Projections of Food Demand and Supply to 2010 and 2020
Implications of FAO's 2010 and IFPRI's 2020 Studies for CGIAR Priorities
National Research Systems
Gender Issues and Institution Strengthening
Poverty Issues
Advanced Institutions
Natural Resources (Land and Water)
20. The Chair opened this item by noting that at TAC 66 the Committee had recommended that in considering future priorities and strategies for the CGIAR, specific attention should be given at the present meeting to current projections on future food demand and supply, poverty issues, strength of NARS, the role of advanced research institutions, resource management research, gender issues, and institution strengthening with inputs from Centres and outside participants.
21. The session began with a retrospective look at the framework used by TAC for the 1992 priority setting exercise. The intent was to examine the extent to which the parameters used in 1992 were relevant to the current exercise in light of past trends and projections of the global food situation toward 2010/2020 with a view to possibly incorporating new parameters. Dr. Gryseels began the discussion by presenting an overview of the 1992 framework. A formal analytical/quantitative model comprised of three dimensions was employed. These were: (1) activities (resource management, breeding and germplasm enhancement, policy and management, production systems, institution building); (2) spatial (regions and agroecological zones); and (3) products (four major production sectors - crop, livestock, fisheries, forestry and their corresponding commodities). An analysis was done for 45 commodities in developing countries spread over nine agroecological zones and four regions. The unit of analysis was the agroecological zone, regionally defined. The timeframe was 1991-2010 using FAO projections. The quantitative analysis started with a baseline composed of value of production, number of poor people, and land in use. The baseline was then weighted using a number of modifiers (yield gap, malnutrition, GDP per capita, need for production growth, deforestation, number of scientists, soil degradation, small countries, food imports). The calculation yielded a modified baseline giving greater weight to Africa and less to Asia. A prioritized list of commodities was also calculated based on value of production adjusted by value of the baseline and of the modifiers. The analysis yielded a ranked priority index, which was then compared with actual resource allocations. This served as a basis for discussion of priorities and, in turn, yielded TAC recommendations on priorities for regions, commodities and activities as a framework for further recommendations on resource allocation.
22. Recently, the TAC Secretariat had recalculated the commodity priority index using alternative sets of developing countries, specifically, those with annual per capita incomes of less than US$ 1,000 and those considered the 39 poorest countries by the United Nations. Calculations for these two sets of countries were then compared to the 1992 index and implications were drawn for changes in priorities of specific regions. By giving more weight to the poverty dimension, these more recent calculations resulted in substantial increases in the priority indexes of Africa and Asia and substantial decreases in those of Latin America and WANA. The purpose of this exercise was not to arrive at specific recommendations, but rather to show the possible implications for CGIAR priorities of introducing more explicitly the poverty alleviation dimension. Similarly, limiting the analysis to the poorest 39 countries significantly affected the priority rankings of specific commodities. Gainers included cassava, livestock, banana, bean, and sorghum; losers included potato, wheat, and maize. Dr. Gryseels stressed that these analyses related only to CGIAR commodities and regions, and not to activities such as policy and management research. He closed by saying that the Secretariat would welcome TAC's comments and guidance for the ongoing analyses which it was conducting in support of TAC's current priority setting exercise. The task before TAC was to decide upon a methodology for recommending priorities to the Group. This might involve such considerations as updating the projections from the last exercise, conceptualizing the linkage between productivity improvement and poverty alleviation, identifying alternative sources of research supply, factoring in research on natural resources management, and incorporating gender analysis into the overall research programme. Finally, it would be necessary to link the priorities to resource allocation in a transparent way based upon interaction with all concerned stakeholders.
23. The Chair brought this discussion to a close, remarking that introducing the poverty variable could have a dramatic effect on priorities with potentially large implications for the allocation of System resources. He also alluded to the extensive interaction with various groups which TAC undertook in the 1992 exercise with a view to improving the analysis, and expressed TAC's intention to consult similarly during the present exercise. He then called on Dr. Nikos Alexandratos, Chief of FAO's Global Perspectives Studies Unit, to present FAO's global food projections for 2010.
24. Dr. Alexandratos based his presentation on the recently published FAO study, 'World Agriculture: Towards 2010'. The study documented the recent and projected slowdown in the overall rate of growth of world agricultural production, as well as a continued decline in the per capita rate of growth of agricultural production. The latter trend has caused concern among some analysts as to whether major food shortages will be a global prospect within the next 15 years. Most of the recent and projected decline in the per capita rate of agricultural growth was attributable to slower growth in the developed countries which, in mm, was related to expected low rates of population growth and policy reforms to reduce large food stocks and gear agricultural production more closely to market demand. For developing countries, however, there had been no pronounced trend toward a decline in per capita production except in the cereals sector, which in the last decade had experienced a sharp reduction in the per capita rate of growth which is now essentially zero. The decline in cereals has been compensated for by rapid per capita growth in the rate of livestock production. The question then was whether the declining trend for cereals should be taken as a sign of alarm or as a reflection of diversification of production out of cereals spurred by economic growth in certain developing countries. Dr. Alexandratos felt that this question should be an important dimension of TAC's deliberations over priorities.
25. Population growth rates have been declining since the 1950s and annual increments to world population, while at an all time high of 90 million, were about to begin to decline very slowly. In this context, economic growth rates over the next decade were expected to grow strongly in East Asia, with a slight improvement for South Asia and low rates of growth for all other developing countries, particularly sub-Saharan Africa and WANA. These trends in food production, population growth, and GDP per capita provided the backdrop for FAO's projections of food availability over the next decade and a half. Viewed in terms of calories per capita per day, individual levels of consumption in developed and developing countries were projected to vary by some 500 calories, a difference which Dr. Alexandratos characterized as enormous. His analysis also showed that sub-Saharan Africa will experience virtually no improvement in daily caloric intake per capita, which is to say that the region will not solve its food problem by the year 2010.
26. Cereals yields in developing countries (wheat, rice, and maize) were projected to continue to grow toward 2010, but more slowly than in the past. This production outcome for developing countries would be more or less sufficient to meet the growth of their own demand as determined by their rates of population and economic growth. However, their need for further net imports of cereals from developed countries will continue and was projected to rise from the current 90 million tons per year to approximately 160 million by 2010. Developed country cereals exporters were likely to continue to be able to meet projected developing country demand. However, financing these levels of imports was likely to pose a serious problem for sub-Saharan Africa.
27. IFPRI's global food projections to 2020 and their implications for investment were presented by IFPRI Research Fellow, Dr. Mark Rosegrant. The projections were essentially in agreement with those of FAO. The aggregate food supply/demand situation globally was favourable. Food production will grow fast enough to cause prices to fall but despite the world's productive capacity to meet effective demand, there will not be much improvement in food security in certain regions of the developing world, particularly sub-Saharan Africa and South Asia. Net imports of cereals by developing countries will increase to 185 million tons by 2020. Sub-Saharan Africa's share of these net imports will rise from nine million metric tons today to 27 million in 2020, a level considered unsustainable through domestic resources and requiring large increases in financing or food aid from developed countries. Small progress will be made in the developing world in reducing the absolute number of malnourished pre-school children from the present 184 million to 150 million. In sub-Saharan Africa there will actually be large increases in the number of malnourished children. Like the FAO study, IFPRI predicted that there was little scope for increases in agricultural production through expansion of areas under cultivation, except for certain sub-Saharan African countries. Dr. Rosegrant acknowledged this was a controversial prediction because some analysts felt the predicted rates of expansion would cause rapid degradation of the land and therefore may not be sustainable.
28. On the demand side, there was a very rapid structural shift underway in demand for food in fast growing areas, especially East Asia, resulting from rapid urbanization, changing tastes and preferences, and rising incomes which were causing a shift to more diversified diets (from cereals to livestock products, fruits, and vegetables.) However, the most rapid growth in demand for food will be in areas that were not expected to grow economically, but which had very high population growth rates. These included sub-Saharan Africa, WANA, and South Asia. In absolute terms, between 1990 and 2020 effective global demand for food could only be met by increases of 62 percent more rice production, 58 percent more wheat, and 55 percent more maize. But even these increases would not lead to food security and adequate diets in all regions.
29. In the ensuing discussion, TAC Members and other participants posed questions to the two presenters focusing primarily on the research and policy implications of supply/demand projections for various commodities, the interactions between agricultural production and natural resources degradation, the relationship between global and regional level projections, the feasibility of basing research resource allocations on highly aggregated food production/consumption data, and the potential effects of agricultural research in alleviating poverty by promoting economic growth.
30. A question of particular concern centred on the paradox of ample projected food availability at the global level in the context of rising demand for cereals in sub-Saharan Africa and South Asia. In addition to the issue of financing increased net imports, the problem of increasing domestic production in the context of an overall supply situation which appeared to be sufficient seemed problematic. Agricultural research breakthroughs might further lower world food prices for cereals. While this would help poor consumers, particularly in urban areas, it could have negative effects on poor farmers by decreasing their incomes. In a more general sense, however, it was argued that the growth linkage effects of increased agricultural productivity were generally associated with accelerating growth in the rural non-agricultural economy, particularly in countries where agriculture was the largest sector. But for purposes of priority setting, one would need country level studies to confirm the autonomous impact of research on food security compared to alternative investments in other sectors. Furthermore, it was argued that in considering the potential effects of research on increasing agricultural productivity growth, there would be gradual exhaustion of gains from conventional breeding early in the next century; further yield growth would be generated as conventional breeding was combined with wide-crossing, transgenic breeding, and other tools resulting from biotechnology research.
31. By way of trying to clarify the importance of this discussion for TAC's deliberations, one Committee member suggested that it might be helpful to think of the CGIAR's priorities for resource allocation in terms of three complementary dimensions: (1) increased food production, and hence consumption, through commodity research; (2) increased agricultural production in areas where the rural poor were concentrated; (3) development of the agricultural sector and promotion of non-agricultural rural employment through so-called growth linkage effects. These three dimensions were said to be complementary rather than mutually exclusive, but each required that priorities for resource allocation be tailored to different objectives. This was a complicated issue that warranted TAC's attention.
32. There was considerable debate on the question of the potential for increased agricultural production through area expansion, particularly in marginal zones where lands may become quickly degraded and the costs of restoring or rehabilitating them may be too large a constraint to consider them a notable source of growth. This issue was felt to be particularly important in sub-Saharan Africa. It was acknowledged that data constraints made it virtually impossible to quantify the extent of land degradation and its impact on agricultural productivity at the global level. Given variations in natural resources endowment by country, the need for micro level research on this question was emphasized.
33. A presentation complementing the IFPRI 2020 projections was delivered by IFPRI's Director General, Dr. Per Pinstrup-Andersen. He offered for TAC's consideration a six-point plan of recommended action which followed from the recent IFPRI conference, 'A 2020 Vision for Food, Agriculture, and the Environment'. The actions included strengthening the capacity of developing country governments to perform their appropriate functions; enhancing productivity, health, and nutrition of low-income people and increasing their access to employment and productive assets; strengthening agricultural research and extension systems in and for developing countries; promoting sustainable agricultural intensification and sound management of natural resources, with increased emphasis on areas with fragile soils, limited rainfall, and widespread poverty; developing effective, efficient, and low-cost agricultural input and output markets; and expanding international assistance and improving its efficiency.
34. The presentation prompted debate on two issues of particular relevance to TAC's priority setting: whether there was sufficient justification for the CGIAR System to begin investing further in research on postharvest technology, and whether it would be cost-effective to invest in agricultural development in marginal environments. Neither question was resolved, but it was observed that the effects of research on postharvest activities were problematic in light of poor marketing, transport, processing and storage infrastructure in many developing countries - all of which required large investments. While one TAC Member argued that the recommendation on investing in marginal environments should be heavily qualified, it was countered that if most of the developing world's poverty and environmental degradation were concentrated in these areas, the problem could not be ignored by the CGIAR. The only alternative to promoting agricultural growth in these areas was migration to high potential zones or urban areas.
35. The first presentation on priority setting for National Agricultural Research Systems (NARS) was made by Dr. Cyrus Ndiritu, Director, Kenya Agricultural Research Institute. Dr. Ndiritu began by noting that national systems typically faced the dilemma of setting priorities in a context of scarce resources and changing policy environments, factors which often diverted them from their mandates and weakened their capacity to implement the activities which a rational calculus would indicate were most important. Thus, in selecting priorities NARS had to take into account not only such variables as the value of production and the potential for research impact on increasing production and sustaining food security, but sectoral and/or macroeconomic policies that impinge on farmers' incentives to grow food. NARS generally did not have the human and financial resources to address all of these considerations. Dr. Ndiritu illustrated the complexities involved in NARS priority setting by citing specific examples of the effects of structural adjustment programmes and the actions of parastatals that had vitiated the beneficial effects of national research activities by creating disincentives for farmers to adopt technologies that had been successfully developed by the national programmes. In the case of his own organization's priorities, he emphasized that food security was conceptualized in broader terms than increasing production through improved technology. It was also necessary to address the potential contribution agricultural research could make to generating rural employment, particularly through the development of agribusiness. He stressed, however, that this was not simply a technological problem for research, but involved as well understanding the trade policy environment within which agribusiness could become viable.
36. The second presentation for this item was made by Dr. Bie. His remarks focused primarily on the implications for TAC priority setting of the need to translate technological packages now available to narrow the yield gap into improved productivity for poor farmers. The role of NARS, with assistance from the international centres and others, was seen as key to this process. Dr. Bie began by emphasizing the need for continued public support by developing country governments for their NARS, in particular so that they would remain competitive with other sectors and possess the requisite scientific and management skills and infrastructure needed to perform effectively. Given their role in generating public goods for agriculture, the private sector was unlikely to assume their functions.
37. Dr. Bie identified three ways in which the CGIAR could help to strengthen NARS. First, it would be useful to treat them as part of the larger structure of institutions, including universities, that created intellectual outputs relevant to improving agricultural production. Neglect of these institutions by developing country governments in recent years meant that the multidisciplinary approaches needed to develop technologies and policies to help to close the yield gap remained to be developed. The CGIAR could contribute to strengthening these institutions, particularly in concert with Advanced Research Organizations (AROs) of the developed countries. The Centres could play a pivotal role in transmitting to NARS the latest science from the AROs and relaying back from NARS the research problems which warranted the attention of the advanced institutions. Secondly, the last few years have seen increased South-South collaboration among NARS, a development which FAO supported, as well as increased regional integration of NARS through various entities. The CGIAR had an opportunity to strengthen the national programmes by collaborating with these regional groupings through its ecoregional approaches. Finally, the Centres could become more involved in assisting NARS to engage in formal priority setting, with ISNAR playing a key role. Dr. Bie also identified a number of potential constraints to strengthening IARC-NARS collaboration. Some developing country governments did not share the CGIAR's goal of alleviating poverty and, instead, would give priority to the production of cash or export crops. TAC and the Centres should help NARS to resist such pressures.
38. The third presentation for this session was delivered by ISNAR's Director-General, Dr. Christian Bonte-Friedheim. He emphasized that the CGIAR's priorities vis-à-vis the NARS needed to be adapted to the latter's growing strength and heterogeneity. Because NARS had substantial differences in resource endowments, political contexts, and internal and external relations, the Group and the Centres needed to be more closely attuned to their specific needs rather than taking a global approach. Perhaps the most important new dimension in the CGIAR's relationship with NARS was the latter's expression of a desire to be partners in priority setting and, in that sense, participate in ownership of fee System's objectives. Another manifestation of this was the desire of stronger NARS to bid jointly with Centres for projects and to implement them as equal partners. They had also begun to relate directly to Advanced Research Organizations in both the public and private sectors. These developments should attract TAC's attention as it set System priorities and attempted to determine the most appropriate role for Centres in the newly emerging context. For example, the growth of regional cooperation among NARS was a healthy sign, but prerequisite to effective cooperation with the CGIAR at this level was the creation of national level fora in which the various elements of the national systems could be represented. The CGIAR had an important role to play in this regard.
39. Following these presentations the Chair invited comments from TAC Members and other participants. The ensuing discussion addressed such questions as the sources of change in NARS' needs and outlooks; the appropriate division of labour between the international centres, advanced institutions, and the national programmes; how best to relate to NARS in a regional context; how to involve NARS in the CGIAR's thrusts in the areas of poverty alleviation and natural resources management; and how to involve NARS in TAC's priority setting.
40. On the question of the antecedents of NARS' new orientation, it was noted that in the last five to ten years there had been a significant increase in the number of NARS researchers holding M.Sc. and Ph.D degrees (although this appeared to be less true for the social sciences and policy research). Thus, there were more scientists to work with, but at a time when reductions in public funding for agricultural research meant that NARS scientists lacked the appropriate equipment and infrastructure needed for effective collaboration. Concern was expressed that this large investment in NARS human resource development was being underutilized, yet there was substantially increased demand for national agricultural research compared to five years ago. A perhaps related development was the rapid privatization of some NARS activities, at least in Latin America.
41. There was considerable debate on the issue of devolution to NARS of research activities carried out by the Centres. In the days of traditional commodity research this was treated as a matter of comparative advantage. Today, it was complicated by the research implications of the CGIAR's priorities on natural resources management and poverty alleviation which were not necessarily shared by developing countries, and by domestic political pressures for NARS to conduct research in support of agricultural commercialization. These political complexities were matched by technical problems involved in conducting natural resources management research across broad ecoregional zones lying beyond the scope of NARS national mandates. Another concern was that traditionally devolution implied a reallocation of financial resources from the Centres to NARS. But in the framework of the new CGIAR funding mechanism, devolved research might no longer be considered an international public good eligible for CGIAR resources, raising the question as to the future stability of the research. In the past, devolved research had sometimes resulted in losses in the stock of knowledge when changes had subsequently occurred in the strength of NARS. Finally, even assuming continued strength, there remained the issue of the international mobility of genetic resources' responsibility for a particular crop devolved to the national level.
42. These complexities clearly implied that there was need for greater consultation between the CGIAR and NARS on priority setting. Various views were expressed on this issue. From the standpoint of process, NARS apparently preferred to interact with TAC in the context of their various regional groupings. However, it was argued that until agricultural research strategies were in place at national levels, a regional framework for interaction would be problematic given the heterogeneity of individual NARS. Although TAC did not reach consensus on this issue, the Chair indicated that the Committee shared many of the concerns expressed by participants. He reminded the members that CGIAR guidelines on IARC-NARS interaction on research itself were premised on the principles of comparative advantage and the production of international public goods. He considered this framework a viable one for pursuing this issue in the current round of priority setting. He expressed the sentiment of TAC Members that there should be greater interaction with NARS on CGIAR priorities and strategies. While the modalities of this interaction were not identified, the Chair indicated that TAC would share its analyses with NARS as the process moved forward.
43. The Chair invited Dr. Hilary Feldstein, Programme Leader of the CGIAR's Gender Programme, to make the first of two presentations on the implications for the CGIAR's priorities and strategies of incorporating gender issues in its research agenda.
44. Dr. Feldstein presented the case for integrating gender analysis into CGIAR-sponsored research in terms of the increased knowledge it would generate of key aspects of agricultural production that were women-specific or of policies that limited women's productivity. Since gender-based research was often location-specific, the use of macrolevel indicators such as female illiteracy rates or distribution of female heads of households to identify gender-related production or policy constraints were problematic. Instead, Dr. Feldstein proposed a "user perspective" to research on such factors as the influence of women's preferences on choice or adoption of technology or women-specific constraints which could be addressed by policy research, such as property rights and access to credit, inputs, and infrastructure.
45. In conducting such research, she proposed a division of labour between IARCs and NARS, whereby the former would undertake strategic and applied research having an international character and the latter applied and adaptive research. The most important task for the CGIAR was to give priority to the development of methodologies that would further strengthen use of a gender and user perspective for research below the national level, but which had international spillover effects. This was particularly important given the microlevel nature of such research. Among the topics which Dr. Feldstein identified as women-specific were postharvest storage and processing, seed storage and selection, germplasm collection and in situ conservation, and fuelwood and water for domestic use. Given the uneven exposure of IARCs and NARS to gender analysis, it was recommended that priority be given to a review of current CGIAR activities in this area, methodological development, training, and communication.
46. The second presentation on this item was given by Dr. Bie. He began by noting that while the member states of FAO, for cultural and religious reasons, had diverse views on gender as a development issue, the Organization had taken a proactive position and created a division for Women in Development. Its purpose was to ensure that the gender dimension would be incorporated into FAO projects. The kinds of indicators FAO used in this work included male/female child ratios at one year, female literacy, percentage of girls with more than six years of schooling, and tertiary education of women.
47. Dr. Bie endorsed the topics for research proposed by Dr. Feldstein and suggested, in addition, the following: aspects of mechanization in agriculture having a gender dimension; byproducts of plants having informal and not only commercial uses; the impact of the AIDS pandemic on family structures and women's roles in agricultural production and the transmission of agricultural knowledge to children; and the need to increase the number of female scientists in NARS.
48. In closing, Dr. Bie conveyed the sentiment of FAO's Director General on the importance which the Organization attached to the women's issue and its desire to work closely with the TAC and CGIAR on this issue.
49. The Chair then invited TAC's comments on the two presentations. These focused on the implications of some of the research methodologies and topics proposed by the speakers for TAC's priority setting.
50. In terms of research foci, it was suggested that research on small ruminants, milk production, agroforestry, and female illiteracy were particularly important for sub-Saharan Africa. While some of these topics were beyond the mandate of the CGIAR, it was suggested that the Centres might link up with other organizations whose work on these issues could be integrated with their own. Discussion of the relative utility of female-headed households and illiteracy rates as indicators of poverty highlighted the location-specificity of the gender issue. Caution was urged on the allocation of CGIAR resources to the study of such global indicators. The implications of potential demographic changes resulting from the AIDS pandemic were explored in terms of their potential effects on agricultural production from a reduction in the length of production years of populations and from erosion of the agricultural knowledge base due to disruption in the transmission of technical expertise from parents to children. The latter had important implications for research, extension, and training to strengthen labour input.
51. The question was raised as to how the international dimension of gender-based agricultural research could be articulated for purposes of priority setting, given its location-specificity. This was said to be a difficult issue to deal with because of cultural, biological, and agroecological variations at study sites. Since the external validity of individual cases would be problematic, it may be necessary to undertake comparative analyses to identify trends and patterns across countries and extrapolate the implications for national and international policies. Given the uneven capacity of CGIAR Centres to undertake gender analysis, it was suggested that TAC might wish to recommend to the Group the analytical tools needed to do the job, e.g., the participatory methods used in Rapid Rural Appraisal. Finally, in response to a question as to whether the rate of return to agricultural research was higher when a gender dimension was included, it was stated that there were currently no findings across projects that would confirm such a hypothesis. However, anecdotal evidence on the beneficial effects of the user perspective on, e.g., technology adoption, was suggestive of the possibilities.
52. In closing this discussion, the Chair highlighted the recurrent theme of the need for research on postharvest technology which has emerged in other contexts of TAC's consideration of priorities.
53. The Chair called upon Dr. Gaurav Datt of the World Bank to open the discussion on poverty as an issue in TAC's priority setting.
54. Dr. Datt based his discussion on the World Bank's work in the area of global poverty monitoring, focusing on what was known about global poverty and the implications for alleviation targeting. He began with a methodological caveat that differences in the standards or indicators used to measure poverty made international comparisons problematic. Any priority setting exercise based on such measures should take this caveat into account, although greater confidence could be placed in regional estimates than international ones.
55. The incidence of poverty at the global level based on a US$ 1 per person per day poverty line was estimated by the World Bank at 34 percent, but there was considerable diversity at the regional level: South Asia (59 percent), sub-Saharan Africa (55 percent), Latin America/Caribbean (28 percent). East Asia (15 percent), and Middle East/North Africa (5 percent). While alternative poverty measures resulted in slightly different distributions - notably, a rank-reversal between South Asia and sub-Saharan Africa - from a poverty alleviation perspective, the regional priorities implied in the above distribution were said to be fairly straightforward. By comparison, the CGIAR allocated its 1993 budget regionally per thousand population as follows: sub-Saharan Africa (US$ 255), West Asia/North Africa (US$ 139), Latin America/Caribbean (US$ 112), and Asia (US$ 42). Dr. Datt concluded that while the relatively high priority accorded to SSA was consonant with the World Bank's analysis, there appeared to be some scope for reallocation of resources to Asia, particularly South Asia, which had the lowest per capita allocation.
56. The second presentation on this topic was given by Dr. Shiv Nath Saigal of the International Fund for Agricultural Development. He described IFAD's poverty alleviation programmes as being targeted to specific population groups - e.g., poor farmers, landless labourers, women, indigenous peoples - rather than to regions. To the extent that these programmes had an area focus, it was based on the criterion of resource endowment which was seen as major cause of poverty. Thus, emphasis was given to less endowed areas, particularly in South Asia and sub-Saharan Africa, although an attempt was made to maintain a balance between investment in high and low potential zones. IFAD's strategy was to promote labour-intensive forms of technology generation, including use of traditional practices, that increased yields of crops typically grown by poor farmers with a view to directly alleviating poverty at the farm and household levels. It had evolved techniques to identify and reach the poor directly and developed location-specific activities and institutions to achieve these objectives. The poor were identified through diagnostic methods that tap into the perceptions of the poor themselves, rather than through statistical surveys. A number of insights had been gained using these procedures: the poor were not homogeneous, but needed differentiated treatment; solutions to poverty were location-specific; vulnerability mapping was an essential tool to identify where the poor were located.
57. Drawing out the implications of IFAD's experience for TAC's priority setting and the research activities of the CGIAR Centres, Dr. Saigal suggested that it may not be possible for the Centres to engage in narrow targeting of the poor. A more feasible approach might be to target farming systems, particularly those located in low potential areas. He advocated a strategy of agricultural intensification for these areas and the development and extension of biotechnology research geared to the needs of small farmers.
58. The Chair opened the floor to comments by TAC Members. These centred primarily on the feasibility of using poverty - whether defined by regions, population groups, or zones - as a criterion for setting priorities and allocating resources. The essence of this debate was whether in strategic terms it would be better to concentrate CGIAR resources by regions or attempt to target them more precisely to the poor.
59. Dr. Winkelmann began the discussion by posing to TAC Members the question: Given the CGIAR's commitment to poverty alleviation, if, as some have argued, more poor people lived in low potential than high potential areas, was it not necessary for TAC to know with some level of precision the geographic distribution of poverty in order to establish where the System's resources should be allocated?
60. In the ensuing discussion, a number of members endorsed the idea of addressing more specifically the problems of marginal areas because of their degree of poverty and natural resources degradation. Much of the damage now being done may be irreversible and, as had been noted in prior discussion, these areas were a major source of migration either to high potential areas or to already crowded cities. However, there was disagreement among TAC Members and other participants on whether the best way to treat these problems was through direct interventions in marginal areas, or by continuing to invest in high potential areas and promoting economic growth in the rural sector more broadly. Without such investments, the rates of migration to cities in some countries might have been even greater than they were. Diverting resources away from high potential zones might actually cause increased rural to urban migration and result in upheavals in both areas.
61. The various views expressed on this issue reflected variations in country experiences, again suggesting the location-specificity of the research problem. To some this seemed to imply the need for in-depth poverty profiles of low potential areas as a basis for resource allocation. Others felt that as a producer of research with the character of an international public good, the CGIAR required an allocation and programmatic framework that cut across national boundaries and different ecoregional zones.
62. While the issue of an allocative framework was not resolved, there was extended discussion of the methodological question of the feasibility and utility of measuring poverty. Some participants considered it risky to base allocative decisions on statistical estimates of the regional distribution of poverty, because the aggregate nature of the data tended to mask the actual incidence of poverty even at the regional level, and certainly below it. For example, existing data sets on the incidence of poverty in rural versus urban areas were available for only half of the developing countries and many of these used different poverty lines, preventing comparability of the data across countries. (This observation was questioned later.) The availability of statistics on poverty for Africa were actually declining with more available for urban than rural areas, making it difficult to fine tune resource allocation to degree of poverty. It was argued that even if a satisfactory method and reliable data were available, fee concept of stratifying the developing world's poor by income or nutritional levels in order to allocate agricultural research resources would add little to our understanding of where to invest those resources with greatest effect, given the highly heterogeneous nature of the populations and geography involved. Such a calculus required informed judgement as much as quantitative indicators.
63. In summarizing this session, the Chair drew a number of conclusions and implications for the TAC's and the Group's consideration. Although fee CGIAR was concerned wife poverty everywhere in fee developing world, fee agricultural focus of its mission implied a logic to giving greater priority to rural than to urban poverty, although it was also clear that creating a vibrant rural sector could have beneficial effects for poor urban consumers. In operational terms, however, fee targeting of resources accurately depended very much on how well poverty can be measured. The discussion gave TAC a better sense of fee possibilities and constraints in this regard. Correlatively, giving more priority to marginal areas was premised on fee assumption feat fee magnitude and depth of poverty in these environments was greater than feat in high potential zones, feat gains are possible through research, and, perhaps, feat fee latter, in any case, had available to them non-agricultural solutions to poverty feat marginal ones did not (except migration). While something was known about fee percentages of poor in these two zones, little was known about their absolute numbers. Thus, if fee CGIAR's commitment to poverty alleviation were to be operationalized and its resources allocated cost-effectively, more precise guidance from fee Group on fee issue of targeting must be sought.
64. The Chair introduced three speakers from advanced research organizations who were asked to share with TAC their views on the potential for the CGIAR to develop working relationships with such organizations. They were Dr. Jozef Schell of the Max-Planck Institut, Germany; Dr. B. Bachelier, Ministère de l'Enseignement Supérieur et de la Recherche, France; and Dr. Rudy Rabbinge, Scientific Council for Government Policy, The Netherlands.
65. Dr. Schell's presentation dealt with European institutions generally and stressed the potential for cooperation in biotechnology research, one of the major strengths of the advanced institutions. Since very little of such research in the advanced organizations in either the public or private sectors covered commodities important to the developing countries, involvement with the CGIAR could help to focus and apply biotechnology research to developing countries' needs. The principal contribution that biotechnology could make to the CGIAR's twin objectives of poverty alleviation and protecting the environment was to develop, through molecular biological research, plants which produced higher yields while requiring lower inputs. Dr. Schell provided three examples of leading institutions currently carrying out biotechnology research of relevance to the developing world. They were the International Laboratory for Tropical Agricultural Biotechnology (ILTAB); Advanced Molecular Initiative in Community Agriculture (AMICA); and International Service for the Acquisition of Agribiotech Applications (ISAAA). He stressed the importance of the role the CGIAR could play in ensuring that the worldwide effort in biotechnology research was brought to bear on the agricultural and environmental problems of developing countries.
66. Dr. Bachelier provided an overview of the French agricultural research and development system and its potential contribution to the CGIAR's work, touching upon research policy, topics for research, and modes of collaboration. Overall responsibility for the system at the political level was lodged in an interministerial National Coordination Committee (NCC). Relations with the CGIAR were handled by the Commission of International Agricultural Research (CIAR). France had increasingly established linkages with international, regional, and national research organizations. For example, it participated in CGIAR Task Forces and provided support to the regionalization of research in developing countries, particularly in Africa (e.g., CORAF, ASARECA, SACCAR). The primary disciplines embraced by the system were genetic improvement and plant pathology, water management and soil fertility, agricultural livestock systems, socioeconomics, and agroprocessing. Cash crops and forestry, inter alia, figured significantly in the system's research agenda. In discussing modalities of collaboration, Dr. Bachelier touched upon past and ongoing cooperation with IARCs and ecoregional programmes involving NARS, AROs, IARCs, and NGOs. He saw this approach as increasingly key to planning, implementing, and securing funding for future programmes in which France was involved.
67. Dr. Rabbinge briefly reviewed the evolution of the CGIAR from its early focus on crop improvement through genetic research, through its emphasis on agronomic and farming systems research, to the more recent thrusts on natural resources management and ecoregional approaches. The System, in Dr. Rabbinge's view, was now entering a new phase requiring integration of strategic, basic, applied, and participatory research and tailoring it to the specific needs of ecoregions to achieve sustainable agricultural development. The need for advanced methodologies and specialized knowledge to address these new concerns will require an enormous investment and the lead time between research and development will, in some cases, be quite long. Participation of the CGIAR in tripartite collaboration with NARS and AROs would thus become increasingly important. Dr. Rabbinge stressed the urgent need to create more synergism between these three research systems in order to maintain the same level of impact which the CGIAR has had with its traditional research. Biotechnology and systems-research for sustainable development were seen as the most promising fields for collaboration. The suggested approach was felt to be not only the soundest way to proceed scientifically, but also the most cost-effective way of ensuring continued core support for the international centres. Dr. Rabbinge closed by emphasizing that in the future agricultural research would have to be increasingly demand- rather than supply-driven.
68. The Chair opened the floor for comments by TAC Members. The central focus of these comments and the ensuing discussion was on the appropriate division of labour between the advanced institutions, international centres, and national systems. Consideration was also given to mechanisms for collaboration and attendant issues such as intellectual property rights and funding arrangements.
69. One TAC Member saw the opportunity of tapping into the expertise of the advanced institutions as an excellent way for the CGIAR Centres to address the more complex research issues now facing them. This prompted an extended debate on the question of comparative advantage. Would the suggested tripartite mode of collaboration mean that the AROs would perform basic research, with the IARCs and NARS focusing on applied and adaptive research, respectively? What was the research niche of the Centres in an emerging situation of direct ARO-NARS collaboration and how would such bilateral arrangements impact the CGIAR's priorities? Consensus was not reached on these issues and they triggered a number of subsidiary questions which would warrant TAC's and the Group's attention as new CGIAR priorities and strategies were put in place. If research was to be demand-driven, would new priorities such as biotechnology and systems-research be added to the Centres' agenda, or would they substitute for traditional commodity research? Was there evidence that biotechnology could improve yields sufficiently to justify greater weight to this research in CGIAR priorities? When joint research was undertaken with private AROs, would issues of intellectual property rights impede transfer of the technology to NARS? Was such research best undertaken on-site in the developing countries or in ARO laboratories? If NARS were to absorb and apply the new science, did this not imply a need for their developing some in-house capacity for basic research? If so, what were the implications for training of NARS in the new technologies and who should undertake it?
70. While various views were expressed on these individual questions, the Chair expressed the Group's consensus that the shifting comparative advantage among research systems now underway at the global level could be most effectively exploited to the benefit of the world's poor if the actors involved achieve an appropriate division of labour. TAC must ensure that mechanisms were in place to make that process an expeditious one.
71. The Chair introduced the first of three speakers for this session, Dr. Pierre Crosson of Resources for the Future. Dr. Crosson presented what he considered to be the principal issues in resource management for agriculture in the developing countries. His analysis was premised on the assumption that the management of natural resources was not an end in itself, but a means to contributing to agricultural development at acceptable costs, including environmental costs. Further, natural resources should not be seen as something apart from knowledge, which was a resource embodied in people, technology, and institutions. Thus, knowledge and natural resources constituted a "unity of resources" employed in agriculture to be managed in the most cost-effective way. Environmental costs in this context were understood as externalities, i.e., on-farm and off-farm costs. In sum, meeting the demand challenge in agriculture involved mobilizing natural resources and knowledge in ways that provided socially acceptable outcomes in terms of production.
72. Dr. Crosson posed the question of how much increases in the supply of natural resources would leverage the supply of food and fibre given the present state of knowledge of how these resources should be managed. Drawing on the FAO 2010 Study, he pointed out that the projected amount of land currently in range and forest which could be converted cost-effectively to agricultural production by 2010, would contribute to only 12 percent of the production needed for global food security. In qualitative terms, the amount of currently degraded agricultural land which might be restored to production was so small that it cannot be considered a significant source of production increase to meet the global food needs projected in the FAO and IFPRI studies. Dr. Crosson, therefore, argued that it was only through the mobilization of knowledge, as previously defined, that the world's agricultural production needs could be met. This was not to say that the CGIAR should not be concerned with natural resources management, particularly in marginal lands in Africa and Latin America. There might very well be payoff to research in these areas through increased understanding of why improving these lands for agricultural production had yielded increases that fall far short of their potential productive capacity. On the other hand, payoffs to enhanced knowledge (i.e., technology) depended very much on improvements in soil and water management. Dr. Crosson concluded that achieving complementarity of knowledge and natural resources management was key to addressing the complex production and sustainability issues of the marginal areas.
73. The second presentation in this session was given by Dr. Richard Harwood of Michigan State University. He stated the case for expanding the CGIAR's portfolio to include work in marginal areas on the assumption that food security and the cost of food were projected to be reasonably under control for the next 10 to 20 years. However, he qualified this by saying that because the growth of cereals (rice and wheat) yields were expected to continue to decline, in the long run the CGIAR should maintain a reasonable part of its research portfolio in these commodities, with particular emphasis on the genetics and production systems needed to maintain high yields. The latter will require new thrusts in soil biology and biochemistry and the genetic shifts of pests.
74. Dr. Harwood then turned to the primary focus of his presentation, the targeting of poverty in marginal areas. In his view, such interventions can be justified only when these areas had deteriorated to the point that they were a drain on national development in terms of the severity of their poverty, environmental loading of downstream production facilities, and migration of the poor to urban areas. There would otherwise not be sufficient national political will to initiate and successfully sustain such interventions. There must also be potential for improved productivity at reasonable cost and lack of alternative development options for these areas.
75. The objective of targeting was first to satisfy local food needs, thereby creating the conditions for farmers to move out of traditional food and feedgrain production and into perennial crops, animal agriculture, and other value-added crops. The conditions for development success included a progressive community structure, start-up investment capital, a lengthening of farmer planning horizons, and either a power source for tillage of annual crops or a market for high value-added, non-food perennial crops. Given these conditions, several new technologies would be needed for success, including those for management of soil biology to ensure efficient nutrient utilization, for carbon husbandry, and for local water harvesting systems. The principal areas for carefully targeted (i.e., to productivity increases) research investment included soil ecology, production systems, management of biological processes, and policy and economic constraints.
76. The final presentation of this session was delivered by Dr. Rabbinge. It was based on the report of the CGIAR Task Force on Sustainable Agriculture which he chaired. His remarks focused on the concept of sustainability and the methodological and institutional changes which the CGIAR should undertake to consolidate and strengthen its activities in the area of sustainable agriculture.
77. In the Task Force's view, the meaning of sustainable agriculture varied according to socioeconomic and environmental conditions; thus, there was no single methodology that would be valid for all ecoregions. It suggested that the problem of sustainability be approached by continuing to focus on issues of productivity and efficiency, explicitly taking into account long- and short-term environmental and equity concerns. A number of methodological recommendations were made, including the need to collect better information on the geographic distribution of declining productivity; to adopt an integrated systems approach to analysing sustainability issues at the field, ecoregional, and perhaps even global levels; and to create integrated research programmes for soil/water/nutrient and pest/disease/weed management. The Task Force also recommended a number of institutional changes which would help to consolidate and strengthen CGIAR ecoregional and Systemwide initiatives and research consortia, socioeconomic and public policy research, and collaborative links with NARS, NGOs, and AROs. Dr. Rabbinge concluded his remarks by re-emphasizing that sustainability research should focus sharply on promoting productivity and efficiency. He disagreed with the notion that sustainable production necessarily excluded the use of external inputs.
78. The Chair invited TAC's comments on the three presentations. These centred on such issues as the relationship between land degradation and agricultural production losses, whether degradation was a linear or non-linear process and the implications for policy interventions, the feasibility of measuring degradation accurately over extended periods of time, the degree of correlation between land degradation and the incidence and depth of poverty, the geographic location of degraded areas, and the relative priority to be given to researching various issues.
79. One TAC Member endorsed Dr. Crosson's analysis that it was difficult to find strong economic evidence on the costs of natural resources degradation in terms of decreased production. He also noted that while attention had been focused for many years on the downside effects of degradation, there was growing interest in measuring the upside opportunities, i.e., the possibilities for improving soil productivity. Finally, he acknowledged that while the logic of Dr. Crosson's argument was compelling, the perception in the donor community was influenced by concerns prompted by public opposition to various forms of environmental degradation. Priority setting for natural resources management needed to take various realities into account, particularly on issues as data-poor as the ones under discussion.
80. Another TAC Member inquired whether further weight should be given to investing in prevention of natural resource degradation relative to investing in rehabilitation. He also asked whether the knowledge resource central to Dr. Crosson's analysis was new or existing knowledge, and if the latter, whether the real question for priority setting was one of investing enough in extension to put the knowledge into use. Finally, he inquired whether the notion that movement toward a market system and rights in property ownership would help both to prevent resource degradation and alleviate poverty. Other TAC Members queried whether natural resource degradation was a linear or non-linear process and the long-term implications for research. A related concern was whether, in addition to focusing on the costs to production of on-site degradation, attention should also be given to the off-site, downstream effects which could be even more costly. Finally, it was observed that there may be differences of opinion between developed and developing countries on the relative significance of natural resource degradation owing to the substantially greater media coverage of this issue in developed countries. More public awareness needed to be generated in the developing countries; there was need for social science research in this area.
81. Dr. Crosson agreed that extension of existing knowledge or technology to farmers was a key institutional issue susceptible to research. He did not think that the trend toward privatization, particularly in sub-Saharan Africa, would necessarily lead to a reduction in poverty, but it could help to improve land management practices. On the linearity question, little empirical research had been done on this issue, but anecdotal evidence indicated that natural resource degradation was often reversible; the real question was whether restoration of degraded lands was a cost-effective way of improving productivity. On the issue of on-site versus off-site degradation, evidence from the United States suggested that off-farm costs were much the higher of the two. Whether this was true for other countries was, again, an important research issue.
82. Dr. Harwood, responding on the question of public awareness, argued that in many developing countries the externalities of degradation were readily apparent to the downstream recipients of pollution, since the damage was often obvious and economically costly; this was less a research issue than a matter of mobilizing public support at the national level for policies to deal with these effects at their source. He felt that private ownership in land and privatization, or at least stable access to land, were crucial to providing economic incentives to farmers to improve on-site land management. In Asia and Africa, and indeed globally, population pressures and migration into tribal areas were breaking down traditional land management systems. This was occurring in some very high risk areas. In the absence of private property, community action could address these problems by encouraging growers to have a long-term investment strategy with respect to the land.
83. In the ensuing discussion, a number of the points raised above were pursued in greater detail. The notion that the global production cost of degradation was relatively small was qualified by acknowledging that in many locales it had a severe impact on large numbers of people. At present, the location and size of degraded areas could be gauged only through fairly crude indicators using highly aggregated data. Much more refined information covering the locational and temporal dimensions of degradation was needed before the CGIAR could begin to think in terms of targeting its resources to the most severely affected areas and/or populations. There was an urgent need to reduce uncertainty on this issue through research.
84. The association between degradation and poverty was said to be "lumpy" in the sense that it was not uniform across regions. This had strong implications for TAC's prioritization. The CGIAR's working assumption appeared to be that if research to increase agricultural productivity was targeted to areas where these conditions were highly correlated, poverty would be reduced. In some cases, however, environmental conditions may be so fragile that intensification or the application of new biological technologies may make an already delicate situation much worse, perhaps increasing poverty. The scientific evidence on the feasibility of such a strategy seemed very problematic and required more definitive answers than appeared to be currently available. On a related point, while it was claimed that much of the basic science relevant to problems of natural resources management for tropical agriculture was already known, adapting it to specific agroecological environments represented a large research gap which should be addressed in TAC's prioritization. This was said to be reflected in the lack of specificity in the discussion of the component technologies that needed to be researched, for example, in the areas of soil and water management.
85. In drawing this session to a close, the Chair shared with participants the perspective of Dr. Serageldin on investment in natural resources management which currently figures in the World Bank's approach to this issue. In his view, there were four classes of capital important to humans: natural resources, physical resources, human capital, and social capital. The last included the institutions available to manage natural resources. The question that emerged in managing these resources was whether it was necessary to preserve for future generations the total stock of inherited natural resources or a stock of total capital larger than that which was inherited. Historically, he observed, humans have traded natural resources for human capital. The key question confronting TAC and the CGIAR was what kinds of trade-offs among these classes of capital should inform its priority setting as it pursues the System's objectives?
86. TAC would prepare a follow-up to these discussions and an issues paper on each topic for discussion by the Group at ICW'95. TAC had also initiated development of a framework and criteria for setting CGIAR priorities which would also be discussed with the Group at ICW'95. A comprehensive report on CGIAR priorities and strategies would be completed at TAC 69 in March 1996 and submitted to the Group at MTM'96.
