This study was expected to address CGIAR priorities for research on marginal lands. The final point in the terms of reference called for suggested priorities for research on these lands and an assessment of the appropriate balance of effort between marginal and favoured lands (FL). As it delved into the issues, the Panel discovered there was: considerable confusion on definitions and concepts of marginality; a strong body of conventional wisdom on the relationship between poverty, marginal lands (ML) and resource degradation, the role of the CGIAR in poverty alleviation, and the potential of ML to produce food; and an abysmal lack of data to validate this conventional wisdom. As a consequence the study has focused exclusively on strategic questions bearing on how the CGIAR might more systematically address poverty alleviation. In the process the Panel felt obliged to diverge from the initial hypothesis that research strategy would be determined primarily by the biophysical characteristics of ML.
Perceptions Driving International Concern with ML
The concern of the Lucerne Declaration that the CGIAR ensure that its research programmes address the problems of the poor in less-endowed areas derives from widely held views on the nature and scope of the issue. The axis is that research efforts by the System have done little for the poor on biophysically marginal lands. The view that most of the poor live in these fragile areas and are a cause and consequence of a downward spiral of resource degradation and deepening poverty, lends urgency to corrective action. Definition of strategy with respect to overall allocation of resources within ML and between ML and FL would require: forecasts of the poor, and the rate of resource degradation with associated yield decline and externalities; and an estimation of the potential poverty alleviation impact and the related environmental protection impact of research options applicable to both ML and FL. These premises and requirements were taken as the point of departure for the study.
Definitions and Concepts
Marginal lands: In order to provide a basis for mapping ML by biophysical characteristics, from which one could go on to determine numbers and incidence of rural poor and production potential for each area, the TAC Secretariat undertook an extensive analysis of existing data leading to a six-quadrant matrix of three land types: FL, ML and lands in low or zero intensity use (LZI), each sub-divided into two areas - those with a high expectation of a productivity response to research and those with a low expectation of response. The results from this exercise did not provide operationally relevant criteria for CGIAR decisions. Accordingly, a second approximation was made whereby biophysical characteristics establishing marginality were qualified by their current use in agriculture. This yielded estimates for favoured agricultural lands (FAL) - 800 million ha - and marginal agricultural lands (MAL) - 1,800 million ha.
Location and incidence of rural poverty: From the above figures estimates were made of rural population on the two land types - 930 million on FAL and 1,760 million on MAL. No data are available for rural poor on each land type. However, from examination of data on poverty incidence in countries with high and low proportions of MAL and available case studies, there is no evidence that poverty incidence is higher on MAL than FAL. Applying available estimates of rural poverty incidence to rural population on these two land types results in 325 million poor on FAL (35%) and 630 million (65%) on MAL.
Role of the CGIAR in MAL: A review of the project portfolio suggests that about two-thirds of the System's resources are invested in research relevant to MAL. But there are no data on what may be targeted to poverty alleviation. The centres have had many successes on marginal lands. If one accepts the above proportion of investment in MAL as well as the view of limited research impact on poverty, then a question of cost-effectiveness clearly arises. However, there is no evidence to support or refute the latter view. It is nevertheless a key question which can only be addressed by progressive accumulation of information along the lines proposed in Recommendation 1 discussed below.
Poverty and resource degradation: There is no evidence to support the view implying that the poor are the prime cause of resource degradation on MAL. Again, clarification of this issue would derive from data assembled in response to Recommendation 1 and the study called for in the TAC Soil and Water Study to address "myths" on linkages between poverty, resource degradation and externalities.
From Marginal Lands to Marginal People
In reviewing the data presented above on the location of poverty, poverty incidence and the allocation of CGIAR resources to MAL and FAL, the Panel came to the conclusion that the inconsistencies and lack of data on the underlying site-specific forces driving the poverty process were such as to invalidate their usefulness in guiding strategy towards poverty alleviation on marginal lands.
The concept identified as relevant to evaluation of strategy was "marginal area" (MA), which is defined firstly by presence of high numbers and high incidence of marginal people, and secondly is subject to a relatively homogeneous set of conditions which explain why they are poor - institutions, policy, infrastructure, human capital, biophysical characteristics of land, etc. This definition of the "unit of account" for analytical purposes poses two problems for the CGIAR:
by redefining the issues in terms of MA which may incorporate any combination of ML and FL, poverty alleviation is being placed in a global context, thus moving away from the original intent to deal only with poverty on MLMAs, by definition, will be site-specific and likely to be heterogeneous, and subject to more internal diversity, complexity and institutional constraints than would be expected for non-MAs. This is information-intensive and clearly presents a challenge in structuring research to provide international public goods. It also establishes a context for poverty alleviation which goes beyond what the System is now doing or is likely to do.
The Panel is aware that over the past two years the CGIAR has fully adopted poverty alleviation as the objective of its research efforts and the MA approach proposed may be regarded as merely a restatement of this commitment. Nevertheless, it is felt that working through the ML/MAL/MA sequence has highlighted a number of dimensions which clarify strategic questions on approach - the data gaps (poverty/resource degradation linkages, poverty processes), targeting marginal rural people and site-specific characteristics of MAs which condition the relevance of research. The Panel is convinced that the System must come to grips with the site specificity and both the biophysical and non-biophysical elements which explain poverty if it is to design research in expectation that poverty impact can be evaluated.
The entry point to an evolving strategy derives from Recommendation 1: The CGIAR needs to sharpen its strategic focus on poverty alleviation, particularly in setting priorities for research related to marginal rural areas. A prerequisite is development of a geo-referenced database linking biophysical land conditions with poverty and the processes that produce it (i.e. the dynamics of poverty).
The focus on poverty alleviation in marginal areas involves challenges in identification, development and organization of research. Chief among these is the need to develop new forms of partnerships with NARS, NGOs, developed country institutions and multi and bilateral agencies. The CGIAR's role will be restricted, albeit important. The key to success is establishment of the right kinds of partnerships, hence:
Recommendation 2: The centres should establish new forms of partnership in order to effectively address their roles in a broader poverty alleviation strategy related to those who live in marginal rural areas.
System Research on MAs
The research thrust on poverty alleviation is seen to be in three inter-related areas:
Institutional and policy change:
Work in this area has been covered by other TAC studies on social science, institutional, policy and management research. The main focus is on analyses of: poverty processes; opportunities and constraints to institutional change; and viable options for improving welfare within existing constraints.
Generating new and improved technologies: Technologies need to fit the institutional and socio-economic characteristics as well as biophysical conditions of MAs.
Diversifying land use systems and income opportunities: Research will follow lines already established for natural resource management and clarification of linkages between resource degradation and poverty. The focus would incorporate diversification of production from renewable resources and off-farm income opportunities.
Two recommendations derive from the above research agenda:
Recommendation 3: The CGIAR System should develop improved mechanisms by which centres can be involved with other partners in generating and interpreting improved scientific evidence on (i) the extent and magnitude of the impacts of agriculture, forestry and fisheries on the degradation or enhancement of natural resources and the consequences for production and food security; and (ii) the linkage between poverty and observed resource degradation.
Recommendation 4: Expanded collaborative mechanisms and activities should be developed among centres and between centres and their non-CGIAR partners, to help focus research and institutional strengthening on issues related to adoption, adaptation and utilization of research results that so far have remained unused.
Implications of a Broader Research Focus
The Panel recognized that it is recommending a broadening of the research focus to include non-traditional (for the CGIAR) types of research related to off-farm employment, postharvest technology, niche opportunities, and forestry, agroforestry and fisheries. Such expansion implies a possible readjustment in use of existing resources. On the question of whether there should be an increase in resources, the Panel is fully aware of the constraints. In this era of donor fatigue and contraction of funds for both national and international R&D for agriculture, any expansion may seem unrealistic. The nature of research to target poverty alleviation on MAs dictates the forging of new linkages between the CGIAR and other actors. High transaction costs are inevitable. However, there is an urgent need for an expanded innovative research effort to address poverty alleviation for those who live on MAs. These marginal people have been disregarded in the past. There appears to be a unique opportunity for the CGIAR to contribute towards correcting this neglect.
