Updating report on the World Bank research program (Vol. 2 of 6) : Report of the research advisory panel on agriculture and rural development

20310 Vol. 2 REPORT OF THE RESEARCH ADVISORY PANEL ON AGRICULTURE AND RURAL DEVELOPMENT May 1979 FILE COPY TO THE WORLD BANK FINAL REPORT OF THE EXTERNAL ADVISORY PANEL ON AGRICULTURE AND RURAL DEVELOPMENT Dr. Gelia T. Castillo Professor Jean Marc Boussard professor of Rural Sociology Charge de Recherches University of Philippines Institut National de la Recherche Los Banos, Philippines Agronomique Paris, France Dr. James Boles Dr. Lucio G. Reca University of California Buenos Aires Berkeley, California Argentina Dr. Vijay Shankar Vyas, Chairman Director Indian Institute of Management Vastrapur, Ahmedabad, India May 1979 CONTENT Page No. I. Introduction 1 II. Overview of Research in Agriculture and Rural Development 2 III. Research Needs and Strategy 8 IV. Comparative Advantages and Disadvantages 10 V. Collaboration with Outside Research Organizations 12 VI. Research Organization and Management 14 VII. Dissemination of Research Results 18 VIII. Summing Up 21 Appendix I - Research Agenda 23 I. Introduction 1.1 The Bank's lending in Agriculture and Rural Development (ARD) is a substantial activity - not only in its own right, but also for the overall impact that it can make on the development strategy of the borrowing countries through its direct and induced effects. Looking to the magnitude and complexity of the task, lending decisions in these areas need to have proper support from well designed research. As it was succinctly stated in a review document, "The Bank needs knowledge to formulate intelligent policy and research can supply that knowledge." The Bank has assembled a well qualified group of researchers that, in their respective Departments and Divisions, have acquired an excellent reputation in professional and, to an extent, operational circles. Currently, the activities of these various research groups within the Bank are being assessed. 1.2 The Research Advisory Panel on Agricultural and Rural Development (RAPARD), one of several similar Panels each reporting on a specialized area of research, has been charged with the task of evaluating the research program of the World Bank in its specialized field. While each of the Panel Reports will have intrinsic interest and value, they will also provide a partial basis for the work of the General Research Advisory Panel. 1.3 Without presuming to know completely the rationale that led to the formation of these Panels at this time, this Panel (RAPARD) interpreted its task broadly. Is the research in this field relevant to the operational needs of the Bank as well as the economic policy requirements of member countries? Is it of high quality? Does it cover all or most of the imortant researchable questions? Is the research mix appropriate? Are the comparative advantages of the Bank fully utilized? Is the scope for collaboration with other research institutions properly exploited? Are the research results adequately disseminated and used? To what extent should the research program be charged with the additional task of assisting the development of research institutions in developing countries? 1.4 These, at least, are some of the questions we have considered. We recognize, however, that the ability of this Panel to respond in a meaningful way to all of these questions is limited. One obvious limitation is that we have, for the most part, reviewed only "externally" funded research projects; projects funded at least in part from research funds administered by the Research Committee, though we are aware that there are other research acti- vities not so funded. We are also limited by our consideration of only those research projects identified as Agricultural and Rural Development, even though other projects exist which address issues also relevant to Agriculture and Rural Development. Additionally, we have had direct access to only one class of potential research users, namely members of the various Regional divisions of the Bank which administer the actual lending operations. Finally, although Panel members bring to its deliberations a variety of administrative and research experiences, the Panel has no direct information on the usefulness of the Bank's research product to member countries, and especially to borrowing countries. Despite these limitations, this Panel has taken its charge seriously and will exercise its best judgement in responding to the set of questions outlined above. -2- II. Overview of Research in Agricultural and Rural Development 2.1 To obtain some indication of the magnitude of our evaluatior problem, during the past six years there have been some 26 "external" r projects classified as ARD. Annual research expenditures on this set c projects ranged from US$225,000 to US$504,000, approximately 15% of "e,, research expenditures. During this period annual "external" research e diture has been approximately constant in real dollars even though Bank lending for Agriculture and Rural Development projects has approximatel doubled, from US$957 million to US$3.2 billion. This represents an inc in lending for Agriculture and Rural Development from 12% to about one- of the total Bank lending. 2.2 Research efforts in the Bank in the area of Agricultural and Rural Development (ARD) have, for the purpose of this review, been broa classified as research on: (1) Project Design and Management; (2) Secto: Policy Issues; and (3) Developmental Policy Issues. With the recent em- on Rural Poverty, the projects in this area can be treated as a special separate category. 1. Studies on Project Design and Management 2.3 These studies aid in improving the current modes of designinc operational lending projects. Two distinct approaches are discernable: (i) Quantitative analysis and modelling - The objective of these studies is to develop a formal framework for del ing interrelationships between various elements of a p and evaluating the direct and induced effects of opera tional lending projects; the latter, generally, are di cussed in terms of income generation and income distri. bution under different institutional settings. A good example of such research is the Indus Basin Irrigation Project which is basically an attempt to formalize the design and management of a large scale irrigation proj, Analysis of rural change falls also into this category Several "internal" projects are operated by the DRC in same vein. Using extensively quantitative and econome: procedures, these studies are generally of high profess quality despite their obvious reliance on comparative statics at the expense of the dynamic dimension. At tf time, they are not always well received in the operatic departments of the Bank, precisely because of their de of sophistication which may limit routine use. Nevertt it is quite natural that the methods for tomorrow do nc with the commonly used routines of today. The shadow pricing system itself was not better received a decade than, for instance, the Social Accounting Matrix of toc One problem with this type of investigation is avoiding large a commitment in unfruitful directions, a difficul far avoided. -3- ii) Non-quantitative, comparative studies of various rural development projects; The prototype of this kind of research project is the African Rural Development Study. Several other projects such as the Study of Lilongwe Land Development Program, the Agricultural Innovation and Rural Development Study, the Study on Management and Operation of Irrigation Projects also belong to the same line of investigations. Such studies can be described as in-depth, ex-post, comparative analyses of agricultural and rural development projects. Some of these projects particularly the African Rural Development Study, resulted in outstanding pieces of analysis, and are generally appreciated by the personnel in the operational departments. Some of these projects approach the subject in a non- conventional way. For example, the study on Agricultural Innovation and Rural Development rejects the conventional micro-economic framework for analyzing technological changes as being too restrictive and argues for a more complete systems framework. 2. Sectoral Policy Studies 2.4 Research in this category aims at an analysis of important problems at the national level. The projects in this area include two kinds of studies: (i) Several studies of key policy variables belong to this category. The main research objective of these projects was to increase knowledge of the institutional and economic variables affecting development process. These include the studies of Smallholder Agriculture in Yugoslavia, Land Reform in Latin America and the country case studies in Agricultural Process and Subsidies. The report on Land Reforms in Latin America is one of the more balanced pieces have been of value to users within the Bank, is a systematic effort to identify particular types of agricultural projects suited to the conditions emerging in each country after land reform. The study on Agricultural Prices and Sub- sidies shows, basically, the same strengths and weaknesses as the Land Reform Project. However, in the latter case the time overrun was less important. The study on Yugo- slavia was impeded by the difficulty of finding competent researchers sufficiently interested in the project and by difficulties in obtaining data. (ii) National Agricultural Models: The objective of this type of research is to construct and validate sectoral models. An outstanding example of such a study is the CHAC model which was followed by similar models for Brazil and Portugal. The project on Agricultural Pricing and Storage Policy in East Africa was also attempted along the same lines. Emphasis of the researchers is on methodological issues; -4- i.e., how to build easy-to-construct, adaptive, and v. models of the agricultural sector of a given economy. studies gave rise to a great number of publications, a contributed to the reputation of the Bank within the tific community. At the same time, they could be usec operational purposes. For these reasons, they are ger considered to be successful within the Bank. Neverthe (and without diminishing their merits), it must be poi out that they are examples of comparative statics. T. absence of a time dimension not only deprives the user important information, but also casts doubt on the val of their results. In view of the importance of change over time adding a dynamic dimension would have signi- ficantly modified the models and perhaps the conclusio drawn from them. Similarly, a major shortcoming of many of these models the fact that the links between the agricultural secto the rest of the economy have been largely neglected, e through prices and quantity requirements. More generally, these models are extremely difficult t. test, therefore, one is not sure whether their predict- will be realized. For this reason extreme care needs be exercised in interpreting their results. 3. Development Policy Issues 2.5 These studies aim at deepening our understanding of the develc mental process by selecting some key determinants. Many of these studies located in more than one country. Examples of such studies are provided projects such as Land Reform in Latin America; Farm Mechanization in Inc Pakistan; Rural and Urban Public Works; Raising Productivity on Small I Price Intervention in the Agricultural Sector; Research and Productivit,, Changes. 2.6 Some of these projects, i.e., Farm Mechanization in India and Pakistan, and Rural and Urban Public Works, producted good results from g operational angle. On the other hand, few others, e.g., Raising Product' on Small Farms could not serve its original purpose (although it producec a computer program which was not without value at the time it was writter 2.7 Finally, two other projects deserve a special mention. Price Interventions in Agriculture is concerned with the new distributive impac of subsidies and taxes on products and factors in the agricultural sectoi It deals with an unquestionably important topic and was carried out by a consultant, within the time and resources originally committed to the prc However, the project apparently suffered from lack of supervision, the consequence of which is a lengthy final version, inappropriate for Bank use at the operational level. This is particularly painful because, as indicated above, the subject matter was adequately covered from a concept point of view. -5- 2.8 The project on Agricultural Research and Productivity Changes deals with a basic question pertaining to agricultural development, and is a good example of a "successful project". Several reasons help to explain the good results: a clear specification of the topic and the choice of tvo highly qualified scientists to develop the subject. The study also provides some insights for research organization with the Bank: it is a good example of collaborative research as well as an indication that scale and quality of research are not necessarily correlated. 4. Poverty Issues 2.9 Although this topic may be considered as a special case of the preceeding category, it needs a separate mention owing to its importance in the Bank's lending program and the large number of research projects in this area. The latter include projects such as Agricultural Development, Rural Income and Poverty, Food Deficits in Target Groups, Distribution of Income Through the Extended Family System, Structure of Rural Employment, Income and Labor Market and Rural Savings and Investment Behavior. These projects are mainly concerned with the functioning of economic institutions in low income rural societies. 2.10 Similarly, the Bank has been increasingly involved in the area of Food and Nutrition research in the recent past, which, given the magnitude and importance of the subject is a proper field of research activity in the Bank. Several aspects of the work on these issues are worth mentioning: (i) carefully designed studies which make excellent use of far-from-ideal data coupled with rigorous economic analysis; (ii) the methodological contributions to the analysis of probems of food and nutrition, as for example, in the case of the study on Food Deficits in Target Groups; (iii) the positive interaction among different research units in the Bank (for example DED and CPS); (iv) the production of research results which give insights in the area of direct relevance to the operational side of the Bank. 5. Summation 2.11 The overall impression which emerges from a review of research in the field of ARD in the Bank is that of a large output of high quality. These research projects have yielded a large number of interesting and excellent research publications. At least 20 research papers based on these projects were published in top level professional journals or were brought out as full-length monographs. Due recognition should also be given to the contribution of the external consultants in achieving these satisfactory results. It can be safely concluded that on purely professional grounds, the overall performance of researchers in the Bank is very satisfactory. 2.12 This is not to say that the research efforts are not capable of improvement. The collection of research projects in Agricultural and Rural -6- Development does not appear to constitute a comprehensive research prog There is evidence that in some cases inadequate attention has been give: other relevant research, not only from other research organizations but from the Bank itself. One further gap is the relative absence of studi. that delve into the rich lending experience of the Bank to derive a bet understanding of the development process, although the problems of hand large amounts of disparate data relating to such experience is clearly a difficult and time consuming task. 2.13 Mention has been made of the reliance on comparative statics the consequential absence of dynamic elements in the models. The Bank 2 is aware, not only of this omission, but also of the difficulty and comr of adding the time dimension in appropriate ways. 2.14 Reference has been made to problems encountered in research administration, particularly in respect to the use of outside consultant research organizations, and their adequate supervision and coordination. Given the limited resources available, these problems suggest an explana if not a justification, for the virtual absence of efforts to build rese capacity in the developing countries. 2.15 Nevertheless, the Bank's research has quite likely contribute to the shift in lending policy toward the small farmer and the rural poc it has entered the debates on development issues besides providing some macro-policy indications and evidences of direct and indirect project outcomes. Small advances on the methodological front have likewise been achieved and a few more observations about the development process added Moreover, there is promise of operational guidance from certain studies. In general, it is not that the Bank's research is poor or that it is not useful, but that it is not enough to address the multitude and magnitude of problems encountered in the lending program, nor the knowledge needed to make rural development feasible in diverse situations. 2.16 Moreover, despite the claims and counter-claims within the Ba: it cannot be truthfully said that the Bank's research has no operational significance, for research can exert influence in inconspicuous ways. It giving advice, assistance and comment and when participating in country economic missions, sector work, and project related activities, DPS and most probably rely on their insights from research to carry them through, The influence is, therefore, likely to be more circuitous than straight- forward and perhaps never in its particular research purity but in combit with previous knowledge and partly with speculative judgment. 2.17 In this context, however, it is notable that (as suggested in para 2.1) the resources devoted to "external" research in ARD is relative small. While the Research Committee nominally allocates some 20% of its annual budget to ARD research, in practice only-11 to 13% is used. The total staff time allocated to ARD research may be as low as 6 manyears pE year. One reason why the Research Committee funds are not used may be related to the limited in-house capability to plan and supervise their us More importantly, when spread over 6 regions, 25 operational projects diu -7- ,d some 120 developing member countries, the 26 ARD studies completed over the past 6 years looks very thin. It is perhaps not surprising that operational staff see few research results of direct relevance to their o work or to the countries with which they are concerned. 2.18 Clearly, producing good, publishable research is one thing. Coping with the many and specific needs of the Bank, fully exploiting its ccaparative advantages, and properly disseminating the results of research is another. The Bank's research efforts have to be evaluated in terms of their fulfilling these tasks. We will now address these questions. -8- III. RESEARCH NEEDS AND STRATEGY 3.1 The research program in the Bank should be directly basica to fulfill three tasks: (1) having a good general understanding of t- developmental process; (2) having reliable tools for designing and a- development projects; and (3) being able to take a position in natio: policy issues whenever the Bank is involved. A significant amount of in the Bank is conducted with these purposes in view, although is is r always clear as to which one of the objectives listed above is sought fulfilled, and sometimes the studies made may have "fallen between stc 3.2 The Bank will derive obvious advantages if the research is sciously focused on major issues and subject areas. At present the i7 given is that of highly diffused research efforts without a clear focL In an institution whose main task is to promote research, such multipl is permissible, probably desirable. But in an institution like the Ba research efforts should have an operationally relevant focus and be sc more structured. Such structuring does not mean a lack of innovative in different areas; it only pleads for periodic efforts to develop maj themes that will serve to guide the research program. 3.3 A more purposeful research program has the following import elements. First, a conscious effort is necessary to identify the next of problems that developing countries are likely to face. At present international organizations as well as national governments merely rea the problems. The Bank is not an exception to this rule. A group in Bank with the assistance of a few knowledgeable outsiders may be able identify the set of problems that are likely to acquire serious dimens in, say, the next decade. This will enable the Bank as well as the me countries to be better prepared to cope with these problems. Such a g may be constituted every 3 to 5 years, to assist in identifying the en tasks. This group will not indulge in merely a speculative exercise. will review the current situation, identify emerging trends and assess governments' capabilities in coping with the problems in key areas suc population, food, non-renewable resources, technologies, and instituti factors. It should identify the nature and variety of tasks, as well existing and emerging bottlenecks. Fortunately, a lot of work on emer tasks and bottlenecks is being done by specialized institutions. The researchers should bring the major findings of these studies together relate these to its principal activities. With the back drop of such the Bank will be able to develop major strategies for research and len 3.4 Second, in the context of these emerging problems in the de world, the main thrust of Bank research should be to identify and spel alternative strategies and policy options, keeping in view the socio-e milieu in different countries. Obviously, such options will change ov period of time. However, without bringing in policy options in differ sets of circumstances, developmental efforts are likely to remain part and ad hoc. -9- The third group of research activities, which should be the king- of the research program in the Bank, pertains to those providing "guide- s for lending strategies". The projects in this category should aim at: an understanding of intersectoral and intra-sector linkages, and structural , systematic problems within the sector; b) improving behavioral aspects of research design, which at present are treated inadequately; and c) a focus on ,.ganizational and administrative issues related to project implementation. It is not unusual to find that the social and physical infra- structure created at high cost do not serve the rural poor. This is due, near alia, to various administrative and institutional "leakages", an 'ect on which very little systematic work is done. In brief, the main purpose of this type of research is to find out how a given amount of loan cr credit can be made more effective in serving the stated objectives. .3 an institution like the Bank such research is of immediate importance .nd can promise high payoff. 3. The fourth category of research pertains to project level research i.e., procedures and processes for identification, formulation and evaluation or lending projects. Various stages in the project cycle are carefully i"entified in the Bank, but the research support for undertaking tasks associated with different stages of a project is not yet firm. Systemati- zation of knowledge in this field will also be of help to the member govern- =ents in their own economic operations. 3.3 Finally, a large developmental institution like the Bank should :n:ribute to the understanding of the process of social and economic growth. :evelopment is a complex phenomenon, it does not proceed along neat lines or in regular stages. The interplay of various social and economic institutions, th-e nature and limitations of public intervention, and the response of institutions to intervention need careful study. Equally important in this regard is understanding individual motivation and incentives. Without an understanding of the process of change, development will continue to be viewed in a mechanistic way and effectiveness in influencing its course ill be marginal. For example, though there is a recognition that economic development influences population growth, there is little knowledge as to which element of development and in what manner it influences, say, fertility behavior. 3.9 To recapitulate, the Bank's need is to have well researched ideas and views on: (i) the nature and magnitude of emerging tasks in agriculture and rural development; (ii) various options available in key areas of ARD; (iii) possibilities of making lending more effective; (iv) methodologies for identifying and formulating bankable projects; and (v) understanding the process of change in order to draw useful inferences for policy options. 3.10 Not all of this research needs to be in-house. There are signi- licant opportunities for collaborating with other research bodies. Before we give our views on the nature of such collaboration, it will be advisable to examine comparative advantages, and limitations, that the Bank has in conducting the type of research which can maximize the Bank's contributions to ARD. -10- IV. COMPARATIVE ADVANTAGES AND DISADVANTAGES 4.1 It is at least a debatable point that the World Bank has absolute advantage in the field of economic research in Agricultura Rural Development when compared to individual universities and spec research institutions. Even if this point were to be conceded, the still the matter of comparative advantage. To the fullest extent p the Bank should utilize the data and research products generated by the public and private institutions working in the field of economi development. Furthermore, in developing and managing a research pr of its own, the Bank should consider carefully its comparative adva- and disadvantages in deciding how to proceed. 4.2 Some of the Bank's comparative advantages are: (i) There exists in the Bank a substantial number of cor researchers, not only in the field of ARD but also the general field of economic development. The var: formal and informal communication networks existing within the Bank serve to identify problems, appraisc proposed solutions, and disseminate research resultF By definition, development is an integrated process. From this perspective, the position of a researcher in ARD in an institution which copes with the whole developmental process is indeed an advantageous one. (ii) Over a period of time the Bank has developed relatic with a large number of countries. It can easily set up an information network or mobilize expertise from diverse sources. Numerous contacts between the Bank member governments and local institutions facilitate access to data and specialized knowledge. The resea in the Bank can cope with diverse conditions by adop a comparative approach which may expand the number 0: cases for analysis and insights. This is possible d, to the operational involvement of the Bank in a numb countries in the same region. (iii) The Bank has a veritable gold mine of development ex7 waiting to be analyzed, refined and shared. Where e- one find 363 lending projects approved during the fi: period 1974-78 with 210 of them defined as rural devt ment projects in which more than half of the direct 1 are expected to accrue to the rural poor. (iv) Finally, relative to most other institutions, the Bar more capable of allocating adequate funds to importar and well conceived research projects. This permits t further possibility of hiring external consultants ir to benefit from the association of outstanding specia in a given field. It also permits effective collabor with local institutions having access to reliable dat a stronger feel for the local scene. -11- There are, at the same time, several comparative disadvantages: (i) The organizational culture of the Bank is understandably dominated by the lending project syndrome, since this is the major function of the Bank. Those who are not involved in lending activities may feel a degree of alienation. The demand for usable and timely results may place undue pressure on the researchers. (ii) By the very nature of the Bank's relationships to the governments and institutions of the borrowing countries and the role which its projects play in these countries, the Bank is vulnerable to their political and socio- cultural sensitivities. Research results carry their own hazards. (iii) Finally, and perhaps of greatest importance, the Bank is based in Washington while agricultural and rural development is taking place in the hinterlands of the borrowing countries. Although there is virtue in maintaining objectivity and distance from the phenomena being studied, there is a risk that Washington may be too distant to allow the intuitive and practical insights which determine the operational relevance of research findings. These comparative advantages and disadvantages of the Bank, as as the ultimate limitation of staff numbers should be clearly in mind A deciding whether a research task should be completed in-house or Ma"brtd with outside consultants or developed in collaboration with other organizations. -12- V. COLLABORATION WITH OUTSIDE RESEARCH ORGANIZATIONS 5.1 In a sense, collaboration of the Bank with outside research organizations is both implicit and explicit. It is implicit in that it should be aware of and use, where appropriate, research results obtained and published by other organizations in the field of agricultural and rur development. It is explicit when, in an effort to develop the most cost- effective and efficient research program, comparative advantages and disac ages of the Bank relative to outside research organizations result in dire collaboration. 5.2 There is, however, another set of reasoning that would justify collaboration even at the potential risk of loss of timeliness, research quality and cost effectiveness. In its own long-term interest, the Bank should assist in the development of an institutional capacity for research particularly in the borrowing countries. Presumably, the Bank is or shoul responding primarily to lending project proposals originating in the devel countries. To the extent that the quality of these proposals can be impro over time, the processes of review, approval, disbursement, monitoring and evaluation by the Bank staff would be greatly simplified. In the most rec Annual Report concern is also expressed that disbursement in certain major sectors, particularly agriculture, will be somewhat lower than expected. Reference is made to the fact that "new style" projects are technically complex and that new agencies and institutions are needed to discover ways benefit low-income groups of people previously not reached by most governme programs. To ensure that institution building to improve indigenous capab4 gets the attention it deserves, it should be included as a major element in the set of criteria used in evaluating research (as well as other) projects 5.3 While each specific project will require particular attention to the issue of collaboration, it is possible to identify a number of gener subjects that are likely to be largely in-house and others that are likely prove more effective in a collaborative setting. 5.4 The Bank is well equipped to prepare sectoral and economy- wide models to test alternative policy options. The work is complex, time consuming and requires large resources. At the same time it is necessary t a special effort be made to simplify the techniques for easier replication such exercises, and to assist in dissemination and training both of Bank st and also of outside research organizations. 5.5 The projects on lending strategies have again to be essentially in-house research, although on some of the aspects, particularly on the behavioral and organizational dimensions, collaboration with others may be desirable. It is also possible to farm out some research in this catego to institutions in developed countries but particularly also to those in developing countries where a suitable capability to conduct such research exists or might easily be developed. 5.6 Another category of research, aimed at lending project identifi- cation, formulation, monitoring and evaluation, should be basically an in-hi -13- activity. Here one should particularly emphasize the effort to review and analyze the rich resource of project experience in order better to understand the process of agricultural and rural development. 5.7 A core program of research in all these areas should be located in the Bank yet in each of these areas, barring the research on lending strategy and on project related methodology, the Bank should encourage involve- nent of outside expert groups to a significant extent. There are several aspects of research under each of these categories where collaboration with other knowledgeable individuals or specialized institutions would be useful. 5.8 There are at least four situations where a project might be executed outside the Bank. First, if highly specialized technical knowledge is needed, the Bank might not prove to be a suitable location. For example, in the area of ARD, there are several problems which can be better researched in the International Agricultural Research Centers than in the Bank. Since the Bank has taken a leading role in the establishment and development of these centers, and since most of them have excelled in their chosen fields, close collaboration between these institutions and the Bank is both feasible and desirable. Second, the projects necessitating a large amount of primary data collection can be profitably located in the institutions of the country concerned. Because of its geographical location, its high cost operations and the given style of functioning, the Bank is not the right agency for large scale data collection in the developing countries. With proper planning and monitoring, these tasks can easily be entrusted to the local institutions. Third, when the research needs a thorough and intimate knowledge of the local situation in a country, i.e., in case studies, the Bank has hardly any comparative advantage. The institutions located closer to the field can do this task more efficiently. Finally, the Bank may also support local institutions, particularly those with competence in behavioral, organizational, and sociological areas, even when only a part of the research effort is directly operational. 5.9 It is clear from the above that a variety of arrangements will have to be made in order to make satisfactory use of the collaborative arrangements. Carelessness in this regard can lead to failures and frustrations, as has already been pointed out. On the other hand, with the support of the Bank, and more importantly, the involvement of its staff in the research projects located in different centers, the Bank can obtain high quality results and in the process strengthen local research capabilities. There cannot be a uniform type of collaborative arrangement with outside research groups. Much depends on the strength of the collaborative organization and the nature of the project. The arrangements can range from contracting with local instituitons to collect primary data to making research grants in areas which might be of mutual interest and benefit. In between come arrangements such as visits of external collaborators as consultants, placement of the Bank's staff in local institutions and joint project teams named by the Bank staff and the collaborative institution. -14- VI. RESEARCH ORGANIZATION 6.1 Most of the "external" research in Agricultural and Rural Dev ment is performed in the Development Research Center, the Development Ec Department, and the Agricultural and Rural Development Department. The f two units are in the Development Policy Staff (DPS) and are very much re oriented, while the last unit is one of the departments of the Central P Staff (CPS) and has other important functions in addition to research. ! in the DPS has been typically longer term and methodological and discipl- oriented, while the research performed in the CPS has been typically int, mediate term and more immediately "problem solving", with perhaps more er on operational issues. External research is funded and projects reviewec approved by a Research Committee that has jurisdiction over both CPS and research projects seeking external funding. 6.2 The existing Research Committee is in a sense an apex organi- zation, but by its composition, nature of functioning, and largely, due t the mandate given to it, the Research Committee concentrates mainly on th screening of projects which are proposed for funding by different units i Bank. Its role as a priority setting, task allocating, monitoring and in acting body is conspicuous by its absence. What is needed, therefore, is a new committee or group but a redefinition of the functions of the prese Research Committee along the following lines: (i) Reviewing periodically the research needs of the Bank, in collaboration with outside experts and with research and operational staff. (ii) Identifying and defining major themes to be addressed in the research program. (iii) Exercising initiative in the generation, development and implementation of research projects on the defined themes (iv) Monitoring and periodically reporting on both ongoing and completed research. (v) Assisting in the task of building research capacity in th developing countries. (vi) Reviewing and approving research proposals submitted by t panels. In view of these redefined and expanded functions, the Researc Committee should be designated as the Research Council. The Council will be able to perform these functions only if it has a regular staff, workin exclusively for it, to assume routine administrative functions. -15- 6.3 The Council should also be assisted by standing Research Panels to screen and strengthen research proposals prior to final review and decision by the Council. These Panels, organized on the basis of subject matter interest and competence, would replace the current ad hoc Panels organized for each proposed project. They would also monitor on-going research projects in their respective areas and should also facilitate communication between researchers and users, both within and outside the Bank. In order to ensure that "new blood" is brought into the process, which is an important advantage of the ad hoc Panels, a small number of ad hoc members should be added to the Panel for the review of each research proposal. 6.4 While the reorganization of the Research Committee and the redefinition of its research function will accomplish a great deal, there is also the need to delineate more clearly the categories of research activities for which the various units of the Bank are likely to have comparative advantage. Any clarification of the territorial jurisdiction of different units is likely to arouse passion. Despite this danger, and in view of the five fold classification of research activities already developed, the following correspondence is suggested: Research Area Unit Primarily Responsible 1) Emerging Tasks Research Council and Research Themes 2) Alternative Strategies DRC and Policy Options in ARD; Methodological Issues in Project Design 3) Issues Related to Bank Policy AGR on ARD, Strategies, Behavioral and Organizational Aspects 4) Processes of Change and DRC/DED Intersectoral Relations 5) Impact Studies, Intercountry Regions Comparisons, Project Related Research 6.5 These suggested principal areas of responsibility do not have well defined boundaries and, in any case, should not be regarded as mutually exclusive. The rationale for this division is provided by the existing "division of labor" as well as the nature and functions of the various units. It needs to be reiterated that not all the research needs to be in-house. 6.6 The emerging tasks in ARD have to be identified and transformed into major themes for the research program (with due regard for the con- tributions and findings of other institutions). In our view this can be best performed by the Research Council with its several subject-matter Panels. -16- 6.7 An examination of policy alternatives in the area of ARD require, a comprehensive and systematic look at the whole economy. Various types of modelling will be an aid to such a discussion. With its available compete, and the traditions already established, DRC is the proper location for this type of work. The methodological issues in project design could be tackled by the DRC in close collaboration with operational departments. The existing practice of inviting outside experts as consultants is a welcome device and can be better utilized once the main thrust of research in the DRC is programmed. 6.8 Guidelines for lending strategies and project design involve issues which are of direct relevance to the Bank's operation, and therefore should be considered the king-pin of the research program. AGR, in close touch with operations and, at the same time, having key staff functions, has a comparative advantage for this research area. 6.9 The aspects of agricultural and rural development that can be better understood in an intersectoral context (urban-rural linkages) and which may also involve in-depth study of the development process may be better pursued by DED. Here also there are opportunities for collaborative research with other institutions. 6.10 Research on issues of immediate need and operational relevance should be addressed more actively by the operational departments concerned. They should be provided with resources to permit them to review, interpret and execute studies of project impact and program strategies, as well as those involving inter-country comparisons within their Regions. The purpose here should be to facilitate the additional work that would make the overall research program more relevant to operational concerns. 6.11 An organization provides only a formal structure. If it is to achieve its objectives, certain procedures should be followed. Procedures which are relevant for research objectives include: (i) development and phasing of research projects; (ii) research collaboration with outside agencies; (iii) monitoring and evaluation of on-going research. 6.12 The initiative for research projects can come from a variety of sources, both within and outside the Bank. Regardless of its source all research will be more meaningful if the sponsoring units are clear about the particular clientele, either inside or outside the Bank, who will benefit from the research product. 6.13 The research initiative of the staff in the operating departments is dampened because the scrutiny to which research proposals are subjected is excessive and dilatory. In these circumstances it is suggested that the operational departments should be encouraged to develop acceptable proposals jointly with, or with help from, DRC, DED, AGR or consultants. -17- 6.14 In the previous section, we have outlined the scope for collaborative research. There are, however, certain pre-conditions for the success of any such arrangement. These are: (i) there should be a mutual respect and trust, (ii) the institution should already have some competence in the area where the Bank's interests lie, (iii) the research design is clear and explici: and there is a provision for periodical consultation and monitoring and, (iv) both institutions are assured of some mutual gain from the collaborative activities. 6.15 There are, however, some situations where even when these conditions are fulfilled, collaboration may be neither possible nor desirable. If the research problem is considered "sensitive" by the host government, e.g., problems of land reform in some countries, the local institutions will be seriously handicapped in pursuing research. The credibility of the local institution vis-a-vis the member government is also an important factor to be considered, and this may vary depending on the nature of the problem. 6.16 Most research projects require a multi-stage phasing of development and implementation which makes it possible to assess intermediate output from each stage before proceeding to the next. For example, available information from past research, practice and theory may be reviewed and tentative hypotheses proposed. Either with available data, or following collection of new data, the hypothesis is tested. This in turn may lead a reformulation and retesting or to a successful conclusion. 6.17 As in the case of initiating new projects, the Research Council and its subject-matter Panels will have a major role in monitoring the successive phases of on-going research, at present one of the weakest links in the research process. -18- VII. DISSEMINATION OF RESEARCH RESULTS 7.1 Even if the research program generates results that are of hIgh quality, are relevant, cover the field, and are potentially usable, there still remains the problem of dissemination and utilization. 7.2 It is easy to say that good research will be its own salesman and that no special effort is needed. If this were true, nothing further would need to be said. As evidenced by publications in referred journals books, and working papers, there is a large quantity of high-quality researa from the Bank available to each of the several groups of potential users. These publications have been supplemented by seminars and workshops both in-house and in the relevant geographical regions. 7.3 This problem of dissemination and utilization of research results is not unique to the Bank. Particularly in Agriculture and Rural Developme:!: the problem has been approached by extension programs of various types. Even within the Bank, some additional effort is needed. For the potential user in the developing countries, an even greater effort is required. From the producer-user dialogues, two general views were expressed: (i) that operational needs of the Regions are not being addressed by the Bank's research; (ii) that important policy papers and research reports are not being read by operational people within the Bank, and by developing country officials outside. 7.4 This occurs despite the fact that the research product is considered as "interesting", even by those people who complain that they do not benefit from it. In fact, the research program is interesting because it addresses questions that they, themselves, have addressed. It is not of much help because they cannot usually read the research reports at the time they need it. Research reports are always either forgotten or forthcoming. Besides, the language of the researchers seems often tedious and too abstract. 7.5 Difficulties of this kind are not specific to the research in the Bank. They are encountered in any research service associated with an operational institution. They are painful, but probably unavoidable. Even so, conflicts arising between researchers and operational departments do not seem to occur in the Bank with the same frequency as in many other instit- utions. To a certain extent, some credit for this must be given to the formal and informal communication links inside the institution, and to the pro- fessionalism and relative open-mindness of both research and operational staff members. What can be done, nevertheless, to improve the situation? 1. Inside the Bank 7.6 The first thing to do is to avoid destroying what is already functioning. The dispersion of the research activities within several -19- rtents is certainly regretable from the point of view of research ,,Ageament. it is nevertheless a very efficient way to take account of - g~3emspectrum of needs and should be maintained. Similarly, the facilities c epportunities which already exist for a researcher to participate in ,perational mission, and for a member of the operational staff to stay a while in a research unit should be encouraged and expanded. Obviously, te great majority of the operational staff members will never be interested : research, and some researchers will never be attracted by operational But even if only a few professionals can be exchanged between the two stes, this will facilitate personal acquaintances and communication. The documentation within the Bank could probably be improved by -.:mputerized document retrieval system, which would allow, for instance, s :ebody having tractorization problems in the Ivory Coast to know about ,:her studies on similar questions done, for example, in Pakistan and in india, without painful bibliographical investigation. Such a system gives rtsearchers access to unpublished documents and data from the project files. To some extent this would compensate for the difficulty of setting up a fuil data bank in agricultural and rural development. Such a documentation system, limited to documents internal to the Bank, would probably not have an undue cost, and could significantly improve the informal communication system, by enabling all staff members to identify colleagues sharing the sa=e interest in a particular field. Finally, as noted earlier, it must be pointed out that communication be:veen research staff and operational staff members of the Bank is not necessarily direct. Indirect ways of communication may involve, for instance, te identification of a problem by a researcher belonging to the Bank's staff, its solution found outside the Bank, and the application of the solution being done by the operational staff on the basis of information provided by a third source. In such cases, the Bank's research will never be credited for the results, although it may have played a very significant role in their attainment. This must not be forgotten when judging research efficiency. From this point of view, information flowing from the Bank's researchers to the whole scientific community are just as essential as connections inside the Bank. The best way of ensuring communication in that respect is through the normal channels of research result dissemination, i.e*, professional papers, participation in scientific meeting and congresses. This is in effect one of the better ways of maximizing the multiplier effect of the Bank's research expenditures. At the same time, this implies that a continuous flow of information proceeds from the operational staff to the research staff. In practice, this is perhaps the most difficult thing to achieve, for the operational staff is (probably rightly) convinced that a new problem submitted to research will never be solved in due time. Thus, nobody in the operational staff is tempted to lose time discussing with the researchers. It is therefore essential to create communication patterns which would enable the researchers to forecast the needs likely to appear within the operational staff.. Again, informal links are probably more efficient in this respect than formal ones. -20- 2. Outside the Bank 7.9 A major obstacle to research dissemination outside the Bank, in the developing countries, is the lack of cooperating institutions capable of interacting with the Bank in identifying problems, translating them into research projects, in contributing to them and receiving research results from them that can then be adopted or adapted in the particular context of each country. 7.10 Consequently, major effort -- even at the potential expense of some reduction in research quality and cost effectivenss -- should be made to incorporate some element of institution building in virtually every researd project. Particularly where implementation implies a time-frame longer than the research project, and even longer than the disbursement period, institution must be developed to guarantee continuous and effective work. In the design and review of research and lending projects as currently practiced, an additional criteria should be used, namely the project's contribution toward the building of research and analytical capacity in the country being studied. 7.11 Possible actions to achieve this objective include the following: (i) increased participation of research staff in EDI activities; (ii) the identification of a cooperating institution, wherever possible, for each project processed through the Research Council; (iii) the involvement of researchers from the Bank staff in lending operations aimed at creating research institutions; and (iv) the sending of some researchers of the Bank staff on sab- batical leave for one or two years in LDC universities. -21- VIII. SUMMING UP The above review has indicated that we have found the Bank's AD research to generally be of high quality, and that it is widely acclaimed a professional circles. At the same time we have identified a number of prcblems, or potential problems, which it may be useful to enumerate specifi- i) Despite the major importance of agricultural lending for the Bank, it seems to us that the direct impact of ARD research on this activity has been marginal. ii) Despite the quality of research projects, it has been difficult for us to perceive any internal logic or clear focus to the ARD research portfolio. iii) Despite a number of interesting and stimulating projects, we do not feel that the Bank's ARD research has yet con- tributed a significant break-through in our understanding of the development process. iv) Despite individual examples of close cooperation by specific people at the working level, we feel that the current organization of the Bank's research does not bridge the gap between research and operations; it does not close the loop from perceived need, through research to implementation of results. 8.2 Despite these perceived shortcomings, which cannot be fixed by any one simple recommendation, our overall recommendation is that the Bank's research support for ARD studies should be expanded considerably - roughly in proportion to the increase in Bank lending for the ARD sector. This is partly due to our perception that much of the transfer of research results to operational practice is indirect. (As an example, the Nairobi speech and consequent "new style" projects, cannot be traced to any one study, but did reflect, to some extent, the cumulative impact of a number of individual studies and memos, which had the effect of questioning the impact of the previous pattern of Bank lending.) As the Bank's lending for ARD continues to increase while maintaining the slant towards the rural poor, the problem of designing useful projects and suggesting appropriate implementation strategies Will become more complex. Without proper research support, it will be very difficult to cope with the emerging tasks in an effective and innovative manner. In this context, the resources available for ARD research in the past seem relatively inadequate. 8.3 Despite the specific concerns mentioned above, the existence- of the Bank's research establishment, manned as it is by qualified and sensitive researchers, is a major asset. In view of the increasing needs for well researched ideas in the field of ARD, we sugggest the Bank should strengthen and systematize the related research activities and such to improve -22- the bi-directional flow of useful information between researchers and users both within and outside the Bank. It is in this spirit that we ha developed the following set of recommendations. h (i) The administration of research in the Bank should be strengthened by modifying the form and function of its existing machinery as follows: (a) a Research Council should replace the present Research Committee with a much wider mandate to strengthen research activities in the Bank; (b) the Council should identify major themes for research in ARD (and other sectors) and the research units in the Bank should give due importance to these themes while planning their activities; (c) the screening and monitoring of projects should be entrusted to the subject matter research panels. (ii) The organization of research should be improved to take full advantage of available resources. Toward this end: (a) each unit in the Bank entrusted with the task of research in ARD should have a well-defined mission in the context of a total research program; (b) given a mission orientation, each unit should then be encouraged to form research groups across jurisdictional lines in order to gain the critical mass needed for larger studies, and to take advantage of existing staff skills; (c) increased use should be made of joint studies with outside institutions, especially those in developing countries. (iii) The dissemination and utilization of research results should be given further attention. Toward this end: (a) the operational departments should be given resources to enable them to carry out certain research activities, especially to enable the Regions to make better use of research results; (b) a computerized information storage and retreival system should be considered; (c) specific responsibility should be placed on researchers to develop their output to the stage where it is readily accessible and relevant to operational staff. (iv) The Bank should make conscious effort to build up research capabilities in its developing member countries through its lending and research programs. (v) In developing the overall research program for ARD, gaps in research identified earlier in this report should be given priority attention, within the broader agenda suggested in Appendix I. (vi) In the light of increasing complexities in ARD, the growing magnitude of the lending operations in this area, and the need to develop a critical mass of researchers, Bank support for research in ARD should be expanded considerably. -23- Appendix I. Research Agenda According to the previously defined categories of need, possible of new or expanded areas of research follow. Nature and Magnitude of Emerging Tasks. (a) Links between demography and ARD. The various research projects that we have examined were, for the most part, vit.hout any explicit links with demographic studies conducted elsewhere !n the Bank or in the scientific community. However, demography, rural and urban balance, labor migration, age and sex composition of the rural household A,d rural sector, are all of importance to ARD, and warrant expanded efforts. (b) Water and Energy Balance. These agricultural inputs should be given particular attention in view of their limited supply and their critical importance. (c) New production techniques. The "book of blueprints" is a familiar idea for any production economist. !:evertheless, it is yet a theoretical idea, with little empirical content. ';*., in relation to agricultural development, the set of new techniques available is continuously changing, often with dramatic consequences, even for the poor traditional farmers. Research in this area might involve (1) a continuous review of new techniques as they emerge, particularly with close collaboration with the international centers such as IRRI and CIMMYT, and (2) the building of a set of models aiming at forecasting the dynamic consequences of their adoption on income distribution, crop location, and on national as well as international trade patterns. These studies must also incorporate social and behavioral elements. II. Options in Key Areas. (a) Relationships between techniques and institutions. A prototype for this type of research is the study of sharecropping recently done in DRC and DED. It should be expanded toward analogous investigation of other forms of institutions. Especially, socialist economic patterns should be examined from this point of view, in order to assess the conditions under which they are efficient. (b) Risk, uncertainty and finance. Outstanding research has been done on the effects of uncertainty on the level of production and the adoption of new techniques. But these studies -24- have not been related to a complementary aspect, namely finance. In effect the financial status of a production unit is one of the major determinants of its risk aversion, however the latter is measured. A clarification of this issue would cast important light on a major determinant of loan demand in rural areas and consequently on the adoption of capital intensive techniques. (c) Food and nutrition. Nutritional needs for low income populations is obviously an important question. The recent pilot study in AGR should be expanded and its con- clusions checked, in order to assess the efficiency of alternative forms of public intervention. (d) Factor shares. Technical progress in agriculture often involves three separate effects, an increase in production, a substitution of capital for labor and/or land, and economies of scale. What are the long term relative importance and consequences of these effects? This question is of major importance in evaluating the cost of helping the rural poor. III. Making the Lending Program More Effective. For the most part, the Bank's lending program in ARD aims at promoting new techniques. But the rationale for the adoption of these techniques by farmers and people living in rural areas is far from being clear. Research in this area might include: (a) Incentives of all kinds including both taxes and subsidies. In relation to household behavior the sociological and anthropological points of view are likely to beome important supplements to the neoclassical economic framework. All approaches require a theoretical basis, to be verified by appropriate surveys and model experiments. -These attempts need to be expanded, particularly in a variety of institutional and cultural contexts. (b) Rural infrastructure. The improvement of rural infrastructure, including rural roads and water supplies is an important concomitant of development, and may often greatly influence the dynamics of regional development. The impact of such infra- structure needs to be evaluated in a dynamic context. (c) "Software" aspects of projects. In most rural areas, the success or failure of a project depends on individual decisions which, in turn, depend on cultural patterns and institutions. So far this problem has often been neglected by the research as well as by the operational staff. Sociological and anthropological considerations would be of assistance. -25- (d) Learning from prior lending projects. Si=ary research priority for the Bank is to learn from its own experience, Apr athe "gold mine" that it has in its past history, not only in an cnal way by each staff member, but with serious application of methodo- ical sophistication, conceptual talents, analytical expertise, plus time (ots of time) and resources. (e) Strategies for reaching the rural poor. '-ere are two aspects of this, program content and methods of implementation, ,,rticularly to minimize the "leakages" that may divert a significant part o the impact away from the targeted population. *V. Methodology for Projects. Computer programs for data handling and management have been developed in various services of the Bank in an almost clandestine fashion, vihout it being recognized as a research problem. Nevertheless, the concep- :ual difficulties associated with these endeavours are far from being trivial. A- the same time, quick procedures for processing the large amount of data used by the various services could improve their efficiency in a very signifi- cant way. Moreover, if, in the future, the Bank were to enter into the business of using these data for other purposes than their immediate uses for projects, these problems should be solved. Therefore, it is recommended ::t some identified research projects in that field should be added to the research project portfolio. V. The Process of Change. The major question here is to elaborate a theory of rural develop- ment. One cannot imagine a better place than the Bank for the elaboration of this needed theory. (a) Intersectoral models. A prototype of these is the Social Accounting Matrix. It should be developed and expanded, especially in a more dynamic direction. Many other sorts of such models could be imagined, at the regional level, casting light on regional development issues, and their effects on production and income distribution. (b) Rural industries. Agricultural and Rural Development relies not only on farmers' behaviour, but also on the possibility of developing a significant industrial sector in the rural areas involved in projects. What kinds of industries are suitable in a variety of contexts? What are the obstacles against their birth and survival? Such are some of the questions that a rural industries research program could address itself. -26- (c) Insertion of traditional farmers in the market economy. This topic should be investigated in the light of the results drawn from the "incentive" topic already referred to. (d) Public utilities. The role of public utilities in ARD is self evident. Nevertheless, few studies have attempted a quantitative approach to measure its impact. If successful, such studies would make an invaluable contribution to project appraisal.