2.1. Marginal Lands
2.2. Initial Framework for Evaluating the Issues
2.3. Marginal Lands as a Moving Target
2.4. Contexts in which the Term "Marginal Lands" is Used
2.5. From Biophysically Marginal Lands (ML) to Marginal Agricultural Land (MAL) Based on Existing Use
2.6. Location and Incidence of Rural Poverty on MAL
2.7. Linkage of FAL to Poverty on MAL
2.8. Current CGIAR Activity in Poverty Alleviation on MAL
2.9. From Marginal Lands to Marginal Population
2.10. Poverty Processes on Marginal Areas
2.11. Implications: Recommendations for the CGIAR
In responding to the Phase I report of this study, most of the centres raised questions on: the interpretation of "marginal lands"; the advisability of using very coarse-grained indicators of biophysical capability of land as one basis for programming the System's activities; the credibility of estimates relating to rural population and land types; and, what is meant by poverty alleviation with respect to its links to research delivered by the CGIAR.
Many different names are used to designate lands in terms of their production capacity - favoured, fertile, marginal, low potential, resource poor, high potential, fragile, vulnerable or degraded. Terms which relate to "marginal" areas are frequently used interchangeably and often without definition. The difficulty in formulating a clear definition stems from the fact that "productivity" varies according to the type of land use. A tract of land that is "marginal" for crop production may be well suited for grazing. "Fragile" lands may be sensitive to degradation under cultivation but may be sustainably used for forestry. Further more, productivity is not only based on the biophysical characteristics of land, but also depends on the socio-economic parameters of a specific environment. Technologies may be known but the necessary incentives, institutions or inputs may be missing. Farmers are generally aware of the physical benefits of improved seeds and of mineral fertilizers, but may not have access to inputs or be unwilling to accept the climatic and price risks. The range of possible uses of land is so wide and socio-economic conditions are so diverse that no definition can encompass all the relevant factors. However, in order to ensure a common understanding, the general terms used in this study are briefly described below.
Attention is drawn here to the concept of "land" which is broader than just soil or terrain. Although soil is an important and relatively stable element of land it is only one component. Land is an area of the earth's surface which comprises the major biophysical attributes which influence its use. These include the atmosphere, soil, geology, hydrology, plant and animal populations, and the results of past and present human activity such as terracing, drainage or irrigation. The evaluation of land does not only refer to its productive potential, but also to the sustainability of its use, that is the maintenance or enhancement of its productivity over the long term, while at the same time conserving its potential as a resource base.
Table 2.1 lists the four terms - favoured, marginal, fragile and degraded land - which are commonly used in this study. It should be noted that the constraints of a biophysical and socio-economic nature may apply separately or simultaneously. Marginality can be the result of different combinations of constraints. For instance, biophysically "good" land can be marginal on account of its isolation from markets, the unavailability of inputs, or the small size of holdings. The nature, composition and interaction of the factors which determine land marginality differ widely.
Table 2.1: Proposed Definitions of Land Types
|
Definition |
Biophysical Constraints |
Socio-Economic Constraints |
|
Favoured land: Land having no, or moderate limitations to sustained application under a given use. Moderate limitations will reduce benefits but an overall advantage will be gained from the use of inputs. Wide options for diversification. With proper management, risk of irreversible damage is low. |
No or moderate constraints related to soil, climatic and terrain conditions. Soil fertility, if adequately maintained, is favourable. Relatively reliable rainfall and/or irrigation water. |
The level of yields depends not only on favourable biophysical conditions, but on accessibility to inputs, market and credit facilities, and beneficial output/input ratios. |
|
Marginal land: Land having limitations which in aggregate are severe for sustained application of a given use. Increased inputs to maintain productivity or benefits will be only marginally justified. Limited options for diversification without the use of inputs. With inappropriate management, risks of irreversible degradation. |
Soil constraints (low fertility, poor drainage, shallowness, salinity), steepness of terrain, unfavourable climatic conditions1. |
Absence of markets difficult accessibility, restrictive land tenure, small holdings, poor infrastructure, unfavourable output/input ratios. |
|
Fragile land: Land that is sensitive to land degradation, as a result of inappropriate human intervention2. Sustained production requires specific management practices. Land use is limited to a narrow choice of options. |
Soils of low fertility, erodible, steep terrain, high groundwater levels, flood-prone. |
Population pressure, food deficits, competition for land from other sectors, unavailability or high cost of inputs. |
|
Degraded land: Land that has lost part or all of its productive capacity as a result of inappropriate human intervention. Various forms and degrees of degradation, both reversible and irreversible, may occur. Rehabilitation of reversible forms of degradation requires investment. |
Erosion, salinization, fertility depletion, lack of adequate drainage on soils and terrain prone to deterioration. |
Population pressure, land shortage, inadequate support to agriculture, lack of institutional framework, high cost of rehabilitation, lack of investment. |
1 The soil, terrain and climatic constraints applicable to marginal lands are described in Annex II. The constraints may apply separately or cumulatively.2 A distinction needs to be made between reversible and irreversible forms of degradation. Some soils are vulnerable to nutrient depletion, but are sufficiently resilient for soil fertility to be restored through good management.
The nature and extent of the challenge faced by the CGIAR in addressing poverty alleviation on marginal lands may be illustrated from Table 2.2. Three broad land types are identified:
(i) favoured lands with high present agricultural use values (HPUV),
(ii) marginal lands with low present agricultural use values (LPUV) and
(iii) lands at low or zero intensity of agricultural use (LZI).
For each land type one may speculate that there are those with a high potential for expansion of production based on research (HPEPR) and those with a low potential for expansion of production based on research (LPEPR). Thus; there is a six-quadrant matrix which may be addressed by research. It is generally held that research undertaken to date, focused on land types in quadrant I and quadrant II, has largely benefited quadrant II, particularly the wheat and irrigated rice lands. As it stands this table is an approximation of potential to expand agricultural production through research. It does not reflect options for research on forest systems, coastal fisheries related to terrestrial systems or the dynamics of movement of land types among the quadrants (see Annex II).
In their response to the Phase I report, most of the centres commented on the criteria used in establishing area estimates for each quadrant, and approximating rural population and rural poor associated with each. In addition, a concept was introduced for marginal lands which identified those areas where both the numbers and proportion of rural poor were low, in contrast to what might be considered "poverty dense" areas. In general, the centres do no design their programming to address soil/climatic domains or rural populations classified by geographic area. Rather they tend to target problem areas or specific land use regimes. In spite of the difficulties in defining terms used to characterize the biophysical productivity of lands and in estimating "poverty density", the Panel considered the six-quadrant approach to be a useful concept in sorting out the issues and priorities in allocating research resources between high productivity areas and marginal lands associated with persistent poverty. It is a particularly appropriate framework in light of the following widely held views which characterize thinking on rural poverty and the role of the CGIAR:
ML are defined in biophysical terms which establish them as: having low inherent productivity for agriculture; fragile and therefore susceptible to degradation because of slope and/or climate; and subject to high agricultural risk due to climate and disease;ML support a high proportion of the rural poor, particularly the poorest of the poor;
the combination of fragility and high density of poor people who place a premium on current consumption (resulting in over-exploitation of natural resources) is leading to accelerated erosion and vegetation destruction; the consequence is a downward poverty spiral with significant negative externalities because of the large areas classified as ML relative to those considered favoured;
the impact of CGIAR research on poverty alleviation and related productivity increase and environmental protection in ML has been low.
Table 2.2: Present Use Value (PUV) of Lands and Potential to Expand Production Based on Research (PEPR)
|
Present Use Value |
High PEPR |
Low PEPR |
|
Favoured Agricultural Lands |
QUADRANT I |
QUADRANT II |
|
Marginal Agricultural Lands |
QUADRANT III |
QUADRANT IV |
|
Lands at Low or Zero |
QUADRANT V |
QUADRANT VI |
It is recognized that lands do not neatly define themselves in the above quadrants and that their classification is not definite. In practice, there is a continuum among the quadrants with changes in economic conditions, improvement of infrastructures, innovative research, fluctuations in population pressure. For example, the existence of a continuum is particularly common between quadrant V (forest), quadrant III, and quadrant I. Defining the LZI type is a hazardous exercise where one is moving from the extensive desert margins in quadrant VI to potential opportunities for intensification of forestry in quadrant V. In addition, the definition of lands which are responsive or unresponsive to research (in production terms) is a static concept. In practice, research innovation should enable LPUV lands in quadrants III and IV to be transferred to HPEPR lands in quadrant I and LPEPR lands in quadrant II.
Any number of factors may lead to shifts of land from one category to another. These shifts may be upward, through applications of improved techniques, or downward as a result of land degradation or inappropriate development of lands formerly at low use levels. Hence, marginality is not a static concept.
When dealing with issues of natural resources management, it is essential that these spatial and temporal dimensions be taken into account. Therefore, "marginality" has to be assessed in terms of specific types of land use. Land that is marginal for a crop requiring a long growing period could be highly productive for more drought tolerant, short-season crops. What is marginal land for cropping may be very suitable for livestock production. The severity of soil constraints may vary with the climatic conditions under which they occur. For instance, under low rainfall, lighter soils may be more productive than heavier soils on account of easier tillage and more rapid uptake of available water. Shifts between quadrants may also result from management practices. Infertile soils which are marginal can become highly productive through judicious application of organic and mineral inputs.
A key characteristic of marginal, as distinct from favoured lands is the location specificity of terrain, climate, soils and socio-economic conditions. Many technologies to remedy biophysical marginality are well-known. The dynamic dimension is critical - marked shifts of land quality will result from the use of resources with the necessary incentives to apply results of research, i.e., overcoming the policy and institutional constraints.
In summing up, the Panel found that the definition of "marginal" depends on so many qualifiers that it becomes meaningless in an operational sense. This is illustrated by the fact that land can be "marginal" depending on:
· its use (what is marginal agricultural land may be highly productive forest land);· its natural biophysical characteristics (which can be altered by investment);
· its location relative to infrastructure such as roads, railroads, harbors, and cities (a road into a region can completely alter the economic returns from land near the road);
· the institutional and policy context which influences access of inhabitants to land, water, credit, markets, outside inputs (development of market access can completely alter the economics of land use);
· population pressure (e.g., size of land holdings; from a cattle rancher's perspective, his or her large area of land is not "marginal", even though the biophysical yield per ha is low; at the same time, a farmer with only one ha. in the midst of the most favoured agricultural area may feel that he or she is on "marginal land");
· technology development (Jojoba development in arid environments; acid tolerant rice in the Cerrados of Brazil);
· taking advantage of niche opportunities (spices, flowers, vegetables, special fibres)
Lands move out of and into marginal status depending on which of the above dimensions are applied in the definition. It only makes sense to define "marginal land" in terms of a clearly defined, specific situation.
Because of the wide variety of ways in which "marginal" lands can be defined, the term is used to mean quite different things, depending on the context. In fact, the Panel reached the conclusion that the concept, in a biophysical sense alone, is not adequate to identify operational responses by the agriculture, forestry and fisheries sectors to poverty alleviation - the fundamental goal of the CGIAR.
The Panel identified three common threads that run through the documentation on "marginal lands," broadly defined. They are:
· A concern with poverty - under the assumption that most of the rural poor live on "marginal" lands; thus a concern for "marginal" lands is a proxy for concern with poverty alleviation for those who happen to live on the myriad forms of less favoured areas of the world; and a further concern for the fact that the CGIAR may somehow have missed these poor people, because it has not targeted poverty in the "marginal" lands areas. (This concern dominates in the development literature)· A concern with vulnerable and fragile lands and the problems of irreversible destruction or degradation of sensitive natural areas - the problems of desertification; deterioration of mountain environments; the destruction of other natural environments such as mangroves and natural forests, and pollution and destruction of biodiversity in coastal zones. (This concern dominates in the environmental literature)
· A concern that the more favoured rainfed lands, the irrigated lands, and some of the "rice" and "wheat" baskets of the world, as well as highly productive bottomlands and hillsides, were being "marginalized" through overuse or misuse. (This concern shows up in the agricultural development literature, as well as in the environmental literature)
The Panel further noted that, even within each of the interpretations of why "marginal lands" are important, there is uncertainty over what the term means, how lands can be classified as "marginal", and how they can move out of marginal status into productive lands contributing to sustainable development.
In light of the above reasoning, the Panel came to the conclusion that, although the soil/climatic domains specified in Table 2.2 provide a scientifically coherent basis for identifying land capability with respect to its agricultural potential and its expected productivity response to agricultural research, it was not an operationally relevant concept. While retaining the biophysical criterion for classification, it was considered that a better approximation of the marginal land issues faced by the CGIAR would be provided by qualifying biophysical characteristics of lands according to their current use in agriculture (See Table 2.3). This enabled a specification of a scientifically defensible area of favoured agricultural lands (FAL). Thus, they excluded forest and woodland areas with high agricultural potential. FALs were considered to be in rainfed and irrigated agriculture (800 million ha) in areas which are fertile (with or without chemical subsidies), well-drained, with even topography and (if unirrigated) with adequate rainfall. They are in comparatively intensive use, and generally within their use-capability. They are at risk of degradation if mismanaged but risk of externalities from mismanagement is low, as is vulnerability to irreversible damage.
Marginal agricultural lands (MAL) were identified as those areas which are currently used for agriculture, grazing or agroforestry (1.8 billion ha). Such areas typically encompass mountains and tropical and sub-tropical lowlands or plateaux with low, unstable rainfall or higher rainfall areas in intensive use relative to use-capability under existing population densities, traditional technologies and institutional structures. In most cases, in absence of external inputs, they have reached or exceeded the threshold limits to maintenance or enhancement of agricultural performance. They are characterized by: poor soil fertility (nutrient deficiencies, acidity, salinity, poor moisture holding capacity, etc.), inaccessibility (poor communications, immobility with all its social and economic implications); fragility (low input absorptive capacity, high input-output ratios, limited capacity to withstand disturbance, vulnerable to irreversible damage); and heterogeneity (physically and culturally diverse with site-specific constraints and opportunities which restrict applicability of general technological or institutional measures to remove constraints or exploit opportunities). Aside from the above inherent characteristics, marginal low-productivity lands may also result from degradation of non-marginal lands or inappropriate development of lands formerly at low or zero use levels. All areas are at risk of further degradation with high expectation of negative externalities.
The quadrants V and VI were redefined in Table 2.3 such that quadrant V was considered forested area regardless of its agricultural potential and quadrant VI was classified as arid with very limited rural population regardless of its irrigation potential. Because of the dynamics of movement among quadrants, both of these - particularly quadrant V - are relevant to rural poverty alleviation for the current and future generations in marginal areas.
A second modification in the conceptual ML approach centred on the location of rural poverty. Since there are no data for rural population in forest or arid areas, estimates were made for FALs and the residual population was assigned to quadrants III - VI (Table 2.4) with the expectation that the majority would be located on MALs. The distribution is 35% (930 million on FALs) and 65% (1,760 million) on other lands (MAL, forest and arid lands).
Table 2.3: Extent of Different Land Types in the Developing Regions (million ha and share within region)
|
|
|
BREAKDOWN OF OTHER LAND TYPES |
|
|||||
|
Land types |
Favoured agricultural |
Other land types |
Marginal agricultural |
Forest & woodland |
Arid lands |
Irrigation in arid lands |
TOTAL |
|
|
Quadrant |
I/II |
III-VI |
III/IV |
V |
VI |
(I/II) |
|
|
|
Region |
|
|
|
|
|
|
|
|
|
SUB-SAHARAN AFRICA |
200 |
2155 |
545 |
1030 |
550 |
3 |
2355 |
|
|
40 countries |
Share (within row) |
8.5% |
91.5% |
23.7% |
43.7% |
24.6% |
0.1% |
100% |
|
ASIA |
305 |
1530 |
550 |
635 |
340 |
28 |
1835 |
|
|
20 countries |
Share |
16.6% |
83.4% |
30.0% |
34.6% |
18.5% |
1.5% |
100% |
|
CENTRAL AND SOUTH AMERICA |
190 |
1780 |
400 |
1220 |
160 |
4 |
1970 |
|
|
26 countries |
Share |
9.6% |
90.4% |
20.3% |
61.9% |
8.1% |
0.2% |
100% |
|
WEST ASIA AND NORTH AFRICA |
100 |
1185 |
290 |
50 |
845 |
12 |
1285 |
|
|
19 countries |
Share |
7.8% |
92.2% |
22.6% |
3.9% |
65.8% |
0.9% |
100% |
|
TOTAL |
795 |
6650 |
1785 |
2935 |
1925 |
46 |
7445 |
|
|
Total: 105 countries |
Share |
10.7% |
89.3% |
24.0% |
39.4% |
25.9% |
0.6% |
100% |
Table 2.4: Rural Population and Rural Poor on Different Land Types (in million)
|
|
|
TOTAL POPULATION |
RURAL POPULATION |
RURAL POOR |
AVERAGE RURAL POVERTY |
||
|
Land types |
All land types |
Total |
Favoured agricultural |
Other land types |
Other land types |
|
|
|
Quadrant |
I-VI |
I-VI |
I/II |
III-VI |
III-VI |
|
|
|
Region |
|
|
|
|
|
|
|
|
SUB-SAHARAN AFRICA |
530 |
375 |
101 |
274 |
176 |
64% |
|
|
40 countries |
Share (within row) |
|
100% |
27% |
73% |
|
|
|
ASIA |
2840 |
2044 |
755 |
1289 |
375 |
29% |
|
|
20 countries |
Share |
|
100% |
37% |
63% |
|
|
|
CENTRAL AND SOUTH AMERICA |
430 |
117 |
40 |
77 |
48 |
61% |
|
|
26 countries |
Share |
|
100% |
34% |
66% |
|
|
|
WEST ASIA AND NORTH AFRICA |
345 |
156 |
37 |
119 |
35 |
29% |
|
|
19 countries |
Share |
|
100% |
24% |
76% |
|
|
|
TOTAL |
4145 |
2693 |
933 |
1759 |
633 |
36% |
|
|
Total: 105 countries |
Share |
|
100% |
35% |
65% |
|
|
The question then arises as to whether there is a greater incidence of rural poverty on MALs than FALs. National figures available for rural poor, as a percent of total rural population, were compared with the percent of rural population on MALs to assess whether countries with a high proportion of population on MAL also showed high incidence of rural poverty. No correlation was found. This conclusion also is suggested by the work of Kelley and Rao in India3. An IFPRI study, also in India, suggests that both the absolute numbers and incidence of rural poverty are greater on FAL than MAL4. Accordingly, rural poverty in the two areas was approximated by applying national percentages - the result 325 million poor on FAL and 630 million on MAL5.
3 Kelley, T.C. and P.P. Rao. 1995. "Marginal Environments and the Poor: Evidence from India". Economic and Political Weekly. Vol. xxx. No. 40.4 Fan, S. and P.B.R. Hazell. 1996. "Should the India Government Invest More in Less Favoured Areas?". IFPRI, Washington D.C.
5 This figure contrasts with an estimate made in the mid 1980s - 370,000 poorest of the poor (using the criterion of the poorest 20% for each country) on low potential lands, in: Leonard, H.H. (ed.). 1989. "Environment and the Poor: Development Strategies for a Common Agenda". Transaction Books. Oxford.
Similarly, in forecasting poverty on MAL, the only proxy for future numbers in the latter area would have to be derived from projections of overall rural poverty. Using trend estimates for the 1990-2000 period, one might speculate that, unless a major effort is made, numbers are unlikely to decline significantly at a global level in the near future. However, at the regional level, in sub-Saharan Africa the situation is likely to deteriorate such that its share of the world's rural poor may increase from about 30% to 45% by early next century. Asia would continue to have the highest absolute number of rural poor, but its world share may drop from around 60% to 50%6.
6 Pinstrup-Andersen, P. and R. Pandy-Lorch. 1994. "Alleviating Poverty, Intensifying Agriculture, and Effectively Managing Natural Resources". Food, Agriculture and the Environment Discussion paper 1. IFPRI. Washington, D.C.
Introduction of the degradation and productivity dimensions to the poverty-marginal land linkages requires a focus on the dynamics of movement, from favoured agricultural lands downward to ML and the reverse through upgrading. It also requires considering the possibilities for: expanding the stock of favoured agricultural lands by transformation from the 2.9 billion ha of forests and development of irrigation on a small fraction of the 2 billion ha in arid land; or, expanding the stock of marginal lands, assumed to be driven largely by rural poverty, at the expense of forest lands. These dimensions are discussed below with respect to rural poverty alleviation (in MALs) linked to FALs.
There are three reasons for considering favoured agricultural lands (FALs) in a study of poverty alleviation on marginal lands:
Through degradation they can shift into the low productivity category (marginal) of either rainfed or irrigated lands. As pointed out in comments by several of the centres on the Phase I report of the study, the clearest example is salinization of irrigated lands. The impact of this process is increasing poverty on an increasing stock of marginal lands.Through further productivity increase, particularly in quadrant I, they should provide employment opportunities for rural poor who would otherwise contribute to pressure (degradation) on MALs or to accelerated conversion of forest to MALs.
Similarly, the processes of reduction of pressure on natural resources as a means to poverty alleviation would also apply to employment opportunities generated in value-added or off-farm activities in FALs.
In addition to the role of favoured lands in poverty alleviation for present and future rural populations in marginal areas, there is also the question of research directed to the poor on favoured lands themselves. This is a fundamental strategic question for the CGIAR in terms of the balance among research investments which address a range of rural population target groups from a poverty alleviation perspective. There is also a major question of yield maintenance research in FALs. There is a continuum from investments designed to: rehabilitate the productive capacity of the resource, i.e. reverse the degradation process; conserve the "state" of resources, and implicitly their yield capacity; and enhance yields. The first two are land improving investments, the third is "technological change"7.
7 Scherr, S.J. and S. Yadav. 1996. "Land Degradation in the Developing World: Implications for Food, Agriculture, and the Environment to 2020". Food, Agriculture and the Environment. Discussion Paper 14. IFPRI, Washington, D.C.
It is evident that priority of research related to poverty alleviation in all areas (favoured and marginal) through changes in the cultivated area, employment and income generated on FALs by reducing shifts to lower productivity, investment in productivity increase, human capital, physical infrastructure and institutional change, should be judged by the cost-effectiveness of research expected to have poverty alleviation impacts for the various target groups. The Panel is aware that we are a long way from specification of these target groups. Nevertheless, the linkages and dynamics of movement among classes of land and changes in the numbers and location of the rural poor, dictate that any strategy of rural poverty alleviation on marginal lands take into account the social cost-effectiveness of research on favoured lands, and should incorporate the spillovers to other areas. The framework for cost-effectiveness analysis is discussed in Annex III.
The dilemma of research on MALs vis-a-vis FALs has been characterised by de Wit8 - "well endowed regions are able to meet demands at relatively low prices... this marginalizes less endowed regions because their terms of trade are eroding. Scarcity of funds, and the lack of political power of (these) regions makes it unlikely that the transfer of money that would be needed to revert such marginalization processes will occur in the foreseeable future..... Agricultural research which is oriented to improvement of least endowed regions may open up new possibilities in some situations. However, in many cases its results are more readily applied in regions that are better off." Accepting this premise, a key issue for the CGIAR is the weighting of poverty in the formula discussed in Annex III.
8 De Wit, C.T. 1990. "Understanding and Managing changes in Agriculture" in J.W.G. Jones (ed.) "System Theory Applied to Agriculture and the Food Chain", Elsevier.
Based on the foregoing definitions of MAL and FAL, the CGIAR invests about one-third of its resources in the favourable agricultural lands of quadrants I and II (FALs), and the remaining two-thirds on the marginal agricultural lands of quadrants III and IV (see Annex IV). Considering, however, that targeted MALs include productive areas such as the cracking "black cotton" soils of India and western Africa, the poorly drained "inland valleys" of West Africa and the infertile "Cerrados" of Brazil, it would also seem appropriate to draw the conclusion that three quarters of CGIAR resources are being applied to increase the sustainable productivity of lands with high agro-ecological potential.
In terms of the three primary thrusts of research (see section 3.4), estimates of resource allocation indicate that around 30% support activities aimed at generating improved biological opportunities, close to 50% go to the intensification and diversification of production systems, and the remaining 20% to policies and institutional issues.
Current activity categories, however, are not sensitive enough to indicate actual resources allocated to alleviating constraints for the marginal poor. An assessment of objectives, outputs and beneficiaries of the 374 projects endorsed for 1997 showed that 25% are fully targeted at poverty alleviation on MALs and 7% in FALs (Annex IV).
In reviewing the figures presented in Sections 2.6 and 2.8 derived from a biophysical definition of MAL and FAL, the Panel came to the conclusion that the inconsistencies and lack of data on the underlying site-specific forces driving the rural poverty process were such as to invalidate their usefulness in guiding CGIAR strategy towards poverty alleviation on marginal lands.
On the surface, the existence of two-thirds of the rural poor on MAL would fully justify the CGIAR's concern with poverty on these lands. This is further reinforced if one accepts an underlying hypothesis of the Lucerne Declaration, that Green Revolution techniques have had limited impact on productivity for poverty alleviation in MALs, so defined. However, on further reflection the Panel concluded there was no evidence to support this hypothesis. There are areas where CGIAR research has not contributed to rural poverty alleviation but they are not confined to MAL. In addition, there are MALs which have shown significant response to research by the System in terms of productivity and, by inference, poverty alleviation. MAL is simply not a good proxy for where the poor people live or where the CGIAR has had little impact. Under these circumstances it was decided to discard biophysical productivity of lands, whether in the pure sense of world soils and climate, i.e. ML, or in the restricted sense of those areas currently in agricultural use, i.e. MAL, as the point of entry for assessing CGIAR strategy with respect to rural poverty alleviation.
The Panel had considerable difficulty in finding a substitute term for ML or MAL which respected an interpretation of the intent of the terms of reference to be that the study address strategic issues in research aimed at poverty alleviation on lands which were intuitively considered as biophysically marginal and which clearly were believed to support high concentrations of rural poor who had not benefited from Green Revolution research e.g. the Himalaya, Andes, desert margins of WANA or tropical forest margins. In fact, the Panel believes that the above interpretation of the intent has been respected. For want of anything better the term adopted as the "unit" for evaluation of strategy was "marginal area" (MA), i.e. one with a high incidence of rural poor subject to a relatively homogeneous set of conditions which determine why they are poor (see section 2.10). This obviously includes biophysical conditions.
In spite of the fact that MA may be applied to any combination of biophysically marginal or favoured land, the Panel elected to use this term in Chapters 3 to 6 in examining options open to the CGIAR to address rural poverty through change in: policies and institutions; new and improved technologies and diversification of land use and income sources in MAs.
The Panel considered the forces driving the processes which retain people in poverty on MAs, in terms of the inter-relationships between the resources and knowledge available to these people and their incentive to use them for sustainable enhancement of livelihoods. This relates to the question of constraints to access of the poor to knowledge, natural resources, markets and off-farm income opportunities (including the options to migrate). Constraints are generally regarded as rooted in institutions and policies.
Marginal areas, because of characteristics such as their isolation, perceived low and risky productive potential, and the insignificant economic and associated political power of their inhabitants, typically have been neglected by central governments. As a result, only limited public investment has been made in education, health, infrastructure, etc. Little interest has been shown in determining the aspirations of marginal people or their knowledge of how to cope with harsh environments as a basis for focusing public action responsive to their capabilities and needs. In view of the high costs of providing quality services to such areas, it has been argued that accelerated development in other sectors -non-marginal rural areas, industry, services - would be sufficient to attract migration on a scale which would increase the resource base (and income) per capita for the residual population.
The "trickle down" theory of development and poverty alleviation in most marginal areas has not worked. Populations in many are increasing in absolute numbers in spite of migration, and increase is likely to continue, probably at a decreasing rate, for some decades. Without investment in the "resource base" of these people, with expanding populations the expectation can only be progressive extension of poverty and degradation of the soils and forests on which they depend for a large part of their sustenance.
In agriculture, inappropriate research has been blamed for not taking into account indigenous knowledge and the opportunities and constraints which apply to the site-specific characteristics of MAs. Thus, it is asserted that few research results have led to widespread or significant improvement in the welfare of marginal people. This is undoubtedly true in many instances but is merely one symptom of the underlying cause of marginality, i.e. the institutional and policy arrangements. Without any negotiating leverage by marginal communities on the decision makers responsible for the amount and type of investment in MAs, these investments tend to be small and poorly adapted to the requirements of the people. This applies equally to agricultural research and extension investments (and other areas such as education, health or infrastructure) which have been unable to respond to complex diverse requirements for sustainable exploitation of opportunities deriving from natural resources and value adding micro-enterprises.
However, the problem is not technology as such. Rather, it is the institutional arrangements which determine: what and how agricultural research and extension is carried out, together with the whole range of support services (credit, roads, communications, schools, etc.); resource entitlements; the functioning of markets; and local capacity to manage their own affairs9. A matter of concern is the breakdown of common property systems in some areas where they comprise an important part of the risk-sharing and survival strategies of poor farmers and herders on lands with highly variable rainfall (see Box 1).
9 Jodha, N.S. 1995. 'Enhancing Food Security in a Warmer and More Crowded World: Factors and Process in Fragile Areas'. In Climate Change and World Food Security (ed. T.E. Downing).
It is evident that the institutions and policies which drive the poverty process are country-specific. Nevertheless, the CGIAR appears to have a role to play, albeit small, in helping to introduce poverty alleviation strategies for the poor living in MAs. Among the principal avenues open to the CGIAR are those already outlined in the TAC report on policy and management research10.
10 TAC. 1996. 'Perspectives on Policy and Management Research in the CGIAR'. Document No. SDR/TAC:IAR/95/26.1. Rome.
The Panel is fully aware that by discarding the biophysical criteria for defining ML or MAL, it is violating the spirit of the terms of reference. By redefining the issues in terms of MAs which may occur on any combination of FAL and MAL, the issue of poverty alleviation is being placed in a global context, rather than being confined to marginal lands and the rural poor who derive a large part of their livelihoods from them, and who a priori are believed to have been by-passed by the Green Revolution. Nevertheless, it is sustained that this is merely another way of cutting the pie. The key question is still where are the concentrations of poor who have been by-passed by CGIAR research. The criterion for distinguishing the by-passed from the non-by-passed rural poor is different. It carries the implications that conditions for marginality of people also exist on FALs as is implied in Table 2.1, e.g. landless and land poor (farm size issues).
|
Box 1: Common Property Resources and Marginal Lands Rural common property resources (CPR) described simply are community resources where every member of the community has access and reciprocal obligation to help in their regulated use, without any exclusive private right to claim and sale any part of the resources. In the developing countries they include community grazing lands, forests, watering points, watersheds and (in a de facto sense) most of the public lands not assigned by the state for private use or public reserves. Much of these latter areas may be considered de facto "open access", i.e., there is no management by a community, individual or the state. CPRs play an important role in: consumption by the rural poor as a source of supplies (e.g., fuel, fodder and food); incomes through collection as well as processing and sale of wild products - especially during the seasons and years when crop incomes are not available; and enhancing nutrient availability to draft animals needed for crop farming. In some areas rural poor derive as much as 30 percent of their consumption from CPRs. Moreover, in villages with productive CPRs, they provide crop income as well. Thus, CPR-PPR (private property resource) based activities have significant complementarity. However, in most countries due to institutional, policy and demographic changes, CPRs are under severe stress. Area has declined due to legal or illegal privatisation and with weakening of the culture of group action productivity has deteriorated. In practice, CPRs have become open access. This suggests research aimed at rehabilitation of CPRs with two inter-related components: (i) technology oriented measures (in terms of germplasm, agronomic practices adapted to communal management) and (ii) institutional arrangements (which could help in promoting group action and better regulated use, including lands currently in open access exploitation. Research focused on enhancing the range of options for the poor in MAs should include the CPRs. Similarly, local level initiatives through participatory research could address the institutional dimension of CPR rehabilitation and management. The above possibilities can be incorporated into the Systemwide initiative on property rights and collective action coordinated by IFPRI. As a part of this initiative, the question of selective privatisation of CPRs can also be examined. Especially in the African context, there is a push for land-titling supported by the donors. The justification is to ensure proper incentives for land users and create collateral for credit for the farmers using common lands. The CPRs - due to their wider spread in terms of access and spatial location, offer multiple options to the poor to cope with risk. By promoting individual land entitlements, the policy may tend to restrict the range of options and access to resources and opportunities available to the poor. There is room for research to examine the trade-offs. |
One way to reduce the scope of a "marginal lands" study in the initial stages would be to exclude any consideration of MAs on the 800 million ha classified as FAL. One might justify this on the grounds that these are the areas which most clearly have benefited from the Green Revolution - if there is residual poverty on this area, this could be addressed as part of a global rural poverty study.
However, the conclusion was that, because of:
- heterogeneity of ML and FL (or MAL and FAL) in landscapes, the likely overriding importance of institutions in defining MAs,- the linkages between MAs and non-MAs in poverty alleviation within MAs,
- the critical importance of income in poverty alleviation which dictates a research focus on: food and non-food products (and associated value added); opportunities and constraints to off-farm sources of earnings and productivity from forest areas, coastal zones and savanna woodlands (not incorporated in the definitions of FAL and MAL), and
- the likelihood that definition of MAs on non-favoured areas will probably spill over into definition of MAs on FALs, again because of country or region-specific institutions driving the poverty process,
it would be more efficient to specify MA without regard to an arbitrary distinction between biophysically favoured and marginal lands. Given the existing mix of work in the centres (see Chapters 3-6) it is not clear the extent to which this explicit focus on poverty alleviation would imply changes in the actual research project work done by the centres.
With the above in mind, the Panel puts forth the following first recommendation:
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 with the processes that produce it (i.e., the dynamics of poverty).
In order to meet this challenge, the System, and individual centres within it, will need to develop a database relevant to the design and assessment of options that lead to more effective impacts on income of the rural poor in highly diverse marginal areas. Elements of this database will be the soil, climatic and terrain conditions. However, as the centres have pointed out in their comments on the Phase I report, there are many other elements which are equally or more important - production systems, human capital, market access and infrastructure, institutional and policy constraints and the number and location of the rural poor with respect to any or all of the above characteristics. A logical place to start this activity would appear to be the centres' expertise on the geographic areas with which they deal; perhaps initially by developing an intuitive Geographical Information System (GIS) which can be progressively improved by formal or informal methods. The issue of site specificity, diversity and complexity must be addressed to identify points of entry which are relevant and consistent with the CGIAR's scale. The question here is the extent to which centres may move towards targeting site-specific (or through a typology, situation-specific) rural poverty through research which meets the "international public goods" criterion.
Meeting this recommendation is regarded as a prerequisite to action on the three recommendations put forward in Chapter 3. The Panel is convinced that the centres' scientific expertise, coupled with their field experience in most, if not all, of the diverse MAs, represent a unique resource in coming to grips with the information/analytical challenge. Thus the work should be undertaken in a coordinated fashion by the centres themselves. IFPRI and ISNAR could take leading roles in providing support to the activity. An opportunity exists to initiate this process in the up-coming MTP exercise.
