3.1 Priorities and Priority-Setting
3.2 Quality of Research
3.3 Relevance and Impacts of IFPRI's Work - What for Whom
3.4. Research Environment
3.5 Institutional and Scientist Linkages
The preceding chapters provided an overview of and an initial assessment of divisional activities in research and outreach. This chapter focuses on cross-cutting programmatic themes. Organization and Management themes are addressed in Chapter 4.
Priorities for policy research that bear on poverty alleviation, food security and environmental sustainability are not easy to establish, especially when global, regional and even country level issues fall within its purview. The credibility of the exercise will then lie in the process by which policy issues were identified, the criteria that were applied in assessing their relative importance, and the expertise and breadth of experience of those who made the final choices. The research agenda that emerges from this process will no doubt depend also on the donor resources available during the period involved. Under the circumstances, what matters most in determining the credibility of the choices resulting from priority-setting, is the quality and commitment to developing country policy concerns of the people involved in the process, rather than further refinements of the process. It is not surprising, therefore, that reasonable people tend to agree with a process for priority-setting, and yet differ on the actual outcomes in terms of priorities.
After an assessment of the priority-setting process that IFPRI uses, the range of consultations that have taken place, and the criteria deployed, the Panel is satisfied that the way the Centre has gone about priority-setting, as well as the research agenda its Board finally approved are both appropriate and timely, given the outputs available from other policy research bodies, the comparative advantage of the staff of the Centre, and the uncertainly about donor resources. While donor preferences have no doubt influenced the final research agenda, there are also cases where the Centre has accorded high priority to some projects (e.g., research on water), although it has not been able to mobilize adequate funds for them.
IFPRI uses a pragmatic process that recognizes the impossibility of carrying out a highly sophisticated quantitative priority-setting exercise in policy research. Rather, IFPRI depends on client input, and input from peers in the policy research field, and the IFPRI researchers' judgement. The Panel fully accepts the good quality of the latter two inputs. At the same time the Panel suggests that IFPRI more fully exploit the means available to gain input in its priority-setting from developing country policymakers, researchers and analysts. For example, there appear to be opportunities to design outreach activities that are more explicitly two way communication exercises and that go beyond the specific research in question. This could generate new ideas for research modification or for new projects and provide input for priority-setting.
At the same time the Panel cautions that, as a CGIAR Centre, IFPRI should be focusing on international public goods research, i.e., research that would not be undertaken by any one country in isolation, yet would be of benefit to many if undertaken. In such cases, there can be some direct conflicts between listening to a given country's priorities for research as opposed to synthesizing the ideas from many countries where a given issue is important, yet not the top priority in any one country. In this case, IFPRI is balancing a "quality" of choice criterion related to international public goods research against criteria that any given country would apply in their own national priority-setting framework.
A final question is whether IFPRI's resources are being allocated in harmony with CG goals and priorities. This is not an easy question to answer because programmes and projects have multiple goals not all of which will be limited to specific CG goals and priorities. Some activities such as 2020 have broad goals that serve the entire spectrum of priorities. An approximate fit between IFPRI's resource allocation and CG goals and priorities can be had by reviewing the pattern of allocation of funds among the relevant groups of programmes in IFPRI. IFPRI's research on consumption and nutrition, for example, which has a direct bearing on the goals of poverty alleviation and food security accounted for the use of 31% of the Centre's funds during the period under review. Its research on environmental and technology issues accounted for the use of 27% of the funds. This work was largely concerned with the themes of environmental sustainability and poverty alleviation. IFPRI's research in respect of trade and macroeconomic issues and markets (inputs and outputs) which are more general themes added up to only 24% of its funds, while outreach accounted for 18%.
3.2.1 The Selection of Research Inputs
3.2.2 The Quality of Research Methods at IFPRI
3.2.3 The Quality of Published Research Output: Internal and External Quality Control
3.2.4 The Incentives for Quality at IFPRI
One can think of quality of research at different stages in the research process. Thus, there is quality of the problem formulation (discussed in Section 3.1). quality of the inputs used in the research process (related to the data collection processes), quality of the research itself (methodology), quality of the output (mainly the publications and other written form in which the output is presented, where quality is measured in terms of the intended audiences and their abilities to comprehend and use the results). All these dimensions of quality are relevant in the case of a CGIAR Centre, and thus in the case of IFPRI.
The Panel concludes that IFPRI's work is of good quality at all stages in the research process. It also appears that quality at IFPRI receives good marks from other institutions. These high marks are a consequence of a careful selection of research problems and methods, quality standards being applied to inputs - researchers and data, and use of an efficient internal quality control system. The system of incentives used within IFPRI also rewards quality in output. Care should be taken, however, not to smother new and original ideas which would not conform, initially at least, to current mainstreams of thinking and thus conventional quality criteria.
The two essential inputs for economic research are researchers and data. Researchers are recruited through careful procedures, described and commented on in Chapter 4. The Panel's judgement is that IFPRI researchers are of high academic quality. IFPRI is known for designing primary data collection methods which stand as models for others long after their completion. It is one of a few economic research institutes producing a significant proportion of the primary data that they utilize. It must be stressed that these primary data also are utilized, and highly demanded by, numerous other academic institutions and their Ph.D students. This relates to the linkage issue addressed in Section 3.4.
The Panel commends IFPRI for its foresight in assembling a distinguished list of independent experts, chosen for their familiarity with the themes addressed by IFPRI and for their methodological rigour, to carry out a thorough Internally Commissioned External Review of IFPRI's research methodology. As discussed earlier, the Panel cannot improve on the ICER review of quality and appropriateness of the research methods used by IFPRI. The study was detailed and covered all research programmes. In general the ICER gave IFPRI high marks, with some specific minor reservations or differences of opinion here and there that IFPRI for the most part has addressed and either adopted or rejected for clear reasons. The Panel reviewed the ICER output in great detail, as well as the IFPRI responses and is fully satisfied with the work of the ICER and the IFPRI's responses to its recommendations.
The Panel encourages IFPRI to continue with the ICER process, perhaps focusing on research issues in addition to quality of methods. Also, when the new OD is fully up and running and has an operational plan in implementation, an ICER looking at the quality of outreach also would be in order.
Overall, the Panel judges IFPRI's published research output to be of high quality.
The internal quality control begins when a research project is being planned, long before any publication is considered. Before any project is accepted in the framework of an MP or of a GRP, the methodology is reviewed by an external reviewer, proposed by the head of the division and nominated by the DG. Usually a seminar is organized, with participants from the World Bank or from universities being invited to comment. The high quality of collected data relies mainly on these procedures.
During the process of execution, projects are submitted to regular review. An internal review at the whole institution level is conducted annually, with any staff member invited to comment on the operations of divisions and selected programmes. In some cases, when a programme is complex (for instance MP17), external interdisciplinary advisory boards are set up and meet annually to provide advice.
Finally, any IFPRI labelled paper, even a discussion paper, must go through an internal quality control process. IFPRI staff members are free to write whatever they like in their own name and researchers are encouraged to publish any article in any refereed journal. Reprints are submitted to the Information Programme and are approved for publication by the Director General. The nature of the control varies with the series the paper is published in: discussion papers are reviewed by one staff member; research report manuscripts contemplated for publication have to be presented at a seminar, thus allowing for the collection of criticisms, remarks and suggestions from the audience, and are subsequently examined along with books by a publication committee. The latter does not perform the entire review by itself, but nominates at least one internal and two external individuals before giving the IFPRI stamp to any document. Usually, a dialogue takes place between the author(s) and the reviewers, with successive versions of the document being prepared before publication occurs. Furthermore, books (especially the famous John Hopkins University Press IFPRI series) are submitted to the publishers' scrutiny regarding sales and other commercial aspects.
All these procedures not only avoid publication of inferior material but also guarantee that the audience of a document will understand the true nature of the delivered message. The Panel commends IFPRI for its thorough internal review and quality control process.
Other items that deserve mention are the short policy briefs and project briefs that currently are being produced. They are of high quality and extremely useful for the non-research person involved in policy-making. The Panel commends IFPRI on the decision to produce these. However, a gap has been noted by some development policy analysts between broad policy statements and publications that are highly theoretical and full of equations. They have expressed a need for intermediate publications that explain the assumptions used to help clarify analytical methods and facilitate their application.
The Panel urges all IFPRI divisions to systematically consider the balance in its written output, making sure not to neglect the needs of those who wish further details on research methods and assumptions used.
There is no formal incentive system for research and publication quality. Yet, it is clear that for a researcher to have a good publication record is a major asset in salary discussions, either within IFPRI or outside, as well as for future employability. The internal institutional culture also plays a large role in maintaining a high demand for quality, as was made apparent by the number and the nature of interventions during the 1997 internal review.
A bibliometric study, showing that IFPRI's publications are often quoted and given authority provides quantitative evidence of the relevance and implied quality of these criteria (see also Section 3.3 on impacts).
However, the system of quality control also has its costs, both financial and intellectual. In terms of financial costs, the time devoted to reading and correcting manuscripts can be significant in IFPRI, and to some extent contributes to the allegedly "high cost" of an IFPRI internationally recruited researcher.
In terms of intellectual cost, the review process, although extremely efficient in avoiding dramatic mistakes and the publication of unwarranted statements, is also at risk of preventing new "shocking ideas " to be expressed publicly. For instance, it can very well be imagined that the early works of Louis Pasteur on microbiology would have been refused by the IFPRI reviewing system, had the latter been operational at the time, simply because his ideas were unacceptable then. There are no indisputable ways of avoiding the risk of rebuffing a good manuscript, except by carefully choosing broad and open minded reviewers. The Panel is, however, satisfied that IFPRI has often called into question commonly accepted principles.
Whatever the financial (opportunity) and intellectual costs of quality control because of the necessary large commitment of time, these costs, must be incurred, otherwise research would lack all justification. Furthermore, there are some hidden benefits that researchers derive from review of their peers' work. First, they keep abreast of what others in their fields are doing and thinking. Second, even an experienced researcher can learn from a thorough review of a colleague's work.
3.3.1 IFPRI's Overall Work in Impact Assessment
3.3.2 IFPRI's Impacts
3.3.3 Impacts on Capacity Strengthening in Developing Countries
3.3.4 Case Studies of Impacts on Policy Decisions in Developing Countries
3.3.5 Concluding Comments
The ultimate objective of IFPRI is to have an impact on poverty reduction, food security, and sustainable development. In IFPRI's case, these impacts will come via changes in policies, which is its research and outreach domain. The impacts will come through two types of outputs: directly through the policy relevant information it provides policymakers; and indirectly through capacity strengthening (that strengthens the ability of local policy researchers to produce needed information on policy options).
In both cases, impacts will depend on (1) the relevance (and quality) of the information and techniques developed and extended to target groups; and (2) the number and nature of such target groups effectively reached with relevant knowledge.
The CGIAR System increasingly is focusing on impacts of its investment in agricultural and related research. It rightly requires centres to develop improved and expanded measures of their impacts, both for accountability purposes and because work in this area helps centres focus on the key objectives and the indicators and milestones that can help them determine whether they are on the right course as they proceed down the research path. IFPRI is taking very seriously this charge from the System, partly because it has been involved in work on impact assessment for some time, and partly because it agrees with the benefits that can be derived from a better understanding of impacts - both in guiding ongoing research and in setting priorities for future research.
While considerable progress has been made over the years in measuring the impacts of biophysical research, e.g., plant breeding and production research, such results cannot be applied readily to social science and particularly policy research that is of quite a different nature in terms of final impacts. In fact, the Panel believes that it will be a long time off before definitive, quantitative assessment of the final impacts of policy research on the human condition and sustainable development will be possible. As a result, the Panel cautions against expecting such impact assessment results. Rather, IFPRI's clients should expect good progress from IFPRI in providing measures of the intermediate impacts of its work - measures of the impacts of its written output (in terms of usefulness and extent) and impacts of its capacity strengthening activities in terms of the quality and relevance of work produced by those clients involved in such strengthening activities. In addition, as is already the case, IFPRI's clients should expect more anecdotal evidence of impacts, or at least evidence that the results indicated by IFPRI's work are consistent with policy changes that actually have or are taking place. In fact, as indicated below, the Panel concludes that IFPRI is making good progress on all these fronts.
IFPRI is involved in two types of work related to impacts. One, which is more of a research activity, deals with the development of social science methods for assessing the impacts of agricultural, food, and natural resources research. That work is taking place mainly in EPTD.
The other activity is related to assessing the impacts of its own policy research, or social science research more broadly considered. For this work a new position has been created, but not yet filled. The focus in this section is on the latter thrust.
Progress is being made in understanding the conceptual underpinnings of the impacts of policy research. In fact, IFPRI is on the forefront of this work and has sponsored several recent activities that bring together the best thinking on this subject. The resulting papers are being edited for publication in a book. IFPRI is to be commended for its activity in this area. IFPRI researchers have recognised that the topic is a multidimensional one, and that specification of the possible paths of impacts is an important task at hand. Researchers at IFPRI also have explored the various issues surrounding the problem of attribution when it comes to linking policy research to ultimate impact on poverty alleviation and other measures of human welfare. At the same time, IFPRI has been exploring impact assessment potential from different perspectives, and in terms of different stages along the path from research input to final impact.
As mentioned above, the Panel considers that the impact of IFPRI's research in a practical sense depends on two conditions First, the research must be relevant to the problems at hand, i.e., the perceived priority issues and options faced by the countries to which the research is addressed. Given that necessary condition, the impact then is the result of how well and how widely the research results - either substantive results or methods and techniques - are extended and integrated into the target country research and policy arenas. This is where the outreach function becomes a critical one (see Section 2.5).
The issue of relevance of IFPRI's work in its various research divisions was discussed under each of the division headings in Chapter 2. The general conclusion that emerges is that nearly all of IFPRI's chosen topics during this MTP are of high priority and very relevant to the agricultural, food, nutrition, and natural resources issues faced by developing countries. Thus, in terms of the discussion of impacts, the Panel concludes that the first necessary condition for having significant impact has been met in IFPRI; it is working on issues and research methods that are relevant to resolving the policy issues that exist in developing countries. Below we discuss the second dimension - the effectiveness of IFPRI's extension of its research and technical knowledge to the relevant actors in the research and policy arenas of its target countries and regions.
The Panel believes that a reasonable approach to, and emphasis on impact assessment is being taken by IFPRI. On the one hand, it is at this stage focusing on intermediate impacts, recognising that final impacts on poverty alleviation and sustainable development are difficult if not impossible to measure at this stage in our understanding of the processes involved. On the other hand, IFPRI is moving ahead aggressively to develop with its partners further conceptual underpinnings and methods for assessing the impacts of policy research at increasingly more distant points along the continuum from research input, through research output, transfer, and ultimate use in policies that contribute to poverty alleviation, food security and environmental protection.
While the Panel agrees in general with this approach, it also wishes to point to the need for IFPRI to devote more effort to developing effective milestones for all its projects, i.e., key intermediate targets necessary for achieving and/or delivering a project's output within an agreed timeframe. Such milestones along the research path indicate whether in fact IFPRI is moving towards achievement of the ultimate objectives of the research projects. Such milestones could include the holding of training sessions, workshops, and conferences, reference to IFPRI research in local media sources, number of persons trained at given levels, number of publications and papers, and so forth.
At present, IFPRI is exploring its intermediate impacts at the stage of its outputs - its publications, training courses, and outreach activities. Thus, IFPRI recently has been involved in a number of activities that further an understanding of its publications and the audiences and uses of its publications. It has been exploring through an independent survey organization the nature of its readers and their satisfaction with its written outputs; it has been looking at the use of its science output by other researchers; and it has a continuing process of evaluating its courses by participants. It also is developing some anecdotal information that further links its activities, at least in a qualitative sense, to actual policy change. The evidence gathered so far from these various activities is discussed below together with the Panel's assessment of the activities and their results.
Capacity strengthening has become an increasingly important goal of IFPRI over the review period. What is the evidence of the impacts of such activity? First, one proxy measure is the number of policy researchers that have been involved in such activity. Second, the impacts of IFPRI's written work - publications and other papers - can be assessed through reader surveys and through analysis of the use of IFPRI publications. The Panel considered a number of different measures of such intermediate impacts. The overall conclusion is that IFPRI is doing a good job, but that it could improve the impacts of its outreach activities as discussed below.
Training courses and other related activities. IFPRI has formal training activity associated with most of its research activity, in addition to its more informal approach of involving researchers from developing countries in training through association and joint work with IFPRI researchers. In the area of training, a number of examples of activity provide a perspective on the dimensions of IFPRI's involvement. For example, IFPRI conducted courses for more than 110 Malawi policy analysts and regional training for 30 participants from other Southern African countries. Virtually all the trained people are now working in governments in those countries. In Pakistan, IFPRI collaborated with the government in training more than 60 people in policy-related data collection and analysis, including econometrics; and data generated by IFPRI's collaborative project in Pakistan has been used in more than 100 master level dissertations related to food policy.
The Panel concludes that the anecdotal evidence available indicates that IFPRI is doing a good job in capacity strengthening, but that it could do a better and more consistent job of it, if it developed a more focused approach within the context and guidance of a general strategy for capacity strengthening and outreach.
The capacity strengthening impact from IFPRI's projects appear to be highly variable. Thus, some additional structuring of this function across the Institute could have benefits. In fact, the Panel notes that IFPRI intends to develop such focus and structure once it has its new OD fully staffed and operating. A key issue, as discussed elsewhere is generating researcher motivation to participate by building ownership for outreach into the research divisions.
IFPRI has recently commissioned three studies on its publications, one on research citations and two on readership surveys as described below:
Bibliometric Study of IFPRI Scientific Output: Citation Analysis Results. IFPRI recently completed a thorough study of the use of its scientific writings in English-speaking countries. Comments on the implications of this study regarding linkages between IFPRI scientists and other researchers in the field are provided in Section 2.5. As indicated, those linkages are good in the case of linkages with researchers in advanced institutions. One can thus infer that the impacts of IFPRI on the scholarly community of policy researchers are satisfactory.
However, the Panel notes that this bibliometric study of interactions was limited to the scientific output, and particularly to the journals included in the Science Citations Index and the Social Science Citations Index (SCI and SSCI). Further, the database includes few journals that publish in languages other than English and that come from the developing world. Thus, the picture that emerges provides indications of intermediate impacts in terms of use of IFPRI's scientific output mainly in developed countries, and considering a very narrow range of outlets and citation evidence.
IFPRI Readership Survey. In 1996, IFPRI commissioned a study to learn about readers' views on the Institute's publications. A four-page survey was mailed out in November 1996 to about one-third of the general mailing list and to all of the subscribers affiliated with the CGIAR, and by the time of writing the report 394 surveys had been returned (16%). Of these, the majority were from North America (34%) and Western Europe (20%), followed by Asia (17%), sub-Saharan Africa (14%) and Latin America (9%). The most regularly read IFPRI publications are Abstracts of Research Reports (65%), IFPRI Report (62%), 2020 Vision publication (56%) and Food Policy Reports (55%), The least read are the John Hopkins University Press publications (16%) The publications are found to be most useful for general interest, research and writing articles, but more than half also use the publication to prepare policy advice, formulate policy and for teaching and business purposes, 30% of respondents believe the publications play a major role and 40% a minor role in helping developing countries devise their food policy. Only a third of respondents were aware that the 'IFPRI Newsletter' and Abstracts are available in other languages. The survey found that 63 percent of the respondents use the Internet now and 24 percent plan to do so in the next few years. Yet only 27 percent were aware of IFPRI's site on the World Wide Web and only 12 percent have accessed it. IFPRI have noted that this indicates that an effort should be made to promote awareness of the Web site. The mean age of the respondents to the survey was 50 years, which could mean that younger readers are accessing IFPRI's information electronically, IFPRI noted that this needs to be confirmed, and concludes that IFPRI should develop strategies and programmes to broaden its subscriber base to include younger people who are involved in food policy development and decision-making. The Panel supports this conclusion.
A Survey of 2020 Vision Readers. This recent survey was independently carried out for IFPRI through a mail insert into the News and Views newsletter to 3,356 readers of whom 452 (13%) responded in time for inclusion. The majority of respondents were North American (32%) or Western European (20%), but 20% were from sub-Saharan Africa, and 14% Asian.
They were fairly equally divided between teachers, researchers, policy advisors and general managers working in higher education (34%), government (18%), non-government (14%) and research (13%) organizations, with 16% being affiliated with the CGIAR. As might be expected from the survey method the respondents were well aware of the publications, especially the 2020 briefs, News and Views and discussion papers. More telling is the response that they found the 2020 Vision publications to be useful, or very useful (74%), for general interest, formulating policy, research and article preparation and for stimulating dialogue within or outside their organization, and that they found the writing style and visual approach to be effective. The Panel concludes that although the audience is mainly North American and Western European, a relatively high proportion of the publications also reach readers in developing countries who find them very useful for policy and research purposes.
As mentioned above, a second way (in addition to capacity strengthening) that IFPRI can have an impact is through use of the substantive results of its research in actual policy debates and decisions. IFPRI has developed a number of anecdotal indications of its impacts on policy changes in its target countries. These indicators are limited to testimonials and correlations between IFPRI research results and policy changes. In some cases there are concrete indications that IFPRI outputs entered directly into policy debates. At the same time, IFPRI is careful not to focus on attribution of policy change to IFPRI research. That is for obvious reasons: A policymaker who makes a positive impact through policy change naturally is reluctant to give up ownership of the change by attributing it to IFPRI research. If IFPRI makes claim to the change, (i.e., suggests causal relationship between its research and the change), then it may risk reduced acceptance and future impact of its work. (As an aside, if the policy change results in a negative impact, then attribution to IFPRI obviously would be much more likely).
A case study has just been completed of the association of IFPRI research with a critical decision made by Pakistan policymakers in 1987 to eliminate the wheat flour ration shops in Pakistan. The case study illuminates the workings of the policy process, examines the role of IFPRI's research in policy-making, and suggests how IFPRI might improve the relevance and impact of its research activities. The study was conducted in collaboration with the Pakistan Institute of Development Economics (PIDE) in Islamabad, Pakistan, and funded by a grant from the Dutch Government. The study is a first step in IFPRI's broader attempt to research and understand the role information in the policy process. The study also indicates, through the input of a number of Pakistani decisionmakers, that IFPRI output and involvement played a significant role in the policy change.
Less formal assessments have indicated similar results. For example, IFPRI work in Pakistan and Bangladesh was assessed by outside agencies to have resulted in a saving of 200 million dollars in fiscal costs. The research dealt with policy changes associated with ration shops, food subsidies, credit programmes and other policies related to food and agriculture. IFPRI research in Vietnam was utilized in policy changes even before it was published. This work related to liberalization of Vietnam's export quota system. In Mozambique, results from collaborative IFPRI research have formed the basis for current restructuring of a poverty relief programme. Famine prevention research results have been used by several countries, including Ethiopia and Sudan.
IFPRI should develop further case studies correlating its work to specific policy changes, in the context of a given hypothesis on the use of information in the policy process. As a focal point on policy impact, the Panel again commends IFPRI on its policy briefs. Comments and evidence from a wide variety of sources indicate that they are widely read and appreciated, Their impact remains to be assessed.
Based on the Panel's assessment, it concludes that:
First, definitive, quantitative assessment of the impacts of policy research on improvements in the human condition (poverty alleviation, food security, environmental improvement) is difficult if not impossible to do with the present state of knowledge. Thus, IFPRI should continue as it is to focus its empirical impact assessment work on intermediate outputs, using a variety of techniques as indicated above.
Second, IFPRI is doing a good job in terms of the advancement of the field of policy research impact assessment. It should continue to work in this area, making contributions as possible to the other centres in the CGIAR System and to the newly established CGIAR Impact Assessment and Evaluation Group (IAEG).
Third, IFPRI is addressing issues that are relevant to solving the key agricultural and food problems facing developing countries. Thus, its potential for more widespread impact in these countries is significant. However, to meet this potential IFPRI needs to focus more systematically on developing a strategic framework for its outreach activity. This brings us back to the recommendation made in Section 2.5 that suggests, in general agreement with the conclusion of the ICER, that IFPRI develop a more proactive, explicit, and systematic approach to integrating capacity strengthening directly into research projects. Accomplishing this implies development within IFPRI of a clear strategy for outreach and measurement of its impacts, strong support services, including improved communication capacity, incentives for research leaders and researchers to take on board outreach functions, and a culture that treats effective information dissemination and capacity strengthening as equal objectives with research. Given the understandably strong career market orientation of most IFPRI researchers, this latter need will be a significant challenge, since an academic career focus dictates that the key output for the researcher is peer reviewed publications; and many of these will be rather distant from the priority information and capacity strengthening needs in developing countries.
The Panel recommends that with respect to Impact Assessment. IFPRI should:
a) continue to assess the impact of its activities mainly in terms of the outputs for which it can be held accountable (intermediate impacts of its publications and capacity strengthening efforts, for example);b) continue its activities related to developing improved understanding of, and means for assessing, impacts of policy and social science research; and
c) develop further in-depth case studies of its impacts in partner countries, and also surveys of the use of its outputs in other, non-partner countries.
3.4.1 The Characteristics and Motivation of IFPRI's Research Staff
3.4.2 The Constraints on Individual Researchers
3.4.3 Interactions among Researchers
IFPRI's research staff is increasingly young, of high calibre, and highly motivated (in certain directions to be discussed below).
The typical IFPRI researcher is recruited from the U.S. and other developed countries' academic market (interpreting the term to include the more research-oriented parts of the World Bank). If, for any reason, he (38 out of 52 senior research and outreach staff are men) has to leave IFPRI, he would return to this market. He would also most likely belong to the economics profession, which dominates IFPRI's staff (of the senior research staff, 90 per cent are economists). All of these are true regardless of whether he was originally from a developed or a developing country (a quarter of the senior staff are from developing countries).
With these characteristics, IFPRI researchers are driven to maximize their value in the academic marketplace by responding to its demands. In one respect, IFPRI's Management has further reinforced this responsiveness by maintaining a limited term appointment. Consequently, the output from IFPRI's research staff has been of consistently high quality (see Section 3.2). In this respect, IFPRI has fulfilled an important part of CGIAR's mission: to bring the highest quality science to work on the problems of poverty.
But scientific quality is surely not the sole criterion by which IFPRI is to be judged, for if it were, we have the universities which can provide work of the same or better quality at a cheaper cost than IFPRI. IFPRI was set up precisely because the CGIAR's stakeholders expect more from it. They expect it to have an impact on the welfare of the poor people. They expect it to spread new ideas about policies to the governments of the world, to contribute to and stimulate policy debates within many developing countries, to participate in international debates concerning global policies, and a great deal more. And they expect it to find the finances to do these wonderful things, an exercise which is not entirely without cost.
From the point of view of the individual researcher, none of these activities will earn him additional kudos in the academic marketplace. It therefore falls on IFPRI's management to fashion a system of motivation within the Institute that will guide individual researchers somewhat away from a single-minded pursuit of their academic goals.
One method open to IFPRI's management is to bring in a different mix of skills to IFPRI's research divisions. It can reduce the heavy academic presence by bringing in, for example, people with extensive exposure to the process of policy-making in developing countries. At one time, IFPRI used to have more such personnel, but now that number is reduced to a few, all of whom are quite productive in the academic sense of the term. The Panel suggests that IFPRI makes a more concerted effort to recruit such personnel who have experience in policy-making and implementation.
Another option is for IFPRI to hire more non-economists. Of the 52 senior research and outreach staff, 46 are primarily economists. The non-economists include one nutritionist, epidemiologist, rural sociologist and political scientist, and two geographers. IFPRI has attempted over the years to hire in researchers in such other non-economist disciplines. These attempts have not been trouble-free, with lack of a critical mass leading to isolation, disaffection and departure, and some divisions have switched to hiring in expertise from the universities as and when needed. The advantage of external collaborators is that they maintain their own support system in their field of specialization, but access to their advice is less flexible unless they are long-term visitors and full integration into IFPRI work is less likely. The ICER addresses this issue of increasing diversity of methods, it commends IFPRI for being firmly rooted in economics and, although it recognizes the need for greater diversity of staff, does not recommend increasing staff in other disciplines as the numbers needed to achieve a critical mass would be financially too great. It recognizes the problems IFPRI has had in the past with short-term consultancies and suggests partnerships with other high quality organizations not dominated by economists but working on similar issues. The Panel expects that IFPRI will have to experiment with different approaches before it attains a happy medium.
The first and most palpable constraint put on individual researchers is the requirement that they raise the bulk of their research funds. This certainly was not the wish of IFPRI's management, but was merely the consequence of the funding constraint faced by it, as IFPRI could obtain only half of its resources on an unrestricted basis. From the point of view of the staff, the fund-raising imposes a major burden, which in some cases is enormous - as much as a quarter or a third of their time.
Such a close relationship with the donors would make IFPRI researchers much more responsive to the donors, and therefore "relevant" (at least from the donor's point of view, if not necessarily the poor's or their governments'). There is a danger that such a decentralized system of fund-raising would lead to a highly opportunistic choice of research topics, and a gradual decline in IFPRI's status to that of a consulting house. IFPRI's management has effectively prevented such a development by requiring that all research activities in the Institute fall into one of the agreed-upon MPs or GRPs, which have undergone a thorough vetting both for content and method.
The second mechanism to influence research in the direction of the Institute's goals is the use of unrestricted funds. The Institute uses these revenues first of all to pay for the administrative overheads not covered by those charged to special projects (See Section 4.4.1). The remaining amount is then transferred to the research divisions in proportion to their shares as laid out in the Medium-Term Plan. Note that this proportionate topping up merely adds to the amount raised by each division, which amount may or may not be proportional to the priorities of the plan, depending on funding exigencies.
Once a division is allocated its share of the unrestricted funds, it then passes on these to the programme leaders, after deducting for its own administrative overheads. One division tends to allocate the amounts in proportion to the budgets shown in the MTP. The others tend to make up for the shortfalls of each programme, so as to maintain the programmatic integrity of the division - the "free-rider" disincentive problem being averted by peer pressure among a small group of about five or six persons. In both cases, the DG does not influence these division-level decisions, except to ensure that the process remains transparent.
Because the unrestricted funding available to IFPRI is not very large, the Management is quite constrained in its ability to maintain institute-wide programmatic integrity. Occasionally what is deemed very important by the Institute has gone unfunded for years. Thus far, the SMT, which oversees such decisions, has been extremely reluctant to play the gap filling role.
The third mechanism used by the Institute is the evaluation of the research personnel. At the beginning of each year, each research fellow submits his or her own work programme for that year, at the end of which a self-evaluation form is submitted (together with the following year's work programme). This procedure is used by the senior management to pass on the Institute's policies, as they pertain to the particular person. Thus, at present, the Director General is concerned about the low output of IFPRI Research Reports and has therefore let it be known that he will value the staffs contribution in that area highly in future evaluations.
IFPRI's management has thus effectively combined the use of incentives, peer pressure and negotiations to direct individual researchers somewhat away from a single-minded pursuit of academic excellence toward other goals that the Institute may deem important. The one area where researchers may be encouraged a little bit more is outreach. As discussed in Section 2.5, IFPRI needs considerable rethinking in this area. The Panel's suggestion there is toward a far more active participation of the researchers in outreach.
A productive research institute requires constant cross-fertilization of ideas among its personnel. The drive of each individual researcher to achieve academic excellence detracts somewhat from this goal, as interaction with fellow researchers take time away from their work, without clear gains. At IFPRI, there are brown-bag seminars that discuss some specialized topics among five or ten researchers, but other than this there is less of the sort of in-house seminars which discuss some broader policy issues which would bring different divisions together.
IFPRI does have policy seminars which also bring in outsiders. While such seminars may be useful as one of the many ways that IFPRI disseminates its results, they cannot be as successful in encouraging intra-IFPRI exchanges. Whatever inter-divisional cross-fertilization exists, is in the context of joint work on an MP or GRP, which as indicated in Chapter 2, is typically a collaboration among more than one division. Another very useful device is the annual Internal Program Review which has been an important means of communication within IFPRI.
The ideal that should be sought after in an institute the size of IFPRI is that every research fellow has a clear understanding of what other research fellows are working on. It is not our impression that this ideal has been attained or has neared attainment. This is surprising as IFPRI's intellectual climate is dominated by economists, who comprise the overwhelming majority of the research staff. With such near uniformity of backgrounds, the low level of staff interaction is hard to explain.
3.5.1 IFPRI Linkages with Other CGIAR Centres: Systemwide and Other Joint Activities
3.5.2 Systemwide Activities
3.5.3 Other Inter-Centre Joint Activities
3.5.4 IFPRI Linkages with Non-CGIAR Advanced Institutions
3.5.5 IFPRI Linkages with Institutions and Individuals in Developing Countries
As mentioned in Chapter 1, one of the concerns that led to the Stripe Study on Policy Research in the System was the issue of IFPRI's linkages with other centres in the System. Some considered IFPRI as a "service" centre for the rest of the System. The Stripe Study identified two issues that it felt the IFPRI EPMR should address: Synergy that could be derived from inter-centre projects; and the role of IFPRI in ecoregional initiatives.
Additional questions have arisen since the Stripe Study concerning:
(i) the strength of IFPRI's linkages upstream with advanced policy research institutions relative to its linkages downstream with the more applied, developing country policy research and analysis groups; and(ii) IFPRI's participation in Systemwide Activities. These various issues are addressed below.
At this point, it suffices to say that in general the Panel is satisfied that:
· IFPRI's linkages with other CGIAR Centres, both in terms of formal and informal linkages are good. It is one of the most active collaborators in the System;· IFPRI's links to advanced institutions are excellent. It works closely and smoothly with a great variety of organizations, mainly in North America, but also in other parts of the world. IFPRI's work is widely cited in scholarly journals.
· IFPRI's linkages with less developed country policy research and analysis institutions are adequate in a relative sense, but could be strengthened, as in fact is the intention of IFPRI during the current MTP period.
A particular problem faced by IFPRI in the past, and that was flagged in the Stripe Study on Policy and Management Research, was the perception of other CGIAR Centres that IFPRI should serve in an advisory, and in some instances service role to the other centres. As an example, IFPRI received far more requests than could reasonably be handled from other CGIAR Centres to work with them, particularly in developing their ecoregional and other Systemwide policy-related programmes. It was necessary to develop a set of research priorities that best enabled IFPRI to meet its own MTP objectives, and also to only link up when funding and staff resources permitted. The most difficult cases are Systemwide activities that demanded frequent participation in planning meetings, but which never actually led to new sources of funding or to new research.
While this confusion existed a few years ago, other Centres now appear to accept the fact that IFPRI is not a service centre, and that IFPRI has a legitimate research mandate of its own that it must pursue. Probably, the Stripe Study helped to clarify the role of IFPRI in the System. IFPRI staff are less frustrated now than they were when the Stripe Study was initiated, since the Institution has clarified the conditions under which inter-centre activity will be initiated and terminated.
At the same time, the Panel points out that IFPRI probably is the most active CGIAR Centre in terms of inter-centre activity, since it has working relations with all the centres. Some of this interaction is informal, as in the case of an agricultural policy research priority-setting exercise in Malawi convened by IFPRI and its Malawi partners and involving some six other CGIAR Centres working in a collaborative mode to identify the particular policy related topics on which they had expertise and interest. Other interaction is formal, for example through shared staff appointments (e.g., with IIMI, ILRI, and ICARDA) and through participation in Systemwide activities.
Within the CG System, both IFPRI and ISNAR are the two centres who specialize in the areas of policy and management research. There are collaborative projects between them, but also some potential for overlaps. In a world in which the policy terrain is truly vast, and the resources and staff of any one Centre are small, some degree of duplication of efforts must be tolerated, and may even be desirable, as indeed the work of IFPRI and FAO on similar issues has shown. The recent Strategic Plan of ISNAR will provide the two centres an occasion to assess whether there is in fact more than a marginal overlap between the research agenda of the two centres. Some of the titles of research programmes look similar. On the other hand, the target audience of ISNAR is NARS, whereas IFPRI works primarily with economic research institutes, universities and ministries in developing countries. The range of issues and their implications for these audiences could be different even when the themes look similar. Thus, for example, ISNAR may look at the implications of environmental policies for the management tasks of NARS, IFPRI may assess the implications for the concerned ministries (Agriculture, planning, etc). The Panel did not have access to the needed documents to probe this matter further. It is an issue for the IFPRI and ISNAR Boards to examine.
At present, IFPRI is a partner in seven Systemwide programmes and convenor of two of these. These include:
· The Property Rights and Collective Action Systemwide Programme, convened by IFPRI, began in 1995. The goal of the programme is to contribute to policies and practices that alleviate rural poverty by analyzing and disseminating knowledge on the ways that property rights and collective action institutions influence the efficiency, equity, and sustainability of natural resource use. In addition to the 15 other CGIAR Centres that are participating, membership includes researchers at NARS, NGOs, and universities. Much of IFPRI's work in MP11, Property Rights and Collective Action in Natural Resource Management, will be carried out in conjunction with this Systemwide programme. One of the innovative activities initiated by IFPRI is the interactive network related to the programme. As indicated in the discussion of the EPTD, a number of publications are coming out of this activity already.· The Agricultural Research Indicators Systemwide Initiative, co-convened by IFPRI and ISNAR, the objective of this initiative is to keep national, regional, and international decision-makers fully apprised of the global state of agricultural research and development investments and provide a periodic series of policy assessment on these trends. These activities are linked to IFPRI's work in Agricultural Research, Extension, and Education Policy, GRP1. The initiative has not been funded at a level that would permit much activity. IFPRI and ISNAR are, therefore, no longer pursuing it as a Systemwide initiative, but have consolidated it into a normal collaborative research project between both Centres. The Panel believes that this is an activity that should be more centrally funded by the System, since it is producing information that will be useful across the System, and more to some other centres than to IFPRI.
· Alternatives to Slash-and-Burn Agriculture Systemwide Programme (ASB) is a worldwide research and development project to reduce tropical deforestation and promote rehabilitation of degraded land by providing farmers who practice slash-and-bum agriculture with alternative ways of using the land. IFPRI will continue to provide socioeconomic and policy guidance, particularly through its work in Brazil (MP8), in 1998-2000. IFPRI collaborates with other ASB research efforts that are under way in Indonesia and Cameroon.
· Water Management Systemwide Programme, convened by IIMI, is designed to increase the productivity of water in agriculture in an environment of growing scarcity and competition. IFPRI leads two of the four components (multiple users and river basin allocation issues). Up to now, this has not been an active initiative due to lack of funding, although through the EPTD there has been associated joint activity with IIMI, including a shared staff appointment. The water area in general is one in which the EPMR Panel believes there is opportunity for considerable expansion in work.
· Livestock Research Programme works to improve the sustainable production and utilization of feed resources. IFPRI will continue to participate by providing policy linkages to livestock issues. If the programme expands into dairy issues during the next MTP, IFPRI will increase its involvement through its work under MP7. IFPRI has a shared staff appointment with ILRI.
· Systemwide Genetic Resources Programme, formed in 1994, contributes to the global effort to conserve agricultural, forest, and aquatic genetic resources and promotes their use in ways that are consistent with the Convention on Biological Diversity. IFPRI is not very active in this activity at present, although the intention to become more active in it during the coming years. IFPRI is expected to provide the Systemwide programme support on policy and socio-economic issues in genetic resources conservation and use, particularly on intellectual property rights. These activities are linked to IFPRI's research work in agricultural research policy, GRP1.
· Desert Margins Initiative was started to increase the food security of poor, rural populations and contribute to poverty alleviation by halting or reversing desertification. IFPRI, one of eight CGIAR Centres involved in the initiative, provides research on policies and institutional options for improved natural resource management. This initiative is linked to IFPRI's work in less-favoured lands, GRP5.
In addition to the Systemwide programmes, IFPRI is involved in several of the ecoregional activities in the System, e.g., the Central American Hillsides, and the East African Highlands.
From 1992 to 1997, IFPRI has established numerous joint research activities with other CG Centres, involving joint fund raising and the hiring of jointly appointed staff in complementary to its involvement in the Systemwide programmes. For example, in the case of the EPTD, these include collaboration
(i) with ICARDA on the development of sustainable intensification strategies for the low-rainfall areas of North Africa and West Asia, with two jointly appointed Post-Doctoral Fellows based in Aleppo;(ii) with ILRI on analysis of the links between property rights, risk and livestock production systems in Africa, with a jointly appointed Post-Doctoral Fellow based in Nairobi;
(iii) with ILRI on the development of sustainable intensification strategies for the East African Highlands, with a Post-Doctoral Fellow to be recruited later this year for placement in Ethiopia;
(iv) with CIAT on development of GIS methods and databases for impact analysis of past agricultural research in Latin America and for priority-setting for future agricultural research, with a jointly appointed senior research scientist based in Cali; and
(v) with IIMI on water management policies, with a jointly appointed senior researcher based in Washington. There are also joint projects with other Centres which do not involve jointly appointed staff.
The Panel commends IFPRI on its Inter-Centre activities, which range from intense collaboration to policy support in areas where IL. RI also has research ongoing or completed.
At the same time, the Panel cautions IFPRI on expanding such involvement, since the transactions costs can be substantial,
IFPRI has good linkages with the policy research community in advanced institutions, both governmental and non-governmental ones. A recently completed citation analysis of IFPRI's work in relation to that of other similar institutions places IFPRI solidly in the middle in terms of output and citations by others in the scholarly journals in developed countries. Thus, the Panel concludes that IFPRI is doing well in this area. During the past five years, IFPRI staff have published more articles in peer reviewed scholarly journals than during any other period in the organization's history. Analysis of co-authorship - an indication of communication and linkage with other scholars - indicates that IFPRI is actively involved with leading scientists in the policy analysis research community. In the case of research reports, the links are strongest with researchers from developing countries. In the case of peer reviewed scholarly journals listed in the International Science Index, the linkage is with World Bank and top echelons of U.S. academic research institutions in the field. The Panel notes that this fits well with the dual roles of IFPRI as a member of the policy research community of scholars as well as an active participant in the application-oriented policy communities, primarily in developing countries. IFPRI serves the role of a "knowledge intermediary" between the scholarly community and policy clienteles.
IFPRI has relatively good links with policy researchers and their institutions in developing countries. The dilemma in many countries is the scarcity of such institutions and trained researchers. It is with that in mind that IFPRI has increased its focus on capacity strengthening and training, including longer term presence in key partner countries.
At present IFPRI has some 125 individual and institutional collaborators, with considerably more than half of these in developing countries. Over the 1998-2000 MTP period, IFPRI plans to increase this number to more that 250, with over 70 percent being with developing country entities.
In an attempt to gain some further insight on IFPRI collaboration with developing country researchers, the EPMR Panel gathered data on field projects and the production of working papers and other in country documents prepared under the project auspices by developing country researchers. This review led the Panel to the following conclusions: under the surface of IFPRI's scholarly output in formal publications - which quite naturally is dominated by IFPRI staff authorship - lies a large volume of country level working papers, briefs, theses, and other outputs, a majority of which are authored or co-authored by researchers from the partner countries. These provide evidence that complementary to the scholarly publications that are necessary to maintain IFPRI's credibility and links with the top minds in the policy research field, there is a strong IFPRI linkage to developing country policy researchers, most of them being formally or informally involved in the capacity building activities undertaken by IFPRI research divisions.
The Panel believes that IFPRI is active with, and is providing benefits to the policy research communities both in its partner countries and in other developing countries. However, at the same time, the Panel urges IFPRI to intensify its efforts to improve its linkages with developing country researchers and to carry it on to the stage of involving more developing country researchers as co-authors in its own formal publication series. The benefit can be two ways: relatively less experienced developing country researchers can benefit from the experience such exacting activity involves, while IFPRI researchers can benefit from the special knowledge held by those from the countries with which the research deals. Thus, IFPRI's twin objectives of improved outreach and integration of such with research can benefit.
Improved capacity building depends naturally on more than jointly authoring papers and publications. An example of good collaborative links with national researchers and institutions in developing countries is that of the EPTD, which defines its working style in terms of such links. A good case is the Arresting Deforestation (MP-8) activities in Brazil, where the collaboration with EMBRAPA is sufficiently strong that it is helping to pay for the research, in addition to providing office space, vehicles and administrative support at the Amazon field site. Other examples are IFPRI's activities in Vietnam and Bangladesh which are being financed partly by the Governments of these countries.
