link_rural_poverty

8/7/2019 link_rural_poverty

1/13


2/13

environmental problems can have important implications for understanding the links.

Third, the environment literature does not usually treat poverty measurement issues -- level, distribution(over households), and time path (whether it is transitory or chronic), and how these can affect theenvironment/poverty links. We show that for environment/poverty analysis, these poverty measurementissues matter, and that it is inadequate to limit the measurement of poverty to "welfare-poverty", measuredaccording to income, consumption, or nutrition criteria as is common in the poverty and food security

literature. Rather, we argue that the criterion for poverty in environment/poverty analyses should be theability to make minimum investments in resource improvements to maintain or enhance the quantity andquality of the resource base, to forestall or reverse resource degradation. A household below this line weterm "investment poor" to differentiate it from being "welfare poor". Households or villages above a welfare-determined poverty line may still be too "investment-poor" to make e.g. needed soil conservationinvestments.

Fourth, the strength and symmetry of the causal links between poverty and environment are rarelydiscussed in the literature. Is poverty alleviation necessary or sufficient to redress environmental problems?Does redressing environmental problems alleviate poverty? Under what circumstances? The questions arecritical for policy formulation and sequencing. We show that the answers depend on the type and level ofpoverty and the type of investments required to address particular environmental problems. Moreover, thedynamics of the links are relatively neglected in the literature, apart from recent treatment of population-environment dynamics (e.g., Dasgupta and Mler, 1994). But the links are likely to change over time, and

causal directions may reverse in unexpected ways. For example, the poor may "mine" the soil throughintensive cropping without accompanying investments in soil conservation, and then use the profits todiversify their incomes away from risky agriculture to protect their medium-term food security; this strategymight in the longer run reduce pressure on the land.

Fifth, insights from the literature on farm household economics and household food security strategies (e.g.,Ellis, 1993) have not been brought sufficiently to bear on understanding environment/poverty links. Yet ruralhousehold and village income, land use, and investment strategies determine the links between environmentand poverty. In turn, this behavior is conditioned by factors such as price and interest rate policy, villageinfrastructure, and technology -- hence the practical link to policy and research strategies.

To address these shortcomings, a new conceptual framework is needed to move the debate on this link intoa more practical and strategic domain. Such a framework should: a) differentiate types of poverty and typesof environmental change into categories useful for guiding food security, poverty, and environment policies,

and monitoring the effects of policies and projects. Such differentiation is taking place in the debate aboutthe links between general economic growth and agricultural growth on one hand, and environment on theother (see e.g., de Haen, 1992), and the debate about environment and poverty should be deepened andenriched in the same way; b) show the household behavior determinants of the links, and how this behavioris conditioned by policy and other factors.

We present such a conceptual framework as a flow diagram in Figure 1, relating four blocks of variables: a)categories of assets of the rural poor; b) household behavior pertinent to environment/poverty links (incomegeneration, investment, and land use); c) categories of natural resources (soil, water, etc.); d) conditioningvariables (market conditions, etc.). There are several links between the boxes: the asset categories ofpoverty affect household and village behavior, which in turn affects the (quality and quantity) of naturalresources as well as household/village assets. The conditioning variables influence the links between thetypes of poverty and behavior, as well as the links between behavior and natural resources.

We proceed as follows. Section 1 presents typologies of poverty and of the environment. Section 2discusses behavior of households and villages (in particular, income, investment, and land use strategies).Section 3 links the typologies of poverty and of environment with household/village behavior, withillustrations from various agroecological zones. Section 4 discusses the conditioning variables. Section 5summarizes and draws policy implications. We limit our discussion to rural areas, and focus on resourcedegradation because it is the major environmental problem in most poor countries.

TYPOLOGIES OF POVERTY AND ENVIRONMENT RELEVANT TO THEIR LINKS


3/13

Poverty

Typology

We begin by examining the asset portfolio of the rural poor, asking the question "poor in what?" Householdsand villages have the following assets: a) natural resources, composed of water (ground and surface),ground cover and its (bio)diversity (trees, bushes), wild fauna and flora, and soil/land; b) human resourceendowment, composed of education, health, nutritional status, skills, and number of people; c) on-farmresources (livestock, farmland, pastures, reservoirs, buildings, equipment); d) off-farm resources (local off-farm enterprise capital and migration activity capital); e) community-owned resources such as roads, dams,and social institutions. (See Figure 1.) These assets (stocks) are used to generate flows of product and/orcash income. The level and composition of that income determines whether households are poor and howpoor they are.

An asset decomposition of poverty will be useful only where the fungibility of assets is limited. Theexistence, conduct, and performance of labor, capital, and product markets conditions this fungibility, whichin turn determines how easily households can convert one form of wealth to another (say household laborinto farm capital, or land into cash).

Where markets are absent or underdeveloped (even just seasonally), or where there are constraints tomarket access (tied to resource endowments), one asset market or holding can be isolated from another.Asset-specific poverty can influence livelihood activities and investment decisions with possible adverseimplications for some components of the environment. For example, where a family is poor in land, but landis the collateral to get credit, the household's ability is limited to acquire cash to make farm and nonfarmcapital investments, as well as land improvements to protect soils. This in turn limits its ability to diversifyinto nonfarm activities or to intensify farming. The household then is left to resort to extensive farming,implying the need to push into fragile lands.

Criterion: "welfare poverty" versus "investment poverty"

The literature treating households suffering from what we term "welfare poverty" uses criteria based onincome, consumption, and nutrition, e.g. based on a benchmark minimum income sufficient to attainminimum caloric intake (below which one enters "ultra-poverty" according to Lipton, 1983) or to meet ananthropometric standard (Lipton and van der Gaag, 1994), or to buy a diet just sufficient given a regionaldiet level and composition (Greer and Thorbecke, 1986). These measures may be appropriate for assessing

human misery, but may not be the appropriate benchmark for use in assessing poverty levels in the contextof analysis of poverty-environment links. Purely "welfare-poverty" criteria can miss the potentially largegroup of households that are not "absolutely poor" by the usual (consumption-oriented) definition, but aretoo poor -in that their surplus above the minimum diet line is still too small- to make key conservation orintensification investments necessary for their land use practices to not damage the resource base or leadthem to push into fragile lands.

Rather, for analysis of poverty-environment links, we suggest the use of a measure of "investment-poverty",the cut-off point for which we define as the ability to make minimum investments in resource improvementsto maintain or enhance the quantity and quality of the resource base -- to forestall or reverse resourcedegradation. Unlike welfare measures of poverty, the cut-off point for investment poverty is site specific, afunction of local labor and nonlabor input costs, and of the types of investment that are needed for theparticular environmental problems or risk faced.

Of course, the welfare-poor are usually also investment-poor. The converse (that the investment poor arenecessarily welfare poor) is not necessarily true. Hence anthropometric and welfare poverty maps would notnecessarily detect the kind or level of poverty that may be important to environmental links. If a household isabove the "welfare-poverty" line, it can still be "investment-poor" in three situations: a) if the marketconditions are such that the household cannot convert its assets or products into enough cash to makeconservation investments, or b) the household can obtain the cash but cannot buy labor or other inputsneeded for conservation measures because their supply is constrained, or c) while household income maybe somewhat above the welfare-poverty line, it is not sufficiently above it to generate an adequate surplusfor investment.


4/13

Moreover, in the long-run, if a household is investment-poor but not welfare-poor, it may lead to naturalresource degradation that eventually causes the household to become welfare-poor -- the vicious circle canbe realized. Also, a household might be above an investment-poverty line on average but have severelyunstable income and thus be more averse to risky investments in land improvement.

Typology of environment

The natural environment can be decomposed into the following (ecologically interdependent) categories (inFigure 1): soil, water (ground and surface), biodiversity (of flora and fauna), and air. The overlap can beincomplete between assets owned by the rural poor and the components of the natural resource base, andthis can affect poverty/environment links.

Depending on which attributes rural households' value and what these values are, a particular naturalresource (a household asset) may be either harvested, left undisturbed, or even enhanced. Yet someattributes of particular natural resources may not matter to the rural poor, at least in the short-medium run,but may greatly concern policymakers monitoring natural resource stocks -- that is, be valued differently andseen as serving different purposes when viewed from the perspective of the society in general and that ofthe rural poor. For example, a large tree can be a valuable "carbon sink" to society, but the tree may beviewed by the rural poor just as a source of ash to enhance soils and as timber for sale. This is part of thegeneral problem of negative environmental externalities of a particular farm household's behavior to the restof society. Society may even view a household as rich in a resource that society values (e.g. biodiversity)

but this resource may not be considered valuable (at least under current circumstances) by the poorhousehold, since markets for converting social value into private wealth may not exist.

BEHAVIOR OF THE POOR: THE DETERMINANTS OF ENVIRONMENT-POVERTY LINKS

Figure 2 shows the poor household's resource allocation to income earning activities and investment and theenvironmental consequences of this allocation, in a simple two-period dynamic framework. a) Thehousehold starts with a set of household food security and livelihood objectives. b) In period 1 the householdhas access to a set of natural resources, human capital, and on-farm and off-farm physical and financialcapital. c) The household is also faced with a set of "external conditioning factors" (prices, policies,technologies, institutions, community assets). d) The household allocates its labor, land, and capital toincome-earning activities and investments in two sectors, agriculture and nonagriculture. e) The livelihoodactivities and investments have environmental consequences on-farm and via externalities, off-farm. f) Theactivities and investments, plus the environmental consequences, alter the household's access to resources

and capital, hence the household has a new stock of assets as it enters period two, and begins theallocation again.

Note that the level and type of poverty (in terms of asset categories) affects what income and investmentstrategies the household follows, and how it carries them out. Below we discuss these strategies and howthey can affect environment/poverty links.

Income strategies

To manage risk (ex-ante) and cope with loss (ex-post), faced with the high risk that typifies agriculture inenvironmentally-fragile areas, the rural poor earn income from a variety of sources, shown in Figure 2: in thenonagricultural sector (in local and migration activities), in gathering (local flora and fauna), and in farmingand livestock husbandry. But the poor's strong incentive to diversify income is not always matched bycapacity, which can be undermined by their asset poverty. Below we examine what the poor do and how

they do it, and posit how this affects the environment.

Nonagriculture

The rural poor in fragile areas diversify incomes and assets into the nonagricultural sector to manage riskand cope with cropping shortfall, as well as alleviate poverty. When the poor engage in nonagriculturalactivities, they are usually labor-intensive, with little use of capital, thus low entry barriers. Such activitiesinclude e.g., small commerce, portage, farm labor, unskilled labor for construction, in addition to long-haul orseasonal migration. Yet there can be an inverse relation between the fragility of the resource base and theavailability of off-farm activities. The latter tend to be plentiful where they spin off from a dynamic agriculture


5/13

(rural growth linkages) or urban-rural exchange. (Matlon, 1979; Haggblade et al., 1989; Reardon et al.,1994) Thus, environmentally-fragile areas that might benefit from increased off-farm activities may be lesslikely to create opportunities for them.

Diversification of activities and assets into the nonfarm sector can have implications for the environmentdepending on the nature and the severity of the environmental problem, and the nature of investmentsrequired to address it. The effect can be either positive or negative. On the one hand, where the poor

successfully diversify into nonagricultural activities, they can become less directly dependent on land, henceless vulnerable to land degradation. Cash from nonfarm activities can also be used to finance soilconservation investments and use of fertility-enhancing inputs. (See Reardon et al., 1994b for review ofevidence.) On the other hand, resource conservation investments or allocation of labor to natural resourcemanagement may not be a priority use of investment funds or own-labor for the poor household.Investments in off-farm activity can compete with investments in land conservation, a competitionexacerbated by the poor's lack of investable cash or available labor beyond their subsistence needs.

Whether diversification away from cropping or from livestock husbandry leads to less resource degradationalso depends partly on whether past agricultural activities started a "downward spiral" of degradation (e.g.,removal of top soil and bush cover) that would not naturally be reversed without resource conservationinvestments. But if no critical threshold has been crossed (overstressing the ecosystem's resilience) thenreduction in pressure on the land from income diversification might reduce or arrest degradation of theenvironment. Hence, both the degree of degradation and the degree of poverty matter to poverty-

environment links.

Livestock husbandry and wild food/fuel gathering

When the poor engage in small livestock husbandry and gathering of wild food and fuel, they do so mainly inthe commons or in open access lands, and the activities are usually labor-intensive and require little capital.In relative terms, income in cash or in kind from these activities tends to be more important to the poor thanto richer rural households in their diversification and coping strategies. (Jodha, 1992; Webb and Coppock,1992). This greater importance makes the rural poor more vulnerable than the rich to the disappearance ofedible/marketable wild fauna and flora, and to removal of bush cover (but not necessarily biodiversity perse).

This greater reliance on livelihood activities based on use of the commons and on open-access lands hasoften led observers to blame the poor for overgrazing and overforesting open access lands. Livestock

indeed are important to the poor, but the poor household usually cannot afford to own many animals. Forexample, the poorest tercile of households owns far fewer animals per household than do richer householdsin West Africa; see Christensen, 1989. Thus individual poor households put less pressure on semiaridpasturelands than do individual rich households.

The conventional argument is thus turned on its head: as absolute importance of livestock holding increaseswith household income, and as incomes rise in rural areas, we should expect households to invest in morelivestock and place greater pressure on the commons. Analogously, reducing poverty in tropical forest areassuch as the Amazon may induce technical change (e.g. adoption of chainsaws) in forest conversion evenamong small farmers, thus increasing deforestation rates.

Cropping

In general, poverty has the most direct effect on the environment via cropping where poverty influences the

household's technology and investment path in intensification of cropping. Boserup, 1965 outlines a numberof technology and investment paths to agricultural intensification that farmers follow in the wake of increasedland constraints and demand for land -- conditions that result from population growth, increased demand foragricultural products, or reduced transportation costs (Boserup, 1965; Pingali et al., 1987). Two broad pathscan be distilled from Boserup's framework: a) a labor-led intensification path where farmers merely add laborto the production process on given land, allowing them to crop more densely, and weed and harvest moreintensively; b) a capital-led intensification path where farmers augment their labor with variable and capitalinputs, in particular fertilizer, organic matter, and capital that facilitates land improvement. Boserup identifiesthe second path as having higher land productivity than the former. Similarly, Matlon and Spencer, 1984note that the capital-led path is more sustainable and productive in fragile, resource-poor areas as the


6/13

fertility-enhancing input use helps the farmer to avoid exhausting the soil during intensification and thecapital (land improvements) help avoid erosion and runoff. By contrast, in much of the African tropics forexample, the labor-led path to intensification is unsustainable, and leads to land degradation and stagnationof land productivity. In situations such as the tropical highlands where demographic pressure anddegradation are severe, farm households that follow only the labor led path are in for long-run ecologicaldisaster and further immizeration. (Matlon and Adesina, 1992; Cleaver and Schreiber, 1994).

Hence, in situations of fragile and degrading environments and land constraints, hence lack of opportunity toextensify, households too "investment-poor" to make the requisite investments for the "capital led"intensification path will find themselves both increasing the rate of degradation and vulnerable to itsproductivity consequences. On the other hand, where there is still sufficient extensive margin (such asforest), these investment-poor can convert forest to farmland, and offset land degradation with more landunder cultivation -- until they run up against the end of the land frontier.

Illustrations from Rwanda show the divergent tendencies inherent in the poor's cash and land constraintscoincident with their reliance on the land. Poorer households participate less than do richer households incoffee farming, partly because of land quality and input requirements. They fallow less because of pressureto crop all available land to meet food security needs. Yet poor farmers crop more intensively than do larger,wealthier farmers, protecting the soil from hard rainfall through denser planting. Small and poor farmers tendto invest more in resource conservation measures (e.g., terraces) when they have cash available from off-farm work. (Clay et al., 1995)

Moreover, being poor can lead a priori to a distribution of assets skewed toward certain types and qualitiesof natural resources. Sometimes poorer families are constrained to take the most fragile land or to clearforest and work with easily eroded land. Sometimes they are relegated to the end of irrigation canals andsuffer the combined residual silting problems of the other users (Yudelman, 1989)

Investment strategies

Investments (such as in land improvements) embody land use strategies to enhance the environment, andinvestments in on-farm and off-farm assets affect crop choice and the type and rate of agriculturalintensification. We focus on land improvements (such as bunds, terraces, windbreaks, culverts), becauseland is generally the most important natural resource available to poor rural households.

Land improvement investments are determined by the following three things.

First, Incentives specific to the household, which include: a) net returns of the investment, which depend onyields and input requirements per unit, and the prices of inputs and outputs; b) riskiness of the investment --which includes short- to medium-term risk associated with price and yield variability, and long-term riskstemming from political and policy instability, and insecure land tenure; c) relative returns and risks --compared to alternative farm and nonfarm investments; d) the household-specific discount rate, proxied byhousehold characteristics that heighten the importance of immediate survival to increase the discount rate;the poorer the household, the more immediate survival counts, for example.

Second, Capacity to invest, specific to the household, which include: a) the four categories of assetsdiscussed above, and the flow of cash earned from them; b) "complementary assets" on-farm (such as awell that provides water to maintain a live windbreak, the latter being the investment in question, and thewell being the on-farm complement).

Third, "External conditioning variables" common to households in a particular agroclimatic/policy context,which include: a) technologies for production and input or output processing affect the set of availableinvestments and their profitability and riskiness; b) agricultural and macroeconomic policies, which affectinput and output prices; c) institutional environment (legal system including land tenure customs and laws,markets, extension services). The economic and institutional environment affects not only the household'sobjective function (e.g., how much food to grow versus to buy) but also the feasibility and relative profitabilityof nonfarm and farm activities (intersectoral terms of trade), through its effects on output and input prices(via output, input, credit, land, and labor markets), on the type of technology available, and on access toresources; d) the physical environment (soils, rainfall, temperature, diseases and pests) affects the technicalfeasibility of potential investments, and their profitability and riskiness; e) transport and communication


7/13

infrastructure determines the availability of information, access to markets, and costs and returns ofinvestment; community level watershed management infrastructure (such as dams, culverts, and farm-levelbunds) are key complementary investments. (Christensen, 1989; Reardon and Vosti, 1992; Reardon et al.,1995).

The level and type (by asset category) of poverty affect the determinants of investment. The assets of poorhouseholds tend to be mainly labor. Investments that require inputs other than labor, or that are large or

"lumpy" are less likely to be undertaken by the poor. The movement of a household from the category ofbeing poor in an asset (such as land) to being "investment poor" in general (not being able to make specificinvestments required for natural resource maintenance or enhancement), depends on the level of risk (fromprice and rainfall instability, or from insecurity of land tenure hence risk of appropriation of capital), on thenature of markets, and on whether these translate into lack of sufficient liquidity (to buy labor or materials forcertain land improvements), and on the willingness to use the liquidity for the investment in question.

ILLUSTRATIONS OF LINKS BETWEEN TYPES OF POVERTY, THE POOR'S INCOME ACTIVITIES ANDINVESTMENT, AND THE ENVIRONMENT

Below we illustrate the above discussion with three cases.

The poor in tropical rainforests of Brazil: rich in biodiversity, and poor in soil, financial assets, and labor

Poor farmers in the Amazon region have access mainly to forest cover and biodiversity, both of which arelocally abundant but globally scarce. Markets do not generally exist to translate global demand for theseresources into secure income streams for farmers. The poor are also constrained by lack of soil nutrients, oflabor, of good health, and of cash to meet food production and purchase needs. Poverty in all assetcategories but forest cover and biodiversity lead farmers to convert forests into farmland by burning trees togenerate nutrients to enhance soils. That is, farmers use biophysical processes rather than markets toconvert plentiful assets into inkind income flows to meet food security needs. To break this cycle, the poormust move away from biophysical processes for converting their principal asset (forest cover) and reducedependence on that asset. This diversification will require markets for forest cover and biodiversity, newtechnologies to increase the productivity of labor and of alreadycleared land, and alternative offfarm incomesources.

But much of the past deforestation in the Brazilian Amazon had little to do with poverty, though thoseresponsible for future forest conversion and their motives might be quite different. Rather, in response to

fiscal and other incentives, large landholders using sophisticated technology laid waste to millions ofhectares of tropical forest. While the decomposition of assets and the environment, and households' andcommunities' objectives and constraints regarding income and asset smoothing proposed in this paper shedlight on the motives behind such forest conversion activities, clearly, these actors had graduated wellbeyond any reasonable definition or type of poverty. (Hecht, 1984).

The poor in the Sahel: Rich in land area, and poor in land quality and on-farm and off-farm physical andfinancial assets

Sahel lands are fragile and degrading. Despite increasing constraints on land, the poor have relatively goodaccess to rangeland and cropland (though of poor carrying capacity), but are relatively poor in livestock andoff-farm income. Thus they are relatively dependent for food security on their low quality farmland and onopen access lands for gathering and small scale animal husbandry in the commons, which are over-grazed.Note that the responsibility for overgrazing is weighted toward richer households that own much more

livestock than do poorer households. Moreover, there is a need to intensify cropping to reduce pressure onthe fragile margins, but poor households lack cash and credit to purchase nonlabor variable inputs. Poorcommunities lack complementary infrastructure (such as roads or depots) to obtain these inputs. (Stryker,1989; Lele and Stone, 1989; Matlon and Adesina, 1992; Reardon et al., 1994 and forthcoming)

Labor, the poor's chief asset, is a key input in the dry season for bund construction (a key land improvementto protect fields against flashfloods and runoff over the hardpan surface). Capital to augment their labor(carts or trucks to haul rocks or laterite for bunds) is expensive. There is little investment in rock bundsexcept where government-supplied trucks have permitted. Labor is not easily substitutable for these key


8/13

materials or equipment.

The poor in the Rwanda highlands: rich in labor, poor in land and off-farm physical and financial assets

In Rwanda there is intense pressure on the land, and the poor depend on micro plots to survive, and havefew animals. There is very skewed distribution of nonfarm income. Where the land-poor are also poor in off-farm capital, they can make few soil conservation investments because of lack of cash for materials andlabor hire. Those with cash crops such as coffee or with nonfarm income have both the incentive (relianceon little land) and the cash to make conservation investments. Moreover, the poor practice labor-ledintensification (see section 3), and lack the means to buy fertilizer or mulch, and own few animals togenerate manure. Limits to yield increases are reached early in such a system, and the soil can beexhausted from lack of amendments such as fertilizer. This gives rise to a vicious circle of poverty, labor-ledintensification, degradation, and more poverty. (Clay et al., 1995)

CONDITIONING VARIABLES

The income and investment strategies of poor households, hence the links between poverty andenvironment, are conditioned by the following factors, which are relevant to policymakers.

First, the existence, structure, and performance of markets condition the prevalent type of poverty, as well

as the substitutability of assets in investment and income generation, as discussed in section 1.

Second, production and resource conservation technologies embody the substitutability among assets inboth production and investment. Changing technical rates of substitution among assets, especially betweenhuman-made assets and natural resources (Ruttan, 1992), can alter household decisions and environmentalconsequences. For example, technology change in agriculture that increases land productivity can helprelieve pressure on surrounding forests and hillsides.

Third, relative input prices, output prices, wages, and the interest rate affect farm resource use andinvestment. Raising the price of a low-valued asset (e.g. biodiversity) can alter its use as well as reduce theincidence of poverty. Complementary "hard infrastructure" (such as culverts, dams, wells, market facilities,and roads) and "soft infrastructure" (such as extension, schools, and medical services) at the village levelaffects the cost of transactions and of inputs and outputs, and thus private costs of investment in resourceconservation. Infrastructure also influences the development of nonfarm activities, the commercialization of

agriculture, and urban-rural links, which are important determinants of income opportunities for the poor.

Land improvement technology and input price changes together affect the cutoff point for "investment-poverty". For example, a welfare-poor household might not be investment-poor if it can use its (relatively)abundant labor to build bunds from easily-available local materials. But say that bunds need to be built onlywith laterite pieces that are far from the farm, requiring trucks to fetch. Then the labor-rich household isinvestment-poor. A change of technique that allows bunds to be built with local materials would reverse theinvestment-poverty, as would a decrease in the cost of transporting bund materials from afar. Conversely, amacroeconomic change such as the recent devaluation of the franc CFA in Francophone West Africacurrency increases transport costs which may drive some households into investment poverty by making thecosts of certain investments out of reach.

Fourth, community wealth (physical and social assets) conditions the poor household's options and behaviorin three ways: a) Community wealth affects insurance and wealth distribution mechanisms. Intra-community

distribution of wealth also conditions the level and incidence of poverty as well as its effect on theenvironment. The distribution of poverty in the community can affect the cost faced by and the effectivenessof redistribution and joint-investment institutions in the community. The distribution is governed by rulesrelated e.g. to land tenure and inheritance rights, and to credit collateral. Land tenure institutions caninfluence the perceived incentive and risk facing the poor weighing the decision to invest in landimprovement. b) The community may be forced to dedicate resources for (welfare) insurance to the welfare-poor, which means diversion of resources from production and from resource conservation investments, aswell as from enforcement of watershed management. On the one hand, the community can compensatewelfare-poverty with community insurance mechanisms (e.g., interhousehold transfers) and throughtransfers can mitigate investment-poverty. These institutions and investments can help the poor to avoidcoping activities or resource mining that degrades. Shared poverty can make the enforcement difficult, as


9/13

can very poorly distributed income. Property rights set and enforced by communities are intimately linked tothe externalities of farming and herding that affect the environment of the commons. (Schmid, 1987). c)Community investments in infrastructure influence two key things related to the incidence of "investmentpoverty". On the one hand, infrastructure -- both physical and social -- affects economic opportunities for thepoor by its effect on demand levels for the poor's farm products as well as products and services ofdiversification activities off-farm. On the other hand, community complementary infrastructure in awatershed, such as in culverts and dams or trucks or carts to transport bund materials, affect the cost of

natural resource management investments by the poor. These community and government complementaryinvestments can make private land improvement investments more affordable to the poor.

The level and distribution of wealth in the community can also affect the enforceability of regulations tocontrol access to the commons.

Fifth, the population growth rate determines pressure on the land and fuels degradation if there are notincome alternatives or technical change available to relieve the pressure. Human fertility decisions at thehousehold level, aggregated over households, determine this conditioning factor. (Lipton, 1992; Das Guptaand Mahler, 1994)

CONCLUSIONS

Not all environmental degradation in developing countries is linked to poverty; for example, pollution as anexternality of the agriculture of richer farmers or forest or commons over-exploitation by large and capital-intensive lumber and cattle operations can ravage the environment without the poor's lifting a hand. Butwhere there are links between poverty and the environment, they are often complex and to address them ischallenging. To date, efforts to analyse the link between poverty and environment have been too general onthe poverty side, too general on the environment side, and thus have not been able to sort out seemingly-conflicting evidence. This paper addressed this issue by: a) decomposing poverty into asset categories; b)decomposing the physical resource endowment/environment; c) showing how poverty types and levelsaffect household livelihood activities and investment decisions, which in turn affect the environment; d)showing what factors (external to the household) condition the links.

Several conclusions emerge regarding the poverty-environment links, referred to as "links" below.

First, the level of poverty conditions the links. Households can have incomes above an established welfarepoverty line -have enough to eat, have good health- but still be too poor in key assets and thus overall cash

and human resources to be able to make critical investments or follow key land use practices to maintain orenhance their natural resource base. They might thus be better off than the "welfare poor" but still be"investment poor"; the former is usually also the latter. In any case the important measure for poverty in theanalysis of links is "investment poverty".

Second, the type of poverty affects the links. We defined "type" according to the asset category orcategories in which households are poor. The potential for substitutability among assets in incomegeneration and investment influences the relation between poverty in a given asset and household behavior;substitutability itself is conditioned by the existence and performance of asset and output markets.

Third, the distribution of poverty across households within a community affects the links. It does so directly,as well as indirectly through affecting the wealth of the community and its ability to manage risk, tocompensate the poor, and to build and maintain infrastructure that can help the poor to invest in naturalresource enhancement.

Fourth, the type of environmental problem conditions the links. Poverty and soil degradation interactdifferently from the way poverty and rural pollution do, or as poverty and loss of biodiversity do.

Fifth, the income, investment, and land use strategies of the rural household and community determine thelinks. Poverty affects what kinds of activities and investments are undertaken, and in what way, and thushow natural resources are used and enhanced.

Several policy implications and research gaps emerge from the conclusions. Conditioning variables -


10/13

markets, natural resource use and conservation technologies, relative prices, infrastructure and institutions,the level and distribution of community wealth, and population growth- affect the poor's strategies. Many ofthese conditioning variables can be influenced by policy.

First the links between poverty and environment in a given setting depend on the level, distribution, and typeof poverty, the type of environmental problem, and conditioning variables. As these change over contexts,the direction of causality and the strength of the links can change. Given the diversity of types of poverty and

types of environmental problems, policy prescription will be site-specific, and the proposed conceptualframework can provide guidance in policy formulation and implementation.

Second reducing poverty can reduce resource degradation where poverty is driving extensification ontofragile hillsides or forests. But alleviating poverty will not necessarily lead to less resource degradationwhere the only insurance available is investment in more livestock, and insurance demand increases withhousehold income. Or alleviating poverty will not reduce pollution from overuse of agricultural chemicals, theuse of which increases with farmer wealth.

Third, enhancing the natural resource base can reduce poverty, where for example soil degradation isreducing yields on the farms of the poor. But conserving natural resources can also increase poverty, forinstance in cases where poor households are barred from gathering of wild flora and fauna but depend onthis as a key income strategy for survival.

The upshot of the second and third implications is that, in the short term, reducing poverty will notnecessarily protect the environment, nor will protecting the environment necessarily alleviate poverty.Specific policy action to affect the set of conditioning variables will be needed to maximize the achievementof both goals at once. Moreover, reducing poverty just to the point where households are over the "welfare-poverty" line may not be enough in many cases for households to be well-off enough to be able to afford keyinvestments in natural resource management; the latter require that households be above the "investmentpoverty" line, a line that depends on the costs and types of investment needed, and the composition ofassets of the poor household.

Fourth, the most effective way to simultaneously reduce poverty and enhance the resource base is tounderstand what categories of asset poverty and conditioning variables are driving households' behavior(e.g. degrading land use practices or lack of conservation investments), and focus effort on these. Promotionof key markets, investment in complementary infrastructure, and research to make resource managementtechnologies more productive and affordable are examples of such efforts. Policy should aim at affecting

household and community behavior with the aim of helping poor households to attain their main objective,food security, while as much as possible maintaining or enhancing the resource base.

Fifth, production and investment of households and communities can generate environmental externalities(both positive and negative). The latter can have a direct and large impact on the asset holdings of the ruralpoor's neighbors (who themselves might also be poor), thereby potentially reducing and changing thecomposition of their wealth, and promoting compensatory income earning activities and investment. Whilenot explicitly treated in this paper, our conceptual framework can be extended to incorporate theseexternalities, and policy actions to address them analyzed in that framework.

Sixth, a key gap in research is understanding how community and government complementary investments(in physical and social infrastructure) can make private land improvement investments more affordable tothe poor.

BIBLIOGRAPHY

Boserup E., 1995. The conditions of agricultural growth: The economics of agrarian change underpopulation. Aldine. Chicago.

Christensen G., 1989. Determinants of private investment in rural Burkina Faso, PhD dissertation. CornellUniversity. Ithaca, N.Y.

Clay D., Byringiro F., Kangasniemi J., Reardon T., Sibomana B., Uwamariya L., 1995. Promoting food


11/13

security in Rwanda through sustainable agricultural productivity: Meeting the challenges of populationpressure, land degradation, and poverty, Staff Paper No. 95-08. Michigan State University, Dept. ofAgricultural Economics, February 1995. East Lansing, Michigan.

Cleaver K.M. and Schreiber G.A., 1994. Reversing the spiral: The population, agriculture, and environmentnexus in Sub-Saharan Africa The World Bank. Washington, D.C.

Dasgupta, P. and K.G. Mler, 1994. Poverty, institutions, and the environmental-resource base, World BankEnvironment Paper, No. 9, The World Bank. Washington, D.C.

de Haen H., 1992. Environmental consequences of agricultural growth, In Vosti S., Reardon T. and von UrffW. (Eds.), Agricultural sustainability, growth, and poverty alleviation: Issues and policies. DSE, Feldafing,Germany.

Ellis F., 1993. Peasant economics: Farm households and agrarian development, Second Edition WyeStudies in Agricultural and Rural Development, Cambridge University Press. Cambridge, U.K.

Greer, J. and Thorbecke E., 1986. Food poverty and consumption patterns in Kenya International LaborOffice. Geneva, Switzerland.

Haggblade S., Hazell P., and Brown J., 1989. Farm-nonfarm linkages in rural Sub-Saharan Africa, WorldDevelopment, Vol. 17, no. 8, August 1989, pp. 1173-1201.

Hecht S.B., 1984. Cattle ranching in Amazonia: Political and ecological considerations, In: Frontierexpansion in Amazonia, M. Schmink and C.H. Wood (Eds.), University of Florida Press. Gainesville, Florida.

Jodha N.S., 1992. Agricultural growth and sustainability: perspectives and experiences from Himalayas, In:Agricultural sustainability, growth, and poverty alleviation: Issues and policies, S. Vosti, T. Reardon, and W.von Urff (Eds.), DSE, Feldafing, Germany.

Lele S.M., 1991. Sustainable development: A critical review, World Development, Vol. 19, No. 6, pp. 607-21.

Lele U. and Stone S., 1989. Population pressure, the environment and agricultural intensification: variations

on the Boserup hypothesis, MADIA Discussion Paper, No. 4 The World Bank. Washington, D.C.

Leonard H.J., 1989. Environment and the poor: development strategies for a common agenda, In:Environment and the poor: development strategies for a common agenda, H.J. Leonard et al., U.S.-ThirdWorld Policy Perspectives, no. 11 Overseas Development Council. Washington, D.C.

Lipton M., 1992. Accelerated resource degradation by Third World agriculture: Created in the commons, inthe West or in bed? In: Agricultural sustainability, growth, and poverty alleviation: Issues and policies, S.Vosti, T. Reardon, and W. von Urff (Eds.), DSE, Feldafing, Germany.

Lipton M., 1983. Poverty, undernutrition, and hunger. World Bank Staff Working Paper, no. 597, World Bank.Washington, D.C.

Lipton M. and J. van der Gaag (eds.), 1994. Including the poor. World Bank Regional and Sectoral Studies,World Bank. Washington, D.C.

Matlon P., 1979. Income distribution among farmers in Northern Nigeria: Empirical results and policyimplications, African Rural Economy Paper, no. 18. Michigan State University. East Lansing, Michigan.

Matlon P.J. and Adesina A.A., 1992. Agricultural growth and sustainability -- prospects for semi-arid WestAfrica, In: Agricultural sustainability, growth, and poverty alleviation: Issues and policies, S. Vosti, T.Reardon, and W. von Urff (Eds.), DSE, Feldafing, Germany.


12/13

Matlon P. and Spencer D.S.C., Increasing food production in Sub-Saharan Africa: Environmental problemsand inadequate technological solutions, American Journal of Agricultural Economics, volume 66, no. 5(December).

Mink S.D., 1993. Poverty, population, and the environment, World Bank Discussion Paper, no. 189. TheWorld Bank. Washington, D.C.

Pearce D.W. and Warford J.J., 1993. World without end: Economics, environment, and sustainabledevelopment, Oxford University Press. New York, New York.

Pingali P.L., Bigot Y., Binswanger H.P., 1987. Agricultural mechanization and the evolution of farmingsystems, Johns Hopkins University Press. Baltimore, Maryland.

Reardon T., Sustainability issues for agricultural research strategies in the Semi-Arid Tropics: Focus on theSahel, Agricultural Systems, forthcoming.

Reardon T., Crawford E, and Kelly V., 1994. Links between nonfarm income and farm investment in Africanhouseholds: Adding the capital market perspective, American Journal of Agricultural Economics, Volume 76,No. 5 December 1994, pp. 1172-1176.

Reardon T., Crawford E., and Kelly V., 1995. Promoting investment in sustainable intensification of Africanagriculture, Staff Paper No. 95-18: Michigan State University, Dept. of Agricultural Economics, March 1995.East Lansing, Michigan.

Reardon T., Fall A.A., Kelly V., Delgado C., Matlon P., Hopkins J., and Badiane O., 1994. Is incomediversification agricultureled in the West African SemiArid Tropics? The nature, causes, effects, distribution,and production linkages of offfarm activities, In: Economic policy experience in Africa: What have welearned?, A. Atsain, S. Wangwe, and A.G. Drabek (Eds.), African Economic Research Consortium. Nairobi,Kenya.

Reardon T., Matlon P.M., and Delgado C.L., 1988. Coping with household-level food insecurity in drought-affected areas of Burkina Faso, World Development, Vol. 16, No. 9, September, 1988), pp. 1065-1074.

Reardon T. and Vosti S.A. Issues in the analysis of the effects of policy on conservation and productivity atthe household level in developing countries, Quarterly Journal of International Agriculture, vol. 31, no. 4,October-November, 1992, pp. 380-396.

Ruttan V.W., 1992. Sustainable growth in agricultural production: poetry, policy, and science, In: Agriculturalsustainability, growth, and poverty alleviation: Issues and policies, S. Vosti, T. Reardon, and W. von Urff(Eds.), DSE, Feldafing, Germany.

Rwalinda P., Tardif-Douglin D, and Uwamariya L., 1992. Aspects de la caficulture au Rwanda: rsultats del'enquete sur la sensibilit-motivation des caficulteurs rwandais, Division de Statistiques Agricolespublication no. 25. Ministry of Agriculture, Division de Statistiques Agricoles, November 1992. Kigali,Rwanda.

Schmid A.A., 1987. Property, power, and public choice: An inquiry into law and economics, Second Edition

Praeger. New York, New York.

Spencer D.S.C. and Polson R.A., 1992. Agricultural growth and sustainability: Conditions for theircompatibility in the humid and subhumid tropics of Africa, In: Agricultural sustainability, growth, and povertyalleviation: Issues and policies, S. Vosti, T. Reardon, and W. von Urff (Eds.), DSE, Feldafing, Germany.

Stryker J.D., 1989. Technology, human pressure, and ecology in the arid and semi-arid tropics, In:Environment and the poor: development strategies for a common agenda, H.J. Leonard et al., U.S.-ThirdWorld Policy Perspectives, no. 11, Overseas Development Council. Washington, D.C.


13/13

Vosti S.A. and Witcover J., 1993. Gender differences in levels, fluctuations, and determinants of nutritionalstatus: Evidence from Southcentral Ethiopia, In: Effects of selected policies and programs on women'shealth and nutritional status, E. Kennedy and M. Garcia (Eds.), International Food Policy Research Institute.Washington, D.C.

Webb P. and Coppock D.L., 1992. Prospects for pastoralism in semi-arid Africa, In: Agriculturalsustainability, growth, and poverty alleviation: Issues and policies, S. Vosti, T. Reardon, and W. von Urff

(Eds.), DSE, Feldafing, Germany.

Yudelman M., 1989. Sustainable and equitable development in irrigated environments, In: Environment andthe poor: development strategies for a common agenda, H.J. Leonard et al., U.S.-Third World PolicyPerspectives, no. 11, Overseas Development Council. Washington, D.C.

link_rural_poverty

Documents