tamil Nadu WRCP 700,000 636,500 P003596 China Yangtze Flood Emergency Rehabilitation 442,000 P003954 Vietnam Irrigation Rehabilitation Project 338,388 P010501 China Songliao Plain 240,000 240,800 Indonesia Nusa Tenggara Development 40,000 166,000 P034212 Pakistan Private Sector Groundwater 234,000 150,000 P057271 Peru Irrigation Subsector Project 50,000 135,000 P008277 Albania Irrigation Rehabilitation 56,800 110,300 P003218 Philippines Agrarian Reform na 73,000 P051171 Peru Pe Sierra Natural Resources Management 31,061 31,061 P035717 Lebanon Lb-Irrigation 175,943 26,528 P058877 Brazil Rural Poverty--Ceara na 11,126 P009964 Estonia Agriculture 5,000 10,515 P010476 Kyrgyz Flood Emergency 10,000 P034891 Pakistan Balochistan Community Irrigation 8,922 P040521 Kazakhstan Irrigation & Drainage 6,622 P056595 Tanzania RBM and Smallholder irrigation 7,000 5,317 P055974 Niger Pilot Private Irrigation 4,000 3,469 P051386 Yemen Taiz Water Supply Pilot 1,949 P059055 Burkina Faso Pilot Private Irrigation 900 1,600 P010461 Dominica Irrigated Land & Watershed 1,500 1,300 P008270 Armenia Irrigation Rehab 100,000 P038695 Madagascar Irrigation II 40,000 P009122 Sri Lanka Northeast Irrigated Agriculture Project 24,000 Total 1,710,204 2,410,397 Average per project 106,888 114,781 Note: na = not available. 1 2 7 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 Table F9. Typology of Lessons from Completed Projects (in percentages) Class Class Typology EAP LAC ECA MNA SAR AFR Total Total Project Design 9.0 2.5 4.5 6.0 2.5 3.5 28 Designing & Implementation & Managing Procedures 4.9 4.9 4.1 4.1 2.0 2.0 22 57 Projects Bank Processes 1.4 0.7 1.2 1.0 1.7 1.0 7 Community-Driven Development 3.9 2.5 2.5 0.9 1.7 0.5 12 Institutional 2.9 2.2 2.7 1.7 1.0 0.5 11 Development Enhancing Technology 1.4 0.6 0.6 1.0 1.4 1.0 6 Lessons Cost Recovery/O&M 1.5 0.5 1.5 0.5 4 43 about Disaster-Related 0.5 2.0 0.5 3 Professional Targeting 1.2 1.0 0.4 0.2 0.2 3 Knowledge Knowledge and Management Technologies Mobilization 1.4 0.2 1.2 0.2 3 Economics 0.3 0.3 0.5 1 Total 28 17 17 15 13 9 100 1 2 8 ENDNOTES Chapter 1 3. A number of bilateral donors have given 1. FAO 2002. Cereals and cereal products ac- grants and loans for agricultural water man- counted for 58 percent of food consumed by agement, particularly France, Japan, the weight. Netherlands, the United Kingdom, and the 2. Abdel-Dayam et al. 2004. United States, but the amount is not easily de- 3. This was because investment in drainage was termined. deferred until poor drainage became a prob- 4. "CDD is broadly defined as giving control of lem--a rational economic decision. In most decisions and resources to community groups cases in South Asia, for example, drainage and local governments. CDD programs op- only became an issue 40­60 years after sur- erate on the principles of local empower- face water canal irrigation was introduced. ment, participatory governance, demand- 4. IEG 2004a. With total construction costs of responsiveness, administrative autonomy, $750 per hectare and maintenance costs of greater downward accountability, and en- $10 per hectare per year, the payback period hanced local capacity" (CDD Web site, World is only three to four years. Bank). 5. Lipton, Litchfield, and Faures 2003. 5. Other client services include: technical as- 6. IEG 1992. sistance and aid coordination; country pro- 7. IEG 1994. This report covers Bank activities gram support; client training; knowledge in the period 1950­93. management; sector strategy; quality assur- 8. World Bank 1997a. ance; external partnerships and outreach; 9. IEG 1999 and 2000. business development; network, council, and 10. IEG 2002a. sector board activities and research. 11. World Bank 2004f. 6. The Bank budget for lending preparation was $99.7 million in 2001; by 2005 this had in- Chapter 2 creased to $146.3 million. The U.S. GDP de- 1. Costs are expressed in constant 2002 U.S. flator was 88.4 in 1993 and 112.1 in 2005 dollars. Full details of how the amount lent (year 2000 = 100). for agricultural water was determined are in 7. There were 256 loans in 1993 and 223 in appendix A. 2000; average direct preparation costs per 2. Camdessus Panel 2003. The Camdessus re- project fell from $349,000 to $288,000 over the port notes: "There are no reliable estimates same period. of global investment in irrigation. Large pub- 8. For example, during the period FY01­FY05, lic sector schemes are funded mainly by local typical public-sector governance/rule-of-law public agencies and international aid, with projects took 11.5 months to prepare and smaller schemes and on-farm investments 3.7 years to implement. Conversely, rural mainly financed by farmers, informal credit projects took 18 months to prepare and 6.8 and banks." years to implement. Water supply and sani- 1 2 9 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 tation projects take longer still: 25 months for percent, industry 3.9 percent, services 3.7 preparation and 7.3 years to implement. percent, and export of goods and services 8.2 9. The Bank offers two basic types of lending in- percent. In the period 1980­90, agricultural struments: investment loans (which typically value added was 3.4 percent (World Bank have a long-term focus of 5 to 10 years and 2000a). finance goods, works, and services) and de- 16. IRRI 2005; from data on population, labor velopment policy loans (which typically have force, and wages. a shorter time horizon of 1 to 3 years and pro- 17. United Nations 2005. vide quick-disbursing external financing to 18. Picciotto, van Wicklin, and Rice 2001. support policy and institutional reforms). For 19. World Commission on Dams 2000. an overview of these two lending instru- 20. IEG 1995. ments, see World Bank 2000b and 2004e. 21. IEG 2002a. 10. The increase in staff dealing with "other client 22. This analysis is based on 129 cases covering services" was also significant between 1997 47 countries in which the Bank had water op- and 2000: external affairs staff increased from erations. 42 to 124; legal staff increased from 91 to 23. MNA is the only region in which agricultural 119; and information technology staff in- GDP increased in the 1990s. More detailed re- creased from 485 to 699. gional analysis is given in appendix D. 11. The 2002 rural strategy states: "another 24. The typology of borrowers was based on the staffing issue is how long and to what extent number of operations per borrowers. A high the Bank should maintain expertise in the correlation was found among the number of technical aspects of agriculture, irrigation operations, consistency of operations per and natural resources management. The Bank borrower (number of years with at least one will keep this expertise at "core" level [cur- operation), and amount committed on irri- rently 16 staff] ... and use FAO/CP to provide gation and drainage per borrower. specialized expertise ... " (World Bank 2002c, 25. For agriculture and irrigation in the period p. 99). 1988­91, the Bank covered 53 percent of 12. Survey by the Bank's Human Resources De- total project costs and 10 percent of Bank fi- partment in 2005. nancing was from IBRD; in the period 13. Its share of project preparation costs has 1995­2000, the Bank covered 45 percent of doubled since 1994 and is now 21 percent. total costs and 76 percent of Bank financing 14. The combined share of adjustment lending was from IBRD. in total IBRD and IDA lending has exceeded 26. World Bank 2002d. In FY98, 54 percent of all one-third since FY98, reaching 53 percent in projects in the country portfolio were classi- FY99, peaking during the East Asian crisis fied as "at risk," and water projects featured and again in FY02 with the Turkey crisis, and prominently in that list. dropping back to a 33 percent share in mid- 27. The assessment separated policy reform from FY04. The share of adjustment lending on a institutional reform, although sometimes the three-year rolling average peaked at 42 per- two are hard to treat separately. cent, and remained flat at about 39 percent 28. A better practice was the Albanian Irrigation in the ensuing years. The share for IBRD and Drainage Rehabilitation Projects I (1995) alone has exceeded 37 percent since FY98, and II (1999). Both projects received strong reaching 63 percent in FY99, and receding to policy support from the government. With 33 percent in mid-FY04. IDA's share since every change in the institutional setting the FY98 has been in the range of 15 to 27 per- government changed or amended legislation cent. and action was taken. 15. Value added to economic growth in the pe- 29. World Bank 2003c. riod 1990­99 was as follows: agriculture 2.2 30. Berkoff 2003. 1 3 0 E N D N O T E S 31. World Bank 2005e. 1994 Tunisia Agricultural Sector Investment 32. IEG 2005c. Project cost $11,321 per hectare for dams and conveyance, which also provided water Chapter 3 supplies in addition to modernization and 1. Draft paper on Ethiopia, Country Water Re- extension or irrigation. sources Assistance Strategy, World Bank. 7. Kikuchi and Inocencio (forthcoming) ana- 2. IEG 2005a. lyzed the 314 World Bank projects imple- 3. India Planning Commission. mented during the period 1967­2003, with 4. The rural nonfarm sector includes small-scale more than 95 percent of them designed be- food-processing plants, machinery repair fore 1990; the mode peaked in 1981. shops, and increasingly modern and tech- 8. This is based on analysis of 2,908 projects nology-intensive industries. Bank-wide and includes 550 rural sector proj- 5. Indian agriculture accounts for about 25 per- ects, of which 161 are classified by IEG as cent of GDP. Instead of growing at its his- agricultural water. toric rate of about 4 percent, agricultural 9. For example, the Bulgaria and Jordan Agri- growth slowed to 2 percent in the late 1990s. cultural Sector Loans; the Bolivia and Kenya 6. IEG 2005b. El Niño Projects; the Dominican Republic's Hurricane George Recovery Project; and the Chapter 4 Tajikistan, Kyrgyzstan, Tajikistan, and Yemen 1. This is discussed in detail with respect to fig- flood emergency projects. ures 2.7 and 2.8 and in the discussion of re- 10. Regression of entry on exit ERRs yields: sult chains in chapter 5. ERR (entry) = 0.852 (exit ERR) + 4.254. The 2. Assuming that a tenth to a quarter of 12 mil- same exercise on all AWM that were designed lion households in the nondedicated projects before FY94 and completed later yields: participate in agricultural water-management ERR (entry) = 0.427 (exit ERR) + 15.472. In activities. the earlier cohort designed between 1982 3. Such as in Yemen (2004), Tunisia (2000), Peru and 1993, 78 percent of ERRs predicted at ap- (1997), Lebanon (1994), and Mexico (2004). praisal were not achieved. 4. This figure is taken only from PADs where 11. The world price for cotton was $2,079 a ton hectares are reported; because we think there in 1994 and fell to $922 a ton in 2001. is more reporting of this in the larger projects, 12. The example is taken from the Tieshan sub- this may bias the true average upward. project, which improved 14,272 hectares of 5. Therefore, the 1998 Bolivia El Niño Emer- early-rice production and converted 1,707 gency Project reported only the length of hectares of mixed rain-fed grains to 1,446 flood protection provided but not the area hectares of irrigated early-season rice. benefited; the 1999 Turkey Flood Recovery 13. The new PADs were introduced in 1998 and Project only reported that 31 villages bene- Section C3 requires a summary statement of fited. Others, such as the 1994 Indonesia benefits and target population. Dam Safety Project only reported 94 dams and 14. M&E systems typically have four components: dam safety institutions were established. (1) monitoring of financial inputs and phys- 6. Two pilot projects were extremely costly: the ical outputs during implementation; (2) 1997 Mali Pilot Private Irrigation Promotion process monitoring; (3) post-implementa- Project cost $12,787 per hectare for small-scale tion monitoring, including inspection of com- groundwater irrigation; and the 2000 pleted works, physical and financial audits; Bangladesh Agricultural Services and Inno- and (4) outcome and impact evaluation. vation Project cost $10,774 per hectare for 15. To facilitate this more-detailed analysis of demonstration of high-tech irrigation. The project design, a stratified random sample 1 3 1 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 of 80 projects was taken from the portfolio to 6. Beside the difficulties at the political level, the give representative subsamples of dedicated poverty focus is still weak, despite a certain and nondedicated projects. Because of the degree of evident awareness there is not a smaller sample size and two data subsets, clear poverty strategy with respect to WUA the time comparison was reduced to two support and processes. time periods, FY94­FY98 and FY99­FY04, as 7. Thurman 2002. compared with the analysis in chapter 2. 8. This means that if the preproject average 16. The detailed evaluation is presented in ap- cost recovery is 20 percent, the average proj- pendix D. It is based on 12 questions on the ect plans to set up a system to recover more design quality of ICR M&E systems, and an than 85 percent of the O&M cost. additional three questions for completed 9. For example, in the 1994 India Haryana Water projects. Resources Consolidation Project, in the Mem- 17. A t-test of the difference between dedicated orandum of Understanding with WUAs, there and nondedicated projects on this measure is nothing mentioned about fees other than of M&E quality was significant at the 99 per- a reassertion of the right of the government cent level. of Haryana (GOH) in "levying irrigation fees 18. Ezemenari, Owens, and Soto 2000. as fixed by GOH." However, an SAR annex, not 19. An ordered probit analysis of the 18 variables the main report, mentions that, "as further in- associated with IEG's M&E analysis is formation relating to operating costs be- reported in appendix C. The four variables comes available [adjustments in rates] will reported above are significant with be made, particularly to the extent that 0.0370.003. farmers undertake operation and mainte- nance of facilities below the distributary Chapter 5 level." This statement did not find its way into 1. From Van Koppen and Safilios-Rothschild the Memorandum of Understanding with 2005: food-short months; school attendance communities. So, in this case, for WUA mem- changes; wage employment; income (from bers, there could not be any evident rela- existing household surveys); levels of in- tionship between contributions to O&M debtedness; changes in key assets; inclusion through the WUA and reduced fees to the of those identified as marginalized in com- government. munity associations (but this is a means rather 10. A notable exception is Banyopadhyay, Shyam- than an end); changes in cash crop produc- sundar, and Xie 2005. tion; use of health services (and in some 11. Hussain and Wijerathna (2004) identify four cases exemptions from payments); receipt main charge methods in Asia: (1) cultivated of funds from relatives; input purchase; pros- area by crop type (much of South Asia); (2) titution; home improvements; and, more area and output by crop type (Vietnam); (3) generally, participatory wealth ranking and multiple criteria: area, productivity, location, wealth perception changes. In Pakistan, Je- service level, and capacity to pay (Indonesia-- hangir et al. (2004) found the presence of transferred systems); (4) layered criteria: vol- flush toilets and the number of people per umetric at the main canal and cropped area room to be correlated with poverty level. at lower levels (China). 2. This is based on an open-ended survey of 18 12. 1994 Jordan Agricultural Sector Adjustment Bank managers and senior staff responsible Operation. for AWM. 13. This is the essence of the warabundi system, 3. For example, the Egypt Matrouh Natural Re- spreading water thinly and thereby forcing sources Management Project (IEG 2003). farmers to make individual intensive/extensive 4. World Bank 1996, annex II. land/water ratio choices, none of which can 5. World Bank 2004f; IEG 1995. ever cover 100 percent of the land up to the 1 3 2 E N D N O T E S crops' maximum (or even optimal) water keting. The difference was significant at the need (duty). 10 percent level. 14. In one case in the early 1990s, Bank staff showed that a proposed Tanks Improvement Appendix A Project in Tamil Nadu, India, would be un- 1. "CDD is broadly defined as giving control of economical because the net water saved decisions and resources to community groups would not be sufficient to cover the high and local governments. CDD programs op- cost of lining watercourses. A high proportion erate on the principles of local empower- of water lost to the ground was recovered by ment, participatory governance, demand- wells or tanks lower in the system (some responsiveness, administrative autonomy, tanks were actually managed as percolation greater downward accountability, and en- tanks) and nearly all incremental water in- hanced local capacity" (CDD Web site, World tercepted by any tank whose holding capac- Bank). ity was enlarged would simply be at the 2. IEG 2004b. expense of another tank lower in the basin. 15. These measures include a judicious mix of the Appendix B following: some lining, low-pressure pipes, 1. If the additional 153 CDD-type projects groundwater recharge facilities, conjunctive that may include a minor agricultural use, improvements to on-farm surface sys- water component were added to overall tems, sprinkler systems, microirrigation sys- project numbers, the number of projects re- tems, land leveling, reduced or differently ported for LAC and AFR would increase dra- timed tillage, some deep plowing, soil fertil- mati-cally, although the Africa region would ity improvement and balanced fertilization, then account for almost the 20 percent of the mulching, seed improvement, improved total number of operations in both time pe- planting techniques, changed cropping pat- riods. terns, shelterbelts to reduce evaporation, 2. Note, though, that the largest part of the de- volumetric pricing, and community man- veloping world's irrigation area (70 percent) agement. The key, however, is the comple- is in Asia. mentary package. 3. Most lending operations were recoded back 16. Of the 40 dedicated irrigation projects re- to fiscal year 1990 because it was chosen as viewed, the appraisal documents showed the base year for measuring progress toward that for 57 percent of them there was a link the Millennium Development Goals. This to extension, for 19 percent a link to mar- should allow us to be quite confident about keting, and for another 19 percent a link to these figures. credit. Nondedicated projects scored better: 58 percent showed links to extension, 39 Appendix E percent to credit, and 39 percent also to mar- 1. IEG 2004b. 1 3 3 BIBLIOGRAPHY Abdel-Dayam, S., J. Hoevenaars, P. Mollinga, Bell, C., P.B.R. Hazell, and R. Slade. 1982. Project W. Scheumann, R. Slootweg, and F. van Steen- Evaluation in Regional Perspective: A Study bergen. 2004. Reclaiming Drainage: Toward of an Irrigation Project in Northwest Malaysia. an Integrated Approach. Agriculture and Rural Baltimore, Maryland: Johns Hopkins University Development Department, Report 1, World Press. Bank, Washington, DC. Berkoff, J. 2003. "A Critical Look at the Irrigation Alston, J., G.W. Norton, and P.G. Pardey. 1995. Sci- Sector." Paper presented at the World Bank ence Under Scarcity: Principles and Practice conference, "Water Week 2003: Water and De- for Agricultural Research Evaluation and Pri- velopment," Washington, DC, March 4­6. ority Setting. Ithaca, New York: Cornell Uni- Besley, T., R. Burgess, and B. Esteve-Volart. 2005. versity Press. "Operationalizing Pro-poor Growth: A Coun- Aw, D., and G. Diemer. 2005. Making a Large Ir- try Case Study on India." Paper presented at the rigation Scheme Work: A Case Study from U.K. Department for International Develop- Mali. Directions in Development Series. Wash- ment, London School of Economics, and World ington, DC: World Bank. Bank workshop, "Operationalizing Pro-poor Bandyopadhyay, S., P. Shyamsundar, and M. Xie. Growth in India," New Delhi, India, June 9. 2005. "Understanding the Economic Impacts http://siteresources.worldbank.org/INTINDIA/ of Irrigation Management Transfer in the Philip- Resources/oppgindia.pdf pines: A Case Study of the Magat River Inte- Bhatia, R., and M. Scatasta. 2003. "Study on the grated Irrigation System." East Asia Rural Multiplier Effects of Dams: A Review of Past Development and Natural Resources Depart- Studies and Methodological Issues." Unpub- ment, World Bank, Washington, DC. lished. Barker, R., and F. Molle. 2004. Evolution of Irri- Bhatia, R., M. Scatasta, and R. Cestti. 2003. "Study gation in South and Southeast Asia. Compre- on the Multiplier Effects of Dams: Methodol- hensive Assessment of Water Management in ogy Issues and Preliminary Results." Paper pre- Agriculture. Research Report Series No. 5. sented at the Third World Water Forum, Kyoto, http://www.iwmi.cgiar.org/assessment/FILES/ Japan, March 16­23. pdf/publications/ResearchReports/CARR5.pdf. Bhattarai, M., and A. Narayanmoorthy. 2003a. "Im- Barker, R., C. Ringler, N.M. Tien, and M.W. Roseg- pact of Irrigation on Rural Poverty in India: An rant. 2004. Macro Policies and Investment Aggregate Panel-Data Analysis." Water Policy 5 Priorities for Irrigated Agriculture in (5): 443­458. Vietnam. Comprehensive Assessment of Water ------. 2003b. "Impact of Irrigation on Agricul- Management in Agriculture. Research Report tural Growth and Poverty Alleviation: Macro Series No. 6. http://www.iwmi.cgiar.org/ Level Analysis in India." International Water assessment/files_new/publications/CA%20 Management Institute, Water Policy Research Research%20Reports/CARR6.pdf. Paper 12. http://www.iwmi.cgiar.org/assessment/ BBS (Bangladesh Bureau of Statistics). 1999. Cen- files_new/publications/Workshop%20Papers/ sus of Agriculture, 1996. National Series, Vol 1. TATA_2003_Bhattarai.pdf. Statistics Division, Ministry of Planning, Dhaka, Bhattarai, M., A. Narayanmoorthy, and R. Barker. Bangladesh. 2002. "Irrigation Impact on Growth, Returns 1 3 5 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 and Performance of Agriculture in India: State Century." Irrigation and Drainage Systems Level Panel Data Analysis for 1970­94." Un- 11 (2): 83­101. published, International Water Management Cestti, R., and R.P.S. Malik. 2004. "Indirect Eco- Institute, Colombo, Sri Lanka. nomic Impacts of Dams." Paper presented at Bhattarai, M., R. Sakthivadivel, and I. Hussain. the "International Workshop on Impacts of 2002. "Irrigation Impacts on Income Inequal- Large Dams," Istanbul, Turkey, October 25­27. ity and Poverty Alleviation: Policy Issues and Op- Datt, G., and M. Ravallion. 1997. "Why Have Some tions for Improved Management of Irrigation Indian States Performed Better Than Others at Systems." Working Paper 39, International Reducing Rural Poverty?" Food Consumption Water Management, Colombo, Sri Lanka. and Nutrition Division Discussion Paper No. 26, Binswanger, H., and S. Aiyar. 2003, "Scaling Up International Food Policy Research Institute, Community-Driven Development." Policy Re- Washington, DC. search Working Paper 3039, World Bank, Wash- De Haan, A., and M. Lipton. 1998. "Poverty in ington, DC. Emerging Asia: Progress and Setback." Uni- Binswanger, H., and J. Quizon. 1986. "What Can versity of Sussex, U.K. Agriculture Do for the Poorest Rural Groups?" Dhawan, B.D. 1988. Irrigation in India's Agri- Discussion Paper No. 57, Research Unit, Agri- cultural Development: Productivity, Stabil- culture and Rural Development Department, ity, Equity. Delhi: Institute of Economic World Bank, Washington, DC. Growth; London: Sage Publications. Brabben, T., C. Angood, J. Skutsch, and L. Smith. ------. 1999. Studies in Indian Irrigation. New 2004. "Irrigation Can Sustain Rural Livelihoods: Delhi: Common Wealth Publications. Evidence from Bangladesh and Nepal." U.K. De- Dhawan, B.D., and H.S. Datta. 1992. "Impact of Ir- partment for International Development, Im- rigation on Multiple Cropping." Economic and perial College, London. Political Weekly, March 28, pp. A15­A18. Bruinsma, J., ed. 2003. World Agriculture: To- Dinar, A., and A. Subramanian, eds. 1997. Water wards 2015/2030, An FAO Perspective. Lon- Pricing Experiences: An International Per- don: Earthscan Publications. spective. World Bank Technical Paper No. 386. Bruns, B. 2004. "Irrigation Water Rights: Options World Bank, Washington, DC. for Pro-poor Reform." Paper presented at the Dorosh, P.A., S. Haggblade, and P. Hazell. 1998. Regional Workshop and Policy Roundtable "The Impact of Alternative Agricultural Strate- Discussion on "Pro-poor Intervention Strate- gies on Growth of the Rural Nonfarm Econ- gies in Irrigated Agriculture in Asia." omy." Paper presented at the International http://www.adb.org/water/actions/REG/ Food Policy Research Institute Conference on ABSTRACT-irrigation-water-rights.asp#notes "Strategies for Stimulating Growth of the Rural Bucknell, J., I. Klytchnikova, J. Lampetti, M. Lun- Nonfarm Economy in Developing Countries," dell, M. Scatasta, and M. Thurman. 2003. "Irri- Arlie House, Virginia, May 17­21. gation in Central Asia: Social, Economic and Elbers, C., P. Lanjouw, J. Mistiaen, B. Ozler, and Environmental Considerations." Europe and K. Simler. 2004. "On the Unequal Inequality of Central Asia Department, World Bank, Wash- Poor Communities." World Bank Economic Re- ington, DC. view 18 (3): 401­421. Camdessus Panel. 2003. Financing Water For All: Estache, A., Q. Wodon, and V. Foster. 2002. "Ac- Report of the World Panel on Financing Water counting for Poverty in Infrastructure Reform: Infrastructure. World Water Council and the Learning from Latin America's Experience." Global Water Partnership. http://www.financing World Bank Institute Development Study, World waterforall.org. Bank, Washington, DC. Carruthers, I., M.W. Rosegrant, and D. Seckler. Ezemenari, K., J. Owens, and G.R. Soto. 2000. 1997. "Irrigation and Food Security in the 21st "Progress on Impact Evaluation Plans in Bank 1 3 6 B I B L I O G R A P H Y Projects: Comparison of Fiscal Years international workshop on "African Water Laws: 1998­2000." Unpublished, Poverty Group, Plural Legislative Frameworks for Rural Water Poverty Reduction and Economic Management Management in Africa," Johannesburg, South Network, World Bank, Washington, DC. Africa, January 26­28. Facon, T. 2002. "Downstream of Irrigation Water Findley, C. 2005. "China Country Assistance Eval- Pricing: The Infrastructure Design and Oper- uation: Agricultural Sector." Background paper ational Management Considerations." Irriga- prepared for IEG, World Bank, Washington, tion Water Policies Conference, Agadir, June DC. 15­17. Frederiksen, H., J. Berkoff, and W. Barber. 1993. Fan, S., and P. Hazell. 2000. "Should Developing Water Resources Management in Asia. Main Countries Invest More in Less-Favored Lands? Report. World Bank Technical Paper No. 212. An Empirical Analysis of Rural India." Eco- Washington, DC: World Bank. nomic and Political Weekly 35 (17): Fung, A., and E.O. Wright. 2003. Deepening 1455­1464. Democracy: Institutional Innovations in Fan, S., P. Hazell, and S. Thorat. 1999. "Linkages Empowered Participatory Governance. Lon- between Government Spending, Growth, and don: Verso. http://www.ssc.wisc.edu/~wright/ Poverty in Rural India." Research Report No. Deepening.pdf. 110. International Food Policy Research Insti- Gardner, B.D. 1983. "Water Pricing and Rent Seek- tute, Washington, DC. ing in California Agriculture." In Water Rights, ------. 2004. "Investment, Subsidies and Pro- Scarce Resources Allocation, Bureaucracy, poor Growth, and Rural Poverty in India." In and the Environment, ed. T.L. Anderson, Doward, et al., Institutions and Policies for 83­116. Cambridge, Massachusetts: Ballinger Pro-poor Agricultural Growth. Report on Pro- Publishing. ject 7989, U.K. Department for International Gardner-Outlaw, T., and R. Engelman. 1997. "Sus- Development, Science Research Unit. taining Water: A Second Update." Population Fan S., P.L. Huong, and T.Q. Long. 2003. "Gov- Action International, Washington, DC. ernment Spending and Poverty Reduction in http://www.populationaction.org/resources/ Vietnam." Unpublished, Project report pre- publications/water/water97.pdf pared for the World Bank­funded project "Pro- Gautam, M. 2003. Debt Relief for the Poorest: An poor spending in Vietnam," by International OED Review of the HIPC Initiative. Indepen- Food Policy Research Institute, Washington, dent Evaluation Group. Washington, DC: World DC, and the Central Institute for Economic Bank. Management, Hanoi. Gibb, A. 1974. "Agricultural Modernization, Non- Fan, S., L. Zhang, and X. Zhang. 2002. "Growth, farm Employment and Low Level Urbaniza- Inequality, and Poverty in Rural China: The tion: A Case Study of Central Luzon Subregion." Role of Public Investments." Research Report PhD diss., University of Michigan. No. 125. International Food Policy Research In- Haggblade, S., and P. Hazell. 1989. "Agricultural stitute, Washington, DC. Technology and Farm-Nonfarm Growth Link- FAO (Food and Agriculture Organization of the ages." Agricultural Economics 3 (4): 345­364. United Nations). 2002. World Agriculture: To- Hazell, P., and L. Haddad. 2000. "Agricultural Re- wards 2015/2030. Rome: FAO search and Poverty Reduction." 2020 Vision ------. 2004. FAOSTAT agricultural statistical for Food, Agriculture, and the Environment database. http://faostat.fao.org/faostat/; ac- Discussion Paper No. 34, International Food cessed on April 22, 2005. Policy Research Institute, Washington, DC. Ferguson, A.E., and W.O. Mulwafu. 2005. "Irriga- Hazell, P., and S. Haggblade. 1991. "Rural-Urban tion Reform in Malawi: Exploring Critical Land- Growth Linkages in India." Indian Journal of Water Interactions." Paper presented during the Agricultural Economics 46 (4): 515­529. 1 3 7 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 Hazell, P., and C. Ramasamy. 1991. The Green Rev- Water_for_All_Series/Water_Poverty_Realities/ olution Reconsidered: The Impact of High- Agricultural_Water.pdf. Yielding Rice Varieties in South India. Hussain I., and M. Hanjra. 2003."Does Irrigation Baltimore, Maryland: John Hopkins University Water Matter for Rural Poverty Alleviation? Ev- Press. idence from South and South-East Asia." Water Heyd, H. 2004. "Water Governance and Poverty." Policy 5 (5): 429­442. Unpublished. International Food Policy Re- ------. 2004. "Irrigation and Poverty Alleviation: search Institute, Washington, DC. Review of the Empirical Evidence." Irrigation Hossain, M. 1996. "Agricultural Policies in and Drainage 53 (1): 1­15. Bangladesh: Evolution and Impact on Crop Hussain, I., and D. Wijerathna. 2004. "Irrigation Production." In State, Market and Develop- and Income-Poverty Alleviation: A Comparative ment: Essays in Honor of Rehman Sobhan, eds. Analysis of Irrigation Systems in Developing A.A. Abdullah and A.R. Khan. Dhaka: University Asia." International Water Management Insti- Press Limited. tute. Hossain, M., M.L. Bose, A. Chowdhury, and R. IEG (Independent Evaluation Group). 1992. Meinzen-Dick. 2002. "Changes in Agrarian Re- "Water Supply and Sanitation Projects: The lations and Livelihoods in Rural Bangladesh: In- Bank's Experience, 1967­89." IEG Report No. sights from Repeat Village Studies. In Agrarian 10789. World Bank, Washington, DC. Studies: Essays on Agrarian Relations in Less- ------. 1994. "Review of World Bank Experience Developed Countries, eds. V. K. Ramachangran in Irrigation." IEG Report No. 13676. World and M. Swaminathan. New Delhi: Tulika Books. Bank, Washington, DC. Hussain, I. 2004a. "Pro-poor Intervention Strate- ------. 1995. The World Bank and Irrigation. A gies in Irrigated Agriculture in Asia. Poverty in World Bank Operations Evaluation Study. Irrigated Agriculture: Realities, Issues, and Op- Washington, DC: World Bank. tions with Guidelines. India, Pakistan, ------. 1999. "Rural Development: From Vision Bangladesh, China, Indonesia, and Vietnam." to Action?" IEG Report No. 19448. World Bank, International Water Management Institute Washington, DC. (IWMI), Asian Development Bank (ADB). ------. 2000. "Rural Development: From Vision ------. 2004b. "Assessing Impacts of Irrigation on to Action? Phase II." IEG Report No. 20628. Poverty: Approaches, Methods, Case Studies World Bank, Washington, DC. and Lessons." IWMI, Paper prepared for work- ------. 2001 Strategic Directions. Washington, shop on IWMI-BOKU-Sieberdorf-EARO-Ar- DC: World Bank. bamintch University Study, Ethiopia, April. ------. 2002a. Bridging Troubled Waters: As- ------. 2004c. "Have Low Irrigation Charges Dis- sessing the World Bank Water Resources Strat- advantaged the Poor?" International Water Man- egy. Washington, DC: World Bank. agement Institute paper prepared for ADB's ------. 2002b. Social Funds: Assessing Effec- Water Week, January. tiveness. Washington, DC: World Bank. ------, ed. 2004d. "Pro-poor Intervention Strate- ------. 2003. "The Egypt Matrouh Natural Re- gies in Irrigated Agriculture in Asia: Poverty in sources Management Project." Project Perfor- Irrigated Agriculture: Issues and Options." In- mance Assessment Report. World Bank, ternational Water Management Institute, ADB, Washington, DC. Country Report. ------. 2004a. "Egypt Second Pumping Stations Hussain, I., M. Giordano, and M. Hanjra. 2004. Rehabilitation Project (Loan 3198-EGT) and "Agricultural Water and Poverty Linkages: the National Drainage Project (Loan 3417-EGT Case Studies on Large and Small Systems." and Credit 2313-EGT)." Report No. 28003. In Water and Poverty: the Realities. Experi- World Bank, Washington, DC. ences from the Field. Water for All Publication ------. 2004b. "Use of Atlas Ti in OED Research." Series No. 5. Asian Development Bank. Prepared by Silke Heuser for IEG. World Bank, http://www.adb.org/Documents/Books/ Washington, DC. 1 3 8 B I B L I O G R A P H Y ------. 2005a. "The World Bank's Assistance For gated Ecosystems: Village Studies in Chhattis- Water Resources Management In China." Back- garh." Economic and Political Weekly 35 (Dec. ground paper prepared for the China Coun- 30): 4664­4669. try Assistance Evaluation. World Bank, Jayaraj, D. 1992. "Determinants of Rural Non- Washington, DC. Agricultural Employment: A Village Level Analy- ------. 2005b. "Mexico Irrigation and Drainage sis of the Data for Tamil Nadu." In Themes in Sector Project, the On-Farm and Minor Irriga- Development Economics: Essays in Honour of tion Networks Improvement Project, and the Malcom S. Adiseshiah, ed. S. Subramanian. Agricultural Productivity Improvement Pro- Delhi: Oxford University Press. ject." Report No. 32732. World Bank, Wash- Jehangir, W., M. Ashfaq, I. Hussain, M. Mudasser, ington, DC. and N. Aamir. 2004. "Pro-poor Intervention ------. 2005c. The Effectiveness of the World Strategies in Irrigated Agriculture in Asia: Bank Support for Community-Based and Poverty in Irrigated Agriculture: Issues and Community-Driven Development. Annex l. Options." IWMI, ADB. Washington, DC: World Bank. Johnson, S., M. Svendsen, and F. Gonzalez. 2004. ------. 2006a. "The Economic Impact and Via- "Institutional Reform Options in the Irrigated bility of Large Scale Irrigation Projects: An Sector." Agriculture and Rural Development Analysis of the Second and Third Andhra Discussion Paper 5, World Bank, Washington, Pradesh Irrigation Projects." Draft, World Bank, DC. Washington, DC. Jones, W. I. 1995. The World Bank and Irrigation, ------. 2006b. "Republic of Yemen: Country As- Vol.1. IEG Report No. 14908. World Bank, Wash- sistance Evaluation." Report No. 36527. World ington, DC. Bank, Washington, DC Jordans, E., and M. Zwaterteveen. 1997. A Well of IFPRI (International Food Policy Research Insti- One's Own: Gender Analysis of an Irrigation tute). 2005. "Lessons Learned from the Dragon Program in Bangladesh. International Irriga- and the Elephant." Essays in the IFPRI Annual tion Management Institute and Grameen Krishi Report. December. Foundation, Bangladesh. India Planning Commission. 2002. India's Tenth Khan, M. 2003. "Responding to Millennium De- Five Year Plan: 2002­2007. New Delhi: Gov- velopment Goals: A Strategy for Water Security." ernment of India. Department of Economic and Social Affairs, IPTRID (International Programme for Technol- United Nations. ogy and Research in Irrigation and Drainage). Kikuchi, M., and A. Inocencio, with D. Merrey, 1999. "Poverty Reduction and Irrigated Agri- M. Tonosaki, and A. Maruyama. Forthcoming. culture." Issues Paper No. 1. Rome: FAO. Costs and Performance of Irrigation Projects: IRRI (International Rice Research Institute). A Comparison of Sub-Saharan Africa and 2005. World Rice Statistics. http://www.irri.org/ Other Developing Regions. Colombo, Sri Lanka: science/ricestat. International Water Management Institute. IWMI (International Water Management Insti- Lipton, M., J.A. Litchfield, and M. Faures. 2003. tute). 2003a. "Pro-poor Irrigation Management "The Impact of Irrigation on Poverty: A Frame- Transfer? The Importance of Small Farmer Par- work for Analysis." Journal of Water Policy 5 ticipation in Irrigation Management Transfer in (4): 413­427. India." Water Policy Briefing, IWMI, Colombo, Löfgren, H. 1996. "The Cost of Managing with Sri Lanka. Less: Cutting Water Subsidies and Supplies in ------. 2003b. "Pro-poor Intervention Strategies Egypt's Agriculture." Trade and Macroeco- in Irrigated Agriculture in Asia." Six-country nomics Division Discussion Paper 7. Interna- review. ADB-funded IWMI Project. IWMI, tional Food Policy Research Institute, Colombo, Sri Lanka. Washington, DC. Janaiah, A., M. Bose, and A. Agarwal. 2000. "Poverty McCartney, M., E. Boelee, O. Cofie, F. Ameras- and Income Distribution in Rainfed and Irri- inghe, and C. Mutero. 2005. "Agricultural Water 1 3 9 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 Development in Sub-Saharan Africa: Planning and Implementation Process." Paper prepared and Management to Improve the Benefits and for the Collaborative Program of Investments Reduce the Environmental and Health Costs." in Agricultural Water Management in Sub- Paper prepared for the Collaborative Program Saharan Africa, International Water Manage- of Investments in Agricultural Water Manage- ment Institute. ment in Sub-Saharan Africa, International Water Mundlak, Y., D. Larson, and R. Butzer. 2003. "Agri- Management Institute. cultural Dynamics in Thailand, Indonesia and McGee, R., J. Levene, and A. Hughes. 2002. "As- the Philippines." Australian Journal of Agri- sessing Participation in Poverty Reduction Strat- cultural and Resource Economics 48 (1): egy Papers: A Desk-Based Synthesis of 95­126. Experience in Sub-Saharan Africa." Research OECD (Organisation for Economic Co-operation Report 52, Institute of Development Studies, and Development). 1999. "Agricultural Water Brighton, U.K. http://www.ids.ac.uk/ids/ Pricing in OECD Countries." Unpublished. bookshop/rr/rr52.pdf. OECD, Paris. Meinzen-Dick, R., M. Mendoza, L. Sadoulet, G. Olson, D. 1996. "The Columbia Basin Project: Abiad-Shileds, and A. Subramanian. 1997. "Sus- Project Operations and Economic Benefits, A tainable Water User Associations: Lessons from Regional Overview." Report prepared for the a Literature Review." In User Organization for Pacific Northwest Project. Sustainable Water Services, eds. A Subraman- Omilola, B. 2005. "Rapid Poverty Reduction Ap- ian, N.V. Jagannathan, and R. Meinzen-Dick. praisal of the Impact of Small-Scale Fadama Washington, DC: World Bank. Irrigation Investment in Nigeria." Paper pre- Meinzen-Dick, R., and W. Van der Hoek. 2001. pared for the Collaborative Program of In- "Multiple Uses of Water in Irrigated Areas." Ir- vestments in Agricultural Water Management rigation and Drainage Systems 15 (2): 93­98. in Sub-Saharan Africa, International Water Man- Mejia, A., L. Azevedo, M. Gambrill, A. Baltar, and agement Institute. T. Triche. 2003. "Brazil: Water, Poverty Reduc- Oweis, T., and A. Hachum. 2001. "Coping with In- tion and Sustainable Development." In Brazil creased Water Scarcity in Dry Areas: ICARDA Re- Water Series, Volume 4: Water Poverty Reduc- search to Increase Water Productivity." tion, and Sustainable Development. Wash- International Center for Agricultural Research ington, DC: World Bank. in the Dry Areas, Aleppo, Syria. Mellor, J. 2001. "Irrigation, Agriculture and Poverty Pardey, P.G., R.K. Lindner, E. Abdurachman, Reduction: General Relationships and Specific S. Wood, S. Fan, W.M. Eveleens, B. Zhang, and Needs." In Managing Water for the Poor: Pro- J.M. Alson. 1992. The Economic Returns to In- ceedings of the Regional Workshop on Pro- donesian Rice and Soybean Research. poor Intervention Strategies in Irrigated AARD/ISNAR Report. Agriculture in Asia, August 9­10, 2001. Patel, M. 1981. "Modeling Determinants of Health." Colombo, Sri Lanka: International Water Man- International Journal of Epidemiology agement Institute. 10 (3):177­80. Minten, B. 2006. "The Role of Agriculture in Poverty Penning de Vries, P., H. Sally, and A. Inocencio. Alleviation Revisited: The Case of Madagascar." 2005. "Opportunities for Private Sector Partic- Research sponsored by the Poverty Reduction ipation in Agricultural Water Development and and Economic Management Group of the Management." Paper prepared for the Collab- World Bank, and by the Norwegian Trust Fund. orative Program of Investments in Agricultural Mody, J. 2004. "Achieving Accountability through Water Management in Sub-Saharan Africa, In- Decentralization: Lessons for Integrated River ternational Water Management Institute. Basin Management." Policy Research Working Perry, C. 2001. "Charging for Irrigation Water: The Paper 3346. World Bank, Washington, DC. Issues and Options, with a Case Study from Morardet, S., J. Sehoka, H. Sally, and D. Merrey. Iran." IWMI Research Report 52, International 2005. "Review of Irrigation Project Planning WaterManagementInstitute,Colombo,SriLanka. 1 4 0 B I B L I O G R A P H Y Pezzey, J. 1992. "The Symmetry between Con- Rosegrant, M.W., and H.P. Binswanger. 1994. "Mar- trolling Pollution by Price and Controlling It by kets in Tradable Water Rights: Potential for Ef- Quantity." Canadian Journal of Economics 25 ficiency Gains in Developing Country Water (4): 983­991. Resource Allocation." World Development 22 Picciotto, R., W. van Wicklin, and E. Rice. 2001. (11): 1613­1625. Involuntary Resettlement--Comparative Rosegrant, M.W., X. Cai, and S.A. Cline. 2002. Perspectives. World Bank Series on Evaluation World Water and Food to 2025: Dealing with and Development. Volume 2. Transaction Pub- Scarcity. Washington, DC: International Food lishers. Policy Research Institute. Pinstrup-Andersen, P., and P. Hazell. 1985. "The Im- Rosegrant, M.W., and S. Cline. 2002. "The Politics pact of the Green Revolution and Prospects for and Economics of Water Pricing in Developing the Future." Food Reviews International 1 (1): Countries." Water Resources Impact 4 (1): 6­8. 1­25. Rosegrant, M.W., and P. Hazell. 2000. Transform- Qiuqiong, H., D. Dawe, J. Huang, S. Rozelle, and ing the Rural Asian Economy: The Unfinished J. Wang. 2005. "Irrigation, Poverty and Inequality Revolution. Hong Kong: Oxford University Press. in Rural China." Australian Journal of Agri- Rosegrant, M.W., F. Kasryno, and N. Perez. 1998. cultural and Resource Economics 49 (2) "Output Response to Prices and Public In- 159­175. vestment in Agriculture: Indonesian Food Raju, K.V., A. Gulati, and R. Meinzen-Dick. 2003. Crops." Journal of Development Economics 55 "Innovations in Irrigation Financing: Tapping (2): 333­352. Domestic Financial Markets in India." Rosegrant, M.W., C. Ringler, D.C. McKinney, X. Cai, Markets and Structural Studies Division Dis- A. Keller, and G. Donoso. 2000. "Integrated cussion Paper No. 58. International Food Pol- Economic-Hydrologic Water Modeling at the icy Research Institute, Washington, DC. Basin Scale: The Maipo River Basin." Agricul- http://www.ifpri.org/divs/mtid/dp/papers/ tural Economics 24 (1): 33­46. mssdp58.pdf. Rosegrant, M., R.G. Schleyer, and S. Yadav. 1995. Ravallion, M. 2001. "Growth, Inequality and "Water Policy for Efficient Agricultural Diversi- Poverty: Looking Beyond Averages." World De- fication: Market Based Approaches." Food Pol- velopment. icy 20 (3): 203­223. ------. 2004. "Pessimistic on Poverty?" The Econ- Rosegrant, M.W., and M. Svendsen. 1993. "Asian omist April 10: 74. Food Production in the 1990s: Irrigation In- Ravallion, M., and S. Chen. 1997. "What Can New vestment and Management Policy." Food Pol- Survey Data Tell Us about Recent Changes in icy 18 (1): 13­32. Distribution and Poverty?" World Bank Eco- Saleth, R.M. 1996. Water Institutions in India: nomic Review 11 (2): 357­382. Economics, Law and Policy. New Delhi: Com- Ringler, C., C. Rodgers, and M.W. Rosegrant. 2003. monwealth Publishers. Irrigation Investment, Fiscal Policy and Water ------. 1997. "India." In Water Pricing Experi- Resource Allocation in Indonesia and Viet- ences. An International Perspective, eds. A. nam. Summary Synthesis Report. ADB-RETA Dinar and A. Subramanian, 54­60. World Bank Project 5866. Environment and Production Technical Paper No. 386. World Bank, Wash- Technology Division, International Food Policy ington, DC. Research Institute, Washington, DC. Sawada, Y., and N. Shinkai. 2002. "Do Infrastruc- Ringler, C., M.W. Rosegrant, and M.S. Paisner. ture Investments Help Reduce Poverty? Eval- 2000. "Irrigation and Water Resources in Latin uation of an Irrigation Project in Sri Lanka." America and the Caribbean: Challenges and Japan Bank for International Cooperation, Strategies." Environment and Production Tech- Tokyo. http://home.hiroshima-u.ac.jp/jea2002/ nology Division Discussion Paper No. 64, In- summary_html/SAWADAyasuyuki.pdf. ternational Food Policy Research Institute, Scobie, G.M., and R. Posada. 1978. "Technical Washington, DC. Change and Income Distribution: The Case of 1 4 1 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 Rice in Columbia." American Journal of Agri- UNESCO (United Nations Educational, Scientific, cultural Economics 60 (1): 85­92. and Cultural Organization). 2003. Water for Seckler, D. 1990. "Private Sector Irrigation in People, Water for Life. United Nations World Africa." In Irrigation in Sub-Saharan Africa: Water Development Report 1. http://www. The Development of Public and Private Sys- unesco.org/water/wwap/wwdr. tems, eds. S. Barghouti and G. Le Moigne. United Nations. 2005. World Population Prospects: World Bank Technical Paper No. 123. World The 2004 Revision--Highlights. Department Bank, Washington, DC. of Economic and Social Affairs, Population Di- Seckler, D., D. Molden, and R. Barker. 1998. Water vision. New York. Scarcity in the Twenty-First Century. IWMI Ut, T., M. Hossain, and A. Janaiah. 2000. "Modern Water Brief 1. Colombo, Sri Lanka: Interna- Farm Technology and Infrastructure in Vietnam: tional Water Management Institute. Impact on Income Distribution and Poverty." Shuval, H. I. 1990. Wastewater Irrigation in De- Economic and Political Weekly (Dec 30): veloping Countries: Health Effects and Tech- 4638­4643. nical Solutions. Summary of World Bank Vaidynathan, A., A. Krishnakumar, A. Rajagopal, Technical Paper No. 51. Water and Sanitation and D. Varharajan. 1994. "Impact of Irrigation Discussion Paper Series No. 11433, World Bank, on Productivity of Land." Journal of Indian Washington, DC. School of Political Economy 6 (4). Smith L. 2004. "Assessment of the Contribution of Van de Walle, D. 2000. "Are Returns to Invest- Irrigation to Poverty Reduction and Sustainable ment Lower for the Poor? Human and Physical Livelihoods." International Journal of Water Capital Interactions in Rural Vietnam." Policy Resources 20 (2): 243­257. Research Working Paper No. 2425, World Bank, Soussan, J.G. 2002. "Poverty and Water Security." Washington, DC. Forthcoming in the Review Of Concept paper prepared for the Water and Development Economics. Poverty Initiative, Asian Development Bank, Van Koppen, B. 1999. "Targeting Irrigation Support Manila. to Poor Women and Men." Water Resources De- Svendsen, M., A. Gulati, and K.V. Raju. Forth- velopment 15. coming. "Reform Options for Construction Van Koppen, B., R. Parthasarathy, and C. Safiliou. and Rehabilitation." In Financial and Institu- 2002. "Poverty Dimensions of Irrigation Man- tional Reforms in Indian Irrigation. Bangalore, agement Counselor in Large-Scale Canal Irri- India: Books for Change. gation in Andrah Pradesh and Gujarat, India." Thompson, R.L. 2001. "The World Bank Strategy Research Report 61, International Water Man- to Rural Development with Special Reference agement Institute, Colombo, Sri Lanka. to the Role of Irrigation and Drainage." Keynote Van Koppen, B., and C. Safilios-Rothchild. 2005. address on the occasion of the 52nd IEC meet- "Poverty Considerations in Investments in Agri- ing of the International Commission on Irri- cultural Water Management." Paper prepared gation and Drainage, September 16­21, Seoul. for the Collaborative Program of Investments Thurman, M. 2002. "Irrigation and Poverty in Cen- in Agricultural Water Management in Sub- tral Asia: A Field Assessment." World Bank, Saharan Africa. Washington, DC. Vermillion, D.L. 1999. "Property Rights and Col- Timmer, P. 2003. "Agriculture and Pro-Poor lective Action in the Devolution of Irrigation Growth." Pro-poor economic growth research System Management." In Collective Action, studies. Contract No. PCE-1-02-00-00015-00. Property Rights and Devolution of Natural Tsur, Y., T. Roe, R. Doukkali, and A. Dinar. 2004. Resource Management: Exchange of Knowl- "Pricing Irrigation Water: Principles and Cases edge and Implications for Policy. Proceed- from Developing Countries." Resources for ings of the International Conference, Puerto the Future, Washington, DC. Azul, the Philippines, June 21­25. 1999, eds. 1 4 2 B I B L I O G R A P H Y R.S. Meinzen-Dick, A. Knox, and M. di Grego- ------. 2003a. Reaching the Rural Poor. Wash- rio. Feldafing: German Foundation for Inter- ington, DC: World Bank. national Development. ------. 2003b. Philippines: Country Water As- Winpenny, J. 2003. Financing Water for All: Report sistance Resources Strategy. East Asia and of the World Panel on Financing Water In- Pacific Department, World Bank, Washington, frastructure. World Water Council and the DC. Global Water Partnership. http://www.financing ------. 2003c. Brazil Water Series, Volume 4: waterforall.org. Water Poverty Reduction, and Sustainable World Bank. 1990. World Development Report Development. Washington, DC: World Bank. 1990: Poverty. Washington, DC: World Bank. ------. 2004a. Irrigated Agriculture in the Brazil- ------. 1992. Wapehans Report. Task Force on ian Semi-Arid Region: Social Impacts and Ex- Portfolio Management. Washington, DC: World ternalities. Report No. 28785-BR. Washington, Bank. DC: World Bank. ------. 1993. "Water Resources Management: A ------. 2004b. Dominican Republic: Country World Bank Policy Paper." World Bank, Wash- Water Assistance Resources Strategy. South ington, DC. America and Caribbean Department. Wash- ------. 1994. Jordan Agricultural Sector Adjust- ington, DC: World Bank. ment Operation. World Bank, Washington, DC. ------. 2004c. Honduras: Country Water Assis- ------. 1996. "The World Bank Participation tance Resources Strategy. South America and Sourcebook." World Bank Institute, Washing- Caribbean Department. Washington, DC: World ton, DC. http://www.worldbank.org/wbi/ Bank. sourcebook/sbhome.htm. ------. 2004d. Peru: Country Water Assistance Re- ------. 1997a. Rural Development: From Vision sources Strategy. South America and Caribbean to Action. Report 97/25, Washington, DC: World Department. Washington, DC: World Bank. Bank. ------. 2004e. "From Adjustment Lending to De- ------. 1997b. Strategic Compact. Washington, velopment Policy Lending: Update of World DC: World Bank. Bank Policy." Operations Policy and Country ------. 2000a. World Development Report Services, World Bank, Washington, DC. 2000/2001: Attacking Poverty. Washington, ------. 2004d. Water Resources Sector Strategy-- DC: Oxford University Press. Strategic Directions for World Bank Engage- ------. 2000b. "World Bank Lending Instruments: ment. Washington, DC: World Bank. Resources for Development Impact." Opera- ------. 2005a. Madagascar: Review of Agricul- tions Policy and Strategy, World Bank, Wash- ture and Environmental Sectors. Washington, ington, DC. DC: World Bank. ------. 2001. Water Resources Sector Strategy. ------. 2005b. Madagascar: The Impact of Pub- Washington, DC: World Bank. lic Spending on Irrigation Perimeters, ------. 2002a. China: Country Water Assistance 1985­2004. Washington, DC: World Bank. Resources Strategy. East Asia and Pacific De- ------. 2005c. Agricultural Growth for the Poor: partment. Washington, DC: World Bank. An Agenda for Development. Washington, DC: ------. 2002b. Reaching the Rural Poor in the Eu- World Bank. rope and Central Asia Region. Europe and ------. 2005d. Shaping the Future For Water in Central Asia Department, Environmentally and Agriculture: A Sourcebook for Investment in Socially Sustainable Development Unit Rural Agricultural Water Management. Washing- Development Strategy Paper, World Bank, ton, DC: World Bank. Washington, DC. ------. 2005e. "Feedback Links between Econo- ------. 2002c. Rural Poor I: Strategy and Busi- mywide and Farm-Level Policies with Applica- ness Plan. Washington, DC: World Bank. tion to Irrigation Water Management in 1 4 3 WAT E R M A N A G E M E N T I N A G R I C U LT U R E : T E N Y E A R S O F W O R L D B A N K A S S I S TA N C E , 1 9 9 4 ­ 2 0 0 4 Morocco." Policy Research Working Paper 3550, World Bank, Washington, DC. ------. 2006. Reengaging in Agricultural Water Management: Challenges and Options. Wash- ington, DC: World Bank. World Commission on Dams. 2000. Dams and De- velopment: A New Framework for Decision- making. London: Earthscan. 1 4 4 IEG PUBLICATIONS Study Series 2004 Annual Review of Development Effectiveness: The Bank's Contributions to Poverty Reduction Addressing the Challenges of Globalization: An Independent Evaluation of the World Bank's Approach to Global Programs Agricultural Extension: The Kenya Experience Assisting Russia's Transition: An Unprecedented Challenge Bangladesh: Progress Through Partnership Brazil: Forging a Strategic Partnership for Results--An OED Evaluation of World Bank Assistance Bridging Troubled Waters: Assessing the World Bank Water Resources Strategy Capacity Building in Africa: An OED Evaluation of World Bank Support The CIGAR at 31: An Independent Meta-Evaluation of the Consultative Group on International Agricultural Research Country Assistance Evaluation Retrospective: OED Self-Evaluation Debt Relief for the Poorest: An OED Review of the HIPC Initiative Developing Towns and Cities: Lessons from Brazil and the Philippines The Drive to Partnership: Aid Coordination and the World Bank Economies in Transition: An OED Evaluation of World Bank Assistance The Effectiveness of World Bank Support for Community-Based and ­Driven Development: An OED Evaluation Evaluating a Decade of World Bank Gender Policy: 1990­99 Evaluation of World Bank Assistance to Pacific Member Countries, 1992­2002 Financial Sector Reform: A Review of World Bank Assistance Financing the Global Benefits of Forests: The Bank's GEF Portfolio and the 1991 Forest Strategy and Its Implementation Fiscal Management in Adjustment Lending IDA's Partnership for Poverty Reduction Improving the Lives of the Poor Through Investment in Cities India: The Dairy Revolution Information Infrastructure: The World Bank Group's Experience Investing in Health: Development Effectiveness in the Health, Nutrition, and Population Sector Jordan: Supporting Stable Development in a Challenging Region Lesotho: Development in a Challenging Environment Mainstreaming Gender in World Bank Lending: An Update Maintaining Momentum to 2015? An Impact Evaluation of Interventions to Improve Maternal and Child Health and Nutrition Outcomes in Bangladesh The Next Ascent: An Evaluation of the Aga Khan Rural Support Program, Pakistan Nongovernmental Organizations in World Bank­Supported Projects: A Review Poland Country Assistance Review: Partnership in a Transition Economy Poverty Reduction in the 1990s: An Evaluation of Strategy and Performance The Poverty Reduction Strategy Initiative: An Independent Evaluation of the World Bank's Support Through 2003 Power for Development: A Review of the World Bank Group's Experience with Private Participation in the Electricity Sector Promoting Environmental Sustainability in Development Putting Social Development to Work for the Poor: An OED Review of World Bank Activities Reforming Agriculture: The World Bank Goes to Market Sharing Knowledge: Innovations and Remaining Challenges Social Funds: Assessing Effectiveness Tunisia: Understanding Successful Socioeconomic Development Uganda: Policy, Participation, People The World Bank's Experience with Post-Conflict Reconstruction The World Bank's Forest Strategy: Striking the Right Balance Zambia Country Assistance Review: Turning an Economy Around Evaluation Country Case Series Bosnia and Herzegovina: Post-Conflict Reconstruction Brazil: Forests in the Balance: Challenges of Conservation with Development Cameroon: Forest Sector Development in a Difficult Political Economy China: From Afforestation to Poverty Alleviation and Natural Forest Management Costa Rica: Forest Strategy and the Evolution of Land Use El Salvador: Post-Conflict Reconstruction India: Alleviating Poverty through Forest Development Indonesia: The Challenges of World Bank Involvement in Forests The Poverty Reduction Strategy Initiative: Findings from 10 Country Case Studies of World Bank and IMF Support Uganda: Post-Conflict Reconstruction Proceedings Global Public Policies and Programs: Implications for Financing and Evaluation Lessons of Fiscal Adjustment Lesson from Urban Transport Evaluating the Gender Impact of World Bank Assistance Evaluation and Development: The Institutional Dimension (Transaction Publishers) Evaluation and Poverty Reduction Monitoring & Evaluation Capacity Development in Africa Public Sector Performance--The Critical Role of Evaluation All IEG evaluations are available, in whole or in part, in languages other than English. For our multilingual selection, please visit http://www.worldbank.org/ieg THE WORLD BANK