sustainable development t.q8 114 3 work in … · the first draft, prepared in 1995 by derek...

_ E N V ENVIRONMENTALLY AND SOCIALLYSUSTAINABLE DEVELOPMENT

Rutral I)cvc/opmil,n

Work in progress 114 | 3for public discussion t.q8

Strengthening NationalAgricultural ResearchSystemsPol0/ Isscy uesadl Gooo( Pratire

r, E. . .

I)erek9/( WB,v,-/a'Alex1 .j/t\

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

ENVIRONMENTALLY AND SOCIALLYSUSTAINABLE DEVELOPMENT

,Rural Development

Strengthening NationalAgricultural ResearchSystemsPolicy Issues and Good Practice

Derek ByerleeGary E. Alex

The WorldBankWashington, D.C.

Copyright (C 1998The International Bank for Reconstructionand Development/THE WORLD BANK1818 H Street, N.W.Washington, D.C. 20433, U.S.A.

All rights reservedManufactured in the United States of AmericaFirst printing February 1998

This report has been prepared by the staff of the World Bank. The judgments expressed do notnecessarily reflect the views of the Board of Executive Directors or the governments they represent.

Cover: Pigeonpea farmers near Makindu, Kenya, learn that they have much to gain from adoptingextra-short-duration varieties. Photo from the International Crops Research Institute for the Semni-AridTropics (ICRISAT).

Derek Byerlee is principal economist in the Rural Development Family of the Environmentally andSocially Sustainable Development Network in the World Bank. Gary E. Alex is a consultant to theAgricultural Research and Extension Group (ESDAR), also in the Rural Development Family.

Library of Congress Cataloging-in-Publication Data

Byerlee, DerekStrengthening national agricultural research systems: policy

issues and good practice / Derek Byerlee and Gary Alex.p. cm. - (Environmentally and socially sustainable

development series. Rural development)Includes bibliographical references.ISBN 0-8213-4173-11. National agricultural research systems-Developing countries.

2. Agriculture-Research-Government policy-Developing countries.3. World Bank. I. Alex, Gary E. II. Series.S542.3.B94 1998630'.7'201724-dc2l 97-51583

CIP

®3 The text and the cover are printed on recycled paper, with a flood aqueous coating on the cover.

Contents

Acknowledgments vi

Foreword vii

Abbreviations and Acronyms ix

Executive Summary 1

Introduction and Objectives 9

Chapter 1 Recent Evolution of National Agricultural Research Systems 11Investment in National Agricultural Research Systems 11Payoffs to Research Investments 13The Changing Agricultural Research Agenda 14Emerging Institutional Challenges in National Agricultural Research Systems 16

Chapter 2 National Research System Development: Issues and Good Practice 18Conceptual Overview of the Emerging National Agricultural Research Systems 18Coordination in the Emerging National Agricultural Research Systems 21Competitive Funding of Research 22Integrating Universities in the NARS 24Increasing the Role of the Private Sector 25Linking the National Agricultural Research Systems to Regional and International

Research Systems 28

Chapter 3 Strengthening Public Research Institutes: Issues and Good Practice 33Responding to the Funding Crisis 33Organizing National Agricultural Research Institutes for Increased Efficiency 39Institutionalizing Research Planning and Evaluation 42Upgrading Human Resources Management 47Improving Client Orientation and Technology Transfer in National Agricultural

Research Institutes 50

Chapter 4 The World Bank's Role in Strengthening National Research Systems 57Evolution of Bank Support to Research 57Key Elements of Current Bank Strategy 61

iii

iv Strengthening National Agricultural Research Systems

Supporting New Research Priorities 64Implications for the Bank 66

Annexes1 Main Recommendations from Reviews of World Bank Research Projects 722 World Bank Funding for Agricultural Research by Year

and by Geographic Region 763 Illustrative Project Performance Indicators for Agricultural Research Projects 78

Bibliography 83

Boxes1.1 Crop improvement research: An institutionalized success story 141.2 Crop and resource management research: A continuing challenge 151.3 The diversity of National Agricultural Research Systems 172.1 Funding versus execution of research in Australia 202.2 Building consensus: The importance of a strategic vision 232.3 Why have universities been underutilized in national agricultural

research systems? 252.4 Research foundations: A diversity of approaches 272.5 Technology spill-ins: Spreading research benefits beyond borders 292.6 Collaborative partnerships between IARCs and NARSs 302.7 Latin American Regional Research Fund 313.1 Funding reliability in the short term 343.2 Five key questions to ask on cost recovery in public research institutes 373.3 Why is research different from many public sector activities? 393.4 Agricultural research and poverty reduction 473.5 Farmer-run research: Experience with the Comite de Investigaci6n

Agropecuaria Local 523.6 Links to farmers' organizations 523.7 Improved research-extension linkages: The case of Ghana 533.8 Serving the "policy client" 554.1 Major recommendations of the Operations and Evaluation Department review

of agricultural research projects 604.2 Problems identified in a 1997 review of agricultural research projects 614.3 An emerging challenge: The national agricultural research systems of Eastern

Europe and Central Asia 644.4 Global Forum on Agricultural Research 644.5 Geographic information systems: A potentially powerful new tool

for agricultural research 66

Tables1.1 Number of scientists employed by the public sector of national agricultural

research systems 121.2 Public agricultural research expenditures: Global trends 121.3 Intensity of public sector investment in agricultural research 133.1 Definition of "autonomy" in terms of the types of flexibility required

for various management functions 403.2 Summary of monitoring and evaluation indicators and their implementation

in agricultural research 46

Contents v

4.1 Summary data on World Bank agricultural research projects 584.2 Regional share of World Bank loans for agricultural research 594.3 List of operational strategies identified in 1980 World Bank policy guidelines 604.4 Trends in institutional issues emphasized in the current World Bank portfolio

of research projects 61

Figures2.1 Traditional view of a national agricultural research system 192.2 New perspective on national agricultural research systems 194.1 Percentage of World Bank agricultural loans for research and extension 584.2 Geographic distribution of World Bank agricultural research funding

between 1981 and 1996 59

Acknowledgments

P reparation of this report was coordi- ongoing agricultural research and extensionnated by the Rural Development Family projects.(RDV) and the Agricultural Research As a result of these discussions, this report

and Extension Group (ESDAR) of the Environ- was extensively rewritten. This version bene-mentally and Socially Sustainable Develop- fited from the comments of many personsment Network of the World Bank. The report inside and outside the Bank. Special thanksrepresents the input of Bank staff and review- are due to Trevor Sykes, Alex McCalla, Jockers. The first draft, prepared in 1995 by Derek Anderson, Carl Eicher, Cornelis De Haan,ByerleewithcontributionsfromJockAnderson, Tjaart Schillhorn-van Veen, Michel Petit,resulted from discussions with colleagues Johannes Ter Vrugt, Jacob Kampen, Gunterin RDV and task managers of agricultural Steinacker, John Nickel, Moctar Toure, Umaresearch projects in regional operations. Lele, Hawanty Page, Andreas Springer-These discussions continued during 1995 and Heinze, and Willem Zijp.1996 through a workshop on financing agricul- The authors also wish to acknowledge thetural research in Latin America, several meet- editorial contributions of Donna Allen, Leoings of agricultural research task managers Demesmaker, Alicia Hetzner, and Virginiaheld in preparing the Agricultural and Rural Hitchcock. Gaudencio Dizon desktopped thisSector Action Plan, and a portfolio review of report.

vi

Foreword

T he financing of agricultural research is The report contains four parts. Chapter 1one of the most important components provides a brief overview of the recent evolu-of the World Bank's rural sector portfo- tion of national research systems. Chapters 2

lio. Support to the research subsector has seen and 3 are the main body of the report. Chaptera sharp change in emphasis in the 1990s toward 2 synthesizes policy issues and good practicessupporting institutional reform and pluralism for developing national agricultural researchin funding and executing research and toward systems within an emerging global agriculturaldeveloping sustainable funding mechanisms research system. While there is growing par-for public sector research. In light of these ticipation from diverse partners in researchchanges and in response to a recent compre- funding and execution, the public sector willhensive review of the subsector by the Opera- continue to be central in undertaking researchtions Evaluation Department (OED), the Bank on the emerging challenges of sustainable ag-has been developing new forms of support to ricultural intensification, poverty alleviation,agricultural research. and conservation of natural resources. Chapter

This report has two objectives. First, it pro- 3 focuses on the key policy and institutionalvides a brief review of recent trends and key reforms needed to strengthen these public sec-policy issues in strengthening national agricul- tor research institutions. The report providestural research systems-broadly defined to in- the underlying rationale for selecting goodclude national research institutes, universities, practices and discusses their applicability inthe private sector, and nongovernmental specific situations. Chapter 4 discusses impli-organizations. Second, it synthesizes "good cations for the Bank in its ongoing efforts topractice" in ongoing institutional and policy strengthen national research systems.reforms in the subsector. The report explores Bank support for agricultural research will beagricultural research policy issues and pro- central to its objectives of alleviating povertyvides a resource on selected research policy and conserving natural resources. However, ag-and management issues for Bank staff and ricultural research policy and best practices willpartners in borrowing countries. Most of the continue to evolve in response to the changingpractices presented here have been incorpo- roles of the public and private sectors, new in-rated in recent Bank projects in various forms. stitutional mechanisms for funding and execut-For complex issues, such as agricultural re- ing research in the public sector, and changingsearch policy, best practices are often very situ- demands on research systems. Through its re-ation specific. cently formed thematic team on agricultural

vii

Foreword viii

knowledge and information systems, the perience in research projects, and modify bestWorld Bank will closely monitor ongoing ex- practices in light of these experiences.

Alex F. McCallaDirector, Rural Development Family

Michel PetitDirector, Agricultural Research

and Extension Group

Abbreviations and Acronyms

AFR Africa Region of the World GIS Geographic informationBank system

AgGDP Agricultural gross domestic GREAN Global Research on theproduct Environment and

ARF Agricultural research fund Agricultural NexusARI Agricultural research intensity program proposalASARECA Association for Strengthening IARC International agricultural

Agricultural Research in research centerEastern and Central Africa IDB Inter-American Development

CAS Country assistance strategy BankCEO Chief executive officer IFPRI International Food PolicyCG Competitive grant Research InstituteCGIAR Consultative Group on Interna- IPM Integrated pest management

tional Agricultural Research IPR Intellectual property rightsCIAL Local agricultural research ISNAR International Service for

committee (Comite de Investi- National Agriculturalgaci6n Agropecuaria Local) Research

EAP East Asia and Pacific Region LAC Latin America and theECA Europe and Central Asia Caribbean Region of the

Region of the World Bank World BankEDI Economic Development M&E Monitoring and evaluation

Institute of the World Bank MNA Middle East and North AfricaESDAR Agricultural Research and Region of the World Bank

Extension Group of the NARI National agricultural researchWorld Bank institute

FAO Food and Agriculture NARS National agricultural researchOrganization system

FFA Framework for Action NGO Nongovernmental organizationFSR Farming systems research NRM Natural resource managementFUNDAGRO Foundation for Agricultural NRMR Natural resource management

Development of Ecuador researchFUNDEAGRO Foundation for Agricultural OED Operations Evaluation

Development of Peru DepartmentGDP Gross domestic product PRA Participatory rural appraisal

ix

Abbreviations and Acronyms x

PROCI Program of Regional R&D Research and develop-Cooperation ment

PROCIANDINO Andean Program of SAS South Asia Region of theRegional Cooperation World Bank

RARA Regional agricultural SPAAR Special Program forresearch association African Agricultural

RELC Research-Extension Liaison ResearchCommittee TFP Total factor productivity

RDC Research and Development UNEP United Nations Environ-Corporations of Australia ment Programme

RDV Rural Development Family USAID U. S. Agency for Inter-(formerly Agriculture and national DevelopmentNatural Resources Depart- WANA West Asia and Northment, or AGR) of the Africa Region of theWorld Bank World Bank

Executive Summary

T his report reviews key issues facing re- research policy approved by the Bank in 1980,search managers, policy makers, and that emphasized quantitative targets for ex-the development assistance community pansion of research systems.

in positioning national agricultural research The paper discusses current issues in the sus-systems to address the challenges of the next tainable development of NARSs and identifiesdecade. The review is motivated by the sharp "good practice" for selected issues in agricul-fall in investment in agricultural research in tural research policy and management. Goodmany countries in the 1990s, following a period practice in agricultural research varies widelyof rapid growth since 1970, combined with across countries depending on the stage of de-management and organizational problems in velopment of the NARS, institutional history,many research systems that challenge the sus- the type of agriculture, and the size and politi-tainability of agricultural research capacity. cal organization of the country. The paper

Strengthening agricultural research systems (1) reviews recent changes in NARSs capacitywill be central to current World Bank efforts to and performance; (2) examines organizationalrevitalize agriculture and rural development and conceptual challenges and good practiceover the next decade. Sustainable intensifica- in moving from a focus on public research in-tion of agricultural systems through rapid and stitutes to emphasizing the development of na-broad-based technological change is critical tional research systems that recognize key rolesto solving global problems of food security, for universities and the private sector; (3) ex-poverty alleviation, and conservation of natu- amines key funding, institutional, organiza-ral resources. tional, and management issues and good

The World Bank has been an active partici- practice in reforming public research institutes;pant in the growth of national agricultural re- and (4) discusses the consequent implica-search systems (NARSs), lending nearly $4 tions for World Bank support to agriculturalbillion for their support since 1981. The empha- research.sis in these projects has steadily evolved fromone of investment in research infrastructure Fostering the Evolution of Nationaland human resources development to improv- Research Systemsing research management and institutionbuilding within a pluralistic national research NARSs have evolved rapidly over the pastsystem. Recent Bank lending programs have three decades, especially in the quantitative di-often included downsizing and consolidation. mension. In aggregate, public research systemsThis is a sharp departure from the agricultural in the developing world employ over 100,000

1

2 Strengthening National Agrictultutral Research Systems

scientists and manage an annual investment of tion to developing national agricultural re-over $8 billion. This investment now exceeds search institutes (NARIs) to strengthening na-that in industrialized countries. But over the tional research systems, broadly defined topast decade, public research systems almost include the NARI, universities and the privateeverywhere have faced a crisis of confidence sector (both for profit and nonprofit). The chal-reflected in stagnant or reduced funding and lenge is to develop a well-articulated researchsevere restrictions on operating costs. The pri- system to meet national objectives for the sectorvate sector has potential to fill some of the gap by developing innovative institutional modelsand has expanded rapidly from a very small that encourage participation of alternative re-base to account for about 10-15 percent of research funders and suppliers, bringing more re-sources invested in agricultural research in de- sources into the research system, and exploitingveloping countries. complementarities among various participants

at the national and international levels.The Changing National AgriculturalResearch Systems Environment in the 1990s Key Elements of the New National

Agricultural Research Systems ParadigmIn recent years the economic and technical en-vironment for agricultural research has also The emerging concept of a NARSs is charac-changed drastically. Trade liberalization has terized by seven key elements:shifted economic incentives and improved A conceptual and increasingly, institution-technology is now a critical factor in a coun- alized separation of researchfunding fromtry's ability to exploit its comparative and com- research execution, each of which requirespetitive advantages. Technological needs are distinct skills and inputs. Research fund-changing toward more knowledge- and skill- ing and the setting of broad priorities forintensive agriculture and more complex farm- research is a policy issue, while researching systems. Important changes in the execution by alternative suppliers is an ef-technology for research itself, especially the ficiency issue.new biotechnologies and informational tech- . A recognition of the pluralistic instituttionalnologies, are raising new issues in organizing structure of a NARS that includes universi-national research systems, related to econo- ties, the private sector, farmer organiza-mies of size, international collaboration, and tions, and nongovernmental organizationspublic-private linkages. (NGOs), that allows additional scientific

These increasing demands on research sys- skills to be tapped and matched withtems contrast with the lack of institutional in- needs, thus increasing research efficiency.novation in research-system management and * A sharper focus of public funding for re-organization and the stagnation or decline in search on public goods and diversification offunding for agricultural research. The substan- funding support for public research insti-tial investment in NARSs over the past two tutes. There is a strong case for public-sec-decades, although providing high payoffs, has tor funding of basic and strategic researchnot yet resulted in the institutional capacity to (long-term research with uncertain payoffssustain those payoffs in a rapidly evolving and high spillovers), research on problemstechnical and policy environment. Many sys- of small-farm agriculture (high transac-tems are suffering a crisis of management, with tions costs for farmers to organize theirtop-heavy bureaucracy, centralization of deci- own research), and research on natural re-sionmaking, and lack of incentives for the in- source management (positive environ-novation process so essential for research. mental externalities).

Emphasis on building agricultural research * A recognition of the complementary rolescapacity has now shifted from exclusive atten- of public and private sector research and

Exectitive Suimmary 3

development (R&D) and the potential effi- and private sectors and their interaction; theciency gains through private-public sector comparative advantage and mandates for cen-collaboration in both the funding and execu- tral, state, and local research institutes; sustain-tion of research due to the complementary able funding mechanisms; and linkages withskills and resources in each sector. clients, technology transfer agencies, and theIncreased flexibility and institutional auton- global agricultural research system. Such a vi-omy in public research institutions com- sion will require considerable dialogue amongbined with increased accountability to all stakeholders to ensure broad consensus onpromote results-oriented institutions, the future direction of the NARS. This visioncharacterized by business-oriented man- may be developed by a small apex body thatagement of human, financial, and physical is broadly representative of key stakeholdersresources. and that can act as a national forum for devel-

. The involvement of stakeholders, especially oping and coordinating research strategies,the clients of the research system, in re- formulating policies, and financing of research.search governance, priority setting, execu- There are many other less formal institutionaltion, financing, and evaluation so that mechanisms for system coordination, and ininstitutions within the NARSs become large countries several mechanisms may bemore responsive to their needs. The insti- needed.tutional models for achieving this in- Second, all NARSs will have to seek meansclude full or partial funding of research by to increase private-sector investment in R&Dfarmers and other clients, involvement of and enhance public-private sector complemen-farmers and farmers' organizations in gov- tarity and partnerships. A variety of mecha-ernance of research organizations, and nisms can be employed, including investmentvarious types of contractual relationships in strong basic and strategic research programswith clients in executing research. in the public sector, implementing institutional

* New modelsfor technology transfer that move mechanisms and legal frameworks for public-beyond the traditional research-extension private joint ventures, strengthening and en-chain to involve farmers, NGOs, and the forcing intellectual property rights, andprivate sector in a variety of formal and relaxing regulations on approval and releaseinformal partnerships, and information of new technologies and on importing technol-dissemination and feedback mechanisms. ogy (but ensuring that key public health and

national food security concerns are protected).Good Practice in the Sustainable In some situations, private foundations canEvolution of National Agricultural successfully support agricultural research, es-Research Systems pecially if they are well endowed and have

strong local support.In the new way of thinking, NARS are almost Third, universities are a valuable but under-invariably complex with wider and more di- utilized resource that have much potential toverse institutional participation in both re- contribute to the evolution of a strong NARS.search funding and execution. Such complexity This will require policy and structural reformsrequires increased attention to strengthening such as competitive funds, and collaborativecoordination and interaction within the research networks to expand university in-system. volvement in research. Universities themselves

First, each NARS needs to develop a broad will have to adopt policies and develop capa-"vision" of the future evolution of the research bilities to promote and conduct research. Insystem over the next 10-20 years that suc- addition, upgrading of universities will becinctly outlines the changing demands for tech- critical to improving the human resource ca-nology; the expected future roles of the public pacity for scientific research at all levels of the

4 Strengthening National Agricultural Research Systems

NARS. Strengthening of universities in both share of global human and financial resourcesresearch capacity and human resource devel- for agricultural science, will be key to strength-opment is, therefore, integral to strengthening ening most agricultural research systems. TheNARS. pressure on research funding over the past dec-

Fourth, governments can shift more of their ade and erosion of confidence in public re-funding through competitive grants in order to search organizations is a serious threat to themobilize underutilized resources, promote re- viability of many national agricultural researchsearch partnerships, improve research quality, institutes (NARIs). Institutional reform of pub-and move research resources toward agreed lic research institutes is needed to improvenational priorities. Competitive grant systems their efficiency anc effectiveness throughwork best in institutional structures where streamlining research priorities, reformingthere is substantial management flexibility and management and incentive systems, and build-independence from political and bureaucratic ing links to external partners.interference. However, competitive grants can-not, and should not, substitute for institutional Developing Sustainable Fundingdevelopment and longer-term investments indeveloping research infrastructure, nor are Most NARIs depend largely on public fundingthey as relevant to small NARSs, where higher and/or donor support, both of which are sub-administrative costs and limited potential for ject to long-term erosion and uncertainty. Thiscompetition reduces their value. requires two responses. First, NARIs must

Finally, all NARSs will have to integrate with build political support for public funding ofthe emerging global agricultural research research by increasing public awareness- atsystem (composed of international centers, re- all levels-on the role and impacts of research,gional centers and networks, large multina- developing strong and articulate client organi-tional R&D programs, and a diverse array of 'zations that can act as a lobby for agriculturalNARSs) to keep abreast of rapid advances in research, and by reforming the managementscientific knowledge, and to improve the cost and effectiveness of NARIs to make them moreeffectiveness of technology generation by cap- attractive investments.turing "spill-ins" and through collaborative re- Second, all public research organizationssearch. A variety of institutional mechanisms, will need to diversify funding from their al-such as international and regional networks most complete dependence on governmentand collaborative research programs with in- budgetary appropriations. One approach isternational centers and advanced research in- through implementing cost recovery for somestitutes, can be employed. Regional research products and services, such as royalties on re-initiatives, including regional networks, joint search products, user fees for nonresearchresearch initiatives, and the establishment of products and services, and joint ventures withregional research funds, will also play a role, the private sector. Although this commerciali-provided they are cost effective and are NARS- zation of research products and services candriven and funded. provide important contributions to operating

costs , it should be approached with caution,Increasing the Efficiency as it can quickly distort program priorities andand Effectiveness of Public attention. Public sector institutes require clearResearch Institutions and transparent policies for ensuring that com-

mercialization is consistent with the public in-Although NARSs will become more institu- terest, applying intellectual property rights totionally diversified in both the funding and protect technologies, providing access to pro-execution of research, public research institutes tected technologies for research purposes,(the NARIs), which command a considerable and sharing the revenues generated. Special

Execuitive Summary 5

caution is needed in commercializing nonre- research (with substantial economies of size),search products and services to ensure that this and management of national research net-is complementary, not competitive, to the main works (to promote spillovers), while much ofbusiness of research. the applied and adaptive research will be

NARIs can also diversify funding by encour- decentralized to administratively independentaging, where feasible, farmer financing of re- regional research centers, in which farmerssearch through levies, together with are involved in funding, priority setting, andmechanisms to ensure farmer participation in govemance.setting priorities for the research that they Finally, efficiency of most research organiza-fund. This approach will work best if farmer tions can be enhanced through cutting redundantcontributions are matched by government and nonperforming staff, and consolidatingfunds as an incentive for farmers to fund research programs and infrastructure. Suchlonger-term research, or research with broader downsizing should be as transparent as possi-societal benefits. ble and should be conducted within a well-

defined NARI strategic plan. It is critical thatOrganizing Public Research Institutions research managers are in agreement from the

highest levels that budget savings from down-Most public research organizations are part of, sizing and consolidation will be retained by theor closely associated, with government minis- research institute to improve operating bud-tries. Excessive political and bureaucratic inter- gets and salaries of remaining staff.ference in their management is one reason forlack of effectiveness of some NARIs. Institu- Strengthening Research Managementtional reform requires that NARIs be providedgreater autonomy (up to the creation of inde- Systematic research priority setting and moni-pendent research corporations) to give them toring and evaluation of research programs arethe flexibility to adopt business practices for important elements in good research manage-financial and personnel management (includ- ment. In contrast with past emphasis on one-ing setting salary levels) more conducive for off development of research master plans withattracting high-quality scientists and utilizing outside technical assistance, future emphasisthem effectively. Success in providing auton- should be on developing analytical capacityomy will depend on (1) the appointment of within the research system to undertake re-members of governing boards in their individ- search priority setting and research policyual professional capacity, (2) the involvement analysis, together with strong commitment byof diverse stakeholders in governance, (3) ob- research managers to ensure that research re-taining expert advice on public sector reform .3ources shift in ways consistent with identifiedto define rules and regulations for managing priorities. The processes for priority setting arefinancial and human resources and assets, arnd as important as the analytical methods used,(4) developing performance indicators to en- especially the use of participatory processessure accountability to research funders. that involve a wide cross section of scientists,

Many NARIs have highly centralized organ- and the main stakeholders -policymakers, re-izational structures. Each NARI will have to search partners, and clients.rationalize the level of centralization in re- Effective monitoring and evaluation (M&E)search funding and execution according to the of agricultural research programs and projectspotential for technology spillovers, economies requires simple nonbureaucratic systems thatof size in research organizations, and the need provide timely feed back to research managers.to build local political support for research Monitoring and Evaluation (M&E) shouldfunding. Centralized NARIs will probably be- be implemented in a highly decentralizedcome more specialized in basic and strategic manner, although in large organizations, a

6 Strengthening National Agricultutral Research Systems

very small central unit can be established to Role of the Bank to 2000 and Beyondpromote M&E, provide training, and developstandards. Since 1981 support to NARSs has been a prior-

ity in the Bank's lending portfolio. The BankContinuing Effort on Human has financed 458 projects with agricultural re-Resources Development search components in 91 countries, providing

a total of $3.9 billion for support to research.The quality of agricultural research is only as Development and dissemination of improvedgood as the quality of the scientific human re- agricultural technologies is central to agricul-source base. This implies continuing effort to re- tural and rural development strategies and willvamp performance appraisal and promotion be critical to meeting the challenge laid out insystems, train research managers, decentralize the Bank's action plan for the rural sector. Thusproject management and budgets to individual it is vitally important that the Bank maintainscientists, and provide funding for sabbaticals, its strong commitment to the development ofexternal travel and greater interaction with the national agricultural research systems.international scientific community. Special atten- The focus on strengthening national researchtion is needed to mentor the professional devel- systems and the rapidly changing environmentopment of young scientists. Training will in which those systems operate, have a numbercontinue to be a priority for NARIs for replace- of implications for Bank support to agriculturalment staff and for upgrading skills to enable re- research.searchers to stay current in their disciplines andin emerging fields such as biotechnology, com- Policy Dialogue on National Agriculturalmunications technologies, and systems model- Research System Strategiesing. The further development of in-country (orregional) postgraduate training capacities will Creating and maintaining policy dialogue withbe essential to system sustainability. borrowers on constraints, opportunities, and

strategies for national agricultural research sys-Forging Stronger Links with Clients tems and sustainable funding, based on economic

and sector work, including the country assis-All public sector organizations must develop tance strategy, will be a key part of efforts tostrong linkages with other research organiza- strengthen NARS. This process must be highlytions, technology transfer institutions, and es- participatory to build borrower consensus onpecially with their clients. Research institutions feasible policy options and institutional reforms,must be increasingly demand-driven through and a policy environment conducive to private-the involvement of farmers in setting the sector investment in R&D and technology transfer.research agenda and in research financing.Farmers and farmers' organizations will take Changing Priorities for Project Lendingincreasing responsibility for on-farm adaptiveresearch. The clientele of research organiza- Project lending will shift in accord with thetions will also become more diverse with a new priorities and will likely include the fol-growing role for agribusiness and consum- lowing elements:ers in setting the research agenda. Finally * Supporting policy and institutional re-research-extension links need to be broadened forms to improve the efficiency and effec-to include other mechanisms for technology tiveness of public research institutestransfer and information dissemination and * Building mechanisms for sustainablefeedback through farmer organizations, di- funding, including cost recovery pro-verse types of NGOs, and the private sector. grams, farmer financing, joint ventures,

Executive Summary 7

and establishing research foundations and Finally, in almost all NARS there is a need toresearch corporations strengthen social-science research, which is

* Providing and maintaining critical equip- central to the new priority research areas andment and materials, investing in scientific to improved research management.information systems, and providing train-ing and technical assistance in new re- Emphasis on Qualitysearch areas (for example, biotechnology) and Knowledge Management

. Increasing investments in universities tosupport both research and the upgrading These changes imply that future Bank projectsof educational quality in the agricultural in agricultural research will probably besciences smaller, but more complex as well as knowl-

* Supporting programs to facilitate regional edge- and management-intensive in prepara-and international collaboration and staff tion and supervision. Thus a decrease inexchanges, and technology spill-ins project size (and possibly lending volume) will

. Supporting the development of farmer or- not likely be accompanied by a decrease inganizations and mechanisms by which Bank staff input. Many activities that stress in-they can influence the research agenda, stitutional change will have to be piloted be-including participation in cost-sharing ar- fore expansion. Developing partnerships withrangements to fund agreed-on research other donors will be an important part of apriorities. strategy to bring increased technical and re-

search management skills to project prepara-Supporting Priority Research Areas tion and supervision.

The synthesis and application of good prac-Future projects will also support changing re- tice in agricultural research will require con-search priorities. A first priority in most project tinuing dialogue within the Bank and withlending will be to support shifts in research country partners to identify practices most ap-paradigms toward sustainable intensification propriate to a given country situation. Givenof agricultural systems in both favored and the rapid evolution of good practice in supportless-favored areas. This will require support to to agricultural research, the Bank must be in aboth strategic and applied research on knowl- strong position internally and through partner-edge-intensive practices, such as integrated ships, to evaluate and synthesize ongoingpest and nutrient and water management to experiences with new initiatives, such as com-promote input efficiency. Sustainable intensifi- petitive grants, autonomous research corpora-cation will also require more emphasis on tions and foundations, farmer financing, jointnatural resources management research, in- ventures with the private sector, and links tocluding organizational forms for conducting NGOs. The formation of Bank networks andthe research, developing links with new sets of families can provide a vehicle for both synthe-stakeholders, and initiating research on associ- sis and dissemination of good practice, includ-ated institutional and policy issues. ing arrangements for systematically tracking

Biotechnology research will also be a priority change and monitoring new developments.for support. This should include support to the The dissemination of good practice within andestablishment of strategies, policies and asso- across NARSs through case studies, seminars,ciated public safety and intellectual property and study tours, should be a priority in build-rights issues to promote biotechnology re- ing borrower commitment to institutionalsearch, as well as development of external link- and policy changes, and contributing to aages with advanced research centers and the global knowledge base on agricultural researchprivate sector to gain access to the technology. policy.

Introduction and Objectives

J nvestment by both governments and do- opment projects in which agricultural researchnors in agricultural research in developing was a component. The emphasis in thesecountries grew rapidly from about 1970. projects can be broadly classified into three

This was in part stimulated by the successes of periods:the Green Revolution, in which broad-based 1. A period of expansion up to the early 1980s-technical change in agriculture was becoming the "bricks and mortar" phase-when theever more apparent as an engine of overall eco- main emphasis was on increasing the size ofnomic growth. At the same time, the high republic-sector research organizations throughturns to investment in agricultural research investment in experiment station and labora-were noted in the writings of influential econo- tory infrastructure, equipment, and human re-mists such as Schultz (1964) and Hayami and source development.Ruttan (1975). Overall, investment in agricul- 2. A period of transition from the mid-1980stural research in developing countries grew at when more emphasis was placed on im-6 percent annually from 1961 to 1985 (Ander- proving the management of existing re-son, Pardey, and Roseboom 1994). search resources in the public-sector

Then, in the mid-1980s, rapid growth in na- research institutes through better planning,tional agricultural research systems (NARSs) improved financial management, greater ac-began to slow. From 1981 to 1991 the growth countability, and attention to increasing therate of public research expenditures in devel- relevance of the research program to its im-oping countries slowed to 3.8 percent per year mediate clients, the farmers. However, as in(Alston, Pardey, and Roseboom 1997). Since the first period, most resources in project1985 there has been a sharp drop in funding in loans for agricultural research were pro-many countries, especially in Latin America vided for further expansion and rehabilita-and Africa. Meanwhile, the size of research sys- tion of research infrastructure.tems, measured by the number of scientists, 3. The period from the mid-1990s when Bankcontinued to expand, resulting in reduced ex- projects began to emphasize measures to en-penditures per scientist and a critical shortage hance the institutional sustainability of agri-of operating funds for research. cultural research systems, defined to include

The World Bank has provided substantial all actual and potential participants in thesupport for the growth of NARSs. Since 1981 technology-generation process, such as uni-the Bank has lent nearly $3.9 billion for agricul- versities, the private sector, research founda-tural research through specialized agricultural tions, and some rural-based NGOs, as wellresearch loans and loans for broader devel- as the governmental national agricultural

9


research institutes (NARIs) targeted earlier. these systems through policy and institutionalIn this period, there has been little emphasis reforms. Such reforms will be critical to ad-on system expansion: in many cases the dressing the global problems of food security,Bank has supported downsizing and con- poverty alleviation, and conservation of natu-solidation of public research institutions. ral resources as highlighted in the 1996 WorldEmphasis has now moved from expanding Food Summit (FAO 1996). Food production in

the NARIs toward building a more diverse developing countries will have to almost dou-NARS that incorporates a range of institutional ble by 2025, and much of this increase will haveoptions for conducting agricultural research to come from biological yield increases in de-and development (R&D) and a diversity of veloping countries (McCalla 1994). Sustainablefunding mechanisms that foster competition development of agricultural research systemsand improved articulation among the various will be central to current World Bank efforts toparticipants in the expanded system (McMahon revitalize agriculture and rural development1992; Echeverria and others 1996). These recent over the next decade.developments represent a sharp departure The primary purpose of this paper is to dis-from the agricultural research policy approved cuss key strategic issues in the sustainable de-by the Bank in 1980 (World Bank 1980). They velopment of NARSs and to identify, wherealso increasingly recognize that the appropriate possible, "good practices" in agricultural re-focus is the "agricultural knowledge system" or search policy. The paper (1) provides a review"technology innovation system," terms which of recent changes in NARSs capacity and per-explicitly recognize that extension and educa- formance (chapter 1), (2) examines organiza-tional systems and user involvement are asso- tional and conceptual challenges in movingciated with effective research institutions. from a focus on public research institutes

Over the past two decades the environment (NARIs) to a focus on research systems (NARSs)for agricultural research has changed in other (chapter 2), (3) examines key funding and in-ways as well, especially with the introduction stitutional issues facing public sector researchof policy reforms and more liberalized trade in programs (chapter 3), and (4) discusses impli-the 1980s. Some of these reforms have promoted cations for the World Bank for dealing withprivate sector participation in R&D. However, these emerging issues (chapter 4). We recog-the structural adjustment process, which stimu- nize that the issues are necessarily discussedlated institutional reform in many public sector in summary form and cover only selected as-enterprises, hardly touched the public sector pects of agricultural research policy. DespiteNARIs, except through reduced budget alloca- many commonalties in NARSs problems andtions, which accompanied the general trimming their potential solutions, there is no single modelof government budget deficits. of institutional development for NARSs. Best

It is appropriate now to examine agricultural practices are country- and time-specific. Accord-research policy in light of the evolving de- ingly, we identify some general principles formands on research systems and the potential designing best practices and identify situationsto introduce substantial efficiency gains in under which specific practices may apply.

CHAPTER 1

Recent Evolution of National AgriculturalResearch Systems

N ational agricultural research systems agricultural research, a level 15 percentare facing a range of new challenges higher than the comparable aggregate in-

= that will shape their future develop- vestment in industrialized countries (tablement. Though investment in national systems 1.2). In contrast, in 1971 developing coun-is substantial and this investment has generally try investment in agricultural research wasprovided good returns, the NARSs are now only 70 percent of that of developed coun-confronted by a more complex and demanding tries. From 1981 to 1991 the annual rate ofresearch agenda and must deal with second- growth in research expenditures for devel-generation issues in institutional development. oping countries (3.8 percent) continued to

exceed that of developed countries (1.7Investment in National Agricultural percent) (table 1.2).Research Systems * By the late 1980s the period of rapid expan-

sion of NARSs was ending and a slow-From Pardey, Roseboom, and Anderson (1991), down in spending on research had becomeAlston, Pardey, and Roseboom (1997), and oth- a worldwide phenomenon (table 1.2). Theers, a number of broad quantitative generali- most severe effects were felt in Latin Amer-zations can be made about investment in ica, Sub-Saharan Africa, and the formerpublic-sector agricultural research systems in Soviet Union. The least affected were thedeveloping countries over the past three dec- Asian NARSs where research expendi-ades: tures continued to expand rapidly. Almost

* By the 1980s more than 80,000 agricultural everywhere the number of scientists ex-researchers were working in the NARSs of panded more rapidly than total researchthe developing world (table 1.1). (Recent expenditures, resulting in a declining ex-estimates put the number of researchers in penditure per researcher. This is reflecteddeveloping countries at well over 100,000). in reduced operating budgets and in lowerDuring the period 1961-85 the average size real salaries for scientists.of a developing country NARS increased . There is wide diversity across countries infrom 155 to 630 full time equivalent re- the size and maturity of NARSs. China andsearchers. India have the largest research systems in

* By 1991 developing countries were invest- the developing world, with tens of thou-ing US$8,000 million annually in public sands of scientists, and account for more

11


Table 1.1 Number of scientists employed by the public sector of national agricultural research systems

Agricultural research personnel (full-time equivalents) Growth rate

Region 1961-65 1971-75 1981-85 (percent per year)

Developing countries (130)a 20,256 37,221 81,848 7.2Sub-Saharan Africa (43)a 1,323 2,416 4,941 6.8China 7,469 11,781 36,335 8.2Asia and the Pacific (28)a 6,641 12,439 22,576 6.3Latin America and the Caribbean (38)a 2,666 5,840 9,000 6.3West Asia and North Africa (20)a 2,157 4,746 8,995 7.4

tndustrial countries (22)a 40,395 48,123 56,376 1.7

Total (152t 60,651 85,344 138,224 4.2

a. Bracketed figures indicate the number of countries in the regional totals.Note: The growth rate represents the compound annual average growth rate between 196145 and 1981-85.Source: Anderson, Pardey, and Roseboom 1994.

than half of developing country research * The intensity of research investment iscapacity. But 95 of 130 NARSs in develop- generally lower for large NARSs and foring countries still employ fewer than 200 major food crops, such as rice and wheat,researchers, and 39 systems employ fewer reflecting economies of size and scope inthan 25 researchers. conducting many types of research (Byer-The intensity of investment by the public lee and Traxler 1996). Lack of these econo-sector in research in developing countries, mies of size and scope presents specialmeasured as a share of agricultural gross problems for small NARSs.domestic product, is about 0.5 percent (ta- * The quality of scientific manpower, asble 1.3). Although this has risen from 0.38 measured by the proportion of scientistsin the early 1970s, it is relatively low and with postgraduate degrees, has increasedhas remained nearly unchanged since the sharply in all countries, and especially inlate 1970s. This is far below the two percent Sub-Saharan Africa, which had very fewtarget recommended by the World Bank trained scientists in the 1960s. Currently(1981) and the average for industrialized more than half of the agricultural re-countries of 2.39 percent (Alston, Pardey, searchers in developing countries hold aand Roseboom 1997). postgraduate research degree. In spite of

Table 1.2 Public agricultural research expenditures: Global trends

Expenditures Annual growth rates(millions of 1985 intemational dollars) (percent per year)

Region 1971 1981 1991 1971-81 1981-91

Developing countries 2,985 5,535 8,017 6.4 3.8Sub-Saharan Africa 699 927 968 2.5 0.8China 457 934 1,494 7.7 4.7Asia and the Pacific (excl. China) 862 1,922 3,502 8.7 6.2Latin America and the Caribbean 508 1,008 951 7.2 -1.1West Asia and North Africa 459 738 1,102 4.3 4.0

Industrial countries 4,298 5,713 6,941 2.7 1.7

Total a 7,283 11,248 14,958 4.4 2.8

a. The total excludes Cuba and the Russian Federation.Source: Alston, Pardey, and Roseboom 1997.

Recent Evolution of National Agricultural Research Systems 13

Table 1.3 Intensity of public sector investment in agricultural research(percentage of agricultural GDP)

1971-75 1976-80 1981-85 1986-90 1991

Developing countries 0.38 0.47 0.50 0.49 0.51Sub-Saharan Africa 0.78 0.84 0.86 0.74 0.70China 0.40 0.48 0.41 0.38 0.36Asia and the Pacific 0.26 0.36 0.44 0.50 0.55Latin America 0.43 0.51 0.59 0.49 0.54West Asia and North Africa 0.50 0.49 0.52 0.52 0.52

Industrial countries 1.38 1.60 1.98 2.18 2.39

Totala 0.67 0.76 0.81 0.79 0.81

a. The total excludes Cuba and the Russian Federation.Source: Alston, Pardey, and Roseboom 1997.

this, research systems in many countries that total factor productivity (TFP) in develop-have difficulty recruiting or retaining ing countries has grown at 1-2 percent perqualified staff. year-only slightly less than in industrializedIn recent years private sector investment in countries-and that research accounts for one-agricultural research has increased rap- third to one-half of that growth (Pingali andidly. Currently the private sector accounts Heisey 1996). There is some evidence, how-for 10-15 percent of resources invested in ever, that TFP growth has been slower in thedeveloping countries (Bonte-Friedheim, post-Green Revolution period (Morris andTabor, and Roseboom 1994; Pray and Byerlee forthcoming). This may indicate di-Umali-Deininger 1997), compared to 50 minishing returns to Green Revolution strate-percent in industrial countries. Much of gies of technical change and of emergingthis research is located in Latin America problems with sustaining the quality of the re-and Asia, where it is concentrated in a few source base.large countries, such as Brazil, Mexico, Ar- Studies have found a high rate of return togentina, and India. The bulk of private re- investments in research in developing coun-search is performed by local and tries (see Echeverria 1990 and Evenson andmultinational companies engaged in farm- Rosegrant 1993 for a review of more than 100input supply industries (for example, such studies). While there are always questionsseeds, fertilizers, pesticides, and machin- about the scope and methods of these studies,ery) with some research occurring in the there is little doubt that investment in agricul-food processing sector (James 1996). tural research has, on aggregate, been a huge

success. Even in Sub-Saharan Africa, wherePayoffs to Research Investments yield increases from new technologies have

been relatively modest, studies indicate thatOn the output side there is little doubt that research has paid off (Oehmke and Crawfordagricultural research has made important con- 1996). Rates of return may, however, havetributions to overall productivity growth. fallen in recent years, as marginal gains fromYields of most major crops in the developing some types of research have declined in theworld (with the important exception of Sub- post-Green Revolution period (Lipton 1994;Saharan Africa) have shown unprecedented Byerlee and Traxler 1995). In crop research,rapid growth in the past three decades and much of the benefit to date has been generatedyield growth rather than area expansion has through varietal development (box 1.1). De-now become the major contributor to increased spite the increasing importance of improvedagricultural production. The evidence suggests crop and resource management for sustainable


Box 1.1 Crop improvement research: An institutionalized success story

Plant breeding has been the major success story of support to national breeding programs through theagricultural research in developing countries over the supply of advanced genetic materials, often with spe-past three decades. These successes now go well be- cific traits, such as pest resistance, which require in-yond the widely documented cases of development tensive breeding efforts. One-half or more of theand adoption of modern varieties of rice and wheat varieties released in developing countries are basedin irrigated areas.1 directly or indirectly on IARC-linked germplasm.

More than 70 percent of the cereal area in the de- The success of genetic contributions to productiv-veloping world is now sown to so-called modem va- ity growth also reflects the high priority that NARSsrieties (Byerlee 1996). The remaining areas covered placed on developing strong crop-breeding pro-by traditional varieties are usually in more marginal grams. The number of scientists engaged in cropareas, although there is still considerable scope for breeding has increased rapidly. By 1990 the intensityadoption of modern varieties in medium- and high- of investment in crop breeding in developing coun-potential areas in Sub-Saharan Africa. Since the tries, measured by the number of scientists per unitspread of the first semi-dwarf rice and wheat varie- of production, was as high as in industrializedties in the 1960s, at least two generations of new va- countries (Bohn and Byerlee 1993). Institutional in-rieties have been adopted in most areas of Asia and novations in the form of nationally coordinatedin comnmercial agricultural areas of Latin America. commodity research programs have provided anThese successive generations have provided impor- integrated approach to highly focused crop-improve-tant benefits in higher and more stable yields. ment work. This has enabled NARSs to capture

The success of genetic improvement research can within country "spill-ins" of varieties and germ-be attributed to local innovations, international agri- plasm developed in other regions or states within acultural research center (IARC) support, and long country.term political and institutional commitment to food-crop research. The IARCs have provided strategic 1. Much of this box is based on Byerlee and Pingali 1995.

productivity improvement, research in this export crops (for example, cut flowers and hor-area has generated few successes (box 1.2).' In ticultural products). Improved technology is athe 1990s NARSs are being challenged to main- critical factor in a country's ability to exploittain the productivity of successful research its comparative and competitive advantagesprograms (such as crop breeding), revamp or (Bathrick and others 1996).drop unproductive research areas, and re- Second, agricultural research will be evenspond to new research challenges, all within more important in increasing productivity inthe context of ever tighter research budgets. the future than it has in the recent past. In the

course of agricultural development, technicalThe Changing Agricultural change generally provides an increasing shareResearch Agenda of output growth, relative to resource-based

strategies (for example, expanded land areaIn recent years the economic and technical en- and investment in irrigation), and input inten-vironment for research has changed drasti- sification. In contrast with the past three dec-cally. Three major changes have far-reaching ades, when the major emphasis was onimplications for the NARSs. First, trade liber- intensification of input use and cropping pat-alization has shifted economic incentives toward terns, the next stage of productivity increasesactivities in which countries have comparative will depend on pushing the production fron-advantage. Some traditional crops are now ex- tier outwards, increasing the efficiency of inputposed to competition from imports and may use, and sustaining the resource base. Theseno longer be profitable (for example, maize in research activities are more knowledge- andmuch of hillside Mexico and Central America), skill-intensive, since scientific knowledge em-while liberalization has opened opportunities bodied in new seeds (for example, a pest-resis-for nontraditional crops, especially high-value tant variety) or in the form of improved

Recent Evolution of National Agricultural Research Systems 15

Box 1.2 Crop and resource management research: A continuing challenge

Research on crop and resource management has not the resource base, especially for intensive productionhad the expected impacts.1 Most crop-management systems, where input use is already high (Cassmanresearch has focused on input intensification, espe- and Pingali 1995). There are unlikely to be quickcially the use of external inputs to exploit the yield "technological fixes" for natural resource degrada-potential of the new varieties. However, other fac- tion. For the immediate term at least, increasing pro-tors, such as the adoption of modern varieties and ductivity (and maintaining past productivity gains)the development of more effective input markets, is likely to come from incremental gains from a widehave probably been more influential than crop man- array of management practices, rather than any oneagement research in fostering higher levels of input technological breakthrough. While individually lessuse. newsworthy than the Green Revolution technologies

Crop-management research has been slow to en- of the past, these sources of growth, taken together,hance the efficiency of input use and more sustain- will nevertheless be significant.able production systems. One of the few success In low-potential areas, a "plant breeding first"stories demonstrating the potential of this type of strategy has generally failed, and it is now widelyresearch is the adoption of integrated pest manage- agreed that improvements in crop and resource man-ment practices on rice in some countries in Asia. For agement offer the best scope for increasing produc-example, between 1987 and 1990, the introduction of tivity and conserving the natural resource base (foran IPM program on rice in Indonesia resulted in a 50 example, Janssen 1993). There are few documentedpercent decrease in use of pesticides, a 15 percent success stories attributable to the adoption of im-increase in yields, an increased net profit per farmer proved crop management technologies in marginala season of US$18, and a government savings of areas. Adoption of such practices often requiresUS$120 million a year in pesticide subsidies (Schill- changes in local institutions (for example, land-ten-horn van Veen and others 1997). ure security) and policies (for example, removal of

In high-potential areas, the importance of empha- input subsidies).sizing more efficient use of inputs is reinforced byheightened concern about sustaining the quality of 1. Much of this box is based on Byerlee and Pingali 1995.

information for farmers (for example, IPM) is Byerlee and Eicher 1997 for the case of maize),substituted for higher levels of input use. These these have been relatively localized, dependentchanges imply greater exploitation of scientific on one technology such as hybrid maize, andknowledge at the farm level and more complex have not resulted in sustainable long-runfarming systems. Agricultural research sys- growth. The development of agriculturaltems will have to provide much more specific knowledge systems to respond to the urgenttechnical information, and linkages with tech- need to increase food production in the diversenology dissemination will have to move be- and risk-prone agroclimatic environments ofyond simple delivery of messages or "recipes" Africa must be a high priority for the develop-to be passed to farmers. All participants in the ment community into the twenty-first century.technology development and transfer sys- Finally, there are important changes in thetem-scientists, extensionists, and farmers- technology for research itself, especially in thewill have to make greater efforts to maintain new biotechnologies and informational tech-current skills. nologies. Institutionally, the new biotechnolo-

Third, a major challenge is to increase pro- gies raise important issues for research systemsductivity in areas that have benefited little due to the considerable size economies of muchfrom research investments in the past three of the research and the fact that the privatedecades. Slow agricultural growth and low and sector in industrialized countries dominatesvariable yields have thwarted efforts to reduce the applied end of this type of research.2 Sev-poverty and improve food security in Sub- eral large countries (for example, India) haveSaharan Africa. While there are some partial moved aggressively to establish centers of ex-success stories in Africa (see, for example, cellence in biotechnology, and the IARCs are


also shifting resources toward biotechnology, there are considerable potential efficiencyalthough their levels of investment may not yet gains in NARIs from streamlining research pri-be adequate. New information technologies orities, reforming management and incentive(such as crop modeling, geographic informa- systems, and involving a broader range of in-tion systems, remote sensing) will be important stitutions in the research process.research tools and will have important impli- Concerns also arise about the relevance ofcations for research organization and linkages. current institutional structures and govern-New communication technologies will sharply ance. The issue now is whether the highlyreduce the cost of communication among sci- centralized research management that charac-entists in the same NARS or internationally, terizes many NARIs is the appropriate organ-and these new technologies may reduce econo- izational mode. There are also questions aboutmies of size and the critical mass needed for the future of semiautonomous national re-some types of research. The new technologies search councils. These were established to co-may also help close the knowledge gap be- ordinate the financing of research and to givetween scientists in developing and industrial- agricultural researchers independence fromized countries. the "tyranny of civil-service regulations" on

hiring and remuneration (Antholt 1994). InEmerging Institutional Challenges in practice, few have been able or willing to breakNational Agricultural Research Systems away from the rigidities of government civil-

service rules. Consequently, many researchThe increasing demands being placed on re- systems are suffering a crisis of management,search systems contrast starkly with the lack with top-heavy bureaucracy, centralization ofof institutional innovation in research-system decisionmaking, and lack of incentives formanagement and organization and the stagna- the innovation process so essential for re-tion or decline in funding for agricultural research (Antholt 1994; McIntire 1994). Althoughsearch. Unstable research funding makes it these problems are generic, situations varydifficult for national and international research considerably and require country-specificprograms to comnmit human and financial analysis (box 1.3).resources to address the long-term strategic As a result of these problems and the diffi-issues crucial for sustaining productivity culties in making the transition from a culturegrowth. With few notable exceptions, such as of program expansion (building, training, gear-the Punjab of India, the public research systems ing up) to that of production (of ideas, researchthat will continue to be the key source of tech- papers, and extension recommendations),nical change for basic food staples have failed most NARIs have failed to develop the scienceto muster political support for funding agricul- environment needed to spur innovation andtural research. discovery and which largely determines effec-

Reform of public research system organiza- tiveness. One result is a decline in the qualitytion and management is long overdue. The of scientific output, since it has become difficult1970s and early 1980s were golden years for to attract and retain the best scientists in publicagricultural research. Agricultural research sector agricultural research (Purcell 1994). Fewwas a prestige investment and, with rapidly NARSs have adequately bridged the gap be-rising budgets, research programs and insti- tween public research suppliers and technol-tutes proliferated. The relative abundance of ogy users. The public-good nature of researchresources flowing to NARIs led to compla- has not been adequately defined and the pri-cency in research management and lack of de- vate sector and farmers have not been suffi-mand for accountability of management to ciently integrated into the technologythose who financed research. In the austere systems-either as financiers of research or asbudgetary climate of the 1990s, it is clear that partners in the conduct of research.

Recent Evolution of National AgricuLltural Research Systems 17

Box 1.3 The diversity of National Agricultural Research Systems

National agricultural research systems' development market-oriented technology support systems. In mostchallenges vary considerably by region (Purcell and cases, a sharp downsizing and restructuring will beAnderson 1997). Research institutions in South Asia required to bring the number of scientists in line withhave generally benefited from a longer period of sup- needs and financial support (Pray and Andersonport and have developed greater capacity. In Sub-Sa- 1997). In all regions, but especially in Africa, smallharan Africa investments in the NARSs began later NARSs present special challenges. Small countries re-and much of the earlier investment was focused on quire systems to foster technological innovation, butcommercial farmers and export crops. In Latin Amer- must build low-cost systems that emphasize technol-ica commercial agriculture is more developed and the ogy import and adaptation. 1

private sector, including farmers themselves, cansupport much of the needed technological innova- 1. Small NARSs are defined by ISNAR as countries with lesstion. However, public NARI capacity has frequently than five million population and three of the following char-not been well maintained in Latin America and is acteristics: agriculturally active population of 20 percent orgenerally not adequate to the needs of smallholders more of the economically active population; per capita incomeand of natural resource management in the region; of less than $2,000 in constant 1980 dollars; agricultural GDPnor is it well able to foresee and respond to changing per capita for economically active agricultural population ofcompetitive challenges resulting from economic lib- less than $2,000; and agricultural GDP of 20 percent or moreeralization. The NARSs of Eastern Europe and Cen- of total GDP. Okello and Eyzaquirre (1992) identify 53 small-tral Asia have yet to make the transition to flexible, country NARSs, 24 of which are in Sub-Saharan Africa.

In summary, the substantial investment in NotesNARIs over the past two decades has not yet

resulted in the institutional capacity and output 1. Less is known about the impacts of livestock re-expected. A major human and physical infra- search. Undoubtedly the major impact of livestock re-structure is now in place, but many NARIs that search has been in livestock health. Livestock nutritiondeveloped rapidly have suffered erosion of ca- research appears to have provided few benefits, de-pacity over the past decade because of uncer- spite considerable activity (Cornelius de Haan, World

tain funding, lack of articulation with other Bank, personal communication).actors in the NARSs, lack of transparent long 2. Since high investment costs are required to estab-run strategies and priority-setting mechanisms, lish biotechnology research capacity, initial researchand difficulties in maintaining the quality of projects are costly, but additional research projects canhuman resources (McMahon 1992; Purcell and be added with relatively little additional investment inAnderson 1997). facilities.

CHAPTER 2

National Research System Development:Issues and Good Practice

Tl nhe paper now reviews selected key is- With the swing toward privatization, decen-sues facing research managers, policy- tralization, and competitiveness in the latemakers, and the Bank in positioning 1980s and 1990s, the public sector monopoly of

national research systems to address chal- the NARI is now an obsolete institutionallenges into the twenty-first century. We divide model for building NARS capacity (Echeverria,these into the major issues for the evolving Trigo, and Byerlee 1996). The challenge forNARSs (discussed here in chapter 2) and key most countries is determining how to capturechallenges for the public sector research insti- the potential of alternative suppliers to bringtutes (that is, the NARIs) within those NARSs more resources into the formal research sys-(discussed in chapter 3). tem, and identifying how to exploit the com-

An important change in the 1990s has been plementarities among various participants toto move beyond equating the national agricul- develop a well-articulated research system.tural research system with the public research This requires that NARS institutions, public orinstitute (NARI) to viewing it as institutionally private, develop strategic linkages with otherpluralistic system that includes: private and public sector R&D efforts at the

* Universities, which often command access regional and global level. Such changes bringto well-trained human resources, but increased institutional complexity and requirewhich have oftenbeen considered as solely significant institutional innovation. The majorteaching institutions challenge for the NARSs is, therefore, how to

. Private companies, which seek to develop involve and link these various actors at theand sell technological products for profit national and international levels to meet na-

. Agricultural research foundations, which tional objectives for the agricultural and ruralenjoy varying degrees of private and pub- sectors.lic support

. Farmer organizations and cooperatives, Conceptual Overview of the Emergingwhich might organize their own research National Agricultural Research Systemsinstitutions or support research by othermeans The conceptual view of national agricultural

* NGOs, some of which have the capacity to research systems has evolved, and continues toundertake certain types of research, espe- evolve rapidly. Prior to 1960 most agricul-cially adaptive research. tural research was carried out in government

18

National Research System Development: Issues and Good Practice 19

departments and experiment stations that were spects (figure 2.2). First, it distinguishes be-not integrated into a coordinated national sys- tween the funding of research and the execu-tem. Beginning in the 1950s many developing tion of research as separate functions requiringcountries formalized public sector involve- different skills and inputs. Research fundingment in agricultural research with the forma- and the setting of broad priorities for researchtion of national agricultural research institutes, are policy tasks that relate to policy objec-which grew rapidly in the period of rapid ex- tives and client needs. Once funding levelspansion of funding for agricultural research and research priorities have been establishedthrough the 1970s. In this period a national research execution by alternative suppliers be-research system (that is, the NARS) was comes an efficiency issue (Echeverria, Trigo,viewed largely as a linear system with the min- and Byerlee 1996). While funding and execu-istry of agriculture providing funds to a na- tion may continue to be under the same organi-tional agricultural research institute, which zation, there is a strong case for separatingdeveloped new technologies that were passed them institutionally. A separate funding bodyto extension for dissemination to farmers (fig- can use competitive grants to seek the mosture 2.1). In this view, private sector R&D, if it efficient research providers for a given researchwas recognized at all, was seen as being some- product, avoid conflicts of interest in awardingthing apart, largely associated with supply of research contracts by not itself conducting re-certain inputs to commercial farming where search, and oversee the quality control of re-the benefits of investment in R&D could be search through setting research standards andappropriated (for example, hybrid seed, ma- organizing external reviews (box 2.1).chinery, and agricultural chemicals). Second, the new concept of a NARS recog-

The emerging view of NARSs in the 1990s nizes the importance of diversity in the fundingdiffers from this traditional view in several re- and execution of research (McMahon 1992).

Figure 2.1 Traditional view of a national agricultural Figure 2.2 New perspective on national agriculturalresearch system research systems

Public sector Private sector

Ministry of ii i iAgriculture ResearchMnistry of Agrbusiness,

funding ~Agnculture, foundations,othersfarmers -p

Fund(srPrivate Public Private i

agribusiness goods "I goodsNARI with research g :

and development E aResearch NARI Agnibusiness, a i

Technology execution universities NGOs es

g g ~~~~~~aExtension n

_ _ ~~~~~~~~~~~~~~~~~~~~Extension--diverse dpublic-private mix t

l ~~~~~Input_ _ u

nInputs mxd

Farmers Farmers

20 Strengthening National Agricultuiral Research Systems

Box 2.1 Funding versus execution NARS a reality include budgetary pressures onof research in Australia research agencies, greater scientific capabilities

within a range of institutions, increased clientOne model that clearly distinguishes research fund- and stakeholder pressures on institutions,ing and priority setting from research execution is greater sophistication of institutional struc-that of the research and development corporations tures, and changing research agendas.in Australia. These corporations were set up as Third, in the new view of NARSs, the privatesemi-autonomous bodies as a partnership betweengovernment and farmers, with funding provided sector is recognized as a full player in the de-equally by each partner. The sixteen research cor- velopment and dissemination of agriculturalporations, representing most major commodities as technology: the public sector is focused morewell as research on land and water management, sharply on public goods (Umali 1992). In re-are each governed by a board with representation search, public goods are characterized by tech-from government, farmers, agribusiness, scientists,and the wider community. The corporations oper- nologies where it is impossible or very costlyate with small staffs and low administrative costs for private firms to appropriate all the benefits(less than five percent of total research funds), set of investment in R&D, since it is not possiblestrategies and priorities, and allocate funds, mostly to exclude potential users once a research prod-through competitive bidding. The corporations uct becomes available. In addition, public sec-have full autonomy to contract research to public tor intervention in agricultural research is oftenand private sector institutes, including foreign dinstitutions. desirable for other reasons (Alston, Pardey,

and Roseboom 1997):e High fixed costs and uncertain, long-term

Instead of funding being solely dependent on payoffs to research that make it unprofit-the government treasury, a variety of means of able for individual small firms (that is,funding research are exploited, including pri- farms) to conduct researchvate sector and farmer contributions. Even in * Positive or negative environmental exter-the public sector, funding may be diversified nalities of much agricultural technologyfrom the ministry of agriculture to other min- (for example, soil conservation, agrofor-istries, such as science and technology and en- estry, and pesticide use).1

vironment. This diversification of funding For these reasons there is a strong case for pub-sources can potentially increase funds for re- lic sector involvement in basic and strategicsearch and at the same time stabilize funding research (long-term research with uncertainlevels, since the risk of sole dependence on one payoffs and high spillovers), research on prob-source is reduced. Likewise, various public and lems of small-farm agriculture (high costs ofprivate institutional forms, especially universi- collective action for organizing their own re-ties and various types of NGOs, including search), and research on natural resourcefoundations, may be used to execute research. management (positive environmental exter-The participation of other actors in research nalities). However, the role of the public sectorexecution allows the human resources and continues to evolve with changes in technologyskills available for research to be increased (for and institutions. For example, with emphasisexample, by tapping underused skills in union biotechnology and with new forms of intel-versities) and efficiency to be enhanced by lectual property protection for biological prod-matching scientific skills with needs. Reorient- ucts and processes, the private sector is now aing thinking so that a NARS is viewed as a major player in basic and strategic biologicalcollection of institutions and programs rather research that was previously almost exclu-than a monolithic NARI will remain a chal- sively in the public sector.lenge to development practitioners and re- Fourth, despite the sharper delineation of thesearch program leaders. However, various public and private sector roles in funding re-forces that combine to make the pluralistic search, there are often efficiencies in private-


public collaboration in both the funding and looking for research institutions to becomeexecution of research due to the complemen- more "results-oriented."tary skills and resources in each sector. The Finally, the traditional research-extensionprivate sector may find it efficient to contract link will be less relevant, since much researchthe provision of research to public research in- will be done with the direct involvement ofstitutes that have specialized research infra- potential technology users as financiers and ex-structure and skills: public research institutes ecutors of research. In addition, technologymay find it valuable to collaborate with private dissemination and information transfer pro-sector firms that have appropriate business grams are undergoing a similar transformationcontacts and experience in commercialization with a conceptual distinction between fundingof research products (figure 2.2). and execution of technology dissemination ac-

Fifth, for public sector research institutes to tivities and with a greater diversity of provid-operate effectively in this new environment, ers of extension services. These servicethey require increased flexibility in financial providers are forced to be more accountableand human resource management-for exam- and active in developing linkages to sources ofple, autonomy to engage in contracts with pri- new technology. Effective research organiza-vate firms, or to engage specialized skills. This tions will need to develop new models of tech-implies that public sector research institutes nology transfer that involve formal andmust seek new institutional models that pro- informal partnerships with a variety of provid-vide some level of autonomy from traditional ers of extension and information disseminationcivil service models. services.

Sixth, and critically important, stakeholders,especially the clients of the research system, Coordination in the Emerging Nationalhave an increased role in research priority set- Agricultural Research Systemsting and implementation to ensure that re-search is driven by user demands. The The complexity of the new NARSs raises theinstitutional models for achieving this may issue of how to coordinate research to meetrange from full or partial funding of research national development objectives. There cannot,by stakeholders to involvement of farmers in and probably should not, be a complete coor-the governance of research funding and exe- dination or integration of various research pro-cuting bodies, and to farmers' organizations grams: institutional mandates, objectives, andcontracting research organizations to execute interests will naturally diverge. Private sectoradaptive research. research frequently requires some degree of se-

Implicit in the increased importance of stake- crecy and independence because of the profitholders is the need for institutions within the motive. University research may (but need not)NARSs to become more responsive to the be directed toward more ad hoc and academicneeds of the stakeholders and to become ac- research activities. Stakeholder interests maycountable for producing results. Research conflict (that is, small-scale producers versusagendas will be less shaped by national plan- large producers; producers versus consumers).ning and national policies and more by the Yet, to meet national objectives, some degreerequirements of markets (domestic and export) of coordination and integration is needed toand the country's comparative and competitive link varied programs in a national system thatadvantages. Private sector financiers of re- achieves efficiencies through collaborative re-search in the public sector necessarily demand search activities, networking, sharing of re-accountability and results (profits) from their sults, and improved allocation of researchinvestments. Public funding agencies facing funds.difficult choices among alternative invest- Mechanisms for governance of multi-institu-ments and ever scarcer public resources are tional NARSs are not yet well developed. As


NARSs mature and evolve more complex or- ways runs the danger of adding another layerganizational structures, research leaders and of bureaucracy to research decision making.policymakers should give explicit attention to Other institutional mechanisms may need toestablishing systems that enable researchers be considered, such as:from various institutions to share information, * A research coordination committee to al-exploit complementarities between institu- low regular consultation and communica-tions, and avoid duplication of efforts. Many tion among various NARS participantscountries have attempted to do this through * National coordinated research programsthe formation of apex research councils respon- and networks in which various institutionssible for: collaborate and exchange information for

* Developing national research strategies a particular research areaand plans * Research workshops that bring together all

* Linking research to broader agricultural interested researchers, clients, and stake-policy discussions holders for a particular type of research

* Channeling funds to priority research ar- and provide an opportunity to share re-eas, and thus coordinating research across search results, compare plans, and coordi-research institutions, especially in a fed- nate future workeral-state system * Competitive research grant programs to

. Promoting collaboration and exchanges direct research efforts of various institu-among the various parts of the NARS tions and to promote linkages between dif-

* Coordinating external linkages, especially ferent programswith NARSs of neighboring countries and . Professional associations, research jour-with donors. nals, annual research reports, newsletters,

Such a research council, if broadly repre- and other publications that keep scientistssentative of the clients of research (for example, throughout a NARS informed of currentpolicymakers and farmers), should provide a research efforts and accomplishments.national forum for developing research strate- Each country will have to work out its owngies and allocating research funds. However, specific institutional model for coordinationin practice, many apex research bodies have within the NARS. In large countries severalbecome large research institutes in themselves models may be needed to represent the inter-with major involvement in research execution. ests of different commodities or regions in re-In other cases apex organizations have been search funding and priority setting. Finally, itcharged with research planning and coordina- is very useful for a country to a develop broad-tion, without control over funds. Both situ- based consensus or vision for the future evo-ations typically lead to failure of the lution of the NARS (box 2.2). Such a vision ororganization to undertake its primary respon- technology policy is the starting point for ef-sibilities of directing research policy. forts to reform research systems and increase

In countries with an apex research council, efficiencies.public funds should be administered by thiscouncil, and the council should divest itself of Competitive Funding of Researchadministrative activities in research executionby giving more autonomy to its research insti- Competitive grants (CGs) can help coordinatetutes. Councils should ensure that producers research across different institutions in linehave input into the process of research plan- with national priorities as well as generate ef-ning and evaluation. Where there is no apex ficiencies and stimulate innovation in researchresearch organization, the creation of a small, programs. Competitive funds can also be usednonbureaucratic council should be considered. to consolidate funding from different sourcesHowever, the creation of a new institution al- to address national priorities.


Box 2.2 Building consensus: The importance of a strategic vision

Research investments should be consistent with and changes in the public sector NARI to promote greatersupportive of national objectives. Each NARS needs efficiency in use of resources, including institutionala broad "vision" of the future evolution of the re- autonomy and governance and linkages with clients,search system over the next 10-20 years that is pref- technology transfer agencies, and the global agricul-erably developed within the framework of a national tural research system.science and technology policy. This NARS vision The NARS vision statement can be developed byshould succinctly outline the changing demands for the apex research body, the NARI, or a specially con-technology, the expected future roles of the public stituted committee or comunission. In each case con-and private sectors and their interaction, the com- siderable dialogue among all stakeholders is neededparative advantage and mandates for central, state, to ensure broad consensus on the future direction ofand local research institutes, and the role of univer- the NARS. Such a vision statement should also be asities. The vision should also analyze future sources foundation for Bank support to a new lending pro-of funding, new funding mechanisms, and needed gram for agricultural research.

CGs are generally advertised, with selection example, biotechnology and multidiscipli-of grants for funding based on a peer-review nary research).process with criteria defined by research pri- Several potential difficulties with CGs thatorities and scientific merit. Although CGs have should be recognized:been used in many NARSs for many years, * Most grants are for a fixed period (usuallythese have accounted for only a small propor- no more than three years) and are lesstion of the total research budget. appropriate for supporting long-term

CGs have a number of potential advantages. research.They can: * It is often difficult to establish an inde-

* Allow funds to be channeled to the most pendent peer review system in situations inproductive researchers (as judged by the which a research culture is subservient topeer reviewers). They can thus improve seniority and administrative hierarchies.the productivity, job satisfaction, and com- a In small NARSs there maybe few potentialmitment of scientists competitors for funds, so CGs may fail to

* Enhance the quality of research by requir- promote competition. It may also be diffi-ing detailed proposals from scientists and cult to find peer reviewers who do not haveby technical review of the proposals prior personal interests in the projects beingto funding reviewed.

* Draw a wide range of participants into the * The peer review system is inherently con-research system by making competition servative and may discount projects out-open to all, including NGOs, universities, side the currently accepted view of science.and the private sector * Administration of CGs may get mired in

. Mobilize established research infrastruc- the normal bureaucracy of governmentture and human resources, which might civil service, resulting in long delaysotherwise be under-used because of short- and administrative barriers to accessingage of operating funds funds.

* Promote partnerships in research by en- * CG systems are generally more costly,couraging collaborative research propos- since scientists may spend considerableals across institutions, including joint time in writing proposals that are notpublic-private sector research funded and peer review is a time-consum-

* Reallocate research resources in the short ing and skill-intensive activity.term by tying grants to high priority re- * CGs usually only fund operating costs andsearch areas and types of research (for essential equipment. Undue reliance on


CGs may be at the expense of maintaining CG scheme should have maximum flexibilityexisting research infrastructure. and independence from political and bureau-

CGs have become popular in recent years in cratic interference. Where this is not possiblemany NARSs, but CGs are not a substitute for within the public sector, a special board or foun-institutional development and longer-term in- dation may be needed to administer the CG.vestments in developing research infrastruc-ture. Successful CG program management Integrating Universities in the NARSrequires good monitoring and evaluation sys-tems. In countries with underused research ca- Despite substantial R&D capacity many uni-pacity, CGs can be an efficient means for versities in developing countries remain on thefinancing research, but CG programs may periphery of agricultural R&D (Rukuni 1996)become less effective over time, if research in- (box 2.3). This contrasts with some countriesfrastructure (both human and physical) depre- (for example, the United States and India)ciates and if grant managers and recipients where universities predominate in publicbecome "entrenched" so that the programs are sector agricultural research. Universities pro-less competitive. vide a pool of well-trained scientists with sub-

These characteristics of CGs suggest a num- stantial capacity for executing agricultural4ber of good practices. CGs are most appropri- research. Universities also train the next gen-

ate in mature and larger NARSs seeking to eration of scientists.widen participation in the research process and The basic strength of universities is that theirprovide incentives for increasing productivity.3 research and training functions are comple-In small NARSs (with less than 100 scientists), mentary to each other since research is anthe high administrative costs and limited po- integral part of postgraduate education. Uni-tential for competition reduces the value of versities frequently have an institutional cul-CGs. CGs should not be the sole mechanism ture and a relatively autonomous statusfor funding, but should be used to complement conducive to research, but difficult to establishfunds from regular budget appropriations. within a NARI. The greater flexibility in oper-CGs are most appropriate to fund new research ating procedures and regulations in universi-areas and research and development that re- ties may make it easier for university scientistsquires collaboration, such as public-private to obtain funding and engage in collaborativepartnerships. Long-term research requiring research with private sector firms and othercontinuity (for example, a plant breeding pro- funding entities. Indeed, in some countries itgram) as well as the building and maintenance may be most appropriate to place universitiesof research infrastructure are best funded in the lead in executing research and to givethrough annual budget appropriations. them the status, responsibility, and funding

The use of CGs should be introduced on a usually associated with NARIs.pilot basis to fund selected priority areas and The move toward looking at research sys-then be evaluated for effectiveness to guide any tems as a whole will inevitably elevate the roleexpansion of the program and to refine proce- of universities in research. The following meas-dures and institutional arrangements to ad- ures can also be used to better integrate uni-minister the CGs. In order to make the process versities into NARSs:transparent and to widen participation, intro- . Shift more funding to competitive grantsduction of a CG system should be accompa- to tap university skills in research.nied by a detailed manual and appropriate * Develop collaborative research programstraining programs on procedures for soliciting, (for example, through a special fund) be-preparing, and evaluating proposals, criteria tween NARIs and universities, especiallyfor selecting proposals, and guidelines for in the common situation where these insti-monitoring and evaluation. Management of a tutions are physically close.


Box 2.3 Why have universities been underutilized in national agricultural research systems?

For several reasons, universities have often been mar- * National agricultural research strategies andginalized from agricultural research: plans have not recognized the potential role of

* Universities are generally administered under universities and have not provided mechanismsministries of education, which have no mandate to link university research into the overall na-for and little interest in promoting agricultural tional plan.technology. * Institutional and sometimes personal rivalries

* Universities have little funding for research. In sometimes exist between universities and theNigeria the 53 percent of agricultural scientists ministries of agriculture that oversee both thein the universities had access to seven percent financing and execution of research.of the funding for research, while the 47 per- * In many countries universities are frequentlycent of the scientists in government institutes disrupted by political demonstrations, strikes,had access to 93 percent (Hoste and others and shutdown of university operation, a situ-1995). ation incompatible with continuity in research

* Universities provide inadequate incentives for work and prejudicial to sustained governmentresearch. funding.

. Provide opportunities for staff exchanges, strategic plans, monitoring systems, and sys-

such as graduate students undertaking the- tems for setting priorities. Many must give at-

sis research in the NARI and NARI scien- tention to research management and research

tists taking sabbatical leaves in universities. policies and will need to address operational

* Develop interlinkages in governing bodies issues similar to those in NARIs (see chapter

with representatives of universities sitting 3). These issues include: providing incentives

on governing boards of NARIs, and vice for research, maintaining and upgrading re-

versa. search facilities, encouraging contract and

* Reduce donor support for foreign post- grant research funded by the private sector andgraduate training and channel this support other sources, obtaining intellectual property

to developing local university capacity in rights and commercializing some research

both undergraduate and postgraduate products and services, and better use of their

training. This change may begin with mas- most valuable asset-relatively low-cost post-

ter's degree programs, followed by general graduate students.

Ph.D. programs (for example, in plant

breeding and agronomy). Increasing the Role of the Private SectorNARSs must institute policy and structural

reforms to expand university roles in research, The gap in funding for research in the public

and universities themselves must adopt poli- sector can partly be alleviated by seeking an in-

cies and develop capabilities to conduct re- creasing role for the private sector. The private

search. In addition, upgrading of universities sector can invest in its own R&D capacity expect-

will be critical to improving the human re- ing to earn profits from that investment, or it can

source capacity for scientific research in both mobilize resources for profit or nonprofit goals

the public and private sectors of NARSs. Given through private research foundations.

the high cost of foreign postgraduate training,

a sustainable NARS must have capacity to pro- Investment by the Private Sector in Researchduce most of its replacement scientists. and Development

To ensure NARS development much more

effort is needed to establish and support re- Although private-sector investment in R&D is

search in the university sector. Larger univer- increasing in developing countries, it still only

sities with substantial research may need accounts for 10-15 percent of total agricultural


R&D expenditures (Bonte-Friedheim, Tabor, These changes are already well under wayand Roseboom 1994; Pray and Umali-Deiniger in many countries and have stimulated private1997). Privatization of agricultural research has sector R&D. However, private sector R&D inproceeded most rapidly in Latin America, most developing countries will probably bewhere the private sector contributes up to 40 limited for many years for some of the follow-percent of the R&D expenditures in some coun- ing reasons:tries (Falconi and Elliott 1995; Echeverria and * Many products of research will continue toothers 1996). In contrast, private sector agricul- be public goods that do not attract privatetural R&D is estimated to account for half of sector investment. These include crop,the research expenditures in the industrialized livestock, and resource management prac-countries (Alston and others 1997). tices, where the major product of research

In developing countries private-sector in- is information (for example, IPM).vestment in R&D has been dominant in agri- . Even with implementation of IPRs, privatecultural chemicals and machinery, where sector research will focus on commercialmuch of the technology has been directly trans- agriculture. For example, in plant-breed-ferred from industrialized countries. Private ing research, it is difficult to see how it willsector R&D is also growing rapidly in other be cost-effective for private firms to en-areas, especially for hybrid seed. In Asia and force IPRs for self-pollinated crops in theLatin America private sector R&D in the maize small-holder sector, where seed can easilyseed industry now probably exceeds that by be passed from farmer to farmer.the public sector.i Although the R&D that es- . Developing private R&D is a long-termtablished the market for hybrid seed has process, which requires considerable inputlargely been carried out in the public sector, from public R&D. Even proprietary hybridsparticipation of the private sector in R&D in- areoftenbasedonprebreedingandbreedingcreases as the market matures. Since many de- research in the public sector.6 The mainveloping country markets are reaching this source of scientific personnel for private-sec-phase of development, and given the develop- tor R&D efforts is the public sector NARI.ment of hybrid seed technology for other crops * Market size is an important determinant of(for example, rice and cotton) and the imple- private R&D. This places small countriesmentation of plant variety rights, the impor- and secondary commodities at a disadvan-tance of the private sector in crop improvement tage in attracting private R&D. Since mostR&D will probably increase rapidly. countries still restrict technology imports

Private-sector R&D can be promoted by a (by varietal release procedures and regula-number of policy and institutional reforms: tions on imported seed), private R&D in

* Encouraging private investment in general small countries will only prosper wherein the agricultural sector, including foreign countries within a region adopt open-bor-investment. der policies for technology flows, so that

* Removing barriers to private sector par- private companies can target a regionticipation (including rules on access to rather than a single country.public-sector research products) and relax- These comments point to the complemen-ing rules on approval and release of new tarities of public and private sector R&D. How-technologies (such as new varieties) and on ever, too often the public sector has viewed theimporting technology private sector as a competitor. A much more

* Strengthening legal structures to allow pri- constructive approach would develop strate-vate appropriation of the benefits of pri- gies to exploit complementarities. These strate-vate R&D, including legislation on trade gies could include:secrets and intellectual property rights * Backing private sector R&D with strategic(IPR). and applied research in the public sector in


ways that foster a competitive and efficient rations, agribusiness, philanthropists, and for-private sector, including the development eign donors. Foundations, as a type of NGO,of R&D capacity in local agricultural input- have considerable flexibility in allocating andsupply firms administering funds, and may be able to re-

. Focusing public sector research in areas spond quickly to new research challenges.unlikely to attract the private sector, be- Foundations with an endowment are assuredcause of the nature of the technology (for of funding stability at least for core activities.example, many crop-management tech- Finally, foundations may promote close linksnologies) or the nature of the farmer (for and responsiveness to client needs since theseexample, farmers in more marginal areas) are often managed by boards with majorityto avoid the potential of undermining in- representation from private industry.centives for private R&D Agricultural research foundations have been

. Developing institutional mechanisms (for established in various ways and for differentexample, specialized competitive funds) purposes (box 2.4). Large private corporationsand appropriate legal frameworks for joint have helped to establish foundations for socialventures between the public and private or public relations reasons.7 Others allow pri-sectors, especially mechanisms that allow vate firms that lack sufficient resources or skillsthe private sector to adapt and test-market to pool efforts and undertake joint R&D pro-public sector technologies, with a share of grams (Trigo 1987). Commercial farmers maythe profits being returned to the public support the establishment of foundations, suchsector research institution as the Agricultural Research Trust of Zim-

. Rationalizing legal and institutional barri- babwe. In Latin America bilateral donorsers on technology imports and releases so (especially the U. S. Agency for Internationalas to clearly define issues of public interest Development) have helped establish research(for example, reduced public health haz-ards).

To fully harness these complementarities and Box 2.4 Research foundations: A diversitysynergies will require more attention to each of approachessector's comparative advantage. However, it Foundation operations are as varied as their meanswill be many years before the private sector of formation. Many foundations, reflecting theirassumes a large part of the R&D needs in most support from private business and from the U.S.developing countries. In the meantime, public Agency for Intemational Development (USAID), fo-support to research must increase. cus on agribusiness development. The Jamaican

Agricultural Development Foundation undertakesa varied program, including management of a com-petitive research grant fund and direct, venture-capital-type investments in agriculture. The

Private foundations (for example, the Ford and CIBA-Geigy Foundation (in Mali) and the Agricul-Rockefeller Foundations) have been significant tural Research Trust of Zimbabwe both operate re-in supporting agricultural research, especially search stations. The Agricultural Developmentinternational research. Although private foun- Foundations of Peru (FUNDEAGRO) and Ecuador

(FUNDAGRO) support development and operationdations exist throughout the developing of the NARI, as well as carry out their own activi-world, this institutional approach has to date ties. The Fundacion Chile, the Foundation for thebeen most widely applied in Latin America. Development of Polish Agriculture, and many

Foundations have several advantages for other foundations promote agricultural investmentsupporting agricultural research. They repre- and provide technology support to private agri-sentuon agriculturalmbusiness. Some foundations have been established

in response to the frustration, especially of donors,raising and managing research funds from with instituting reforms in the public NARIs.nontraditional sources, such as private corpo-


foundations as alternatives to NARIs in several broadened mandates emphasizing natural re-countries. sources as well as traditional germplasm prod-

There are, however, problems with the founda- ucts; (2) regional centers and networks; (3)tions and some doubt whether many will be a private sector firms, many with large multina-sustainable long-term addition to the tool kit of tional R&D programs; and (4) an extremely di-NARSs (Sarles 1990; Echeverria, and others 1996): verse array of NARSs, many with both federal

* Foundations established without sizable and state systems. In a world in which costs ofendowments have encountered funding some types of R&D are high, especially forproblems and found it difficult to raise some basic and strategic research (for example,funds from domestic sources. Some foun- biotechnology), there is a strong rationale fordations started under donor projects had NARSs to link into a global system to captureno long-term financial strategy and faced a spill-ins of technology and reduce the cost ofcrisis at the end of donor support. Endow- technology development.ments provide some independence andstability for a foundation (Horkan and Jor- Technology Spill-ins and the Globalizationdan 1996).8 of Research

• Foundations that depend on agribusinessand farmer support are likely to address A country may capture spill-ins by importingonly the needs of a minority of farmers, several types of technological products fromtypically the commercial farmers. As a the global system. These include improvedfoundation's program becomes more com- technologies that can be released directly tomercial, it may crowd out legitimate pri- farmers after initial in-country screening (di-vate sector R&D activities. rect spill-ins) or adapted through local research

* Foundations are often not well linked to to fit local conditions (indirect or adaptivethe rest of the NARS, so that duplication or spill-ins). They also include scientific informa-lack of complementarity in research may tion, new knowledge, and methods for im-be a problem (Jarvis 1994). There may be proving the efficiency of its own technologylittle net additional benefit if foundations development efforts (knowledge spill-ins).simply take over public sector research.9 The ability to capture research spill-ins is

Experience shows that foundations can be an particularly important for small NARSs andefficient and sustainable means of enriching the for research institutes serving a small region orinstitutional mix of NARSs. However, these commodity that cannot justify investment in asuccessful experiences show that to ensure sus- full technology development program. Thetainability and effectiveness, foundations must best sources of spill-ins are neighboring re-be well endowed and have strong local support. gions within a country, the IARCs, and coun-Opportunities to establish foundations and the tries with similar agricultural systems andorientation of foundation activities in research agroecological features. A strategy emphasiz-will be highly situation-specific in capitalizing ing capture of research spill-ins can be veryon unique funding opportunities, committed productive, but requires flexibility in the re-leadership, and specific development agendas. search system to identify and act upon oppor-

tunities arising from developments elsewhere.Linking the National Agricultural It demands that NARSs recognize that carefulResearch Systems to Regional and screening of imported technologies is a legiti-International Research Systems mate scientific activity and an efficient use of

scarce scientific talent.The global agricultural research system has Recent work has shown that potential spil-evolved today into a complex system com- lovers, even for biological technologies, may beposed of (1) international centers (IARCs) with large (box 2.5). However, realization of these


Box 2.5 Technology spill-ins: Spreading research benefits beyond borders

Conventional wisdom has maintained that biologi- Recent work for wheat has shown that both poten-cal technologies are location-specific and must be tial and actual spill-ins of wheat varieties are muchadapted to local conditions, so that direct importa- larger than expected (Maredia and Byerlee 1996). Thetion of technologies is not feasible. In the 1960s and cost of local research can be considerably reduced if a1970s this argument was used to justify investment country, especially a small country, depends on im-in research in tropical and subtropical areas, since ported varieties, since a testing program to screen im-technologies from temperate countries, where most ported varieties is only about one-third the cost of aagricultural research on food crops had been full breeding program.1 It is clear that private R&D isconducted, were poorly adapted to many develop- organized to maximize spill-ins. Multinational seeding countries. Since investment in research in companies, for example, usually establish only onedeveloping countries now exceeds that in indus- major research station in a large country, or one re-trialized countries, the potential for spillovers gional station to serve several smaller countries. Im-between developing countries with similar agro- ports of seed across countries with similarclimatic conditions is much greater. The growing environments can be substantial (Echeverria 1991).ability to use new information technologies, suchas Geographic Information Systems (GIS), to de- 1. Other recent studies, such as Davis, Oram, and Ryanscribe like agroecological zones should greatly (1987); Evenson (1997); and Thirtle and others (1995), alsoincrease the precision with which spillovers can demonstrate the considerable potential for internationalbe targeted across countries, as for example be- spill-ins of technology. A similar situation prevails in indus-tween highland regions of the Andes and Himala- trial countries where, in large countries such as Australia andyas. Research in some new technologies, especially the United States, half or more of the benefits of research inbiotechnology, is less location-specific than con- one state are captured by other states (Evenson 1994; Brennanventional technologies. forthcoming) and other countries (Thirtle and others 1995).

spillovers depends on geographic proximity * Eliminate import duties and nontariff re-and historical and cultural links between coun- strictions on the import of agricultural andtries. It also depends on policies, such as open- research inputsness of the economy in exposure and incentives * Make seed certification voluntary andfor private import of technology, as well as optional, while facilitating certificationregulations and licensing requirements for im- to the extent possible and enforcingporting chemicals and planting materials. "truth-in- labeling" laws for seed and otherQuarantine laws and local rules on testing and inputsrelease of agricultural technologies also influ- * Streamline seed varietal registration pro-ence the degree and speed of spillovers. Fi- cedures. (Varietal registration might benally, for importation of technologies to occur made voluntary or, as an interim step,successfully and rapidly, considerable local ca- countries in a region can adopt commonpacity is needed to identify and screen tech- variety lists so that varieties approved innologies from abroad. Many NARSs do not yet neighboring countries are automaticallyhave this capacity and indeed have been en- added to the approved list.)couraged by donor support and national pride * Rationalize regulations that restrict avail-to follow a "self-sufficiencw" policy with re- ability of agricultural chemicals, fertilizers,spect to agricultural R&D.' and pharmaceuticals by minimizing regu-

Governments can do much at the policy level lations designed to "protect" farmers andto encourage technology spill-in. Specifically, emphasizing those that address publiccountries need to minimize regulations that health and environmental concernsimpede the import of technologies, especially * Promote linkages to potential sources offor those embodied in production inputs, such spill-ins, such as IARCs, regional and inter-as seed of improved varieties (Gisselquist national networks, and advanced research1994). Thus governments can: institutes (for basic research).


For any one commodity, the choice of re- realize other indirect, but equally important,search strategy with respect to technology benefits from such international collaboration,development or screening of imported technol- including:ogy, will depend on the degree of local adap- . Staff development through exposure to ex-tation needed for imported technologies. It perienced researchers and other researchmay be most appropriate to import widely programsadapted varieties from the international sys- * New perspectives and approaches to ad-tem and focus local research efforts on crop dressing research problemsand resource management, which is likely to * Enhanced professional recognition in thebe more location-specific. The potential to re- global scientific communityalize spill-ins should be an integral part of re- * Access to advanced and specialized re-search priority setting exercises. search facilities

. Access to international sources of fundingLinkages with International for collaborative work.Research Institutions NARS linkages with international research

programs may be based on a variety of indi-Participation in regional networks and linkages vidual or institutional arrangements. Potentialwith national, regional, and international re- partners include public sector research insti-search centers can help to reinforce NARS re- tutes of industrial countries, private sector for-search program quality and can be mechanisms profit businesses, university researchto promote technology spill-ins (box 2.6). Most programs, development assistance-funded re-developing country NARSs are net importers search programs, NARSs within the region orof basic research knowledge from advanced re- in other developing countries, and the IARCs.search institutes in both developing and indus- Benefits from research linkages flow bothtrialized countries and from the IARCs. 1 The ways (Lele and Coffman 1995) and researchneed for linkages is often dictated by the small institutions in both the developed and devel-size or budget of a NARS, but linkages are of oping countries should have incentives for col-benefit to both parties (Lele 1995). laboration. However, costs in the initial

While technology spill-ins represent the formation of collaborative research activitiesmost obvious rationale for international re- are high and often hard to defend in relationsearch linkages, NARS research programs may to other budget priorities.

Box 2.6 Collaborative partnerships between IARCs and NARSs

The IARCs were created to generate spillovers by jointly (Maredia and Byerlee 1996). Recent moves toproducing international public research goods and develop consortia and two-way contracting arrange-services. They provide invaluable support for tech- ments between NARSs and IARCs are a useful movenology development by developing country NARSs.I in this direction.2

IARCs may often be low-cost providers of researchproducts due to their ability to reap economies of 1. LARCs are relatively more important to small NARSs.size, but the IARCs too are facing budget pressures However, the greatest absolute advantage is captured byand are being forced to focus programs and reduce large NARSs (Maredia and Byerlee 1996).some of their training, assistance, and technical ad- 2. The CGIAR renewal process and the Global Forum forvice and support. Discussions between the IARCs Agricultural Research recognize regional associations as theand the NARSs to exploit complementarities can im- desired mechanism for liaison between IARCs (and ad-prove resource allocation. Complementarity is fos- vanced research institutes) and the many individual NARSs.tered by joint decision-making on the types of This provides consultation and NARS input into developingproducts to be provided by each party, and also by research agendas and should assist in disseminating IARCthe fact that, in practice, most products are developed research results.


In summary, NARS linkages to the global re- grams, to joint technical meetings and re-search system are important to facilitate techview of work of common interest. Suchnology spill-ins. Extensive international sharing of information is relatively low-contacts and linkage arrangements between sci- cost and worth promoting to the fullestentists, often with support from donor funding, extent possible.will promote sharing of technologies, as well as * Integration of research programs throughefficiencies in technology development. A vari- joint development of research strategiesety of institutional mechanisms is needed to and plans, with different cooperating part-exploit international complementarities and de- ners undertaking specific aspects of thevelopment of such institutional linkages is a work, sharing results, or agreeing to use asignificant aspect of maturation of NARSs. For common facility (laboratory, station,problems and commodities not addressed by germplasm bank, equipment, or library).IARCs or advanced research institutes and in Regional research networks sponsored byregions with small countries, regional research some IARCs are examples of such jointinitiatives may play a greater role. planning and cooperation.

Joint funding of research in a fully inte-Regional Agricultural Research Associations grated program. The most ambitious

scheme of this type is the Regional FundIn recent years regional research collaboration for Agricultural Technology in Latinhas received renewed interest as a way of re- America, which will distribute funds foralizing efficiencies in NARSs, especially in re- regional research priorities through com-gions with many small countries (Eyzaguirre petitive grants (box 2.7).1996). Regional agricultural research associa- The level of activity of the RARAs andtions (RARAs) have evolved as formal organi- the heavy donor support in many cases posezations formed by NARSs (usually by the challenges to the sustainability of such ini-NARIs) to provide a structure for integrating tiatives:regional technology programs under NARS- * Over the long run RARAs must developdirected management and capturing the effi- strategies for funding and operation thatciencies offered by regional approaches.Regional associations have become more active Box 2.7 Latin American Regionalas both donors and NARSs face problems of Research Funddeclining budgets and as the increased maturitygives NARSs the confidence to seek regional A good example of a regional research undertakingcollaboration. RARAs have also been stimulated exists in Latin America, where the Inter-American

Development Bank and NARSs of Latin Americabyteusugnro om groare establishing a Regional Fund for Agricultural

and free trade associations. Technology, which will be operated under a re-Several new initiatives are directed towards gional endowment to provide funding through

strengthening and working with the RARAs." competitive grants for priority regional research ac-These initiatives represent a major shift in pri- tivities (IDB 1996). The fund will encourage andority to RARAs and imply a substantial in- finance strategic research of regional applicability

on a medium- and long-term basis, provide a forumcrease in support to and through these for discussing agricultural technology policy, facili-institutions. The regional collaboration and tate exchange of information and technology, andlinkages can operate at various levels of inte- facilitate representation of regional views in inter-gration, including: national forums. Priorities for competitive grants

* Informal and formal networks to share re- will be based on the potential to maximize regionalsearch results, ranging from exchange of spillovers. National, regional, and international or-

search results, ranging from exchange ganizations and private sector firms will be eligibleresearch publications and journals, to to submit proposals for funding.regular visits between different NARS pro-


overcome the current heavy dependence 4. An ISNAR study (Hoste and others 1995) foundon donor support. This dictates a need for that in Benin and Nigeria the number of scientists wasregional centers and networks to be kept about equally split between the governmental researchsmall, to operate on modest levels of fund- institutes and the agricultural universities. In Chile theing, and to seek sustainable sources of number of scientists and their qualifications are higherfinance. in the universities than in the NARI (E. Venezian,f Regionance initiatives must address high-pri Catholic University of Chile, personal communicationoRegional initiatives must address high-pri- 1995) ority problems in the region and not donor 5. By 1992 fully one-third of the maize breeders in

fads. Greater NARS control of the govern- Latin America and Asia (outside China) and over one-ance of RARAs combined with financial half of related research expenditures were in the pri-contributions by the NARSs themselves vate sector (Byerlee and Lopez-Pereira 1994).are necessary elements to ensure that 6. In the U.S. hybrid seed industry, public sectorRARAs address NARSs needs. inbreds accounted for one-half of the inbreds used up

• Regional research priorities are best devel- to the 1960s, some 30 years after the initiation of pri-oped in conjunction with national research vate sector R&D. In developing countries outsidepriorities (Spurling and others 1992). Re- China hybrids based on private R&D make up only 60gional programs should be complemen- percent of the sales of hybrid maize seed by the private

taryto and supportive of NARS programs, sector; the rest of the hybrid seed sold by the privatetary .t or °rt sector is based on hybrids developed in the public

andmstidntif appro t oera sector (Lopez-Pereira and Filipello 1994).roles and mechanisms to avoid becoming 7. For example, the United Brands Company andan additional layer of bureaucracy. the Honduran Agricultural Research Foundation,

To ensure the long-term relevance of re- the CIBA-Geigy Foundation and the Cinzanasearch programs, RARAs must broaden their Research Station in Mali, and Fundaci6n Polar inbase to include research participants other than Venezuela.the NARIs, such as public and private univer- 8. Larger endowments have generally contributedsities, private firms, and others. to a stronger organization, although there may be a

RARA development and sustainability will trade-off between a large endowment, which providesrequire confident leadership from the NARSs, independence and more modest funding that forcespatience, innovation in seeking sustainable fi- the organization to remain active in seeking additionalnancing, and stable support from donors. The funding and being responsive to clients.NAncings m . . 9. In some NARSs there have been jealousies be-NARAs must develop real ownAsrerh ofsthe tween relatively well-funded foundation programsRARAs and ensure that RARAs are a cost- and government programs, which may have a similareffective mechanism for executing research. mandate but are encumbered by limitations on salary

levels and operating flexibility.Notes 10. Only recently and mainly in Africa, where there

is a prevalence of small NARSs (24 out of 48 countries),1. For example, private companies may overinvest has there been much emphasis on regional research

in developing new pesticides relative to investment in initiatives (Spurling and others 1992; Weijenberg andintegrated pest management technologies and new, others 1993; Weijenberg and others 1995; and Taylorpest-resistant varieties. and others 1996).

2. Although this discussion uses the term "competi- 11. Industrialized countries are also major importerstive grant," the principles are equally valid for com- of agricultural technology.petitive contract research. The difference between the 12. The CGIAR Global Plan of Action to strengthentwo is simply the degree of control the funding agency global agricultural research is based on consult-exercises after the work is awarded. ation with regional forums of NARSs and includes a

3. Even in such systems, some recent evidence ques- strategy of strengthening these regional forumstions the productivity of CGs (Huffman and Just 1994). (RARAs).

CHAPTER 3

Strengthening Public Research Institutes:Issues and Good Practice

P ublic research institutes-the NARIs- the number of scientists (and sometimes theirare faced with the dual challenge of ob- salaries) have grown faster than the total re-taining increased funding support while search budget. The result has been manifested

improving their relevance and efficiency in severely reduced operating costs, fewer pur-within the evolving national and global re- chases, and less maintenance of capital items.search system. Most of the effort in reforming This has led to declines in staff morale andNARIs over the next decade must be focused serious threats to the productivity of many re-on making more efficient use of the available search systems. Coupled with the lack of fundsphysical and human resource infrastructure for is the problem of extreme volatility in funding.research. Relatively little effort should be ex- Uncertainty in funding is especially damagingpended on their further expansion.! in agricultural research because of the long-

The key issues for the NARIs are as follows: term and continuous nature of much research* Increase funding and diversify funding (box 3.1).

support The reasons for the funding crisis are varied* Undertake institutional and organiza- and include the following interrelated factors:

tional changes to allow more efficient and * Implementation of fiscal austerity leadingflexible management to cuts in general government budgets

* Enhance efficiency by improving methods . Undue dependence on a single source ofand capacities for planning, priority set- funds (such as annual appropriations fromting, and evaluation general tax revenues)

* Implement human resources management . Lack of understanding by decisionmakerssystems to attract and motivate scientists of the public-good nature of much agricul-and develop research leadership tural research and poor links between

* Strengthen the client orientation of re- NARS managers and policymakers re-search and expedite technological transfer sponsible for budget allocationsto users. . Lack of evidence of the results of research

or failure to communicate that evidence toResponding to the Funding Crisis policymakers

* Disenchantment of policymakers becausePublic research institutions in almost all coun- of the perceived inefficiency of public re-tries have suffered from a funding squeeze as search systems

33


Box 3.1 Funding reliability in the short term States is highly pertinent. There, investment bythe public sector in agricultural research grew

In many NARSs reliability in allocation of the ap- in real terms at 4 percent annually over a 100-proved research budgets is a serious problem. Ag- year period: even during the austere 1980s andricultural research is seasonal and late release of 1990s public funding for agricultural researchfunding can mean complete loss of a year's experi- continued to increase, although slowly (Fugliements and waste of funds already spent. Problems anotes19;AtnadPrey96)with dependability of fund releases are embedded and others 1996; Alston and Pardey 1996).in the general budget procedures of many coun- Research intensity in developing countries,tries: occasionally the problem is worsened by ad- averaging about 0.5 percent of agriculturalministrators' lack of appreciation for research. gross domestic product, is significantly lowerFailure to address financial management problems than in industrialized countries (more than 2can jeopardize the viability of research institutions. percent). This reflects the low tax base and theIn Sub-Saharan Africa this problem is especiallyacute and no general solution has been found. Prac- high share of agriculture in developing countrytices that can at least partially alleviate the problem economies. Poorer countries typically allocateinclude: improving the overall govemment budget an equivalent proportion of the agriculturalrelease system, advancing fund releases on a six- sector budget to agricultural research as domonth basis, giving greater autonomy to research richer countries (Elliott and Pardey 1988), butinstitutions (including flexibility in financial man-agement), streamlining procedures for approving spread over a relatively large agricultural sec-expenditures, and using revolving funds. tor, this results in low research intensity. How-

ever, the lower research intensity in developingcountries often implies a low priority for agri-

* Lack of political support for research from cultural research.farmers, farmers' organizations, and agri- Appropriate research intensity is a functionbusiness of the stage of development: it is not useful to

X A decline in agricultural commodity prices assign arbitrary quantitative targets, such asinducing a reduction in research funding.2 1-2 percent of agricultural gross domestic

The funding problem can be tackled at two product (AgGDP), for all developing countries.levels: by building political support for public The main task is to ensure that both real re-funding and by diversifying away from de- search budgets and research intensity are in-pendence on public funds by tapping alterna- creasing over time in a consistent andtive sources of funding such as those generated sustainable manner. At present, most countriesby the commercialization of technologies or by do not even comply with this more modestcost-sharing with farmers. Increases in donor objective.funding can provide temporary relief, but all Despite the clear rationale for public sectorNARI institutions must develop sustainable fi- funding, few countries have built an adequatenancing strategies based on local sources of political base for agricultural research over thefunding (Ellsworth 1997). long term. There is no universal recipe for de-

veloping such a support base, but better com-Building Political Supportfor Public munication of the needs and benefits ofSector Funding research is critical. Some potential elements of

a strategy to build a support base are:The first order of business must be to restore * Documenting and publicizing the impactspolitical support for public funding of agricul- of research. Although agricultural re-tural research. The current slowdown in fund- search is integral to achieving the nationaling should be seen only as a temporary objectives of food security, poverty reduc-aberration in what should be a long-term trend tion, and environmental conservation,toward increased public investinent in agricul- NARIs have generally done a poor job oftural research. The example of the United presenting the case in a form readily un-

Strengthening Public Research Institutes: Issues and Good Practice 35

derstandable to political leaders and ad- ing the importance of a well-functioningministrators. There is still little evidence of agricultural research system.the returns to specific types of research * Broadening the stakeholders in agricul-and, even in some large and relatively tural research beyond the traditional agri-strong NARSs, such as India, there have cultural base. Research programs canbeen few recent studies of research im- exploit funding from other ministries, suchpacts. Research institutes must put more as environment, science and technology,emphasis on advocacy and public aware- and commerce and trade. The recent in-ness for policymakers, farmers, and other crease in funding for environmental issuesconstituents. offers an opportunity for agriculture toDeveloping strong and articulate client or- persuade environmental groups that agri-ganizations that have a political voice and cultural research can contribute to solvingthat can act as lobby groups for agricul- environmental problems.tural research. While organizations ofcommercial farmers are common and in Commercialization of Research Productsmany cases contribute directly to research and Servicesfunding, organizations of small-scalefarmers with political influence are rare. There are alternatives for increasing funding ofYet farmer organizations have been impor- research in public research institutes throughtant in speaking out for agricultural re- commercialization of research products andsearch funding in most industrialized services, including: sales of research productscountries. Agribusiness and environ- and services (for example, sale of basic seed),mental groups can also be important sup- sales of nonresearch products and services (forporters of research programs. The recent example, soil tests), and various forms of jointdemocratization of the political system in ventures. The benefits of these options are con-many countries has encouraged the devel- tingent on revenue from commercialization be-opment of such organizations, which have ing returned directly to research institutionspotential in the long term to provide a (and being additional to their budgets) ratherconstituency for agricultural research. than being returned to the general treasury.Reforming the management and effective- The sale of research products from the publicness of public sector research institutions sector, through some form of royalties andto make these attractive to investors. While backed by adequate intellectual propertythis issue is addressed in detail elsewhere rights, has sometimes been successful in fund-in this paper, the lack of funding support ing research operating costs. For example, thein part reflects perceived inefficiencies and Brazilian maize program receives a large sharelack of impacts of research organizations. of its operating funds from the sale of basicDeveloping close relationships and good seed to private seed companies (Lopez-Pereiracommunication between NARI adminis- and Garcia 1994).trators and policymakers charged with Research systems can also generate revenuesbudget allocations. In countries where through sales of nonresearch products andlinks to budget officials are strong, re- services, such as soil and chemical testing, di-search budgets have suffered less. Having agnostic tests, sale of commercial seed andpolicymakers and clients of the research vaccines, staff consultancies, and even com-system on the boards of research institu- mercial agricultural production. Such activitiestions is one way to improve communica- can sometimes be justified, where there is sur-tion. For the World Bank, the ongoing plus capacity in research establishments, anddialogue with ministries of finance pro- revenues from commercial activities can helpvides an excellent opportunity for clarify- maintain the basic research infrastructure.


However, in the longer run, the most efficient sector institute should develop a clear andapproach will be for the public sector to sell off transparent policy for ensuring that such com-excess research infrastructure. mercialization is consistent with the public in-

Finally, public research institutes are increas- terest, applying intellectual property rights toingly entering joint ventures with the private protect technologies developed fully or partlysector either through a private company con- with public funds, providing free access to pro-tracting research in the public sector or through tected technologies to other public research in-agreements with private firms to adapt a re- stitutes, and sharing revenues between thesearch product from the public sector to spe- central administration of a research institute,cific markets, test a product widely, or the department that undertook the work, andundertake market development. Joint ventures the scientists generating the technology. Spe-that share the costs and benefits of R&D are cial caution is needed in commercializing non-being developed in many countries for genetic research products and services to ensure thatimprovement, seed production, plant propaga- this is a complementary, not competitive, ac-tion, and veterinary products. While currently tivity to the main business of research institutesaffecting only small segments of the overall (box 3.2).research system, such schemes are bound togrow as market mechanisms become more Farmer Financing of Researchprevalent in guiding agricultural development through Industry Leviesand R&D activities.

With the prevailing constraints in public Funding for research in public institutes canfunding for research and with the increasing also be generated through some form of indus-commercial orientation of agriculture in devel- try levy on agricultural output, or by contribu-oping countries, commercialization of public tions from producer associations, usually fixedsector research products and services will as a certain percentage of total sales of a prod-clearly be significant in future funding. This uct.4 From an economics perspective, researchalso represents a means for NARIs to serve the levies can be an efficient way to fund researchneeds of expanding agribusiness. A reasonable (Alston, Pardey, and Roseboom 1997) and areexpectation is that, in low-income countries, 10 also equitable since the cost of research is bornepercent of the budget of public sector research by those who benefit (that is, large producersinstitutions might be generated through these contribute more but also receive more of themeans, and up to 20-25 percent in middle- benefits). This form of funding is increasinglyincome countries. Although these funds can common for some commodities and for com-potentially provide an important share of mercial agriculture. It works best for commer-operating costs, they are often linked to specific cial products that pass through a concentratedtopics or commodities and to commercial ag- marketing channel making it administrativelyriculture and do not resolve the problems of feasible to collect the contributions or levy. Thefunding other types of research, especially for most common examples are for export crops,small-scale agriculture. where a marketing monopoly often facilitates

The main policy issue for such arrangements the collection of the levy. It is also feasible foris to ensure that public sector research is seen food crops produced commercially and withas being, and remains, motivated by broader strong producer organizations, and even forsocietal objectives and is not perceived as hav- basic food staples produced by small-scaleing "sold out" to industry. For this reason com- farmers, if the product passes through a nar-mercialization of technology should be row marketing or processing channel (for ex-approached with caution, as it can quickly dis- ample, a food marketing corporation or a fewtort program priorities and attention. Before large grain mills).5 It is least appropriate forengaging in commercial activities, each public traditional food crops (for example, cassava)


Box 3.2 Five key questions to ask on cost recovery in public research institutes

While there may be potential for generating signifi- services can quickly distract scientists fromcant additional funds from research programs (espe- their main task of developing new knowledgecially in areas where commercial agriculture is and technologies, and result in research institu-important), a number of critical questions have to be tions acting as state farms and businesses.faced in establishing a policy for commercializing re- * If technologies are to be protected by intellec-search: tual property rights, research managers must

* If the sale of research products is feasible and ask what is the cost of collecting revenues fromprofitable, why should the public sector be in- contracts and royalties relative to the funds gen-volved in the research in the first place? The erated? A research institute that seriously pur-private sector should be encouraged to take sues intellectual property protection will needover these areas of research directly. Indeed, to establish a specialized office with the appro-public sector participation in such activities priate legal and business skills for negotiatingwith support from the public purse may under- with private companies. The revenues raisedmine private sector entry into the R&D activity. from these sources often do not pay the extra

* If the public sector is motivated by financial costs incurred.rewards, will its research be directed to more * If research is successful in the sale of new tech-favored regions and farmers at the expense of nologies, should some of the funds received besmall-scale farmers and more marginal areas used as incentives for rewarding the scien-that may be the primary target of national pol- tists responsible for developing the product?icy?' Reliance on the sale of research products This provides a logical incentive mechanismwill also bias research away from crop- and for researchers, but immediately raises issuesresource-management research, where there is of equity and incentives for public sectorlittle opportunity to eam revenues from the sale employees.of research products.

* If a public institution sells nonresearch prod- 1. Recent sales of basic seed as a source of revenue for maizeucts, will this reduce research outputs? Sales of researchers in China and Mexico appear to be shifting incen-research products may complement research ef- tives away from research for the marginal areas that areforts, but the sale of non-research products and ostensibly targeted by government policy.

that pass through informal marketing chan- try representatives have a controlling or majornels. To be effectively implemented it should vote can be the vehicle for distributing funds.be initiated and supported by farmers who see This is the model used in Australia, where 16it as in their interest, as beneficiaries, to con- research corporations have been set up as part-tribute to funding agricultural research. How- nerships between industry and government toever, legislation is usually needed to ensure administer the research funds collected fromthat the levy is obligatory on all farmers and all major agricultural industries.to avoid "free riders." Where feasible levies for research should be

While farmer financing represents a valuable encouraged with appropriate institutionalpotential source of funds for public research means to ensure farmer participation in deter-institutes, its real value is only realized if farm- mining the level of contribution as well as set-ers' contribution to funding is combined with ting priorities for the research to be funded.farmers' participation in setting the research This approach will work best with matchingagenda, thus producing a more demand- govemment funds. This is because farmers willdriven research system. Where funds are col- typically focus on short term research prioritieslected through producer associations, the at the expense of more basic and long-termassociation can determine or influence the re- research, or research with broader societalsearch priorities for expenditure of the funds. benefits (for example, protection of the envi-In other cases some type of research board or ronment). Government provision of matchingcorporation in which farmers and other indus- funds can be accompanied by scientists, poli-


cymakers, and the wider community joining research strategies, long-term support for tar-farmers in determining research priorities, to geted research programs, development ofrepresent these long-run and broader interests. mechanisms for sustainable financing for re-

search, more effective use of existing resources,The Role of Donors and International careful prioritization of donor-funded inputsDevelopment Banks (especially of costly technical assistance), and

flexible linkage of research funding to institu-The NARI funding crisis raises difficult issues tional and policy reforms. All of this requiresfor donors who have been important contribu- strong NARI leadership (Byerlee and Alextors to the growth in investment in agricultural 1997). This suggests that donors should sup-research over the past three decades. Initially port research by using every opportunity to:donor contributions were largely limited to * Shift from project funding to providingcapital development. 6 However, because do- block funds or competitive grants for capi-nor-driven expansion of the research system tal and operating expenditures for agreed-often resulted in a large physical and human on high-priority research activities. Thiscapital base that could not be maintained or system is being tested in "Frameworks forused effectively because of shortage of operat- Action" between donors and NARSs ining funds, donor contributions soon extended some African countries, including Malito operating budgets.7 Donor support to agri- and Tanzania (SPAAR 1996).cultural research is most pronounced in Africa, * Change focus from investing in capital-where donors accounted for nearly one-half of development projects to stimulating insti-the total research budget of NARIs in 1990 tutional changes that improve the use of(Pardey, Roseboom, and Beintema 1997). existing resources; this is already under

Such extensive donor support has been criti- way in many countries.cized for its unintended negative impacts on * Emphasize projects that have an explicitNARI development. Some potential pitfalls objective of increasing local funding forfrom donor financing of research systems in- NARIs. These might include developingclude: creating donor dependency, causing farmer organizations, sponsoring impactfragmentation of research efforts, creating lack studies, and developing mechanisms forof continuity in programs, increasing require- tapping private industry funding.ments for "counterpart" funding for recurrent * Foster greater private sector R&D invest-costs, making ineffective use of some donor- ments through supporting policy reformsprovided inputs, and failing to link research aimed at encouraging private initiative infunding to policy reforms (Byerlee and Alex all sectors, and through projects specifi-1997). It also has been argued that NARS man- cally targeting private-sector R&D (for ex-agers find it easier to obtain funds from donors ample, support to establish an enablingthan to develop political support for local legal and administrative environment).funding especially in NARIs where donor sup- * Encourage dialogue between ministries ofport covers the major part of the national re- finance and the managers of the NARIs tosearch budget (Trigo 1987; Rukuni 1996). create more sustainable funding arrange-

Donor funding of agricultural research pro- ments.grams will undoubtedly remain an important In summary, donors, including developmentelement of development assistance programs, banks, must give explicit attention to sustain-as research investments offer high rates of re- ability of research funding. This implies reduc-turn and provide mutual benefits to donors ing total dollar contributions and givingand developing countries. To minimize the greater emphasis to improving research qual-negative effects of donor support, assistance ity as opposed to expanding programs. Second,should be based on well-developed national donors must closely coordinate their activities


with each other and with NARS priorities to cial needs of research. It also provides the flexi-ensure that the support is complementary and bility needed for financial management, becauseaddresses key strategic technological and re- research operating costs as a proportion of totalsearch management issues. budget are higher tian for many other govern-

mental departments, and costs and types of ex-Organizing National Agricultural penditure are more variable over time.Research Institutes While many agree in principle on the needfor Increased Efficiency for autonomy in research organizations, there

is little agreement or understanding of whatTwo dimensions of research organization pro- this means in practice. The degree of autonomyvide the context in which efforts to increase granted may vary from delegating research or-efficiency must be made. These are the degree ganizations limited power on financial and hu-of autonomy of research organizations and the man resources decisions to full autonomy indegree of centralization. research governance that allows the organiza-

tion to set and change its own rules of business.Autonomy of Research Organizations A research organization may be completely

privatized (as in New Zealand), even thoughIn many countries the pursuit of science in the much of its funding still comes from the publicpublic sector within the rules and regulations sector. Table 3.1 provides a summary of theof the civil service has resulted in low produc- various administrative, financial, and per-tivity of scientists. Given the special require- sonnel decisionmaking powers that need to bements of research (box 3.3), there has been a considered in granting autonomy to a re-worldwide move to grant greater autonomy to search organization. In practice, in setting upresearch organizations outside civil service "autonomous" research organizations, only aregulations. Autonomy is expected to give in- few of these powers have been delegated tostitute management and staff the flexibility these organizations.needed to tailor business rules and regulations Institutional autonomy (or semi-autonomy)and human resources management to the spe- may be provided to an apex research policy

Box 3.3 Why is research different from many public sector activities?

* Research is a creative process with uncertain gency, can wipe out years of experimentaloutcomes and cannot be micro-managed from work.day to day. Good scientists require flexibility . Research managers need flexibility to shift re-and independence over a long period to achieve sources among the major budget categories ofresults. operating costs, capital equipment, and salaries

* Recruitment and promotion of scientists re- to ensure overall efficiency and adequate oper-quires different standards than those for civil ating costs.service employees. Scientific skills are highly . Research institutions require flexibility to diver-specialized and scientists require opportunities sify their funding support by soliciting fundsfor advancement in rank and salary within their from various ministries, the private sector, orspecialized areas. In a competitive international internationally, and by commercializing re-market special incentives and rewards are search products.needed to attract and retain the best scientists. . Diverse stakeholders (government, producers,

• Research often requires lumpy recurrent and agro-industry, and the broader scientific com-capital costs (for example, setting up an experi- munity) should be active in setting the researchment) that demand considerable flexibility in agenda. When research is controlled and man-financial and procurement arrangements. A aged directly by line ministries or departments,broken irrigation pump that is not fixed because many stakeholders may be marginalized fromof rigid procurement rules or financial strin- these processes.


Table 3.1 Definition of "autonomy" in terms of the types of flexibility required for various management functions

Function Flexibility needed

Govemance

* Board membership and appointment of chief Selection of Board members in their personal capacity, development and approvalexecutive officer of rules of business, and open and transparent selection of a chief executive officer

free of political intederence.

Research programs

* Program formulation, funding, and Development and approval of annual research plans, allocation of research bud-implementation gets, and monitoring and evaluation of research activities.

Personnel management

Hirng and firing, incentive and reward Independent, transparent, and merit-based selection of research leaders andsystems, and human resources development scientific staff, performance-based salary and promotion, dismissal of unproductive

staff, and merit- and skill-based selection of staff for training.

Administrative procedures

* Procurement of supplies, equipment, and Independent procurement of supplies and equipment with internal accountabilitypersonal services mechanisms and flexibility to employ short-term consultancies.

Financial management

* Funding, disbursement, and asset Flexibility to obtain funding from any source (including borrowing), retain and usemanagement generated income and acquire and sell capital assets, and set procedures for

disbursement.

Intemational collaboration

* Negotiation of agreements for collaborative Power to enter into agreements with international institutions for researchresearch and donor assistance collaboration and scientific exchanges.

Source: Nickel 1997.

advisory committee, an apex research funding leading to tensions with the civil service overentity, a research executing institution, or a which rules always apply.combined funding and executing agency. The Autonomous research bodies must recog-general trend has been toward forming semi- nize that, in return for freedom to manage theirautonomous apex research councils to provide financial, human and physical resources, theyindependence in strategy and programming, must be held accountable in delivering re-or to create autonomous research organiza- search products. This requires medium-termtions or corporations that combine apex poli- and annual plans that include specific meas-cymaking roles with research execution. In ures of research output and indicators of per-practice, many of these so-called autonomous formance that can be used by the governingresearch organizations have not lived up to ex- board and funders to evaluate the organiza-pectations because governments have been re- tion. In return for freedom in day-to-day man-luctant to delegate real powers and because agement, research organizations will have tomany scientists themselves feel more secure devote more time to planning and assessmentworking under civil service regulations. Even than they have done in the past.research organizations that have been given Granting of autonomy to research organiza-considerable autonomy have often not used tions must start with governance of the organi-these powers so that business rules differ little zations. Few research organizations havefrom those in the civil service. In other cases succeeded in establishing the professionallybusiness rules have not been well defined, competent and independent Board of Directors

Strengthening Putblic Research Institutes: Issutes and Good Practice 41

basic to functional autonomy. The usual prac- * In establishing autonomous bodies, experttice has been to appoint ex-officio members advice in public sector management is re-representing government departments, with quired to carefully define and agree onthe minister of agriculture as chair. rules and regulations for managing finan-

The most successful examples of autono- cial and human resources and capital as-mous research organizations have created re- sets. Granting autonomy in some areassearch corporations that fit all or most of the without flexibility in other areas will se-requirements of table 3.1. These corporations, verely constrain the operation of the newwhich now exist in countries as diverse as New organization.Zealand, Colombia, and Uruguay, have . Autonomous bodies need well-definedadopted business practices for financial and mechanisms to ensure accountability topersonnel management, and salaries, similar to those who fund them. The mechanismsthose of private companies. Their creation has should define specific outputs over a givenoften been associated with a significant down- period with built-in milestones that can besizing of staff, made possible by increased monitored at regular intervals to assessflexibility in personnel management. progress.

The privatization of research in the form ofresearch corporations offers much potential for Centralization of Research Fundingimproving the performance of agricultural re- and Executionsearch institutions, especially for research exe-cution. However, it represents a radical Public sector research in many countries hasdeparture from the conventional view of public typically been highly centralized with the de-sector research and experiences with this gree of centralization tending to increase overmodel should be carefully monitored as a time with the formation of a large national re-guide to implementing it more widely.8 In the search institute (that is, the NARI) and centralmeantime, the following guidelines represent research institutes for particular commoditiesthe current state of knowledge in granting or agro-ecosystems. Another favorite vehicleautonomy to research organizations: for centralizing and coordinating research ef-

* The respective tasks of autonomous or- forts has been the national coordinated re-ganizations in research execution and search programs, usually organized alongfunding and in research policy should be commodity lines.clearly defined. One rationale for centralizing research is the

. The composition of governing boards, in- potential for economies of size and scope, aris-cluding the Board chair, should shift from ing from high fixed costs for much research,institutional representation toward ap- specialization of scientific expertise and equip-pointment of professional members who ment, and opportunities for team interaction.are respected scientists and stakeholders But several factors may lead to diseconomiesappointed in their individual professional of size in research, including higher transactioncapacity. costs in conducting research for a larger and

* Autonomous organizations should be more dispersed mandate area, lack of compe-linked to the different stakeholders, who tition among rival research programs, and in-should be active in its governance and fi- creased institutional and technological risk (fornancial support. example, genetic uniformity).

* Alternative mechanisms, such as policy The potential for economies of size in re-advisory committees, should be explored search is largely an empirical question. Becauseas these can provide policy input into set- of the substantial overhead cost of experimentting priorities for autonomous research stations, libraries, and laboratories, the majororganizations. factor determining economies of size is the


market size over which a technology will be management of national research networks,applied (Byerlee and Traxler 1996). This argues while much of the applied and adaptive re-for centralizing those research activities which search increasingly becomes decentralized andproduce technologies with wide applicability administratively independent in regional re-(for example, much basic and strategic research search centers in which farmers are involvedand some applied research such as varietal im- in funding and priority setting.provement). For example, national coordinatedprograms have been effective for conducting Institutionalizing Research Planningapplied crop-improvement research over di- and Evaluationverse agro-ecosystems. However, crop- and re-source-management research tend to be more In the past decade many NARIs, usually withlocation-specific and there is little indication World Bank or other donor support, have un-that centralizing this research is effective. dertaken strategic planning and priority-set-

Another issue involves centralizing funding ting and introduced more formal methods forof research-especially the relative share of re- evaluating research projects before fundingsearch to be provided by national, state, and (Odame and Setshwaelo 1997). Typically, thelocal governments. The main rationale for cen- results of this work are embraced in a mastertralized funding is that state and local govern- plan for the NARI, which details overall re-ments have few incentives to fully fund search system strategy, including details on re-research, which generates spillovers beyond search priorities, organizational structure,their jurisdiction. However, research that has physical facilities, staff development, and insti-mostly local applicability, especially adaptive tutional linkages needed to implement theresearch, can and should be funded locally, strategy.where possible. For adaptive research, thereis a strong case for decentralizing to the dis- Research Planning and Priority Settingtrict or county level, with local governmentspicking up a share of funding. There is also Research "master plans" have often been re-much potential for beneficiaries to fund quired as a condition for a project loan oradaptive research and participate in research credit. A recent survey of 37 countries in Sub-priority-setting. The closer that research is to Saharan Africa (Maredia and others 1996) re-the user, the more likely it is that users can ported that 18 had completed research plans,participate in this way. These considerations 12 had plans under preparation, and threesuggest that some advantages of economies were going to prepare plans. However, despiteof size and market size for centralized research the proliferation of master planning in NARSssystems will be offset by the potential for user and the apparent benefits in terms of donorparticipation in a more decentralized research project implementation, the impact on researchsystem. efficiency and funding has been less than ex-

In summary, the optimum organization of pected. Several reasons account for this:research will be a complex tradeoff of effi- * Strategic planning and priority settingciency issues, duplication versus competition, have often been seen as being imposedinstitutional pluralism versus central control, from outside and in too many cases, masterand centralization versus proximity to and in- plans have been developed by outside con-teraction with users. The optimal organization sultants with little local input (Ravnborgof research will probably be different for fund- 1993; Anandajaysekeram and Rukuniing research and for executing research: these 1994). Most NARSs lack the analytical ca-will be increasingly separated in the future. pacity to undertake planning and priority-Centralized NARIs will probably become more setting work as a continuing activity. As aspecialized in basic and strategic research and result, most exercises have been one-shot


affairs with a tendency for the plan to be and outcomes, and adjustment to strategy andforgotten once the donor has signed off. action plan, as appropriate. Major stakeholders

* Too much attention has been placed on should be involved as full partners at each stepdeveloping quantitative estimates of re- in the process to increase the quality of infor-source allocation within the public-sector mation used and to help build political supportNARI, at the expense of analysis of broader for the strategy as well as for the research sys-subsectoral and science policy issues, in- tem (Ricks and others 1997a).cluding postharvest issues, the potential As NARIs develop more experience in plan-role of the private sector and its links to the ning and priority-setting, there will be morepublic sector, and the opportunity to im- examples of good practice. Experiences overport rather than develop new technologies. the past decade provide lessons that can beAs a result most master plans, despite the applied in developing capacity in future prior-range of detail on the research system, lack ity-setting exercises to address weaknesses invision for the future development of the the master planning approach:whole technology system for the subsector * Analytical work to guide research plan-(Boughton and others 1995). ning is best undertaken from within the

* Nearly all analysis of research priorities research system, preferably by a small unithas been carried out at the macro level to with ready access to senior research man-analyze priorities across programs. How- agers (Alston, Norton, Pardey 1995). Thisever, research decisions are made opera- implies that donor support for priority-set-tional at the project level and there has ting should shift from providing up-frontbeen no effective mechanism to move from consultancies for master planning effortsthe top-down program planning to the bot- to developing institutional capacity build-tom-up annual cycle of project formulation ing within NARIs. This will necessarily befor funding. As a result, the funded pro- a longer-term exercise, but will result injects in aggregate are often inconsistent more effective ownership of the results.with macro-level priorities. * The establishment of an analytical unit for

- It has proved very difficult to make the research priority setting should be accom-hard decisions to close down unneeded panied by strong commitment of researchfacilities and reallocate staff and budgets managers to ensure that results will be con-in line with priorities, so that priority set- sidered in decisions on research-resourceting has often not been translated into allocation and that, in fact, research re-practice.9 sources shift in ways consistent with iden-

Success in research-planning and priority- tified priorities.10setting depends much on the processes for un- * Analysis should be broadened beyonddertaking the planning. In the past too much mechanistic allocation of resources by pro-emphasis was placed on refining the analytical gram to include work on linkages betweentechniques employed. Institutional processes agricultural policies and agricultural re-for research planning should include the fol- search policy, research resource allocation,lowing general steps (Collion and Kissi 1995): and the economic evaluation of the effectsassessment of the external environment; as- of previous research. Analysis should takesessment of the current status of the organiza- a comprehensive view of the R&D system,tion, projection of desired future state of the including the rapidly growing participa-organization, analysis of the gap between ex- tion of the private sector in many researchisting and desired state, determination of a areas, which may influence the priorities ofstrategy to close the gap, formulation of an ac- NARIs.tion plan, implementation of the action plan, . The process should be participatory in-monitoring and evaluation of implementation cluding a wide cross section of scientists,

44 Strengthening National Agrictltural Research Systems

and the main stakeholders-policymak- little effort to estimate expected costs anders, research partners, and clients. At lower benefits for non-commodity programs,levels of planning and priority-setting, or to assess resource allocation over agro-participation of users should increase rela- ecological zones. In practice the latter re-tive to participation of policymakers, so source allocation decisions are often morethat in the design and selection of research strategic.projects, the main interaction should be Economic analysis can also be applied at thebetween scientists and farmers. At the project level to estimate ex-ante cost-benefit ra-project level successful design of projects tios as a basis for selecting projects. This ap-depends more on using relevant participa- proach is being applied in some NARIs, and intory process and appropriate scientific Australia is a requirement for project submis-method than on the employment of formal sion for many funding agencies. At this stagepriority-setting methods. in the development of analytical capacity in

priority-setting, the use of ex-ante cost-benefitThe Role of Economic Analysis analysis at the project level should be ap-in Priority Setting proached cautiously because:

. Research is an inherently risky activityIn recent years economic methods for research with outcomes that are difficult to estimatepriority-setting have evolved rapidly and can ex-ante.now provide valuable information to decision- . The impacts of some types of research aremakers for more effectively allocating research difficult to measure or value (for example,resources to meet national objectives. basic research or research on natural re-

Master planning exercises typically employ sources management).some form of economic model to analyze re- * The success of a research project dependssearch resource allocation. However, there critically on its scientific merit as well as onhave been a number of difficulties with the its relevance to solving real problems. Un-approaches used: due emphasis on ex-ante economic analy-

* The methods used for analysis of research sis may detract from efforts needed inpriorities have often been inappropriate, in areas such as peer review and client inter-part because of lack of consensus within action.the economics discipline on appropriate * Resources required to conduct a rigorousmethods. In particular, weighted scoring economic analysis of a research programmethods have been widely used, even or project are considerable if the results arethough the criterion variables and the to be credible to scientists and funders.weighting methods have sometimes Economics skills in most NARSs are scarcehad little, if any, basis in economic theory and the opportunity costs of employing(Alston, Norton, and Pardey 1995). these in economic analysis of projects rela-

* There has been a lack of data to compare tive to other activities (for example, eco-actual resource allocations to indicated pri- nomic evaluation of technologies andorities, or to monitor shifts in resource al- diagnostic surveys) are high.locations over time in relation to the In summary, economic analysis in setting re-priorities. Without information on current search priorities should be encouraged at theresource allocations, it is impossible to program level and should be broadened to in-know the direction in which resources clude noncommodity research. NARIs need toshould be shifted, even after the formal develop the capacity to conduct such analysispriority-setting exercise is carried out. on an ongoing basis, rather than as a one-off

. In most cases analysis has focused on re- exercise. An important starting point is to de-source allocation across commodities, with velop a project information system to track


current resource allocations (both financial and Indicators of research outputs are used forhuman) and monitor shifts in resource alloca- both monitoring and evaluation. These may betion in accord with priorities. At this stage in measured at both the project and program levelthe state of the art, economic analysis has not and will be different for basic and applied re-been shown to provide added value in select- search. At the planning stage both researching individual research projects. projects and research programs need to iden-

tify the expected outputs of the research as wellMonitoring and Evaluating Research Programs as the intended users or beneficiaries of the

research results.In this era of tight budgets, NARIs increasingly Finally, impact indicators are largely used forrequire performance monitoring and evalu- evaluating research programs. Impact indicatorsation (M&E) systems that highlight research are rarely practical or necessary at the projectprogram impacts on national policy objectives level, but should be applied to the program leveland provide a basis for assessing progress and at regular intervals, although, because of themaking needed adjustments to programs and long-run payoffs to much research, these inter-policies. Through monitoring (the routine col- vals should usually not be less than five years.lection of data on program performance and Impacts may be measured up to the system level,program adjustments to identify problems) for example, through studying the economic re-and evaluation (the analysis of monitoring data turns to overall research investments (for exam-and system performance), NARI leaders and ple, econometric estimates of research impactsadministrators can track research performance on changes in total factor productivity). Researchagainst plans and provide feedback for revis- funders sometimes call for even more in-depthing plans and strategies. M&E also provide the evaluation of agricultural research on nationalbasis for measuring accomplishments and de- policy objectives, such as food security or pov-termining the success of program activities erty alleviation. However, because of the multi-(Murphy 1993). tude of factors influencing these objectives, this

Progress toward the desired objectives is is difficult to do in practice (box 3.4).measured by indicators of program progress. In summary, the priority information re-Given the difficulties of measurement, it may quirements for M&E are as follows:not be possible to quantify some program out- . Within a program, each project proposalcomes, and qualitative or intangible indicators should include a few simple monitorablemay have to be used. Table 3.2 provides a sim- milestones that quantify implementationple framework for thinking about an M&E sys- progress as well as progress in achievingtem for agricultural research. Different types outputs. Progress against these milestonesof indicators are applied at different frequen- can then be measured in annual progresscies at each level of a research system. Process reports, perhaps supplemented by a veryindicators are largely for monitoring individ- brief six-month report.ual research projects and are the building . At the program or institute level, each pro-blocks for monitoring overall progress in im- gram should define long-run strategiesplementing an agreed-on research program. and a five-year work plan that providesProcess indicators should be quantified as measurable indicators of implementationfar as possible. Good practice requires that of the research program and of researchresearch proposals include milestones that ex- outputs. Performance against these indica-plicitly lay out specific progress in implemen- tors can then be measured annually as parttation. A relevant example is provided by the of the reporting requirements for projectmilestone indicators required in projects and program leaders.funded by the Research and Development Cor- * Regularly-about once every five years-porations in Australia. each program should have a comprehen-


Table 3.2 Summary of monitoring and evaluation indicators and their implementation in agricultural research

Type of indicator Examples Level at which applied Frequency of implementation

Process indicators for monitoring implementation performance

Inputs Funds expended, scientist time Project level, with aggregation to Annually, with six-monthused programs, institutes, and systems progress updates

Outputs Tnals conducted, crosses made, Project level, with aggregation to Annually, with six-monthsurveys completed, vaneties and programs, institutes, and systems progress updateother technologies developed,and recommendations made

Impact indicators for monitoring and evaluating research

Research outcomes (for Applied research: Tech- Project level and program level, On completion of a projectmonitonng and evaluation) nology adoption rates and with aggregation to institutes (for example, after three

publications and systems years)

Basic and strategic research: Annually at the programNew research methods, tools, level (for example, annualtechniques, hypotheses report)supported or rejected, andpublications More in-depth at fixed

intervals (for example,external reviews ofprograms)

Research impacts Applied research: Yield and Program, institute, and Once every five years for(for evaluation) production increases, cost system levels each program

reductions, economic return onresearch investment, impact ontotal factor productivity

Basic and strategic research:Use of new knowledge toincrease the efficiency andefficacy of applied research

sive external review of technical aspects * Designing a simple system that mini-and an analysis of its impacts. mizes data collection to a few critical

All research organizations must have an in- variablesstitutional capacity for M&E and for feeding . Using a variety of mechanisms besidesthe results back into decisionmaking. Building quantitative indicators as an integral partsuch a capacity requires only a modest invest- of M&E. These include planning and re-ment and should be a priority for NARIs. The porting workshops, field visits, externalmajor problem in M&E systems for research technical reviews of research programs,has been the tendency to collect too much in- and regular visits to observe laboratoriesformation in a highly centralized and bureau- and experimentscratic manner. To date, there are few examples * Decentralizing the implementation as farof good practice in this area, but the key ele- as possible. For example, project monitor-ments of successful institutional capacity ing should be decentralized to program orappear to be: subprogram leaders.


Box 3.4 Agricultural research and poverty reduction

Agricultural research is increasingly called upon to address particular needs of women or disadvantagedaddress equity issues, especially poverty reduction. groups.This might be done through targeting research to In many cases policymakers and donors call forregions and commodities important to the poor, or more research on marginal areas as a way of reducingthrough developing technologies that favor the poverty and arresting resource degradation. Whilepoor-especially labor-intensive technologies. How- this may be appropriate, the situation is usually muchever, the direct link between agricultural research more complex. In some of the most difficult areas,and poverty reduction is complex. Research that gen- the rate of gain from improved technology may beerates broad-based productivity increases is one of very slow. Providing infrastructure, education andthe most effective means of reducing poverty through nonfarm opportunities may bring earlier, greater andrural employment and income generation. One major more sustainable benefit (to both poverty alleviationimpact of agricultural research is increased availabil- and conservation of natural resources) than is possi-ity and lower prices for food, this is especially bene- ble with agricultural intensification in many areasficial to the poor, who spend a large share of their that have very high production risks and/or a fragileincome on food. The main emphasis in agricultural resource base. (Of course, research may well be aresearch systems should be on broad-based produc- priority for such areas for reasons other than produc-tivity enhancement in general. As a secondary con- tion intensification.) Finally, there are often substan-sideration, research orientation can be modified by tial spillover benefits from productivity gains inchoosing particular regions or commodities that offer favored areas through the operation of labor marketsstrong prospects of direct and sustainable pro-poor (that is, migration from marginal areas) and foodimprovements, including research on staple foods markets (that is, lowered cost of food) (Renkow 1993;consumed by the poor. Research may also directly David and Otsuka 1994).

Establishing a very small unit to promote Human Resource ManagementM&E, provide training and develop stand- and Developmentards, but undertake M&E only at the macrolevel Problems with human resource management

* Ensuring that there are mechanisms for in NARIs can be grouped into three sets ofresearch managers to receive timely infor- interrelated issues: administration, incentives,mation from the M&E system and take and training. Administrative problems centercorrective actions around the problem of maintaining a produc-

• Contracting some M&E activities, espe- tive scientific environment that encourages sci-cially impact studies, to independent agen- entists in innovation and independent thoughtcies or individuals. and initiative. Management approaches must

be more conducive to such an institutional en-Upgrading Human Resources vironment than are the hierarchical structuresManagement found in many institutions (Grindle 1997).

Lack of development of a "science culture" dis-Ultimately, good research depends on the courages inquiry. In some systems few scien-quality of human resources of the NARIs. tists are given real opportunities beyond theirProbably the most important element in build- graduate studies to seriously practice science.ing NARI capacity is to strengthen the man- Scientists are often frustrated by bureaucraticagement of the human resource base (Eicher requirements established to prevent real or1990). Evidence that the quality of the human perceived abuses (for example, numerous sig-resource base may be declining in some re- natures to authorize small purchases, oneroussearch systems is cause for alarm. restrictions on the use of vehicles, or inability


to obtain approvals for travel to international tists the opportunity for salaries, benefits,meetings). and prestige equivalent to those enjoyed

There are no easy solutions to these prob- by managers and administrators. Such ar-lems, although autonomy from civil service rangements are sometimes known asrules is often sought by research institutions science-service schemes and are widely into provide needed management flexibility. place in industrialized countries.Management reviews of institutions by out- * Putting emphasis on professional develop-side experts or management firms can also ment of young scientists, including men-help to identify bureaucratic bottlenecks and toring by leading researchers.establish the need for training in research * Decentralizing budgets and project man-management. agement to individual scientists. Manage-

The second key element in human resources ment skills-learnt through long experi-management is an incentive system to encour- ence-can be fostered by providingage and reward researchers. Most NARIs still independence to scientists, once prioritiesfollow rigid civil service rules, which evaluate are agreed upon and there is an adequateand reward researchers on seniority rather monitoring and evaluation process inthan merit and which do not require account- place to ensure accountability. One way toability from scientists and research managers. encourage independence and research en-Few research institutions have good mecha- trepreneurship is through competitivenisms for personnel evaluation, whether meas- grants awarded to individual scientists orured in research outputs or the adoption of teams, rather than to institutions.these outputs by farmers and other clients. Sci- . Developing opportunities for consult-entists generally assert that they want merit- ancies by scientists to boost earnings andbased incentive systems, but in practice obtain wider professional experience. Afrequently appear more comfortable with tra- sensible balance is required if institutionalditional systems based on job security and productivity is to be significantly and sus-equal pay for all. Politics and economic reali- tainably boosted. Such consultancies canties often thwart any move to establish special be efficient vehicles for transfers of tech-salary schedules for researchers. Civil service nologies (and even policies).rules, labor laws, and local attitudes limit flexi- . Providing funding for sabbaticals, externalbility for introducing performance-based in- travel, training, and greater interactioncentives, and for dealing with poor performers. with the international scientific commu-

Some of the most effective incentives to en- nity to overcome the intellectual in-breed-courage good research can be nonmonetary ing common in many institutions.and can include altruistic motivation; profes- Third, training will continue to be importantsional recognition by peers; or the satisfaction for NARIs to maintain their human resourceof scientific interest and curiosity. Nonetheless, base for productive and sustainable researchNARIs need to continue to work to provide programs. Future emphasis will be on trainingresearchers with adequate, competitive salaries for replacement staff and for upgrading skillsand other incentives to reward output. Man- to enable researchers to stay current in theiragement tools that can increase the effective- disciplines and in emerging fields, such asness of human resource use include the biotechnology, communications technologies,following: and systems modeling. Meeting the costs of

Revamping performance appraisal and even modest foreign trairiing programs will bepromotion systems. Research institutions an issue for many countries, and the develop-need regular staff appraisal systems that ment of local training capacities will be essen-recognize merit and give research scien- tial to system sustainability.


Research Program Leadership Downsizing and Consolidationand Management

Downsizing through reducing and consolidat-Improved leadership and sound management ing research infrastructure and staff is one po-are key requirements for effective research in- tential solution to the problems of low scientificstitutions (Rukuni 1996; Nickel 1989). Leader- salaries and acute scarcity of operating costs,ship gives direction to an institution and its provided that budget savings are used forprograms, motivates staff, and represents the these purposes. Some argue that downsizing isinstitution with outside stakeholders. Sound inconsistent with the findings of high returnsmanagement ensures that what is done is done to investment in agricultural research, imply-efficiently-planning, conducting research, ing that underinvestment in agriculturalprocurement, and reporting. research must be prevalent. However, down-

It is not always clear what constitutes good sizing relates specifically to staff numbers andleadership and good management; nor is it infrastructure, rather than to budgets. The staffclear what can be done when these are lacking. targeted are generally not senior scientific staff,Changing leadership may be the only option, but support staff, who constitute a high pro-but even this carries no guarantee. In fact, portion of total staff in many NARIs and whosuspicions of political motivation in change are increasingly becoming redundant becauseof leadership, long-term power struggles, long of improvements in office technology, labora-vacancies in key positions, or frequent changes tory procedures, or mechanized experiment-in leadership positions may worsen a bad station operations. In some cases (for examplesituation. in much of the former Soviet Union), national

Few are truly born leaders, and leadership systems have expanded to a size well beyondand management skill training can be impor- national needs and affordability and, therefore,tant to research systems. For some, the major a reduction in scientific staff will also beneeds will be in human resource management needed. Having a personnel monitoring sys-and development. For others, training priori- tem that permits ready identification of non-ties may be in strategic planning, project performers is a minimal requirement forplanning, proposal writing and project man- rational decision making in this regard.agement, or financial management. Unfortu- Even where the returns to agricultural re-nately, this training is often less valued than search are high in the aggregate, there may betechnical training and may be considered by good opportunities for reductions in specificpotential trainees as an admission of weakness. programs. It is not difficult to find researchNARIs need to institute regular management programs that have been unproductive fortraining and leadership development pro- many years and that merit hard scrutiny. Ifgrams as an integral part of human resources there is an alternative supplier (for example, indevelopment. the private sector or abroad), the foregone re-

Although training is important in develop- turns from eliminating the program may being leadership skills, it is also essential to en- quite small. Finally, there is frequent overlapsure that the right people get appointed to top and duplication in research programs, or pro-positions. Transparent and open recruitment grams that target small areas or crops of lim-processes have to be in place to appoint re- ited significance. Recent work suggests thatsearch leaders and senior managers. This is many of these programs are inefficient users ofoften difficult to carry out under standard civil valuable research resources (Maredia and By-service rules, and progress in achieving insti- erlee 1996; Traxler, Byerlee; Jain 1996). In mosttutional autonomy may be part of the solution NARIs it is not difficult to find programsto the leadership problem. whose marginal contribution is small.


Consolidating research infrastructure is an- pecially small-scale farmers and farmers inother aspect of downsizing. Although there has marginal areas. Failure to involve the main us-always been recognition of the need for a "criti- ers in the research process led to several inef-cal mass" of scientists in a research program, ficiencies: failure to adequately identifythere have often been political or donor-in- problems and priority technology needs, exclu-duced drives to open new research stations, sion of an important source of innovations-programs, or laboratories. NARI expansion has the farmers themselves and their indigenousleft many countries with more stations, field knowledge; inability to access farmer fundingoffices, and laboratories than they can main- for research; and delays in the disseminationtain. Station or program consolidation will of research results.then usually be part of staffing reductions. This led to a strong emphasis on farming

There are efficiency arguments for consider- systems and on-farm research (FSR) during theing downsizing as an option in many NARIs. 1980s, much of it with World Bank support.To avoid irreparable damage to staff morale, The major contribution of this movement wasthe downsizing process should be as transpar- to provide methods for diagnosing farmers'ent as possible, and be conducted within a problems, setting research priorities to addresswell-defined NARI strategy and research plan. those problems, and screening potential tech-It is critical that research managers have agree- nologies for their relevance to small-scale farm-ment from the highest levels that budget sav- ers' circumstances (Merrill-Sands and Collionings from downsizing will be retained by the 1992). In recent years more participatory meth-research institute to improve operating budg- ods have been developed that seek to directlyets and salaries of remaining staff. Donors can involve farmers in the design and execution ofhelp this process by providing downsizing experiments (Tripp 1989; Okali, Sumberg, andconditionality and the funds needed to com- Farrington 1996; Ashby and others 1995). Allpensate departing staff and to improve the this has, of course, implied considerable invest-working conditions of those who remain. ment of researchers' time in working directly

with farmers.Improving Client Orientation and Reviews of the impact of these methods con-Technology Transfer in National cluded that they were only partially successful,Agricultural Research Institutes as measured by more rapid adoption of tech-

nologies by farmers (Merrill-Sands and othersStrategies to involve users in the research proc- 1990; Tripp and others 1990). Difficulties ofteness include: informal research-farmer collabo- arose from treating FSR as a separate disciplineration, such as in farming systems research; and creating new units or departments to con-interaction with farmer organizations; empow- duct FSR rather than trying to introduce aerment of farmers through farmer re- farming systems perspective throughout thepresentation in research governance; farmer- research system through active collaborationfinancing of research and participation in de- among existing disciplines (especially amongcisions on the allocation of research funds; technical and social scientists) and with farm-closer research-extension linkages; and link- ers. FSR approaches cannot and should not beages with private industry associations. a substitute for strong disciplinary capacity.

Other problems with FSR have included:Improving Client Orientation * Lack of technologies on the shelf to adoptof Research Institutions because of weak commodity and discipli-

nary programsFrom the 1970s there was a growing awareness * Poor links between research, extension,that much agricultural research in NARIs was and policy, so that research recommenda-not relevant to the needs of most farmers, es- tions often did not reach farmers, or were


not adopted because of lack of access to nied by mechanisms by which farmers in-appropriate inputs 11 fluence how these funds are spent.

* Difficulties in feeding back information * Through participation of farmers in thefrom farmers to define priorities for ap- execution of research, especially adaptiveplied and strategic research conducted on research designed and implemented bythe experiment station. farmers themselves (box 3.5).

Merrill-Sands and Collion (1992) aptly summa- . Through participation of farmers or farmerrize the main outcome of this movement as lead- organizations in M&E of research, espe-ing to "client-oriented researchers but not to cially applied and adaptive research,client-oriented research organizations." FSR per- where farmer feedback and technology usespectives and methods are important for con- can be part of performance evaluation.ducting effective research programs, especially Although there is a high potential for farmerfor small-farm agriculture, and continuing ef- organizations to define the research agenda,forts are needed to strengthen farming systems effective organizations usually exist for only aperspectives and methods in agricultural re- limited number of commercial and export-search programs. FSR and community participa- oriented crops. The development of farmer or-tory approaches are especially relevant to natural ganizations with necessary technical, manage-resources management and eco-regional re- rial and political skills to represent the bulk ofsearch. Farming systems approaches also need small-scale farmers will continue to be a chal-to be incorporated into university degree pro- lenging task (box 3.6) (Eponou 1996; Colliongrams to provide future researchers and techni- and Rondot 1997). In addition, there are othercal staff with a better understanding of how and limitations on farmer input and direction ofwhy farmers do what they do, and with appro- research. First, farmers lack knowledge ofpriate tools for working with farmers. some new technologies and options. They tend

Greater client orientation in research systems to seek incremental changes to known produc-should result from direct influence of clients on tion systems and may oppose broader changesthe content of research. Farmers and their or- (for example, early resistance to adoption ofganizations can exert pressure in several ways: hybrid maize). Second, farmers may take a

e Through the political system, by appro- short term perspective and limit support to ba-priation of government revenues for agri- sic or strategic research with longer term pay-cultural research. This assumes that the offs.interests of farmers' and other clients of the There are, of course, other clients for theresearch system are broadly represented in NARSs besides farmers (box 3.7). Large agri-the political process. business firms may have their own research

. Through the participation of representa- programs and become partners rather thantives of farmers organizations (general or strictly clients for public sector research. Evencommodity-specific) on governing boards in such cases private sector firms usually relyand research advisory and review commit- on public sector research institutions for moretees of research institutions. Farmers in- basic research and for some technical backstop-cluded on such boards should represent ping of their research programs. In any event,the spectrum of intended clients of the re- identifying the technology needs of agribusi-search system. In some cases it may be ness requires attention to clients other than theappropriate to include representatives farmer.of processors, marketing, or consumers The less obvious client for agricultural re-to represent post-harvest and consumer search programs is the consumer. Many con-interests. sumers' demands and interests are reflected in

* Through contributions by farmers to the the market (for example, price premiums forfunding of agricultural research, accompa- grain quality) and routine collection of such

52 Strengthening National Agrictiltural Research Systems

Box 3.5 Farmer-run research: Experience with the Comite de Investigaci6n Agropecuaria Local

Moving from the traditional, station-based research and advises on the research. Technical staff visit 2-3system to a participatory, demand-driven approach times per season after the first 2-3 seasons.of on-farm testing presents potential problems. On- CIALs operate with a small CIAL Fund (US$500farm research can be costly in researcher time and per community) to cover the risks of crop failure oroperating costs. Even the best-intentioned efforts to subsidize the costs of trials. These CIAL funds havemay fail to draw farmers into a participatory research been consolidated into a corporation at the nationalprocess or provide adequate supervision for dis- level, but each CIAL manages its own fund. Thepersed field trials. Turning responsibility for on-farm funds, like the whole program, are "owned" by thetesting over to farmers is an attractive alternative be- community and managed by the Committee.ing tested by CIAT in several NARSs of Latin Amer- By 1995 55 CIALs had been established in Colom-ica (Ashby and others 1995). bia and others were formed in Bolivia, Ecuador, Hon-

Under the Comites Investigaci6n Agropecuaria duras, and Peru. Communities have been veryLocal (CIAL) program, begun in 1990, an institution responsive in taking on local research responsibili-with interest in technology dissemination (usually a ties. After two seasons CIALs are able to take onstate agency, NGO, or cooperative) facilitates a meet- almost all responsibility for on-farm experiments anding in which a comrnmunity analyzes potential needs paraprofessionals can provide almost all training andfor local technology testing.1 If the community is in- monitoring for the program. Research quality hasterested in undertaking local research, it selects a been good, with 75 percent (of 273 plots) statisticallyfour-member Committee (the CIAL) from the com- analyzable. Six CIALs have evolved into commercialmunity to coordinate the research work. Outside seed enterprises.technical staff from the organizing institution assistin planning and analysis of research trials and a para- 1. Comites Investigaci6n Agropecuaria Local are local agri-professional (a CIAL-experienced farmer) monitors cultural research committees.

market information can be important to re- also increasingly need to recognize other cli-

search priority setting. Other consumer needs ents, such as agribusiness and consumers. Thisthat are not as easily detected include issues of implies that institutional structures for public

food safety, nutritional quality, and environ- research systems must:mental interests. In some countries consumer * Develop incentive mechanisms for scien-advocacy groups and other NGOs are active tists to be rewarded for work that makesand can provide an interface between the re- explicit efforts to diagnose and solve realsearch system and the consumer. In such cases,or where markets do not function well, con-sumer surveys can be a valuable input into Box 3.6 Links to farmers' organizations

research* decisionnaking, especially for gaug- A recent study of linkages between research anding the strength of preferences about particular farmer organizations in Burkina Faso, Ghana, andquality traits (Ricks 1997b; Rubey, Ward, and Kenya (Eponou 1996) found that most are ineffec-Tschirley 1997). Consumer concerns, especially tive. Farmer organizations may not consider link-food safety issues, such as pesticide residues, ages to technology programs as an organizationalwill be increasingly important in setting the objective, or may be unaware of the possibility of aresearch agenda. linkage, or may not see a possible benefit. Research

institutions have only recently recognized farmerA NARS's client base is varied and not well organizations in the research process; may still be

organized. Building a farming systems per- unconvinced of potential benefits; and are hinderedspective into agricultural research systems has by existing operational policies. The study con-contributed to the relevance of research to the cluded that more effective linkage mechanisms are

main clients, the small-scale farmer. However, needed between farmer organizations and research,main and that the power of farmer organizations must

in future more attention must be given to em- be increased to strengthen their influence in the re-powering farmers to provide direct input into lationship.the research agenda. Research systems will


Box 3.7 Improved research-extension linkages: The case of Ghana

Research-extension linkages in Ghana have always * Regional meetings for each agroecologicalbeen complicated by the fact that research is in the zone to consider results from the districtMinistry of Science and Technology and extension in PRAs. Extensionists, researchers, and farmersthe Ministry of Food and Agriculture. But recent Bank all participate in these meetings, and assignsupport to both research and extension has provided identified problems to research and exten-an opportunity to strengthen research-extension link- sion. The regional meetings also identify ex-ages. Research-Extension Liaison Committees tension training needs and plan joint(RELCs) have been formed in each major agroecologi- research-extension workshops and monitor-cal zone to jointly plan research and extension priori- ing tours for the year.ties and to promote joint training sessions, field visits, * A national meeting of research institutes andworkshops, field days, and on-farm trials. programs-with extension participation-de-

These RELCs sponsor a series of annual planning velops the research program plans for the com-meetings and joint activities at multiple levels in the ing year.system. Wherever possible private agribusiness, Although the system is still in its early stages, thefarmers, and NGOs are invited to join. The annual RELCs have clearly led to greater collaboration andplanning cycle includes: communication between research and extension. Fur-

Participatory rural appraisals (PRAs) of farmer ther evolution of the system must address issues ofproblems at the district level. Extensionists, reducing the number of meetings and joint activities,farmers, and researchers take part in these ap- which are costly and time-consuming, and improvingpraisals, which are used to draw up district ex- monitoring to ensure that planned activities are car-tension plans for the year. ried out.

problems of farmers and other clients and research institutes link with this wider range

that achieves on-the-ground results of actors.

* Develop mechanisms for ensuring a real Research-extension linkages remain a critical

voice by users in setting the research weakness in many countries (Merrill-Sands

agenda. The appropriate choice of mecha- and others 1990; Spurling and others 1992;

nism is very situation-specific, depending Antholt 1994). In theory, the dividing line be-

on the type of user and the type and level tween research and extension blurs in the

of research stages of adaptive research and verification of

e Involve farmers in decisions on allocating technology. As agriculture becomes more

public funds for research, and link farmer knowledge-intensive, these linkages will be

financing of research to farmer empower- even more critical because research must gen-

ment in research decisionmaking. erate greater amounts of (and better quality)

information to pass on to farmers.

Enhancing Technology Transfer Institutional solutions to the research-exten-

sion gap, such as research-extension liaison

Technology transfer or dissemination has been persons, high-level research-extension coordi-a weak link in most NARSs. The reviews of nating committees, and other measures de-

World Bank experience in research have all signed to strengthen linkages, have rarely

identified this as a limitation on realizing succeeded. The reasons for this failure include

the benefits of research. In the past this has the institutional separation of research and

been associated with weak research-extension extension, the fact that many research recom-

linkages: nowadays, however, technology mendations are not relevant to farmers' cir-

transfer is being realized through other mecha- cumstances, the tendency for researchers to see

nisms, including diverse types of NGOs and their job as done when the technology is de-

the private sector. Strategies to enhance tech- veloped, rather than when it is adopted, and

nology transfer will increasingly require that the low professional preparation and status of


extension staff, which limit their ability to feed from NGOs (Kaimowitz 1993). NGOs haveinformation back to the research system. considerable potential to supplement public

The World Bank has recently tried to solve extension services that are under budget pres-these difficulties by designing joint research sures and to serve areas and farmers not cov-and extension projects, sometimes labeled ered by the public system."technology" or "agricultural services" pro- The strong rural presence, dedication, andjects. This will only succeed, however, if the commitment to impact on the poor, commondesign of these projects is truly a joint activity to many NGOs, makes them potentially valu-and not simply the sum of two components. In able partners in technology-transfer programs.the long run a truly decentralized extension NGOs also have some comparative advantagesystem with real accountability to farmers is in adaptive research, since they tend to empha-likely to be the best means of allowing exten- size participatory methods and have the flexi-sion to place demands on the research system. bility to undertake and integrate activities,In the short run some institutional means that especially in resource conservation where com-have been useful in strengthening research-ex- munity decisions may be required. But mosttension links include: NGOs have little experience in conducting re-

. Joint operation of on-farm adaptive research and often lack technical expertise. Somesearch with strong involvement of farmers, examples of successful research program-NGOincluding the joint conduct of diagnostic collaboration are available from Central Amer-surveys and on-farm experiments ica, where the adoption of resource-conserving

* Joint formulation of technological recom- technologies has been promoted by NGOs withmendations and extension priorities (box close ties to these research programs (for ex-3.7) ample, Buckles and Erenstein 1996).

* Introduction of reward systems based on Community-based organizations and farmerfarmer adoption of technology for research organizations are specialized types of NGOsand extension staff formed by rural people-the major users of the

* Bureaucratic mergers that simplify life for research system's products. Community-basedall concerned. organizations are now becoming more in-

Even these short-run solutions are unlikely volved with natural resource management ac-to succeed unless a real effort is made to ensure tivities (for example, irrigation managementthat farmers themselves are at the center of and community-based forest resource use),efforts to reinforce research-extension linkages. many of which must be undertaken by theReward systems for research and extension community to be effective. These rural organi-must be firmly based on the involvement of zations can be important users of (or con-farmers and performance on the ground. duits for) technical information from research-Experience shows that effective research- extension systems to their members.extension linkages are better promoted by pro- The links between research and NGOs are inviding incentives and rewards to both parties their early stages and much remains to beto solve farmer problems, than by institutional learned about how to effectively and sustain-solutions in the form of new committees or ably integrate NGOs into the process of tech-specially created positions. With the essential nology transfer. In the future, NARS strategiesharmony of interests, there will be real pres- will need to consider a significant involvementsure to work together. of NGOs, especially community and farmers

Stronger links to NGOs offer another means organizations, in on-farm adaptive research,of strengthening technology transfer. NGOs community-based resource management, andare increasing rapidly and, by one estimate, 15 in providing a conduit for feedback from userspercent of farmers in Latin America (primarily to research priority setting and design. To fullysmall-scale farmers) now receive some services use this synergy, research institutions, espe-


cially regional research stations, must develop sector may also work with private companiesinstitutional mechanisms for establishing part- to disseminate technical information from thenerships with NGOs and incorporating them public research system; for example, in usinginto decisionmaking. chemical companies to disseminate some IPM

The private sector, including input dealers, technologies.agribusiness enterprises, and private consult- Finally, government policies are often criticalants, will become increasingly important in to farmers' ability to adopt and profit from newtechnology transfer, implying the potential for technologies and research programs must rec-the development of NARI-private sector part- ognize policy constraints to technology adop-nerships. Even large agribusiness firms often tion and provide input to policy formulationfind it cost effective to seek technical support (box 3.8).for local adaptation and transfer from the pub-lic sector.12 To promote competitiveness and Notesbenefits from such enterprises, NARIs need toensure that organizationally there is a window 1. Although this section focuses on the national

or point of contact for private agribusiness to research institutes, which exist in most countries,

seek technical information and support from many of the principles articulated here are equally

the NARI. Such a contact point may serve to relevant to universities and other public sector re-

coordinate assistance from various units in the search organizations.goermet(,and quaran- 2. See Hayami, Kikuchi, and Morooka (1989) for evi-

gove sernmce) ant (rsarch,lp eteson, feedprivatese dence of the relationship between expenditures on ricetine service) and can help to feed private sector reachndwldiepies

d b k h 1 i Th 11 ~~~~~research and world rice prices.needs back into research planning. This will 3. For example, India spends about 0.26 percent ofrequire increased capacity for post-harvest agricultural GDP on research (Singh, Pal, and Jha 1997)

handling and processing technologies for non- while it spends the equivalent of 7 percent of agricul-

traditional crops, with much of this work done tural GDP on input subsidies alone (Gulati and Sharma

under cost-recovery mechanisms. The public 1995). In Zimbabwe in many years expenditure on

Box 3.8 Serving the "policy client"

The agricultural policy environment has major impli- technologies, which are highly dependent on the pol-

cations for technology transfer, and NARIs are well- icy environment.

placed to provide information and analysis to change The policy environment also affects investment in

the policy environment. Socioeconomic research research; there is little incentive to invest in technology

often makes the most important and direct contribu- generation when the results are unlikely to be adopted

tion to policymakers, but biological research also pro- or benefits generated. However, the policy environ-

vides valuable inputs. During the research process ment for future adoption of any research result is

much information is generated that could be useful highly uncertain and the long time lag in producing

in policy formulation (for example, input response useful research results provides one rationalization for

data, potential of new crops, and factors affecting the research investment even in a poor policy environ-

efficiency of input and resource use at the farm level). ment. With an unfavorable policy environment, in-

Experience has shown that providing this type of vestment in research should be continued in specific

information can help shape policies and facilitate the highly focused areas, such as for technologies whose

adoption of technologies (for example, Martinez and adoption success is less sensitive to the policy envi-

Sain 1991) and NARIs should attempt to package and ronment (for example, new varieties), long-term

report such information in a form useful to policy- research that requires continuity (for example, main-

makers. There is a critical need for NARIs to increase taining genetic resistance to diseases), and research on

their input to the policy process. Ultimately, it is in high-priority issues such as emerging pest problems.

the interests of NARIs to increase their involvement Meanwhile research on impacts of technology devel-

with policy issues, as judgment of NARI effectiveness opment can provide important input to government

will be based on the adoption and impact of new policy decisions on investment in research.


research is dwarfed by support to the grain marketing tension systems themselves are being privatized andcorporation. decentralized.

4. Levies also include check-offs (as used in the 9. Donors have been unwilling or unable to adjustUnited States). Such research financing mechanisms their priorities to conform to the new master plan. Thisvary from what is essentially an additional tax to fi- may be especially problematic when one donor cham-nance government research programs to funds col- pions the development of a research plan, which placeslected (with government assistance) on behalf of a low priority on topics or regions in which other do-producers themselves for use by a nongovemmental nors have on-going activities.producer organization to fund research and promo- 10. Linking results from the priority-setting work totional activities (Gilles 1997). the allocation of competitive grants is one way to

5. In this case the method has the advantage that the quickly shift resource allocation in line with identifiedlevy is paid in proportion to farmers' marketed surplus priorities.and is likely to be strongly progressive in terms of 11. There has always been a tension in farming sys-farmers' incomes. tems research (FSR) especially in Africa, between de-

6. The term "donor" is used here to include intema- veloping technologies for the current situation oftional lending agencies. farmers, often characterized by limited access to inputs

7. At first this support was for incremental operating and credit, and using the results of FSR to influencecosts associated with capital development, with the the policy agenda. Much of this tension relates to theexpectation that this budget would be assumed by the debate between low-input and high-input strategiesgovernment at the end of the project. More recently, for agricultural intensification. It is now apparent thatthere have been many examples of donor support to there is a middle ground, which recognizes that someoperating budgets that are independent of capital external inputs, especially fertilizer, will be required.development. 12. One response to the problem of technology sup-

8. Creation of autonomous research corporations port for agribusiness has been project support forfor supplying research products may also widen the agribusiness development (for example, throughinstitutional and cultural gap between research the USAID). Such projects have followed varied strate-and extension and lead to difficulties in disseminat- gies of institutional development, market develop-ing technology. However, this is becoming less of a ment, and enterprise support, but have typicallyproblem as technology dissemination is increasingly included consideration for technology support (Kumarbeing managed through diverse suppliers, and ex- 1995).

CHAPTER 4

The World Bank's Role in StrengtheningNational Research Systems

Jt is vital that the Bank maintain its commit- agricultural research, the Bank should stronglyment to developing national agricultural commit to the development of the NARSs. Thisresearch systems. The Bank is by far the final section summarizes key elements in current

largest donor to the sector, and other tradi- Bank strategies to support agricultural research.tional bilateral donors are reducing their fund-ing for agricultural technology programs.- The Evolution of Bank Support to Researchdevelopment and dissemination of improvedagricultural technologies will be critical to From 1981 to 1996 the Bank financed 458 pro-meeting the challenge that the Bank has laid jects with agricultural research components inout in its Action Plan for the rural sector to 91 countries. Of these projects 50 were free-reduce poverty, improve food security, and standing research projects and 408 broaderbetter manage natural resources (World Bank sector projects with research components. Re-1997a). Such a transformation must depend on search funding under these projects totaleda more productive and sustainable agriculture, US$3.87 billion. In recent years annual commit-which requires a continuous flow of new tech- ments have averaged about $200 million (tablenologies and improved management tech- 4.1 and Annex 2). The Bank's lending for agri-niques. cultural research now accounts for a large

The discussion of issues facing the research share of all external support for agriculturalsubsector repeatedly identified this subsector research in developing countries. By the end ofas having changing needs and opportunities. 1996 91 countries had received assistance forThese have significant implications for Bank agricultural research under the lending pro-support to national agricultural research sys- gram. Many countries had sequential agricul-tems. To some extent, the conventional wis- tural research projects: in some cases the Bankdom that national agricultural research has supported agricultural research continu-systems (NARSs) have become stronger is now ously for 15 years or longer.questioned and it is realized that much of pastcapacity development is not sustainable. It is Bank Lending to Agricultural Researchevident that the capacity of many NARSs hasbeen eroded by the current funding crisis and Both the research and extension shares of totalby institutional inertia. Given the level of past Bank lending for the agricultural sector in-investments and the emerging challenges to creased significantly since the late 1970s and

57


Table 4.1 Summary data on World Bank agricultural research projectsProjects with research Free-standing research Projects with research Total research lending

Year funding projects components ($US billion)

1981 32 3 29 0.361982 27 4 23 0.201983 36 2 34 0.211984 32 2 30 0.131985 39 4 35 0.20

1986 28 2 26 0.161987 36 4 32 0.211988 20 2 18 0.341989 25 4 21 0.251990 32 9 23 0.29

1991 22 4 18 0.181992 26 3 23 0.341993 24 2 22 0.241994 27 1 26 0.211995 26 3 23 0.391996 26 1 25 0.16

Total 458 50 408 3.87

Source. Pritchard 1994 for 1981-87 and ESDAR (World Bank) database for 1988-96.

especially since the mid-1980s (figure 4.1). This with Africa, Latin America, and East and Southis a function of both a modest increase in fund- Asia each receiving between 29 and 31 percenting for research and a decrease in the total of the total funding from 1981 to 1996. Europefunding for the agricultural sector. and Central Asia and West Asia/North Africa

Research project financing has been reason- each received only 5 percent of the total. Overably well distributed geographically (figure 4.2) this period there has been a major geographical

shift in the share of lending from Latin Amer-Figure 4.1 Percentage of World Bank agricultural loans ica and East Asia to Africa and South Asiafor research and extension (table 4.2).

Percent Bank Policies and Strategies

40

The World Bank developed its only formal pol-icy statement on aFricultural research in 1980

30 (World Bank 1980). This policy clearly empha-sized expansion of the NARSs, with particularemphasis on the NARls. The policy identified as

Extension a long-term objective the increase of public ex-20 penditure on agricultural research from the 1980

level of less than 0.5 percent of agricultural GDP(AgGDP) to 2.0 percent. This expanded program

10I of research envisioned a corresponding increaseof 9,000 additional trained scientists.

Research The Bank's strategy, outlined in 1980, was to

o I U I 1 I LJ l U I L increase Bank lending for research programs(at the expense of lending for extension) with

1977-80 1981-84 1985- 88 1989-92 1993-96 emphasis on food staples for low-income con-

The World Bank's Role in Strengthening National Research Systems 59

Figure 4.2 Geographic distribution of World Bank cent years. Most of these operational strategiesagricultural research funding between 1981 and 1996 still guide support to NARIs today.

Since approval of the 1980 agricultural re-South Asia search policy paper there have been several

t5% _ internal reviews of Bank experience with agri-cultural research projects (World Bank 1983;

Middle East Pritchard 1990). These reviews largely con-& Nort~h Africalrgl

5%/O A i firmed the directions set out in the policy guide-lines and noted significant successes ininstitution building, although the challenges as-sociated with developing capacity for priority-setting, research-extension linkages, budgetarysupport and incentives for scientists, and re-search monitoring and evaluation continued.

Latin America East Asia Recommendations from the various reviews of& Caribbean & Pacific the Bank research project portfolio are summa-

Europe & rized in annex 1.Central Asia In 1996 the Operations and Evaluation Depart-

5%ment (OED) of the Bank did a comprehensive

sumers, cash and export crops, farming sys- review of completed agricultural research pro-tems research, sustainable production systems, jects. This was more guarded in its conclusions.environmentally sound production practices, It found that, although Bank intervention in ag-natural resources management, and nutrition. ricultural research has had significant positiveThe strategy also provided for support to more impacts, serious deficiencies persist in sustainingbasic and strategic research at the IARCs and research institutions and in establishing institu-strengthening NARSs in fisheries research, tional capacity in research planning, priority set-tropical soils management, irrigation and ting, and evaluation. The review noted particularwater management, pest management, nitro- concern with the sustainability of research fund-gen fixation, photosynthesis, and genetic ma- ing, especially the weak capacity for policy andnipulation. The 1980 policy also laid out economic analysis in most NARSs, and stronglyoperational strategies for strengthening agri- recommended the wider use of economiccultural research management and incentives analysis to guide research planning efforts andthrough project lending (table 4.3). The grow- increase research efficiency. The major recom-ing importance of these operational strategies mendations of the review are given in box 4.1.was reflected in the number of agreements and In the 1990s the Bank responded to the newconditionalities on organization, structure and realities of NARSs and significantly departedmanagement attached to project lending in re- from certain aspects of the 1980 policy guide-

Table 4.2 Regional shares of world bank loans for agricultural research(percent)

Region 1981-84 1993-96

Africa 6 50Latin America 36 8South Asia 15 27Southeast and East Asia 39 9West Asia/North Africa and Europe/Central Asia 4 6

Note: Percentages are based on 127 projects in 1981-84 and 103 in 1993-96.


Table 4.3 List of operational strategies identified in 1980 World Bank policy guidelines

Theme Strategy

Planning and priority setting * Encourage preparation of national research plans and develop capacity in planning andmanaging research.

* Review overall agricultural sector policy in relation to incentives for technology adoption.Financing * Provide financial support for long-term development of research programsReview and evaluation * Institute periodic internal and external reviews of research programs

* Encourage research organizations to establish effective monitoring and evaluationsystems to assess impacts.

Incentives * Ensure adequate incentives for researcher performance.Linkages * Design operational strategies to strengthen researcher-extension-farmer linksTraining * Incorporate both research management and technical specialization into training

programs and link training to the needs of research plans.

Source: World Bank 1980.

lines. The emphasis now has shifted to a focus probably had less success than with the quan-

on quality dimensions of agricultural research, titative dimensions of expanding the size of the

especially management, incentive systems, and system.

accountability, with conscious efforts in most Overall, since 1980 agricultural research has

cases to avoid further creation or expansion of been considered a priority in the Bank's lend-

public research organizations. Downsizing and ing portfolio. The Bank has responded to the

consolidation of research infrastructure is the 1980 policy guidance and provided a com-

order of the day. Table 4.4 shows how empha- mendable level of attention and support to ag-

sis in Bank programs has changed in the past ricultural research and development, although

three years to emphasize these institutional results may not have always been in line with

and qualitative issues, while aspects such as expectations. A 1997 Bank review of "at-risk"

master planning have been de-emphasized. In projects indicated that the research project

these institutional dimensions the Bank has portfolio has performed relatively well, but

Box 4.1 Major recommendations of the Operations and Evaluation Department reviewof agricultural research projects

The 1996 Operations and Evaluation Department re- 5. Encourage the use of economic analyses forview made ten major recommendations for strength- borrowers to prioritize research activities andening future lending to agricultural research: for the Bank to evaluate research programs ex-

1. Review the roles of the agricultural sector and ante and ex-postagricultural research in Country Assistance 6. Use a core of specialized Bank staff to enhanceStrategies (CASs) support for research program development

2. Continue strong support for international, re- 7. Treat dissemination of research findings as angional, and national agricultural research, but integral part of technology developmentwith comprehensive support for NARSs con- 8. Design projects that develop NARS procedurestingent on borrower commitment to sustain- to make research programs more relevant toable funding for research and to sound research clients' needsmanagement principles 9. Include monitoring and evaluation systems

3. Provide limited research support combined as a mandatory element of Bank researchwith nonlending services to strengthen bor- projectsrower commitment, when favorable sectoral or 10. Include measures to enhance scientific rigorresearch policies are lacking (training, technical assistance, external reviews,

4. Foster external NARS linkages and Bank alli- competitive grant funding, scientific network-ances with development partners to enhance ing, and institutional linkages) in all Bank-sup-the effectiveness of research programs ported research projects.


Table 4.4 Trends in institutional issues emphasized in the current World Bank portfolio of research projects

Percentage of projects approved by Board

Issue 1990-93 After 1993

Emphasis on institutional pluralism 50 86Promotion of private-public interaction in research and development 12 71Support for new funding sources 6 87Support for competitive funding 12 86Support for downsizing and consolidation 25 57Involvement of farmers in research governance and funding 38 87Development of master plans for national agricultural research institutes 50 14

Source: Based on a review of World Bank porHfolios of research projects.

that there are some common problems in re- increased its support to the Consultativesearch projects (box 4.2). Group on International Agricultural Research.

The Bank has responded to the challenges ofbuilding effective NARS capacities in other Key Elements of Current Bank Strategyways. In 1985 the Bank with other donors initi-ated the Special Program for African Agricul- The Bank and other donors have shifted em-tural Research (SPAAR) to meet the special phasis of support for national agricultural re-needs in African agricultural research (for ex- search systems. The earlier focus on systemample, uncoordinated donor programs and expansion has given way to greater attentionlack of local funding). SPAAR is helping to to management of resources within the re-build consensus and cooperation to coordinate search system. This emphasis will need to con-and focus donor support to NARSs.3 In 1994 the tinue, but greater support will be required forBank established the Office for Agricultural Re- policy reforms that enable NARSs to developsearch and Extension (ESDAR), a multidonor sustainable financing systems and that inte-forum to coordinate support to agricultural re- grate international and national institutionssearch and extension programs and create the into efficient research systems.linkages necessary for a more efficient and ef-fective global agricultural research system Expanded Rolefor Policy Dialogue(Petit and others 1996). The Bank, recognizingthe strategic importance of international agri- Recognizing the strategic issues now faced bycultural research for NARSs, has in recent years NARSs, the Bank must expand support to policy

Box 4.2 Problems identified in a 1997 review of agricultural research projects

In 1997, under the Portfolio Improvement Program, * Difficulties in establishing institutional auton-the Bank reviewed research and extension projects omy for research institutions because of lack ofcategorized as at risk of producing unsatisfactory understanding of the special needs for researchoutcomes. Common problems found were: management, and lack of imagination in formu-

* Lack of a consensus in-country on a strategic lating the mandates and rules of business of thevision for public sector research institutions and new organizationsthe evolution of the national agricultural re- * Inadequate attention to sustainable financingsearch system (NARS) for research, especially for new research initia-

* Ineffective national leadership for many re- tives after project completionsearch institutions, resulting in both internal * Weak monitoring and evaluation systems formanagement problems, as well as lack of politi- both research programs and institutionalcal support, especially for funding research changes.


and institutional change through economic and mented to "protect" farmers or maintain publicsector work and conditionality on loans. The sector monopolies. Since private R&D invest-accurate assessment of constraints, opportuni- ment is often made on the basis of regionalties, and strategies for a national agricultural markets, harmonization of policies at the re-research system will be central to future eco- gional (supra-national) level will often be morenomic and sector work. This process must important than at the national level.be highly participatory to build borrowerconsensus on feasible policy options. Much dis- Emphasis on Sustainable Fundingcussion will be needed to change traditionalattitudes and relationships within NARSs and Both Bank lending and nonlending work shouldto build a country commitment to investment give explicit attention to the level and long termin agricultural research and institutional sustainability of research funding. New invest-change appropriate to that country's situation. ments in research capacity building must be

Economic and sector work should give par- conditioned on borrower commitment to sus-ticular attention to sustainable funding of ag- tainable financing (as in the Operations andricultural research and to promotion of private Evaluation Department (OED) Report). The em-R&D. Sustainable funding is best addressed in phasis should be on mobilizing resources forthe Country Assistance Strategy (CAS) that is support to research programs after completiondiscussed with ministries of finance that have of donor assistance. Bank support may include:a major influence on the allocation of public * Economic and sector work and associatedfunds. Such dialogue between the country and policy dialogue that reviews the allocationthe Bank should explicitly identify the public- of public funds to competing claimsgood component of agricultural research as the * Support in Bank projects to sustainablebasis for public funding. Conditionality may funding mechanisms for a more regularbe used selectively to influence borrowers' re- and predictable level of financing for re-search policies. Conditions may relate to research programs, such as farmer financing,strictions on system expansion, promotion of cost-recovery, and endowed researchregional coordination, and implementation of foundationsinstitutional and management reforms in * Development of mechanisms for fosteringNARIs. These conditions must be used care- public-private interaction in agriculturalfully, recognizing that project conditions are research, including the legal frameworknot a substitute for, or necessarily indicative of, for establishing joint venturesborrower commitment; rather these should be . Explicit analysis of sustainable researchseen as mutually-accepted guidelines for funding levels in preparation of projectschange developed through participatory proc- for Bank support to ensure that researchesses. Where possible, conditionalities on re- activities and infrastructure can be main-search funding and institutional reform can be tained after project completionpart of broader sector investment and adjust- * Analysis of the research portfolio to clearlyment loans whose success ultimately depend identify the public-good research activi-on availability of improved agricultural tech- ties, relative to those activities that couldnologies, rather than as conditions specific to be undertaken by the private sector oragricultural research projects. other alternative suppliers.

Economic and sector work must also analyzepolicies and regulations that may be restricting Renewed Attention to Universities and Humanprivate investment in R&D. In this it will be Resources Developmentimportant to distinguish between policies andregulations that relate to public safety and the Recent Bank projects have given more atten-national interest versus those that are imple- tion to the place of universities in executing


research, mainly as potential recipients of com- and define regional research activities. Thispetitive research grant funding. In the future may require special regional funds adminis-this focus must be broadened to strengthen tered through Bank regional programs. Once auniversity capacities to undertake research and real commitment to regional activities has beento recognize the key task of universities in established, support for regional activitiestraining the next generation of scientists. More could be included in country loans for agricul-scientists will have to undertake postgraduate tural research, with the NARSs taking the leaddegrees locally, and this will require a substan- in defining the priorities and activities for thistial upgrading of capacities in local universi- support. The Bank is also leading efforts to fa-ties. In some NARS that have already shifted cilitate more active participation of NARSs into local training, this capacity has been eroded international agricultural research policyand it will require long-term support to revi- issues through the Global Forum on Agricul-talize the university system. The Bank has less tural Research (box 4.4).experience in projects aimed at university ca-pacity building and will need to begin cau- Donor Coordinationtiously and engage outside experts to helpbuild such capacity. Bilateral donors can provide complementary

grant financing for training and technical as-Continuing Institutional and Policy Reforms sistance, which countries are often reluctant toin National Agricultural Research Institutes fund from Bank loans. The complementarity of

these inputs can be especially important in im-While NARIs are no longer so central to efforts proving the quality of research and strength-to strengthen agricultural research capacity, in ening NARSs linkages to the global researchmost countries, the bulk of the human resources system. Donor inputs can also be an importantand research infrastructure is located in the complement to Bank work on diagnosingNARI. So, continued efforts are required to NARS constraints and developing system im-strengthen the productivity and cost effective- provements through participatory processes.ness of NARIs through a range of management Donor coordination is the clear responsibilityand organizational reforms, including defini- of NARS leaders, but the Bank, as the majortion of mandate, diversification of funding lender for agricultural research, can encouragesources, decentralization of some research activi- greater coordination of all donor investmentsties, planning and priority setting, monitoring and activities in agricultural research. The Bank'sand evaluation, human resources development prominence in financing agricultural researchand management, and links to major clients. and its recognized capabilities can help to con-The institutional structure of NARIs in relation vene donors and research agencies to a commonto civil service rules and the governance of the purpose. Such coordination is most important atNARI will be major factors in implementing the country level, although regional and interna-these reforms. Policy and institutional reform tional coordination is also needed.in the formerly centrally planned countries of One good model of coordinated donor fi-Europe and Central Asia represent a special nancing is the consolidated funding mecha-challenge with high priority (box 4.3). nism, whereby donors provide support for an

agreed-upon national research plan. The Agri-Fostering Regional and Global Collaboration cultural Research Funds (ARFs) in Kenya, Tan-

zania, and Zambia are a practical approach toThe Bank needs new approaches to promote a consolidated funding mechanism. These al-cost-effective regional research activities. In- low donor funds to be channeled (on a com-itially, the Bank can help build commitment by petitive or noncompetitive basis) to priorityfunding workshops and meetings to establish research activities. The ARF standardizes man-

64 Strengthening National Agricultuiral Research Systems

Box 4.3 An emerging challenge: The national agricultural research systems of Eastern Europeand Central Asia

One of the emerging challenges for the Bank and for istries to regional centers more closely linked to pro-the global agricultural research community is that of ducer needs. Many staff left voluntarily and joinedrevitalizing agricultural research in the countries of the private sector; some institutes were privatized orEastem Europe and Central Asia. The agricultural re- simply ceased to function. For other countries of thesearch systems have not fared well during the transi- region the reform process is taking much longer.tion to market economies. Many countries have major Governments of the region must first recognize theagricultural production potential and a solid scientific importance of restructuring their agricultural tech-base. The research systems, however, have not made nology innovation systems to support market-basedthe transition to more market-driven systems. They economy and a growing private sector. With demon-are oversized and isolated from the world scientific strated commitment to reform, the Bank and othercommunity. With the drastic fall in funding, younger donors may initiate support vital to the reform proc-staff are abandoning the research institutions. ess. A regional initiative to share experience and

Integration of the research system of former East build consensus for reform may also be needed.Germany, admittedly a special case, provides someinsights. After reunification the number of scientists Source: Basler, A. 1997. "Lessons Leamed from Changes inoutside the universities is to be reduced by more than Research Investment Strategies in Germany." Paper for an80 percent (from 6,200 to 1,081) to bring it in line with ESDAR workshop on new investment strategies for agricul-the numbers in the West. Staff and institutions were tural and natural resource research. October 1-3. London,shifted from centralized, science academies and min- U.K.

agement of funds, but allows for separate ac- as market liberalization continues, there willcounting and directing of support to specific be growing demand for research on commodi-programs in line with the different accounting ties with export potential that will have torequirements and program priorities of indi- address special problems of post-harvest han-vidual donors. dling and quality standards, including pesti-

cide residues. Bank support will also beSupporting New Research Priorities dictated by changes in science itself. Four ma-

jor areas of research for increased Bank supportA large part of Bank loans for agricultural are briefly discussed below.research is used to support priority researchareas. These are established by NARSs, increas- Linkages to Basic Researchingly through the use of more systematic andformal priority-setting methods. Such priori- The past decade has produced major changesties will necessarily be dynamic. For example, in the opportunities for agricultural research to

Box 4.4 Global Forum on Agricultural Research

A 1996 meeting of NARSs, Consultative Group on (2) improving priority setting for international agri-International Agricultural Research (CGIAR) cen- cultural research, (3) strengthening regional organi-ters, advanced research institutes, NGOs, the pri- zations and NARS-NARS partnerships, (4) fosteringvate sector, development assistance donors, and research partnerships, and (5) mobilizing funding forfarmers organizations developed a Declaration of agricultural research. The forum is scheduled to meeta Global Forum on Agricultural Research. The every three years and will operate through a Secretar-Global Forum seeks to bring these various partners iat, Steering Committee, and electronic forum in thetogether to promote increased partnerships and interim. The Global Forum provides an opportunitydiscussion of major issues affecting international for NARSs to participate more actively in shapingagricultural research. The Global Forum focuses on debates within the global agricultural research systemfive goals: (1) enhancing the capacities of NARSs, and for developing collaborative linkages.

7The World Bank's Role in Strengthening National Research Systems 65

make use of advances in basic sciences. This is season-specific management options (for ex-especially true in biotechnology, which is now ample, integrated pest, nutrient, and waterbecoming a reality in farmers' fields in indus- management). In lower potential areas, wheretrialized countries. Only the largest and most variability in space and time tends to be high,mature NARSs will have the resources or ca- knowledge-intensive approaches are needed atpacity to undertake basic research, but all will an even earlier stage of technical change.have to ensure that their farmers have access to The move toward knowledge-intensive man-the products of this research. Universities often agement will be stimulated by environmentalmay be important in basic research, especially concerns as well as by the need to reduce costsas they become more fully involved in research and remain competitive under market liberaliza-in general. However, most NARSs will have to tion. Developing such knowledge-intensive sys-depend on external links with regional, inter- tems will entail activities, including more basicnational, and advanced research centers, and research on crop-soil-water-nutrient interactions,with the private sector, to gain access to much decentralized adaptive research, use of new in-of this new technology. Bank projects can facili- formation technologies (for example, geographictate these links by giving support to: information systems (GIS) and crop models) (box

* Funding to facilitate global links and col- 4.5), and adaptation of decision aids and infor-laborative research programs. Even mation technologies to the needs of small farmerssmaller NARSs are important participants (Byerlee 1997). The appropriate mechanismsin the more applied end of the basic re- to organize and manage research and technol-search spectrum. ogy dissemination for knowledge-intensive ag-

• Developing policies and associated legisla- riculture is still being debated and the Banktion on intellectual property rights. needs to be permanently active in that debate.

- Developing-with full participation of thepublic-policies on biosafety and potential Research on Natural Resources Managementrisks associated with the release of newtransgenic technologies. With heightened awareness of environmental

* Developing country-specific strategies for concerns, strengthening capacity in research onbiotechnology, including developing local natural resource management is a now a highexpertise, mechanisms for accessing pro- priority. However, it is important to recognizeprietary technologies, and participation in the complementarity between productivity-en-international networks. hancing research and research on natural re-

* Support to local capacity building in sources management. Commodity researchbiotechnology issues, especially biosafety may make important preventative contribu-and environmental impact assessment. tions to natural resource management through

productivity increases in favored areas that al-Researchfor a Knowledge-Intensive Agriculture leviate pressures on marginal areas. However,

technology is only one element in the solutionIn all regions a priority will be to strengthen of natural resource management and environ-both research (strategic and applied and adap- mental problems. Institutional and policytive) and technology dissemination in the changes, such as property rights and appropri-emerging knowledge-intensive agriculture. In ate price incentives, may be more critical forhigh potential areas that have undergone rapid many natural resource problems.technical change, returns to further intensifica- Bank projects can support research on natu-tion are diminishing and future gains will come ral resources management through:from using inputs more efficiently. This implies * Encouraging capacity building in selectedreducing the level of external inputs through disciplines, especially soil science, geo-developing improved information on site- and graphic information systems, and social-


Box 4.5 Geographic information systems: A potentially powerful new tool for agricultural research

Geographic information systems (GIS) represent a * Identifying "recommendation domains" or ar-potentially important new tool for agricultural re- eas for dissemination of new technologiessearch to organize and manipulate geographically * Monitoring the spread of new technologies,referenced data. Computer technology and method- pest outbreaks, or changing social or environ-ologies for GIS technology are evolving rapidly. GIS mental conditionssystems offer the potential to access vast amounts of . Serving as a powerful tool for before and afterremote sensing data, manage and integrate large data the fact impact assessment of research and newsets, and develop models to predict changes in de- technology.fined variables. Agricultural research applications ofGIS include: GIS technology is especially useful for natural

. Helping to diagnose and prioritize research prob- resources management applications and can be effec-lems by mapping and characterizing areas with tive for presenting and explaining data to decision-specific problems (that is, identifying climate, al- makers. To be effective, the relatively modesttitude, and soil conditions for a particular pest) investments in GIS hardware, software, and training

. Planning regional testing by identifying areas must be balanced by appropriate attention to avail-with appropriate or similar conditions for trials ability of geo-referenced data (UNEP 1997).

science, which are some of the weakest client orientation. Yet social science is one ofdisciplines in most NARSs. the weakest disciplines in NARSs, even in rela-

* Supporting new organizational models to tively strong NARSs.5

facilitate natural resources management This implies either increased staffing of so-research, which is by nature systems-ori- cial scientists or increased funding for contract-ented, multi-disciplinary, and long term ing economic and rural social science researchresearch. This does not imply the creation at universities, NGOs, and other institutionsof new institutes focused on natural re- outside the NARI. It also requires increasedsources-although this may sometimes be attention to training, both in-country and ex-justified-but may require new mecha- ternally, and strong leadership for social sci-nisms, such as the matrix approach, to ence research at the highest level in NARSs tobring together required skills from com- promote social scientist input into researchmodity and disciplinary programs within programs and research management.the existing organizational structure(Crosson and Anderson 1993). Implications for the BankSupporting the involvement of other stake-holders, including government agencies, While there is a strong case for continued fi-NGOs, universities, IARCs, and local com- nancing of agricultural research, the size andmunities, to resolve resource management composition of loans in the future is likely toproblems. This may require greater organ- be different. Although good research and tech-izational and wider networking skills than nology development require sustained invest-conventional commodity research. ment, it is more dependent on scientific quality

than on money.Social Science Research

Implicationsfor Project LendingMany of the issues confronting the NARSs re-quire strong social science input (Byerlee and Many of the research system costs that haveFranzel 1993). These include: broadening the been financed in the past are now in much lessagenda of NARSs (for example, NRMR), im- need of Bank financing. These include:proving efficiency (for example, research pri- * Construction of new research infrastruc-ority setting and evaluation), and improving ture. It is now clear that, for the present,


most NARIs have adequate-or even ex- * Investment in information systems is nowcess-physical infrastructure. Future a necessity to ensure that all scientists havestrategies have to consider options for re- ready access to the Internet and to scientificductions in physical plant: information and databases, nationally and

* Financing of operational costs. In the past internationally.external funding of operational costs de- * Foreign training, international exchanges,layed the development of sustainable in- and technical assistance, at lower levelsdigenous funding mechanisms for NARSs. than in the past, will remain important toHowever, operational costs may be neces- keep researchers abreast of advances insary and legitimate investments in smaller, science and to develop expertise in newpoorer NARSs, where there are few alter- priority research areas.natives. Even in some of the larger systems, * The costs of regional research and intema-operational cost financing may be justi- tional collaborative research will need tofied in the short term to influence re- be incorporated into NARS budgets.search system programs and organization Investments in local universities to supportand to shift attention to new research pri- both research and agricultural scienceorities. However, the total volume of op- training.6 The upgrading of human re-erational cost financing should decline sources in NARSs will be critically depend-over time. ent on the quality of local undergraduate

* Foreign training to expand the capacities and graduate teaching programs.of scientific and management staff of * Financing of management and policy re-NARSs, and technical assistance to con- forms, including building new mecha-duct research in NARSs. System expansion nisms for funding, training in researchhas now slowed, and qualified and capable management, and outside contracting oflocal staff are now much more available. management expertise for specific areas.

Although much of the costly, big-ticket items These shifts in project funding toward lessare less needed now than in the past, there costly items, and the critical need to focus onremains some important financing needs for financial sustainability, imply that future Bankmany NARSs, especially in public sector re- projects will be smaller, but more complex andsearch institutes: managerially intensive in preparation and su-

* Equipment and materials-usually in- pervision. Overall lending to agricultural re-cluding some imported materials-will be search may fall (although this depends oncontinuing needs as NARSs adopt new lending to countries in Eastern Europe andtechnologies and equipment and take on Central Asia that have, to date, requested littlenew research priorities. Modest funding support for agricultural research), but thisfor equipment can have a substantial im- may not be accompanied by a decrease in Bankpact on the productivity of research staff input, given the complexity of the pro-systems. jects. Since many of the projects will stress

* Facility maintenance and repair will be drastic institutional change, there is a strongsorely needed to correct errors of the past case for piloting certain activities before fur-in overbuilding and under-maintaining ther expansion.facilities. Upgrading-and in some cases Financing of sequential projects over a pe-relocating-experimental farms to im- riod of 10-20 years is still needed in many re-prove research quality will be important. gions, especially in Africa. This providesSuch investments should always be condi- greater stability of funding for NARS pro-tioned on governmental commitment and grams and sufficient time for institutional de-capacity to provide adequate funding for velopment and the completion and realizationfuture maintenance. of the benefits of research. Long-term projects


may reduce overall lending costs and help tural knowledge and information systems is amaintain current levels of finance for the sub- promising vehicle to improve technical back-sector, despite lower annual disbursements. stopping and portfolio monitoring. Continued

Bank financing of research is increasingly be- training on research management and policying included as a component of larger projects, issues is essential, especially as this relatesespecially projects that combine research and to the evolution of institutionally plural-extension and sector investment programs. It istic NARSs and the facilitation of researchis essential that these be monitored carefully, spill-ins.in view of Pritchard's (1990) finding that:

"problems in implementing research com- Improving Monitoring and Evaluation

ponents in agricultural and rural develop-ment projects were sometimes severe and Monitoring and Evaluation in agricultural re-the limited success of these components is search projects requires renewed attention. Theah causted fucncern."the development of M&E capacity of research pro-

a cause for concern.' grams in research institutions provides the ob-In research components of projects particular vious mechanism for improving the M&E of

care is needed to: research supported under Bank projects; that* Ensure adequate (that is, relatively high) is, there should be no need for a separate M&E

research management and technical input system for Bank-supported research. Gener-and consultation on design and supervi- ally, however, Bank-funded projects will havesion institutional development objectives addi-

* Include funding for international linkages, tional to the objectives of the research systemtechnical assistance, and training and that require additional performance indi-

* Provide support within a framework that cators. These are often difficult to quantify andleads to the development of sustainable measure. Ultimately, of course, successfulNARS institutions achievement of research program objectives

, Provide adequate levels of support to should be taken as an indicator of the successmake a real impact on research programs. of institutional changes. However, because of

the long time lag, intermediate indicators areBank Staffing and Project Supervision needed.

A preliminary set of indicators (annex 3) forLevels of staff inputs into both design and su- research projects has been established for re-pervision of agricultural research projects need search management, research execution, andto be protected, regardless of whether research research capacity building (World Bank 1995).is supported in free-standing projects or as a These indicators are now being implementedcomponent of a larger project. Research and in several projects and the Bank should reviewextension projects are staff-intensive because their application in practice and developthe technical and institutional issues are com- guidelines for best practices in monitoring theplex. The level and technical content of super- performance of research projects.vision can be important to the performance of Most Bank projects are too short to measureagricultural research and extension projects impact during the project life cycle. However,(World Bank 1997b). The Bank will need to find in most cases, agricultural research is fundedways to maintain direct staff supervision, com- in sequential projects and an ongoing processplemented by external expertise. to measure impacts from research funded in

With the rapid changes in the subsector, staff earlier projects is required to provide con-must be well supported to maintain effective- tinuing feedback to research planning andness. The emerging thematic group for agricul- operations.


Knowledge Sharing: Dissemination across regions to discuss and develop initia-of Good Practices tives in agricultural research, extension, and

education. As with any evolving system, oneThe promotion of good practices in agricul- clear imperative for this thematic group willtural research presents special challenges, as be to set in place arrangements for systemati-best practice may vary widely across countries, cally tracking change and for monitoring newdepending on the stage of development of the developments.NARS, the type of agriculture, and the size and The dissemination of good practice resultspolitical organization of the country. The Bank within and across NARSs is critical, especiallyshould continue to develop principles of good in building borrower commitment to institu-practice to guide research project design. The tional and policy changes. A variety of mecha-application of those principles will require nisms, including case studies, study tours, andcontinuing dialogue both within the Bank regional seminars, are needed to promote theseand with country partners to identify those activities. Since seeing is believing, the use ofpractices which are most appropriate to a study tours should be more widely encouragedgiven situation. Given the complexity of the so that staff may visit research systems in otherphenomena under consideration here, any rea- developing and industrialized countries,sonable analysis of good practice must neces- where specific aspects are regarded as modelssarily be multipronged, involving a mix of of best practice.active participation in relevant project work in The Bank is unlikely to have all the requiredthe regions, research-like investigations of re- skills in-house to evaluate and apply best prac-cent experience around the world, and on- tices. The research group will need to developgoing interaction with others engaged in ana- mechanisms for regularly interacting withlyzing the evolution of agricultural research others, such as the International Service forsystems. National Agricultural Research (ISNAR),

Good practice in agricultural research has actively engaged in agricultural researchbeen evolving rapidly over the past few years. policy and management. Partnerships, espe-To feed lessons back to future project design, cially those that access private sector experi-the Bank and others must be able to evaluate ence, may provide new perspectives and beand synthesize ongoing experiences with new particularly beneficial in technology systeminitiatives, such as competitive grants, research development.corporations and foundations, farmer financ-ing, and joint ventures with the private sector. NotesSpecial studies or reviews are needed to de-velop options for long-standing issues that 1. One example is United States assistance to NARShave bedeviled Bank and other projects. The development programs, which fell by 71 percent fromnew priorities should include: providing ap- 1985 to 1996 (Alex 1996).propriate incentives to researchers, estab- 2. A 1981 publication Agricultural Research: Sectorlishing effective research-extension linkages, Policy Paper was based largely on the 1980 Bank Reportmonitoring research impacts, developing effec- 2966, "Agricultural Research Systems: Sector Policytive priority-setting mechanisms, and linking Paper."NARIs to the private sector and NGOs. 3. SPAAR has promoted the development of Frame-

The formation of Bank networks and families works for Action (FFAs). These are comprehensive re-can provide the vehicle for both synthesis and gional action plans for improving NARS operations.

7 . Common elements or principles of the FFAs include:dissemination of best practice. A thematic developing country research strategies, developing sus-group for agricultural knowledge and informa- tainable financing systems, improving NARS manage-tion systems is being established across the ment capacities, establishing country research advisoryBank to bring all operational staff together groups, improving research-extension-farmer linkages,


and facilitating regional research spill-ins (Spurling 6. Such investments will confront some of the sameand others 1992; Weijenberg and others 1993; Weijen- limitations in financing need as for the NARS as aberg and others 1995). whole. Emphasis in university development programs

4. Although this principle is sound, developing in- must be on training students oriented to employmentstitutional and financial sustainability for many re- in private sector occupations or a pluralistic NARS andsearch institutions will be a long-term objective. There not traditional training, which emphasized prepara-will often be, as in Africa, a need to support interim tion for future government service.steps for establishing sustainable capacities. 7. A new World Bank Group initiative to improve staff

5. Explanations for this widespread situation in- professional development organizes staff in networks,clude historical lack of emphasis on social-science families, and thematic teams. An Agricultural Knowledgetraining within aid programs and strong job-market and Information Systems Thematic Team is being formedcompetition from the private sector and international in the Rural Development Family of the Environmen-assistance agencies. tally and Socially Sustainable Development Network.

Annexes

ANNEX 1

Main Recommendations from Reviewsof World Bank Research Projects

Issue 1980 Research Policy Paper 1983 Research Review 1981-87 Research Review 1996 0E Research Report 1997 Vision/action paper

Increasing resources for research

Mobilizing donor funds Lead an intemational effort to Take lead in donor Ensure consideration ofmobilize resources for coordination for agricultural agricultural research inagricultural research research intemational forums

Bank lending for research Increase Bank lending for Continue Bank lending to Consider support for targetedresearch projects develop strong NARSs research programs, if full

support to NARS developmentis not warranted

Financing operating costs Provide financing of Finance NARS operating costs Encourage increase in NARSincremental operating costs, only when borrowers operating cost funding inas needed demonstrate commitment to relation to salary costs

research agenda

Consideration of research in Provide fuller treatment of Define research linkages to Include consideration of Ensure consideration ofcountry dialog agricultural research issues in country's overall agricultural agricultural research issues in agricultural research in CAS

sector studies and policy dialog development in project CAS and in policy dialogue and in policy dialog withanalyses with borrowers borrowers

Undertake in-depthassessment of research needsand institutions for policy dialog

Govemment commitment to Require govemment Require demonstration ofresearch funding commitment to research and borrower commitment to

funding of research system research

Public awareness of research Expand analysis and Assist NARSs to developimpacts evaluation of research projects public awareness programs on

to consider research impacts value of research, and supporton employment and nutrition publication of research results

and accomplishments

Diversification of funding for Explore alternative funding Encourage research cost Assist NARSs to diversityNARSs mechanisms for providing recovery initiatives in NARSs research funding sources

operating funding for research

Improving research system efficiency and quality of research

Training for NARS staff Expand training programs to Emphasize training, especially Define research staffing needs Emphasize needs-basedstrengthen administrative and for research program leaders and provide adequate training training and technicaltechnical capacities of NARSs to meet those needs assistanceand tailor training to ensurethat staff return to NARSs

Incentives for researchers Provide improved incentives Provide incentives for retention Insist on performance Focus on institutional reformsand terms of service for of high-quality staff within incentives for NARS staff to improve researcherresearch scientists NARSs incentives

Long-term research program Provide longer-term (10- Expand time frame of researchsupport 15-year) support to NARSs projects

Monitoring and evaluation Assist NARSs to establish Emphasize monitoring andsystems monitoring and evaluation evaluation and socioeconomic

systems analysis within NARSs

Research and experiment Provide training in research Emphasize support for Support improvements in thestation management management and experiment research management and management of research

station management experiment station facilitiesmanagement

Input to quality of research Increase supervision and Improve supervision of Emphasize routine internal Emphasize technical contentevaluation of research technical quality of research and external review of of research in projectprojects, including annual and of institutional changes in reseanch programs preparation and supervisionworkshops, progress reviews, NARSsand external reviews

University development Strengthen graduate teaching Consider research-education Strengthen university capacityand research at developing sector linkages and their for research and teachingcountry universities importance for research and

training

Intra-NARS coordination Provide funding for contract Promote interagency Encourage university andresearch to use resources in coordination within NARSs NGO research through fundingvarious agencies of the NARSs competitive research grants

Other issues * For small countries, develop Increase assessment of As appropriate, rationalize Encourage increase in NARScapacity for adaptive research sociopolitical and cultural existing facilities and operating funding in relation toon small number of crops factors and increase borrower resources rather than salary costs

* Provide support for more participation in project design expanding systembasic and strategic research

(fabte continues on the foltowing page)

Main Recommendations from Reviews of World Bank Research Projects (continued)

Issue 1980 Research Policy Paper 1983 Research Review 1981-87 Research Review 1996 OEO Research Report 1997 Vision/action paper

Improving research priority setting and client orientation

National agricultural research Require development of an Encourage development of Continue to promote researchplans overall strategy for national national agricultural research master plans, but with greater

research systems plans national ownership of plans

Applied and adaptive research Target applied and adaptive Support adaptive and applied Require formulation of a client-research directly applicable to research for the whole farming responsive research plan asfarmers' problems system part of project preparation

Research on natural resources Support research onconservation and the conserving productivity ofenvironment natural resources and on

environmental consequencesof high input productionsystems

Farming systems approach Require a farmer-oriented Emphasize research aimed at Emphasize client-responsivephilosophy using FSR, the whole farm research through FSR, on-commodity, and farm research, andmultidisciplinary teams for stakeholder involvement inadaptive research on farmers' research planningproblems

Other issues * Target research on food Emphasize research for low- Emphasize ex ante economic Increase client involvement incrops for small farmers and potential areas evaluation and use of research research planning andthe poor and on nutrition and performance indicators implementation.food system impacts of Provide guidance on userresearch involvement in researchTarget farming systems design and evaluationresearch on neglected orresource-poor areas

Improving mechanisms for technology transfer

Research-extension linkages Strengthen research-extension Review concepts and pro- Consider technology Increase integration oflinkages by combining within a cedures for research/extension generation, acquisition, and research and extension.single agency linkages and develop opera- adaptation as an integral

tional guidelines system

Policy environment for Require borrower adoption oftechnology adoption policies conducive to

generation and adoption ofefficient technology

Improving linkages with the global agricultural research system

Support to CGIAR Continue support to IARCs Continue support to CGIARNARS-IARC linkages Link NARSs research work Take maximum advantage of Use Bank influence to in- Facilitate NARS linkages to

with the IARC programs NARS linkages to IARCs crease IARC support for lARCs and other researchresearch with NARSs under entitiesBank projects

Regional initiatives Support regional research Insist on and support regionalinitiatives research cooperation

Policy environment for private Ensure a favorable policy Encourage private sectorsector research environment for private sector research by removing

research and technology restrictions on R&D andacquisition technology imports.

Improving Bank support to agricultural research system development

Specialized Bank staff Increase the number of Train Bank staff in supervision Establish core of specialized Establish core staff to enhancespecialized Bank staff and appraisal of research research staff quality of agricultural research

projects projects and provide special-ized training relevant toagricultural research projects.

Portfolio review Examine causes for lack ofeffectiveness of researchcomponents of projects andevaluate effectiveness ofpolicy-based lending fordevelopment of researchcapacities

Project start-up problems Analyze administrative Provide support forproblems at project start-up procurement at beginning of

projects

Other issues Expand EDI course materials * Determine Bank policy withon the management, regard to biotechnologyorganization, and development researchpotential of research . Evaluate the contribution of

long- and short-term consult-ants to research projects

Note: Reviews are in Wodd Bank (1980, 1983, 1997a), Pritchard (1990), and Purcell and Anderson (1997).

ANNEX 2

World Bank Funding for Agricultural Researchby Year and by Geographic Region

South Asia Africa Latin America and the Caribbean Middle East and North Africa

Research Research Research ResearchTotal projects Full research funding Totalprojects Full research funding Totalprojects Fullresearch funding Totalprojects Fullresearch funding

Year with research projects (US$million) with research projects (US$million) with research projects (US$million) with research projects (US$million)

1981 6 1 27.75 14 1 26.25 4 2 243.17 2 0 2.16

1982 4 0 10.25 8 1 79.29 8 2 52.80 3 1 19.27

1983 12 0 10.50 13 1 31.30 4 2 154.40 3 0 3.801984 6 1 23.20 12 3 71.40 5 0 9.30 0 0 0.001985 7 0 5.20 13 3 76.70 4 3 34.50 5 1 9.90

1986 5 1 79.30 14 2 38.90 3 3 26.70 2 0 2.00

1987 5 1 24.40 18 4 87.60 8 3 67.10 1 0 0.20

1988 3 2 67.03 13 4 260.92 0 0 0.00 2 0 1.21

1989 2 1 24.65 14 8 127.95 2 1 19.80 3 1 31.38

1990 4 1 86.90 11 6 53.34 6 1 35.20 3 2 56.80

1991 1 0 3.00 13 4 104.41 3 3 74.80 0 0 0.00

1992 4 1 36.50 10 3 42.70 4 3 165.90 4 2 18.14

1993 6 2 33.86 7 4 94.42 2 1 12.23 3 1 39.00

1994 4 4 92.90 4 l 18.63 5 3 47.67 3 1 12.48

1995 5 3 40.00 8 2 78.14 6 3 147.40 1 0 0.791996 4 2 17.40 5 1 26.82 5 4 58.39 0 0 0.00

Total 78 20 582.84 177 48 1218.77 69 34 1149.36 35 9 197.13

Europe and CentralAsia East Asia and the Pacific Total

Total projects with Full research Researchfunding Totalprojectswith Full research Research funding Total projects with Full research Research fundingYear research projects (US$million) research projects (US$million) research projects (US$million)

1981 4 2 15.19 2 1 47.84 32 7 362.36

1982 2 0 4.64 2 1 33.50 27 5 199.75

1983 0 0 0.00 4 1 11.09 36 4 211.09

1984 2 1 18.20 7 0 11.90 32 5 134.00

1985 4 1 10.70 6 3 66.70 39 11 203.70

1986 1 0 0.40 3 1 9.60 28 7 156.90

1987 1 1 7.50 3 1 25.90 36 10 212.70

1988 0 0 0.00 2 2 15.60 20 8 344.76

1989 1 0 4.00 3 1 24.80 25 12 232.58

1990 1 1 10.80 7 4 37.03 32 15 280.07

1991 0 0 0.00 5 1 11.04 22 8 193.25

1992 1 1 77.60 3 0 2.50 26 10 343.34

1993 2 0 0.80 4 1 60.30 24 9 240.61

1994 3 1 8.27 8 2 25.50 27 12 205.45

1995 3 0 3.54 3 3 121.94 26 11 391.81

1996 5 3 35.83 7 1 17.53 26 11 155.97

Total 30 11 197.47 69 23 522.77 458 145 3,868.34

Source: ESDAR database on Wodd Bank research projects.

ANNEX 3

Illustrative Project Performance Indicatorsfor Agricultural Research Projects

P roject performance and impact are meas- chased, and training completed. The secondured by indicators, which provide direct level of process indicators is that of measuringor indirect measurement of program direct project activity accomplishments (output

progress. Project performance indicators indicators), such as trials conducted, technolo-are now required for all Bank projects. Objec- gies developed or released, buildings con-tively quantifiable indicators are desirable, structed, research plans completed, staff trainedbut it is often not possible to quantify program in an agency, and laboratories equipped.outcomes and qualitative or subjectively meas- Indicators of project impacts are more diffi-ured indicators may have to be used. Perform- cult to measure as these are less directly con-ance indicators are broken out in a cause- nected with project activity. A first leveland-effect hierarchy including: includes outcome indicators (or intermediate

* Impact indicators to measure develop- impact indicators or intermediate result indi-ment progress and impact of the project on cators). These measure changes that are fairlydesired social goals direct results of the project and may include:

- Project outcome indicators to measure re- technology adoption rates, measures of pro-sults directly attributable to the project ductivity of research institutions, and meas-

* Project output indicators to measure direct ures of the quality of research work. Changes"products" of the project in outcome indicators should be expected to be

* Project implementation indicators to measurable within a project lifetime. Impactmeasure inputs to project activities. indicators relate to the effects of the project on

These performance indicators essentially national social and economic objectives, suchmeasure performance at two major levels: proc- as productivity, poverty alleviation, and natu-ess indicators that measure the inputs and out- ral resource conservation. As the time lag forputs of research and are closely related to achieving these impacts is great and attributionresearch program activities and impact indica- of cause and effect between a project and thetors that reflect the development impacts of re- indicators is difficult, there is considerable de-search on national policy objectives. bate as to whether and how these can be meas-

Process indicators typically include measured and related to individual projects. Theseures of input usage (input indicators), such as are, however, important in that such impactsfunds expended, technical assistance provided, provide the rationale needed by those con-scientist and staff time used, equipment pur- cerned with funding research investments.

78

Illustrative Project Performance Indicators for Agricultural Research Projects 79

In designing a project performance monitoring jects relating to research management, researchsystem, it is important to keep the system prac- execution, and research capacity buildingtical and relatively simple, with limited numbers (table 3A.1). This list and systems for monitor-of performance indicators to be tracked in any ing project performance will need revision inparticular project. A preliminary checklist of po- response to further Bank experience and deter-tential indicators has been established for pro- mination of best practice.

Table 3A.1 Potential project performance indicators for research projects

Research infrastructure and human resourcesIndicator level/type Research management components Research program implementation components component

Social and economic impact To the extent that they can be associated with To the extent that they can be associated with To the extent that they can be associated withagricultural technology: agricultural technology: agricultural technology:* Increased factor productivity . Increased factor productivity * Increased factor productivity* Social rate of return to research investment * Social rate of return to research investment * Social rate of return to research investment* Trends in environmental indicators * Trends in environmental indicators * Trends in environmental indicators* Reduction in poverty * Reduction in poverty * Reduction in poverty* Improved nutrition * Improved nutrition * Improved nutrition

Project outcomes Objective measures: Objective measures: Objective measures:* Trends in recurrent costs relative to total . Adoption of released technologies and manage- * Number of peer-reviewed publications a year per

budget ment recommendations researcher* Resource allocation shifts toward established . Benefits of adoption of new technologies . Trained technicians per scientist

priorities . Adoption of resource conservation technologies . Research staff tumover rate* Share of private financing of total research * Proporion of laboratories equipped and functioning

expenditure* Share of budget of public sector self-generated* Number of joint ventures with the private sector* Percentage of budget allocated through competitive

00 grants* Ratio ot professional to support stafl* Share of contracted research

Subjective measures: Subjective measures: Subjective measures* Improved coherence of research program and . Regional and intemational collaboration estab- * Improved experiment station density and siting in

sector policy lished as integral part of research process relation to research needs* Improvement in management (reflected in staff . Disciplinary balance of research (for example, * Utlization and effectiveness of information systems

surveys) involvement of social scientists) (including electronic connectivity and libraries)t Improved staff morale and behavior (as perceived in * Staff motivation and moralestaff surveys)

* Capacity to undertake priority setting and planning* Improved articulation of objectives and strategy of

research program* Extent of stakeholder participation in research

govemance, priority setting, funding, and execution

Project outputs Objective measures: Objective measures: Objective measures:• Research plans established . New varieties released * Number of researchers trained* Research plans financed . Extension recommendations published . Number of female researchers trained* Peer-review mechanisms in place . Number and type of early adopting farmers and . Degrees awarded

estimated productivity impacts

. Monitoring and evaluation systems established, * Other research products, completed research . Laboratories builtincluding management information systems tools/methods developed . Journals/publications acquired

* Staff performance appraisal and promotion system * Papers published . Trainees' assessed job performance before andrevamped to reward productivity * Number of collaborative research programs after training

* Policy unit established established with national and intemational partners . Percentage of scientists linked to electronic informa-* Changes in orientation of research tion systems

(multidisciplinary, multi-institutional)

Subjective measures: Subjective measures: Subjective measures:* External management reviews . Improved knowledge of farming systems . Quality of construction* Quality of research-extension linkages . Improved knowledge of natural resources status * Number of trainees expressing satisfaction with* Adequacy of regulatory and legal frameworks and trends training

for research and research organizations . Improved research relevance through use of on- * Trainees' job performance* Adequacy of financial accounting procedures farm participatory methods * Quality of training* Evidence of feedback from M&E system to re- * Quality of civil works

search and management decisions* Degree of research networking

Project inputs/implementation Objective measures: Objective measures: Objective measures* Participatory forums established . Timely provision of funds, equipment, and staff time * Curriculums developed* Training and technical assistance requirements * Number of proposals processed and average time * Attendance records

identified and contracted to process . Training programs conducted* On-time flow of funds to research units * Implementation according to milestones established * Works contracted* On-time reporting of expenditures for each research project and program * Staff trained as per plan* Annual research reports produced a Proportion of researcher time in field or laboratory * Performance against construction schedules

* Completion of diagnostic reports on famer . Progress against timeline in competitive biddingconstraints and priorities processes

* Meetings of researchers and clients* Completion of scientific program reviews, inctuding

extemal reviews

Subjective measures: Subjective measures:* Quality of contacts with stakeholders * Degree of farmer participation in research* Stakeholders' satisfaction with research system assessments

* Farmer assessment of research process quality ofinformation and dialogue between researchers andclients

* Meetings of researchers with clients

Source: Substantialy modified and developed from Wodd Bank 1995, Annex 2.

Bibliography

Alex, G. 1996. "USAID and Agricultural Research: Re- Agency for International Development, LAC TECHview of USAID Support for Agricultural Research." Project, Washington, D.C.U. S. Agency for International Development, Office Basler, A. 1997. "Lessons Learned from Changes inof Agriculture and Food Security, Washington, D. C. Research Investment Strategies in Germany." Draft

Alston, J., G. Norton, and P. Pardey. 1995. Science under page for an ESDAR workshop on New InvestmentScarcity. Ithaca: Cornell University Press. Strategies for Agricultural and Natural Resources

Alston, J. M., and P. G. Pardey. 1996. Making Science Pay: Research, October 1-3, 1997. London, U.KEconomics of Financing, Organizing and Managing Bohn, A., and D. Byerlee. 1993. The Wheat Breeding In-Pttblic-Sector Agricultutral R&D. Washington, D.C.: dustry in Developing Coutntries: An Analysis of Invest-American Enterprise Institute. ments and Impacts. Part I of 1992/93 International

Alston, J. M., P. G. Pardey, and J. Roseboom. 1997. Centerfor Maize and Wheat Improvement (CIMMYT)"Financing Agricultural Research: International In- World Wheat Facts and Trends. Mexico: CIMMYT.vestment Patterns and Policy Perspectives." Paper Bonte-Friedheim, C., R. Tabor, and J. Roseboom. 1994.presented to the Conference of the International As- Financing National Agricultural Research: The Challengesociation of Agricultural Economists, August 10-16, Ahead. Briefing Paper 11. The Hague: ISNAR.Sacramento, Calif. Boughton, D., E. Crawford, J. Howard, J. Oehmke, J.

Anandajayasekeram, P., and M. Rukuni. 1994. "Agri- Shaffer, and Staatz, J. 1995. "A Strategic Approach tocultural Research in Eastern and Southem Africa: Agricultural Research Program Planning in Sub-Issues, Policy and Institutional Challenges." Paper Saharan Africa." Policy Synthesis 8, U. S. Agency forpresented at the International Conference for International Development, Office of Sustainable De-Agricultural Economists, August 22-29, Harare, velopment, Washington, D.C.Zimbabwe. Brennan, J. P. Forthcoming. "Efficiency in Wheat Im-

Anderson, J. R., P. G. Pardey, and J. Roseboom. 1994. provement Research: A Case Study of Wheat Im-"Sustaining Growth in Agriculture: A Quantitative provement Research in Australia." In M. MarediaReview of Agricultural Research Investments." and D. Byerlee, eds., Efficiency of Investment in Na-Agricultural Economics 10: 107-23. tional and International Wheat Improvement Research.

Antholt, C. H. 1994. Getting Ready for the Twenty-First International Center for Maize and Wheat Improve-Century: Technical Change and Institutional Modern- ment (CIMMYT) Research Report, Mexico, D.F.:ization in Agriculture. World Bank Technical Paper CIMMYT.217. Washington, D.C. Buckles, D., and 0. Erenstein, 1996. Intensifying Maize-

Ashby, J. A., T. Gracia, M. Guerrero, C. A. Quiros, J. I. based Cropping Systems in the Sierra de Santa Marta,Roa, and J. A. Beltran. 1995. "Institutionalizing Veracnrz. Natural Resources Group Paper 96-07,Farmer Participation in Adaptive Technology Test- Mexico D.F.: CIMMYT.ing with the 'CIAL'." Network Paper 57. Agricultural Byerlee, D. 1996. "Modern Varieties, Productivity andResearch and Extension Network, Overseas Devel- Sustainability: Recent Experience and Emergingopment Institute, London. Challenges." World Development 24(4): 697-718.

Bathrick, D. D., K. J. Byrnes, J. G. Stovall, and D. R. . 1997. "Knowledge-Intensive Crop Manage-Podems. 1996. "Technology Institutions for Agricul- ment Technologies: Concepts, Impacts, and Pros-tural Free Trade in the Americas (TIAFTA)." U. S. pects in Asian Agriculture." In M. Hossain and

83


A. Siamwalla, eds., Impacts of3O Years of Rice Research Echeverria, R. G. 1990. "Assessing the Impact of Agri-in Asia. Los Banos, The Philippines: International cultural Research." In R. G. Echeverria, ed., MethodsRice Research Institute. for Diagnosing Research System Constraints and Assess-

Byerlee, D., and G. Alex. 1997. "Towards More Effective ing the Impact of Agricultural Research. Vol. II AssessingUtilization of External Assistance in Building Agri- the Impact of Agricuiltural Research. The Hague:cultural Research Systems." Sourcebook on Financing ISNAR.Agricultutral Research. The Hague: ISNAR. . 1991. "Impact of Research and Seed Trade on

Byerlee, D., and C. K. Eicher. 1997. Africa's Emerging Maize Productivity." In P. Pardey, J. Roseboom, andMaize Revoltution. Boulder, Colo.: Lynne Rienner. J. R. Anderson, eds., Agricultural Research Policy: In-

Byerlee, D., and S. Franzel. 1993. Institutionalizing the ternational Quantitative Perspectives. Cambridge:Role of the Economist in National Agricultutral Research Cambridge University Press.Institutes. Economics Working Paper 93-01. Mexico, Echeverria, R. G., E. J. Trigo, and D. Byerlee. 1996.D.F.: CIMMYT. Institutional Change and Effective Financing of Agricul-

Byerlee, D., and M. Lopez-Pereira. 1994. Technical ttural Research in Latin America. World Bank TechnicalChange in Maize: A Global Perspective. Economics Paper 330, Washington, D.C.Working Paper 94-02. Mexico: CIMMYT. Eicher, C. K. 1990. "African Scientific Capacity for Ag-

Byerlee, D., and P. Pingali. 1995. "Asian NARSs: Frus- ricultural Development." Agricultfure Economics 4trations and Fulfillments." In G. H. Peters and D. (May): 117-43.Headley, eds., Agricultural Competitiveness: Market Elliott, H., and P. G. Pardey. 1988. "Determinants ofForces and Policy Choice: Proceedings of the Twenty- Support for National Agricultural Research Sys-Second International Conference of Agricultural Econo- tems." In E. Javier and U. Renborg, eds., The Changingmists, Harare. Dartmouth, U. K.: Aldershot. Dynamics of Global Agriculture. The Hague: ISNAR.

Byerlee, D., and G. Traxler. 1995. "National and Inter- Ellsworth, L. 1997. The Road to Financial Sustainability: Anational Research in the Post-Green Revolution Report of the Stustainable Financing Initiative. U. S.Period: Evolution and Impacts." American Jouirnal of Agency for International Development, AgriculturalAgriculttural Economics 77(2): 268-78. Policy Analysis Project III. Washington, D.C.

. 1996. "The Role of Technology Spill-overs and Eponou, T. 1996. Partners in Technology Generation andEconomies of Size in the Efficient Design of Agricul- Transfer: Linkages Between Research and Farmer Organi-tural Research Systems." Paper presented at the con- zations in Three Selected African Countries. Researchference, Global Agricultural Science Policy to the 21st Paper 9. The Hague: ISNAR.Century, August 26-28, Melbourne. Evenson, R.E. 1994. "Analyzing the Transfer of Agricul-

Cassman, K. G., and P. L. Pingali. 1995. "Extrapolating tural Technology." InJ. R. Anderson, ed., AgriculturalTrends from Long-Term Experiments to Farmers Technology: Policy Issues for the International Commu-Fields: The Case of Irrigated Rice System in Asia." In nity. Wallingford, U.K.: CAB International.V. Bamett, R. Payne, and R. Steiner, eds., Agricultural . 1997. "Rice Varietal Improvement and Inter-Sustainability in Economic, Environmental, and Statisti- national Exchange of Germplasm," In M. Hossain andcal Terms. London: John Wiley & Sons, Ltd. A. Siamwalla, eds., Impacts of 30 Years of Rice Research

Collion, M. H., and A. Kissi. 1995. Guiide to Program in Asia. Los Banos, The Philippines: International RicePlanning and Priority Setting. Research Management Research Institute.Guidelines 2E. The Hague: ISNAR. Evenson, R. E., and M. W. Rosegrant. 1993. "Determi-

Collion, M. H., and P. Rondot. 1997. "Lessons from nants of Productivity Growth in Asian Agriculture:Experiences and Conclusions from a Workshop on Past and Future." Paper presented to the AAEAPartnerships between Agricultural Services Institu- Preconference Workshop on Post-Green Revolutiontions and Producer Organizations: Myth or Reality?" Agricultural Development Strategies: What Next?World Bank, Africa Regional Office, Washington, D. C. July 30-31, Orlando, Fla.

Crosson, P., and J. R. Anderson. 1993. Integration of Eyzaguirre, P. 1996. Agriculture and Environmental Re-Natural Resource and Environmental Issues in the Re- search in Small Cotuntries: Innovative Approaches to Stra-search Agendas of NARS. Briefing Paper 7, The Hague: tegic Planning. Chichester: John Wiley & Sons.ISNAR. Falconi, C., and H. Elliott. 1995. "Public and Private

David, C., and K. Otsuka, eds. 1994. Modern Rice Tech- R&D in Latin America and the Caribbean." In G. H.nology and lncome Distributtion in Asia. Boulder, Colo.: Peters and D. Headley, eds., Agricultural Competitive-Lynne Rienner. ness: Market Forces and Policy Choice: Proceedings of the

Davis, J. S., P. A. Oram, and J. G. Ryan. 1987. Assessment Twenty-Second International Conference of Agriculturalof Agricultural Research Priorities: An International Per- Economists, Harare. Aldershot, U. K.: Dartmouth.spective. Canberra: Australian Centre for Interna- FAO (U.N. Food and Agriculture Organization). 1996.tional Agricultural Research. "Role of Research in Global Food Security and Agri-

Bibliography 85

cultural Development." Background paper for Technology: Policy Issues for the International Commu-World Food Summit, FAO, Rome. nity. Wallingford, U.K.: CAB International.

Fuglie, K., N. Ballenger, K. Day, C. Klotz, M. Ollinger, Kaimowitz, D. 1993. "The Role of Non-GovernmentalJ. Reilly, U. Vasavada, and J. Yee. 1996. Agricultural Organizations in Agricultural Research and Technol-Research and Development: Public and Private Invest- ogy Transfer in Latin America." World Developmentments under Alternative Markets and Institutions. Agri- 21(7): 1139-50.cultural Economics Report 735. Washington, D.C.: Kumar, K. 1995. Generating Broad-Based Growth throughU. S. Department of Agriculture, Economic Research Agributsiness Promotion: Assessment of USAID Experi-Service. ence. U. S. Agency for International Development,

Gilles, J. L. 1997. Checkoffs: New Approaches to Funding Center for Development Information and Evalu-Research, Development, and Conservation Programs. ation, Washington, D.C.Sustainable Development Publication Series Techni- Lele, U. 1995. "Building on the NARS-CGIAR Partner-cal Paper 53. Washington, D. C.: USAID. ship for a Doubly Green Revolution: A Framework

Gisselquist, D. 1994. "Import Barriers for Agricultural for the IFAD-Led Initiative." Paper prepared for theInputs." U.N. Development Programme-World meeting on Strengthening NARS-CGIAR Partner-Bank Trade Expansion Program Occasional Paper ships: NARS Outline Action Plan. October 28. World10. World Bank, Washington, D. C. Bank, Washington, D. C.

Grindle, M. S. 1997. "Divergent Cultures? When Public Lele, U., and R. Coffman. 1995. Global Research in theOrganizations Perform Well in Developing Coun- Environmental and Agricultural Nextusfor the 21st Cen-tries." World Development 25(4): 481-95. tury: A Proposal for Collaborative Research among U.S.

Gulati, A., and A. Sharma. 1995. "Subsidy Syndrome in Universities, CGIAR Centers, and Developing CountryIndian Agriculture." Economic and Political Weekly Institutions. Report of a Taskforce on Research Inno-30(39): A93-102. vations for Productivity and Sustainability. Gaines-

Hayami, Y., and V. W. Ruttan. 1975. Agricuiltural Devel- ville, Fla.: Cornell University and University ofopment: An International Perspective. Baltimore: Johns Florida.Hopkins University Press. Lipton, M. 1994. "Agricultural Research Investors'

Hayami, Y., M. Kikuchi, and K. Morooka. 1989. "Market Themes and Issues: A View from Social Science Re-Price Response of World Rice Research." Agricultural search." In J. R. Anderson, ed., Agriciultural Technol-Economics 3(4): 333-34. ogy: Policy Issues for the International Community.

Horkan, K. M., and P. L. Jordan. 1996. "Endowments as Wallingford, U. K.: CAB International.a Tool for Sustainable Development." U. S. Agency Lopez-Pereira, M. A., and M. P. Filippello. 1994. "Maizefor International Development, Center for Develop- Seed Industries Revisited: Emerging Roles of thement Information and Evaluation, Washington, D.C. Public and Private Sectors." Part 1 of International

Hoste, C. H., H. Michelson, H. Nouwakpo, L. Olug- Center for Maize and Wheat Improvement (CIMMYT)bemi, and L. Zuidema. 1995. A Framework to 1993/94. World Maize Facts and Trends: Maize SeedStrengthen the Role of Universities in National Agricul- Industries, Revisited: Emerging Roles of the Public andtural Research Systems. ISNAR Briefing Paper 24. The Private Sectors. Mexico, D. F.Hague. Lopez-Pereira, M. A., and J. C. Garcia. 1994. The Maize

Huffman, W. E., and R. E. Just. 1994. "Funding, Struc- Seed Industries of Brazil and Mexico: Past Performance,ture, and Management of Public Agricultural Re- Current Issues, and Futuere Prospects. CIMMYT Eco-search in the United States." American Journal of nomics Paper 97-02. Mexico, D. F.Agricultural Economics 76(4): 744-59. Maredia, M. K., and D. Byerlee. 1996. "Efficiency of

Inter-American Development Bank. 1996. Toward a Re- Wheat Improvement Research: A Comparativegional System of Technology Innovationfor the Food and Analysis of National and International Research Sys-Agricultural Sector. Washington, D.C. terns in Developing Countries." Paper presented at

James, C. 1996. "Agricultural Research and Develop- the conference, Global Agricultural Science Policy toment: The Need for Public-Private Sector Partner- the Twenty-First Century, August 26-28, Melbourne,ships." Issues in Agriculture 9. CGIAR (Consultative Australia.Group on International Agricultural Research), Maredia, M. K.,J. Howard, and D. Boughton. Forthcom-Washington, D.C. ing. "No Shortcuts to Progress: An Assessment of

Janssen, W. G. 1993. "Economic and Agricultural Devel- Agricultural Research Planning and Priority Settingopment in West Asia and North Africa-The Need in Africa." International Development Workingfor Agricultural Research." Food Policy 18(6): 507-22. Paper, Michigan State University, East Lansing.

Jarvis, L. S. 1994. "Changing Private and Public Roles in Martinez, J. C., and G. Sain. 1991. "Toward Farm-basedTechnological Development: Lessons from the Chil- Policy Analysis: Concepts Applied in Haiti." Agricul-ean Fruit Sector." In J. R. Anderson, ed., Agricultutral tural Economics 5(3): 223-35.


McCalla, A. 1994. "Agriculture and Food Needs to 2025: Pardey, P. G., J. Roseboom, and N. M. Beintema. 1997.Why We Should Be Concerned." Sir John Crawford "Investments in African Agricultural Research."Memorial Lecture. CGIAR, Washington, D. C. World Development 25(3): 409-23.

Mclntire, J. 1994. "Reviving the National Agricultural Re- Petit, M. J., G. Alex, H. Blackburn, W. Collins, J. Doyle,search System of Mexico." In J. R. Anderson, ed., Agri- R. Freed, F. Heidhues, U. Lele, G. Persley, and H.cultural Technology: Policy Issues for the International Rouille D'Orfeuil. 1996. The Emergence of a GlobalCommunity. Wallingford, U.K.: CAB International. Agricultural Research System: The Role of the Agricul-

McMahon, M. 1992. "Getting Beyond the 'National In- tural Research and Extension Group (ESDAR). ESDARstitute Model' for Agricultural Research in Latin Report 1. Washington, D.C.: World Bank.America: A Cross-Country Study of Brazil, Chile, Pingali, P. L., and P. W. Heisey. 1996. "Cereal CropColombia, and Mexico." Regional Studies Program Productivity in Developing Countries: Past TrendsReport 20,World Bank, Latin America and the Carib- and Future Prospects." Paper presented at the con-bean Technical Department, Washington, D. C. ference, Global Agricultural Science Policy to the

Merrill-Sands, D., and M. H. Collion. 1992. "Making the Twenty-First Century, August 26-28. Melbourne,Farmer's Voice Count in Agricultural Research.." Australia.Quarterly Joutrnal of International Agriculture 32(3): Pray, C. E., and J. R. Anderson. 1997. "The Agricultural260-79. Research System of the Former Soviet Union: Past

Merrill-Sands, D., D. Kaimowitz, K. Sayce, and and Future." Joturnal of International Development 9(3):S. Chater. 1990. The Technology Triangle: Linking Farm- 517-27.ers, Technology Transfer Agents, and Agricultural Re- Pray, C. E., and D. Umali-Deininger. 1997. "Privatesearchers. The Hague: ISNAR. Agricultural Research: Will it Fill the Gap?" Paper

Morris, M., and D. Byerlee. Forthcoming. "Maintaining presented at the Conference of the International As-Productivity Gains in Post-Green Revolution Agri- sociation of Agricultural Economists, August 10-16,culture." In C. K. Eicher and J. Staatz, eds., Agricul- Sacramento, Calif.tural Development: The Third World (3rd ed.). Baltimore: Pritchard, A. J. 1990. Lending by the World Bankfor Agri-Johns Hopkins Press. cultural Research: A Review of the Years 1981 through

Murphy, J. 1993. Monitoring and Evaluation in Agricul- 1987. Washington, D.C.: World Bank.t ural Research: Concepts, Organization, and Methods. . 1994. "World Bank Investment in AgriculturalInternational Service for National Agricultural Re- Research Policy and Strategy: Past and Future." Insearch, Informal Report. The Hague. J. R. Anderson, ed., AgriculttLral Technology: Policy

Nickel, J. 1989. Research Management for Development: Issues for the International Community. Wallingford,Open Letter to a New Agricultural Research Director. San U.K.: CAB International.Jose, Costa Rica: Inter-American Institute for Coop- Purcell, D. L. 1994. "The World Bank Experience witheration on Agriculture. NARS-Building." In J. R. Anderson, ed., Agricultural

. 1997. "Institutional Aspects Related to Agricul- Technology: Policy Issuesfor the International Commu-tural Research in Pakistan." World Bank, South Asia nity. Wallingford, U.K.: CAB International.Regional Office, Washington, D. C. Purcell, D. L., and J. R. Anderson. 1997. Agricultural

Odame, H., and L. Setshwaelo. 1997. Agricultural Re- Extension and Research: Achievements and Problems insearch Plans in Sub-Saharan Africa: A Stattus Report. The National Systems. Washington, D.C.: World Bank.Hague: ISNAR. Ravnborg, H. M. 1993. Planning and Priority Setting in

Oehmke, J. F., and E. W. Crawford. 1996. "The Impact Agricultural Research: The Case of Tanzania. Workingof Agricultural Technology in Sub-Saharan Africa." Paper 93.5. Copenhagen: Centre for DevelopmentJoturnal of African Economies 5(2): 271-92. Research.

Okali, C., J. Surnberg, and J. Farrington. 1996. Farmer Renkow, M. 1993. "Differential Technology AdoptionParticipatory Research : Rhetoric and Reality. London: and Income Distribution in Pakistan: Implications forIntermediate Technology, on Behalf of the Overseas Research Resource Allocation." American Journal ofDevelopment Institute. Agricultural Economics 75(1): 33-43.

Okello, A., and P. Eyzaguirre. 1992. NationalAgricultuhral Ricks, D. 1997a. Apple Industry Strategic Planning. StaffResearch in a Regional Context: The Small Countries of Paper 97-13. Department of Agricultural Econom-Southern Africa. ISNAR Small Countries Studies Pa- ics. East Lansing: Michigan State University.per 7. The Hague: ISNAR. . 1997b. Use of Consumer Market Researchfor Apple

Pardey, P. G., J. Roseboom, and J. R. Anderson. 1991. Industry Planning. Staff Paper 97-8. Department of"Topical Perspectives on National Agricultural Agricultural Economics. East Lansing: MichiganResearch." In P. G. Pardey, J. Roseboom, and J. R. State University.Anderson, eds., Agricultural Research Policy: Interna- Rubey, L., R. W. Ward, and D. Tschirley. 1997. "Maizetional Quantitative Perspectives. Cambridge, U.K.: Research Priorities: The Role of Consumer Prefer-Cambridge University Press. ences." In D. Byerlee and C. K. Eicher, eds., Africa's

Bibliography 87

Emerging Maize Revolution. Boulder, Colo.: Lynne Agricultural Economics Association. Aldershot, U.K.:Reinner. Gower.

Rukuni, M. 1996. "A Framework for Crafting Demand- Traxler, G., D. Byerlee, and K. B. L. Jain. 1996. "LinkingDriven National Agricultural Research Institutions Technical Change to Research Effort: An Examina-in Southern Africa." Draft paper. U. S. Agency for tion of Aggregation and Spill-over Effects." Depart-International Development, Food Security II Project, ment of Agricultural Economics, Auburn University,Washington, D.C. Auburn, Alabama.

Sarles, M. 1990. "USAID's Experiment with the Private Trigo, E. J. 1987. "Agricultural Research OrganizationSector in Agricultural Research in Latin America and in the Developing World: Diversity and Evolution."the Caribbean." In R. G. Echeverria, ed., Methods for In V. W. Ruttan and C. E. Pray, eds., Policy for Agri-Diagnosing Research System Constraints and Assessing cultural Research. Boulder, Colo.: Westview.the Impacts of Agricultural Research. The Hague: Tripp, R. 1989. Farmer Participation in Agricultural Re-ISNAR. search: New Directions or Old Problems? Institute of

Schillhom van Veen, T. W., D. A. Forno, S. Joffe, D. L. Development Studies, Discussion Paper 256, Sussex,Umali-Deininger, and S. Cooke. 1997. Integrated Pest U.K.Management: Strategies and Policies for Effective Imple- Tripp, R., P. Anandajayasekeram, D. Byerlee, andmentation. Environmentally Sustainable Develop- L. Harrington. 1990. "Farming Systems Research Re-ment Studies and Monograph Series 13. Washington, visited." In C. K. Eicher and J. W. Staatz, eds., Agri-D.C.: World Bank. cultural Development in the Third World. (2d ed.)

Schultz, T. W. 1964. Transforming Traditional Agriculture. Baltimore: Johns Hopkins University Press.Chicago: University of Chicago Press. Umali, D. L. 1992. Public and Private Sector Roles in Agrictul-

Schweikhardt, D. B., and J. T. Bonnen. 1992. "Financing tutral Research: Theory and Experience. World Bank Dis-Agricultural Research in the Presence of Intemational cussion Paper 176. Washington, D. C.Benefit Spill-overs: The Need for International Coop- UNEP (U.N. Environment Programme). 1997. "To-eration and Innovation." In G. H. Peters and B. F. wards More Efficient Use of GIS Analysis by theStanton, eds., Sustainable Agricultural Development: The CGIAR System and its Partners." Global ResourceRole of International Cooperation. Proceedings of the Information Database (GRID) CGIAR Project Discus-Twenty-First International Conference of Agricultural sion Paper. Arandal, Norway.Economists, Tokyo, Japan. London: Gower Publications. Weijenberg, J., J. Dione, M. Fuchs-Carsch, A. Kere, and

Singh, A., S. Pal, and D. Jha. 1997. "Public Investment J. Lefort. 1993. Revitalizing Agricultural Research in thein Agricultural Research and Education in India: Sahel: A Proposed Framework for Action. World BankAgenda for the Ninth Plan." National Center for Discussion Paper 211. Washington, D.C.Agricultural Economics and Policy Research, New Weijenberg, J., M. Dagg, J. Kampen, M. Kalunda, A. M.Delhi. Mailu, S. Ketema, L. Navarro, and M. A. Noor. 1995.

SPAAR (Special Program for African Agricultural Re- Strengthening National Agricultural Research Systems insearch). 1996. "The SPAAR Consolidated Funding Eastern and Central Africa : A Framework for Action.Mechanism Concept Takes Shape in Zambia." World World Bank Technical Paper 290. Washington, D.C.Bank, SPAAR Secretariat. Washington, D.C. World Bank. 1980. Agricultural Research Systems: Sector

Spurling, A. T., T. Y. Pee, G. Mkamanga, and C. Nkwan- Policy Paper. Agriculture and Rural Developmentyana. 1992. Agricultural Research in Souithern Africa: A Department, Report 2966. Washington, D.C.Frameworkfor Action. World Bank Discussion Paper . 1981. Agriculitural Research: Sector Policy Paper.184, Washington, D.C. Washington, D.C.

Taylor, A., G. Boukambou, M. Dahniya, B. Ouayogode, . 1983. "Strengthening Agricultural ResearchR. Ayling, M. Noor, and M. Toure. 1996. Strengthen- and Extension: The World Bank Experience." Wash-ing National Agricuiltural Research Systems in the Humid ington, D.C.and Suab-Humid Zones of West and Central Africa: A . 1995. "Performance Indicators in Bank- FinancedFramework for Action. World Bank Technical Paper Agricultural Projects: A First Edition Note." Agricul-318. Washington, D.C. tural and Natural Resources Department. Washing-

Thirtle, C., V. E. Ball, J. C. Bureau, and R. Townsend. ton, D.C.1995. "Accounting for Efficiency Differences in Euro- . 1997a. Rutral Development: From Vision to Action,pean Agriculture: Cointegration, Multilateral A Sector Strategy Paper. ESSD Studies and Mono-Productivity Indices, and R&D Spill-overs." In graphs Series 12. Washington, D.C.D. Headley and G. H. Peters, eds., Agricultutral Com- . 1997b. "Portfolio Improvement Program: Agri-petitiveness, Market Forces and Policy Choices. Proceed- cultural Research and Extension 'At-Risk' Project Re-ings of the XXII Conference of the International view." Washington, D.C.

Distributors of COLOMBIA GERMANY ISRtAEL NEPAL PORTUGAL SWEDEN

Infenlace Llda. UNO-Vedag Yozmot Lherature Ltd. Everest Media Intemational Services (P) Lid. Livrada Portugal Wenrergren-Williams AB

W oorld Bank Carrerna6 No. 51-21 PoppelsdorlerAllee55 PO. vox 56055 GPO Box 5443 Apartado2681,RuaDoCamno70-74 P.O. Box 1305

Apartado Aereo 34270 53115 Bonn 3 Yohanan Hasandlar Street Kathmandu 1200 Lisbon S-171 25 Solna

Publications Sanltal de Bogota, D.C. Tel: (49 228) 949020 Tel Aviv 61560 Tel: (977 1) 472152 Tel: (1) 347-4982 Tel: (46 8) 705-97-50

Prices and credit terms var from Tel:(57 1)285-2798 Faxn (49 228) 217492 Tel: (9723) 5285-397 Fax: (977 1)224431 Fax: (1) 3470264 Fax: (468) 27-00-71

countrcountrountry. Convrt yosrr Fax: (57 1) 265-2798 URL: http://www.uno-vedag.de Fax: (972 3) 5285-397 E-mail: [email protected]

local dtstributor before placing an E-mail: [email protected] NETHERLANDS ROMANIAlocal dlstributor before placin an COTE D'IVOIRE R.OY Intemational De LindeboorollnOr-Publikalies Compani De Librarii Bucuresti S.A. SWITZERLAND

order. Center d'Edition et de Diffusion Atricaines GREECE PO Box 13056 P.O. Box 202,7480 AE Haaksbergen Str. Lipscani no. 26, sector 3 Librairie Payot Service lnstlautionnel

ARGENTINA (CEDA) Papasotiriou s A. Tel Aviv 61130 Tel: (31 53) 574-0004 Bucharest C6tes-de-Montbenon 30

04 B.P 541 35, Stoumara Str. Tel: (9723) 5461423 Fax: (31 53)572-9296 Tel: (40 1)6139645 1002 Lausanne

Oficivadel Libra Intemacional Abidjan 04 10682 Athens Fax: (9723) 5461442 E-mail: [email protected] Fax: (401) 3124000 Tel: (4121) 341-3229

Iv120 Buenos A1es Tel: (225) 246510;246511 Tel: (301) 364-1826 E-mail: royiltnetvision.net.il URL: http://www .woddonline.nV-lindeboo Fax: (4121) 341-3235

Te:1(541(0815-835 ArFax: (225) 25 0567 Fax: (301) 364-8254 RUSSIAN FEDERATIONTel: (541)815-8354 .Palestinian Authority/Middle East NEW ZEALAND Isdatelstvo <Ves Mirj ADECO Van Diermmen EddionsTechniques

Fax: (541) 815-8156 CYPRUS HAITI Index Infonmation Services EBSCO NZ Ltd. 9a, Kolpachniy Pereulok Ch. de Lacuez 41

AUSTRALIA, FIJI, PAPUA NEW GUINEA Center for Applied Research Culure Dhusion PO.B. 19502 Jerusalem Private Mail Bag 99914 Moscow 101331 CH1807 Bonay

SOLOMON ISLANDS, VANUATU, AND Cyprus College 5, Rue Capois Tel: (9722)62712t9 New Market Tel: (7095)9178749 Tel: (41 21) 943 2673

WESTERN SAMOA 6, D3iogenes Street, Engomi C.P 257 Fax: (972 2(6271634 Auckland Fan: (7 095) 917 92 59 Fax: (41 21) 943 3605

D.A. nfornalio SericesPO. Box 2056 Fuol-as-Prince Tel: (64 9) 524-8119

D.A. ntnlomation Services Nicosia Tel: (509) 23 9260 ITALY Fax: (64 9) 524-8067 SINGAPORE, TAIWAN, THAILAND

648 WhOehorse Road Tel: (357 2)44-1730 Fax: (509) 23 4858 Licosa Commissionaria Sansoni SPA MYANMAR, BRUNEI Central Books Distribution

Mhcham 3132 Fax: (357 2) 46-2051 Via Duca Di Catabna, 1/1 NIGERIA Ashgate Publishing Asia Pac4tic PIe. Ltd. 306 Silom Road

Victoria HONG KONG, MACAO Casella Postale 552 Universsy Press Limited 41 Kallang Pudding Road 604-03 Bangkok 10500

Tel: (61) 3 9210 7777 CZECH REPUBLIC Asia 2000 LId. 50125 Firenze Three Crowns Building Jericho Golden Wheel Building Tel: (66 2) 235-5400

Fail (61) 3 9210 77s 8 National Inlormalion Center Sales & Circulation Department Tel: (55) 645-415 Private Mail Bag 5095 Singapoore 349316 Fax: (66 2) 237-8321

URL: htp:Avwww .dadirect.com.au prodeina, Konviktska 5 Seabird House, unit 1101-02 Fax: (55) 641-257 Ibadan Tel:(65)741-5166

GCS- 11357 Prague 1 22-28 Wyndham Street, Central E-mail: [email protected] Tel: (234 22) 41-1356 Fax: (65) 742-9356 TRINIDAD & TOBAGO

Tel: (422) 2422-9433 Hong Kong URL: htpiJ/www,tbcc.itlicosa Fax: (23422)41-2056 E-mail: ashgalenasianconnect.com AND THE CARRIBBEAN

Geral aad Ca. Fax: (422)2422-1484 Tel: (852) 2530-1409Weirbargganse 26 URL: htp:/twww .nis.czt Fax: (852) 2526-1107 JAMAICA NORWAY SLOVENIA St. Augustine Shopping Center

A-lollNbuWigsen28E-mail: [email protected] Ian Rancde Publishers Ltd. NIC Into A/S Gospodarski Vestnik Publishing Group Easlem Main Road, SI. Augustine

A-1011 Wien DENMARK URL: httpl/www .asia2000.com.hk 206 Old Hope Road, Kingston 6 Book Depadmenl, Postboks 6512 Enerstad Dunajska cesla 5 Trinidad & Tobago, West Indies

Tel: (431)512-47-31- SamfundsLieratur Tel:876-927-2085 N-606 Oslo 1000 4ubtiana Tel: (868)645-8466

Fax: (431) 512-47-31se9oerasAlld I11 HUNGARY Fax: 876-977-0243 Tel: (47 22( 97-4500 Tel: (308661) 133 83 47; 132 12 30 Fax: (868) 645-8467

URL: h(43t)www 4ger73d1c28utine RoseDK-1970 Frederksberg C Euro Into Service E-mail: [email protected] Fax: (47 22) 974545 Fax: (36 61) 338030 E-mail: tobeftdnclad.net

Tel: (45 31) 351942 Margtszgeti Europa Haz E-mail: [email protected]

MicroANdLAstries DESH lnpmenl Fax: (45 31) 357822 H-1 138 Budapest JAPAN PAKISTA U

Micro Indusies Deve(Dpmen) URL: hftpJ/www.sl.cbs.dk Tel: (36 1)111 6061 Eastem Book Service Mirza Book Agency SOUTH AFRICA, BOTSWANA Gusntro Ltd.

Assistance Socdety (6 ) Fax: (36 1) 302 5035 3-13 Hongo 3-chome, Bunkyo-ku 65, Shahrah-e-Quaid-e-Azamn For single itdles: PO Box 9997, Madhvani Building

Dhaansendi RoAre ECUADOR E-mail: [email protected] Tokyo 113 Lahore 54000 Oxford University Press Southem Africa Plt16/4JinjaRd.

Dhakaond 1209 ea Libn Mundi Tel: (81 3) 3818-8861 Tel: (92 42) 735 3601 Vasco Boulevard, Goodwood Kampala

Dhakat 209 LibrerIalnlemacional INDIA Fax: (813)3818-0884 Fax: (92 42) 576 3714 PO. Box 12119, Ni City 7463 Tel: (256 41)251 467

Faa: (8002)811188 PO. Box 17-01-3029 Allied Publishers Ltd. E-mail: orders@ svt-ebs.co.jp Cape Town Fax: (25841)251 468

Juan Leon Mera 851 751 Mount Road URL: hpiAvwww .bekkoame.orjp/-svt-ebs Oxionrd University Press Tel: (27 21) 595 4400 E-mail: [email protected]

Quito Madras - 600 002 5 Bangalore Town Fax: (27 21) 595 4430

BELGIUM Tel: (593 2) 521-606; (593 2) 544-185 Tel: (91 44) 852-3930 KENYA Shamae Faisal E-mail: oxtordrpoup.co.za UNITED KINGDOM

Jean De Lannoy Fax: (5932)504-209 Fax: (91 44) 852-00 Alrica Book Service (E.A.) LId. PO Box 13033 Microinto Ltd.

Av. du Roi2 20 E-mail: librns1 %@fibrminndi.com.ec Quaran House, Mtangano Street Karachi-75350 For subscription orders: PO. Box 3, Aton, Hampshire GU34 2PG

Tel: (322)538-5169 E-mail: [email protected] INDONESIA PO. Box 45245 Tel: (9221)446307 e.O Ebx41Sri England

Tel: (322) 538-5169 ~~~~~~~~~~~~~~~~Pt. Indira Limraed Nairobi Faa: (92 21) 4547640 P.O. Box 41095 Tel: (44 1420) 86848

Fax: (322)538-8841 EGYPT, ARAB REPUBLIC OF Jalan Borobudur 20 Tel: (2542)223641 E-mail: [email protected] Craighall Fax: (44 1420) 89889

BRAZIL Al Ahram Disobution Agency PO. Box 181 Fax: (254 2) 330 272 Johannesburg 2024 E-mail: [email protected]

Publicacoes Tecnicas Intemacionais Ltda Al Galaa Street Jakana 10320 Pak Book Corporatin Tel: (27 11) 880-1443 URL: hop-/www mirointocouk

Rua Peivoto Gamide, 209 a Calm 34~~~~~~~~~Te: (62 21) 390-4290 KOREA, RIEPUBLIC OF Aziz Chambera 21, Queen's Road Faa: (27 11) 880-0248

01409 Saex Paolo, S Tel: (202) 578-6083 Fax: (6221)390-4289 Daejon Trading Co. Ltd. Lahore E-mail:[email protected] VENEZUELA

Tel: (55 11) 259-6644 Fax: (20 2) 578-6833 PO. Box 34, Youida, 706 Seoun Bldg Tel: (92 42) 636 3222; 636 0885 Tecni-Ciencia Libnsa, S.A.

Fax: (55 11) 25894844 IRAN 44-6 YouidD-Dong, Yeongchengpo-Kx Fax: (92 42) 636 232B SPAIN CenNil Culdad Comercial Tamanco

Faa: (55p11)258-6990 The Middle East Observer Ketab Sara Co. Publishers Seoul E-mail: pbc2brain.net.pk Mundi-Prensa Libros, S.A. Nvel C2, Caracas

E-malL: hposmaslerwwiwuol.br 41, Shers Streel Khaled Eslamboli Ave., 6th Street Tel: (82 2) 785-1631/4 Castellb 37 Tel: (58 2) 959 5547; 503560016

URL: hopilwwee ,uol.br Cairo Delalraoz Alley No. 8 Fax:(822)784-0315 PERU 28001 Madrid Fax:(582)9595626

CANADA Tel: (20 2) 393-9732 PO. Box 15745-733 EdRorial Desarrollo SA Tel: (34 1) 431-3399

Renout Publishing gCo. Ltd Fax: (202)393-9732 Tehran 15117 MALAYSIA Apanado 3824, Lima 1 Fax: (34 1) 575-3998 ZAMBIATel: (98 21) 8717819; 0716104 University o Malaya Cooperative Tel: (51 14) 205380 E-mail: libreriarmundiprensa.es University Bookshop, University o Zambia

536 Canotek Road FINLAND Fax: (98 21) 8712479 Bookshop, Limited Fax: (51 14) 286628 URL: htp://www.mundiprensa.es/ GrealEastRoadCampus

Tel: (613) 745-2605 Akateeminen Kirjakauppa E-mail: ketab-saranecda.net.ir PO. Box 1127 PO. Box 32379

PO. Box 128 Jalan Pantal Baru PHILIPPINES Mundi-Prensa Barcelona Lusaka

Fax: (613) 745-7660 FIN-00101 Helsinki Kowkab Publishers 59700 Kuala Lumpur Intemational Booksource Center Inc. Consell de Cent, 391 Tel: (260 1)252576

E-mail: xrder [email protected] Fa: (200 1) 253 952

URL:hoprvv.dpwwrenousbooks.com Tel: (3580)121 4418 PO. Box 19575-511 Tel: (603)756-5000 1127-A Antipolo St, Barangay, Venezuela 08009 Barcelona Fax

Fax: (3580)1214435 Tehran Fax: (603) 7554424 Makati City Tel: (34 3) 488-3492

CHINA E-mail: [email protected] Tel: (98 21) 258-3723 E-mail: umkoopCtm.net.my Tel: (63 2) 896 6581;6505; 6507 Fax: (34 3) 487-7659

China Financial & Economic URL: hap/www.akateeminen.coml Fax: (98 21) 258-3723 MEXICO Fax: (632)896 1741 E-mail: [email protected]

Publishing House FRANCE IRELAND INFOTEC POLAND SRI LANKA, THE MALDIVES

8, Da Fo Si Dong Jie World Bank Publicalions Govemment Supplies Agency Av. San Femando No. 37 Intemational Publishing Servrce Lake House Bookshop

Bel:0ng 3257 66, avenue d'lina ORig an tSolAthair Col. Torello Guerra Ul. Piekna 31137 100, Sir Chinampalam Gardiner Mawatha

Tel: (8610)06333-825 75116 Pads 4-5 Harcourt Road 14050 Mexioo, D.F 00-677 Warzawa Colombo 2

Fax: (86 10) 6401-7365 Tel: (331) 40-69-30-56/57 Dublin 2 Tel: (52 5) 624-2800 Tel: (48 2) 628-6089 Tel: (94 1) 32105

Fax: (331)40-09-30-68 Tel: (353 1) 661-3111 Fax:(525)624-2822 Fax: (48 2)621-7255 Fax: (94 1)432104

Fax: (353 1)475-2670 E-mail: [email protected] E-mail: books%[email protected] E-mail: [email protected] sous

URL: hBpJ/rtn.nel.mx URL: http//www .ipscg.waw.pVips/exporV

THE WORLD BANK

1818 1I Strcet,. N.\\

Washinoton 1).(C. 20433 USA

Iclephloluc: 202-477-1234

FIesimillc: 2(02-477-6391

Tele:: MI1 64145 WORIJWANK

MCI 248423 WR,)VN

\\orld W\ide \\Wh: http:N// w.w-orldbank.r g/

EI-m11aLI: hooks(ii N% orldhank.or,-

ENVIRONMENTALLY AND SOCIALLYSUSTAINABLE DEVELOPMENTRii ae I) ., c Ilp t

9 780821 341735

ISBN 0-8213-4173-1

sustainable development t.q8 114 3 work in … · the first draft, prepared in 1995 by derek...

Documents