Agricultural Millenium Development Goals

REPORT NO. 32729-GLB

Agriculture and Rural Development Department

Agriculture and Achieving TheMillennium Development Goals

THE WORLD BANK

Agriculture & Rural Development DepartmentWorld Bank1818 H Street, N.W.Washington, D.C. 20433http://www.worldbank.org/rural

Agriculture and Rural Development Department

Agriculture and Achieving TheMillennium Development Goals

THE WORLD BANK

This volume is a product of the staff of the International Bank for Reconstruction andDevelopment/The World Bank. The findings, interpretations, and conclusions expressedin this paper do not necessarily reflect the views of the Executive Directors of The WorldBank or the governments they represent. The World Bank does not guarantee the accu-racy of the data included in this work. The boundaries, colors, denominations, and otherinformation shown on any map in this work do not imply any judgment on the part ofThe World Bank concerning the legal status of any territory or the endorsement oracceptance of such boundaries.

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. TheInternational Bank for Reconstruction and Development/ The World Bank encouragesdissemination of its work and will normally grant permission to reproduce portions ofthe work promptly.

For permission to photocopy or reprint any part of this work, please send a request withcomplete information to the Copyright Clearance Center, Inc., 222 Rosewood Drive,Danvers, MA 01923, USA, telephone 978-750-8400, fax 978-750-4470, http://www.copyright.com/.

All other queries on rights and licenses, including subsidiary rights, should be addressedto the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC20433, USA, fax 202-522-2422, e-mail [email protected].

iii

A C R O N Y M S A N D A B B R E V I AT I O N S v

E X E C U T I V E S U M M A RY vii

Introduction 1

Perspectives on the Role of Agriculture in Meeting the MillenniumDevelopment Goals 3

MDG 1—Eradicate Extreme Poverty and Hunger 5Poverty Reduction Through Agriculture-Led Economic Growth 7Agriculture’s Contribution to Food and Nutrition Security 8

MDG 2—Achieve Universal Primary Education 9MDG 3—Promote Gender Equality and Empower Women 10MDG 4—Reduce Child Mortality 10MDG 5—Improve Maternal Health 12MDG 6—Combat HIV/AIDS, Malaria, and Other Diseases 13MDG 7—Ensure Environmental Sustainability 13MDG 8—Develop a Global Partnership for Development 15

Alternative Scenarios to 2015: Prospect and Policies for Meeting the MDGs on Poverty 19

Country-Level Assessment of Selected MDGs: Ethiopia and Zambia 19Ethiopia: Agriculture-Led Growth and 22Zambia: Facing the Challenge of the Millennium Development Goals—

The Role of Agriculture 29Conclusions 34

Global Quantified Assessment of Selected MDGTargets: IMPACT-WATER 36

Results 41Implications for Investment 46Conclusions 48

3

2

1

Contents

Global trends and Emerging Issues in Meeting the MDGs: a Trade,Macroeconomic, and Policy Perspective 49

Policy Actions and Their Impacts on the MDGs 49Trade Policies 50

Policies of Developed Countries 50Policies of Developing Countries 52Policies in Development Assistance: The Case of Food Aid 53Macroeconomic Reforms and National Government Infrastructure

Investments 54Importance of Rural Infrastructure 55On Public Investment and Infrastructure 58

The Role of the Private Sector in Supporting Agriculture to Meet the MDGs 58

The Rural Nonfarm Private Sector 60Effective Legal, Regulatory, and Institutional Environments 61Private-Public Partnerships and Investment Synergies: The Case of

Water and Agricultural Research 61The Role of Foreign Direct Private Investment 65The Role of Governance Structures in Agriculture for Achieving

the MDGs 65Conclusion 67

Conclusions 69

E N D N O T E S 7 5

R E F E R E N C E S 7 7

Appendix 1. The Millennium Development Goals, Targets, and Indicators 81

Appendix 2. IMPACT-WATER Developing Countries and Region 84

5

4

iv Contents

Contents v

Figures2.1 Proportion of Population Living on Less than $1 Per Day (PPP),

Most Recent Estimates 4

2.2 Agriculture Value Added, as a Percentage of GNP, 2001 4

2.3 Progress in Attaining MDG 1 Targets for Poverty and Undernourishment, by Region (Trend Lines Based on Recent Performance and Targets for 2015) 5

2.4 Child Malnutrition by Urban or Rural Residence, Stunting (Low Height for Age) Prevalence Among Preschoolers, Surveys Since 1999 6

2.5 Prevalence of Undernourishment—Proportion of the Population Unable to Acquire Sufficient Calories to Meet their Daily Caloric Requirements, 2003 Estimate 8

2.6 Child Mortality by Urban or Rural Residence, Surveys Since 1999 11

2.7 Estimated Prevalence of Iron Deficiency Anemia in Women Aged 15 to 49,2004 12

3.1 Poverty Rate in Ethiopia Under Different Growth Scenarios 29

3.2 Poverty Reduction Under the Current Growth Path (2001–2050) 32

3.3 Per-Capita Kilocalorie Availability in 1995 and Projections to 2015, Baseline and MDG Scenarios 43

3.4 Child Malnutrition, Developing Countries, 1995, 2015 Baseline, 2015 MDGScenario, and 2015 Target 43

3.5 Child Malnutrition, Developing Country Regions, 2015 Baseline, 2015 MDGScenario, and 2015 Target 44

3.6 Number of People Without Access to an Improved Water Source, 1995, and Projected 2015 Baseline and MDG Scenarios 46

3.7 Cost Estimates for Implementing the Baseline and MDG Scenario, 2015,Developing Countries 47

4.1 Changes in Paved Roads, Telephone Lines, and Electricity Production Over Time, Selected Regions 57

4.2 Sample of Public-Private Investment Arrangements 62

Tables2.1 Summary of Links Between the Agricultural Sector and the Millennium

Development Goals, Principally at Household Level 18

3.1 Comparison of Poverty Across Selected African Countries 20

3.2 Poverty Lines in the Country Case Studies 20

3.3 Economic and Demographic Indicators, 2001 21

3.4 Average Trends Across Countries, 1991–2001 Averages 21

3.5 Cereal Production in Selected African Countries 22

3.6 Growth and Poverty Reduction Under Agricultural Growth Options 24

3.7 Poverty in Zambia During the 1990s 31

3.8 Growth and Poverty Reduction Under the Alternative Zambian GrowthScenarios (2001–2015) 35

3.9 Child Malnutrition Estimates (Underweight), 1990 and 1995, with 2015 MDGIndicator Target 38

3.10 Access to Safe Drinking Water, 1990 and 1995, with 2015 MDG Target Indicatorand 2015 IMPACT-WATER Baseline Estimates 39

3.11 Parameters for Developing Countries, Baseline and MDG Scenarios 41

3.12 Income Growth Requirements for Baseline and MDG Scenarios, 1995–201542

3.13 Southeast Asia, Share of Malnourished Preschool Children, 1995, and 2015Baseline, MDG Scenario, and Indicator Target 45

4.1 Latin America and the Caribbean: Main Features of Changes in AgriculturalPolicies 55

4.2 Possible Scenario Regarding Financing Sources for Agriculture under NEPAD 64

vi

This report was prepared by Mark W. Rosegrant, Claudia Ringler, Todd Benson, XinshenDiao, Danielle Resnick, James Thurlow, Maximo Torero, and David Orden InternationalFood Policy Research Institute and coordinated by Nwanze Okidegbe, Derek Byerlee,and Sanjiva Cooke (World Bank). Editorial assistance was provided by Melissa Williams(World Bank) and Sarah A. Cline and Rowena Andrea Valmonte-Santos (consultants).This report was prepared under the general direction of Kevin Cleaver (sector director)and Sushma Ganguly (sector manager).

vii

ACC/SCN Administrative Committee on Coordination/Subcommittee on Nutrition

AfDB African Development Bank

AIARD Association for International Agriculture and Rural Development

BRI Bank Rakyat Indonesia

CBOs Community-Based Organizations

CPI Corruption Perceptions Index 2004

EC European Commission

EPRDF Ethiopia Peoples’ Revolutionary Democratic Front

FAO Food and Agriculture Organization

FAOSTAT Food and Agriculture Organization Statistical Database

FDI Foreign Direct Investment

GATT General Agreement on Trade and Tariffs

GDP Gross Domestic Product

GNP Gross National Product

HIPC Heavily Indebted Poor Countries

HIV/AIDS Human Immunodeficiency Virus/ Acquired Immune Deficiency Syndrome

IFAD International Fund for Agricultural Development

MAAIF Plan for Modernization of Agriculture

MDGs Millennium Development Goals

MFPED Poverty Eradication Action Plan

MIGA Multilateral Investment Guarantee Agency

MOFED Ministry of Finance and Economic Development

MOPED Ministry of Planning and Economic Development

ODA Official Development Assistance

OECD Organization for Economic Co-operation and Development

PPP Purchasing Power Parity

PRSP Poverty Reduction Strategy Papers

NEPAD New Partnership for Africa’s Development

UN United Nations

Acronyms andAbbreviations

UNDP United Nations Development Program

UNESCO United Nations Educational, Scientific, and Cultural Organization

UNICEF United Nations Children’s Fund

SAPs Structural Adjustment Programs

WANA West Asia and North Africa

WFP World Food Program

WHO World Health Organization

WTO World Trade Organization

Vice President: Ian JohnsonSector Director: Kevin CleaverSector Manager: Sushma GangulyTask Team Leader: Nwanze OkidegbeTask Team: Derek Byerlee, Sanjiva Cooke, Melissa Williams

viii Acronyms and Abbreviations

ix

Relationship Between Agriculture and the Millennium Development Goals

In 2000 the member states of the United Nations adopted theMillennium Declaration as a renewed commitment to human devel-opment. The Declaration includes eight Millennium DevelopmentGoals (MDGs), each with quantified targets, to motivate the interna-tional community and provide an accountability mechanism foractions taken to enable millions of poor people to improve theirlivelihoods. The MDGs are as follows:

1. Eradicate extreme poverty and hunger2. Achieve universal primary education3. Promote gender equality and empower women4. Reduce child mortality5. Improve maternal health6. Combat human immunodeficiency virus/acquired immune

deficiency syndrome (HIV/AIDS), malaria, and other diseases7. Ensure environmental sustainability8. Develop a global partnership for development.

About 70 percent of the MDGs’ target group live in rural areas,particularly in Asia and Africa, and for most of the rural poor agri-culture is a critical component in the successful attainment of theMDGs. Even though structural transformations are important in thelonger term, more immediate gains in poor households’ welfare canbe achieved through agriculture, which can help the poor overcomesome of the critical constraints they now face in meeting their basicneeds. Thus, a necessary component in meeting the MDGs by 2015in many parts of the world is a more productive and profitable agri-cultural sector.

While the linkage with agriculture is particularly strong for thefirst MDG, or MDG 1—halving by 2015 the proportion of those suf-fering from extreme poverty and hunger—all MDGs have direct orindirect linkages with agriculture. Agriculture contributes to MDG 1through agriculture-led economic growth and through improvednutrition. In low-income countries economic growth, which enablesincreased employment and rising wages, is the only means by whichthe poor will be able to satisfy their needs sustainably.

MDG 2, on universal education, has the most indirect linkage toagriculture. A more dynamic agricultural sector will change the

Executive Summary

assessment of economic returns to educating chil-dren, compared to the returns from keeping chil-dren out of school to work in household(agricultural) enterprises. Agriculture contributesto MDG 3 directly through the empowerment ofwomen farmers and indirectly through reductionof the time burden on women for domestic tasks.Agriculture contributes to reduced child mortality(MDG 4) indirectly by increasing diversity of foodproduction and making more resources availableto manage childhood illnesses. Agriculture di-rectly helps improve maternal health (MDG 5)through more diversified food production andhigher-quality diets, and indirectly through in-creased incomes and, thus, reduced time burdenson women. Agriculture also directly helps to com-bat HIV/AIDS, malaria, and other diseases (MDG6) through higher-quality diets, and indirectly byproviding additional income that can be devotedto health services. Agriculture practices can beboth direct causes of and important solutions toenvironmental degradation (MDG 7). More pro-ductive agricultural technologies allow the with-drawal of agriculture from marginal, sensitiveenvironments. Developing a global partnershipfor development (MDG 8) will help maintain thesteady increase in agricultural trade and signifi-cant increases in development assistance offeredto the agricultural sector, increases that help sus-tain the benefits from agriculture in the longerterm

Agriculture and MDG 1: Country-LevelAnalysis for Ethiopia and Zambia

Compared with other regions of the developingworld, Sub-Saharan Africa faces the largest chal-lenge in terms of meeting MDG 1. In 2001, about47 percent of the population was living below theinternational poverty line. Agriculture will be theprimary means of addressing this challenge, as 65 percent of the people in the region derive theirlivelihoods from the sector. However, the contri-bution that agriculture can make in achieving theMDGs in this region depends on the particularconstraints and opportunities prevailing withineach of the countries.

In this report, Ethiopia and Zambia, with ruralpoverty rates of 45 percent and 86 percent, respec-tively, are analyzed according to the level and typeof agricultural growth required to meet the MDGs

based on alternative development scenarios fromeconomy-wide models. Ethiopia is a predomi-nantly subsistence economy with agriculture con-tributing 52 percent to GDP and with almost 85 percent of the population living in rural areas.Productivity in the sector is low, due to frequentdroughts, limited input use, and poor infrastruc-ture. In contrast, mining traditionally has domi-nated economic growth in Zambia, which in turnhas marginalized the country’s agricultural sector.Zambia has both a higher proportion of its pop-ulation living below the poverty line and a higherconcentration of individuals at the low end of theincome distribution.

Model results for Ethiopia show that if thecountry stays on a business-as-usual growth path,poverty will increase by another 10 million people,and food security will be compromised even fur-ther. The largest impact on poverty and food secu-rity can be achieved through a focus on growth instaple crops, which today account for 65 percent ofagricultural value added as well as most small-holder employment. Rapid growth in the livestocksector has the most significant effect on overalleconomic growth but has a relatively smallerpoverty alleviation effect. Very rapid growth in thenontraditional-export sector fuels total economicgrowth as well, but has little impact on ruralpoverty levels. Accelerated growth in all threesectors would help slash the poverty rate by 16 percentage points to 27 percent by 2015. Thesimulated growth in staple food production couldbe achieved through a doubling of the irrigationarea by 2015, and by improving the efficiency offertilizer use combined with enhanced seed use.Moreover, as more than 50 percent of the poor livein food-deficit areas where the availability of foodstaples per household is half the national average,market access and market development need to beintegral parts of a national agricultural develop-ment strategy. Enhanced market access, chieflythrough large investments in improved and ex-tended road networks, would reduce the nationalpoverty rate to 22.7 percent, and thus helpEthiopia reach the MDG 1 target.

In Zambia, the poverty rate under business-as-usual growth would still be 68 percent by 2015,only 7 percentage points lower than the currentpoverty rate of 75 percent. Annual GDP growth of8.8 percent would be required to halve poverty by 2015. Although agriculture accounts for only

x Agriculture and Achieving the Millennium Development Goals

Executive Summary xi

25 percent of GDP, it is still the main source oflivelihood for most of the country’s population,including the majority of Zambia’s poor who livein rural areas where the incidence and severity ofpoverty is greatest. More rapid productivitygrowth under the Agriculture-Led GrowthScenario would lead to higher sectoral growth forboth staples and export crops. Under a focus onnonagricultural growth, rural households wouldbenefit from increased demand in urban areas, butthe overall effect on poverty would be relativelysmall. Agricultural processing within the manu-facturing sector, however, does in fact represent apotential area for growth and poverty reduction.Within the Agriculture-Led Growth Scenario, aStaples-Led Growth Scenario combined with poormarket access would again have little impact onpoverty among small-scale farm households. AnExport-Crop-Led Growth Scenario would favorrural medium-scale households due to market andcredit constraints for the rural poor.

Agriculture and MDG 1: Regional Analyseswith a Focus on Child Malnutrition

Malnutrition affects nearly one-third of all chil-dren under five years of age in developing coun-tries—174 million children in 1990. More than halfof childhood deaths are associated with under-weight, and malnourished children who surviveinto adulthood are more likely to suffer fromchronic illness and disability, and have a higherprobability of reduced physical and intellectualproductivity. The IMPACT-WATER model is usedto project the proportion of malnourished childrenunder business-as-usual and an alternative MDGscenario that attempts to close the gap between the MDG target rate of childhood malnutritionand business-as-usual outcomes. The gap is closedthrough growth in agriculture as well as comple-mentary investments in social sectors, and specialattention is given to those countries and regionsleast likely to reduce malnutrition significantly byother means. Under business-as-usual, the devel-oping-country level of childhood malnutrition isstill 24 percent by 2015, down from 30 percent in1990. Regions least likely to reach the MDG targetindicator on halving childhood malnutrition areSub-Saharan Africa, where the number of mal-nourished children has increased over the last 30 years, South Asia, where substantial progress

was made, but from very high levels, and parts ofSoutheast Asia. Levels in West Asia and NorthAfrica are comparatively low, but have remainedvirtually unchanged over the last 30 years.

The MDG scenario combines two broad coursesfor improving food security and reducing povertyin developing regions: the first way is throughbroad-based and rapid agricultural productivityand economic growth to increase effective in-comes, effective food demand, and food availabil-ity; and the second is through investments ineducation, social services, and health (proxied inthe model by female secondary enrollment rates,the female-male ratio of life expectancy at birth,and access to clean water).

The changes in agricultural and complementarysocial indicators result in a reduction in total childmalnutrition from 24 percent under business-as-usual in 2015 to 17 percent under the MDG scenario,a reduction from 131 million children to 91 million children. Under the MDG scenario, tobring developing countries, particularly South Asiaand Sub-Saharan Africa, within reach of the pre-school malnutrition target indicator, total invest-ments in agricultural and supporting sectors during1995–2015 will have to increase by $161 billion*

based on IMPACT-WATER calculations. The threemain areas of investment for the MDG scenario inpercentage terms are rural roads, irrigation, andeducation. Together these three areas will require$403 billion between 1995 and 2015 to achieve therapid reductions in childhood malnutrition sim-ulated under the MDG scenario. Agriculturalresearch investments account for $109 billion, and$78 billion of investments are required towardincreasing access to safe water. Due to the long lagsin the generation of impacts from agriculturalresearch, increases in research expenditures—evenbeginning now—will have relatively small impactson crop yields by 2015. Increased investments inagricultural research are likely to be essential tomeet crop and animal production needs beyond2015, however. Other investments, such as roadsand irrigation, have significant lags in impact aswell, so implementing the investment portfoliorequired for the MDG scenario will require veryrapid action. As relatively high levels of access toclean drinking water are already achieved in thebaseline scenario, only $15 billion in investmentsare added for the MDG scenario; these investmentshave virtually no lag period in becoming effective.

Role of Trade, Policies, and Governance Systems

In addition to the targeted investments required tohelp agriculture achieve its potential contributionfor the MDGs, policies and governance systemsneed to be supportive for agriculture to achievemaximum impact.

Supportive systems and policies include tradeand domestic support polices for agriculture indeveloped and developing countries, macro-economic reform and public-sector infrastructureand other investments, the role of the private sec-tor and public-private partnerships, and generalgood governance.

Price support policies and border protection ofwealthy Organization for Economic Co-operationand Development (OECD) countries, valued athundreds of billions of dollars each year, causeharm to agriculture in developing countries.Removal of OECD protection could boost ruralvalue added in low- and middle-income countriesby $60 billion per year. The increase in worldprices from removal of OECD protection wouldlead to larger agricultural production in develop-ing countries. An important area for developingcountries to increase participation in internationalmarkets is through the buildup of capacity to pro-duce for exacting standards of importing markets.Developing countries themselves retain substan-tial trade barriers and have varying requirementsregarding agriculture liberalization, due to differ-ences in trade specialization and needs of net foodimports. When developing countries join in agri-cultural trade liberalization, they can achieveoverall welfare gains of $20 billion annually, twicethe gains in national welfare compared to reformsin the developed countries only. Overall, a suc-cessful conclusion for agriculture in the WorldTrade Organization Doha Development Roundtrade negotiations can make an important contri-bution to achieving the MDGs, by establishingsustainable positive incentives for agriculturalproduction among developing countries.

Food aid is another component of internationaltransactions that directly and indirectly affectsrural poverty in a globalized agricultural economyand therefore could have a significant impact inachieving the MDG targets. While the provision ofaid in the form of food is not the optimal form fordevelopment assistance, donors would probablynot provide equivalent cash development assis-

tance in place of food if existing food-aid pro-grams were terminated. Thus, attention needs tofocus on how its effectiveness can be maximizedand potential harms mitigated.

Infrastructure is of particular concern as one ofthe key inputs entering into the “production func-tion” of the MDGs and the achievement of manyof the MDG targets, from poverty reduction toenvironmental sustainability targets. In Sub-Saharan Africa in particular, a lack of adequateinfrastructure, typically attributed to geographyand poor initial conditions, clearly impedes moreproductive agriculture. Achieving the health andeducation MDGs will require more than healthand education interventions; in particular, infra-structure services have a crucial role to play. Pipedwater is crucial to reduce diarrhea in young chil-dren, while electricity allows for more hours ofstudying and road access promotes easier estab-lishment of schools and higher attendance. Wherethe government believes that service should beprovided beyond what a well-functioning marketwill offer, subsidies may be justified to promoteadditional investment to achieve these govern-ment goals.

To improve the effectiveness of public invest-ment, increased coordination at the country,regional, and donor levels is necessary because thelinkages and complementarities of infrastructureinvestment have often not been realized. More-over, the traditional approach of top-to-bottominfrastructure development has to be changed to amore demand-driven approach. Finally, the im-pediments to efficient markets in rural areas needto be addressed through regulatory reforms inorder to increase the availability and effectivenessof resources to address the real access gap in theseareas.

Public intervention alone is not sufficient todeliver the services and investments required toachieve the MDGs. To alleviate rural poverty indeveloping countries, the private sector can con-tribute to economic growth through job creationboth on and off farm. The private sector also hasan important role in supporting timely and effi-cient credit availability to fuel agricultural devel-opment and growth. Finally, the private ruralnonfarm sector is an important engine for ruraldevelopment. In Asia, for example, the rural non-farm economy accounts for 20–50 percent of totalrural employment and 30–60 percent of total ruralincome.

xii Agriculture and Achieving the Millennium Development Goals

Executive Summary xiii

Private financing of investment and MDGs maybe explored as a general way to ensure availabilityof basic services, particularly as the official devel-opment assistance or aid for the water and othersectors has declined in recent years. Financingoptions should also include income redistributionand tax efforts in China, India, and other poorcountries to co-finance the MDGs.

Foreign direct investment (FDI) and other long-term, relatively stable investments have a signifi-cant impact on agricultural and overall economicgrowth. However, the impact of this investmenton hunger and poverty may be limited. Little ofthis investment goes directly into agriculture orrural areas. Moreover, FDI is largely concentratedin just a few countries.

Ideally, the public and private sectors comple-ment each other, with the government providingan appropriate enabling environment for privateinitiatives to develop. Public-private partnershipsare an important way to increase financial, human,and social capital in rural areas. Partnerships caninclude publicly provided training for small- andmedium-scale enterprises, partnerships in educa-tion, agricultural research, the provision of infor-mation and communication technologies, theexpansion of rural infrastructure including roads,and the development of rural industrial clusters.Moreover, through partnerships, public researchinstitutions can gain access to advanced scientificknowledge and technologies held by the privatesector, mechanisms for developing, processing,marketing, and distributing final products tofarmers and consumers, and financial resourcesthat are increasingly difficult to obtain. In the areaof agricultural research, however, a sustained pub-lic role in funding agricultural research will beessential, particularly for crops and regions thatare unlikely to be served by the private sector,such as those in less favorable environments.

Good governance is typically defined under theterms of accountability, transparency, predict-ability, and participation. These principles are onlymeaningful and supportive of agricultural devel-opment toward achieving the MDGs if adequateinstitutional and social structures are available.Countries with a good governance structure andadequate institutions tend to ensure political andeconomic stability, possess reasonable state capac-ity, enforce property rights and contracts, providesufficient public goods, and limit government cor-ruption and predation. Conversely, countries with

poor governance and poor institutions typicallyhave poor public services, including those for agri-cultural extension and research, and have particu-larly poor social services like water provision andeducation. Good governance can be built throughthe development of social capital, particularly inrural areas, where participation by individuals insocial networks increases the availability of infor-mation and lowers its cost, helps enforce propertyrights, and reduces opportunistic behavior in nat-ural resource use—thus supporting all MDGs.

Effective community-based organizations, suchas farmer associations or cooperatives, water usergroups, and farm and other microcredit and lend-ing groups can improve governance, for example,by educating and sensitizing the public about theirrights and entitlements under public programs; byacting as a conduit to the government for publicopinion and local experience; by influencing localagricultural development policies; and by helpinggovernment and donors fashion a more effectivedevelopment strategy through strengtheninginstitutions, staff training, and improving man-agement capacity. At a higher level, donor coun-tries and international aid agencies should focustheir resources on those countries where good orimproving governance structures will ensure thatthe rural poor are reached. Finally, at the interna-tional level, enhanced governance and commit-ment could help bring about improvements in theglobal trade and environmental agendas. Progressin the trade agenda will come from regarding theneeds of the poorest countries, including en-hanced access to both agricultural and other mar-kets; progress in the global environmental agendawill come from the gains of developing countriesfrom enhanced environmental standards, includ-ing protection of remaining biodiversity in thesecountries, or from participating in and gainingfrom climate mitigation policies.

Conclusions

Given that the majority of poor people live in ruralareas or rely on agriculture, and that agriculturepaves the way for economic growth in the poorernations, agricultural and rural development willunderlie progress on the broad array of economicand social indicators emphasized by the MDGs.The most effective strategy for making steady, sus-tainable progress toward the MDGs is to serve allthe goals in an integrated way. However, each goal

will need a well-defined package of technologiesand services for success at the field level.

Of the eight Millennium Development Goals,the first goal is the one whose attainment mostclearly involves the agricultural sector: The pooraround the globe are disproportionately farmersand herders, and, perversely, the hungry also mostcommonly find their livelihoods through agricul-ture. By increasing food availability and incomesand contributing to asset diversity and economicgrowth, higher agricultural productivity and sup-portive pro-poor policies allow people to break outof the poverty-hunger-malnutrition trap. As thecountry-level model simulations revealed, broad-based agricultural growth is the key for decreasingpoverty and increasing growth in Sub-SaharanAfrica. A global assessment of Target 2 of MDG 1(halving child malnutrition levels) shows that thecombination of agricultural and economic growthtogether with larger investments in social sectors,including health and education, can substantiallynarrow the gap between the business-as-usual out-comes for 2015—24 percent of developing-countrypreschool children malnourished—and the targetindicator—15 percent children malnourished—toreach 17 percent. However, the outcome varies sig-nificantly by country and region. Latin America,West Asia and North Africa, and China will, onaverage, likely get close to the target indicator by2015, even under business-as-usual; however, thelikelihood that Sub-Saharan Africa and South Asiawill come close to their respective target rates ismuch smaller. The total increase in investmentsestimated is $161 billion in agricultural and sup-porting sectors during 1995–2015. In addition tothese investments, significant policy and gover-nance reform is required.

To achieve faster agriculture-based growthrates, there must be in place favorable macro-economic and trade policies, good infrastructure,and access to credit, land, and markets. These con-ditions create level playing fields and give farmersincentives to adopt new and sustainable technolo-gies and diversify production into higher-valuecrops, actions that raise incomes and lift house-holds out of poverty. An improved domestic regu-latory framework would intensify competitionamong suppliers of essential inputs, such as seedsand fertilizer. In addition, the elimination of tradebarriers for agricultural products, especially thehigh-value-added products, would encourage a

greater number of private entrepreneurs toexplore opportunities in agribusiness. A healthymarket and private sector would provide value-added, skilled work to the landless poor and gen-erate multiple livelihood opportunities in both thefarm and nonfarm sectors. Other importantreforms in the trade area include the elimination ofexport subsidies; the move toward measures thatsupport income instead of stimulating productionin developed countries, and the reform of theinternational and national governance of food aidprograms.

As the financing requirements for realizing theMDGs are substantial, the private sector is increas-ingly called upon to fill investment gaps. Its com-plementary and supporting role in the provisionof basic services in water, land, health, and otherinfrastructure development that is lacking in most developing countries cannot be ignored. Itwill take a particular kind of private-sectorinvolvement to generate the necessary economictransformations. Private entrepreneurs are nowincreasingly held to environmental, social, andcorporate governance principles that stress sus-tainable business practices and adherence to laborstandards. Without these standards and practices,the private sector and disadvantaged groups cannot mutually benefit from consumer, employ-ment, and entrepreneurial activities. Governmentagencies in developing countries urgently need to revisit the legal, regulatory, political, and in-stitutional framework in agriculture, research,extension, and industrial sectors to facilitate pri-vate-sector involvement. Moreover, both theprivate and public sectors must foster private-public–sector partnerships and cultivate this rela-tionship with the end objective of addressing theMDGs. Moreover, to allow the agriculture sectorto develop its full potential for achieving theMDGs, the share of ODA spent on agricultureneeds to increase significantly.

It is a promising development that the review ofprogress—and lack thereof—in achieving theMDGs has reached global attention. Calls foraccountability and action that have real impact onpeople are growing because of that attention.Policy action and increased investment in the crit-ical arenas of sustainable agriculture productivityand food and nutrition security will be essentialfor responding effectively and responsibly toreach the Millennium Development Goals.

xiv Agriculture and Achieving the Millennium Development Goals

1

Today, 1.1 billion people in the world are living on less than $1* perday—430 million in South Asia, 325 million in Sub-Saharan Africa,260 million in East Asia and the Pacific, and 55 million in Latin America(World Bank 2004a). Although in aggregate, human material condi-tions have improved greatly over the past century, the continueddaily deprivation of the basic needs of millions of people testifies tothe fact that the global challenge of human development is far frommet. Too many children around the world today lead lives character-ized by hunger, illness, and, all too often, early death.

In order to establish a renewed commitment to human develop-ment, in 2000 the member states of the United Nations adopted theMillennium Declaration. Emerging from the Millennium Summit, thedeclaration recommits the community of nations to a broad range ofsteps to lead to a “more peaceful, prosperous and just world” and “tofree our fellow men, women and children from the abject and dehu-manizing conditions of extreme poverty . . . to making the right to de-velopment a reality for everyone and to freeing the entire human racefrom want.” In contrast to previous global policy statements, theMillennium Declaration also established eight goals—the MillenniumDevelopment Goals (MDGs)—each with quantified targets, to moti-vate the international community and provide a mechanism for ac-countability in undertaking action to enable millions of poor people tolive lives of dignity, free from extreme want.

The goals are:

1. Eradicate extreme poverty and hunger2. Achieve universal primary education3. Promote gender equality and empower women4. Reduce child mortality5. Improve maternal health6. Combat human immunodeficiency virus/acquired immune de-

ficiency syndrome (HIV/AIDS), malaria, and other diseases7. Ensure environmental sustainability8. Develop a global partnership for development.

For each goal, one or more targets have been set, most for 2015, using1990 as a benchmark. The goals, targets, and specific indicators aredescribed in Appendix 1 of this document.

Since the Millennium Summit, the MDGs have come to serve askey objectives in guiding the planning and implementation of a broad

Introduction

Today, 1.1 billion people

in the world are living on

less than one dollar per

day and most depend on

agriculture for their

livelihoods.

1

*All dollar amounts are U.S. dollars.

range of global and national development efforts.Progress toward the attainment of the MDG targetsis being monitored, and policies and programs arebeing modified to bring about the changes neededto attain those targets.

The vast majority of people whose lives need tochange the most, in order to attain the targets speci-fied in the MDGs, depend on agriculture for theirlivelihoods. Coming up with strategic options for as-sisting these individuals and their households is anecessary component for improving global perfor-mance in meeting the MDGs. However, there arelittle or no diagnostics of the direct and indirect linksbetween the MDGs and agriculture. Moreover, someof the indirect effects will come about through broad,economy-wide processes, while others, particularlythose operating at the household and communitylevel, will be felt more immediately. Additionally,cause-and-effect relationships between the agricul-tural sector and the MDGs are not all one way. Whilethe agricultural sector provides critical inputs for at-taining the MDG targets, the broad improvements in

human capital needed to reach those targets will alsoresult in a considerably more productive and resilientagricultural sector. There is, therefore, a need toundertake a critical review of the pathways by whichagriculture can contribute directly or indirectly toattaining the MDGs.

Following an explanation of the MDGs and a dis-cussion of their progress to date, we assess goal-by-goal the potential contributions of agriculture toachieving the MDGs in the second chapter. Thethird chapter examines more closely how the targetsof MDG 1 can be achieved in Sub-Saharan Africa(Ethiopia and Zambia), and how agricultural andeconomic growth, together with larger investmentsin social sectors including health and education, cansubstantially narrow the gap between business-as-usual and MDG outcomes for Target 2 of MDG 1(halving child malnutrition levels). This report thenexamines how trade and macroeconomic policiesfor agriculture can improve the attainment of theMDGs, particularly the goal of halving poverty. Thefinal section sets out some conclusions.

2 Agriculture and Achieving the Millennium Development Goals

3

About 70 percent of those living on less than $1 a day live in ruralareas (World Bank 2002).1 For the majority of the poor in the world,agriculture is a critical component in the successful attainment of theMDGs. Although the rural poor pursue a range of strategies to assuretheir livelihoods, the dominant strategy is food production throughcropping or raising livestock.2 The vast majority of people whoselives need to change the most to attain the targets specified in theMDGs are farmers and herders. The material well-being of theseindividuals and households is dependent upon the productivity oftheir cropping and livestock husbandry activities. As shown in figure2.1 the poorest countries of the world are also those in which agricul-ture is the predominant sector of employment. To a large degree, thepoverty experienced in these countries is a product of unproductiveagriculture. Moreover, the dominance of agriculture in the economiesof the poorest countries (as shown in figure 2.2) often is more a reflec-tion of a poorly performing economy in which subsistence agricultureis serving as a safety net of last resort for populations with limitedeconomic options, rather than as an effective engine of economicgrowth. Although in the longer term a broad transformation anddiversification of rural economies away from a strong dependence onagriculture is desirable, more immediate gains in the welfare of poorhouseholds are most likely to come through the poor overcomingsome of the critical constraints they now face in meeting their basicneeds through agriculture. Thus, a necessary component in meetingthe MDGs by 2015 in many parts of the world is a more productiveand profitable agricultural sector.

In this chapter, we will make a broad assessment on an MDG-by-MDG basis to consider how more productive agricultural activitiesand a more vibrant agricultural sector in general might significantlyadvance a country’s efforts to attain each MDG. How might improvedagricultural performance at household, community, and economy-wide levels bring progress toward the MDGs? A summary of thediscussion on how the agricultural sector can contribute to the attain-ment of the MDGs, as well as how, in turn, progress in attaining theMDGs can animate the agricultural sector is provided in table 2.1 atthe end of this chapter.

Direct and indirect effects need to be distinguished. We will paymore attention to those MDGs that are directly influenced by agricul-tural activities—most notably, MDG 1, halving by 2015 the proportionof those suffering from extreme poverty and from hunger. Moreover,some of the indirect effects will come about through broad economy-

Perspectives on theRole of Agriculture in

Meeting the MillenniumDevelopment Goals

The poorest countries of

the world are those in

which agriculture is the

predominant source of

employment. To a large

degree, the poverty

experienced in these

countries is a result of

unproductive agriculture.

2


Figure 2.1 Proportion of Population Living on Less than $1 Per Day (PPP), Most Recent Estimates

Source: World Bank (2003). Note: PPP = purchasing power parity.

no data

less than 20 percent20 – 3030 – 4040 – 5050 – 60more than 60 percent

Figure 2.2 Agriculture Value Added, as a Percentage of GNP, 2001

Source: World Bank (2003).

no data


Perspectives on the Role of Agriculture in Meeting the Millennium Development Goals 5

wide processes across sectors, while others, particu-larly those operating at the household and commu-nity levels, will be more immediately felt. Of course,for the more localized indirect effects of agricultureto contribute effectively to achieving the MDGs,such effects must be spread widely throughout therural population. Additionally, cause-and-effectrelationships between the agricultural sector andthe MDGs are not all one way. While the agricul-tural sector provides critical inputs to attaining theMDG targets, the broad improvements in humancapital needed to reach those targets will also pro-vide an important foundation from which a consid-erably more productive and resilient agriculturalsector can be developed. Finally, while most MDGtargets are complementary, some might actuallyinvolve tradeoffs. For example, enhanced access toimproved drinking water sources might collide, insome regions, with the goal of reduced hungerthrough increased irrigated agriculture. Similarly,several indicators of MDG 7, ensuring environmen-tal sustainability, might well be adversely affectedby efforts aimed at increasing agricultural and eco-nomic development that are important for theachievement of MDG 1.

MDG 1—ERADICATE EXTREMEPOVERTY AND HUNGERProgress in meeting the targets of the first MDG isencouraging in East Asia, generally adequate inSouth Asia and Latin America, but disturbinglypoor in Sub-Saharan Africa. Based on the simpletrend line shown in figure 2.3, the prevalence ofdollar-a-day poverty in East Asia in 2015 shouldbe considerably less than the 50 percent reductionfrom 1990 levels. In Latin America and South Asia,poverty rates should be close to the targets, ifslightly above.

However, for Sub-Saharan Africa, the propor-tion of the population living on less than $1 perday in 2015 is quite likely to have increased fromthat of the reference year 1990. The regional trendsin reduction of undernutrition, while all movingdownward, are not as clear-cut as the trends inpoverty. The undernutrition target is quite likelyto be met in East Asia and in Latin America (seealso the global assessment of childhood malnutri-tion in chapter 4). However, in both South Asia andSub-Saharan Africa, undernutrition will likelyremain at levels considerably above the targets for

Figure 2.3 Progress in Attaining MDG 1 Targets for Poverty and Undernourishment, by Region (Trend Lines Based on Recent Performance and Targets for 2015)

Source: World Bank (2004b) (left), and FAO (2003) (right).

0

10

20

30

40

50

1990 1995 2000 2005 2010 2015

Shar

eof

peop

leliv

ing

onle

ssth

anU

S$1

ada

y(%

)

Sub-Saharan AfricaSub-Saharan Africa

South Asia

South Asia

East Asia & Pacific

East Asia & Pacific

Latin America & Caribbean

Latin America & Caribbean

MDGTargets

0

10

20

30

40

1990 1995 2000 2005 2010 2015

Prev

alen

ceof

unde

rnou

rish

men

t(%

)

2015. Indeed, in Sub-Saharan Africa, using recentperformance as a guide, little progress in reducingundernutrition will have been made.

We have already observed that poverty has astrong rural and, hence, agricultural dimension:Approximately 70 percent of the world’s poor live inrural areas and primarily pursue agriculture-basedlivelihood strategies. Hunger is endemic in mostrural areas of the developing world. Consequently,as shown in figure 2.4, malnutrition levels are con-sistently higher in rural areas than in urban zones.

Improving the productivity of and the economicreturns of agriculture for farming households willhave immediate effects in eradicating extremepoverty and reducing hunger. First, through the

market, increased agricultural income will directlyimprove both household consumption levels today,and household asset levels to improve productionand better weather economic shocks in the future.Increased food production will lead to real reduc-tions in food prices, which will improve the pur-chasing power of the poor throughout the economy,whether they are engaged in agriculture or someother sector. More important, agriculture can serveas the basis for broad pro-poor economic growthto bring about permanent reductions in poverty.Second, with complementary nutritional factors inplace, both subsistence farming households andthose purchasing food in local markets will enjoyimmediate physiological benefits from increased

6 Agriculture and Achieving the Millennium Development Goals6 Agriculture and Achieving the Millennium Development Goals

Figure 2.4 Child Malnutrition by Urban or Rural Residence, Stunting (Low Height for Age) Prevalence AmongPreschoolers, Surveys Since 1999

Source: ORC/Macro, 2004. MEASURE DHS+ STATcompiler. http://www.measuredhs.com. Most recent survey. Accessed May 2004.

0 10 20 30 40 50

Kazakhstan

Armenia

Turkmenistan

Nepal

Cambodia

Bangladesh

Colombia

Peru

Haiti

Egypt

Gabon

Guinea

Mauritania

Mali

Tanzania

Benin

Zimbabwe

Burkina Faso

Rwanda

Ethiopia

Uganda

Malawi

Zambia

Nigeria

Percent

RuralUrban


food production. As the first MDG is the one inwhich the impact of a more dynamic agriculturalsector will be felt most directly, we will discussin more detail these two issues—poverty reduc-tion through agriculture-led economic growth andimproved nutrition through agriculture.

Poverty Reduction Through Agriculture-LedEconomic Growth

Income redistribution and economic growth are thetwo economic mechanisms for reducing poverty.While redistribution is a viable option to addresspoverty in many parts of the developed world, inthe developing countries of the world with largesegments of populations unable to meet their basicneeds, income redistribution policies are unlikelyto have much effect on general welfare levels. Thereare simply insufficient resources in such economiesto assure the basic needs of all. In these countries,higher economic productivity—that is, economicgrowth enabling increased employment and risingwages—is the only means by which the poor willbe able to satisfy their needs sustainably.

For both long-developed countries and for thehandful of more recently developed countries, sig-nificant increases in agricultural productivity werea critical early step in building sustained economicgrowth. Initial growth in staple food production bythe small-scale, labor-intensive agricultural sectorthrough the use of improved technologies resultedin reduced food prices, increased real wages, and,consequently, lower poverty. Reduced food pricesenabled greater access to food, resulting in betternutrition for the general work force while also free-ing up additional household resources from foodto other expenditures, including productive invest-ments. In rural areas, investments initially wentinto cash crop production and agricultural pro-cessing activities, but, as the economy grew, ruralnonfarm and urban activities became increasinglyprofitable. As this process of economic transfor-mation advanced, the agricultural sector tended toplay a decreasing role in sustaining economicgrowth.3 Movement of labor out of the agricul-tural sector occurs as employment opportunities inother economic sectors grow. Depending on ruralpopulation growth, larger-scale commercial enter-prises may become more characteristic of the agri-cultural sector.

There is considerable empirical evidence thatdemonstrates the significant contribution thatgrowth in agricultural productivity can make to

reducing poverty. A recent cross-country analysisby Thirtle and others (2002) found that, at thenational level, a 1 percent increase in agriculturalyields decreases the percentage of population livingon less the $1 per day by 0.64 to 0.91 percent, witha slightly higher reduction for the countries inAfrica. Notably, in analyzing the effect of growth inthe manufacturing and service sectors on poverty,no significant change in poverty was associatedwith growth in these sectors.

This description of the process is necessarilysimplistic and ignores a broad range of potentialimpediments to agriculture-led economic growth.One of the most critical of these at present is lowprices for staple food crops globally, due to the suc-cess of the Green Revolution (as well as high levelsof subsidies in the developed world). As a con-sequence, it now is difficult to generate profitsthrough staple food crop production even at highlevels of productivity, particularly for smallholderswho are unable to achieve economies of scale intheir production. Increasingly, in order to obtainan adequate return from their efforts in agriculture,farmers need to diversify their production intohigh-value, but knowledge-demanding, specializedcrops. Similarly, agriculture cannot be expected tobe an engine of economic growth for those coun-tries that have no comparative advantage for agri-cultural production or face significant barriers toproducing for global markets. Many of the drierinland countries of Africa face important challengesin this regard.

The initial distribution of agricultural assets inthe economy is also a critical feature in whether ornot agriculture-led growth will reduce poverty.Where land ownership is concentrated, as mostnotably in parts of Latin America, such economicdevelopment is unlikely to reduce poverty greatly(Timmer 2003). In contrast, in most of Africa andmany parts of Asia where the poor continue to haveaccess to land, agriculture-led growth should leadto significant reductions in poverty. Finally, wheresignificant movement of the population out of ruralareas to the cities and out of agriculture into urban-based economic sectors has already occurred, as inmany of the middle-income countries of the devel-oping world, there is little potential for the agricul-ture sector to catalyze broad economic growth andreduce poverty. In these countries, poverty reduc-tion efforts should focus on the industrial and ser-vice sectors, while not ignoring the key role thatagriculture plays in their rural economies.


While these caveats are important to recognizein tailoring economic development strategies, nev-ertheless, such a process of agricultural growth hasproven in the past to be a common means by whichto spur broad-based pro-poor economic growth.Yet, for the poorest countries, the process of small-scale agriculture-led economic growth leading tostrong economies and minimal poverty will notoccur by 2015, the target date for most of the MDGtargets. For these countries, the relevant time frameis one of decades and generations. However, aprocess of building sustained economic growthrequires that productivity increases in agriculturebe achieved in the early stages. Moreover, the ben-efits of the initial steps of agricultural productiv-ity increases accrue primarily to the farmers andherders, among whom the poor are concentrated.By pursuing such an economic growth strategywithin the context of the MDGs, we can achieveboth the significant improvements in well-beingthat the MDGs seek to promote, and lay the founda-tion for the sustainable economic transformationsneeded to attain, in the longer term, the broader aimof the Millennium Declaration of “freeing the entirehuman race from want.”

Only a few countries have dramatically reducedpoverty and achieved rapid economic growth with-

out significantly increasing the productivity of theagricultural sector. However, agricultural produc-tivity gains alone are not sufficient to bring aboutsustained economic growth. No country has beenable to sharply reduce poverty only through agri-cultural strategies. Institution building in the agri-cultural sector and parallel developments in othersectors of the economy are needed to transform thefoundational contributions from the agriculturalsector into sustained broad economic growth in theeconomy. Agricultural strategies alone will not leadto success. However, the converse also applies: Forthe poorest countries, economic growth and sus-tained poverty reduction are unlikely to be achievedwithout initially stimulating sustained agriculturalproduction growth.

Agriculture’s Contribution to Food and Nutrition Security

The indicators for the second target of the first MDGinclude the reduction by half between 1990 and2015 of the prevalence of underweight (low-weight-for-age) children and the proportion of the popula-tion whose food intake falls below the minimumlevel of dietary energy requirements (undernutri-tion). As shown in figure 2.5, the levels of under-

8 Agriculture and Achieving the Millennium Development Goals8 Agriculture and Achieving the Millennium Development Goals

Figure 2.5 Prevalence of Undernourishment—Proportion of the Population Unable to Acquire Sufficient Caloriesto Meet their Daily Caloric Requirements, 2003 Estimates

Source: FAO (2003).

no data



nutrition are high throughout the developing world,and particularly in many of the countries in whichagriculture is also the dominant livelihood. To attainthe MDG 1 second-target goals, food and nutritionsecurity needs to be enhanced for the poor. A house-hold is food secure if it can reliably gain accessto food in sufficient quantity and quality for allhousehold members to enjoy a healthy and activelife. It is possible, however, for individuals in food-secure households to have deficient or unbalanceddiets. Nutrition security is achieved when secureaccess to sufficient, safe, and nutritious food iscoupled with a sanitary environment, adequatehealth services, and knowledgeable care to ensure ahealthy and active life for all household members.

Clearly, food and nutrition security is closelytied to agricultural productivity. Increased foodproduction increases local food availability. Higherproduction from one’s own farm or herds increasesone’s access to food and enhances household foodsecurity. The nutritional quality of the food pro-duced is also an important consideration in reduc-ing malnutrition, particularly for households whoacquire most of their food from their own fieldsand herds. For food purchasers, higher productiongenerally means lower food prices and access to agreater quantity of food in the market for a givenincome level. Particularly in South Asia and Africa,the most potent force for reducing malnutritionis raising food availability through increased agri-cultural productivity, as well as trade (Smith andHaddad 2000, p. 84).

Many countries, particularly in Eastern andCentral Africa, are characterized by a declining orslowly growing food crop sector and very low pur-chasing power. While stable access to food throughthe market requires that the food marketing systemis effective in supplying food while also benefitingthose who have food to sell, the systems in thesecountries are unable to provide effective markets.People living on less than $1 per day are unable topay the prices necessary to import all of the staplefood they require.

Consequently, if hunger is to be addressed effec-tively, a range of complementary actions are neededin addition to those aimed at enhancing crop andlivestock production. An important componentof these actions is the agriculture-led economicgrowth described previously, whereby real incomesand access to food are increased. However, a host ofother institutional factors must be addressed, aswell as several cross-sectoral challenges. The latterare particularly the case in going beyond food secu-

rity to attain nutrition security. To reduce malnu-trition in a comprehensive manner, agriculturalstrategies must be implemented as part of a broaderset of actions that involve the health, water and san-itation, and education sectors (see also chapter 5).

Of the eight MDG goals, the first is the onewhose attainment most clearly involves the agri-cultural sector: The poor around the globe aredisproportionately farmers and herders, and, per-versely, the hungry and undernourished also mostcommonly find their livelihoods through agri-culture. The impact that a dynamic agriculturalsector will have on the attainment of the otherseven goals is less direct. Nonetheless, importantgains in achieving these goals can be made throughexplicit attention to agriculture. We now considerthese other MDGs.

MDG 2—ACHIEVE UNIVERSALPRIMARY EDUCATIONPrimarily in an indirect fashion, investments inagriculture will advance progress toward attainingby 2015 the goal of enabling children everywhere,boys and girls alike, to complete primary school.Higher productivity in agriculture leading to higherincomes will enable either the use of hired labor foragricultural operations or the use of labor-savingtechnologies in place of the labor of school-age chil-dren in farming households. However, a criticalcomponent of this equation is the value that farm-ers perceive that they or their children will obtainby sending their children to school. These benefitsare to a large degree determined by the vibrancyof the economy and the extent to which the highereconomic capacities of trained individuals are re-warded. In stagnant economies and particularly inrural areas where the range of employment oppor-tunities is narrow, perceived returns of educationcommonly are judged to be significantly less thanthe opportunity costs associated with keeping achild in school, and so unable to work full time inthe family’s agricultural enterprise. The broad-based agriculture-led growth discussed earlier isnecessary to alter the outcome obtained by farminghouseholds making this comparison.

Such a pattern of growth, particularly as it extends to the nonfarm and urban sectors, willdemand increasingly skilled labor, and will increasethe returns of investment in the schooling of one’schildren. Moreover, the relationship between in-creased educational attainment and a more active

agricultural sector runs both ways. The agriculturalsector in most poor countries is unlikely to continueto expand for very long on the basis of productiv-ity increases of staple food crops alone. Increas-ingly, the sector will have to turn to the productionof high-value cash crops that usually have quitespecific production and marketing requirements.Meeting the requirements to engage profitablyin their production requires a better-trained workforce. As this sector of the local agricultural econ-omy develops, the returns of providing basic edu-cation to one’s children become considerably morecompelling.

MDG 3—PROMOTE GENDER EQUALITY ANDEMPOWER WOMENThroughout the developing world, women arefarmers and find their principal productive activ-ities in agriculture. Considerable research showsthat when men and women are able to use agricul-tural inputs at equal levels of intensity, women areequally effective as men in profitably engaging inagriculture, being responsive to changing marketconditions in the suite of crops they produce, andeffectively utilizing new technologies (Quisumbing2003). Agriculture provides key contributions to theeconomic empowerment of women.

Moreover, the relationship between agricultureand the empowerment of women works both ways.A dynamic agricultural sector that offers broad wel-fare benefits can be expected to emerge only whenwomen are given the opportunity to participateprofitably in the sector. Where they have securityin their access to productive resources and controlof their agricultural production—that is, wherewomen farmers are empowered to achieve theirfull economic potential within agriculture—the wel-fare effects of a productive agricultural sector canexceed the simple economic productivity measuresfor the sector. The economic empowerment ofwomen both in agricultural production and in othereconomic spheres can be expected to advance sig-nificantly efforts to attain several of the MDGs.

Several of the MDGs are directly determined bythe extent to which sufficient resources are pro-vided to children as they develop and grow. Theimmediate provision of these resources—healthcare, feeding, life skills training, and so on—is inher-ently a gendered task. In most societies, womenare the principal caregivers within the household.

This being the case, if the benefits of a dynamic agri-culture sector are to result in sustained improve-ments in the direct determinants of welfare—income, health, education, among others—it is nec-essary that women have an important role indetermining how the fruits of their agriculturalactivities are used.

There is considerable empirical evidence of theimportance of improving the status of women forimproved general welfare. For example, in a broadcross-country analysis, Smith and others (2003)found that women’s decision-making power rel-ative to men’s was significantly associated withimproved nutritional status of their children. Theyconclude that sustainably improving nutritionalstatus requires proactive efforts to improve the sta-tus of women, particularly in South Asia, but also inSub-Saharan Africa. The authors suggest programsthat will enable women to gain access to new re-sources and promote girls’ education and healthcare, subsidize childcare for working parents, andimprove the nutritional status of adolescent girlsand young women.

Moreover, regarding agriculture in particular,an important dimension in the empowerment ofrural women is alleviating the labor burdens theyexperience so that they can adequately providefor their children’s needs. Domestic time demandsupon women are greater than for men. Agriculturaltechnology developed with close attention to alle-viating some of the labor constraints experiencedby rural women has the potential to improve notonly the well-being of the woman farmer, but alsoof others in her household who are dependentupon her care.

MDG 4—REDUCECHILD MORTALITYThe linkages between agriculture and child mortal-ity are indirect but strong. Agriculture is a criticalcomponent in assuring food and nutrition security.As was described earlier, levels of child malnutri-tion are significantly higher in rural areas than inurban areas. It is estimated that 45–55 percent of allchild deaths are due to malnutrition exacerbatingthe negative effects of disease on a child’s health(Pelletier and others 1994). As a consequence ofpoor nutrition security in rural areas, mortalityrates for children under five years of age are signif-icantly higher among rural children than for theirurban counterparts (figure 2.6). Poor nutrition secu-


rity is poor food security coupled with poor accessto quality health services, lower general knowledgeof proper feeding and management of childhoodillnesses, poor sanitation, and unprotected watersources.

A productive rural economy offering sufficientemployment and rising wages is a necessary com-ponent in any effort to reliably and durably reducethe number of children dying in a rural community.Consumption poverty is an important part of theexplanation for why child mortality rates are so highin many parts of the developing world. If efforts toattain MDG 1 by following a broad-based, agricul-ture-led strategy of economic growth bear fruit, weshould see a parallel decline in child mortality rates.

However, poverty does not fully account forchild mortality. Care is also a critical element inreducing child deaths. As women are the primarycaregivers, MDG 3 is relevant here. Time andknowledge are both critical constraints in thisregard. Women need to have the time and theknowledge to appropriately meet the survivalneeds of their children. Education, both the formaleducation sought through MDG 2 and informaleducation from peers and public health services,must be provided to a child’s parents so parentscan take appropriate action to assure the child’spotential to live a healthy and active life.

The economy-wide effects of a dynamic agricul-tural sector can help reduce child mortality, because


Figure 2.6 Child Mortality by Urban or Rural Residence, Surveys Since 1999

0 50 100 150 200 250

Colombia

Peru

Dominican Republic

Haiti

Armenia

Kazakhstan

Nepal

Turkmenistan

Cambodia

Bangladesh

Egypt

Zimbabwe

Gabon

Uganda

Nigeria

Burkina Faso

Benin

Zambia

Rwanda

Tanzania

Malawi

Ethiopia

Guinea

Mali

Deaths of children under five years of age per 1,000 live births

Source: ORC/Macro, 2004. MEASURE DHS+ STATcompiler. http: www.measuredhs.com. Most recent survey. Accessed May 2004.

RuralUrban

more funding will be available for the public provi-sion of medical care, health-care facilities, sanitationand clean water, and public health interventions.While one should not expect that agriculture-ledeconomic growth necessarily will quickly provideeconomic surpluses that can be invested in thisway, making the investments needed to reduce thenumber of children who suffer and die must be apriority as surpluses are generated in the economy.

MDG 5—IMPROVE MATERNAL HEALTHAgriculture can contribute to the goal of reducingmaternal mortality in a way similar to its contribu-tion to attaining the previous three goals. Insofar asagriculture can contribute to the economic empow-erment of women and enable them to participatebetter in decision making in their households andcommunities, women will have greater ability topay attention to their own physical well-being andhave access to increased resources to assure theirown good health.

Agriculture can contribute to improved maternalhealth in another way. For the most part, the qual-ity of the food produced by farmers is assumed tobe irrelevant to how agricultural activities affect

broad nutrition security. Yet, agriculture has consid-erable potential to directly improve maternal healthby improving the diets of both rural and urbanwomen, as well as the other household membersfor whom they are responsible. Micronutrient defi-ciencies are particularly severe among young chil-dren and women. Among women, the health effectsof such deficiencies, particularly of dietary iron, aremost pronounced during pregnancy, at birth, and inthe months following birth. As shown in figure 2.7,the spatial pattern of the prevalence of iron defi-ciency anemia at the national level parallels similarpatterns of poverty and malnutrition. It is estimatedthat more than 65,000 women die annually due tosevere anemia.

By increasing the micronutrient content of foodcrops, deficiencies among women and childrenshould decline, and maternal and child mortalityrates drop. With this goal in mind, a global, inter-disciplinary research program within the Consult-ative Group for International Agricultural Researchseeks to increase the nutrient density of many globalstaple food crops, particularly by increasing thelevels of bio-available iron, zinc, and vitamin A thatthe crops contain. Although this research effort isjust beginning, if such traits can be bred into high-yielding varieties that enjoy wide consumer accept-


Figure 2.7 Estimated Prevalence of Iron Deficiency Anemia in Women Aged 15 to 49, Most Recent Estimates

Source: Micronutrient Initiative & UNICEF (2004).

no data



ability, farmers would be able to offer a direct,low-cost, and sustainable way to improve the nutri-tional status of millions around the world and, inparticular, improve maternal health.

MDG 6—COMBAT HIV/AIDS,MALARIA, AND OTHER DISEASESEfforts made to reduce child and maternal mortal-ity levels to which agriculture contributes will alsocontribute to effectively combating HIV/AIDS,malaria, and other diseases. Although the indica-tors for this goal deal very little with agriculture,and the direct links between agriculture and thesehealth issues are not immediate, they are impor-tant nonetheless. For agricultural households, asfor all households, the productivity of their eco-nomic activities is an important determinant ofwhether people live in an environment that allowsthem to enjoy a healthy and active life and canacquire the health care required to do so.

However, a newly dynamic agricultural sectoralso has the potential to radically alter the diseaseenvironment in a region. For example, wage-labormigration associated with agriculture may exposepopulations to new diseases, increasing the healthburdens they bear. Changes in local water manage-ment for irrigation may alter local disease ecologies,particularly for malaria and water-borne diseases.New health challenges likely will emerge with anevolving agricultural sector. The resources of theagriculture sector, particularly in the public spherethrough extension services, can be used in a coor-dinated fashion with those of the health sector toaddress such issues.

The agricultural sector has an important role inaddressing the important challenge to human dev-elopment of HIV. Of the 25 countries in the worldwith an adult HIV infection level above 5 percentin 2001, all except two have predominantly ruralpopulations. The long and fatal pathway of chronicillness with HIV infection, particularly in youngadults, severely compromises the welfare of farm-ing households. Loss of labor power and farmingknowledge, and increasing nutritional require-ments as the disease progresses, are among the most salient effects within the household (AIARD2003). Professional agriculturalists—researchersand extension workers in particular—must tailortheir work to assist such households to bettermeet the particular farming challenges they face.Households that are able to modify their practices

appropriately to maintain sufficient levels of pro-ductivity in their agricultural activities will be betterable to cope with the burdens of caring for HIV-positive members. Agricultural support servicestaff, working in close coordination with staff fromthe health and education sectors, have a role toplay in stopping the transmission of HIV in farmingcommunities. However, the challenges in accom-plishing this are immense.

Moreover, the agricultural economy as a wholesuffers from a heightened incidence of HIV infec-tion. Declining aggregate agricultural productivitywill result from insufficient labor as farmers andherders fall ill or must devote their time to house-hold members who are ill. Savings will be depletedto meet increased health care costs and, conse-quently, needed inputs for profitable agriculturalproduction will be inaccessible to many farminghouseholds. Any economic gains that might havebeen made by the sector will be negated. In a ruralpopulation ravaged by HIV, there is little scope foragriculture providing the lead in building strongeconomic growth. Although agricultural practicescan be modified and support services provided toincrease the resilience of farming households suf-fering from HIV/AIDS and strengthen householdsthreatened by the disease, the scope of the epidemicis such that it poses a major threat to the alreadytenuous welfare of the poorest, most agriculturallydependent populations.

MDG 7—ENSUREENVIRONMENTALSUSTAINABILITYThis goal covers a broad sweep—biodiversity, crit-ical natural habitats, energy use and global climatechange, unsafe water and poor sanitation, andurban slums. Agriculture is implicated both as ameans to effectively address many of these prob-lems, and as a source of and a contributory factorto the problems that MDG 7 was formulated toaddress. It is unlikely that one can develop the agri-cultural sector in such a manner that only benefitsand no negative externalities accrue. In this light, ajudicious, comprehensive, and participatory assess-ment of the environmental costs and benefits mustbe undertaken in the planning process for any agri-cultural development efforts.

A productive agricultural sector will reduce pres-sure on and contribute to ensuring environmental

sustainability in most of the areas considered byMDG 7. In particular:

• Productive agriculture requires less land perunit yield, leaving marginal agricultural landsto other uses, including forests and other crit-ical habitats.

• Proper agricultural policies will allow the fullcosts of agricultural technologies, includingtheir costs for the environment, to be con-sidered as they are being used. Policies thatinduce transparent assessment of these costswill, for example, reduce the scope for exces-sive nutrient runoff from agriculture, provideincentives for efficient energy and water use inthe sector, and enable the ecologically sustain-able use of a range of technologies, includingpesticides and genetically modified organisms.

• As agriculture is inherently an organic, carbon-based enterprise, the sector is a potentiallyimportant component in any systems estab-lished to manage global carbon stocks.

• A dynamic agricultural sector fostering broad-based economic growth, as with any economicexpansion, should provide additional publicrevenue to enable greater levels of public pro-vision of safe drinking water and improvedsanitation. Increasingly productive and prof-itable farmers in many developing countrieswill be able more and more to provide theseamenities privately for their own households.Moreover, water infrastructure in agricul-ture, whether for irrigation or flood control,will also have important applications forthe provision of safe water and adequatesanitation, particularly in small-scale agricul-tural systems.

• Population pressures in urban slums will bealleviated to a significant degree if profitableagricultural systems are developed in the ruralhinterlands. Although broad agriculture-ledeconomic growth should lead in time to a sig-nificant movement of workers out of rural,agricultural occupations and into the manufac-turing and service sectors located predomi-nantly in urban centers, such migration will beof a different quality than that most commonlyseen at present. Today what is frequentlyobserved is an unproductive agricultural sec-tor that forces many farmers and herders tosearch for employment in urban centers thatare only marginally more attractive in termsof economic opportunity. Moreover, enhanced

urban agriculture can contribute to the in-come growth and nutritional status of slumdwellers.

However, agriculture can also exacerbate envi-ronmental degradation, fuel perverse rural-to-urbanmigration, and deepen poverty. For example:

• Agricultural expansion is a principal factorcontributing to tropical deforestation andincreasing global levels of atmospheric carbondioxide. The underlying causes for agricul-tural expansion are case dependent. Some-times expansion arises from acceleratedgrowth in the agricultural sector (capital-driven), and at other times from poverty aspoor farmers seek land with which to meettheir basic needs (Geist and Lambin 2002).Agriculture-led economic growth in itself isnot environmentally benign.

• A stagnant agricultural sector with low pro-ductivity and profitability will result in theunsustainable use of natural resources for agri-culture. Most commonly this is seen in themining of soil fertility down to a base state inwhich cereal yields, even under suitable agro-climatological conditions and adequate laborinputs, attain only a few hundred kilogramsper hectare. Such environmental exploitationin agriculture can take many other forms,affecting a broad range of the components oflocal ecosystems.

• Poverty can actually increase due to theexpansion of the agricultural sector of theeconomy if the distribution of agriculturalassets—land, in particular—is skewed. If agri-cultural expansion is accomplished in a man-ner in which capital substitutes for labor andlittle growth in employment occurs, aggre-gate welfare will be little improved. More-over, further consolidation of land may occur,resulting in greater pressures on the rural poorto migrate to urban slums.

With appropriate regulatory institutions in placeto safeguard the benefits that society as a wholedraws from the environment, an emergent agri-cultural sector need not lead to environmentaldegradation. The underlying driving force forenvironmental degradation through agriculturalexpansion and the harmful use of farming tech-nologies is frequently poverty, rather than factorsinherent to agriculture itself. If farmers realize per-sonal economic benefits and also recognize the



social benefits from environmental protection, theywill respond to these incentives and employ envi-ronmentally sustainable production techniques.However, sustainable agricultural practices mustbe profitable for this to happen. Whether or not thisis the case depends on the vibrancy of agriculturewithin the overall economy.

MDG 8—DEVELOP A GLOBALPARTNERSHIP FOR DEVELOPMENTAlthough extremely broad in scope, the final MDGhas important implications for agriculture. Thereare several areas where agriculture can contributeto efforts aimed at achieving the targets set.

First, it is under this goal that an open, rule-based, predictable, nondiscriminatory trading andfinancial system is called for. Given the demonstra-ble centrality of the agricultural sector to peoplewhose conditions of life must change if the MDGsare to be attained, agriculture should be among thefocus sectors of initial efforts in this area. Globalagricultural trade must be harmonized and ratio-nalized in a manner that includes consideration ofthe special needs of poor agricultural producersand how they might derive maximum benefit fromsuch trade.

However, even with attention to these importantaspects, the poverty impact of globalized agricul-tural trade remains unclear. Such trade will oftenexclude the smallholder farmer and herder, as itrequires knowledge, capital, and quality assurancelevels, as well as access to marketing networks thatmost smallholder producers cannot attain on theirown. Smallholders may be uncompetitive and un-able to participate in many of the most profita-ble subsectors under a wholly free-trade system.Establishing appropriate institutions is necessary toenable broad welfare gains to be achieved throughtrade. These issues will be addressed further inchapter 4’s discussion of trade and macroeconomicpolicies.

Second, the Heavily Indebted Poor Countries(HIPC) initiative contributes to the targets underthis goal. Poverty Reduction Strategy Papers (PRSP)have been prepared by many of the HIPC countriesto demonstrate to their development partners howthe funds made available through debt relief wouldbe used to reduce poverty. The PRSPs potentiallyare a very effective means by which progress can bemade in achieving the MDGs. It is critical that theframework within which these PRSPs are devel-

oped provide due and relatively detailed attentionto the economic foundation for most of the poorpeople at whom they are targeted—agriculture.Most current PRSPs highlight the need for broad-based economic growth, and such growth is typi-cally one of the four or five “pillars” of most PRSPs.Moreover, agriculture is frequently noted as beingthe most important livelihood for the poor and isprioritized as a key economic sector.

However, the means by which agriculture willlead to broad-based economic growth is frequentlyleft unspecified in the documents. Indeed, a keycriticism of the implementation of the PRSPs hasbeen that social expenditures tend to be givenpriority over the investments, particularly in theagricultural sector, that would accelerate economicgrowth (Gautam 2003). Given the MDG’s strongsocial focus, they perversely provide additional jus-tification for countries privileging social expendi-tures to the neglect of investments that would leadto sustainable pro-poor economic growth. A morebalanced expenditure pattern is needed. In mostHIPC countries that have prepared PRSPs, con-siderably more effort must go into framing themechanisms by which agriculture will bring aboutthe desired improvements in welfare. For exam-ple, Uganda has developed the relatively detailedPlan for Moderniation of Agriculture (MAAIF &MFPED 2000) as the principal cross-sectoral eco-nomic development strategy emerging from thePoverty Eradication Action Plan (MFPED 2000),Uganda’s Poverty Reduction Strategy Paper.Many more HIPC countries need to do likewise.

Measures are currently being undertaken to re-orient the mechanisms for the delivery of bilateraland multilateral official development assistance tobe consistent with efforts to achieve the MDGs.MDG 8 specifies several related to official develop-ment assistance. The sectoral allocation of suchassistance is not specified in these targets. How-ever, clearly this is not a tangential issue. If agri-culture is to be effective in broadly improvingthe human condition, particularly that of rural res-idents, considerably greater levels of resourcesneed to be made available to agricultural develop-ment. Currently, levels of assistance to agriculture,as well as budgetary allocations by governmentsthemselves, are inadequate.

For example, the Comprehensive Africa Agri-culture Development Programme strategy of theNew Partnership for Africa’s Development(NEPAD) proposes investments of $251 billion overthe period 2002 to 2015, or just under $18 billion per

year, to reduce the incidence of hunger and raisefarm output (NEPAD 2002). Such a budget facessome stark constraints. Notably, estimated totalannual government expenditures on agriculture inAfrica in the late 1990s were roughly $6.2 billion(FAO 2001; World Bank 2003). Moreover, throughthe 1990s the major bilateral and multilateral donorsannually committed globally only about $8 bil-lion to agriculture, broadly defined (FAO/IFAD/WFP 2002). The agricultural development needsof Africa alone cannot be met under current levelsof official development assistance, and the humanneeds that could be met through agriculture extendmuch beyond Africa alone. If the poor of the worldare primarily farmers and herders and we want tosee sharp improvements in their well-being in thenear term by 2015, then important gains can bemade if we start with where they are currently

earning their livelihoods—in agriculture. Donorpriorities should reflect this basic element of theglobal poverty profile.

Finally, the roots of deficient agricultural devel-opment found in so many poor countries often liesin power relationships in which the welfare of thepopulation is not served, resulting in poor gover-nance, political and social weakness, and adverseincentives. Although the agricultural sector is notblameless, such problems do not lie fundamentallywithin the agricultural sector, but are reflective ofbroader destructive processes within national polit-ical economies. The incentives for bringing aboutan active agricultural sector are not there because ofthese other problems. Little progress in attainingthe Millennium Development Goals or in vitalizingagriculture can be anticipated in countries that areunable to confront these issues.


Table 2.1 Summary of Links Between the Agricultural Sector and the Millennium Development Goals, Principally at Household Level

Complementary Goal Direct Indirect Nature of Relationship Requirements

1. Eradicate extreme poverty and hunger.

• Increased food production

• increased food con-sumption for sub-sistence farminghouseholds.

• More diverse foodproduction

• higher-quality diets.For farming households:• Increased

production• increased income

through markets• increased consump-

tion and householdassets.

For nonagriculturalhouseholds:• Increased production• reduced prices

for agricultural products

• increased consump-tion or reduction inshare of incomespent on food.

• For both farmingand nonfarminghouseholds,increased income

• increased capitalinvestments in existing economicactivities or diversi-fication into othersectors

• enhanced welfareand increasinghousehold eco-nomic resiliency.

• Two-way, quitestrong, generallypositive.

• Less hunger• more productive

workers in agricul-ture.

• Less poverty• more investment in

agriculture.

• Suitable agriculturalproduction tech-nologies available.

• Relatively equitabledistribution of farm-land across the population.

• Efficient, widespreadrural markets that are linked toregional and international tradecircuits.

• Knowledge onproper diet andnutritional care.Sanitation andhealth servicesavailable.

(continued )


Table 2.1 Summary of Links Between the Agricultural Sector and the Millennium Development Goals, Principally at Household Level (Continued)


• Two-way, princi-pally indirect.

• Possibly some nega-tive ramifications ifincreased returnsfrom agriculture canbe achieved usingchild labor orhigher skills are notrequired.

• Two-way.• Increased willing-

ness of women toinvest in agriculture

• more dynamic agri-cultural sector.

• Possibly negativeramifications ifmore dynamic agri-cultural sector

• increased maledomination of agri-cultural activities.

• Principally one-way.

• Two-way, but notstrong.

• Improved maternalhealth will result inmore productiveagricultural labor,both from womenand from their children.

• Increased returns to skilled labor inagriculture.

• Primary schoolswith adequatequality of instruc-tion are accessible.

• Security of femaleaccess to agricul-tural resources.

• Secure female control over ownagricultural output.

• Knowledge ofproper diet andnutritional care.

• Accessible andeffective health services.

• Degree of controlwomen have overresources to assuretheir own health.

• Availability of nutrient-dense food crops.

2. Achieve universal primary education.

3. Promote gender equality and empower women.

4. Reduce child mortality.

5. Improve maternal health.

• Few.

• Increasingly prof-itable agriculture

• potential to eco-nomically empowerwomen farmers.

• Few.


• higher-quality diets• improved health.

• More dynamic agri-cultural sector willchange assessmentsof the economicreturns to educatingone’s children com-pared to returns ofkeeping childrenout of school towork in house-hold agriculturalenterprises.

• Broader economicimprovementsthrough dynamicagriculture

• increased publicexpenditures onwater and sanita-tion, health, energy sectors

• reduced time bur-den on women fordomestic tasks.


• better nutrition• increased child

survival.• More dynamic agri-

cultural sector• increased income• more resources

available to managechildhood illnesses.

• Primarily throughsame mechanismsas MDG 3 onempowerment ofwomen.

(continued )


Table 2.1 Summary of Links Between the Agricultural Sector and the Millennium Development Goals, Principally at Household Level (Continued)


• Greater and morediverse food production

• higher-quality diets• improved health.

• Agriculture prac-tices can be bothdirect causes of andimportant immedi-ate solutions toenvironmentaldegradation.

• Expanding globalagriculture tradeincreases need forformal trading part-nerships and rules.

• Capital require-ments for compre-hensive agriculturaldevelopment

• Significant increasesin developmentassistance offered to the agriculturesector.

• More dynamic agri-cultural sector

• increased income• more resources to

devote to healthservices.

• More productiveagricultural technologies

• withdrawal of agriculture frommarginal, sensitiveenvironments.

• More profitableagricultural sector

• reduced migrationto urban slums.

• More profitableagricultural sector

• expectation of better governanceand provision ofpublic goods bygovernments to sustain the benefitsfrom agriculture inthe long term.

• Two-way, princi-pally indirect.

• Reduced healthburden enablesmore productiveagriculture.

• Possible negativeramifications if agri-cultural investmentsor labor migrationpatterns exacerbateor extend diseases.

• Two-way. Bothdirect and indirect.

• Agricultural sectoris as likely to havenegative ramifica-tions on the envi-ronment as positive.Unprofitable agri-cultural systemstend to unsustain-ably mine environ-mental resources.

• Declining environ-mental resourcebase is an erosionof the foundationfor the agriculturaleconomy.

• Two-way, but prin-cipally toward agri-culture. Primarilydirect.

• Globalization is aslikely to have nega-tive as positive ram-ifications onagricultural produc-ers, particularlysmall-scale subsis-tence farmers, inthe short term.

• Effective health sys-tem, both curativeand public healthservices.

• Effective interven-tions to limit HIVinfection.

• Particularly for HIVinfection in sub-sistence farminghouseholds, avail-ability of nutritiousfood crops that arenot labor-intensive.

• To minimize nega-tive environmentalexternalities of agricultural invest-ments, participatoryplanning processesrequired.

• Relatively equitabledistribution of agri-cultural assets acrossthe population.

• Environmental costs of agriculturalproduction incorpo-rated into economicassessments of production systems.

• Sufficient knowl-edge, capital, andaccess to markets toenable agriculturalproducers to engagein regional andglobal trade.

6. Combat HIV/AIDS, malaria, and other diseases.

7. Ensure environmental sustainability.

8. Develop a global partnership for development.

19

COUNTRY-LEVEL ASSESSMENT OF SELECTEDMDGS: ETHIOPIA AND ZAMBIACompared with other regions of the developing world, Sub-SaharanAfrica faces a huge challenge in terms of meeting the two targetsconstituting the first MDG—halve, between 1990 and 2015, the pro-portion of people whose income is less than one dollar a day andhalve, between 1990 and 2015, the proportion of people who sufferfrom hunger.

Indeed, as of 2001, approximately 46.5 percent of the inhabitantsof Sub-Saharan Africa, or 314 million people, were living below theinternational poverty line (World Bank 2004b). If the region followscurrent trends, 39.3 percent of the population will remain below thisline by 2015, and Sub-Saharan Africa will be the only developingregion where the number of poor people actually increases from the1990 level (AfDB/OECD 2002; UNDP/UNICEF 2002).

Poverty both contributes to and is reinforced by hunger andmalnutrition. During the last decade, the percentage of under-nourished people in Sub-Saharan Africa decreased only margin-ally, from 35 to 33 percent, while the number of undernourishedpeople increased from 166 million to 198 million. At the same time,the average per-capita energy intake only increased by 100 calo-ries and actually fell in some countries. Children are particularlyvulnerable to the effects of hunger. During the 1990s, the propor-tion of children under five years of age who were underweightremained relatively stagnant, averaging about one-third, whilethe number of underweight children actually increased by 8 mil-lion (UNDP/UNICEF 2002). The extent of child stunting (propor-tion of children with height for age under 2 standard deviationsfrom the reference population median) is even higher, affectingapproximately 41 percent of preschool children in Sub-SaharanAfrica (UNICEF 2004).

Agricultural growth is often identified as a means for ameliorat-ing Sub-Saharan Africa’s hunger and poverty. Indeed, agriculture isthe primary livelihood of approximately 65 percent of people in theregion, represents between 30 and 40 percent of the region’s GDP,and accounts for almost 60 percent of its income from exports (IFAD2003). Increased growth of the agricultural sector offers direct bene-fits to poor farmers, such as income and food, contributes to broaderfood security objectives, and helps establish forward linkages withhigher value-added industries.

AlternativeScenarios to 2015

Prospect and Policies for Meeting the MDGs on Poverty

If Sub-Saharan Africa

follows current trends,

it will be the only

developing region

where the number of

poor people actually

increases from the

1990 level.

3

Nevertheless, the ability of the agriculturalsector to contribute significantly to the MDGs inSub-Saharan Africa will vary according to the con-straints and opportunities prevailing within eachcountry. This chapter focuses on the level and typeof agricultural growth required in two countriesof the region that are currently far from achievingthe MDGs: Ethiopia and Zambia. As shown intable 3.1, the percentage of the population livingbelow the international poverty line of $1 a day ishigher in both countries than for Sub-SaharanAfrica as a whole.

Compared to Ghana and Uganda, progress atreducing poverty in Ethiopia and Zambia has beenrelatively slow over the last decade. In Zambia,the most recent household survey (1998) shows a6 percent increase in the national poverty rate since1991. In Ethiopia, the poverty rate declined from51 percent in 1993 to 44 percent in 2000, but most

of the decline occurred in the early period and thepoverty rate only changed 1 percent between 1996and 2000. Table 3.2 shows the U.S. dollar conversionof the national poverty lines on which these povertyrates are based and highlights that Ethiopia’s poorlive on less than $85 a year.

Economic and demographic indicators, alongwith broad growth trends averaged over the periodfrom 1991 to 2001, further elucidate the similaritiesand differences between the two countries. In U.S.dollar terms, Zambia’s per-capita income is almosttriple that of Ethiopia and lies between the averagefor Ghana and Uganda. This is predominantly dueto the higher urbanization rates in Zambia com-pared with Ethiopia and the higher incomes in theseurban areas. However, once per-capita incomes areadjusted to purchasing power parity (PPP), bothZambia and Ethiopia are far below the averagein the region (table 3.3). In Ethiopia, agricultural


Table 3.1 Comparison of Poverty Across Selected African Countries

Trends in national poverty rates Rural poverty Urban poverty Population under (survey year in parentheses) rate rate $1 US/day

Ethiopia 51.1a 45.5 44.2b 45.4b 36.9b 81.9c

(1992/93) (1995/96) (1999/00) (1999/00) (1999/00) (1999/00)Zambia 68.9d 79.4d 75.4d 85.6d 58.3d 63.7c

(1991) (1996) (1998) (1998) (1998) (1998)Ghana 51.7e 39.5e 49.9e 18.6e 44.8e

(1991/92) (1998/99) (1998/99) (1998/99) (1999)Uganda 56.0f 44.0f 35.0f 39.0f 10.0f n.a.

(1992) (1997) (1999/00) (1999/00) (1999/00)Africa 49.0g

(2000)

Sources: aMOPED (1994); bMOFED (2002); cWorld Bank, World Development Indicators (2003); d Thurlow and Wobst (2004); ePRSP for Ghana (2003); fPRSP for Uganda (2003); and gUN Millennium Development Goals Database.n.a. = Not applicable.

Table 3.2 Poverty Lines in the Country Case Studies

Annual poverty line

Country Local currency U.S. dollarsa Survey year Type of poverty line

Ethiopiab Br 696 84.65 1999–00 absoluteZambiac K 555,432 298.28 1998 basic needs

Sources: aConversions calculated from exchange rates given in the IMF’s International Financial Statistics Database; bMOFED (2002); cThurlow and Wobst (2004).

growth is lower than the country’s populationgrowth, which is worrisome considering that agri-culture constitutes 50 percent of Ethiopia’s GDPand provides the livelihood for approximately 85 percent of its population. By contrast, agricul-ture provides only around 22 percent of GDP inZambia, much lower than the African average.This is because the now faltering copper industryhas traditionally dominated the economy. Althoughagricultural growth has, on average, outstrippedZambia’s population growth, total economicgrowth has not (table 3.4).

The unique challenges facing Ethiopia andZambia partially explain these trends. Since theearly 1990s, both countries have been engaged in aprocess of economic liberalization. In Ethiopia, thegovernment has been trying to transform the coun-try from a centrally planned to a market economyand remains committed to maintaining macroeco-nomic stability. Nonetheless, the country’s growth

has been hindered by a number of factors, includingprotracted conflict with neighboring Eritrea, declin-ing world prices for its coffee exports, and the coun-try’s vulnerability to natural disasters, particularlydroughts (AfDB/OECD 2002). Due to the preva-lence of droughts, Ethiopia remains Africa’s largestrecipient of cereal food aid, which equaled approxi-mately 1.2 million tons in 2002. The effects of foodinsecurity are particularly acute for children, asEthiopia has Africa’s highest rate of children whoare underweight and the second highest rate ofchild stunting after Burundi.

In Zambia, a series of structural adjustment pro-grams have been adopted that emphasize, amongother things, privatization of the copper mines,trade liberalization, and agricultural reforms. How-ever, because of the inadequate design and piece-meal implementation of many of these reforms,coupled with a sizable external debt burden,growth has not improved. In fact, the bias against

Alternative Scenarios to 2015: Prospect and Policies for Meeting the MDGs on POVERTY 21

Table 3.3 Economic and Demographic Indicators, 2001

GDP per GDP PPP per Agricultural GDP Total population Rural population Agricultural share capita (US$) capita (US$) per capita (US$) (millions) (millions) of GDP (%)

Ethiopia 121 810 63.21 65.8 55.4 52Zambia 405 780 94.71 10.3 6.2 22Ghana 421 2250 253.41 19.7 12.5 35Uganda 355 1490 167.79 22.8 19.5 37Africa 567 1826 166.21 673.9 403.7 33a

Source: World Bank (2003).Notes: aExcludes South Africa.

Table 3.4 Average Trends Across Countries, 1991–2001 Averages

Average annual growth rate (%)

GDP Agricultural GDP Population

Ethiopia 5.28 2.31 2.47Zambia 1.18 5.27 2.45Ghana 4.27 3.36 2.18Uganda 6.55 3.82 3.00Africa 3.00 3.00 2.00

Source: World Bank (2003).

agriculture created by the mining sector is stillprevalent, and insufficient input and outputmarkets have circumscribed farmers’ ability toimprove production. Like Ethiopia, agricultural pro-duction in Zambia is also hampered by insufficientand unevenly distributed rainfall. Similar to manyof its Southern African neighbors, the HIV/AIDSpandemic, which affects approximately 23 percentof the population, has severely diminished laborproductivity and placed even greater pressure onfood security. A combination of all of these fac-tors has contributed to a deterioration in per-capitacereal production (table 3.5), culminating in 2002–03in a major food crisis that caused an estimated2.3 million people to require humanitarian assis-tance (Samatebele 2003).

Clearly, addressing the problems of hunger andpoverty deserves priority in both countries. Resultsfrom the scenarios of economy-wide models, whichare discussed in detail reveal that neither countrywill be able to meet the MDGs if they continuealong their current growth trajectories. Improvedperformance in the agricultural sector will be cru-cial for substantial hunger and poverty reductionto occur, and fortunately each country’s govern-ment is promoting development strategies targetedat enhancing the sector’s growth. As the scenar-ios highlight, Ethiopia has the potential to meetthe MDGs through combined growth in the sta-ples, livestock, and nontraditional export sectorsif there are concurrent improvements in the trans-port sector and further market development.Halving poverty in Zambia by 2015 will be morechallenging to achieve. Nonetheless, agriculturalgrowth is a necessary condition for poverty reduc-tion in rural areas. As in Ethiopia, attention to mar-ket access issues in Zambia, particularly through

the construction of more roads, will be essential forgrowth in these areas to materialize.

Ethiopia: Agriculture-Led Growth and Poverty Reduction

With a per-capita income that is only one-fifth ofthe African average, Ethiopia classifies as one ofthe world’s poorest countries. In addition to daunt-ing poverty, persistent food crises have left a largeportion of the population suffering from food in-security. Despite significant amounts of food-aidassistance over recent years, there has been littleprogress at reducing this food insecurity.

Ethiopia’s current circumstances reflect thecumulative challenges faced by the country overthe past decades. In particular, the country is ex-tremely vulnerable to drought and since the early1980s, has experienced seven major droughts, fiveof which resulted in famines. The most recentdrought, which occurred in 2002–03, affected ap-proximately 30 million people (EM-DAT 2004).

In addition to climatic factors, the country hassuffered under the misguided economic policies ofthe socialist Dergue regime, which ruled from 1974to 1991. When the Ethiopia Peoples’ RevolutionaryDemocratic Front (EPRDF) replaced the Dergue in1991, a number of market-oriented reforms wereimplemented, including those aimed at stimulatingagricultural and rural growth (World Bank 2004d).For example, the country liberalized its foreignexchange markets and dramatically decentralizedthe public administration to the woreda (district)level. In rural areas, grain markets were liberalizedand fertilizer markets were opened up to participa-tion from the private sector. In 1992, the EPRDFalso established the Agricultural Development


Table 3.5 Cereal Production in Selected African Countries

Production (kg per capita per year)

1991 1994 1997 2000 2001

Ethiopia 99.55 86.69 143.83 112.94 132.02Zambia 138.48 123.30 109.24 94.38 65.72Ghana 83.29 85.33 83.90 80.38 74.88Uganda 84.61 94.26 72.19 86.27 91.92Africa 130.65 135.17 127.11 120.94 118.50

Source: Calculated from FAOSTAT (FAO 1998) and World Development Indicators data (World Bank 2003).


Led Industrialization Strategy, which emphasizesthe role of the agricultural sector as a catalyst forimmediate improvements in food security and forlong-term growth in the broader economy.

However, the outbreak of conflict with Eritreabetween 1998 and 2000 created a humanitarianemergency in the northern part of the country andreduced the availability of resources to financemany of these reforms. Not only did donors andinvestors reduce their support to the country, but also increases in official defense expendituresdecreased the availability of funds for other sec-tors as well as for antipoverty programs (WorldBank 2004d).

With the return to peace, the government hasreaffirmed its commitment to generating growthand reducing poverty, especially through a strongfocus on the agricultural and rural sector. Sincemore than 85 percent of the country’s populationlives in rural areas where agriculture is the maineconomic activity and where the poverty ratio is sig-nificantly high, and since the nonagricultural sec-tor is extremely small in Ethiopia, any strategy forslashing poverty and hunger has to focus on gener-ating rapid growth in the agricultural sector. To thisend, the Ethiopian government has not only contin-ued to support the Agricultural Development LedIndustrialization Strategy but has also launched aseries of development and poverty reduction pro-grams, such as the Sustainable Development andPoverty Reduction Program in 2001 and the FoodSecurity Program in 2004. Agricultural growth, foodsecurity, and accelerated rural development are thecornerstones for all these programs.

To identify which investments can have thelargest impact on agricultural growth and in turndrive broader growth and poverty reduction, agreater understanding is needed of the linkagesbetween agriculture, growth, and poverty reduc-tion. To this end, researchers at the InternationalFood Policy Research Institute have developed aspatially disaggregated, economy-wide model forEthiopia based on recent national household sur-veys, agricultural sample surveys, global informa-tion system data, and other national and regionaldata. The model analyzes the growth and povertyreduction linkages at both national and regionallevels. The study shows that broad-based growthin agriculture is the key for Ethiopia’s success inmeeting the objective of halving poverty. Withinthe agricultural sector, growth in cereals and otherstaple crops should receive priority, given their

superior role in reducing poverty. The study showsthat increasing national staple food availability by50 percent by 2015 will significantly help povertyreduction. The feasibility of achieving this goalrelies on reducing the productivity gap betweentraditional and modern technologies that havebeen adopted in the country. Achieving sustain-able agricultural growth in Ethiopia also requiressupporting investments in roads and other marketconditions.

In addition, the study emphasizes the need forregionally differentiated strategies, given the coun-try’s size as well as heterogeneous conditions inboth natural resource and economic environments.Indeed, more than 50 percent of the poor live in thefood deficit area, where the availability of food sta-ples per household is half the national average. Thepoverty and food security challenge is huge insuch areas. On the other hand, more than 50 per-cent of food staples are currently provided fromfood surplus areas where food staples availabilityper household is already 70 percent higher than thenational average. In these food surplus areas, thereis a need for greater diversification in agriculturalproduction. Consequently, market access and mar-ket development should be an integral part of anational agricultural development strategy.

The current growth path results in more poverty

In order to first demonstrate the necessity for increased agricultural growth, the InternationalFood Policy Research Institute’s model simulatesthe impact on poverty if Ethiopia continues alongits current growth trajectory. Between 1995 and2002, about 90 percent of increases in total crop pro-duction and 70 percent of increases in cereal pro-duction were due to area expansion. The annualgrowth rate of cereal production was about 2 per-cent, lower than the 2.5 percent population growthrate. The productivity growth rate (yield) for totalcrops and cereals was also low, about 0.2 and0.6 percent per year, respectively. If the crop areaexpansion and growth in yields continue accordingto their current trends for the next 12 years, togetherwith the growth trends in livestock production(4.2 percent per year) and non-agriculture (4.6 per-cent), a business-as-usual scenario shows that in thenext 12 years the national economy (GDP) and theagricultural sector grow at 3.2 and 2.5 percentannually, respectively. Since agriculture, especiallycereal production, will grow either close to or more

slowly than the population growth, the nationalpoverty rate will keep at its high current level of45.7 percent by 2015. With a 2.5 percent populationgrowth rate, the number of the poor living underthe current poverty line will increase to 40 millionby 2015, equivalent to 10 million more than today’snumber. Increases among the poor will mainlycome from the food deficit area, where the currentpoverty ratio is already very high. The majorityof Ethiopians will still be struggling to meet theirbasic food needs, especially since average dailyper-capita caloric intake will be even lower thanits current level.

Growth in staples is the most important for poverty reduction

Cereals and other staple crops account for 65 per-cent of agricultural value-added, and the majorityof small farmers are producers of staple crops.Thus, this sector should have the potential to sub-stantially alleviate poverty. The model simula-tions show that if the average yield of staple cropsgrows by an additional 1.5 percent from its currentrate of 0.6 percent to 2.1 percent annually, and iscombined with the 1.3 percent expansion in croparea assumed in the business-as-usual scenario,cereal production can grow at 3.4 percent per year.By taking into account the demand-supply, agri-cultural-nonagricultural, and cross-sectoral link-ages in agriculture, such a growth rate in staplecrops (combined with the base-run growth trendsin the other sectors) results in an annual GDP andagricultural GDP growth rate of 3.9 and 3.5 percent,

respectively, compared with 3.1 and 2.5 percent,respectively, in the baseline scenario.

More importantly, this staple-crops scenario con-tributes more toward poverty reduction than theother agricultural subsectors or the nonagriculturalsector, even though these other sectors may have asimilar impact on the growth of the overall econ-omy (table 3.6). Growth in staple crops also bene-fits consumers, as staples are the most importantsources of food energy for poor rural and urbanconsumers alike. The national household surveydata indicates that the rural poor, whose income isalready below the poverty line, spend about 70 per-cent of their total income on staple-crop food,which is 30 percent higher than the rural average.In the urban areas, households with incomes belowthe poverty line spend almost 50 percent of theirincomes on staple-crop foods, which is 65 percenthigher than the urban average. Thus, raising pro-ductivity in staple crops will increase the food sup-ply, lower food prices, and help reduce the povertyrate in the urban and rural areas.

However, it needs to be highlighted that in-creased staple production often exceeds farmers’own consumption. Hence, expanding markets forthese commodities is a necessary condition for farm-ers to benefit from growth. Moreover, to improvecrop yield, farmers need to increase the use of mod-ern inputs, which are purchased from markets.Development of input and output markets canstrongly support the growth of agricultural produc-tion. If marketed staple crops increase too rapidlyin the absence of market development, prices forthese crops can be significantly depressed, nega-


Table 3.6 Growth and Poverty Reduction Under Agricultural Growth Options

Add. growth Agric. share Annual growth rate Poverty rate (2015)

Scenarios ratea (2002) GDP Agric. Nonagric. All Rural Urban

Initial value 44.4 45.7 36.91. Current growth 3.1 2.5 3.7 45.7 48.0 33.0Additional growth2. Staple crops 1.5 65.0 3.9 3.5 4.3 36.7 37.7 31.03. Livestock 3.6 26.0 3.9 3.5 4.2 40.0 42.0 26.84. Nontraditional 8.7 4.8 3.6 3.4 3.8 40.2 41.4 33.75. Three sectors (2 to 4) 5.1 5.3 4.9 27.5 27.8 25.66. Market investment (2 to 4) 5.8 5.4 6.1 24.4 25.0 21.2

Source: Results from Ethiopian multimarket model.aAdditional annual percentage yield or productivity growth rate in the relevant agricultural sectors.


tively affecting farmers if declines in prices arelarger than increases in marketed products. It isimportant to realize such risks, and hence to grad-ually rather than abruptly increase production inorder to avoid overwhelming the market.

After staple crops, the livestock sector is thecountry’s second largest agricultural sector. Thehistorical growth rate of 4.8 percent (1995–2002) inthe livestock sector is higher than the growth rate instaple crops or in agriculture in total, implying thatthe sector has a strong growth potential. The “live-stock growth” scenario simulates an additional3.5 percent annual growth rate in the livestock sec-tor, which raises the annual growth rate in the live-stock sector to a total of 8.3 percent. Holding thegrowth rate in the other agricultural and nonagri-cultural sectors the same as in the baseline scenario,this simulation shows that growth in the livestocksector has the most significant effect on overall eco-nomic growth, causing GDP and agricultural GDPto grow at 3.9 and 3.5 percent per year, respectively.However, an 8.3 percent annual growth rate inthe livestock sector has a relatively smaller povertyalleviation effect than a 3.4 percent annual growthrate in the staple crops. Although the urban poorbenefit more, the rural poverty rate falls to 42 per-cent by 2015 in this scenario as opposed to the36.7 percent achieved in the “staples growth”scenario.

While livestock is a relatively large agriculturalsector in the economy, it accounts for a relativelysmall share of income for poor farmers. On theconsumption side, poor consumers in both therural and urban areas consume fewer livestockproducts. Calculated from the household surveydata, the rural households living under the povertyline spend less than 4 percent of their incomes onlivestock and dairy products, which is 40 percentlower than that for an average rural household.Thus, while increased livestock production de-presses prices in domestic markets, farmers do notexperience the same level of benefits that they doin the staples scenario.

The nontraditional-export sector, which includesexportable vegetables, fruits, other horticulturalproducts, chat, cotton, sugar, and sesame seed, cur-rently accounts for 5 percent of agricultural GDP.An additional 8.7 percent annual growth rate in thesector’s productivity is equivalent to the 1.5 per-cent growth rate in staple crops or the 3.5 percentgrowth rate in livestock described in the two pre-ceding scenarios. Nonetheless, growth in some

nontraditional exportable products has been veryrapid in recent years, indicating that it is a boomingsector in the economy. Therefore, a much morerapid growth rate of 13.3 percent is simulated forthe sector in the “nontraditional growth” scenario.To support such production growth, nontraditionalexports have to grow at 29 percent annually. Thisexpansion of nontraditional exports can result in anoverall economic growth rate of 3.6 percent andan agricultural growth rate of 3.4 percent, which iscomparable to the projected annual growth rateunder the “staple crops” scenario.

However, the nontraditional export sector’s con-tribution to poverty reduction is relatively small. Infact, the rural poverty rate falls to 41.1 percent, only4.6 percentage points below the current level. Thisis because nontraditional export growth is oftenconcentrated around cities where there is greateraccess to transportation and other market facilities.Given the technical and financial constraints thatthey face, the majority of poor, rural farmers areunable to adopt the technology required for produc-ing nontraditional crops.

On the demand side, increased production ofnontraditional goods provides few benefits to poorconsumers in both rural and urban areas, especiallysince such products are often intended for exportmarkets. The greatest constraint to growth in such asector is inadequate market access. Rapid growthin the nontraditional-export sector requires a timelyexpansion of markets for such commodities. Indeed,achieving the projected annual growth rate of13.3 percent in the production of nontraditionalexports is unlikely without huge investments ininfrastructure and other market conditions. For thisreason, this discussion regarding the sector’s con-tribution to the overall economy and to povertyreduction may actually be too optimistic.

By focusing on individual agricultural sub-sectors, the preceding analysis emphasizes thatgrowth options among different agricultural sub-sectors have different effects on poverty reduc-tion. Obviously, growth in any single subsectoralone cannot help the country meet the MDGs.Admittedly, growth in staple crops is critical forpoverty reduction, but it needs to be supported bygrowth in other subsectors. Growth in staple cropsand nontraditional exports can increase domesticdemand for livestock products, which helps stabi-lize the livestock prices and raise livestock farmers’incomes. Similarly, growth in the livestock sectorgenerates feed demand. Increased income from

growth in livestock and nontraditional exports canhelp stabilize the prices for food crops. Combiningthe growth rate in the three major agricultural sub-sectors results in a 5.3 percent annual growthrate for agriculture. Such a diversified agriculturalgrowth strategy will significantly reduce poverty,causing the poverty rate to fall by 17 percentagepoints to 27.5 percent by 2015.

Assessing investment options

Achieving the requisite growth rates examined inthese scenarios requires increased public invest-ments. Various public investments, such as agri-cultural research and development, extension, irri-gation, and other infrastructure, can directlyimprove agricultural productivity, raise farmerincomes, and reduce poverty. Although all of theseinvestments are needed, each investment can havea differential impact on agricultural growth andpoverty reduction. Given limited government bud-gets and international donor funds, as well asthe extremely broad areas in which public invest-ments are needed, priority matters in planninginvestment strategies. Accordingly, a broad analy-sis of investment strategies will be helpful in iden-tifying which kinds of investments can bringpro-poor agricultural growth to the country.

Irrigation. We examine irrigation investments firstbecause increased irrigation is important for re-ducing climate risk, which is one of the greatestconstraints to agricultural growth in Ethiopia. More-over, reducing climate risk helps induce the useof modern inputs, such as fertilizers and improvedseeds, which can further increase agricultural pro-ductivity. Currently, there are about 200,000 hectaresof irrigated area in the country, accounting forslightly more than 2 percent of total crop areas.Among the 200,000 hectares of irrigated land,60 percent is dedicated to cereal production.According to the Agricultural Sample Survey dataof 1997 and 2000, the yield gap between irrigationand rainfed crop production is 40 percent; that is,on average, irrigation can increase cereal yield upto 40 percent. Obviously, significantly increasingirrigation area can stimulate growth in cereal pro-duction. However, as irrigation currently accountsfor less than 2 percent of total cereal productionand slightly more than 2 percent of other crop pro-duction, it is unrealistic to expect that investmentin irrigation alone can generate the growth rate

we analyzed above. Moreover, many researchershave shown strong diminishing returns in the large-scale irrigation investment, which implies thatcaution is needed in promoting large irrigationinvestment projects (Fan and Hazell 2001).

We simulate an increase in irrigation area accord-ing to the country’s irrigation development pro-gram drawn from the country’s Water SectorDevelopment Plan (Ministry of Agriculture andRural Development 2005). The model simulates anincrease in irrigated land according to the plans andshows that even by doubling irrigated cereal areasby 2015, irrigation will only account for 3 percentof total cereal production. As a result, the averagegrowth rate of cereal output will increase modestly,from 2.2 percent in the baseline scenario to 2.4 per-cent in the irrigation scenario, which is equivalentto about 0.2 percent of additional annual growth.Irrigated cash crop areas will triple by 2015, andwill account for 5 percent of total cash crop areas,compared with the current 2 percent. This willbenefit exports. For example, horticultural exportswill increase by four-fold by 2015. A similar effectis also observed for coffee exports. Because themedium- and long-term projects are completedonly toward the end of the period analyzed, thepotential returns from the newly irrigated areasare not fully captured in the simulation.

Although irrigation has the potential to raisecrop productivity, the ability of increased output,especially in the export sector, to reach internationalmarkets depends on simultaneously improving theunderlying market conditions. The gains we ana-lyzed here should not be understood as solely frominvestments in irrigation. Without investments inmarkets, national prices for increased output mightdecline, which may reduce the realized gains fromthe investment in irrigation.

With growth in both staple and cash crops due todoubled irrigation area, together with cross-sectorallinkage effects, the annual growth rate for GDP andagricultural GDP rises to 3.6 and 3.0 percent, respec-tively, from 3.1 and 2.5 percent in the baseline. Thenational poverty rate falls to 39.4 percent from itscurrent 44.4 percent. While irrigation has a modesteffect on national poverty reduction, its effect on thefood staples deficit area is significant. Since pro-jected increases in irrigated land are mainly locatedin the food staples deficit area, the rural povertyrate in the area falls to 51.5 percent by 2015 from itscurrent 58.4 percent. There is a much smaller effecton poverty reduction in the food surplus areas dueto a smaller increase in irrigated land in the area.



Adoption of improved seed and greater efficiency infertilizer use. Compared with the international stan-dard, the current yield level of crop production,especially for grains, is quite low in Ethiopia. Boththe low utilization and low efficiency of moderninputs can partially explain such low yield levels.However, there are many factors restraining thedissemination of modern inputs. Survey datashow that while fertilizer has been used on about40 percent of the grain area nationwide, only asmall portion of it is combined with the use of otherinputs, especially with improved seed. As a result,the average yield gap in grain production due tofertilizer use is quite small. Yet, there are significantgains from combining fertilizer use with improvedseeds in grain production, especially in maizeproduction.

Thus, the model simulates a situation in whichthe technology combining improved seeds withfertilizer is expanded to all of the cereal area thatis currently fertilized. At the same time, we increasethe efficiency of fertilizer use by 50 percent over12 years. This technology results in an additionalannual growth rate of 0.9 percent for cereal pro-duction. As a result, the rural poverty rate falls to40.3 percent by 2015, which is 4 percentage pointslower than the current rate.

Increased irrigation combined with the adoption ofmodern seed and improved efficiency in fertilizer use.Since the returns to technology adoption are low ifmodern inputs are used in isolation and not sup-plemented by other technologies, modern technol-ogy needs to be disseminated in a package. Thus,we simulate a situation in which the adoption ofmodern seed varieties is combined with improve-ments in the efficiency of fertilizer use and com-bined with the expansion of irrigated area. Bysimultaneously investing in the three areas, theannual growth rate of cereal production rises to3 percent, resulting in an annual growth rate of3.8 and 3.4 percent for GDP and agricultural GDP,respectively. Growth in the cereal sector, combinedwith increased cash crop production from irriga-tion projects, helps the poverty rate fall to 37 per-cent, which is 7.4 percent lower than its currentlevel and in line with the simulation result thatanalyzed the “staple crop growth” scenario.

The above analysis shows that through technol-ogy adoption and dissemination, combined withincreases in irrigation area, it is possible to have amore than 3 percent annual growth rate in staple

crops. Although meeting the objective of halvingpoverty requires more than improving staple crops’productivity, growth in the staple food sectors isobviously a necessary condition for any significantreduction in poverty. Exploiting the growth poten-tial of staple crops from dissemination of moderntechnology requires not only investment but alsochanges in farm management and a transition fromcurrent farming traditions to more modern farm-ing systems.

Combined with infrastructure investments that reducemarketing costs, the agriculture-led growth strategyhas the potential to meet the MDG of halving poverty by2015. An agriculture-led growth strategy does notimply that investments should only be in agricul-ture. Many studies have shown that poor infrastruc-ture and dysfunctional markets prevent farmersfrom accessing markets and hence diminish agricul-ture’s profitability. It is important to remember thatinstitutional barriers also constrain farmers frombecoming actively involved in market activities andthat market development does not solely implyinfrastructure investment. Nonetheless, in this sec-tion we focus specifically on investments in roadsand other infrastructure that could reduce the trans-portation costs of agricultural trade and improvemarket access for farmers.

The Ethiopian road density is 27 kilometers per1,000 square kilometers, which is only half of theaverage for Africa. Seventy percent of farmers arereported to be more than half a day’s walk from anall-weather road (MOFED 2002). Such poor mar-ket access conditions and high transportation costssignificantly increase the price gap received byfarmers and paid by consumers. The average grainprice gap is estimated to be about 30 to 70 percentacross regions, and domestic marketing costs oftenaccount for more than 50 percent of fertilizer pricespaid by farmers. These costs significantly reducefarmers’ profitability from increased production.

To address the constraints in the road sector, theGovernment of Ethiopia formulated a two-phaseroad sector development program (Ministry ofAgriculture and Rural Development 2005). Underthe first phase, which was completed in 2002,the focus was on the rehabilitation of the corenetworks. Substantial progress was achieved inreopening nearly all of the classified roads, accessi-bility was improved, and the percentage of roadsin good condition also increased. The secondphase of the program involves the rehabilitation of

1,168 kilometers, the upgrading of 2,045 kilome-ters, and the construction of 8,383 kilometers ofroads.

Improving the road network also includes roadmaintenance. For example, the removal of rainwa-ter from the surface of the road as well as fromthe adjacent ground is crucial in Ethiopia becausemost of the roads (about 90 percent) are gravel andearth. Rainwater can wash away the road surfac-ing construction, creating a significant inconve-nience for vehicles and travelers. Because of poormaintenance, many existing roads have becomeimpassable during the rainy season and even inthe dry season.

Lack of market infrastructure also constrainsthe market accessibility of small farmers. Forinstance, lack of storage and marketing facilitiesand less developed formal trading systems signif-icantly increase farmers’ transaction costs and forcemany smallholders back to the subsistence modeof farming.

While it is known that the cost of building andmaintaining roads is high in Ethiopia because of therugged topography and torrential tropical rains,there is no approximate cost information for suchinvestments. Consequently, we have to make twomain assumptions before conducting the simula-tion. First, investment is modeled as a lowering ofthe marketing margins between food staples sur-plus and deficit areas. We assume that the marketprices across zones will converge due to improvedtransportation and market conditions, and theprice gap between surplus and deficit areas willbe 70 percent lower by 2015 than its current level.We further assume that lowered marketing costsare due to the improvement in the service sector’sproductivity, and by 2015 the productivity in theservice sector will be 20 percent higher than thelevel in the baseline’s 2015, which is equivalent toan annual growth rate of 1.5 percent.

Once growth in the agricultural sectors is com-bined with improved marketing margins, the cross-sectoral linkage effects cause the growth rates ofGDP and agricultural GDP to increase to 5.8 and5.4 percent, respectively. Improving marketingconditions makes the terms of trade favorable toagriculture. Specifically, reducing marketing costsmainly benefits smallholders through increasedprices they receive for their goods, which enablesthem to increase their income from producing thesame amount of output. Due to such strong, cross-sectoral linkages and positive price effects, the

poverty rate is significantly lowered, and the coun-try will be quite close to meeting the objective ofhalving the poverty rate by 2015. In fact, the nationalpoverty rate falls to 24.4 percent by 2015.

Summary for Ethiopia

Ethiopia faces serious challenges in attempting tomeet the MDGs. Along a business-as-usual growthpath, the country’s food security will further deteri-orate. Without additional growth in agriculture, thepoverty rate will actually rise, resulting in 12 millionmore people living in poverty by 2015 (figure 3.1).

Growth in staple crops contributes the mostto poverty reduction. With a 3.4 percent annualgrowth rate, (1.5 percent of additional growthrate in productivity), growth in staple food helpsthe economy and the agricultural sector grow at3.9 and 3.5 percent, respectively. The country’spoverty rate will reduce to 36.6 percent by 2015from its current 44.4 percent. Combined growthin staple crops with growth in livestock and non-traditional exports leads to much more rapidgrowth in agriculture and poverty reduction. Sucha growth strategy results in an annual growth rateof 5.3 percent for the agricultural sector, whichwill help the country reduce its poverty rate to26.6 percent by 2015.

Increasing national food staple availability by50 percent by 2015 will significantly advance pov-erty reduction. This goal is feasible by reducingthe productivity gap between traditional and mod-ern technologies that have been adopted in thecountry. By doubling the irrigation area by 2015,improving the efficiency of fertilizer use, and dis-seminating the technology to combine improvedseed with fertilizer use, the model results showthat growth in staple foods is feasible with theright investment strategies. However, as more than50 percent of food staples are currently providedfrom the food surplus area, where per-householdfood availability is already 70 percent higher thanthe national average, market access and marketdevelopment will be especially important andshould be integrated into Ethiopia’s agriculturaldevelopment strategy.

As the model simulations reveal, broad-basedagricultural growth is the key for decreasing pov-erty and increasing growth in Ethiopia. Within theagricultural sector, growth in cereals and otherstaple crops should receive priority. Due to strong,cross-sectoral linkage effects, an agriculture-led



growth strategy combined with investments inroads and other market conditions can significantlyincrease rural incomes. Only under such a scenariocan Ethiopia potentially halve poverty by 2015.

Zambia: Facing the Challenge of theMillennium Development Goals—The Role of Agriculture

While Zambia shares with Ethiopia a high povertyrate, specifically in the rural sector, the circum-stances contributing to Zambian poverty are quitedifferent.4 In particular, Zambia’s economic historyhas been shaped by the misuse of its abundant nat-ural resources. It has failed to translate its consider-able mineral wealth and agricultural potential intosustained growth and the improved well-being ofits population. Even during periods of growth thepoor population has had limited opportunities toparticipate in the growth process. As such, not onlyhas poverty in Zambia remained high, but alsotrends in social indicators suggest that the economyhas continued the declining trend that began three

decades ago. The marginalization of agriculture liesat the center of this development failure.

The marginalization of agriculture

Two policy-induced biases determined the condi-tion of the Zambian economy at the start of the1990s. The first of these was a general bias towardurban areas. During the 1970s the governmentadopted an inward-oriented development strategybased on nationalized and protected state enter-prises. So comprehensive was this strategy that by1991 over three-quarters of GDP was being gener-ated by the public sector (Chanthunya and Murinde1998). The subsequent dependence on copper earn-ings, as a source of both foreign exchange andpublic revenues, created an economy that was vul-nerable to crisis. The first of these crises took placein the mid-1970s when world copper prices felldramatically. Rather than undergo structural re-form, the government chose instead to borrowfrom abroad to maintain current consumption.This marked the beginning of escalating foreign

Figure 3.1 Poverty Rate in Ethiopia Under Different Growth Scenarios

National poverty ratio (%)under different growth scenarios

23

25

27

29

31

33

35

37

39

41

43

45

2003 2005 2007 2009 2011 2013 2015

Current growth path

Additional growth instaples crops

Additional growthin livestock

Additional growth innontraditional exports

Combined growth inall three subsectors

Source: Results from Ethiopian multimarket model.

Growth in all three subsectorscombined with market investments

debt, which would by the early 1990s make Zambiaone of the most indebted countries in the world.5

Despite foreign borrowing, the continued dete-rioration in the terms of trade and falling revenuesled to a reduction in social spending during the1980s. The substantial gains in social outcomes thatwere achieved during the first decade after inde-pendence slowly began to deteriorate. During theglobal recession of the 1980s, the government againrefused to embrace public-sector reform, this timeopting to reduce public investment, first in ruralinfrastructure and later in its own industrial enter-prises. This contraction of public investment and thesmall size of the private sector explain the country’spoor growth performance into the 1990s.

Beyond supporting urban-based industrial andpublic-sector employment, the government’s re-liance on the mining sector directed social spend-ing and political favor toward the urbanizedCopperbelt and Lusaka provinces (Bigsten andKayizzi-Mugerwa 2000). Perhaps the most impor-tant display of this urban bias is the food subsidiesfor urban areas. Ostensibly to ensure food security,these subsides became entrenched and were even-tually perceived as a right by the urban population.Attempts to reduce the scale of subsidies was metwith considerable opposition and led to riots inthe Copperbelt province.

More broadly speaking, the government’s agri-cultural policies, of which food subsidies formedpart, had a profound effect on poverty and vulner-ability in rural areas. Largely driven by its desire toprotect urban food prices, the government chose tosupport maize production throughout the country.This was done through publicly provided inputsubsidies and marketing support, and through pan-territorial price controls. The effect was to distortthe pattern of agricultural production, such thatover 80 percent of the land planted was devoted tomaize (Saasa 2003). Pan-territorial pricing promptedmany farmers to grow maize in areas that were notideally suited to this crop. This was particularlytrue for the more drought-prone southern prov-inces, which are better suited to drought-resistantsorghum and millet, and whose inhabitants there-fore became highly vulnerable to climatic changes(World Bank 2004e).

Apart from concentrating staple production in asingle crop, the maize bias, together with the over-valued exchange rate caused by copper, effec-tively undermined incentives to produce exportablecash crops. At the beginning of the 1990s, Zambia

exported few agricultural commodities and was anet importer of food. This is a significant indicatorof the failure of agricultural policies and of thecountry’s severe food insecurity. Burgeoning pub-lic debt and the resulting fall in rural infrastruc-ture investment exacerbated the situation. Manymore remote areas of the country became isolatedfrom input and output markets. Together the biastoward urban areas and maize production createdan untenable situation, which forced the govern-ment to implement a series of far-reaching struc-tural reforms.

Growth and poverty under structural adjustment

Due to the poor performance of the Zambian econ-omy at the end of 1980s, the newly elected govern-ment in 1991 chose a political platform based onthe implementation of a comprehensive structuraladjustment program. This program, which wasimplemented during the 1990s, included macro-economic stabilization, trade liberalization, priva-tization, and agricultural reforms. Each of thesepolicies was to play an important role in determin-ing the growth and poverty outcomes of the 1990s.

The government implemented a stabilizationprogram aimed at curbing inflation and creatingan environment conducive to private enterprise.Despite eventual success, the impact of deregu-lated financial markets, and the removal of foodsubsidies under agricultural reform, led to rapidincreases in consumer prices. Inflation during theearly 1990s undermined real incomes and raisedthe cost of living, especially in urban areas. Tradeliberalization and privatization also led to wide-spread job losses and contributed to rising urbanunemployment. Many semiskilled workers whohad previously been employed in state enterpriseswere forced into the informal sector, where jobsecurity and wages are substantially lower. Manyunskilled urban workers and their householdsmoved to rural areas, thus reversing the long-standing migrant labor system that had urbanized40 percent of the total population by 1991. Almost10 percent of the urban population moved to ruralareas, mostly into small-scale farming. This col-lapse of the formal economy explains much of thesubstantial increase in the incidence and depth ofpoverty in urban areas during the 1990s (table 3.7).

Agricultural reforms were also pronounced.The government abandoned its support of maizeby removing subsidies and decontrolling prices.



The loss of protection revealed the artificial prof-itability of maize and led to its rapid decline inimportance. Production halved during the 1990s,leading to rising poverty within rural areas. How-ever, farmers shifted production toward moreappropriate crops (Haggblade and Zulu 2003).While the rain-fed Northern and Luapula provincesreverted to cassava, the drier southern provincesplanted millet. Production of these two crops dou-bled in response to agricultural reforms. Further-more, the correction of the overvalued exchangerate—through privatization and falling copperprices—made export agriculture more internation-ally competitive. Cash crop production rose accord-ingly, with cotton, sugar, and horticulture showingrapid growth. Although initially hindered by adjust-ment costs and droughts, the incidence of povertydeclined in rural areas in the late 1990s, drivenmainly by improvements in cash crop production.Perhaps more importantly, the depth of povertyalso declined, reflecting the lower vulnerabilityand improved livelihoods of the poorest amongthe rural population.

Prospects for halving poverty by 2015

Although poverty reduction was evident in the late1990s, especially in rural areas, the observed over-all changes in poverty during the 1990s suggestthat Zambia is facing serious challenges to meetthe MDG of halving poverty by 2015. Using a

spatial dynamic Computable General Equilibriumand microsimulation model for Zambia, by theInternational Food Policy Research Institute analy-sis shows that under its current average of 4 percentGDP growth, the country’s poverty rate will be68 percent by 2015, only 7 percentage points lowerthan the current poverty rate of 75 percent. Lookingfurther into the future suggests that Zambia willnot be able to halve poverty until after 2040 unlesspro-poor growth is accelerated (figure 3.2). Themodel estimates that annual GDP growth of 8.8 per-cent is needed to achieve the target of halvingpoverty by 2015. These findings are consistentwith estimates based on static growth-povertyelasticities, which suggest that the necessary GDPgrowth rate lies between 7 and 9 percent (Thurlowand Wobst 2004).

Rural areas are expected to perform better thanurban areas, given the rising importance and exportpotential of the agricultural sector, and the gradualdecline of copper production and earnings.6 Thenew diversification, achieved at the cost of struc-tural adjustment, appears to have corrected some ofthe long-standing bias against agriculture and ruraldevelopment. Small- and medium-scale householdswithin rural areas benefit the most from diversifica-tion into agricultural export production. At the sub-sectoral level, it is small-scale-intensive cotton andmedium-scale-intensive horticulture that growfastest. Both sectors are expected to grow at around10 percent per year, a rate that is consistent with

Table 3.7 Poverty in Zambia During the 1990s

Incidence of Poverty (P0) Depth of Poverty (P1)

1991 1996 1998 1991 1996 1998

National 68.9 79.4 75.4 41.7 45.4 40.0Rural 88.0 90.1 85.6 61.3 56.7 49.7Urban 46.0 61.2 58.3 18.1 25.9 23.7

Central 69.8 84.1 78.9 38.2 47.8 43.9Copperbelt 55.5 70.8 67.0 22.3 31.7 31.3Eastern 84.3 89.0 82.7 58.8 58.2 45.3Luapula 83.9 88.4 85.4 53.3 51.2 47.8Lusaka 31.0 51.3 54.4 12.1 21.2 22.7Northern 83.6 90.8 85.0 55.5 58.3 47.7Northwestern 77.8 89.0 76.0 48.7 52.8 40.0Southern 78.4 86.1 78.4 52.5 51.1 44.2Western 84.5 89.2 90.3 59.1 59.7 54.3

Source: Thurlow and Wobst (2004).

recent trends in these crops. However, export cropgrowth is largely limited to the Eastern, Central,and Lusaka regions where access to input and out-put markets is better. Accordingly, while the currentgrowth path suggests that agriculture-led devel-opment is desirable, additional efforts targetinga majority of smallholders and broad agriculturalsectors are necessary for achieving significantpoverty reduction.

Agricultural growth is pro-poor growth

Although agriculture accounts for only 25 percentof GDP, it is still the main source of livelihood formost of the country’s population, including themajority of Zambia’s poor who live in rural areaswhere the incidence and severity of poverty is great-est. Apart from shifts toward more diverse stapleproduction, the performance of cash crops hasimproved greatly. However, with the exception ofcotton, which is grown in the Eastern province,most of the farmers engaged in cash crop produc-tion are located close to Lusaka or the country’smain transport routes. Thus, the recent gains fromagricultural growth have tended not to reach themore remote areas of the country. Given the needfor faster economic growth, and in order to assessthe constraints and opportunities for significantpoverty reduction, we consider the impact of accel-erating productivity growth in selected sectors such

that the overall GDP growth increases from its cur-rent 4 percent to 5 percent per year.7

Increased productivity stimulates economicgrowth and lowers poverty. However, there areconsiderable differences in the impact of targetingdifferent sectors. The first two scenarios contrast thepoverty effects of accelerating sector-wide growtheither in all agricultural sectors or all nonagricul-tural sectors. More rapid productivity growth underthe Agriculture-Led Growth Scenario leads tohigher sectoral growth for both staples and exportcrops. Resulting declines in nonagricultural growthare partly offset by cheaper agricultural inputs intodownstream textiles and food processing. Exportagricultural growth stimulates urban investmentby relaxing the foreign exchange constraint. Ruralhouseholds benefit directly from higher incomesand falling poverty, although smaller-scale stapleproducers benefit less than medium-scale exportersdue to the domestic market constraints they facefor their food crops. Urban households also bene-fit from lower food prices, leading to rising realincomes and falling poverty.

Accelerating productivity growth under theNonagriculture-Led Growth Scenario has very dif-ferent implications for poverty reduction.8 Non-agricultural production is more dependent onimported intermediates, thus exacerbating theforeign exchange constraint and underminingimport-intensive investment. High levels of exter-


Figure 3.2 Poverty Reduction Under the Current Growth Path (2001–2050)

68.3

38.2

75.4 78.5

48.9

85.6

51.4

20.4

58.3

0

10

20

30

40

50

60

70

80

90

100

2000 2010 2020 2030 2040 2050

Pove

rty

head

coun

t(P0

)(%

)

Rural

Urban

National

Source: Thurlow and Wobst (2004).


nal debt and limited opportunities for further bor-rowing mean that the large increases in importsmust be matched by more rapid export growth.Although urban poverty declines, the shift towardgreater trade in capital goods offsets private con-sumption growth. Furthermore, nonagriculturalgrowth reestablishes the bias against agriculturalexports and generates few backward linkagesinto agriculture. Therefore, while rural householdsbenefit from increased demand in urban areas, theoverall effect of nonagricultural growth remainsrelatively small.

Despite the less optimistic results for broadnonagriculture-led growth, agricultural process-ing within the manufacturing sector does in factrepresent a potential area for growth and povertyreduction. Agribusiness benefits smallholdersthrough greater demand for their crops, and pro-vides a source of income for the growing nonfarmpopulation. For example, hand-milling in smallerand more remote urban areas has shown consider-able growth in recent years. Furthermore, on-farmprocessing allows smallholders to raise the valueadded of their produce. Scenarios that focus on theagroprocessing sectors take into account these sec-tors’ stronger backward linkages into agriculture.Although not shown here, model simulations sug-gest that increased productivity in the agroprocess-ing sectors is likely to benefit urban and nonfarmhouseholds more than other households, largelybecause they have better access to the required cap-ital (Thurlow 2004). Relieving the credit constraintfacing many farmers should extend the potentialbenefits of this sector.

Since agriculture-led growth appears to offergreater opportunities for broad-based povertyreduction, two additional scenarios contrast theeffects of accelerating economic growth througheither the staples or export crop sectors. The Staples-Led Growth Scenario concentrates agriculturalgrowth within the staples-producing sectors at theexpense of export agriculture. However, increasedproductivity of staple crops has little effect onpoverty among small-scale farm households. Thisis due to current market constraints, which causehigher production to translate into falling prices.Lower prices in staple crops do not suggest thatdomestic demand for staples is insufficient, butrather that the current structure of the domesticmarket limits its ability to absorb substantiallyhigher levels of supply. Poor market access there-fore represents the dominant constraint to growth

and poverty reduction through the expansion ofstaple production.

Exportable commodities allow farmers to getaccess to foreign markets, and, hence, increasedsupply results in increased exports without depress-ing prices. This can be seen in the Export-Crop-LedGrowth Scenario. The changes in poverty rate fol-lowing export crop expansion indicate that ruralmedium-scale households would be better off ifproductivity of traded commodities could be raised.Again it should be stressed that only less-remotehouseholds that already have adequate marketaccess are likely to benefit. This is evident in thestrong declines in poverty within the Eastern andCentral provinces.

The scenarios presented so far suggest thatenhancing productivity within agriculture gener-ates pro-poor outcomes. Given that a majority ofthe poor live in rural areas, it appears that agricul-ture has an important role to play in helping thecountry achieve faster growth and poverty reduc-tion. However, the benefits from agriculturalgrowth are unlikely to reach all rural households.Poor access to markets and credit has been a per-sistent problem in rural areas and continues toundermine rural development. Recently achievedmacroeconomic stability and the increasing par-ticipation of the private sector in the rural econ-omy suggest that market access might already beimproving, at least in the less remote provinces.Private-sector to initiated outgrower schemes haveproven highly successful in the cotton sector and,through credit provision, have allowed smallhold-ers to become the largest suppliers of raw cotton inZambia. However, the high cost of capital neededfor horticultural exports and agribusiness produc-tion has effectively limited the participation ofsmallholders in these sectors. Furthermore, lowproductivity and labor shortages at harvestinglimit production, indicating the need for new tech-nologies and capital investment (Deininger andOlinto 1999). Inadequate financing sources, poormarket access, and low levels of investment there-fore represent the major constraints facing agricul-tural growth in Zambia.

Improving market access to encourage pro-poor growth

The agriculture-led growth scenarios have empha-sized the importance of market access and its rolein determining whether rural farmers can benefit

from economic growth. Transportation costs cur-rently account for between 60 and 70 percent ofthe cost of production (Lofgren, Thurlow, andRobinson 2004). These costs, which are high evenby regional standards, limit farmers’ ability tomarket their produce and contribute to high pricesand, hence, the prevailing poverty in the country.Although Zambia has an extensive road network,which supports the dominant modes of trans-portation, the system has deteriorated over thelast three decades. The government has identifiedincreased investment in transportation infrastruc-ture as critical for growth and poverty reduction.Market access is especially limited in many ruralareas. Only 18 percent of rural households arewithin 5 kilometers of input markets, and few ofthe more remote households have access to healthand education facilities. These households’ accessis limited due to either a lack of roads or the poorquality of the existing network.

This section contrasts the impact of building newfeeder roads in rural areas with new paved andgravel roads in less remote rural and urban areas.The scenarios consider a 10 percent increase inthe provision of either feeder or paved roads.Improved infrastructure provision reduces transac-tion costs in specific sectors in the economy. Inthe case of feeder roads, lower transaction costsare likely to benefit only rural agricultural pro-duction. Paved roads, on the other hand, shouldbenefit nonagricultural production and exportagriculture. The latter is due to the existing con-centration of export agriculture along the coun-try’s main road networks. Furthermore, based onthe current geographic distribution of produc-tion, paved roads will reduce the transaction costsin both the domestic and export markets, whereasfeeder roads will reduce the transaction costs indomestic markets only.

The construction of both feeder and paved roadsincreases growth and reduces poverty. However,improving market access for export crops has amuch stronger growth effect due to the strong pos-itive externalities that less-remote paved roads pro-vide to other nonagricultural sectors. Despite slowergrowth and less impressive export performance,building feeder roads and improving market accessfor staple crops is better at reducing poverty sincethe benefits accrue to the larger and poorer small-scale farm population. Export crops, by contrast,benefit the smaller population of medium-scalefarm households, and to a lesser extent, urban

households. Therefore, while infrastructure andmarket access consistently strengthen povertyreduction, especially in rural areas, there appearsto be a tradeoff between poverty reduction andgrowth.

Summary for Zambia

Zambia has undergone substantial reforms, whichthe country’s recent positive performance suggestsmight have been a prerequisite for renewed growth.Furthermore, the rising poverty and falling socialoutcomes of the pre-reform period suggest thatstructural adjustment might also have been a pre-requisite for poverty reduction. However, short-runadjustment costs raised poverty during the 1990s,moving Zambia further away from achieving thegoal of halving poverty by 2015 (table 3.8). A veryhigh and unlikely growth rate would have to beachieved if the MDG poverty target is to be met.

Although meeting target 1 of MDG 1 appears tobe beyond Zambia’s grasp, the country’s success atencouraging diversification is likely to increase itsrate of pro-poor growth. However, several con-straints remain. The most important of these con-straints is slow agricultural growth, which has beenlargely undermined by low productivity and in-adequate rural infrastructure.

Conclusions

As the two case studies illustrate, reaching theMDGs, and particularly MDG 1, will require muchgreater growth than both Ethiopia and Zambia havecurrently achieved. In Zambia, the GDP growth raterequired to halve poverty is around 8.8 percent,much higher than the rate of 6 percent needed inEthiopia. The unique characteristics of each coun-try’s economy and poverty profile account for thisvariation in growth rates. Specifically, Ethiopia is apredominantly subsistence economy with agricul-ture contributing 52 percent to GDP and withalmost 85 percent of the population living in ruralareas. By contrast, mining traditionally has dom-inated economic growth in Zambia and, in turn,has marginalized the country’s agricultural sec-tor. Indeed, 60 percent of Zambia’s populationlive in rural areas while agriculture contributesonly 22 percent to the country’s GDP. Moreover,Ethiopia’s national poverty rate is 44 percent witha concentration of poverty in the rural areas. InZambia, however, the collapse of mining exacer-



bated urban poverty and contributed to a nationalpoverty rate of 75 percent. Not only is the incidenceof poverty higher in Zambia than in Ethiopia, butthe depth of poverty, which measures the distrib-ution of the poor, is higher as well. In fact, thepoverty gap in Ethiopia is 12 percent comparedwith 26 percent in Zambia. In other words, Zambiahas both a higher proportion of its population livingbelow the poverty line and a higher concentration ofindividuals at the low end of the income distrib-ution. Consequently, Zambia requires faster GDPgrowth than Ethiopia to halve poverty by 2015.

Achieving these rates of growth specificallyrequires greater productivity in the agriculturalsector. The unique structure of each country’s econ-omy determines the impact of this agriculturalgrowth on poverty reduction as well as which agri-cultural subsectors are the most promising. In bothcountries, domestic demand for staples is high,and growth in the staples sector could increase theincomes of the rural poor and improve food secu-

rity in both rural and urban areas. Yet, in order forgrowth in staples to have its intended impact, sig-nificant market access constraints need to be over-come. Otherwise, increased staple production willcause a decline in local food prices and diminishfarmers’ incomes.

In Zambia, the expansion of export crops offersopportunities for increased growth because farmersdo not encounter the same constraints in access-ing foreign markets as they do domestic markets.Therefore, increased supply of export crops, even inthe face of domestic market constraints, would typ-ically not result in a decline in farmers’ incomes,although the poverty impact is lower than for sta-ples crops. Agro-processing also has the potential tocomplement growth in other agricultural sectors.By contrast, in Ethiopia, growth in nontradi-tional exports alone will not have a large impacton poverty reduction. However, a combination ofnontraditional export growth with livestock andstaple production growth can significantly reduce

Table 3.8 Growth and Poverty Reduction Under the Alternative Zambian Growth Scenarios (2001–2015)

Current Agriculture- Nonagriculture- Staples- Staples Export-Crop- Export-Crops growth path led growth led growth led growth market access led growth market access

Average annual growth rate, 2001–2015 (%)

Gross domestic product 4.0 5.0 5.0 5.0 5.1 5.0 5.6Agriculture 4.6 7.7 4.2 7.8 8.1 7.1 8.6Staple crops 4.1 7.3 4.0 8.0 8.4 4.0 3.4Export crops 10.2 13.4 7.1 7.0 3.5 22.8 27.2Mining 1.9 1.9 1.9 1.9 1.9 1.9 1.9Manufacturing 4.5 4.3 5.8 4.3 4.7 4.2 4.6Services 3.9 4.1 5.6 4.0 4.0 4.7 4.7

Final year poverty headcount, 2015 (%)

National poverty 68.3 59.4 63.9 59.5 54.5 62.0 55.8Rural 78.4 68.1 76.4 68.1 61.2 72.3 64.2Small-scale 79.0 68.1 77.2 68.0 60.4 73.0 64.4Medium-scale 69.5 56.3 65.2 59.0 54.5 55.9 45.1Urban 51.4 45.0 42.9 45.2 43.3 44.8 41.8Central 73.8 66.4 70.1 66.5 64.2 67.7 61.9Copperbelt 61.6 55.9 54.5 56.1 54.1 56.2 52.1Eastern 67.1 56.0 68.3 62.5 58.1 51.0 36.7Luapula 79.2 68.4 76.1 65.8 61.8 74.1 68.4Lusaka 45.5 40.3 36.3 40.2 39.1 40.7 39.1Northern 79.7 65.9 76.5 62.5 55.1 75.8 70.9Northwestern 71.1 54.2 67.5 52.6 47.9 66.5 61.3Southern 72.7 65.6 68.4 65.6 58.9 68.0 59.2Western 87.3 77.9 83.9 76.6 56.7 83.3 78.0

Source: Zambia CGE-micro model results from Thurlow and Wobst (2004).

poverty. For both countries, the level of technol-ogy and capital investment required to engage innontraditional export production, as well as agro-processing, circumscribes the benefits to smallhold-ers. Moreover, growth in nontraditional exportsparticularly benefits those farmers concentratedaround cities and where market access and infra-structure are well developed.

Indeed, both case studies suggest that the lack ofphysical infrastructure represents a serious prob-lem that needs to be addressed, not only for agri-cultural growth to occur but also for agriculturalgrowth to reduce poverty, particularly in the mostremote rural areas. By concurrently improvingrural infrastructure and market development, agri-cultural growth might help Ethiopia to achieveMDG 1. In Zambia, on the other hand, reachingMDG 1 will be more difficult, particularly since thecountry is starting with a higher national povertyrate. Nonetheless, increased construction of roads,especially feeder roads, is no less crucial.

Notwithstanding Sub-Saharan Africa’s vastdiversity, the similarities shared by Zambia andEthiopia have important implications for Sub-Saharan Africa as a whole. In particular, agricul-tural growth is absolutely essential for alleviatingthe most severe poverty, especially in rural areas.However, the nonagricultural sector has an impor-tant role in assuring that this growth reaches theintended beneficiaries. Consequently, efforts toincrease yields and expand production areas mustbe appropriately sequenced with investments inroads, irrigation, and storage facilities as well asgreater credit and fertilizer provision for small-holders and an enabling macroeconomic environ-ment that encourages private sector involvement.

GLOBAL QUANTIFIEDASSESSMENT OF SELECTED MDG TARGETS: IMPACT-WATERMalnutrition affects nearly one-third of all childrenunder five years of age in developing countries,174 million children in 1990. Malnourished childrenhave lowered resistance to infection; they are morelikely to die from common childhood ailments likediarrheal diseases and respiratory infections, andfor those who survive, frequent illness affects theirnutritional status, locking them into a vicious cycleof recurring sickness and faltering growth. Theirplight is largely invisible: three-quarters of the chil-

dren who die from causes related to malnutritionwere only mildly or moderately undernourished,showing no outward sign of their vulnerability.More than half of childhood deaths are associatedwith being underweight, and malnourished chil-dren who survive into adulthood are more likelyto suffer from chronic illness and disability, andhave a higher probability of reduced physical andintellectual productivity (de Onis and others 2004;Pelletier and others 1994; Smith and Haddad 2000;UNICEF 2004). Poverty, low levels of education,and poor access to health services are major con-tributors to childhood malnutrition, a complex issuethat requires tackling on a wide number of fronts,including (UNICEF 2004):

• “Ensuring food security for poor households,both enough food and the right kinds of food;

• Educating families to understand the specialnutritional needs of young children, notablythe value of breastfeeding and the impor-tance of introducing suitable complementaryfoods at the right age;

• Protecting children from infections, by suchmeasures as immunization against commonchildhood diseases and provision of safe waterand sanitation;

• Ensuring that children receive quality carewhen they fall ill;

• Shielding them from the micronutrient defi-ciencies that can bring death and disabil-ity, especially iodine, iron and vitamin Adeficiencies;

Paying special attention to the nutritional needsof girls and women, since chronically undernour-ished women tend to bear low-birthweight babiesand so perpetuate the vicious cycle of malnutritioninto the next generation.” These devastating effectsof malnutrition led world leaders to choose “tohalve, between 1990 and 2015, the proportion ofpeople who suffer from hunger” as one of the tar-gets of MDG 1. One of the specific indicators chosenfor this target and goal is the prevalence of under-weight children under five years of age.

We use the IMPACT-WATER model to projectthe proportion of malnourished children under thebaseline and one alternative scenario, titled theMDG scenario.9 Malnutrition is defined here asunderweight (proportion of under-fives falling be-low minus 2 standard deviations from the medianweight-for-age standard set by the U.S. NationalCenter for Health Statistics and the World Health



Organization). This standard is adopted by manyUnited Nations agencies in assessing the nutri-tional status of preschool children in developingcountries. Other, less commonly used indicatorsfor child malnutrition include stunting (propor-tion of children under five falling below minus 2standard deviations from the median height-for-ageof the reference population), and wasting (pro-portion of children under five with weight-for-height falling below minus 2 standard deviationsfrom the median weight-for-height of the referencepopulation).

IMPACT-WATER generates projections of thepercentage and number of malnourished preschoolchildren (zero to five years old) in developing coun-tries. Projections for the proportion and number ofmalnourished children are derived from an esti-mate of the functional relationship between the per-centage of malnourished children, the projectedaverage per-capita kilocalorie availability of food,and nonfood determinants of child malnutrition,including the quality of maternal and child care(proxied by females’ status relative to men as cap-tured by the ratio of female-to-male life expectancyat birth), education (proxied by the share of femalesundertaking secondary schooling), and health andsanitation (proxied by the percentage of the popu-lation with access to safe drinking water). The equa-tions used to project the percentage and numbers ofmalnourished children are as follows:

and

where %MAL is the percentage of malnourishedchildren, KCAL is per-capita kilocalorie availabilityestimated in IMPACT-WATER, LFEXPRAT is theratio of female to male life expectancy at birth, SCHis the percentage of females with secondary educa-tion, WATER is the percentage of the populationwith access to safe water, NMAL is the number ofmalnourished children, and POP5 is number of chil-dren zero to five years old.

The regression equation was derived based ona fixed-effects model of pooled, cross-section time-series data from 63 developing countries coveringthe 1970s, 1980s, and 1990s from a variety of sourcesfor both dependent and independent variables. Themajority of the data on prevalence of child malnu-

NMAL MAL POPt t t= ×% 5

% . .. .

MAL KCAL LFEXPRATSCH WATER

t t t

t t

= ( ) −− − −

25 24 71 760 22 0 08

� ln

trition came from the World Health Organization’sGlobal Database on Child Growth and Malnutri-tion (WHO 2004), with other sources includingthe United Nations Administrative Committee onCoordination–Subcommittee on Nutrition (ACC/SCN 1996) and World Development Indicators (WorldBank 1997). Sources for explanatory factor datainclude calorie availability from the Food andAgriculture Organization FAOSTAT database (FAO1998); female secondary enrollment data from theUnited Nations Educational, Scientific and CulturalOrganization UNESCOSTAT database (UNESCO1998) and World Development Indicators (World Bank2004c); and female-to-male life expectancy ratiosfrom World Development Indicators (World Bank 1997,2004c). For greater detail on sources, data coverage,specific observations used, and model estimationprocedures and tests, see Smith and Haddad (2000).

Table 3.9 presents 1990 and 1995 estimates forchild malnutrition (low weight-for-age) for thereporting regions for this item in IMPACT-WATER,as well as the 2015 MDG target estimates (halving,between 1990 and 2015, the proportion of people[that is, under fives] who suffer from hunger) cal-culated based on 1990 values. Table 3.10 presentsthe same values for access to safe drinking water,related to Target 10 of MDG 7, to “halve, by 2015,the proportion of people without sustainable accessto safe drinking water and basic sanitation.” As canbe seen in Table 3.9, child malnutrition estimatesbetween 1990 and 1995 declined, increased, or stag-nated. This is one of several caveats that need to betaken into account when assessing the possibilityof achieving the MDGs based on the indicators.For example, while the WHO global database onchild growth and malnutrition for Myanmar reportsa proportion of 32.4 percent preschool childrenmalnourished in 1990, the value for 1995 was42.9 percent, and the latter value is used for theIMPACT-WATER base year calculations. Further-more, several countries do not have estimates forchild malnutrition going back to 1990. In thesecases, country assessment reports often suggest ahalving of child malnutrition levels by 2015 withreference to the earliest year with reliable data. Forexample, in the case of Afghanistan, the countryreport suggests a reduction in child malnutrition(underweight) from 48 percent in 2000 to 24 percentby 2015 (UNDP 2004a) as an indicator for MDG 1.

Moreover, the data presented in table 3.9 donot capture rural-urban disparities, but as figure 2.4shows, child malnutrition rates are typically higher

for rural areas. As table 3.9 shows, on a regionalbasis, the incidence of low weight-for-age in 1990was lowest in Latin America and the Caribbeanamong all developing regions, at just under 9 per-cent, followed by the West Asia and North Africa(WANA) region, at around 11 percent. On the otherhand, the level was highest for South Asia, at54.2 percent in 1990, followed by Southeast Asia.The incidence of low weight-for-age was some-what lower in Sub-Saharan Africa, at 26.8 percent.Regional averages mask large differences withinregions; for example, the average for West Asiaand North Africa includes Lebanon, with a share of3 percent malnourished children in 1996, but alsoYemen, where the share was 38.1 percent in the

same year (WHO 2004). Even more important thanpoint estimates are trends in malnutrition over time.

Over the last 30 years, South Asia, where thelargest number of malnourished children reside,has made substantial progress, with the share ofmalnourished children declining from well over70 percent in 1970 to under 50 percent by 1996–97.However, the absolute number of malnourishedchildren declined by only 7 million children overthis period due to continued rapid populationgrowth. The most impressive developments tookplace in East Asia, where the number of malnour-ished children decreased by 50 percent, or 40 mil-lion children, between 1970 and 1997. Latin Americaalso reduced the number of malnourished children


Table 3.9 Child Malnutrition Estimates (Underweight), 1990 and 1995, with 2015 MDG Indicator Target

1990 estimates 1995 estimates 2015 target

Percent

Latin America 8.7 9.1 4.4Sub-Saharan Africa 26.8 13.4

Nigeria 35.3 39.1 17.7Northern Sub-Saharan Africa 40.3 40.0 20.2Central and Western Sub-Saharan Africa 27.9 27.8 13.9Southern Sub-Saharan Africa 28.5 26.9 14.3Eastern Sub-Saharan Africa 26.7 27.6 13.4

West Asia and North Africa* 11.0 13.2 5.5South Asia 54.2 27.1

India 54.7 53.4 27.3Pakistan 40.2 38.2 20.1Bangladesh 67.1 56.3 33.5Other S Asia 49.0 41.2 24.5

South East Asia 35.2 17.6Indonesia 36.8 34.0 18.4Thailand 22.0 18.9 11.0Malaysia 25.0 20.1 12.5Philippines 33.5 29.6 16.8Vietnam 43.6 44.9 21.8Myanmar 32.4 42.9 16.2Other SE Asia 47.0 40.0 23.5

China 19.0 17.4 9.5Developing 30.2 15.1

Notes: *The value for West Asia and North Africa was estimated, as details for several countries in the region are not available. Latin Americaincludes the Caribbean; N SSA stands for Northern Sub-Saharan Africa; C and W SSA stands for Central and Western Sub-Saharan Africa; S SSA stands for Southern Sub-Saharan Africa; E SSA stands for Eastern Sub-Saharan Africa; WANA stands for West Asia and North Africa;Other S Asia stands for Other South Asia; Other SE Asia stands for Other Southeast Asia. The regional disaggregation is presented in Annex 2.No estimates for child malnutrition are made for parts of East Asia (Democratic People’s Republic of Korea, Macao, and Mongolia) and for therest of the world, including Cape Verde, Fiji, French Polynesia, Kiribati, New Caledonia, Papua New Guinea, Seychelles, and Vanuatu. 1995values are incorporated in IMPACT-WATER. The 2015 target is based on 1990 estimates.Sources: 1990 estimates based on the WHO global database on child growth and malnutrition (WHO 2004); 1995 estimates also based on the WHO global database on child growth and malnutrition, but accessed in 1997, as well as additional sources (World Bank 1997 and ACC/SCN 1996).

by half, but the number of malnourished childrenin WANA today is virtually the same as it was inthe 1970s. The most troubling region, however, is Sub-Saharan Africa, where the number of mal-nourished children increased by over 75 percent or12.9 million over the last 30 years. In 1970, roughly1 out of 10 malnourished children resided in Sub-Saharan Africa, today 1 in 5 do.

Despite Sub-Saharan Africa’s discouraging dev-elopments, globally significant improvements inchild malnourishment have been made, as shownby a sharp reduction in the proportion of malnour-ished children. However, because of populationgrowth, the absolute number of malnourished chil-

dren has fallen much less sharply. In addition,despite the long-term improvement in most regions,progress has slowed in recent years, causing addi-tional concern over future prospects for reducingchild malnutrition (Rosegrant and Meijer 2002).

Regarding the proportion of people with accessto safe drinking water, rural-urban disparitiesare important and are typically reported fordeveloping-country regions (WHO/UNICEF 2001).An important issue regarding the access to safedrinking water data is that different surveys under-taken result in very different estimates even forthe same year. For example, in Ghana, the GhanaLiving Standards Survey of 1988 estimated that


Table 3.10 Access to Safe Drinking Water, 1990 and 1995, with 2015 MDG Target Indicator and 2015 IMPACT-WATER Baseline Estimates

2015 IMPACT-WATER1990 estimates 1995 estimates 2015 target baseline estimates

Percent

Latin America 82.0 77.5 91.0 80.2Sub-Saharan Africa 53.0 76.5

Nigeria 53.0 50.0 76.5 64.9Northern Sub Saharan Africa 41.6 37.9 70.8 51.3Central and Western Sub-Saharan Africa 48.8 54.5 74.4 66.6Southern Sub-Saharan Africa 47.8 51.2 73.9 66.7Eastern Sub-Saharan Africa 48.3 47.2 74.2 61.9

West Asia and North Africa 84.0 81.9 92.0 88.2South Asia 72.0 86.0

India 68.0 81.0 84.0 89.6Pakistan 54.2 60.0 77.1 72.5Bangladesh 83.0 79.0 91.5 82.9Other South Asia 54.0 60.3 77.0 71.0

South East Asia 70.1 85.1Indonesia 71.0 60.0 85.5 73.3Thailand 80.0 89.0 90.0 94.3Malaysia 80.9 77.0 90.5 87.2Philippines 87.0 84.0 93.5 90.3Vietnam 55.0 43.0 77.5 62.4Myanmar 64.0 60.0 82.0 69.4Other SE Asia 29.1 59.6 64.5 67.7

China 71.0 67.0 85.5 75.6Developing 74.0 87.0

Notes: Latin America includes the Caribbean; N SSA stands for Northern Sub-Saharan Africa; C and W SSA stands for Central and WesternSub-Saharan Africa; S SSA stands for Southern Sub-Saharan Africa; E SSA stands for Eastern Sub-Saharan Africa; WANA stands for West Asia andNorth Africa; Other S Asia stands for Other South Asia; Other SE Asia stands for Other Southeast Asia. The regional disaggregation is presentedin Annex 2. No estimates are made for child malnutrition for parts of East Asia (Democratic People’s Republic of Korea, Macao, andMongolia), and for the rest of the world, including Cape Verde, Fiji, French Polynesia, Kiribati, New Caledonia, Papua New Guinea,Seychelles, and Vanuatu. 1995 values are incorporated in IMPACT-WATER. The 2015 target is based on 1990 estimates. 1995 and 2015IMPACT baseline estimates were used for the baseline run.Sources: 1990 estimates based on the World Development Indicators (World Bank 2002) and WHO/UNICEF (2001); 1995 estimates alsobased on World Development Indicators, but accessed in 1997.

36 percent of both the urban and rural populationhad access to an improved drinking water source(with different water sources for rural and urbanareas), while the Ghana Demographic and HealthSurvey of 1988 reported that 84 percent of the urbanpopulation and 28 percent of the rural populationhad safe access to drinking water. Finally, theInternational Drinking Water Supply and Sanita-tion Decade Review of National Progress (as ofDecember 1988) estimated that in 1988, 93 percentof the urban population and 39 percent of the ruralpopulation had access to safe drinking water inGhana (WHO/UNICEF 2001 for Ghana).

On a regional basis, in 1990 access to an im-proved source of drinking water was estimated to be highest in West Asia and North Africa at84 percent, followed by Latin America with 82 per-cent and Asia with around 70 percent. Coveragewas lowest in Sub-Saharan Africa with 53 percent.

The data for the other two noncaloric parame-ters incorporated in the child malnutrition equationdescribed above—the ratio of female-to-male lifeexpectancy at birth, and the proportion of femalesattending secondary schooling—are presented inAppendix Table 1 for 1995 and 2015 (estimates forbaseline and MDG scenarios). The indicator for theshare of female secondary schooling is a reflec-tion of Target 3—to “ensure that, by 2015, childreneverywhere, boys and girls alike, will be able tocomplete a full course of primary schooling”—ofthe second MDG, “achieving universal primaryeducation,” and of Target 4—“eliminate gender dis-parity in primary and secondary education prefer-ably by 2005 and to all levels of education no laterthan 2015”—of MDG 3, “promote gender equalityand empower women.”

In 1995, the ratio of female-to-male life expect-ancy at birth for 195 countries with available dataranged from less than 1 (0.9876 for Nepal and0.9982 for Uganda) to a high above 1.2 (Latvia,1.2023, Estonia, 1.2042, and the Russian Federation,1.2305). Among the developing-country reportingregions, Latin America and the Caribbean had thehighest ratio of female-to-male life expectancy at1.0989 in 1995, whereas Bangladesh had the low-est value (1.0072), followed by India with 1.0088(Appendix Table 1). Regarding the proportion offemale secondary schooling, in 1995, among theIMPACT-WATER reporting regions, the share washighest for the Philippines at 78 percent, followedby Malaysia at 63 percent, and China at 62 percent.The share was lowest in Northern and Eastern Sub-

Saharan Africa at 11 and 12 percent, respectively(Appendix Table 1).

For the analysis here, two scenarios were used.The first is a baseline scenario that reflects our bestestimates of future trends in food supply, demand,and trade, as well as changes in the noncaloricparameters determining child malnutrition. Thesecond scenario, titled MDG Scenario, attempts toreach the child malnutrition indicator of MDG 1by 2015 through increases in economic and agri-cultural growth and complementary improvementsin the service sectors, particularly for those coun-tries and regions least likely to make sufficientimprovements for this target if current trends pre-vail. Basic parameters for both scenarios are listedin tables 3.11 and 3.12.

Table 3.12 presents the increases in incomegrowth required between 1995 and 2015 under theMDG scenario compared to the baseline scenario.Nonagricultural income growth rates for LatinAmerica and Asia, with the exception of South Asia,are increased to 25 percent. Rates for West Asia andNorth Africa are increased to 6 percent per year, andrates for the two regions considered least likely tomeet the MDG targets, South Asia and Sub-SaharanAfrica, are increased to 8 percent per year.

Thus, the following items have been includedin the MDG scenario:

• Nonagricultural income growth increasesin developing regions by 25–150 percent,with particularly rapid increases projectedfor those regions least likely to reach themalnutrition indicator, Sub-Saharan Africaand South Asia.

• Annual irrigated yield growth per hectareincreases by 50 percent and rainfed yieldgrowth increases by 25 percent, for all devel-oping countries and regions (cereals, roots andtubers, and soybeans). The increase of cropproductivity is due to a large expansion ofinvestments in agricultural research and irriga-tion infrastructure, enhanced property rightsto land and water, improved coordinationamong agencies, and more transparent andaccountable use of funds. Broad-based growthin production of meat and livestock productsis assumed to be generated by more rapidexpansion in animal numbers, with a 50 per-cent increase in annual numbers growth forall developing countries and regions. Bothitems, crop and livestock production growth,



will increase food availability in developingregions and therefore reduce the share of mal-nourished children.

• Increased investment levels in health andaccess to improved drinking water sourceswill help achieve MDG 7 Target 10.

• Increased investment levels in education willhelp reach higher levels of secondary enroll-ment of girls.

• Increased investment in female and maternalwell-being will help to reach improvementsin the ratio of female to male life expectancyat birth.

Thus, the MDG scenario combines two broadcourses for improving food security and reducingpoverty in developing regions: the first way isthrough broad-based and rapid agricultural pro-ductivity and economic growth to increase effec-tive incomes, effective food demand, and foodavailability; and the second is through investmentsin education, social services, and health (proxied inthe model by female secondary enrollment rates,ratio of female-to-male life expectancy at birth, andaccess to clean water).

Results

The results of the 2015 baseline and MDG scenariosimulations based on IMPACT-WATER show thatthe indicator for halving the prevalence of childmalnutrition can be reached (or even overshot) insome countries and regions, even in the baselinescenario, while it is unlikely to be met, even underthe MDG scenario, for others.

The combined effect of the parameter changesis first, an increase in projected constant-price(rainfed and irrigated) cereal yield growth for alldeveloping regions, 2005–2015, from 1.39 percentannually under the baseline to 1.88 percent per yearunder the MDG scenario; second, an increase ofcrop-irrigated harvested-area growth from 0.16 per-cent per year under the baseline to 0.21 percentannually under the MDG scenario; and third, anincrease from baseline livestock numbers growth of1.43 percent per year during 2005–2015 to 2.17 per-cent annually under the MDG scenario.

Under the MDG scenario, increased per-capitaincome and lower food prices resulting from foodproduction increases lead to higher levels of per-capita food consumption. Increased per-capita food

Table 3.11 Parameters for Developing Countries, Baseline and MDG Scenarios

Parameter Baseline scenario MDG scenario

Nonagricultural income growthPopulation growthLivestock numbers growth

Food crop yield growth

Irrigated area growth

Access to water

Female secondary education

Female-to-male life expectancy ratio at birth

Source: IFPRI-IMPACT parameters.

See table 3.12UN medium variantOutput numbers growth:

1.43%/yr, 2005–2015

Output cereal yield growth:1.39%/yr, 2005–2015

Output irrigated harvested areagrowth: 0.16%/yr, 2005–2015

Interpolated from IFPRI estimatesfor 2020 (see table 3.10)

Interpolated from IFPRI estimatesfor 2020 (see Appendix Table 1)

Interpolated from IFPRI estimates for 2020 (see Appendix Table 1)

UN medium variantIncrease in numbers growth of

animals slaughtered in developingcountries by 50% beginning in2005

Increase irrigated yield growth by50% for cereals, roots, and tubersand soybean in developing regions;increase rainfed yield growth by25% for same crops and countries,all beginning in 2005

Increase irrigated area growth by afactor of 0.1

Following MDG Target 10 (see table 3.10)

Increased from 1995 by between20–150% to reach malnourishmentindicator, but not higher than 95%(see Appendix Table 1)

Adjusted to reach malnourishmentindicator (see Appendix Table 1)

availability translates into higher per-capita caloricintake, which in turn reduces the share of mal-nourished preschool children. Strong growth inpublic investment in the social sectors, includingtotal social expenditures, and particularly expen-ditures for education and water and sanitation,result in further reductions in the share of mal-nourished children.

The projected levels of calorie availability for themajor IMPACT-WATER regions for the two scenar-ios are shown in figure 3.3. Compared to the base-line scenario, per-capita kilocalorie availabilityimproves for all developing regions in 2015 underthe MDG scenario, apart from a very small decline

in Latin America. The daily kilocalorie availabilityincrease is largest by far for the Sub-Saharan Africaregion at 515 kilocalories, driven by very rapidincome growth (8 percent per year) and rapid agri-cultural growth. In South Asia, caloric availabilityincreases by 128 kilocalories, driven by the samefactors as in Sub-Saharan Africa. WANA, whichalready has high levels of kilocalorie availability in1995 and the 2015 baseline scenario, is expected tofurther increase availability by a smaller amount,65 kilocalories.

Figure 3.4 presents the results for child malnutri-tion for the group of developing countries for 1995and 2015 (baseline, MDG scenario, and MDG indi-


Table 3.12 Income Growth Requirements for Baseline and MDG Scenarios, 1995–2015

Country/Region Baseline scenario MDG scenario

Percent per year

Latin America (25% increase)Mexico/Brazil/Other Latin America and Caribbean 3.6 4.5Colombia 3.8 4.8Argentina 4.5 5.6

Sub-Saharan AfricaNigeria 3.8Northern Sub-Saharan Africa 3.3Central and Western Sub-Saharan Africa 3.8 8.0Southern Sub-Saharan Africa 3.2Eastern Sub-Saharan Africa 3.5

West Asia and North AfricaEgypt 4.0Turkey 4.5 6.0Other West Asia and North Africa 3.5

East Asia (25% increase)China 6.0 7.5South Korea 5.0 6.3Other East Asia 3.5 4.4

South AsiaIndia 5.8Bangladesh 4.8 8.0Pakistan 4.5Other South Asia 4.5

South East Asia (25% increase)Indonesia 4.5 5.6Thailand/Malaysia 5.2 6.5Philippines 5.0 6.3Vietnam 5.8 7.3Myanmar/Other South East Asia 4.0 5.0

Notes: Latin America includes the Caribbean; N SSA stands for Northern Sub-Saharan Africa; CW SSA stands for Central and Western Sub-SaharanAfrica; S SSA stands for Southern Sub-Saharan Africa; E SSA stands for Eastern Sub-Saharan Africa; WANA stands for West Asia and North Africa; O S Asia stands for Other South Asia; O SE Asia stands for Other Southeast Asia; O E Asia stands for Other East Asia. The regional disaggregationis presented in Annex 2.Source: IFPRI-IMPACT parameters.


0

500

1,000

1,500

2,000

2,500

3,000

3,500

SSA S Asia SE Asia E Asia LAC WANA

19952015 baseline

Kilo

calo

ries

/cap

/day

2015 MDG

Figure 3.3 Per-Capita Kilocalorie Availability in 1995 and Projections to 2015, Baseline and MDG Scenarios

Source: IMPACT-WATER Simulations, IFPRI 2004.

31

24

1715

0

5

10

15

20

25

30

35

1995 2015 baseline 2015 MDG 2015 target

Perc

ent

Figure 3.4 Child Malnutrition, Developing Countries,1995, 2015 Baseline, 2015 MDG Scenario,and 2015 Target


cator target). On average, the MDG scenario missesthe MDG 2015 indicator by only 1.9 percentagepoints. The changes in agricultural and comple-mentary social indicators result in a reduction intotal child malnutrition from 24.4 percent under thebaseline to 17.0 percent under the MDG scenario.These declining shares reflect large reductions in

the absolute number of malnourished children:from 161 million children in 1995 to 131 millionchildren under the baseline, and 91 million childrenunder the MDG scenario.

As can be seen in figure 3.5, the Latin Americaand Caribbean region basically achieves the targetrate even under the baseline scenario (4.6 percent

proportion of malnourished children compared toa target rate of 4.4 percent). Noncaloric parametersof the child malnutrition equation were changedvery little for the MDG scenario for this region.

Similarly, China, where declines in child malnu-trition have been very rapid over the last 30 years,is projected to come close to the 2015 target indi-cator for child malnutrition even in the baseline sce-nario. Under the baseline, the share of malnourishedpreschool children is expected to reach 10.9 percentcompared to a target rate of 9.5 percent. Under theMDG scenario projecting rapid increases in agricul-tural and economic growth, and moderate increasesin income growth, the target is reached easily,even under very slow improvements in noncaloricparameters, an increase in the ratio of female-to-male life expectancy from 1.0467 to 1.0650, andan increase in female secondary school participa-tion from 62 percent in 1995 to 74 percent by 2015.These increases result in an under-five malnutritionrate of 7.4 percent under the MDG scenario, or7.0 million children malnourished by 2015.

WANA is the third region where reaching thetarget indicator for malnourished children appearsfeasible, on average. Under the baseline scenario,the incidence of child malnourishment declines to8.4 percent compared to the target rate of 5.5 per-cent. Under the MDG scenario, income growth of6 percent annually for all countries in the region,combined with rapid agricultural growth, limitedimprovements in female secondary schooling andthe ratio of female-to-male life expectancy at birth,

and a high target rate for safe drinking waterresult in a rate of 6.8 percent preschool childrenmalnourished.

However, other developing regions, includ-ing Sub-Saharan Africa, South Asia, and parts ofSoutheast Asia, are less likely to reduce the inci-dence of malnutrition in under fives by half from1990 to 2015, although very rapid growth in bothincome and agriculture can help move these coun-tries and regions almost within reach of the tar-get indicator.

Under business-as-usual conditions, Sub-Saharan Africa is unlikely to meet the MDG targetindicator of halving the prevalence of malnour-ished children by 2015 to 13.4 percent. In the base-line scenario, a reduction from 33 percent in 1995to 28 percent in 2015 leaves the region still withmore than double the target rate for child malnutri-tion levels. In absolute terms, this translates into anincrease in the number of malnourished childrenfrom 32 million to 37 million children. Under theMDG scenario, in addition to very optimistic non-agricultural income growth rates and very rapidcrop and livestock production increases, highlyoptimistic developments for the noncaloric param-eters are simulated, including very rapid changesin the ratio of female-to-male life expectancy atbirth and a doubling or more of the share of femalesecondary school enrollment. Even under thesehighly optimistic assumptions, the 2015 MDG sce-nario results in 17 percent of under fives malnour-ished, or 23 million children. Countries and regions


Figure 3.5 Child Malnutrition, Developing Country Regions, 2015 Baseline, 2015 MDG Scenario, and 2015 Target

52

4

0

5

10

15

20

25

30

35

40

45

LAC SSA WANA S Asia SE Asia China Developing

2015 MDG2015 target

Perc

ent

28

8

41

27

11

24

17

7

30

21

7

1713

6

27

18

10

15

2015 baseline



expected to struggle most with the MDG targetindicator are Nigeria and Northern Sub-SaharanAfrica, where child malnutrition rates in 1995 hadbeen 39 and 40 percent, respectively.

In South Asia, the share of malnourished chil-dren is projected to decline from 51 percent in 1995to 41 percent in 2015 under the baseline, a very sub-stantial decline, translating into a reduction in thenumber of malnourished children by 20 million.This process is mostly driven by India, which isprojected to reduce child malnutrition from 53 per-cent in 1995 to 43 percent in 2015, or an absolutereduction of 19 million children. The MDG scenarioincorporates much faster economic and rapid agri-cultural growth, and very optimistic assumptionsfor the noncaloric parameters, such as a jump in theratio of female-to-male life expectancy at birth from1.0088 for India (the fifth lowest rank among 195reported values in 1995) to 1.0850, and an increasein secondary female schooling from 38 percent in1995 to 95 percent by 2015. These optimistic non-caloric parameters necessitate very large invest-ments in the social sector. However, the outcome ofthe MDG scenario is still 30 percent preschool chil-dren malnourished in 2015, compared to the targetrate of 27 percent.

In Southeast Asia, finally, some countries likeMalaysia and Thailand almost reach their respec-tive baseline target reductions of 12.5 percent and11.0 percent, and therefore surpass the indicatortarget rates under the alternative scenario by 2.5

(Thailand) and 1.4 percentage points (Malaysia),despite low improvements in noncaloric malnu-trition parameters. Other countries, includingIndonesia, Myanmar, the Philippines, Vietnam,Cambodia, and Lao People’s Democratic Republicare less likely to halve the proportion of malnour-ished children in the baseline scenario; Myanmarand Vietnam miss the target by 21 and 15 percent-age points, respectively. Compared to the baseline,under the MDG scenario, rapid increases in agri-cultural and moderate additional economic growth,together with additional investments in socialsectors—as proxied by improvements in the ratioof female-to-male life expectancy at birth, higherproportions of female secondary schooling, andenhanced access to water—lead to declines in thenumber of malnourished children by 2015. Thesedeclines are 6 percent in Indonesia, 3 percent inthe Philippines (where many social indicators arealready at high levels today), 11 percent in Vietnam,13 percent in Myanmar, and 3 percent in otherSoutheast Asia, narrowing the Southeast Asiagap to just 3.3 percentage points from the target(table 3.13).

Figure 3.6 presents the number of people with-out access to safe drinking water in 1995, for thebaseline and MDG scenarios, along the lines ofMDG 7, Target 10, to “halve, by 2015, the propor-tion of people without sustainable access to safedrinking water and basic sanitation.” Rates ofaccess to clean water underlying this figure are

Table 3.13 Southeast Asia, Share of MalnourishedPreschool Children, 1995, and 2015 Baseline,MDG Scenario, and Indicator Target

Countries/ 2015 2015 regions 1995 Baseline MDG Target

Percent

Indonesia 34.0 27.8 21.4 18.4Thailand 18.9 12.1 8.5 11.0Malaysia 20.1 13.8 11.1 12.5Philippines 29.6 23.1 20.0 16.8Vietnam 44.9 37.2 26.1 21.8Myanmar 42.9 37.4 24.5 16.2Other SE Asia 40.0 29.9 27.1 23.5South East Asia 34.1 27.5 20.9 17.6

Note: Other SE Asia includes Brunei, Cambodia, and Lao PDR. Myanmar’schild malnutrition indicators worsened considerably between 1990 and1995, but have since improved again.

based on estimates of likely improvements in accessfor the baseline and required improvements toreach Target 2 for preschool children for the MDGscenario. In 1995, an estimated 1.3 billion people indeveloping countries lacked access to safe drink-ing water according to the IMPACT-WATER ratesand population numbers. Under the baseline sce-nario, by 2015, 0.9 billion people still are estimatedto be without access to safe drinking water. Thedecline in the number of people without access tosafe water during 1995–2015 would be largest forAsia at 34 percent, followed by Sub-Saharan Africaat 28 percent, and slowest for Latin America at7 percent. Under the much higher rates projectedfor the MDG scenario, the number of people with-out safe access would drop to 0.7 billion people,which also still falls short of the target. Under thisscenario, only Latin America and parts of Asiawould reach the target.

Implications for Investment

What are the implications for public investment forthe 2015 baseline and MDG scenarios, based onIMPACT-WATER calculations? The most importantpublic investment drivers in IMPACT-WATER areirrigation, rural roads, education, clean water pro-

vision, and agricultural research. Total irriga-tion investments are calculated by multiplying theestimated increases in irrigated area, adjusted forcropping intensity, during 1995–2015 by the aver-age cost of irrigation per hectare, expressed in 1995real U.S. dollars. Rural road investments are calcu-lated by multiplying the incremental road lengthrequired in 1995–2015 by road investment costsper kilometer. The incremental rural road length iscalculated assuming first, that the density of roadsis proportional to cropland area, and second, thatcrop yield growth also contributes to road expan-sion. Expenditures for public agricultural researchare based on expenditure trends and projectionsand their relative contribution to crop yields. Theestimated incremental investment in education is based on the cumulated annual costs of the addi-tional number of female students required to in-crease the percentage of females with access tosecondary education to the levels projected in thescenarios (Appendix Table 1). It is calculated bythe cumulative annual required increase in enroll-ment multiplied by the annual cost of secondaryschool education per student. Incremental invest-ment costs for clean water, finally, are based on theinvestment required to increase the share of peoplewith access to clean water to the levels projected


Source: IFPRI calculations.

10861

276

854

10051

198

567

52 36

165

485

0

100

200

300

400

500

600

700

800

900

Latin America WANA SSA Asia

Num

ber

ofpe

ople

(mill

ion)

19952015 baseline2015 MDG

Figure 3.6 Number of People Without Access to anImproved Water Source, 1995, andProjected 2015 Baseline and MDG Scenarios

under the alternative scenarios. Per-capita costs ofproviding clean water and estimates of the num-ber of people gaining access to clean water are dis-aggregated by urban and rural areas. Additionaldetails on calculation procedures are presented inRosegrant and others (2001). The cost of improve-ments in the five sectors during 1995–2015 under thebaseline and MDG scenarios are shown in figure 3.7.

As expected, the MDG scenario envisions largeincreases in investment in the key drivers. Totalestimated expenditures during 1995–2015 for thegroup of developing countries are $430 billionunder the baseline scenario and $591 billion underthe MDG scenario. Investments in rural roadsaccount for 28 percent of total investment under thebaseline, followed by agricultural research and irri-gation with 24 percent and 21 percent, respectively,and clean water and education with 15 percent and12 percent, respectively. In the MDG scenario, theshare of investments in education increases to20 percent, with levels more than doubling from$51 billion to $118 billion due to the assumed rapidexpansion in female secondary schooling, particu-larly in parts of Asia and Sub-Saharan Africa. Thereare increases in investments for rural roads, irriga-tion infrastructure, and agricultural research, corre-

sponding with the higher yield increases achievedunder this scenario. The average irrigation costper hectare increases from $4,850 in the baseline to$6,204 in the MDG scenario as more expansionoccurs in areas where projects are more costly, suchas Sub-Saharan Africa. The increase in irrigationinvestments is thus not only due to the expansion ofirrigated area, but also to the increase in the cost ofirrigation under the MDG scenario. Compared tothe baseline 2015 net area, area under the MDG sce-nario is only 4 million hectares larger. As relativelyhigh levels of access to clean drinking water arealready achieved in the baseline scenario, only$15 billion in investments are added for the MDGscenario. The increase in investment in agricul-tural research under the MDG scenario is relativelysmall, $5 billion. Due to the long lags in generationof impact from agricultural research, increases in research expenditures, even beginning now,will have relatively small impacts on crop yieldsby 2015. Increased investments in agriculturalresearch are likely to be essential to meet crop andanimal production needs beyond 2015, but donot have large direct impacts on achieving theMDG scenario. Other investments, such as roadsand irrigation, have significant lags in impact as


Source: IFPRI calculations.

0

100

200

300

400

500

600

700

Baseline MDG

Rural roadsAg researchIrrigationClean waterEducation

430

591

US$

billi

on

Figure 3.7 Cost Estimates for Implementing theBaseline and MDG Scenario, 2015,Developing Countries

well, so implementing the investment portfoliorequired for the MDG scenario will require veryrapid action.

The total increase in investments estimatedbased on IMPACT-WATER calculations is $161 bil-lion in agricultural and supporting sectors dur-ing 1995–2015 under the MDG scenario to bringdeveloping countries and, in particular, two majorregions unlikely to achieve the MDG target indica-tor for malnourished children—South Asia andSub-Saharan Africa—within reach of the preschoolmalnutrition target indicator.

Conclusions

The alternative scenarios have shown that thecombination of agricultural and economic growth,together with larger investments in social sectorsincluding health and education, can almost com-pletely eliminate the gap between the business-as-usual outcomes for 2015 (24 percent ofdeveloping-country preschool children malnour-

ished) and the target indicator (15 percent childrenmalnourished) to reach 17 percent.

However, the outcome varies significantly bycountry and region. While Latin America, the WestAsia and North Africa region, and China will likelycome close to or possibly reach the target indicatorby 2015, even under business-as-usual, chancesthat Sub-Saharan Africa and South Asia will reachtheir respective target rates are much smaller. Infact, on average, Sub-Saharan Africa and SouthAsia will not be able to reduce their shares of mal-nourished children by 2015 even under very favor-able agricultural, economic, and social conditions,but improved conditions in these sectors can bringthe countries in these regions much closer to the2015 target and can facilitate further reductions inmalnutrition later on. In Southeast Asia, finally,the picture is mixed, with some rapidly growingeconomies likely to reach the Millennium targetindicator on child malnutrition, while others, in-cluding Myanmar and Vietnam, will need sub-stantial additional investments in agriculture andsocial sectors to get closer to the target.


49

POLICY ACTIONS AND THEIRIMPACTS ON THE MDGSThe prospects for meeting the MDGs of eradicating extreme hungerand poverty and achieving the health and education goals are essen-tial to and directly affected by the agriculture sector. Farm-sectorsupport policies and border protection worldwide are some of thepolicies at the sectoral level that influence both agricultural and MDGoutcomes. Since most of the world’s poor are rural and depend onagriculture for part or all of their livelihoods, agriculture is a criticalsector in which the global trade system must work to their benefit.

Policies directed at achieving the targets specified under theMDGs should be particularly directed at smallholder agriculture.While small individually, in many cases smallholders account for alarge share of agricultural production. In Sub-Saharan Africa, over90 percent of agricultural output comes from smallholders, whoaccount for nearly three-quarters of the poor. In India, farmers withless than 2 hectares account for 40 percent of total food grain produc-tion (Narayanan and Gulati 2002).

Macroeconomic policies of the developing countries also affectagriculture and rural poverty. A number of developing countrieshave implemented structural adjustment programs aimed at correct-ing fiscal imbalances, largely through reductions in public expendi-tures; redefining the role of the state in economic affairs; privatizingmajor sectors of the national economies; accelerating growth; andthrough trade liberalization (Goldin and Winters 1992). These struc-tural adjustment programs affected agriculture and especially small-holders in a number of ways. Beneficial effects often arise from lowerinflation, reduced government debt accumulation, and falling inter-est rates, as well as from depreciation of overvalued exchange rates.Yet, redressing fiscal imbalances also typically involves reductionsin both agricultural and nonagricultural subsidies (Schiff and Valdez1998) and significant reductions in public spending, with potentiallyadverse impacts on agriculture.

Relative impacts on agriculture depend in part on the targeting ofthe remaining allocation of public spending; if public investments aretargeted to public goods and specifically to services that foster eco-nomic growth, then impact on agricultural output may well be posi-tive. For example, Van Blarcom, Knudsen, and Nash (1993) analyzeda sample of 32 countries in which public spending on agriculture hadbeen cut some time after 1970. As mentioned by Schiff and Valdez

Global trends andEmerging Issues inMeeting the MDGs

A Trade, Macroeconomic, and Policy Perspective

Current trade barriers

and subsidies are high

for agriculture.

Successful WTO

agricultural trade

negotiations that lower

protection and subsidies

could boost progress

toward the MDGs.

4

(1998), the 1993 analysis concluded that muchof the public spending on agriculture in these coun-tries had been directed toward relatively unpro-ductive purposes and, therefore, reductions wereappropriate in those cases.

A review of a range of national poverty reduc-tion strategy papers (PRSPs) showed evidence of alink between poverty and lack of infrastructure ser-vices. In particular, agricultural research, education,and rural infrastructure are the three most effectivepublic spending items in promoting agriculturegrowth as shown for India and China (Fan, Hazell,and Hague 2000; Fan, Hazell, and Thorat 2000; Fan,L. Zhang, and X. Zhang 2002; Fan and Hazell 1999;Zhang and Fan 2000).

Fan and Rao (2004) compiled government expen-ditures by type across 43 developing countries from1980 to 1998 and found that structural adjustmentprograms had different consequences for differentsectors. In Africa, governments reduced expendi-ture shares for agriculture, education, and infra-structure, while Asian governments reduced sharesfor agriculture and health. In Latin America, educa-tion and infrastructure suffered from reduction ingovernment expenditures.

Trade liberalization has been another major com-ponent of structural adjustment programs, and lib-eralization of nonagricultural trade has to someextent improved agricultural incentives throughlower industrial prices and through the deprecia-tion of the real exchange rate. However, the poten-tial benefits of agricultural trade liberalization havenot been generally realized because agriculturehas long been treated as a special case left out-side the multilateral trade-liberalization processin the General Agreement on Trade and Tariffs(GATT). As a result, extensive Organisation forEconomic Co-operation and Development (OECD)subsidies and border protection continue to blockopportunities for those poor people who dependon farming for their livelihoods.

The Uruguay Round of GATT negotiations(1986–94) created the World Trade Organization(WTO) and also produced the first comprehensiveframework of multilateral disciplines on agricul-tural subsidies and trade policies. Yet, the UruguayRound Agreement on Agriculture achieved onlymodest agricultural trade liberalization. A criticalquestion is whether the current Doha DevelopmentRound of WTO trade negotiations (2001) can buildon this framework to deliver further market open-ing and opportunities for trade. The overarching

policy issues, as mentioned in Orden, Torero, andGulati (2004), are whether agriculture will bebrought more fully under liberalized trade rulesand how the outcomes will affect the rural poor.

Additional dimensions of the possible disci-plines on agricultural support and protection poli-cies are their effects on food aid and the impact oftechnical regulations and standards on agricul-tural trade opportunities. The latter are posingchallenges to market participation by smallholdersjust as high-value demands are creating potentialnew income streams.

In this chapter we examine four aspects of thepolicies affecting achievement of the MDGs. Theseinclude trade and domestic support polices foragriculture in developed and developing countries,macroeconomic reform and public sector infrastruc-ture and other investments, the role of the privatesector and public-private partnerships, and theimportance of good governance. We argue that sup-portive government investments and well function-ing private and public market institutions, togetherwith foresight in the design of agricultural policies,are required to take advantage of market opportu-nities to sustain increased agricultural output andraise rural incomes that will help achieve the tar-gets formulated for the MDGs.

TRADE POLICIESPolicies of Developed Countries

Support policies and border protection of wealthyOrganization for Economic Co-operation and Dev-elopment (OECD) countries, valued at hundredsof billions of dollars each year, cause harm to agri-culture in developing countries. Evaluating theoverall effects of the subsidies and protectionamong developed countries, assessing the effects ofthese policies specifically on developing countries,or even more specifically assessing their effects onpoverty in developing countries, are complex chal-lenges. The evaluation must rest on counterfactualsimulation of alternative policy scenarios. A diverseset of policies has to be represented, and models toaccomplish these tasks differ in assumptions aboutcrucial parameters, levels of aggregation, scope ofcommodity and country coverage, and many otherdimensions.

A number of model results were reviewedrecently by Beierle and Diaz-Bonilla (2003) withthe objective of describing what is known and the


Global trends and Emerging Issues in Meeting the MDGs 51

remaining knowledge gaps on whether trade lib-eralization (in the form of reduced protectionand export subsidies and lowered import restric-tions) would benefit smallholder farmers and othersin poverty in developing countries. Several keyfindings from their review and other assessmentsfollow:

• Most models demonstrate negative impactsof current developed country (OECD) tradeprotection policies and positive impacts fromdeveloped country liberalization on develop-ing country welfare, agricultural productionand incomes, and food security.

• Impacts vary by country, commodity, and sec-tor, and for regions within countries.

• OECD market access restrictions harm dev-eloping countries, but effects of productionand income-support subsidies are moreambiguous.

• Developing countries tend to gain more fromliberalization of their own policies than fromreforms by the OECD. Consumers in develop-ing countries benefit widely from developingcountry liberalization reforms.

• Model results differ on the basis of assump-tions such as the scope of commodity cover-age, mobility of resources among alternativecrops and between farm and nonfarm employ-ment, availability of underutilized labor, andstatic versus dynamic analysis.

• Multilateral liberalization reduces the benefitsderived from preferential trade agreements,but these losses are relatively small comparedto the overall gains from the broader reforms.

• Most models have not had sufficient resolutionto analyze the impacts of reforms on small-holders, subsistence farmers, and other poorhouseholds, but there is an emerging literatureattempting to do so (Beierle and Diaz-Bonilla2003; Hertel and Winters 2004; Narayanan andGulati 2002; Tokarick 2002).

With the diversity in modeling approaches, nosingle model has all the desired features that wouldallow an examination of the impacts of trade protec-tion and liberalization on smallholder and sub-sistence farmers and food security. Most of thestudies disaggregate only to the regional level—Sub-Saharan Africa, for example. Many analyseswith world general equilibrium models consideronly an aggregate household per country and pre-sent the results mainly in terms of overall national

welfare. However, there are some studies thatdifferentiate impacts by types of households (forexample, agriculture, self-employed, nonagricul-ture; male or female head), focus on food securityor poverty reduction, or, most recently, incorpo-rate detailed household survey data to evaluatenet and distributional effects of reforming tradepolicies.

A few representative results illustrate the pointsabove. The primary mechanism by which OECDtransmits agricultural protection and subsidiesaround the world is commodity prices in worldmarkets. The analytic studies suggest that OECDagricultural subsidies and protection depress worldprices of basic agricultural food crops in the roughrange of 1–5 percent. OECD protection (mainlytariffs and tariff-rate quotas [TRQs]) depresses theprices of other nonstaple commodities by a largeramount, such as 8 percent for sugar, 22 percent forsheep meat, and 24 percent for milk. Averagedacross all commodities, a common estimate of theextent to which OECD policies depress prices is10 percent, with various studies estimating averageprice changes in the range of 5–20 percent.10

The models show that the increase in worldprices from removal of OECD protection will leadto larger agricultural production in develop-ing countries. Beghin, Roland-Holst, and van derMensbrugghe (2002) estimate that removal ofOECD protection could boost rural value added inlow- and middle-income countries by $60 billion peryear. Tokarick (2002, 2003) arrives at lower num-bers, estimating that OECD market access barriersand subsidy policies cost developing countriesas a whole $8 billion in (overall) welfare annually(0.13 percent of developing country GDP).

In a model that allows for unemployment inrural and urban sectors in developing countries andfor slight positive effects of technical change corre-lated with higher levels of trade openness, Diao,Diaz-Bonilla, and Robinson (2003) estimate thatOECD subsidies and border protection reduce agri-cultural exports from the developing world by$37.2 billion (25.3 percent) annually. Agriculturalvalue added among developing countries is re-duced by $23.0 billion annually, while nationalwelfare of developing countries is repressed by$9.4 billion. For specific countries and specific com-modities, the effects can be critical, as in the case ofcotton for the rural poor in a number of Africancountries. Minot and Daniels (2003), using house-hold survey data, find that a drop of world cotton

prices by 20 percent, as might be due to developed-country subsidies, raises poverty in cotton export-dependent Benin by 4 percentage points (anincrease of 10 percent of the population under thepoverty line) through direct and indirect effects onrural incomes.

Beyond the effects of tariff and farm-supportpolicies captured in models as described, the fastestgrowing world agricultural markets are for fruitsand vegetables, livestock products, and other high-value commodities. For these products, regula-tions related to safety and quality play a large role indetermining trade opportunities. The WTO embod-ies agreements to discipline agricultural and foodsafety and quality regulatory decisions that areprimarily sovereign prerogatives. These WTO disci-plines call broadly for countries to achieve legiti-mate regulatory goals in the least trade-distortingmanner. Effectiveness of these disciplines is animportant aspect of a rules-based agricultural tradesystem (Josling, Roberts, and Orden 2004).

Developing countries have a lot at stake in thearea of food regulation. Production of high-valueproducts is potentially a source of higher incomesamong the rural poor. But stringent developed-country regulatory measures to address health,safety, and quality goals can close off market oppor-tunities. There is a trend in quality regulationtoward the required use of certain production meth-ods or required labeling of production and process-ing attributes. These process-focused measuresoften demand complex conformity assessment withhigh compliance costs. Innovations are needed toensure efficient implementation of such measures.There is a great need to build up the capacity ofdeveloping countries to produce up to the exact-ing standards of importing markets. And there is achallenge to the MDG: agricultural and food regu-lations that are well-intentioned in some dimen-sion can have the undesirable effect of reducingincome-earning opportunities or blocking technol-ogy adoption that would benefit the poor.

Policies of Developing Countries

There is growing agricultural trade among devel-oping countries, but these countries also retainsubstantial trade barriers on agricultural products.Developing-country governments (and civil societyorganizations) that are largely united in seekinglower agricultural subsidies and protection in thedeveloped countries have been divided concerning

what to do about the agricultural trade barriers indeveloping countries. Developing countries arenot homogenous with regards to agriculture liber-alization. They have different trade specializations(different export products facing varied degreesof protection or support in developed countries);some are net food importers, and others are facedwith different trade barriers. Those countries withstrong agricultural export potential have called formore open markets, but those fearful of negativeeffects on their poor farmers have been reluctant toendorse such moves. Many development advocatesare adamant that developing countries be grantedroom to retain agricultural trade barriers.

What is at stake in the reduction of agriculturalprotection among developing countries is some-what different than the stakes from developed-country reforms. When developing countries joinin agricultural trade liberalization, reducing highlevels of trade restrictions, relaxing quantitativerestrictions on imports, and lowering import tariffs,they can achieve overall welfare gains of $19.9 bil-lion annually, according to Diao, Diaz-Bonilla, andRobinson (2003). This is twice the gains in nationalwelfare compared to reforms in the developedcountries only. The additional overall gain is pri-marily due to consumers facing lower internalprices as their countries’ own trade barriers arereduced. Developing-country trade policy reformsadd an additional $14.9 billion annually to aggre-gated agricultural exports of developing countries,but do not increase their aggregated agriculturalproduction value added in the Diao, Diaz-Bonilla,and Robinson model.

These results suggest that trade policy reformsamong developing countries boosts their over-all welfare but also creates distributional impactsamong those developing countries that are best ableto gain from trade openness versus those less ableto do so. These reforms also create distributionaleffects between food consumers and producerswithin (poor) countries. There is a need to under-stand these distributional effects better. However,the overall welfare benefits are important as well;they extend to poor food consumers, includingpoor farmers and landless rural workers who usenonfarm income to buy food.

Three leading empirical studies that focus onAfrica’s growth performance—Easterly and Levine(1997), Rodrik (1998), and Sachs and Warner(1997)—conclude that the region performed poorlyafter the reforms of the 1980s, even though the



reasons provided for the poor performance vary.Easterly and Levine emphasize the role of ethnicfragmentation and poor quality of institutions,while Sachs and Warner stress poor trade policiesand geography. Rodrik reports that, contrary toSachs and Warner, the fundamentals for long-termgrowth are human resources, physical infrastruc-ture, macroeconomic stability, and the rule of law.

In an overall context, a successful conclusion foragriculture in the WTO Doha Development Roundtrade negotiations can be understood to make animportant contribution to achieving the MDGs byestablishing sustainable positive incentives for agri-cultural production among developing countries. AJuly 30, 2004 WTO framework agreement to guidefurther negotiations came at the last possible hourto avoid a collapse of the Doha round. For agri-culture, it narrowed the negotiating field some-what, but still leaves wide latitude about howmuch trade will be opened under an agreement’sfinal terms. Because of the way WTO disciplineson tariffs and subsidies are defined and adminis-tered, agreed reductions in protection and domes-tic farm support would have to be quite big tohave much effect on trade (Anderson, Martin, andvan der Mensbrugghe 2004).

In summary, an increase in the share of nationalincome that is exported, or lower importing prices,does not in itself generate growth in per capitaincome and will not on its own set an economy ona sustained growth path unless it is accompaniedby other structural reforms.11 However, excessivelevels of export taxation and import restrictionscan contribute to the relative decline in growth ofsome countries. In addition, to achieve some of thepositive gains, developing countries that will ben-efit from more open markets abroad need to par-ticipate in multilateral agricultural trade policyreforms. Benefits for poor farmers in countries lessable to compete as trade barriers come down willcome not from multilateral trade policy reformitself, but from complementary investments andpolicy improvements.

Policies in Development Assistance: The Case of Food Aid

Food aid is another component of internationaltransactions that directly and indirectly affects ruralpoverty in a globalized agricultural economy andtherefore could have a significant impact in achiev-ing the MDG targets.

Modern food aid emerged after World War II,particularly with the United States P.L. 480, theAgricultural Trade Development and AssistanceAct, of 1954. This law asserted multiple goals forfood aid. Food aid would combat world hungerand malnutrition and promote agricultural develop-ment, but it would also expand trade and developexport markets for U.S. agricultural commodities.Food aid from the United States peaked in the 1960s,but the country has remained the largest singledonor of food aid, accounting for about 55 percentof the total during the 1990s. Food aid now accountsfor less than 5 percent of global trade in agricul-tural and food products. Yet many controversiessurround the use of food aid, either for emergencycrisis relief or as an instrument of humanitarianand development policy. Given relatively fixed orslowly changing budgets, there is a built-in cycle offood aid: the volume of food aid available will belowest when commodity prices are highest (andneed is, in that sense, greatest). Moreover, there ismuch evidence that the provision of food aid issubject to political pressures related to supportingworld commodity prices and other objectives.

Food aid falls under emergency aid, project aid,or program aid. Emergency aid, which has beenincreasing its share in total food aid recently, is aresponse to natural disasters or conflicts that leavevulnerable populations at risk of starvation orsevere malnutrition. Project aid is associated withthe development of specific food security or devel-opment projects, such as school feeding programs.Program aid is the most general use of food as aform of foreign assistance, essentially providing themonetized value of the food as a resource for use bya developing-country government, although some-times with conditionality requirements about howthis aid is utilized.

Food aid can be procured in the donor country,local markets of the recipient country, or from third-country sources. It can be provided through bilat-eral or multilateral channels, and these channels canencompass governments, multilateral agencies suchas the United Nations’s World Food Programme(WFP), and nongovernmental organizations. TheU.S. P.L. 480 requires the use of commodities pro-duced in the United States; in 2002, for example,nearly 90 percent of the total 9.6 million tons offood aid delivered worldwide was procured in thedonor countries. Contributions in financial terms bydonors, rather than in the form of food commodi-ties, allows greater flexibility in providing food

aid, which can be sourced in the donor country orelsewhere using the aid financing. Local or third-country (triangularization) purchases accounted fora higher percentage of the food aid delivered bythe European Commission’s multilateral food aidprogram (70 percent), when food aid is provided foremergency purchases (67 percent), and of fooddeliveries through the WFP (which accounted fornearly 40 percent of total food aid in 2002, nearly60 percent of which was procured in recipient orthird countries) (Hoddinott, Cohen, and Bos 2003).

Once food aid is provided to recipient agencies,it can either be delivered directly as food to targetedpopulations or “monetized” through sale in recipi-ent country markets. When monetized, the cashvalue of the food aid becomes a resource that theagency can use to support various activities. Evenin the case of food aid for emergency purposes,some of the food can be sold to cover nonfood costsof aid delivery. In the cases of project and programaid, a great deal of monetization occurs. There istherefore concern that food aid can depress incen-tives for local food production. But well directedfood aid, or aid programs financed by monetizationof food aid, can also provide essential emergencyrelief and reduce vulnerability of the poor to short-term shocks that undermine their long-term humanand physical assets (Barrett and Maxwell 2004;Hoddinott, Cohen, and Bos 2003).

Several international institutions provide guide-lines for food aid.12 The current rules exempt “bonafide food aid” from restrictions on export sub-sidies, and the nonbinding 1994 WTO MarrakeshMinisterial Decision recommends increased foodaid as a means to help developing countries. Butas direct export subsidies allowed under the Agree-ment on Agriculture are subject to increasingly dis-ciplined negotiation (and are even being phasedout), indirect forms of subsidization, including someuses of food aid, are also under scrutiny. The July2004 framework agreement calls for elimination “bya credible end date” of export subsidies and “paral-lel” elimination of “all export measures with equiv-alent effect.” For food aid, the agreement calls forelimination of “provision of food aid that is not inconformity with operationally effective disciplinesto be agreed. The objective of such disciplines willbe to prevent commercial displacement.”

Food aid remains the subject of ongoing contro-versies in the context of the issues described above.It is widely recognized that provision of aid in theform of food is not the optimal form for develop-

ment assistance, but likewise that donors wouldprobably not provide equivalent cash developmentassistance in place of food if existing food-aid pro-grams were terminated (Hoddinott, Cohen, andBos 2003). Thus, attention has focused on how itseffectiveness can be maximized and its potentialharms mitigated (Barrett and Maxwell 2004; vonBraun 2003).

Macroeconomic Reforms and NationalGovernment Infrastructure Investments

For the last two decades developing countrieshave experienced several changes in their macro-economic policies. Macroeconomic policies changedfrom development strategies aimed at the domes-tic market, with strong public-sector intervention,to more outward-looking strategies with open-ness to trade, deregulation of markets, more fis-cal control, and the private sector as the main agentfor development.

Although it was expected that these new policieswould lead to an increase in growth, the results arenot clear and vary among regions. Results derivedfrom cross-country comparisons may be indica-tive but are not very conclusive. Such comparisonsinvolve differences in the experience of each of theindividual countries, not only in terms of the effectsof the policy reform but also of the initial conditionsat the initiation of reforms and the timing and con-sistency of them. However, many experts agree thatthere is a need to maintain macroeconomic bal-ances, to further trade and financial openness, torestrict activities of the state, and to deregulatemarkets (Devarajan and others 2000; Edwards1995; Stiglitz 2000).

Similarly, agricultural policies changed signif-icantly in the 1990s. An example is shown in table4.1 for Latin America and the Caribbean. Seriousconcerns have been raised regarding the effects ofpolicies on the agriculture sector and on the morevulnerable parts of the population. Demand con-traction, the abolition or reduced role of parastatals,significant reductions in public expenditures, andthe withdrawal of state interventions—combinedwith rigidities in resource mobility, slow responseof the private sector, and lack of appropriate infra-structure and institutions—have resulted in anincrease of unemployment and therefore an increasein inequality between and within rural and urbanareas. Reductions in public investment outlays werenot (fully) compensated by private investments,



and this of course affected expenditures in infra-structure, health and education services, and socialprograms.

Infrastructure is of particular concern as one ofthe key inputs entering into the “production func-tion” of the MDGs and the achievements of manyof the MDG targets, from poverty reduction to envi-ronmental sustainability targets.13 For instance, inSub-Saharan Africa, where less than half of the pop-ulation has access to safe drinking water (Fishbein2001) (see also table 3.10), child mortality maydepend on the availability of clean water (Galiani,Gertler, and Schargrodsky 2005), and attainment ofuniversal primary education for girls may also cru-cially depend on access to piped water (Leipzigerand others 2003).

Importance of Rural Infrastructure

There is an increasing consensus that providing ade-quate infrastructure is an important step in theprocess of poverty alleviation and in providing amore equitable set of opportunities for all citizens.Several authors have studied the aggregate-levellinks between poverty and rural capital-intensiveinfrastructure, including Jimenez (1995), Lipton andRavallion (1995), and Van de Walle (1996). In partic-ular, the literature on specific infrastructure compo-nents, such as the role of rural roads, telephones,or access to electricity on poverty alleviation isvery broad, including Binswanger, Khandker, andRosenzweig (1993), Howe and Richards (1984),Jacoby (1998), and Lebo and Schelling (2001).

Table 4.1 Latin America and the Caribbean: Main Features of Changes in Agricultural Policies

Policies in the 1950s to 1980s Present Policies

Political, institutional, and financial decentralization aimedat serving local projects

Demand-oriented approach, with cofinancing of investmentsin infrastructure and services

Demand-oriented approach, with cofinancing of researchboth in public research institutes and universities

Paid private technical assistance in response to demandfrom producers, subject to presentation of a project forcompetition when subsidies are involved

Elimination of specific lines of credit for the agriculturalsector, of development banks, and of the obligation ofprivate banks to loan a certain proportions of their portfolio to the sector

Deregulation of the labor market, although in the case ofagriculture the rurales had always been less strict andmore difficult to enforce

Freedom of prices

In principle, equilibrium in exchange ratesReduction of tariff rates and application of a uniform rate,

albeit with some exceptions, especially for agriculturalproducts

Elimination of taxes and other hindrances to exportsReplacement of agrarian programs with purchase through

the land markets, aided by loans or subsidies for smallproducers; promotion of formal proof of ownershipthrough the issue of official property titles

Strong state centralism

State planning and financing of “public” goods and servicesConcern for the food security of the country, with research

therefore focused on increasing production and yields of staple foods (rice, wheat, maize, potatoes, beans,lentils, etc.).

Free technical assistance, with programs prepared by statebodies with little or no user participation

Subsidized credit through special lines of credit for the sector

Labor reforms leading to the monetization of the rural econ-omy and the gradual disappearance of sharecropping-type systems, with their labor force being replaced bypermanent or, most often, temporary wage laborers

Price controls for staple foodstuffs in order to check infla-tion and keep minimum wages low so as to encourageindustrialization

Exchange rates favorable to importsTariff rates that were generally high but covered a wide

range of rates

Taxes on agricultural exportsAgrarian reform programs aimed at reducing the inefficiency

of highly heterogeneous production systems (with abundantland and capital for a few agents and abundant labor in the case of the rest) and above all at reducing the possibility of social disorder in rural areas

Source: Dirven (1999).

Recently Renkow, Hallstrom, and Karanja (2002)estimated the fixed transaction costs (those notdependent on commercialized volume) that impedeaccess to product markets for subsistence farmersin Kenya. The authors estimate that high transactioncosts are equivalent to a value-added tax of approx-imately 15 percent, illustrating the opportunities toraise producer welfare with effective infrastructureinvestments. Smith and others (2001) show that,for the case of Uganda, the rehabilitation of roadsincreases labor opportunities in the service sector.

Moreover, based on an infrastructure index thatincludes road, rail, and telecommunications density,Limão and Venables (1999) found that infrastructureis a significant and quantitatively important deter-minant of bilateral trade flows. Improving destina-tion infrastructure by 1 standard deviation reducestransport costs by an amount equivalent to a reduc-tion of 6,500 kilometers at sea or 1,000 kilometers ofoverland travel. According to their findings, most ofAfrica’s poor trade performance can be accountedfor by poor infrastructure. Moreover, lack of ade-quate infrastructure in much of Sub-Saharan Africaimpedes more productive agriculture. In NormanBorlaug’s words, “an effective system to delivermodern inputs—seeds, fertilizers, crop-protectionchemicals—and market output must be established.If this is done, subsistence farmers, who constitutemore than 70 percent of the population in mostcountries there [Sub-Saharan Africa], can have achance to feed their people.” (Borlaug 1999)

In addition, and as argued by Leipziger andothers (2003), achieving the health and educationMDGs will require more than health and educa-tion interventions; in particular, infrastructure ser-vices have a crucial role to play. In health, thereis clear evidence that better access to basic hard-infrastructure services has an import role to play inimproving child health outcomes and therefore incomplying with the three health-related MDGs.For example, Jalan and Ravallion (2001) showedthe importance of piped water to reduce diarrhea inyoung Indian children; and Galiani, Gertler, andSchargrodsky (2005) found that child mortality fell5–9 percent in areas that privatized their water ser-vices. Brenneman and Kerf (2002) showed that elec-tricity allows for more hours of studying and roadaccess promotes easier establishment of schoolsand higher attendance.

Sub-Saharan Africa seems to particularly lagbehind in infrastructure investments, as is shownin figure 4.1 for paved roads, telephone lines,

and electricity production. Increases in populationmoved each group to the left by reducing its ratioof land to labor, and each group of countries alsomoved upwards because of the expansion to accessin each of the infrastructures. It is obvious that thesize of movement for Sub-Saharan Africa was sig-nificantly smaller than those for other countrygroups. The major reason for this significant lag hasnormally been attributed to geography (Diseases,internal distance, and low populated areas were amuch bigger obstacle.) and to the poor initial con-dition of infrastructure in Africa.

Unlike Asia and Latin America, Sub-SaharanAfrica inherited a highly dispersed and unevenlydistributed infrastructure from its colonial past.There was little improvement of infrastructure, ifany, during the colonial era, and “in some impor-tant respects, it can even be said that colonial policyreinforced the handicaps of SSA [Sub-SaharanAfrica]” (Platteau 1996, p. 200). The limited infra-structure that was built during that era was drivenby the objective of connecting natural resources toexport markets. For instance, “two-thirds of theAfrican railways built in the colonial period con-nected mines to a coastal harbor” (Platteau 1996,p. 200). The rest of the continent was virtuallyignored and “only the Union of South Africa withmass immigration of Europeans had more than sixmeters of railways per square kilometer in 1970,and six countries had no railways at all” (Boserup1981, p. 148). The skewed distribution of infrastruc-ture was perpetuated even after independence.

If structural impediments predominate in agri-culture as in the case of Africa, it will be difficult toachieve sustained growth in production by priceincentives alone unless countries develop seriousstrategies to reduce these impediments. In a recentstudy, Fay and Yepes (2003) predicted the demandfor roads, railroads, telecommunications, electricity,water, and sanitation. According to their estimates,in order to meet the predicted demand, the coun-tries of Sub-Saharan Africa will need to investaround $25.9 billion annually between 2005 and2010. Of this sum, $12.6 billion will be needed formaintenance of the existing infrastructure and therest to build new infrastructure. This would requirean annual investment of more than 5.5 percent ofGDP. Wood (2002), taking into account the low pop-ulation density of Africa, predicted that Africa willneed to invest at least twice as much of its GDP ininfrastructure as does low-income Asia, as wellas meet higher recurrent charges for operation and



Paved roads 1990–2000

0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

Roa

d(k

m/1

0sq

kmla

ndar

ea)

0

LAC

OEC

SAS

SSA

2 4 6 8 10 12 14

Telephone lines 1980–2000

0

100

200

300

400

500

600

700

Mai

nte

leph

one

lines

(per

1,00

0po

pula

tion)

0

LAC

OEC

SAS SSA

2 4 6 8 10 12 14 16 18

Electricity production 1980–2000

0

2,000

4,000

6,000

8,000

10,000

12,000

Elec

tric

itypr

oduc

tion

(kW

hpe

rin

habi

tant

)

0

LAC

OEC

SAS SSA

2 4 6 8 10 12 14 16 18

Land area (sq km/100 labor force)

Land area (per 100 labor force)

Land area (sq km/100 labor force)

Note: LAC = Latin America & Caribbean; OECD; SA = South Asia; SSA = Sub-Saharan Africa. Source: Torero and Chowdhury (2004).

Figure 4.1 Changes in Paved Roads, Telephone Lines, and Electricity Production Over Time, Selected Regions

maintenance. Therefore, strategies for infrastruc-ture development are needed to be able to movetoward the required trends to reach the MDGs.

On Public Investment and Infrastructure

Even among countries that have advanced the mostin market-oriented reforms, as mentioned earlier inthis chapter, there is still a significant infrastructureaccess gap, and reforms alone are not enough toprovide complete access in remote poor rural areas.Some sort of public intervention is needed to closethis gap and therefore to achieve the MDGs. Wherethe government believes that service should be pro-vided beyond what a well-functioning marketwill offer, subsidies may be justified to promoteadditional investment to achieve these govern-ment goals. But the government should also seekto improve the functioning of the market so thatsubsidies can provide a maximum benefit whenand where they are needed. In particular, a failureto address the impediments to efficient working ofthe market in rural areas through regulatoryreforms will reduce the availability and effective-ness of resources to address the real access gap inthese areas.

Moreover, the potential best practice to reducethe access gap largely depends on the institutionalframework existing in each country. Countries withsound regulatory institutions and legal frameworkscan adopt some solutions that will be out of reachfor countries with weak institutions. However,there could be institutional solutions adequate toreduce the access gap while simultaneously devel-oping the legal, institutional, and regulatory frame-work needed to advance different strategies.

There are two other issues that concern publicinvestment on infrastructure. The first issue is thelack of coordination in public investments at thecountry, regional, and donor levels. Where the link-ages and complementarities of infrastructure invest-ment have not been realized, it is common to findfragmented approaches, lacking sufficient attentionfor substantive policies and development issues. Infact, in many cases, access to infrastructure hasnot been linked to poverty alleviation strategiesor to the general development goals of countries.Therefore, it is necessary to take an integratedapproach, even if the actual design may vary fromcountry to country.

Second, investment in infrastructure in theseareas is done from top to bottom rather than being

demand driven. At present, the estimation of ruralinfrastructure investment is generally based on theneeds assessed for each sector at the national level,with little or no assessment of demand and coordi-nation at the local level where the service ultimatelywill be provided. More often than not, such invest-ment assessments do not reflect the preferences ofusers of services and the contingencies of ser-vices. For instance, demand for secondary school-ing may be contingent on access roads, and failingto coordinate these two may result in a mismatchbetween availability of a service and its actual use.Furthermore, it is important for communities to beinformed about the technology they want to use,the service level they require, and especially to havea clear understanding of long-term costs and main-tenance implications, so that communities canchoose what is most appropriate for them undertheir budget constraints. There is evidence that ifprovided with appropriate information and tech-nical support, communities can make informedchoices about service options as well as clearlyidentify their willingness to pay, thereby assumingownership and responsibility for the infrastruc-ture (Brenneman and Kerf 2002; Estache, Foster,and Wodon 2001, 2002; Torero, Chowdhury, andGaldo 2003).

Finally, public sector intervention alone is notenough; forging private-public partnerships seemsto be the most efficient way of closing the accessgap in all services covered as we discuss later inthis chapter. Public intervention alone is usuallynot cost effective, and isolated private initiativesalso fail to deliver all services. Despite the rise inprivate-sector involvement in infrastructure pro-vision, the overall investment levels, particularlyconcerning small subsistence farmers, are far fromadequate. The need for the public sector to play afacilitating role has not been addressed. Since thereis little evidence that rural infrastructure is com-mercially viable on a stand-alone basis, the role ofthe public sector needs to be reinvented.

THE ROLE OF THE PRIVATESECTOR IN SUPPORTINGAGRICULTURE TO MEET THE MDGSPublic intervention alone is not sufficient to de-liver the services and investments required toachieve the MDGs as outlined in the MDG scenario



described in chapter 3. Thus, the role of the privatesector for providing investments and deliveringservices is receiving increased attention in the agri-culture as well as the rural nonfarm sector. Ideally,the public and private sectors complement eachother, with the government providing an appropri-ate enabling environment for private initiativesto develop. Good governance principles for bothpublic- and private-service provision are crucial toachieve both effectiveness and efficiency in resourceallocation. The following section describes how theprivate sector can support agriculture’s role inachieving the MDGs.

To alleviate rural poverty in developing coun-tries, the private sector can contribute to economicgrowth through creating jobs both on and off farm.The private sector can also empower the poor byproviding a broad range of products and servicesat lower prices (UNDP 2004b). Large local compa-nies and multinational corporations can channelprivate initiatives into development efforts. Thiscould be done, for example, through the expansionof public-private partnerships in the fields of edu-cation, such as with vocational schools; in the retailsector, through contractual arrangements drawingin small-scale farmers; and, generally, through thediffusion of knowledge and information to small-and medium-scale enterprises working on forwardand backward linkages of agriculture. The privatesector is also important in the finance sector, whereit can provide resources and relationships that canhelp establish these enterprises.

Rural development that helps achieve the MDGscreates the demand for timely and efficient creditavailability to support private investments inagriculture and related services. As developmentproceeds, rural areas will generate large capitalsurpluses that need to be captured and managedefficiently by the financial sector. Appropriatedevelopment of the rural financial sector can alsoreduce household vulnerability through promotionof savings mobilization. Efficient rural financialmarkets that serve the full range of financial needsof farmers and nonfarm enterprises are usuallymore important than targeted or subsidized creditprograms. Such programs should be demanddriven, not donor or supply-side oriented. Thesefinancial programs should provide a broad rangeof financial services, not just savings mobilization;emphasize viable and sustainable financial inter-mediation based on market interest rates; and pro-vide credit for all purposes. The private sector has a

particularly important role to play in the provisionof rural financial services, while the public sectorshould provide an appropriate enabling environ-ment for the private sector to operate efficiently. Inparticular, the public sector should facilitate accessto broader financing options by continuing devel-opment of domestic financial markets coupled withskill building for regulators and private financialinstitutions.

The Bank Rakyat Indonesia (BRI) system hasshown that microfinance services can be providedto low-income and poor households at fully com-mercial terms by a large national commercial bank,thus combining maximum outreach and sustain-ability. The BRI Unit Desa (Village Bank) system,which has developed into the largest internation-ally acknowledged successful microfinance oper-ation in the developing world, reaches almost3 million borrowers and nearly 28 million smallsavers. As BRI Unit Desa practice shows, savingsand credit interest rates are not among the mostimportant considerations for micro-clients. Themassive amount of rural savings mobilized by theunit desas demonstrates the demand for attractivesavings products and the capacity of rural people tosave. The BRI experience further demonstrates thatpolicies and institutions can be designed to achievehigh levels of outreach, serve the very poor, andattain financial and institutional sustainability usingan individual lending technology. The unit desasreach poor clients by using an individual loan tech-nology requiring collateral, but group loans maybe useful for reaching even poorer clients. Amongthe key factors for BRI Unit Desa success areuser-friendly products and services priced for insti-tutional viability; close and regular relationshipswith micro-customers; convenient bank locations,simple loan procedures, quick processes, flexibleterms; offering savers a combination of security,convenience, liquidity, confidentiality, service, andreturn; simple management information systemsand transparent accounting and reporting systems;and close loan monitoring and incentives for repay-ments (SEARCA/IFPRI/CRESCENT 2004; Meyerand Nagarajan 1999).

In Africa the private sector, together with public-sector support, is crucial for expediting the rate ofmarket development; removing or reducing barriersto market access, both by providing special sup-port for markets that are slow to develop and easingmarket participation by poorer producers; andestablishing a more equitable set of market relations

between (small-scale) producers and market inter-mediaries (NEPAD 2002). The following activitiesare suggested to further market development.

• The government can develop the skills andorganization for smallholders, for example,through the promotion of producer groupsor associations.

• The government can help the private sector todevelop and broaden its outreach. A greaterprivate-sector presence will then provide morecompetitive and efficient services to small-holders, particularly for input supplies, pro-duce, marketing, and agroprocessing. Thisapproach is also vital to support the develop-ment of micro-enterprises.

• Both the government and the private sectorcan finance the provision of essential con-necting infrastructure, such as market accessroads, communications, and price and mar-ket information.

• Most important, the private sector can assistsmallholders in promoting a dialogue betweenthe main stakeholders to generate the policy,institutional, and legal context required forenhanced market linkages.

The Rural Nonfarm Private Sector

The rural nonfarm sector is an important enginein economic development. During successful ruraleconomic growth, the emergence and rapid expan-sion of the (mainly private-sector) nonfarm econ-omy in rural areas and the towns they serve is amajor source of growth in incomes and employ-ment. From a relatively minor sector, often largelypart-time and subsistence-oriented in the earlystages of development, the rural nonfarm economycan develop to become a major motor of economicgrowth in its own right, not only for the country-side but for the economy as a whole. Its growthalso has important implications for the welfare ofwomen and poor households, sometimes helpingto offset inequities that can arise within the agri-cultural sector. In Asia, for example, the rural non-farm economy accounts for 20–50 percent of totalrural employment and 30–60 percent of total ruralincome. The rural nonfarm economy is especiallyimportant to the rural poor. Landless and near-landless households everywhere depend on non-farm earnings; those with less than 0.5 hectareearn as much as 90 percent of their income from

nonfarm sources (Rosegrant and Hazell 2000).Nonfarm shares of income are strongly and neg-atively related to farm size. Low-investment manu-facturing and services—including weaving, pottery,gathering, food preparation and processing, domes-tic and personal services, and unskilled nonfarmwage labor—typically account for a greater share ofincome for the rural poor than for wealthier ruralresidents (Hazell and Haggblade 1993). The reverseis true of transport, commerce, and such manufac-turing activities as milling and metal fabrication,which require sizable investments.

As noted, effective financial institutions arerequired in order to promote trade, commerce,and manufacturing in the rural nonfarm economy.Microfinance programs to help women and poorpeople develop nonfarm enterprises may contributeto poverty alleviation. Microfinance also needs tobe accompanied by appropriate training programsto give women and poor people the skills they needto compete in the market.

Rural people need adequate training if they areto obtain relevant technical, entrepreneurial, andmanagement skills. The rural nonfarm economyprovides much of its own training through appren-ticeship schemes and on-the-job learning, but inan increasingly technical and communications-oriented world, specialized training schemes (incomputing and accounting, for example) areneeded, including programs for women, whodominate many service and trading activities.

In much of the developing world, apart fromsome manufacturing activities, the rural nonfarmsector was largely ignored by policymakers untilrecently. Furthermore, because it depended heavilyon agriculture either directly or indirectly for muchof its demand, the rural nonfarm sector also suf-fered as a result of macroeconomic policies that dis-criminated against the agricultural sector. Recentmacroeconomic policy reforms that have benefitedthe agricultural sector should, therefore, have led topositive growth-multiplier benefits for the ruralnonfarm economy. The policy reforms have alsofavored tradable-goods production in general, andthis should have been directly beneficial to muchrural industry. However, these benefits for the ruralnonfarm sector are often limited by a continuingbias toward capital-intensive industry at the ex-pense of trade and services. In many cases, smallfirms have effectively been placed at a competitivedisadvantage to their larger rivals (for example, theydo not receive the same subsidies and tax benefits),



and this has encouraged a more capital-intensivepattern of development than is optimal.

In spite of the constraints on the role of the pri-vate sector, development of this sector is essential toprogress in meeting the MDGs. We now examinethe role of the public sector in enhancing the per-formance of the private sector through provision ofappropriate regulation and oversight, setting fairground rules for the private sector, and through theprovision of public goods.

Effective Legal, Regulatory, and Institutional Environments

Effective legal, regulatory, (for example, to secureproperty rights and enforce contracts) and institu-tional (for example, financial) environments arerequired to promote agricultural trade, commerce,and manufacturing. At the farm level, effective reg-ulatory systems need to be provided, particularlyin the areas of standards and quality control, toensure future competitiveness of farmers both inthe emerging supermarkets nationally and inter-nationally in the global markets, which demandincreasing labeling, standards, and quality control.

At the firm level, in order to ensure the long-term viability of the private sector in development,the creation of an enabling environment is requiredthat includes strong accounting and disclosurestandards, bankruptcy and foreclosure processes,and improved taxation and accounting laws.Garnaut (1998) summarizes the objectives of pru-dential regulation: accuracy, honesty, and trans-parency in financial reporting; the avoidance ofrelated-party and other noncommercial transac-tions; and the maintenance of relatively high capital-to-asset ratios—higher than the norm for developedcountries. Prudent regulation should focus on sim-ple, enforceable targets, rather than attempting to regulate a wide range of indicators (Fane 1998).Corporate governance also needs long-term reform,including, for example, monitoring of enterprisesby commercial banks. Enhanced disclosure andaccounting practices and strengthened enforcementof corporate governance regulations, especiallyas they relate to capital markets, are also essen-tial components of corporate governance reform.Institutions need to be strengthened so that analysisof corporate financing and monitoring of firm per-formance and behavior are comprehensively cov-ered by a combination of private, semipublic, andpublic organizations (World Bank 1998).

Private-Public Partnerships and InvestmentSynergies: The Case of Water and Agricultural Research

How should investments in agriculture and relatedsectors be supported to achieve the MDGs? Public-private partnerships are an important way toincrease financial, human, and social capital inrural areas. Partnerships include publicly providedtraining for small- and medium-scale enterprises,partnerships in education, agricultural research,the provision of information and communicationtechnologies, the expansion of rural infrastruc-ture including roads, and the development of ruralindustrial clusters. Because public funds areincreasingly scarce, private-public partnerships areone important means to enhance investments inrural areas.

Private-public partnerships are considered par-ticularly important to close the access gap in infra-structure and social services. Public interventionalone is usually not cost-effective, and isolated pri-vate initiatives also tend to fail to deliver all ser-vices. Despite the rise in private-sector involvementin infrastructure provision, overall investment lev-els are particularly far from adequate. Accordingto the Commission on Private Sector and Dev-elopment (UNDP 2004b), private-sector manage-ment of small-scale energy production and watersupply projects can be particularly effective. Theseprojects allow management by smaller domesticcompanies whose final product is delivered to therural customer or informal sector. Decentralizedpower production through distributed energy ofvarious kinds can also be contracted to the privatesector through agreements with the public-sectorgrid. Examples are solar power and small hydro-power plants.

In particular, water resources have been recog-nized as under serious threat with growing national,regional, and seasonal scarcities (Rosegrant, Cai,and Cline 2002). Over 1 billion people lack accessto improved water services. Worse, another 2.4 bil-lion people still live without improved water sani-tation. The need for investment, particularly fromthe private sector, is obvious. To facilitate invest-ments, the World Bank, through the MultilateralInvestment Guarantee Agency (MIGA), is engagedin public-private arrangements for the provision ofwater and wastewater. However, the lack of suffi-cient public funds and difficulties in attracting private money in the sector makes achieving the

MDGs more difficult. There are a number of modelsfor public-private investment partnerships, rang-ing from management contract to build-operate-transfer contracts (figure 4.2); their applicabilityand success depend on the local conditions anddevelopment status.

Both public and private companies have beenable to effectively provide water service when theyare operated properly and certain guidelines aretaken into consideration. Gleick and others (2002)spell out several principles that need to be consid-ered for private water provision. Many of these arekey elements of effective public water systems aswell. These principles include ensuring that allbasic human and ecosystem needs for water aremet, with a basic amount of water provided at asubsidized rate for those too poor to afford topay; provision of water at reasonable rates, withincreases in price linked to improved service whenpossible; subsidies that are economically and so-cially sound; and new water projects that arepermitted only when shown to be less costly thanefficiency improvements. It is also crucial thatthere be significant government oversight, withgovernment regulation of water quality, clear con-tracts that spell out the role of all parties, clear rulesfor dispute resolution, and clarity and stakeholderinvolvement during the negotiation process. Finally,the ownership of the water should remain in the

hands of the public to ensure that the rights of thepublic are protected.

The private tubewell revolution in South Asiasince the 1980s is an excellent example of effectiveprivate investment and private-public investmentsynergies. Public investments have been an impor-tant facilitator of private irrigation investment.Private tubewells in South Asia have grown mostrapidly in areas with reasonably good roads, re-search and extension systems, access to electric ordiesel energy, and access to credit; therefore, thewells have been concentrated in and around thecommand areas of large, publicly developed sur-face irrigation systems. Seckler (1990) notes threereasons for the complementarity between publicand private irrigation investment in South Asia:deep percolation losses from the surface systemsrecharge the aquifers for tubewells; tubewells areoften used together with surface irrigation water,which lowers pumping costs and concentratesthese costs in periods of the highest marginalreturns; and the tubewells ride piggyback on theinfrastructure created for the surface systems.

The role and structure of agricultural researchare changing over time. The relative importance ofproductivity-driven growth will increase, becausegrowth in input use is declining as many regions inAsia are reaching high levels of input use. Privateinvestment in agricultural research—which gener-


Figure 4.2 Sample of Public-Private Investment Arrangements

Source: MIGA (2004).

PUBLIC-PRIVATE ARRANGEMENTS FOR THE PROVISION OF WATER AND WASTEWATER SERVICES

PUBLIC PRIVATE

CAPITAL COMMITMENT

Public sector owns the system and makes final investment and funding decisions. Private sector manages the utility and is paid on a profit-sharing basis.

Public sector owns the system and is responsible for investments and tariff setting. Private sector manages utility, collects bills, and pays an operator’s fee.

Private sector finances and develops the project and manages the utility until transfer of ownership to the public sector.

Private sector owns and operates the system.

Managementcontracts

Leasecontracts

Build-operate-transferarrangements

Privatization/concessionarrangements


ates significant public benefits—will increase inimportance, if policy reforms continue to createand improve the incentives for private investmentsby eliminating price distortions and strengtheningproperty rights. Market failures and social objec-tives will continue to call for an important role forpublic investment in agricultural research, however.

Through partnerships, public research institu-tions can gain access to advanced scientific knowl-edge and technologies held by the private sector,mechanisms for developing, processing, marketing,and distributing final products to farmers and con-sumers, and financial resources that are increas-ingly difficult to obtain. There are few examples,however, of successful public-private partnershipsin agricultural research. Part of the problem stemsfrom the sectors’ inherently different research objec-tives: while public institutions conduct researchaccording to their broad social mandates, firmspursue more narrow, profit-maximization goals.Furthermore, public institutions and private firmscompete over the ownership and use of proprietaryscientific knowledge and technologies, over scarcefinancial resources for research, and over markets,clients, and beneficiaries for their outputs. Thiscompetition incurs real and hidden costs andrisks that make partnerships difficult to create or sustain—costs and risks that are exacerbated bydeep-rooted misperceptions and information gapsbetween the sectors. To succeed, public-privatepartnerships require that both parties identifyresearch areas where their objectives are compati-ble, and where both parties are willing to engage indetailed and often difficult negotiations for projectplanning and implementation (Spielman and vonGrebmer 2004).

Agricultural research is often long term, largescale, and risky, and while the returns to newtechnologies are often high, the firm responsiblefor developing the technology may not be able toappropriate the benefits accruing to the innova-tion—as in the case of improved open-pollinatedrice and wheat varieties. The benefits of agricul-tural research often accrue to consumers (throughreduction in commodity prices due to increasedsupply), rather than to the adopters of the newtechnology, so social returns may be greater thanprivate returns to research. Therefore, a sustainedpublic role in funding agricultural research will beessential, particularly for crops and regions in lessfavorable environments, which are unlikely to beserved by the private sector.

New agricultural technologies in Asia—such astechnologies to improve pest management and thenutrient balance and the timing and placement offertilizer applications—are increasingly complex,knowledge intensive, and location specific; theydemand continued investment to create a betterand more decentralized research and extension sys-tem. Because new technologies are more demand-ing for both the farmer and the extension agent,they require more information and skills for suc-cessful adoption compared to the initial adoption ofmodern varieties and fertilizers. Decentralization ofexisting extension services structures that encouragea bottom-up flow from farmers to extension andresearch could also help farmers cope with the addi-tional complexity of efficiency-enhancing tech-nology. Bottom-up information flows, combinedwith adaptive, location-specific research, areparticularly important in the transfer of complexcrop-management technologies. Other moderntechnologies, such as commercial poultry technol-ogy, will be transferred essentially intact fromdeveloped countries, without local adaptation, butwill similarly require higher levels of educationand management skills than traditional livestockoperations. Finally, the increasing importance ofnew, knowledge-intensive technology requires amarket-friendly environment for the adoption andadaptation of new technologies and the removalof restrictions on technology imports, which mustbe encouraged through continued progress in eco-nomic liberalization. Privatization of extensionthrough contracting to private companies can intro-duce incentives for higher efficiency, and has beenapplied by seed companies, including for hybridmaize. Privatization of extension is likely to be suc-cessful when extension is linked to the delivery of aspecific technology (such as hybrid maize or poul-try) and to larger, more homogeneous groups offarmers. For commodities where private extensionservices cannot be self-supported, the govern-ment needs to continue providing assistance andtraining.

In contrast to other developing regions of theworld, the private sector in Africa is not increas-ing its research efforts as government spendingdeclines. Given a share of 2 percent in total spend-ing, the private sector plays an exceptionally smallrole in funding agricultural research in the region.Increasing their contribution is highly unlikelybecause the potential profits from conducting re-search on important crops are not sufficiently

high to attract the interest of either domestic orinternational private firms in Africa (NEPAD 2002).This is a marked contrast to industrial countries,where private enterprises fund over 50 percentof agricultural research. African governments havereduced their support to agricultural research andextension because of the pressures to reduce spend-ing in general. This may also be due to the shift inpriorities when governments question the value ofresearch and extension, given the lack of improve-ment in agricultural productivity in Africa. Donorassistance to agricultural research likewise declinedas a result of priorities that shifted from agriculturalproduction to environmental protection, health,education, water and sanitation, and so forth. Manyquestion the need of continued public funding onagricultural research and extension, thinking thatthe world’s food problems are solved, or con-strained by research systems or extension services,or the private sector will do the job. In order to cor-rect these perceptions, it is necessary to maintainand increase support to these services, which arefundamental to maintain the competitiveness ofagricultural economies in Africa (NEPAD 2002).

Private financing of investment may be exploredas a general way to ensure availability of basic ser-vices, particularly as the official development assis-tance (ODA) or aid for the water and other sectorshas been declining in recent years (Winpenny 2003).

Nonetheless, the U.N. Conference on Financing forDevelopment in Monterrey in 2002 declared thatthe decline in ODA should be reversed and assis-tance should increase by 25 percent or $12.5 bil-lion by 2006. To allow the agriculture sector todevelop its full potential for achieving the MDGs,the share of ODA spent on agriculture needs toincrease significantly.

For the case of Africa, NEPAD (2002) reportedthat estimates on the likely distribution of financ-ing between the public and private sectors remainshighly hypothetical and requires specific countryconditions to be taken into account. Table 4.2 pre-sents one set of assumptions on financing. Thepublic sector is expected to take the lead in waterand land development and rural infrastructure; itis also expected to increase food supplies at anapproximated $7.5 billion, albeit with considerablematching contributions from the farm sector. Thetotal incremental investment requirement, includ-ing operations and maintenance, is about $15.7 bil-lion per year from 2002 to 2015. This figure wasderived from both the national public and privateresources in addition to international cooperation,in line with the Monterrey 2002 commitments onfinancing for development.

A report by Atkinson (2003, cited in Reisen2003) focused on the role of rich countries and onthe sphere of public finances, but urged not to limit


Table 4.2 Possible Scenario Regarding Financing Sources for Agriculture Under NEPAD

Share of total investment (US$ billion)

Immediate future Short term Medium term TotalSource of investment (2002–2005) (2006–2010) (2011–2015) (2002–2015)

AfricaPublic domestic sources 19.6 40.0 37.8 97.4Private domestic 2.8 10.0 14.2 27.0Subtotal 22.4 50.0 52.0 104.4ExternalConcessional assistance

(such as ODA) 25.2 35.0 28.3 88.5Nonconcessional loans 5.6 10.0 4.7 20.3FDI (private) 2.8 5.0 9.5 17.3Subtotal 33.6 50.0 42.5 126.1Rounding adjustment — — — 0.8Total 56.0 100.0 94.5 251.3Annual 14.0 20.1 18.9 17.9

—Not applicable.Source: NEPAD (2002).


perspectives to a resource transfer from rich topoor countries. Rather, options should also con-sider income redistribution and tax efforts in China,India, and other poor countries to co-finance theMDGs. The additional $50 billion considered nec-essary for achieving the MDGs could be financedthrough both traditional and innovative means,such as through the enactment by all major econo-mies of a global tax on carbon use; an agreement bythe U.S. and other major economies to create addi-tional SDRs for development purposes; the enact-ment of a currency transaction tax (Tobin tax); theestablishment of a Global Lottery in agreementwith national lotteries; measures to increase flowsof remittances by immigrant workers for develop-ment purposes; other global taxes, such as a braindrain tax, an international airport tax, taxation ofocean fishing, taxation of arms exports, a “bit” taxon computer use, or a luxury goods tax.

The Role of Foreign Direct Private Investment

International capital flows. Foreign direct investment(FDI) and other long-term, relatively stable invest-ment have a significant impact on agricultural andoverall economic growth. FDI in developing coun-tries by companies based in industrial countrieshas increased substantially with globalization andnow dwarfs foreign aid. Private investment wasfive times greater in 1998 than in 1968, in constant1993 U.S. dollars. However, the impact of thisinvestment on hunger and poverty may be limited.Little of this investment goes directly into agricul-ture or rural areas. Moreover, FDI is largely con-centrated in just a few countries. Only five countriesaccounted for over half of all FDI in 1998 and just12 for all but a small amount of FDI (Runge andothers 2003).

In contrast to FDI, the benefits of short-terminternational capital are relatively small and uncer-tain because, unlike FDI, short-term capital doesnot bring along technology and managementinnovations (World Bank 1998). Moreover, whensavings rates are already high and marginal invest-ment is misallocated, short-term capital greatlyincreases the vulnerability of the economy. Manage-ment of international capital flows should there-fore focus on the creation of an environmentconducive to long-term investments and discour-aging to short-term capital inflows. Tax incentivesand other distortions that favor short-term inflows

over long-term investments should be eliminated.Moreover, both prudential regulations on currencypositions of banks and strengthened supervisionof these regulations and other risk-managementprocedures are required. Finally, short-term andunhedged borrowing by corporations should bedisclosed to reduce the credit risk.

The Role of Governance Structures inAgriculture for Achieving the MDGs

According to the Transparency InternationalCorruption Perceptions Index 2004 (CPI), cor-ruption is considered rampant in many develop-ing countries. Among the 14 worst offenders—expressed as a level of 2 or less with 10 indicatinga clean score—are five countries in Sub-SaharanAfrica, three in Asia, two in Latin America, andfour in the former Soviet Union (TransparencyInternational 2004).

Perceived levels of corruption are but one indi-cator of lack of adequate governance structures inthese countries, and lack of governance is consid-ered one of the major stumbling blocks to achiev-ing the MDGs in general, and to realizing theMDGs through improvements in the agriculturalsector in particular.

Indications abound that poor governance isrelated with bad outcomes for the population, andparticularly the (rural) poor. Conflicts that are oftenassociated with poor governance are linked withhunger, both as a cause and an effect of food inse-curity (Messer and Cohen 2004). Good governanceconveys stability and security, which again haveimportant linkages to agricultural and economicdevelopment (Zhang 2004). Estache and Kouassi(2002) showed that high levels of inefficiency inthe African water sector are linked with weak gov-ernance and institutions.

Good state governance is typically defined underthe terms of accountability, transparency, pre-dictability, and participation. Accountability trans-lates into government institutions following clearlines of responsibility that ultimately end up withthe electorate, from which authority flows. Trans-parency is a concept that makes it possible for gov-ernments (and other institutions) to be accountable.Predictability relates to the knowledge that the rulesof the games are certain and predictable. Partici-pation is both a principle and a means for achievinggood governance. For example, enhanced partic-ipation of communities in local government can

increase accountability, transparency, and pre-dictability. These principles are only meaningfulgiven adequate institutional and social structures,where they can be implemented.

Countries with a good governance structure andadequate institutions tend to ensure political andeconomic stability, possess reasonable state capac-ity, enforce property rights and contracts, providesufficient public goods, and limit government cor-ruption and predation. On the other hand, coun-tries with poor governance and poor institutionstypically have poor public services, including foragricultural extension and research, and particu-larly for social services such as water provisionand education (Wolfensohn and Bourguignon2004). In such countries, the neediest—typicallypoor farmers and other rural dwellers—are leftout of the supply chain, as resources are deviatedto private pockets and to those who can bribe orotherwise coerce the government to supply theirneeds. Moreover, the voiceless rural people inthese cases tend to have no access to recourse orcomplaints. At a higher level, the investment cli-mate in countries lacking good governance andquality institutions is adversely impacted, result-ing in fewer and more expensive private-sectorinvestments in both rural and urban areas. Goodpublic governance should include the attainmentof development targets and delivery of crucialpublic services for the rural poor. On the otherhand, good corporate governance essentially meansresponsible corporate citizenship, that is, respon-sible stewardship of the assets of the company byincreasing the economic value added withoutdamaging the environment or abusing consumersin the process.

How can governance structures in agriculturebe amended for agriculture to play its due role inachieving the MDGs? Good governance can bebuilt through the development of social capital.Social capital affects economic development mainlyby facilitating transactions among individuals,households, and groups in society. This facilitatingfunction can take the following forms:

• Participation by individuals in social net-works increases the availability of informationand lowers its cost. This is true in the rural vil-lages, especially when the information canincrease the returns from agriculture and trad-ing. Examples include crop prices, location ofnew markets, sources of credit, or treatmentsfor plant or livestock diseases. Often, social

capital is the only asset to which the rural poorhave easy access.

• Participation in local networks and attitudesof mutual trust make it easier for any group toreach collective decisions and implement col-lective action. Since property rights are oftenimperfectly enforced in developing coun-tries, collective decisions on how to managecommon or communal resources (such aswatershed land, irrigation, drinking water,and urban waste disposal) can help maximizetheir access, use, and benefits.

• Networks and attitudes reduce opportunis-tic behavior by community members. Socialpressures and fear of exclusion can makeindividuals behave in certain community-beneficial ways. For example, farmers havebeen known to use such networks and atti-tudes to exert mutual pressures to preventindividual diversion of irrigation water, or toprevent loan defaults in group borrowing sit-uations with joint liabilities (SEARCA/IFPRI/CRESCENT 2004).

In addition to the explicit implementation ofgood governance principles, the development andsupport of community-based organizations (CBOs)and general civic education can play an importantrole in good-governance reforms in agriculture.Effective CBOs, such as farmer associations or coop-eratives, water user groups, farm and other micro-credit and lending groups, or other existing ruralCBOs, can improve governance in several ways: byencouraging government ministries to adopt suc-cessful approaches developed within these groups;by educating and sensitizing the public about theirrights and entitlements under public programs; byacting as a conduit to the government for publicopinion and local experience; by collaborating withofficial bodies; by influencing local agriculturaldevelopment policies of national and internationalinstitutions; and by helping government anddonors fashion a more effective development strat-egy through strengthening institutions, staff train-ing, and improving management capacity. Pitfallsfor CBOs can be their own internal governancestructures and processes, particularly their trans-parency and accountability. These organizationsshould strive for financial sustainability, for exam-ple, through developing stronger links with theprivate corporate sector. This is especially workablefor nonadvocacy CBOs, such as those that delivereducational, healthcare, and humanitarian services.



As has been shown in Indonesia, decentraliza-tion can improve governance in rural areas, aslocal governments are usually more adept withand knowledgeable about rural agricultural needsin particular. However, administrative decentraliza-tion without preparatory capacity building and ade-quate fiscal support could force regional and districtgovernments to enact new tax laws and otherregulations that add to the costs of doing businesswithout corresponding increase in public services,or improvement in the quality of governance(SEARCA/IFPRI/CRESCENT 2004).

Improved governance will help lessen con-flict and encourage greater levels of governmentfunding to be allocated to the agriculture sector.Although governance is important to sustainableagricultural development in all countries, the spe-cific political situation in each country will deter-mine the appropriate measures needed to ensuregood governance.

With rapid development of agricultural exportsand trade, as well as with the very rapid growth ofthe retail sector in many developing countries, ithas become more important for the agriculturesector, including smallholders, to be in tune withcorporate and retail sector development, increas-ingly the buyers of agricultural products. Becauseof these developments, improvements in corpo-rate governance should be promoted to supportagriculture as well, such as by consistently enforc-ing laws and providing some incentives for goodgovernance practices. Good governance in the ruralnonfarm sector is also essential, so that the rural pri-vate sector, particularly small- and medium-sizeenterprises, can exploit the emerging investmentand business opportunities in the rural sectoroffered by agriculture and related activities.

At a higher level, donor countries and interna-tional aid agencies should focus their resources onthese countries where good or improving gover-nance structures will ensure that the rural poorare reached. The poverty reduction strategy papers(PRSPs) that the World Bank and others rely onincreasingly include sections on addressing gover-nance objectives. Whereas the MDGs provide along-term perspective, PRSPs can provide medium-term targets to address these long-term goals. BothPRSPs and reaching the MDGs requires strongnational planning and implementation capacity, afeature of good governance. Increased levels anddirection of ODA toward agriculture are also impor-tant (Wolfensohn and Bourguignon 2004).

Finally, at the international level, enhanced gov-ernance and commitment could help improve theglobal trade agenda by making progress regard-ing the needs of the poorest countries, includingenhanced access to agricultural and other markets.Enhanced international governance and commit-ment can also improve the global environmentalagenda, as developing countries would gain fromenhanced environmental standards, including pro-tection of remaining biodiversity in these countries,or from participating in and gaining from climatemitigation policies.

CONCLUSIONTo meet the MDGs, there are several concerns thatall organizations, from the government to the pri-vate sector, in the developed and developing worldneed to face. To begin with, the complementaryroles of government agencies, the private sector,and public-private partnerships cannot be ignoredwhen addressing the lack of basic services in water,land, health, and other infrastructure developmentin most developing countries. Likewise, the assis-tance of international and regional organizationsfor financing investments, extending technologi-cal and intellectual support, and strengtheningskills and capacity will move forward economicgrowth and progress in the developing world.

Government agencies in developing countriesneed urgently to revisit the legal, regulatory, polit-ical, and institutional framework in the agriculture,research, extension, and industrial sectors. A refo-cusing of priorities and policies to meet the spe-cific needs of poor people, especially on propertyrights in land and water to augment their require-ments of food and nutrition security, must beachieved to adequately address MDG 1. In addi-tion, government agencies, through the assistanceof international and regional organizations, mustfacilitate the development of alternative ruralnonfarm livelihoods.

Both the private and public sectors must fosterprivate-public sector partnerships with the objectiveof addressing the MDGs, and specifically MDG 8—to develop a global partnership for development.Several methods have been discussed here on howthe private sector and government institutions canwork together to create mutual opportunities andbenefits. It is also anticipated that the private sectorwill increase its investment in agricultural researchand extension in developing countries. Use of food

aid has to be considered judiciously to avoid nega-tive impacts.

Good governance is also a cross-cutting issue.Governance plays a role in the decision makingand implementation of policies both nationallyand internationally, in public administration, andin service delivery, and therefore plays a role inachievement of the MDGs.

Finally, if the appropriate investments andgovernance are to help reduce global poverty, con-

ducive and sustainable international price incen-tives must be brought about through multilateralreform of agricultural trade and domestic sub-sidy policies. Current trade barriers and subsi-dies are high for agriculture. Benefits will accrueto poor farmers from reforms both in developedand developing countries. Successful WTO agri-cultural trade negotiations that lower protectionand subsidies could thus boost progress towardthe MDGs.


69

The most effective strategy for making steady, sustainable progress onthe Millennium Development Goals is to serve all the goals in an inte-grated way. However, each goal will need a well-defined package oftechnologies and services for success at the field level. Pursuingthe goals separately without acknowledging their interlinkages willreduce the complex process of human and economic development toa series of fragmented, conflicting, and unsustainable interventions.A comprehensive and harmonious development approach is in order.

Given that the majority of poor people live in villages or rely onagriculture, and that agriculture paves the way for economic growthin the poorer nations, agricultural and rural development will under-lie progress on the broad array of economic and social indicators thatthe MDGs emphasize.

In pursuing the MDGs, we should seek ultimately the eliminationof hunger, poverty, and maternal and child malnutrition, with eachMDG being an important step along the way to this ultimate target.An emphasis on healthy, productive individuals means that we mustattend not simply to food security at the aggregate level, but to nutri-tion security (economic, physical, social, and environmental access toa balanced diet and clean drinking water) at the individual level ofchild, woman, and man. In the longer term, the goals should be mod-ified to promote a reduction in the absolute number of people livingin unsuitable conditions across all countries, rather than a reductionin global proportions, and ultimately the elimination of hunger andmalnutrition.

Despite these limitations in framing the task at hand, the MDGscan be used to set a powerful agenda for developing countries andthe international community, because they offer a guide for plan-ning and implementing a broad range of development efforts. Of theeight Millennium Development Goals, the first goal is the one whoseattainment most clearly involves the agricultural sector: The pooraround the globe are disproportionately farmers and herders, and,perversely, the hungry also most commonly find their livelihoodsthrough agriculture. The impact that a dynamic agricultural sectorwill have on the attainment of the other seven goals is less direct.Nonetheless, important gains in these can be made through explicitattention to agriculture.

1. Eradicate extreme hunger and poverty. Improving the productivityof and the economic returns of agriculture will have immediateeffects in eradicating extreme poverty and reducing hunger. In

Conclusions

Agriculture paves the

way for economic growth

in poorer nations and

underlies progress on the

broad array of economic

and social indicators that

the MDGs emphasize.

5

addition, subsistence farming households willenjoy immediate benefits from increased foodproduction. But more importantly, increasedfood production will lead to real reductions infood prices, improving the purchasing power ofthe poor throughout the economy. Furthermore,economy-wide effects will occur when the focuson agriculture increases. Agricultural growthwill, therefore, have ripple effects into othersectors—enhancing the productivity, increasingthe returns, and improving the incomes of thoseworking in sectors quite far removed from farm-ers, herders, and fishermen and women.

2. Achieve universal primary education. Investmentsin agriculture will advance progress towardthis goal primarily in an indirect fashion. At thehousehold level, the profitable use of labor-saving agricultural technologies will reduce theopportunity costs to farmers who allow theirchildren to go to school. Broad economic growthlikely will demand increasingly skilled labor,and will increase the returns of investing inschooling one’s children, both in agriculture andin other sectors.

3. Promote gender equality and empower women.Throughout the developing world, but particu-larly in Sub-Saharan Africa, women are farm-ers. To a greater degree than for men, women’sprincipal productive activities are in agricul-ture. Agriculture also provides key contribu-tions to the economic empowerment of women.A dynamic agricultural sector that offers broadwelfare benefits can only be expected to emergewhen women are given the opportunity to par-ticipate profitably in the sector. However, if thebenefits of a dynamic agriculture are to result insustained improvements in the direct determi-nants of welfare—income, health, and educa-tion, among others—it is necessary that womenhave an important role in determining how thefruits of their agricultural activities are used.

4. Reduce child mortality. The linkages betweenagriculture and child mortality are indirect butimportant. Agriculture is a critical componentin assuring food and nutrition security. It is esti-mated that 45 to 55 percent of all child deathsare due to malnutrition exacerbating the nega-tive effects of disease (Pelletier and others 1994).In addition, increased agricultural productivitymay come at the expense of women’s abilityto offer dedicated care to children in their care.Consequently, agricultural labor and time-

saving innovations are required to eliminate thispotentially zero-sum tradeoff.

5. Improve maternal health. Agriculture can con-tribute to this goal in a similar manner to how itcontributes to attaining all of the previous threegoals. In addition, agriculture can contributeto improved maternal health in other ways.Agriculture has considerable potential to directlyimprove maternal health by improving the qual-ity of the diets of women. Micronutrient defi-ciencies are particularly severe among youngchildren and women. Among women, the healtheffects of such deficiencies are most pronouncedduring pregnancy, at birth, and in the monthsfollowing birth. By increasing the micronutrientcontent of food crops the prevalence of micro-nutrient deficiencies among women shoulddecline and maternal health would improve.

6. Combat HIV/AIDS, malaria, and other diseases.The analysis for several of the other goals alsoapplies here. Although the indicators for thisgoal have no or very little agricultural contentand the direct links between agriculture andthese health issues are not immediate, they areimportant nonetheless. For agricultural house-holds, as for all households, the productivity oftheir economic activities is an important deter-minant of whether or not such households livein an environment that allows them to enjoy ahealthy and active life. Besides, a dynamic agri-cultural sector also has the potential to radicallyalter the disease environment in a region. Newhealth challenges likely will emerge with anevolving agricultural sector. For example, wage-labor migration associated with agriculture mayexpose populations to new diseases, increasingthe health burden they bear. The resources ofthe agriculture sector, particularly in the publicsphere through the extension services, can beused in a coordinated fashion with those of thehealth sector to address such challenges.

7. Ensure environmental sustainability. This MDGcovers a broad sweep, including biodiversity,critical natural habitats, energy use and globalclimate change, unsafe water and poor sanita-tion, and urban slums. A productive agriculturalsector will reduce pressure in all of these areas,including the following:• A productive agriculture requires less land,

leaving marginal agricultural lands to otheruses, including forests and other criticalhabitats.


Conclusions 71

• Proper agricultural policies will allow the fullcosts of agricultural technologies to be consid-ered as they are being used, reducing the scopefor excessive nutrient runoff from agriculture,providing incentives for efficient energy use inthe sector, and promoting the ecologically sus-tainable use of a range of technologies, includ-ing pesticides and GMO materials.

• Agriculture has the potential to be an impor-tant component in any systems established tomanage global carbon stocks.

• A dynamic agricultural sector and the pub-lic revenues that accrue from the sector willenable greater levels of public provision ofsafe drinking water and improved sanita-tion. Increasingly productive and profitablefarmers in many developing countries will bemore able to privately provide these ameni-ties for their own households. For agriculturaldevelopment programs to be environmen-tally sustainable, their long-term environmen-tal costs and benefits have to be taken intoaccount. Plus, policies and regulations need tobe in place to encourage efficient—instead ofexcessive—energy, water, fertilizer, and pesti-cide use. Sound water management in agri-culture is critical for safe drinking water, aswell as for prevention of water-borne dis-eases and wasting of water. At the sametime, agriculture-led economic growth willprovide public revenues that governmentscan use to provide safe drinking water andbetter sanitation, as well as higher incomesthat will allow individual farming householdsto invest in these basic needs. The issue ofbetter resource management arises in urbanareas as well. Urban water subsidies, forexample, go disproportionately to the betteroff in most developing countries because theyare connected to the public system. The urbanpoor, who must rely on water vendors, paymany times more for water than better-off res-idents. Removing such subsidies and usingthe available money to finance wider distrib-ution of piped water would benefit the poor.

• Population pressures in urban slums willbe alleviated to a significant degree if prof-itable agricultural systems are developed inthe rural areas.

8. Develop a global partnership for development. Thereare several areas where agriculture can con-tribute to achieving the targets under this goal,

especially to achieve an open, rule-based, pre-dictable, nondiscriminatory trading and financialsystem. A renewed focus on agriculture couldsupport improvements toward achieving thisgoal. Global agricultural trade must be harmo-nized and rationalized. Moreover, the rationalityused must include a judicious consideration ofthe special needs of poor agricultural producersand how they might derive benefits from globalagricultural trade. The sectoral (re)allocationof reoriented ODA should be specified and, withthe preponderance of poor people in ruralareas, agriculture and rural development shouldreceive significant priority in allocations.

In the remainder of this conclusion, the focus ison policies to meet MDG 1. Eradicating hunger andpoverty requires an understanding of the ways inwhich these two injustices interconnect. Hunger,and the malnourishment that accompanies it, pre-vents poor people from escaping poverty because itdiminishes their ability to learn, work, and care forthemselves and for their family members. If leftunaddressed, hunger sets in motion an array of out-comes that perpetuates malnutrition, reduces theability of adults to work and to give birth to healthychildren, and erodes children’s ability to learn andlead productive, healthy, and happy lives. Thistruncation of human development undermines acountry’s potential for economic development—forgenerations to come.

There are strong, direct relationships betweenagricultural productivity, hunger, and poverty.Three-quarters of the world’s poor live in rural areasand make their living from agriculture. Hunger andchild malnutrition are greater in these areas than inurban areas. Moreover, the higher the proportion ofthe rural population that obtains its income solelyfrom subsistence farming (without the benefit of pro-poor technologies and access to markets),the higher the incidence of malnutrition. Therefore,improvements in agricultural productivity aimed atsmall-scale farmers will benefit the rural poor first.

Increased agricultural productivity enables farm-ers to grow more food, which translates into betterdiets and, under market conditions that offer a levelplaying field, higher farm incomes. With moremoney, farmers are more likely to diversify produc-tion and grow higher-value crops, benefiting notonly themselves but the economy as a whole. Alarger supply of agricultural products also bringsprices down, allowing both the rural and urban

poor to purchase essential foods for less money.Smaller food bills mean that landless poor peoplewill have more money to invest in assets, which willhelp them increase income and survive future eco-nomic shocks. This income and asset security helpsbuild a solid foundation for economic growth, byenabling people to work free from the debilitatingeffects of hunger and undernutrition. A flourishingagriculture sector also facilitates job creation in otherareas, such as the food processing and marketingsectors, and creates secondary economic effects inthe nonfarm economy.

By increasing food availability and incomesand contributing to asset diversity and economicgrowth, higher agricultural productivity and sup-portive pro-poor policies allow people to break outof the poverty-hunger-malnutrition trap. A closerexamination of how the targets of MDG 1 canbe achieved in Sub-Saharan Africa (Ethiopia andZambia) illustrates that countries in the region willneed to achieve significantly larger economicgrowth, and very soon. In Zambia, the GDP growthrate required to halve poverty is around 10.5 per-cent, and in Ethiopia, 5.8 percent. Achieving theserates of growth requires greater productivity in theagricultural sector. The unique structure of eachcountry’s economy determines the impact of thisagricultural growth on poverty reduction as well aswhich agricultural subsectors are the most promis-ing. In Ethiopia, investments in the staple food sec-tor have the largest potential impact on both growthand poverty alleviation, while in Zambia, the expan-sion of traditional and nontraditional goods is moreimportant. Both countries have in common sig-nificant constraints to market access that need tobe overcome to attain projected rates of agricul-tural and economic growth and decreased poverty.Multisector strategies, such as a combination ofnontraditional export growth with livestock andstaple production growth, can significantly reducepoverty in Ethiopia. By concurrently improvingrural infrastructure and market development, agri-cultural growth could help Ethiopia to achieveMDG 1. In Zambia, on the other hand, reaching thepoverty target under MDG 1 will be more difficult.

A global assessment of Target 2 of MDG 1 (halv-ing child malnutrition levels) shows that the com-bination of agricultural and economic growthtogether with larger investments in social sectors,including health and education, can substantiallynarrow the gap between the business-as-usual out-comes for 2015—24 percent of developing-country

preschool children malnourished—and the targetindicator—15 percent children malnourished—toreach 17 percent.

However, the outcome varies significantly bycountry and region. While Latin America, WestAsia and North Africa, and China will, on average,likely get close to the target indicator by 2015, evenunder business-as-usual, the likelihood that Sub-Saharan Africa and South Asia will come close totheir respective target rates is much smaller. In fact,on average, Sub-Saharan Africa and South Asia willnot be able to halve their shares of malnourishedchildren by 2015 even under very favorable agri-cultural, economic, and social conditions; butimproved conditions in these sectors can bring thecountries in these regions very close to the 2015target and can facilitate further reductions in mal-nutrition levels later on.

Under the MDG scenario, to bring developingcountries, particularly South Asia and Sub-SaharanAfrica, within reach of the preschool malnutritiontarget indicator, total investments in agriculturaland supporting sectors during 1995–2015 will haveto increase by $161 billion based on IMPACT-WATER calculations. The three main areas of invest-ment for the MDG scenario in percentage termsare rural roads, irrigation, and education. Togetherthese three areas will require $403 billion between1995 and 2015 to achieve the rapid levels in child-hood malnutrition simulated under the MDG sce-nario. The increase in irrigation investments is notonly due to the expansion of irrigated area, but alsoto the increased irrigation costs under the MDGscenario as a result of expansion in more costlyareas, such as Sub-Saharan Africa.

Agricultural research investments account for$109 billion, and $78 billion of investments towardincreasing access to safe water is also required. Dueto the long lags in the generation of impact fromagricultural research, increases in research expen-ditures, even beginning now, will have relativelysmall impacts on crop yields by 2015. Increasedinvestments in agricultural research are likely tobe essential to meet crop and animal productionneeds beyond 2015, however. Other investments,such as roads and irrigation, have significant lagsin impact as well, so implementing the invest-ment portfolio required for the MDG scenario willrequire very rapid action. As relatively high levelsof access to clean drinking water are alreadyachieved in the baseline scenario, only $15 billionin investments are added for the MDG scenario,


Conclusions 73

but these investments have basically no lag periodin becoming effective.

In addition to these investments, significant pol-icy and governance reform is required. How canthe poor benefit most from greater investmentsand higher agricultural productivity? Experiencehas shown that a number of key conditions helpmaximize the benefits of a growing agriculture sec-tor for poor people. To achieve faster agriculture-based growth rates, favorable macroeconomic andtrade policies, good infrastructure, and access tocredit, land, and markets must be in place. Theseconditions create level playing fields and give farm-ers incentives to adopt new and sustainable tech-nologies and diversify production into higher-valuecrops, actions that raise incomes and lift house-holds out of poverty.

At the macroeconomic level, trade liberalization,particularly in developing countries, will enhancethe targets under MDG 1. To achieve these overallgains, developing countries that will benefit frommore open markets abroad need to participate inmultilateral agricultural trade policy reforms. Im-proved investments in infrastructure are anotherimportant avenue to enhance MDG 1. However,if there is a lack of coordination at the country,regional, and donor level, the linkages and com-plementarities of infrastructure investment willnot be realized. As a result, fragmented approachesthat lack sufficient attention for substantive poli-cies and development issues do not help coun-tries achieve their MDG targets. Moreover, localrecipients need to be integrated fully in any invest-ment plans.

Other important areas of reform in the trad-ing area include the elimination of export subsi-dies; the move toward income-support insteadof production-stimulating support measures indeveloped countries; the expedition of importing-country risk assessments and regulatory changesthat can help open market access for poor countries;and the reform of the international and nationalgovernance of food aid programs. An improveddomestic regulatory framework would intensifycompetition among suppliers of essential inputs,such as seeds and fertilizer. In addition, the elimi-nation of trade barriers for agricultural products,especially the high-value-added products, wouldencourage a greater number of private entrepre-neurs to explore opportunities in agribusiness. Ahealthy market and private sector would providevalue-added, skilled work to the landless poor

and generate multiple livelihood opportunities inboth the farm and nonfarm sectors.

In addition to pro-poor economic and agricul-tural policies, agriculture, like other sectors, needsgood governance, absence of conflict, and well-functioning markets and private enterprise to flour-ish. As the financing requirements for realizing theMDGs are substantial, the private sector is increas-ingly called upon to fill investment gaps. Its com-plementary and supporting role in the provisionof basic services in water, land, health, and otherinfrastructure development that are lacking in mostdeveloping countries cannot be ignored. The devel-opment and business communities must increas-ingly recognize that the MDGs cannot be achievedand private enterprise cannot flourish withoutgreater and more equitable involvement of poorpeople in markets. The idea of enticing global pri-vate enterprise into developing-country markets isnot new but the expectations are different thistime around. In many respects they are driven by agreater understanding that not just any kind of eco-nomic growth will improve the lives of the poor. Itwill take a particular kind of private-sector involve-ment to generate the necessary economic transfor-mations. Private entrepreneurs are now increasinglyheld to environmental, social, and corporate gover-nance principles that stress sustainable businesspractices and adherence to labor standards. Withoutthese standards and practices, the private sector anddisadvantaged groups cannot mutually benefitfrom consumer, employment, and entrepreneurialactivities.

Likewise, international and regional organiza-tions’ assistance in terms of financing investments,extending technological and intellectual support,promoting skills, and strengthening capacity willmove forward economic growth and progress in thedeveloping world. Government agencies in devel-oping countries urgently need to revisit the legal,regulatory, political, and institutional frameworkin agriculture, research, extension, and industrialsectors to facilitate private-sector involvement.Moreover, both the private and public sectors mustfoster private-public–sector partnerships and culti-vate this relationship with the end objective ofaddressing the MDGs. To allow the agriculture sec-tor to develop its full potential for achieving theMDGs, the share of ODA spent on agriculture needsto increase significantly.

The role of good governance systems in achiev-ing the MDGs and agricultural and economic devel-

opment in general has long been recognized. Goodgovernance can be built and strengthened throughthe development of social capital. In addition to theexplicit implementation of good governance princi-ples, the development and support of CBOs andcivic education, in general, can play an importantrole in good-governance reforms in agriculture. Ashas been shown in Indonesia, decentralization canimprove governance in rural areas, as local govern-ments are usually more adept with and knowledge-able about rural agricultural needs in particular.Improved governance will help lessen conflict andencourage greater levels of government funding tobe allocated to the agriculture sector. Although gov-ernance is important to sustainable agriculturaldevelopment in all countries, the specific politicalsituation in each country will determine the appro-priate measures to ensure good governance. At ahigher level, donor countries and international aidagencies should focus their resources on those coun-tries where good or improving governance struc-tures will ensure that the rural poor are reached.Finally, at the international level, enhanced gov-

ernance and commitment could help bring aboutimprovements in the global trade and environmen-tal agenda.

When good governance, equitable markets, andthe other key conditions noted above are absent,poor farmers are unlikely to earn decent incomesand secure adequate diets for themselves andtheir families. If agriculture underperforms orfails, nonfarmers will also feel the negative effects.We need to keep uppermost in our minds thatsignificant gains in agricultural productivity haveprovided the critical first steps in economic dev-elopment in many countries. It is a promisingdevelopment that the review of progress—andlack thereof—in achieving the MDGs has reachedglobal attention. Calls for accountability andaction that have real impact on people are grow-ing because of that attention. Policy action andincreased investment in the critical arenas of sus-tainable agriculture productivity and food andnutrition security will be essential for respondingeffectively and responsibly to reach the MillenniumDevelopment Goals.


75

CHAPTER 21. This is particularly the case in Africa and Asia, but less so

in Latin America where poverty in several countries is con-centrated in the urban centers (Khan 2000).

2. Although here rural statistics are used as a rough proxy forthe population that is dependent on agricultural livelihoods,it is important to highlight that agriculture is also a key liveli-hood strategy for the urban poor, particularly in Africa.

3. Johnston and Mellor (1961), who provided one of the firstcomprehensive descriptions of the role of agriculture in eco-nomic development, note that “insufficient movement outof agriculture will perpetuate, or lead to, excessively smallfarms and serious underemployment of labor as the proxi-mate causes of sub-standard farm incomes (p. 590).”

CHAPTER 34. This section on Zambia draws on Lofgren et al. (2004), Thurlow

(2004), and Thurlow and Wobst (2004).5. Based on per-capita debt to GDP.6. The simulations assume somewhat optimistic projections for

the copper mining sector, including the stabilization of worldcopper prices and the availability of private investment (seeLofgren, Robinson, and Thurlow 2002 and Lofgren, Thurlow,and Robinson 2004).

7. Given that the purpose of this section is to identify differ-ences in the poverty-reduction potential across agriculturalsectors, no attempt is made to account for the fiscal costs ofraising productivity under these scenarios.

8. The Nonagriculture-Led Growth Scenario excludes produc-tivity gains in mining. For a discussion of the effects of re-newed mining growth see Lofgren, Robinson, and Thurlow(2002) and Lofgren, Thurlow, and Robinson (2004) andThurlow and Wobst (2004).

9. IMPACT-WATER is described in detail in Rosegrant, Cai,and Cline (2002).

CHAPTER 410. References and further details are provided in Beierle and

Diaz-Bonilla (2003).11. As mentioned by Rodrik (1998), extensive trade liberaliza-

tion during the 1980s along with other reforms helped someof the region’s leading reformers, such as Uganda andGhana, recover from long periods of economic decline. Butneither Uganda nor Ghana has yet reached the level of in-come per capita it attained in 1970s.

12. The Food Aid Convention is a voluntary agreement amongdonors that has attempted to establish global food aid tar-gets, eligible commodities, and other guidance criteria, buthas no enforcement capacity. The FAO also has an advisorycommittee on food aid. In addition, food aid is subject tolimited rules under the WTO, and those rules may be sub-ject to additional clarification in the ongoing Doha Roundnegotiations.

13. Wharton was one of the first to emphasize the importanceof infrastructure in the generation of positive externalitiesat the microeconomic level. The author recognized thatagricultural development is not exclusively determined bythe “economic behavior of the producers,” but also de-pends on the environment, which according to Wharton in-cludes physical-climatic, socio-cultural, and institutionalcomponents in what he calls “the agricultural infrastruc-ture.” We follow Wharton’s (1967) definition of infrastruc-ture that identifies three categories: one that is capitalintensive (like roads or bridges); one that is capital exten-sive (principally extension services or animal sanitationservices); and institutional infrastructure (consisting of for-mal and informal institutions).

Notes

77

ACC/SCN (United Nations Administrative Committee onCoordination–Subcommittee on Nutrition). 1996. Update onthe Nutrition Situation. ACC/SCN, Geneva, Switzerland.

AfDB (African Development Bank) and OECD (Organization forEconomic Cooperation and Development). 2002. Achieving theMillennium Development Goals in Africa: Progress, Prospects, andPolicy Implications. Paris, France.

AIARD (Association for International Agriculture and RuralDevelopment). 2003. The Agriculture, Nutrition, and HIV/AIDSConnections in Developing Countries. Resource paper preparedfor the US Agency for International Development.

Anderson, Kym, Will Martin, and Dominique van derMensbrugghe. 2004. “Price, Trade and Welfare Implicationsof the Doha Agenda.” Presented at the World Bank confer-ence Agricultural Trade Reform and the Doha DevelopmentAgenda. The Hague, December 1–2 (proceedings forthcoming).

Atkinson, A. 2003. New Sources for Development Finance. Helsinki:Wider.

Barrett, C., and D. Maxwell. 2004. Food Aid After Fifty Years:Recasting Its Role. Draft book manuscript. New York: CornellUniversity.

Beghin, J. C., D. Roland-Holst, and D. van der Mensbrugghe.2002. Global Agricultural Trade and the Doha Round: What are theImplications for the North and South? Working Paper 02-WP308, Center for Agricultural and Rural Development, June.Ames, Iowa: Iowa State University.

Beierle, T., and E. Diaz-Bonilla. 2003. “The Impact of AgriculturalTrade Liberalization on the Rural Poor: An Overview.”Background paper for the RFF/IFPRI Workshop on Agri-cultural Trade Liberalization and the Poor, 3–4 November,Washington DC.

Bigsten, A., and S. Kayizzi-Mugerwa. 2000. “The PoliticalEconomy of Policy Failure in Zambia.” Working Papers inEconomics, Department of Economics, Göteborg University.

Binswanger, H. P., S. R. Khandker, and M. P. Rosenzweig.1993. “How Infrastructure and Financial Institutions AffectAgricultural Output and Investment in India.” Journal ofDevelopment Economics 41(2): 337–366.

Borlaug, N. 1999. “Biotech Can Feed 8 billion in the NextCentury—If Environmental Extremists Permit It.” NewPerspectives Quarterly. Accessed at http://www.digitalnpq.org/global_services/nobel%20laureates/10-06-99.HTML.

Boserup, E. 1981. Population and Technology. Oxford, BasilBlackwell.

Brenneman, A., and M. Kerf. 2002. “Infrastructure & PovertyLinkages, A Literature Review.” The World Bank, Wash-ington, DC.

Chanthunya, C. L., and V. Murinde. 1998. Trade Regime andEconomic Growth. Vermont: Aldershot and Brookfield.

Deininger, K., and P. Olinto. 1999. Why Liberalization Alone HasNot Improved Agricultural Productivity in Zambia: The Role ofAsset Ownership and Working Capital Constraints. Washington,DC: The World Bank.

de Onis, M., M. Blössner, E. Borghi, E. A. Frongillo, and R. Morris. 2004. “Estimates of Global Prevalence of ChildhoodUnderweight in 1990 and 2015.” Journal of the American MedicalAssociation 291(21): 2600–2606.

Devarajan, S., W. Easterly, H. Fofack, D. Go, A. Izquierdo, andL. Pizzati. 2000. A Macroeconomic Framework for PovertyReduction Strategies. Development Research Group and WorldBank Institute. Washington, DC: The World Bank.

Diao, X., E. Diaz-Bonilla, and S. Robinson. 2003. “Tell Me Whereit Hurts and I’ll Tell You Who to Call: Industrialized Countries’Agricultural Policies and Developing Countries.” IFPRI Brief.International Food Policy Research Institute, Washington, DC,September.

Dirven, M. 1999. “The Role of Agents in Agricultural Policies:Intentions and Reality.” Cepal Review 68. Chile.

Easterly, W., and R. Levine. 1997. “Africa’s Growth Tragedy:Policies and Ethnic Divisions.” Quarterly Journal of EconomicsCXII (4): 1203–1250.

Edwards, S. 1995. Crisis and Reform in Latin America. Oxford andWashington, DC: World Bank, Oxford University Press.

EM-DAT. 2004. The OFDA/CRED International Disaster Data-base. www.em-dat.net—Université Catholique de Louvain—Brussels—Belgium

Estache, A., V. Foster, and Q. Wodon. 2001. Making InfrastructureReform Work for the Poor: Policy Options based on Latin AmericanExperience. LAC Regional Studies Program, WBI Studies inDevelopment, Finance, Private Sector and InfrastructureDepartment. Washington DC: The World Bank.

Estache, A., V. Foster, and Q. Wodon. 2002. Accounting for Povertyin Infrastructure Reform: Learning from Latin America’s Experience.Washington DC: World Bank Institute, The World Bank Group.

Estache, A., and E. Kouassi. 2002. “Sector Organization,Governance, and the Inefficiency of African Water Utilities.”World Bank Policy Research Working Paper 2890. The WorldBank, Washington, DC.

Fan, S., and P. Hazell. 2001. “Returns to Public Investments inthe Less-Favored Areas of India and China.” American Journalof Agricultural Economics 83(5): 1217–22.

———. 1999. “Are Returns To Public Investment Lower In Less-Favored Rural Areas? An Empirical Analysis of India.” EPTDDiscussion Paper 43, Environment and Production Technol-ogy Division, International Food Policy Research Institute.Washington, DC, May.

Fan, S., P. Hazell, and T. Haque. 2000a. “Targeting PublicInvestments by Agro-Ecological Zone to Achieve Growthand Poverty Alleviation Goals in Rural India.” Food Policy25(4): 411–428.

Fan, S., P. Hazell, and S. Thorat. 2000b. “Government Spending,Growth and Poverty in Rural India.” American Journal ofAgricultural Economics 82(4): 1038–51.

Fan, S., and N. Rao. 2004. “Public Spending in DevelopingCountries: Trends, Determination, and Impact.” PreliminaryWorking Paper. International Food Policy Research Institute.Washington, DC.

References

Fan, S., L. Zhang, and X. Zhang. 2002. “Growth, Inequality, andPoverty in Rural China The Role of Public Investments.”Research Report 125. International Food Policy ResearchInstitute. Washington, DC.

Fane, G. 1998. “The Role of Prudential Regulation.” In East Asiain Crisis: From Being a Miracle to Needing One? R. H. McLeodand R. Garnaut (eds). London and New York: Routledge.

FAO (Food and Agriculture Organization). 1998. FAO FAOSTATdatabases. Accessed at http://apps.fao.org/default.jsp.

———. 2001. Mobilizing the Political Will and Resources to BanishWorld Hunger. Rome: FAO.

———. 2003. The State of Food Insecurity in the World—2003.Rome: FAO.

FAO/IFAD/WFP (Food and Agriculture Organization, Inter-national Fund for Agricultural Development, and WorldFood Programme). 2002. “Reducing Poverty and Hunger:The Critical Role of Financing for Food, Agriculture, andRural Development.” Paper prepared for the InternationalConference on Financing for Development, 18–22 March,Monterrey, Mexico.

Fay, Marianne and Tito Yepes. 2003. “Investing in Infrastructure:What is Needed from 2000 to 2010?” World Bank PolicyResearch Working Paper 3102, World Bank: Washington, DC.

Fishbein, Robert. 2001. “Rural Infrastructure in Africa: PolicyDirections.” Africa Region Working Paper Series No.18. WorldBank.

Galiani, S., P. Gertler, and E. Schargrodsky. 2005. “Water for Life:The Impact of Privatization of Water Services on ChildMortality.” Journal of Political Economy Vol. 113 (1), pp. 83–120.

Garnaut, R. 1998. “Economic Lessons.” In East Asia in Crisis: FromBeing a Miracle to Needing One? R. H. McLeod and R. Garnaut(eds). London and New York: Routledge.

Gautam, M. 2003. Debt relief for the poorest: an OED review of the HIPC initiative. The Operations Evaluations Division.Washington, DC: The World Bank.

Geist, H. J., and E. F. Lambin. 2002. “Proximate Causes andUnderlying Driving Forces of Tropical Deforestation.” Bio-Science 52(2): 143–150.

Gleick, P. H., G. Wolff, E. L. Chalecki, and R. Reyes. 2002. TheNew Economy of Water: The Risks and Benefits of Globalizationand Privatization of Fresh Water. Oakland, CA: Pacific Institute.

Goldin, Ian and L. Alan Winters (Eds). 1992. “Open Economies:Structural Adjustment and Agriculture.” Cambridge Uni-versity Press, 342p. London.

Haggblade, S., and B. Zulu. 2003. “The Cassava Surge in Zambiaand Malawi.” Conference Background Paper No. 9 preparedfor the InWEnt, IFPRI, NEPAD, CTA conference Successes inAfrican Agriculture, Pretoria December 1–3, 2003.

Hazell, P., and S. Haggblade. 1993. “Farm-Nonfarm GrowthLinkages and the Welfare of the Poor.” In Including the Poor,M. Lipton and J. van der Gaag (eds). Washington, DC: TheWorld Bank. pp. 190–204.

Hertel, Tom, and Alan Winters. 2004. “Putting Development Backinto the Doha Agenda: Poverty Impacts of a WTO Agreement.”World Bank conference held in The Hague, December 2–4 (pro-ceedings forthcoming).

Hoddinott, J., M. Cohen, and M. S. Bos. 2003. “Food Aid in the21st Century: Current Issues and Food Aid as Insurance.”Paper presented at the international workshop on Definingthe Role of Food Aid in Contributing to Sustainable FoodSecurity, Berlin, Germany. September.

Howe, J. D. G. F., and P. J. Richards. 1984. Rural Roads and PovertyAlleviation: A Study Prepared for the International Labour Officewithin the Framework of the World Employment Programme.London: Intermediate Technology Publications.

IFAD (International Fund for Agricultural Development). 2001.Rural Poverty Report 2001: The Challenge of Ending Rural Poverty.Oxford University Press.

———. 2003. “Achieving the Millennium Development Goalsby Enabling the Rural Poor to Overcome their Poverty.”Discussion Paper, IFAD, Rome, Italy.

Jacoby, H. 1998. Access to Markets and the Benefits of Rural Roads.Development Research Group Rural Development. Washing-ton, DC: The World Bank.

Jalan, J., and M. Ravallion. 2001. “Does Piped Water ReduceDiarrhea for Children in Rural India.” Policy ResearchWorking Paper No. 2664. The World Bank, Washington, DC.

Jimenez, E. 1995. “Human and Physical Infrastructure: PublicInvestment and Pricing Policies in Developing Countries.”In Jere Behrman and T. N. Srinivasan (eds.) Handbook ofDevelopment Economics, Volume 3. Amsterdam: North-Holland.

Johnston, B. F., and J. W. Mellor. 1961. “The Role of Agriculturein Economic Development.” The American Economic Review51(4): 566–593.

Josling, T., D. Roberts, and D. Orden. 2004. Food Regulation andTrade: Toward a Safe and Open Global System. Washington DC:Institute for International Economics.

Khan, M. H. 2000. “Rural Poverty in Developing Countries: Issuesand Policies.” IMF Working Paper WP/00/78. InternationalMonetary Fund, Washington, DC.

Lebo, J., and D. Schelling. 2001. “Design and Appraisal of RuralTransport Infrastructure: Ensuring Basic Access for RuralCommunities.” Technical Paper No. 496. The World Bank,Washington, DC.

Leipziger, D., M. Fay, Q. Wodon, and T. Yepes. 2003. “Achiev-ing the Millennium Development Goals: The Role ofInfrastructure.” Policy Research Working Paper 3163. TheWorld Bank, Washington, DC.

Limão, N., and A. J. Venables. 1999. “Infrastructure, GeographicalDisadvantage, and Transport Costs.” Policy Research WorkingPaper 2257. The World Bank, Washington, DC.

Lipton, M., and M. Ravallion. 1995. “Poverty and Policy.” In JereBehrman and T. N. Srinivasan, (eds). Handbook of DevelopmentEconomics, Volume 3. Amsterdam: North-Holland.

Lofgren, H., Thurlow, J., and Robinson, S. 2004. “Prospects forGrowth and Poverty Reduction in Zambia, 2001–2015,” Dev-elopment Strategies and Governance Division DiscussionPaper No. 11, International Food Policy Research Institute,Washington, D.C.

Lofgren, H., S. Robinson, and J. Thurlow. 2002. “Macro andMicro Effects of Recent and Potential Shocks to CopperMining in Zambia.” Trade and Macroeconomics DivisionDiscussion Paper No. 99, International Food Policy ResearchInstitute, Washington, DC.

MAAIF & MFPED (Ministry of Agriculture, Animal Industry, andFisheries and Ministry of Finance, Planning and EconomicDevelopment). 2000. Plan for the Modernisation of Agriculture:Eradicating poverty in Uganda (Government strategy and opera-tional framework). Kampala, Uganda: MAAIF & MFPED.

Messer, E. and M. J. Cohen. 2004. Breaking the links betweenconflict and hunger in Africa. 2020 Africa Conference Brief 10.Washington, DC: IFPRI.

Meyer, R. L., and G. Nagarajan. 1999. Rural Financial Markets inAsia: Policies, Paradigms, and Performance. Oxford UniversityPress.

MFPED (Ministry of Finance, Planning and Economic Dev-elopment). 2000. Uganda’s Poverty Eradication Action Plan—Summary and Main Objectives. Kampala: MFPED.

Micronutrient Initiative & United Nations Children’s Fund(UNICEF). 2004. Vitamin & Mineral Deficiency: A Progress Report.Ottawa: The Micronutrient Initiative & New York: UNICEF.

MIGA (Multilateral Investment Guarantee Agency). 2004. MIGAand Water. Washington DC: World Bank Group.

Ministry of Agriculture and Rural Development. 2005. MillenniumDevelopment Goals Based Needs Assessment: The Rural Dev-elopment and Food Security Sector, draft, Government of Ethiopia:Addis Ababa.


Introduction 79

Minot, N., and L. Daniels. 2003. “Impact of Global CottonMarkets on Rural Poverty in Benin.” Discussion Paper No. 48.Markets, Trade, and Institutions Division, International FoodPolicy Research Institute, Washington, DC.

MOFED (Ministry of Finance and Economic Development).2002. Ethiopia: Sustainable Development and Poverty ReductionProgram. Addis Ababa, Ethiopia.

———. 2003. Ethiopia: Sustainable Development and PovertyReduction Program (SDPRP). Annual Progress Report. AddisAbaba, Ethiopia.

MOPED (Ministry of Planning and Economic Development),Ethiopia. 1994. Poverty Monitoring Mechanism for Ethiopia.Addis Ababa, Ethiopia.

Narayanan, S., and A. Gulati. 2002. “Globalization and theSmallholders: A Review of Issues, Approaches, and Impli-cations.” MSSD Discussion Paper 50, International FoodPolicy Research Institute, Washington DC.

NEPAD (New Partnership for Africa’s Development). 2002.Comprehensive Africa Agriculture Development Programme. Rome:Food and Agriculture Organization of the United Nations(FAO).

ORC/Macro. 2004. Demographic and Health Surveys STAT-compiler. Online DHS survey data extraction tool. Accessedat http://www.measuredhs.com/start.cfm.

Orden, D., M. Torero, and A. Gulati. 2004. “Agricultural Marketsand the Rural Poor.” DCD/DAC/POVNET/A(2004) 1/RD3.DAC Network on Poverty Reduction OECD, Paris. [Numbersare OECD internal tracking system]

Pelletier, D. L., E. A. Frongillo Jr., D. G. Schroeder, and J.-P.Habicht. 1994. “A Methodology for Estimating the Contri-bution of Malnutrition to Child Mortality in DevelopingCountries.” Journal of Nutrition 124(10 suppl.): 2106S–2122S.

Platteau, J.-P. 1996. “Physical Infrastructure as a Constraint onAgricultural Growth: The Case of Sub-Saharan Africa.” OxfordDevelopment Studies 24(3): 189–219.

PRSP. 2003. “Ghana Poverty Reduction Strategy Paper.” Inter-national Monetary Fund: Washington, DC.

PRSP. 2003. “Uganda Poverty Reduction Strategy PaperAnnual Progress Report.” International Monetary Fund,Washington, DC.

Quisumbing, A. (ed.). 2003. Household Decisions, Gender, andDevelopment. Baltimore: Johns Hopkins University Press forthe International Food Policy Research Institute.

Reisen, H. 2003. New Sources of Development Finance.An annotatedconference report. OECD Development Centre. Accessed athttp://www.oecd.org/dataoecd/56/27/28321228.pdf

Renkow, M., D. G. Hallstrom, and D. D. Karanja. 2002. “RuralInfrastructure, Transactions Costs and Market Participationin Kenya.” Journal of Development Economics, 73(1): 349–367.

Rodrik, D. 1998. “Trade Policy and Economic Performance inSub-Saharan Africa.” NBER Working Paper Series, No. 6562.Cambridge, MA.

Rosegrant, M. W., and S. Meijer. 2002. “Appropriate Food Policiesand Investments Could Reduce Child Malnutrition by 43%by 2020.” Journal of Nutrition 132: 3437S–3440S.

Rosegrant, M. W., X. Cai, and S. A. Cline. 2002. World Waterand Food to 2025: Dealing with Scarcity. Washington, DC:International Food Policy Research Institute (IFPRI).

Rosegrant, M. W., M. S. Paisner, S. Meijer, and J. Witcover. 2001.Global Food Projections to 2020. Emerging Trends and AlternativeFutures. Washington, DC: International Food Policy ResearchInstitute (IFPRI).

Rosegrant, M. W., and P. Hazell. 2000. Transforming the RuralAsian Economy: the Unfinished Revolution. Hong Kong: OxfordUniversity Press.

Runge, C. F., B. Senauer, P. G. Pardey, and M. W. Rosegrant. 2003.Ending Hunger in Our Lifetime: Food Security and Globalization.Baltimore: John Hopkins University Press.

Saasa, O. 2003. “Agricultural Intensification in Zambia: The Roleof Policies and Policy Processes.” Institute of Economic andSocial Research, University of Zambia.

Sachs, J., and A. Warner. 1997. Sources of Slow Growth in AfricanEconomies. Boston: Harvard Institute for InternationalDevelopment.

Samatebele, H. 2003. “Overview on the Current Food SecurityCrisis in Zambia.” In S. Drimie and M. Lafon (eds). Food Securityin Southern Africa: Causes and Responses from the Region.Johannesburg: Institut Français d’Afrique du Sud.

Schiff, M., and A. Valdez. 1998. “Agriculture and the Macro-economy.” In B. Gardnet, and G. Rausser (eds). Handbook ofAgricultural Economics. Elsevier Science.

SEARCA/IFPRI/CRESECENT. 2004. Agriculture and rural devel-opment strategy study. Final Report. Prepared for the Ministryof Agriculture, Indonesia. ADB TA 3843-INO. July. Bogor,Indonesia.

Seckler, D. 1990. “Private Sector Irrigation in Africa.” InIrrigation in Sub-Saharan Africa: The Development of Public andPrivate Systems, S. M. Barghouti and G. Le Moigne (eds).World Bank Technical Paper No. 123. Washington, DC: TheWorld Bank.

Smith, D. R., A. Gordon, K. Meadows, and K. Zwick. 2001.“Livelihood Diversification in Uganda: Patterns andDeterminants of Change Across Two Rural Districts.” FoodPolicy 26: 421–435.

Smith, L. C., and L. Haddad. 2000. Explaining Child Malnutritionin Developing Countries: A Cross-Country Analysis. ResearchReport 111. Washington, DC: International Food PolicyResearch Institute.

Smith, L. C., U. Ramakrishnan, A. Ndiaye, L. Haddad, andR. Martorell. 2003. The Importance of Women’s Status for ChildNutrition in Developing Countries. Research Report 131.Washington, DC: International Food Policy Research Institute.

Spielman, D., and K. von Grebmer. 2004. “Public-PrivatePartnerships in Agricultural Research: An Analysis ofChallenges Facing Industry and the Consultative Group on International Agricultural Research.” EPTD DiscussionPaper No. 113. International Food Policy Research Institute,Washington, DC.

Sitglitz, J. E. 2000. Economics of the Public Sector 3rd ed. New York:W. W. Norton & Co.

Thirtle, C., L. Beyers, L. Lin, V. McKenzie-Hill, X. Irz, S. Wiggins,and J. Piesse. 2002. The Impact of Changes in AgriculturalProductivity on the Incidence of Poverty in Developing Countries.Report to Department for International Development (DfID)no. 7946. London & East Kilbride, UK.

Thurlow, J. 2004. “Growth and Market Opportunities in ZambianAgriculture.” International Food Policy Research Institute,Washington, DC.

Thurlow, J., and P. Wobst. 2004. “The Road to Pro-Poor Growth inZambia: Past Lessons and Future Challenges.” ForthcomingDevelopment Strategies and Governance Division DiscussionPaper, International Food Policy Research Institute, Washing-ton, D.C. Input paper into the World Bank project “Opera-tionalizing Pro-Poor Growth,” Washington, D.C.

Timmer, C. P. 2003. Agriculture and Pro-poor Growth. Pro-poorEconomic Growth Research Studies, no. 12. Bethesda, MD:Development Alternatives, Inc. & Boston Institute forDeveloping Economies, Ltd.

Tokarick, S. 2002. “How Do Industrial Country AgriculturalPolicies Affect Developing Countries.” In World EconomicOutlook: Trade and Finance. Washington, DC: InternationalMonetary Fund.

———. 2003. “Measuring the Impact of Distortions in Agri-cultural Trade in Partial and General Equilibrium.” IMFWorking Paper, WP/03/110. International Monetary Fund,Washington, DC.

Torero, M., and S. Chowdhury. 2004. “Infrastructure for AFRICA:An Alternative and a Constraint for Development.” Briefprepared for the 2020 Africa Conference, Assuring Food andNutrition Security in Africa by 2020: Prioritizing Actions,Strengthening Actors, and Facilitating Partnerships. Inter-national Food Policy Research Institute, Washington, DC.

Torero, M, S. Chowdhury, and V. Galdo. 2003. “Willingness toPay for the Rural Telephone Service in Bangladesh and Peru.”Information Economics and Policy 15(3): 327–361.

Transparency International. 2004. Transparency InternationalCorruption Perceptions Index 2004. Transparency Inter-national: Berlin, Germany.

UNDP (United Nations Development Programme). 2004a.Opening Doors to Opportunity. Afghanistan’s Millennium Dev-elopment Goals. New York: UNDP. Accessed at http://www.undp.org/mdg/afghanistan/afghanistan_mdg.pdf.

———. 2004b. Unleashing Entrepreneurship: Making BusinessWork for the Poor. New York: UNDP.

UNDP (United Nations Development Program) and UNICEF(United Nations Children Fund). 2002. The Millennium Dev-elopment Goals in Africa: Promises and Progress. New York:UNDP and UNICEF.

UNESCO. 1998. UNESCO Institute for Statistics. Accessed at[http://www.uis.unesco.org/ev_en.php?ID=2867_201&ID2=DO_TOPIC].

UNICEF. 2004. Monitoring the Situation of Children and Women.UNICEF global database on child malnutrition. Accessed athttp://www.childinfo.org/eddb/malnutrition.

Van Blarcom, B., O. Knudsen, and J. Nash. 1993. “The Role ofPublic Expenditures for Agriculture.” World Bank DiscussionPaper # 216. The World Bank, Washington, DC.

Van De Walle, D. 1996. Infrastructure and Poverty in Viet Nam.Washington, DC: The World Bank.

von Braun, J. 2003. Berlin Statement. Conference summary rec-ommendations drawn from the international workshop onDefining the Role of Food Aid in Contributing to SustainableFood Security, Berlin, September.

Wharton, C. 1967. “The Infrastructure for Agricultural Growth.”In Agricultural Development and Economic Growth. Southworthand Johnston (eds). Ithaca, NY: Cornell University Press.

WHO (World Health Organization). 2004. WHO Global Databaseon Child Growth and Malnutrition. Accessed at: http://www.who.int/nutgrowthdb.

WHO/UNICEF. 2001. WHO/UNICEF Joint Monitoring Prog-ramme for Water Supply and Sanitation Coverage Estimates1980–2000. Accessed at UNICEF Statistics on water: http://www.childinfo.org/eddb/water/database.htm.

Winpenny, J. 2003. Financing Water for All. Report of the World Panelon Financing Water Infrastructure. World Water Council, 3rd World Water Form, Global Water Partnership. Accessed athttp://www.worldwatercouncil.org/download/CamdessusSummary.pdf

Wolfensohn, J. D. and F. Bourguignon. 2004. Development andpoverty reduction. Looking back, looking ahead. Preparedfor the 2004 Annual Meetings of the World Bank and IMF.October. Washington, D.C.: World Bank.

Wood, Adrian. 2002. Could Africa Be Like America? Departmentfor International Development. processed.

World Bank. 1997. World Development Indicators 1997 on CD-ROM.Washington, DC: The World Bank.

———. 1998. East Asia: The Road to Recovery. Washington, DC:The World Bank.

———. 2002. World Development Indicators 2002. Washington, DC:The World Bank.

———. 2003. World Development Indicators, 2003. Washington, DC:The World Bank.

———. 2004a. Partnerships in Development: Progress in the FightAgainst Poverty. Washington, DC: The World Bank.

———. 2004b. “Global poverty down by half since 1981, butprogress uneven as economic growth eludes many countries.”News Release no. 2004/309/S, April 23, 2004. http://www.worldbank.org/data/wdi2004/index.htm.

———. 2004c. World Development Indicators 2004. on CD-ROM.Washington, DC: The World Bank.

———. 2004d. “Ethiopia Country Brief.” Updated June 2004.Accessed at: www.worldbank.org/ethiopia.

———. 2004e. Zambia Country Economic Memorandum: Policies forGrowth, Diversification and Poverty Reduction. Washington,DC: The World Bank.

Zhang, X. 2004. Security is like Oxygen: Evidence from Uganda.DSGD Discussion Paper No.6. Washington, DC: IFPRI.

Zhang, X., and S. Fan. 2000. “Public Investment and RegionalInequality in Rural China.” Discussion Paper No. 71, Envi-ronment and Production Technology Division, Inter-national Food Policy Research Institute, Washington, DC.December.


81

Appendix 1The Millennium DevelopmentGoals, Targets, and Indicators

Appendix 1 The Millennium Development Goals, Targets, and Indicators

Goal Target Indicators

1. Eradicate extremepoverty and hunger.

2. Achieve universal primary education.

3. Promote gender equalityand empower women.

4. Reduce child mortality.

5. Improve maternal health.

6. Combat HIV/AIDS,malaria, and other diseases.

1. Halve, between 1990 and 2015, the proportion of people whose income isless than one dollar a day.

2. Halve, between 1990 and 2015, the proportion of people who suffer fromhunger.

3. Ensure that, by 2015, children every-where, boys and girls alike, will be ableto complete a full course of primaryschooling.

4. Eliminate gender disparity in primary andsecondary education preferably by 2005and to all levels of education no laterthan 2015.

5. Reduce by two-thirds, between 1990 and2015, the under-five mortality rate.

6. Reduce by three-quarters, between 1990and 2015, the maternal mortality ratio.

7. Have halted by 2015 and begun to reverse the spread of HIV/AIDS.

1. Proportion of population below US$1(PPP) per day.

2. Poverty gap ratio [incidence × depth ofpoverty].

3. Share of poorest quintile in national consumption.

4. Prevalence of underweight childrenunder five years of age.

5. Proportion of population below minimumlevel of dietary energy consumption.

6. Net enrolment ratio in primary education.7. Proportion of pupils starting grade 1 who

reach grade 5.8. Literacy rate of 15–24 year-olds.9. Ratios of girls to boys in primary,

secondary, and tertiary education.10. Ratio of literate women to men

15–24 years old.11. Share of women in wage employment in

the nonagricultural sector.12. Proportion of seats held by women in

national parliament.13. Under-five mortality rate.14. Infant mortality rate.15. Proportion of 1-year-old children

immunized against measles.16. Maternal mortality ratio.17. Proportion of births attended by skilled

health personnel.18. HIV prevalence among 15–24 year old

pregnant women.19. Condom use rate.20. Ratio of school attendance of orphans to

school attendance of non-orphans aged 10–14.

(continued )


Appendix 1 The Millennium Development Goals, Targets, and Indicators (Continued)


7. Ensure environmentalsustainability.

8. Develop a global partnership for development.

8. Have halted by 2015 and begun to re-verse the incidence of malaria and othermajor diseases.

9. Integrate the principles of sustainable development into country policies andprograms and reverse the loss of environmental resources.

10. Halve, by 2015, the proportion of peoplewithout sustainable access to safe drinking water and basic sanitation.

11. By 2020, to have achieved a significantimprovement in the lives of at least 100 million slum dwellers.

12. Develop further an open, rule-based, predictable, nondiscriminatory tradingand financial system. Includes a commitment to good governance, development, and poverty reduction—both nationally and internationally.

13. Address the special needs of the least developed countries (LDC). Includes: tariff- and quota-free access for least developed countries’ exports; enhancedprogram of debt relief for HeavilyIndebted Poor Countries (HIPC) Initiativeand cancellation of official bilateral debt;and more generous ODA for countriescommitted to poverty reduction.

14. Address the special needs of landlockeddeveloping countries and small islanddeveloping states.

21. Prevalence and death rates associatedwith malaria.

22. Proportion of population in malaria riskareas using effective malaria preventionand treatment measures.

23. Prevalence and death rates associatedwith tuberculosis.

24. Proportion of tuberculosis cases detectedand cured under directly observed treatment short course (DOTS).

25. Proportion of land area covered by forest.26. Ratio of area protected to maintain

biological diversity to surface area.27. Energy use (kg oil equivalent) per US$1

GDP (PPP).28. Carbon dioxide emissions (per capita)

and consumption of ozone-depletingCFCs (ODP tons).

29. Proportion of population using solid fuels.30. Proportion of population with sustainable

access to an improved water source,urban and rural.

31. Proportion of population with access toimproved sanitation, urban and rural.

32. Proportion of households with access tosecure tenure (owned or rented).

Official development assistance (ODA)33. Net ODA, total and to LDCs, as percentage

of OECD/Development AssistanceCommittee (DAC) donors’ gross national income (GNI).

34. Proportion of total bilateral, sector-allocable ODA of OECD/DAC donors tobasic social services (basic education,primary health care, nutrition, safe waterand sanitation).

35. Proportion of bilateral ODA ofOECD/DAC donors that is untied.

36. ODA received in landlocked developingcountries as proportion of their GNIs.

37. ODA received in small island developingstates as proportion of their GNIs.

(continued )

Appendix 1 The Millennium Development Goals, Targets, and Indicators 83

Appendix 1 The Millennium Development Goals, Targets, and Indicators (Continued)


15. Deal comprehensively with the debtproblems of developing countriesthrough national and international measures in order to make debt sustainable in the long term.

16. In cooperation with developing countries,develop and implement strategies for decent and productive work for youth.

17. In cooperation with pharmaceutical companies, provide access to affordable,essential drugs in developing countries.

18. In cooperation with the private sector,make available the benefits of new technologies, especially information and communications.

Market access38. Proportion of total developed country

imports (by value and excluding arms)from developing countries and LDCs, admitted free of duties.

39. Average tariffs imposed by developedcountries on agricultural products andtextiles and clothing from developingcountries.

40. Agricultural support estimate for OECDcountries as percentage of their GDP.

41. Proportion of ODA provided to helpbuild trade capacity.

Debt sustainability42. Total number of countries that have

reached their HIPC decision points andnumber that have reached their HIPCcompletion points (cumulative).

43. Debt relief committed under HIPCInitiative, US$.

44. Debt service as a percentage of exports ofgoods and services.

45. Unemployment rate of 15–24 year-olds,each sex and total.

46. Proportion of population with access to affordable essential drugs on a sustainable basis.

47. Telephone lines and cellular subscribersper 100 population.

48. Personal computers in use per 100 population and Internet users per100 population.

84

LAC—Latin and Central America

1. Argentina2. Brazil3. Colombia4. Mexico5. Other Latin America: Antigua and Barbuda,

The Bahamas, Barbados, Belize, Bolivia, Chile,Costa Rica, Cuba, Dominica, The DominicanRepublic, Ecuador, El Salvador, French Guiana,Grenada, Guadeloupe, Guatemala, Guyana,Haiti, Honduras, Jamaica, Martinique, Nether-lands Antilles, Nicaragua, Panama, Paraguay,Peru, Saint Kitts and Nevis, Saint Lucia,Saint Vincent, Suriname, Trinidad and Tobago,Uruguay, and Venezuela

Sub-Saharan African

1. Central and Western Sub-Saharan Africa:Benin, Cameroon, Central African Republic,Comoros, Democratic Republic of Congo,Republic of Congo, Gabon, The Gambia,Ghana, Guinea, Guinea-Bissau, Côte d’Ivoire,Liberia, São Tomé and Principe, Senegal,Sierra Leone, and Togo

2. Eastern Sub-Saharan Africa: Burundi, Kenya,Rwanda, Tanzania, and Uganda

3. Nigeria4. Northern Sub-Saharan Africa: Burkina Faso,

Chad, Djibouti, Eritrea, Ethiopia, Mali, Mauri-tania, Niger, Somalia, and Sudan

5. Southern Sub-Saharan Africa: Angola, Botswana, Lesotho, Madagascar, Malawi,Mauritius, Mozambique, Namibia, Reunion,Swaziland, Zambia, and Zimbabwe

West Asia and North Africa (WANA)

1. Arab Republic of Egypt2. Turkey3. Other West Asian and North African countries:

Algeria, Cyprus, Islamic Republic of Iran, Iraq,Jordan, Kuwait, Lebanon, Libya, Morocco,Saudi Arabia, Syrian Arab Republic, Tunisia,United Arab Emirates, and Republic of Yemen

South Asian

1. Bangladesh2. India3. Pakistan4. Other South Asian countries: Afghanistan,

Maldives, Nepal, and Sri Lanka

Southeast Asia

1. Indonesia2. Malaysia3. Myanmar4. Philippines5. Thailand6. Vietnam7. Other Southeast Asian countries: Brunei,

Cambodia, and Lao People’s DemocraticRepublic

China: Includes Taiwan and Hong Kong

Appendix 2Impact-Water Developing

Countries and Regions

Appendix 2 Impact-Water Developing Countries and Regions 85

Appendix Table 2.1 Noncaloric Parameters of the Child Malnutrition Estimation Used for the Baseline and MDG Scenarios

Ratio of female-to-male lifeexpectancy at birth Share female secondary schooling

1995 2015 base 2015 MDG 1995 2015 base 2015 MDG

Latin America 1.0989 1.0989 1.0900 59.0 60.2 70.3Sub-Saharan Africa

Nigeria 1.0603 1.0690 1.0975 28.9 36.6 57.7Northern Sub-Saharan Africa 1.0527 1.0670 1.0920 10.6 11.8 26.4Central and Western Sub-Saharan Africa 1.0534 1.0600 1.0870 17.7 20.9 35.5Southern Sub-Saharan African 1.0498 1.0450 1.0750 20.6 30.3 41.2Eastern Sub-Saharan Africa 1.0335 1.0468 1.0650 11.6 15.6 28.9West Africa North Africa 1.0489 1.0498 1.0510 54.7 66.9 68.3

South AsiaIndia 1.0088 1.0580 1.0850 38.0 50.8 94.7Pakistan 1.0327 1.0388 1.0770 18.5 28.2 46.3Bangladesh 1.0072 1.0550 1.0870 13.4 21.1 33.5Other South Asia 1.0241 1.0380 1.0570 38.7 57.6 46.4

South East AsiaIndonesia 1.0580 1.0598 1.0770 47.6 56.9 71.4Thailand 1.0659 1.0772 1.0800 53.6 63.3 75.1Malaysia 1.0692 1.0699 1.0850 62.6 75.1 78.2Philippines 1.0639 1.0690 1.0746 77.8 88.4 95.0Vietnam 1.0687 1.0777 1.0793 46.0 52.8 92.0Myanmar 1.0849 1.0734 1.0988 33.1 41.0 82.7Other South East Asia 1.0536 1.0690 1.0650 21.4 43.1 53.6

China 1.0467 1.0498 1.0650 61.8 70.2 74.1

Notes: Latin America includes the Caribbean; N SSA stands for Northern Sub-Saharan Africa; C and W SSA stands for Central and WesternSub-Saharan Africa; S SSA stands for Southern Sub-Saharan Africa; and E SSA stands for Eastern Sub-Saharan Africa; WANA stands for WestAsia and North Africa; Other S Asia stands for Other South Asia; Other SE Asia stands for Other Southeast Asia. The regional disaggregation is presented in Annex 2. 2015 baseline values are interpolated from IFPRI IMPACT 2020 estimates. 2015 MDG scenario values are changed toattempt to reach the child malnutrition indicator of MDG 1.Sources: 1995 values for the ratio of female to male life expectancy and for female secondary schooling based on World DevelopmentIndicators (World Bank 2004c).

Documents

Agricultural Millenium Development Goals