7.1. Proposal for a System-wide programme in public policy research
7.2. Centres' policy research capacity for collaborative research
7.3. Resources for inter-Centre collaborative policy and management research
7.4. Medium-term plans and coordination of inter-Centre policy and management research
7.5. IFPRI's System-wide backup role in policy research methodology
7.6. Reporting on policy research in the System
7.7. The special problem of management research
We have seen that policy and management research have assumed increasing importance in the CG System, absorbing a rising share of the total CG budget. This has come as a consequence of the CG system addressing increasingly complex questions for which policy and management solutions are undissociable from biophysical solutions. The result has been a wide range of initiatives in P&M research, not only by Centres explicitly devoted to this type of research, but also, in one form or another, by all Centres. Cooperative projects between Centres has also expanded rapidly, in part stimulated by availability of core financial support for inter-Centre initiatives. This has created a rapidly rising demand for IFPRI's services in the System.
Given the multidimensionality of the policy and management problems to be addressed and the complexity of the CG System, it is obvious that, if the CG is indeed to work as a System and if the delivery of results is to be efficient, there exists a crucial need for coordination both to address the division of labor among Centres in P&M research and the definition of collaborative ventures among Centres in these fields of research. Clearly, IFPRI is not in a position to deliver answers to the policy questions raised by all Centres. Once their research programmes will be at par with those of IFPRI, neither will ISNAR and IIMI be able to address all the management questions that relate to agricultural research and water, respectively. These Centres all have their own research agendas and it is consequently difficult for them to assume additional tasks on demand and inefficient to do so without proper coordination with their own research priorities. Yet, the demands for policy and management research that originate in the other Centres need to be addressed. While considerable inter-Centre cooperation is already happening, it has developed in a spontaneous fashion and efficiency gains may be achieved by some rationalization of the process. Efficiency gains would have a high payoff given the magnitude of excess demand for policy and management research in the System. In this chapter, we analyze the question of how coordination among CG Centres could best be achieved in these areas of research.
TAC has approved five System-wide programmes and 12 System-wide initiatives. They make core resources available for collaboration and coordination of research in those areas, including seed money for the preparation of joint research proposals. Logically, coordination of policy research could be achieved through an additional System-wide initiative in policy research. This would have the advantage of systematizing work in this subject and making resources available for the development of inter-Centre projects through the lead centre, IFPRI in this case.
While improved coordination is essential, a System-wide initiative in policy research is not the solution as it would be too analogous to IFPRI's overall mandate. For the same reason, there are no System-wide initiatives in the commodities which define the programme of specific Centres. The CG cannot have two separate research agendas in policy research, one defined by IFPRI's medium term plan and the other by a System-wide initiative. Or, if the two are the same, this risks creating conflicts over who sets the agenda for CG policy research, IFPRI or the committee in control of the System-wide initiative. In addition, policy research has many different dimensions, and there is no need for overall coordination of policy research as an objective across the whole System. Coordination should only be pursued to avoid duplications, establish comparative advantages in relation to specific questions, and when useful for specific research or outreach initiatives. In general, it seems desirable to avoid adding further administrative layers to an already complex system and to seek instead to enhance inter-Centre coordination and collaboration through decentralized and informal mechanisms while maintaining autonomous Centre initiatives. Managing this complex coordination may, nonetheless, require a purposeful allocation of core resources to this field of research as we shall discuss below.
In the Centres' survey organized for this study, 11 centres responded opposing a System-wide initiative in policy research, two supported the idea, and two abstained. It is thus safe to say that opposition to the initiative is pervasive, not only at IFPRI but also throughout the system as a whole. Four Centres who opposed the idea nonetheless suggested that inter-Centre initiatives on selective P&M research topics may be useful.
The Panel consequently recommends not to pursue a System-wide initiative in policy research as an instrument to enhance collaborations and coordination in this area of research. System-wide initiatives are appropriate when focused on more narrowly defined topics. For this reason, the door should be kept open for selective themes in P&M research to give rise to new inter-Centre initiatives.
One should start by observing that there already exists a considerable degree of collaborative research in the System. IFPRI currently has joint research projects with 11 Centres: CIAT, CIP, CIMMYT, ICARDA, ICLARM, ICRAF, ICRISAT, IIMI, ILCA, IRRI, and ISNAR (IFPRI Report, 1994). It is participating to several System-wide initiatives and has taken the lead in developing two of the approved initiatives, in water management coordinated by IIMI and in property rights coordinated by IFPRI. There are, however, a number of problems that need to be resolved to make this inter-Centre collaboration more effective. Solution to these problems will require alternative instruments if they are not to be managed through a System-wide initiative in policy research.
One problem is that effective research collaboration between IFPRI and other Centres requires more overlap in the capacities of the Centres involved. Effective team work needs enough shared analytical capabilities among team participants. This implies that the participating Centre must have sufficient socioeconomic research capacity and some policy analysis capacity of its own. In part this is because the Centre's social scientists are better able to collaborate with the specialized biological scientists of the problem under study than IFPRI's social scientists since they cohabit with these scientists. The more directly consistent with IFPRI's own long term research priorities the research questions are, the less autonomous policy analysis capacity would a Centre need to develop for effective collaboration. However, all Centres need to have a minimum capacity to engage in collaborative research. Additionally, if the research effort is to be sustained beyond the phase of collaboration with IFPRI, individual Centre capacity to do this is needed. Clearly, the level of sophistication and expertise in policy research need in general not be at the same level as IFPRI's for general themes, but it should be at par with IFPRI's for the specialized themes that concern the Centre and where sustained research will be needed.
The Panel consequently recommends that Centres review their current capacities in socioeconomic and policy research to assess whether this minimum capacity for desired collaborative research is in place and take remedial action if it is not.
In the survey organized for this study, the question was asked whether a Centre plans to strengthen its in-house capacity for P&M research or to rely instead on external collaborations. Out of 13 answers, 10 said that they would increase in-house capacity, although not at the exclusion of enhanced external collaborations. This suggests that the share of CG budget allocated to P&M research may not have peaked yet.
Given overall stagnation in real budgets, one mechanism to enhance collaborative research and strengthen the Centres' policy research capacity would be to outpost more IFPRI scientists in the Centres, at least for the duration of collaborative ventures. IFPRI has been effective in outposting some of its research staff in collaborating institutions. This has been particularly beneficial when policy research implies a strong continuum with socioeconomic research developed at the Centre level. A possibility would be to place these outposted staff in regional units, located in CG Centres, that also regroup outposted staff from ISNAR and IIMI. This does not go against maintaining a critical mass of scientists in Washington D.C. to insure the proper environment for scientific creativity through peer interactions and for interaction with visiting scholars from collaborating NARSs and NAPAs.
The Panel consequently recommends that IFPRI's outposted staff be increased when consistent with research projects and when it enhances a Centre's policy research capacity and interactions with ISNAR and IIMI outposted scientists as a necessary part of regional projects.
A second problem is the lack of resources for the definition of collaborative projects and the preparation of research proposals in policy research. Centres' demand for IFPRI participation has been very high, but it has not been correspondingly supported by resources at the project definition and implementation stages. This has led to superficial participation, protracted periods of project preparation, or inability to follow up on worthwhile potential initiatives. Because they will involve a multiplicity of actors and institutions, these projects are usually particularly complex and costly to organize. Development of these proposals typically takes one or two years of work and negotiations, and they require costs for the gathering of preliminary information and meetings. In principle, the costs of project development should be fully included in project budgets. In practice, this is very difficult to do: project budgets cannot cover retroactively costs already incurred in the preparation of proposals; and few donors are interested in funding project preparation as a separate activity. In general, resources for pre-project activities (as well as for post-project syntheses and derivation of policy conclusions) are lacking under conditions where projects have to be funded by special grants. Preparation of inter-Centre projects in policy research will consequently require continued use of unrestricted funding. If inter-Centre projects are to multiply, availability of unrestricted funding for the development of these projects will correspondingly have to increase.
TAC needs to make resources available for the definition of collaborative projects in policy research and for coordination among projects. In particular, resources need to be made available to IFPRI scientists to help them increase their capacity to respond to requests for participation in multi-Centre research projects or to take initiatives for such projects. Because Centres' ultimate objectives may differ from those of IFPRI, even though they might seek research collaboration, this Panel does not consider adequate that resources for collaborative work be placed at IFPRI. At the same time, TAC should be watchful that non-existence of a System-wide initiative in policy research does not penalize this field of research for access to funds for inter-Centre project preparation and coordination, as done in System-wide initiatives.
The Panel consequently recommends that TAC consider defining a modality that would allow it to allocate resources on a competitive basis in response to requests by two or more Centres for funds to support project preparation and coordination in socioeconomic, policy, and management research. Availability of these funds should be widely advertised in the System, criteria for proposal selection made explicit, and adequate mechanisms for project selection put into place.
Collaborative research between Centres requires that research topics fit within their medium term plans (MTP). If IFPRI is to be at the Centre of much collaborative policy research, adherence of collaborative research to its medium term plans is fundamental to filter excess demand and maintain a cumulative logic to collaborations. However, if the medium term plan is such an important filter for cooperative policy research efforts, other Centres should have an opportunity to participate to IFPRI's process of medium term plan development, if they anticipate the need to collaborate with IFPRI. For the moment, IFPRI's MTP is elaborated through internal discussions, sharing of drafts with NARSs, discussions with donors, regional consultations, and discussion and approval by the Board. Greater consultations would seem necessary to harmonize IFPRI's and potential collaborating Centres' MTPs in anticipation of joint research projects.
Thomas Walker, Director General of CIP, recommends that IFPRI appoint a research fellow to be a liaison person with a specific centre. His observation is that IFPRI social scientists relate well to the profession while Centre social scientists relate to the biological scientists in their Centre. The IFPRI liaison person would help IFPRI access information about what non-social scientists in the other Centers are doing. This would be especially relevant in the drafting of medium-term plans. To quote: "We would like to have someone with whom we could interact on a more routine basis and who would be interested in CDP's commodity and natural resource mandate. IFPRI social scientists know a lot about rice and irrigation, but their knowledge of several Centres' mandates and work areas is sparse. Having a formal contact with a Center would also enhance that person's agricultural experience which is important to the CGIAR as general agriculturalists are increasingly difficult to come by in this age of increasing specialization. This does not imply that the IFPRI contact should be working on the research and raining in the area of the Centre's mandate; only that she or he should be comfortable with it. The spirit of this suggestion is to look for cost-effective ways to bridge the gap between biological scientists in the CGIAR and IFPRI social scientists."
The Panel suggests that Centres which may want to establish collaborative research with IFPRI should be offered greater participation to the development of IFPRI's medium term plan and, reciprocally, opportunities to coordinate their own medium term plans with IFPRI's. It recommends that IFPRI consider appointing a research fellow to be a liaison person with a specific centre.
IFPRI's policy research priorities and those of other Centres are not identical. IFPRI has a global policy research mandate while other Centres tend to be oriented at specific commodities, production systems, or agroecological environments. The policy questions which they can best address deal with production, input use, technology, and farm-level environmental problems related to these commodities, systems, and environments. This research should consequently largely be done by the Centres themselves and their collaborating NARSs, sometimes with IFPRI collaboration, but most often without it. However, even when there is no direct research collaboration, IFPRI could play a methodological back-up and capacity strengthening role when the research calls on methodologies for which there has been a long expertise at IFPRI. An outstanding recent example of this function is the user-friendly econometrics manual for quantitative policy analysis prepared by L. Haddad et al. (1995). In some cases, this type of assistance could well be going from Centres to IFPRI, for instance on research related to management science or environmental economics. In all cases, careful assessment needs to be made of the relative advantages of obtaining this expertise from within the System as opposed to from outside, for instance through joint projects, short courses, or sabbatical leave invitations to university faculties. When this type of collaboration is sought within the System, expertise could be made available through short term consultancies, methodological semi NARSs, and temporary assignment of senior scientists to another Centre to the benefit of the receiving Centre. As observed in chapter 5 of this report, fulfilling this function would induce more IFPRI scientists to move upstream in their research to better assume this technical assistance function. If the System is to perform as a system, this type of collaboration should be induced.
For this purpose, the Panel recommends that TAC consider using the same modality as recommended in section 7.3 above to make resources available on a competitive basis that would have the function of supporting methodological backup activities across Centres. These resources could be managed jointly with funds in support of inter-Centre initiatives in P&M research.
Starting in 1996, the CGIAR research agenda and the budgetary allocation will be presented as a matrix with 17 programme categories. One of these programmes is "policy" which is expected to absorb 13% of the System's 1996 budget. This, however, exaggerates the extent of policy research in the System since, under the "policy" heading, are included socioeconomic research, management research, and policy research proper. In most Centres, the bulk of the "policy" budget goes to socioeconomic research. There is, however, some underestimation since the "protecting the environment" and the "increasing productivity-production systems development and management" programmes includes some policy research. Thus, to adequately report on the extent of policy versus socioeconomic and management research in the System would require further disaggregation of other programme categories.
In general, it would appear desirable to convey to donors exact figures about the extent of policy research in the System. Both the "policy" and the "protecting the environment" programmes should consequently be disaggregated into their respective components to report on the exact amount of on-going policy research. This will be particularly useful to analyze the relative importance of these research categories and observe their respective growth rates. To summarize budget presentation, these categories can subsequently be aggregated in a "socioeconomic, policy, and management" programme as seen fit.
The Panel suggests that budget reporting on the "policy" and "protecting the environment" programmes be disaggregated in their respective components to adequately report on the extent of socioeconomic, policy, and management research in the System.
There are few policy recommendations which do not have management research implications. To be successfully implemented, policies need effective administrative bodies and regulatory agencies, as shown in Figure 2.1. In addition, for a given policy context, there are management options that are important determinants of the effectiveness with which resources are used and of the welfare implications which this use has. In the System, IFPRI's policy research consequently implies delving into public management research. Management research on specialized aspects of the CG mandate are assumed by ISNAR for agricultural research and development, IIMI for water, and CIFOR for forestry. Every Centre also runs into management research questions. For the System, the question of coordination in management research is thus of equal importance as that of coordination in policy research.
One particular field of management research pertaining to the functioning of agricultural research systems may be a relatively small one in the global picture of management work, but is surely a potentially potent one for fostering the improved functioning of NARSs, as well as other entities, such as IARCs. ISNAR is the IARC purposefully dedicated to advancing knowledge in this field, and yet it has been a rather modest participant in research per se dealing with management. The main approach adopted at ISNAR has been to bring together experienced personnel and put them to tasks in advising national systems on how to improve management and organization, based largely on their unique personal experiences and on ISNAR's cumulative institutional wisdom. This is effective as far as such direct and rapid intervention can be useful. What it appears not to be doing sufficently, unless undertaken in a systematically cross-sectional way, is set up a learning situation from which new insights may eventually be derived. The sketched past approach is highly dependent on individual personalities and their particular past, and may not lead to general understanding of the pros and cons of alternative institutional structures, for instance. This raises the question of the optimum balance and synergy between service and research, an issue that ISNAR's forthcoming EPM review will need to address anew.
ISNAR has focused on some selected aspects of NARSs that can contribute to better management, most notably planning, monitoring and evaluation, and this has led to many useful developments of methods that have wide applicability and can usefully inform research managers. Indeed, many of ISNAR's generated ideas have been accepted by donors and international development agencies. This, however, is not the same as a broadly-based research programme in management per se, and such studies need to take place on a more systematic fashion.
The general approach at ISNAR has been to conceptualize management and organization problems at a rather generic level, hire professionals with broad experience to devote themselves primarily to advisory work with requesting NARSs, and to document the experiences in such a way that the preconditions are set for distilling useful new perceptions and hopefully new knowledge. A question that the forthcoming EPR should address is whether this, in itself, amounts to a research programme in the broad area of research-system management, given the importance of this information if agreed new forms of best practice and improved knowledge of bad practices are to be assembled more systematically.
A potentially important research-related activity that spans both policy and management in the field of administering technology generation is research priority setting. ISNAR has supported pioneering efforts that have culminated in the publication of a path-breaking book (Alston, Norton and Pardey, 1994), which will form the basis for orchestrating a proliferation of needed efforts in NARSs to direct scarce research resources to high-payoff areas. Management information systems geared to the particular needs of research systems, also developed at ISNAR, can usefully link to the ex-ante economic analysis work to trace the actual commitment of human and financial resources, monitor the effectiveness of the planning process, and ultimately facilitate the evaluation of the outcomes of research programmes. The whole field of research policy analysis and research strategy formulation is wide and to date still substantially under-attended. The proposed collaborative efforts between ISNAR and IFPRI should be fruitful in this regard. Other groups affiliated to the CG System, such as the newly-established Agricultural Research and Extension group in the World Bank (ESDAR), are also eager to advance practice in this area, and the prospects for the donor community devoting greater resources to this theme seem bright.
Management research in CIFOR takes several different forms. Management, in the sense of the discipline that might be studied in a business school, is conducted primarily in the policy-oriented work, where it is integrated into the socioeconomic research work generally, which has a strong policy bias, and management is interpreted as just one important aspect of policy work.
Much more work at CIFOR on management is done in other programmes, including those with management in the title. But this work is primarily of a biophysical nature and is concerned essentially with the technologies of managing such things as biodiversity in forests, reforestation of degraded land, and so on. It is, of course, quite legitimate for this work to be categorized as management research, but it is of quite a different character from most of the rest of the declared management research in the CGIAR system.
Given the attention to managerial issues in so many of CIFOR's programmes, there will definitely be a reporting problem that underlies any data on the orientation of research activities of CIFOR, and thus of the CGIAR system at large. These special difficulties need to be recognized in the categorization of such research efforts and their corresponding reporting, and these difficulties are likely to be similarly encountered in other resource-based Centres such as ICLARM and ICRAF.
The above comments refer particularly to socioeconomic work concerning the management of various forest resources by local communities, in CIFOR's case especially for non-timber forest products in various parts of the world. The dissection of such activities into socioeconomic research per se, versus policy and eventually management categories, is going to be a challenging task that does not have ready resolution.
Needless to say, a research programme addressed to measuring and, in general, dealing with the sustainable management of forests, either generically or in specific geographical areas. such as is the case in some CIFOR work, is fraught by difficulties ranging from basic measurement through to conceptual and sheer practical aspects such as, with most difficulty for a System with as short as an evaluative horizon as the CGIAR, the long lapses of time before definitive results can be recorded and realized. There is really no way to dodge this latter dilemma except for the supporters of such work to adopt an extremely patient attitude.
Research in water management is, in general, particularly difficult to organize as it is closely linked to the actual operation of water management line agencies and water users associations. This implies that information can hardly be accessed without participation to the operation of these institutions and that research is directly linked to institution building and management advice. This is typical action research which is well known to be difficult to manage methodologically. It is also difficult for the researcher to take a distance from the immediate needs of the agency involved, and hence to extract strategic research implications from management advice. Indeed, IIMI has been effective in management advice as evidenced by the magnitude of the demand for its services, but is to this stage still short in strategic research. At the same time, this link to action gives unique opportunities to access data otherwise beyond the reach of researchers, experiment with alternative organizational systems, engage in collaborative research, and create immediate effective demand for research results.
IIMI has been a late comer to the System. Before this, it had been in operation for some time as a service agency to public and private water management institutions, with an adaptive as opposed to a strategic research mandate. At this date, it is still in full transition from this former function to pursuing a balanced agenda that fulfills both the CG's strategic research mandate and the need to service clients as part of research and institution strengthening. Given the need to expand its strategic research agenda, departing staff will need to be replaced by young economists and management scientists well trained in research methods, implying a move upstream toward applied theory and methods. Collaboration with IFPRI and other Centres, as well as universities and centres of excellence in the field, is also particularly important to reinforce the move toward strategic research. In addition, IIMI has broadened its research perspective from irrigation to the management of water in watersheds, river basins, and irrigation sectors, with the associated problems of environment, health, and broader economic linkages. As a result, IIMI's research mandate necessarily overlaps with that of several other Centres. Several promising collaborative research initiatives with IFPRI have recently been organized for work in Pakistan, India, and Egypt. Initiation of additional collaborative efforts will hopefully be facilitated by the IIMI-convened System wide Water Resources Management Programme.
IIMI is still short in expertise in policy and economic analysis while it is well endowed in irrigation engineers and social scientists specialized in management research. Dimensions of household, community, and cooperative behavior in water users' associations are at best only partially covered and will benefit from collaboration with IFPRI. However, to be effective, this collaboration requires that a critical mass of socioeconomists, institutional economists, and policy analysts be in place at IIMI. Devolution of management of water districts to users in particular extends the range of necessary expertise from largely public agencies to the Grafting of civil cooperative institutions. Subjects like the development of water markets, the economic incentives for the performance of water users' associations, the internalization of externalities in watershed management, the definition of reward systems for managers of water districts, the control of rent seeking in the appropriation of water, and the adoption of improved irrigation and soil conservation technologies are all subjects that have an important dimension of economic and institutional analysis. Among analytical skills, they call on the yet much neglected fields of "new" institutional economics and political economy among CG social scientists.
Improving the efficiency of water use is a crucial theme for the world food situation and for rural welfare in irrigated areas. Yet. obtaining funding for the development of complex joint research projects that involve participation of a multiplicity of disciplines and Centres has proved difficult. In part, this may be due to difficulties in investing sufficiently in project development in an area where research questions are often poorly charted. For instance, requests for funding of pre-projects based on rapid appraisal techniques for the identification of research hypotheses on water markets have proved of little attraction to donors. This stresses again the need for access to unrestricted funding for inter-Centre project development. It also requires the development of medium term research priorities that link IIMI, IFPRI, and other Centres' research agendas.
