Contents - Previous - Next
This is the old United Nations University website. Visit the new site at http://unu.edu
4. Food poverty
of food poverty
Kinds of data
Case study: The importance of non-market entitlements
Policy recommendations to reduce food poverty
Summary and conclusions
Laurie F. DeRose
Food poverty here refers to household-level hunger. Households in food poverty do not have enough food to meet the energy and nutrient needs of all of their members. Depending on patterns of intra-household distribution, at least one member of a food-poor household is always hungry but, potentially, all members are.
This chapter first explores the causes of food poverty. The main issues in the international distribution of food are considered first, but more attention is given to the national distribution of food and why household entitlement to food sometimes fails. It then reviews the evidence for the distribution of food poverty across the world, types of evidence used to measure food poverty, and trends in the incidence of food poverty over time. The chapter then turns to the more piecemeal evidence available on which households within regions are the most likely to be affected by food poverty, evidence that provides further insight into how households succeed or fail in providing food for their members. To emphasize the importance of non-market access to food in protecting households from hunger, a case study of food-sharing networks is included. The chapter concludes with recommendations for reducing food poverty.
Causes of food poverty
Some households live under conditions of chronic or seasonal food poverty. Other households are pushed into food poverty because of changes in area food availability and/or in their own ability to secure entitlement to food.
Food availability in a region is one of the key determinants of the existence and extent of food poverty. In food-short regions, at least some households are food poor. But widespread food poverty persists in the absence of shortage, and many households escape food poverty despite severe food shortage.
Given no shortage of food on the global level, all incidence of food poverty can be attributed to maldistribution rather than underproduction. "Distribution" in this context includes transportation and storage, as well as production patterns and crop choices; it is not simply a matter of shuffling around existing food.
Reducing food shortage at national and sub-national levels is an important tool for reducing food poverty, but not simply because lack of shortage eliminates the necessity of food poverty. Productive and distributive mechanisms - not just food availability - change when there is food shortage. Famines, embargoes, and structural adjustment all cause shifts in the conditions of exchange that lead to entitlement failure (Ravallion 1987). Changes in food production affect at-risk households primarily through changes in income and prices, rather than as a direct result of food supply. Famine may also cause smallholders ordinarily subsisting on the margin to destitute themselves by selling land and other assets - acts that permanently shift the power and social structure of a society.
A special case of food shortage is violent conflict that reduces food availability and changes patterns of food distribution in affected countries. Food imports during times of violence are often restricted by embargoes. During both international and intranational conflict, governments put a high priority on provisioning the military, which tends to decrease civilian access to food. Although it is theoretically possible for local food production to increase enough to offset the food deficit caused by embargoes and diversion of existing supplies to the military, this usually does not happen quickly enough to avoid increased poverty. It is much more common for internal food production to decrease, because land has been abandoned and livestock sold by agriculturists seeking to avoid being plundered (Tschirley and Weber 1994). It is also quite difficult to expand production when economic and human resources are being devoted to the conflict.
Also, different households are likely to suffer from food poverty in times of embargoes and other trade restrictions. Although many of those ordinarily food poor may still suffer, some poor households will benefit from having members in the military and, even if there are no direct benefits, resources then do not have to be shared as many ways. In addition, warfare may increase certain kinds of production, creating new employment and income. However, in most recent conflict situations there are many more losers than beneficiaries, because of destruction, looting, and political control of food distribution. When the share of marketed food goes down, the resources required to obtain it can be prohibitive.
Commerce restrictions also have effects on food poverty that do not operate through shortage. Households whose livelihoods depend on wages from industries that specialize in export goods may be especially vulnerable to food poverty if embargoes lead to unemployment or underemployment for workers in those industries. A fuller discussion of the effects of employment on food poverty follows in the section on entitlements.
People driven across international borders as a result of war or civil strife are refugees. The most fortunate of refugees may be able to contact family members who migrated in less stressful times, but most leave behind possessions, sources of income, and social ties. Abrupt social displacement removes most customary sources of entitlement to food and for many, charitable distributions are the only buffers between them and starvation. Food insecurity among refugee populations is a concern of the United Nations and non-governmental organizations (NGOs), who raise money, food, and resources to relieve poverty and hunger throughout the world.
The situation of refugees highlights the well-known relationship between hunger and infection. Mortality in refugee camps is very high, but little is attributed exclusively to hunger. Concentration of people in refugee camps brings diseases from a wide range of locations, and these diseases spread rapidly among people already weakened by hunger (Seaman 1993). Those whose immunity is compromised by poor food intake are often unable to withstand infections that would not be life threatening under ordinary circumstances.
Government spending and poverty-alleviation efforts influence economic activity, employment, incomes, and household access to food. Exchange rates, balance of trade, debt repayment, and other financial institutions all provide an environment that conditions what economic activities households pursue and whether food is available and affordable. Many countries, prior to the oil crisis of the 1970s, pursued economic policies that relied heavily on debt spending and provided hefty food subsidies and government-sponsored health care for the poor. As debt became an ever-greater burden, the international financial community insisted that governments "adjust" their spending and restructure public enterprise. The set of policies developed to enforce greater fiscal responsibility are called "structural adjustment" or "adjustment" policies. Although the goal of structural adjustment is to build a healthier economy for the long run, the associated policies that cut social welfare spending, readjust exchange rates downward, and insist on greater investment in exports, including export crops, have been blamed for much of the food poverty that currently afflicts the countries affected.
Structural-adjustment policies have been negotiated between international financial institutions - primarily the World Bank and the International Monetary Fund (IMF) - and a large number of developing countries. These allow debt-repayment schedules to be redrawn but they also impose conditions that affect the structure of trade and government spending. The purpose of these is to restructure both production incentives and consumption patterns, so that there is a better match between a country's resources and its economic activities. If structural adjustment achieved all of its goals, adjusted countries would demonstrate sustained economic growth that would allow them to repay their debts and remain out of debt in the future.
The process of structural adjustment changes national food availability because it typically involves currency devaluation, which increases the real prices of imported food. It also changes the price of food and other goods relative to the price of labour; this has important implications for the ability of households to secure access to food. Even households that may benefit from structural adjustment in the longer run are likely to have difficulty securing access to food during rapidly changing social and economic conditions.
External debt and debt financing are important constraints on the ability of nations both to import food and to invest in their own agricultural sectors to enhance future production. The size of the debt burden in developing nations can be better appreciated by comparing the size of the national debt with gross national product (GNP); in 1992 the less-developed countries as a whole produced only about three times as much as they owed. In Latin America and the Caribbean, debt was 38.1 per cent of GNP, and in sub-Saharan Africa it was 88.2 per cent of GNP (World Bank 1994). It is not realistic to believe that countries can expand agricultural production if huge shares of their economic surplus are being spent on debt service and imports.
Unfortunately, even if structural-adjustment programmes would be tremendously successful in the long run, their short-term effects are often devastating. One of the most common macroeconomic policy changes made by adjusting countries is the devaluation of foreign exchange rates. Having an overvalued currency keeps imports artificially cheap, and, after devaluation, less food can be imported, given the same amount of foreign exchange. This particular cause of food shortage is most likely to lead to food poverty for the urban poor. Urban dwellers usually have less flexibility and resources to produce their own food and are therefore likely to rely on markets. Since devaluation is often accompanied by wage freezes in both the private and the public sectors, food prices rise while incomes do not.
Domestic export industries tend to expand with devaluation (because the same volume of exports earns more in the local currency), but the time-lag between devaluation and employment expansion is usually substantial, especially under the conditions of recession that accompany the adoption of structural-adjustment policies. Recession and adjustment are linked for two main reasons. First, recession usually precedes adjustment: countries often agree to the terms of adjustment policies because stagnation in their economies gives them little hope of repaying debt on normal, commercial terms. Second, potential to expand production is limited because debt finance and repayment - even on renegotiated terms - consume surplus that could have been reinvested in the economy.
Inflation aggravates the plight of households in adjusting countries. When faced with both higher food prices and higher prices for other necessary goods, these households must trade off their limited resources. Urban households that have lost formal employment, or who find their income inadequate in the face of inflating prices, often compensate by expanding informal-sector activities, but the profitability of these is especially limited during times of recession. Even though structural adjustment hits the urban poor harder than most rural dwellers, higher prices for agricultural inputs also have severe effects on rural dwellers. In Mexico, growing maize on plots already conditioned to the use of herbicides and fertilizer became unaffordable to many farmers under structural adjustment (Collier 1990).
Overall, structural-adjustment programmes affect the entire economy and decrease private incomes at all levels, but poorer households are disproportionately disadvantaged (Macedo 1988). Even in countries where income per capita has grown significantly, the income-distribution effects of adjustment contribute to income inequality (Serageldin 1990) and increase the likelihood that more households will be in food poverty. Anti-poverty and welfare programmes that directly address the health and food needs of the most disadvantaged can help lessen the pain of structural adjustment, but the poor are still usually left worse off both relatively and absolutely.
Social dimensions of adjustment
The World Bank has tried to limit the impact of adjustment on income distribution. Their Social Dimensions of Adjustment project and other poverty-alleviation programmes include interventions intended to increase access to employment and food for the urban poor and access to land and credit for the rural poor. Public employment schemes and food-for-work programmes have the advantage of targeting scarce resources more effectively to poor households than do direct food-supplementation programmes. In Senegal, the government has sponsored retraining and rural-resettlement schemes for laid-off civil servants and laid-off workers from the manufacturing sector (Serageldin 1990). Such capacity-building programmes are critical in preventing adjustment from miring greater numbers in food poverty. Interventions that relocate underemployed urban workers into rural areas promote agriculture and potentially help reduce the chances of food shortage in adjusting countries.
Social Dimensions of Adjustment programmes have been added to adjustment schemes to compensate in part for health care and nutrition programmes that could not keep up with rising demand for services while government expenditures on them was limited by the adjustment package. Programmes that seek to mitigate the social costs of adjustment are challenged by limited financing, limited knowledge of the most vulnerable groups, and limited capacity to target interventions. These difficulties help explain why they are not always successful. The rapid pace of social and economic change under structural adjustment further complicates these problems. Food hand-outs are more difficult to target to the poor than food-for-work schemes, which tend to attract only those who do not have other viable employment options. However, if interventions do not reach the poor promptly, free food distribution may have to precede work-based programmes because the poor will need improved nutrition before they are healthy enough to work.
Entitlement failure of households
The use of the concept of entitlements for understanding hunger has been popularized by the work of Amartya Sen (examples include Drèze and Sen 1990; Sen 1981a, 1981b). Own production, stored wealth, employment, kinship, and government transfers are all possible sources of food entitlement. Sen distinguished among endowments (e.g. land), exchange (e.g. wage), and social-security entitlements. While other approaches recognize that access to food is as critical a determinant of household hunger as is total production, the entitlement approach does not assume a fixed pattern of inequality in the distribution of food: instead, it focuses on how people acquire food. At low income levels, households whose entitlement to food comes from a single source are much more vulnerable to food poverty than those who have multiple sources of entitlement. Some sources of entitlement are also more stable than others. Diversification across and within food-entitlement categories limits risk of food poverty.
Opportunities for households to diversify risk are somewhat limited where all members typically reside in fairly close proximity to one another, so that labour market and other economic opportunities do not extend easily beyond the production patterns in their place of residence, particularly in rural areas. One way to transcend this limitation is by out-migration of individual members, who may no longer eat with the household on a regular basis but who retain important economic and affective ties. Migrant members who return to the household frequently, to visit, bring money, or contribute agricultural labour, are referred to in the demographic literature as "circular migrants." The contribution of remittances from circular and other migrants toward entitlements to food will be considered in greater detail below.
Access to land (resource-endowment entitlements)
Land, especially land ownership, is perhaps the most critical entitlement for preventing a household's food poverty in rural areas. In addition to providing food in ordinary times, access to land can safeguard against household hunger in situations where food supply is limited by non-environmental factors such as changes in food prices brought about by trade policy. It also provides an asset in protecting against food poverty and destitution if production is limited by factors such as widespread drought.
Land ownership is the most secure means of gaining access to land. Other arrangements, such as patron-client relationships, sharecropping, tenant farming, and use of communal lands also provide production-based entitlement to food, but are less reliable during times of crisis. Chen (1991) showed that, in Gujarat, those with insufficient land were customarily allowed to use private pastures and forest groves, but that these rights were withdrawn during drought.
The value of land for securing entitlement to food is further conditioned by labour availability, which a household must itself provide or otherwise purchase. Purchase-price controls on basic food crops limit earnings from sales of surpluses, pushing farmers toward greater production of non-food cash crops or off the farm into the urban sector. Both responses limit the amount of (particularly male) labour available for food production. The ability of women to sustain food production, where male labour has been diverted to other activities, has been limited by sheer time and energy constraints, as well as by less-frequent contacts with agricultural extension workers, the diversion of their labour to cash crops or other income-earning activities, and their numerous domestic and child-rearing responsibilities.
Additional evidence for how important access to land is for the nutritional status of rural households is reviewed below in the section on within-rural variation in the distribution of food poverty.
Employment (exchange entitlements)
Most labour-market activities (self-employment or employee) are remunerated in cash, which is then exchanged for food. Some forms of employment (especially in agriculture) involve transfer of food directly to employees. Even households engaged primarily in subsistence agriculture frequently have at least one member who is employed off the family farm; this is an important risk-avoidance strategy.
Food entitlements are closely tied to jobs and wage rates, which is why employment statistics are critical indicators of food poverty. Formal-sector jobs in developing economies are difficult to secure; loss of such a job may cause a dramatic drop in family income, since finding other employment with the same pay rate may be a lengthy or impossible process. Employment in informal economic activities presents different challenges, since many of those employed in the informal service sector, cottage industry, and marketing are chronically or sporadically underemployed, even if they never face open unemployment. Compensation may be so pitiful that working even more than full time does not always garner enough income to ensure entitlement to food for all household members.
In crisis situations, successful government initiatives to prevent famine in cases of large-scale food-production shortfalls include employment creation; lost entitlements are re-created through wage-based government employment Drèze and Sen 1990). These programmes allow food distribution to continue through normal economic channels and limit the social costs of widespread unemployment.
Food poverty persists in non-crisis situations and presents one of the key difficulties that developing countries face as they seek to modernize their production systems without increasing hunger. Modern production methods usually replace labour with capital and import the machinery from countries where labour is more expensive. The forces determining wages and employment levels in modern agriculture are different, but the net effect of modernization still is to limit employment-based entitlement to food. Modern crop varieties (e.g. hybrids, high-yielding varieties) that began to be introduced with the Green Revolution in the 1970s have expanded overall demand for agricultural labour. The yield from these new varieties depends heavily on labour-intensive transplanting, weeding, fertilizing, irrigating, and application of pesticides. Modern varieties of wheat and rice require 20-40 per cent more labour than traditional varieties (Mebrahtu et al. 1995). But population growth and in-migration into regions where agriculture has modernized have kept rural labour supply high and rural wages low (Mebrahtu et al. 1995). Mechanization in some cases has reduced rural employment, especially where governments subsidize imports of labour-saving machinery with the intent of increasing agricultural productivity (Mebrahtu et al. 1995). Resulting decreases in rural employment and/or wages can push rural households into food poverty.
All households that rely on employment-based entitlement to food are vulnerable to changes in the structure of production or the competitiveness of labour markets. Surplus labour that spills over from the agricultural to the industrial or service sectors is typically unskilled and untrained, and individual productivity may be limited by poor health and nutritional status. Those who remain unemployed or underemployed for a period of time find that they are less-effective employees and also earn less at self-employment. Under these constraints, households disadvantaged by low earning power at one point in time have little ability to improve their earning potential and need health and nutritional programmes to prevent their further deterioration into food poverty. These are not problems faced by those in more skilled positions or modern-sector industries.
Prices and wages
The degree of food security depends on prices as well as labour productivity and income. Although increased agricultural productivity usually leads to higher incomes and better food security among households that have access to modern inputs and methods, the food security of households that continue to use less-productive methods depends largely on the degree to which production expansion drives down food prices and on how much food they sell rather than buy. Small-scale farmers often consume a share of their own produce, but it is increasingly rare for household food needs to be met entirely by subsistence production. Where prices of other foods, such as legumes, increase dramatically relative to staple grains, some farmers cannot afford to purchase what they do not produce. Even households that are mostly food self-reliant find their food security compromised where lower food prices mean they have to sell a greater share of their own production to meet their non-food needs. More advantageously, higher agricultural production that lowers food prices improves the access to food in non-landed households.
Increases in the prices of essential non-food items, such as fuel and housing, also aggravate food poverty in low-income households, regardless of income source. Households may try to minimize the impacts of rising prices by shifting their consumption patterns, but lack of fuel poses an insurmountable barrier to food preparation.
Prices for all goods are influenced by exchange rates. Many developing countries overvalued exchange rates in an effort to make imports cheaper for their residents; it takes fewer pesos to buy a sack of wheat if the peso is overvalued. In the wake of the debt crisis, and with the widespread adoption of structural-adjustment programmes, fewer countries maintain overvalued currencies. Having imports available only at world-market prices in economies where the currency used to be overvalued means that the same sources of entitlement will not command the control over the same products.
Kinship (social-security entitlements)
Protecting households from sudden jolts and destitution under shifting economic conditions are the traditional (and increasingly non-local) sources of social security. Many households avoid hunger, in periods when their exchange entitlements to food fail, through kinship ties. These food-provisioning networks protect households from sudden drops in their own production and income but are less effective where economic conditions or environmental shocks affect food availability for most in the network. For example, a father who loses his tenant rights to cultivate land can rely on kin for food while he is seeking alternative employment, but if drought depresses all agricultural production, fewer resources are available to share within the kin network.
Kin networks nourish in multiple ways: customary rules of hospitality dictate that kin present at a meal be fed, and food is often sent home with visiting kin as parting gifts. Children may be fed by, or sent to live with, relatives who provide for their food and health needs (Desai 1992). In West Africa, child fosterage occurs in both crisis and non-crisis situations and it is common enough that such arrangements do not create unusual social obligations or even necessitate admitting that the sending family is unable to care for their child(ren) alone.
Kin networks also transfer cash - especially important where someone lives and works in another location. Among the poorest 10 per cent of households in Botswana, 25 per cent of household income is supplied by transfers from other households or family members living elsewhere (Quinn et al. 1988). The role and reliability of migrant remittances are discussed further below.
Kinship links may be created explicitly for the purpose of safeguarding access to food. Marriage ties and "fictive kin" arrangements extend food-sharing obligations and entitlements to non-blood relations. Even very poor families decrease their vulnerability to hunger by giving away food and other gifts in ways that create links to people who will help them in more desperate situations (Adams 1993). It is also customary for age mates in African societies to share food and labour with one another.
Such food-sharing obligations and benefits tend to shrink, however, as the total resources of a community drop. In Gujarat, India, Chen (1991) documented food sharing with a wide variety of caste neighbours when food was plentiful, but non-market exchanges were based primarily on kinship ties during lean times. In rural Africa, sociability declines along with food supply: people eat alone to avoid the obligation to share (Messer 1984).
Female headship in some situations also reduces food flows from kin. Households that are headed by women because the husband is a temporary migrant or lives with one of his other wives may be just as well integrated into kin networks as those who reside with their husbands. This is also likely to be true of female-headed households among matrilineal groups. However, if female headship represents a permanent separation of a woman from her husband or the father of her children, her household may lose the benefits of food sharing with his relatives. For example, divorced women in Ghana have little contact with their husbands' kin, while other women who do not live with their husbands typically live with (or very close to) co-wives, in-laws, and other patrikin (Lloyd and Brandon 1991). In other countries, such as Haiti, marital status matters less because food-sharing networks are entirely female.
Migration that reduces the numbers that need to be fed and/or increases potential income for food is a principal mechanism of risk aversion. Migrant family members, whether they circulate between areas or permanently change residence, give households access to remote labour markets where wages may be far higher, or at least fluctuate at different times. Disadvantages for those left at home in rural areas may include insufficient household labour to sustain home food production and the unreliability of migrant remittances. Many who migrate to urban areas fail to find formal-sector employment, and jobs in the tertiary sector may not earn enough surplus to remit back to the rest of the household. Rural households may even have to provide food for urban members while the latter are seeking employment, particularly where both food and fuel in urban areas are scarce and more expensive (Fuller et al. 1990; Goldscheider 1983).
Regularity of remittances varies and seems to be related to the frequency of return visits. Migrants who return home regularly may be more integrated into their households and feel more accountable. Additionally, the channels for sending money in developing countries are frequently undeveloped, and remittances may not be forthcoming unless the migrant or a trusted friend is personally travelling back to the area of origin. Even when other channels are used, large amounts may not be risked: circular migrants from Mali to France tend to bring more money home when they visit than they send while they are away (Findley 1994).
Given its unreliability, income from migrants typically does not supplement household food budgets as much as would otherwise be expected. Few studies have distinguished source of income from type of income, but Kaiser and Dewey (1991) found that sporadic remittances were associated with decreased food quality and with retarded growth of preschool children in rural Mexico. Lump sums tend to be spent on consumer durables, such as roofing sheets or bicycles, which require significant savings to buy and may be the "target" for which migration was undertaken.
Having migrant household members may help in times of crisis, but the ability to migrate during a crisis is limited by poverty and the numbers similarly displaced and looking for work. Findley (1994) reported that Malians working overseas were expected to increase remittances in times of drought-induced crop failure, but that such long-range migration was difficult to launch in response to drought because of cost. During the 1983-1985 Malian drought there were significantly increased levels of internal migration, but this served primarily to reduce the number of members in rural households, rather than to generate additional income (Findley 1994).
Flawed government policies
Taxation poses another drain on household resources: when taxes rise, the ability of households to meet food and other needs falls. Higher taxes differ in some ways from other price rises, however. Taxes on luxury goods and income can be used to generate government revenue without burdening the poor, but other tax structures can be quite regressive, as is the case with a sales tax.
Taxes can be used to alter the structure of production, as where exporting industries are given tax breaks in order to increase their capacity to earn foreign exchange, or where taxes on rural producers force them to produce cash crops over subsistence crops. As reviewed in chapter 3, taxes on agriculture are favoured by governments, who privilege newly developing urban industries even though these reduce food-production incentives and affect household entitlements and access to food.
Taxation, additionally, is a mechanism by which governments redistribute wealth and food and has the potential to reduce food poverty. Nutritional supplementation programmes reach households not productive enough to be directly affected by a tax burden. Public health programmes reduce nutritional requirements because of the synergism between infection and disease; those that have been particularly effective in a wide variety of contexts include immunization, sanitation, and oral rehydration therapy. Government investments also expand job opportunities in both public and private sectors, thereby raising employment-based entitlement to food. Alternatively, poor investment decisions contribute to economic stagnation or promote capital-intensive industries that employ few workers. Many nutrition and health programmes fail to reach the neediest. In sum, taxation and tax-revenue spending in any particular country clearly influence production choices, household resources, and the distribution of food and health.
Insufficient government safety nets
Direct feeding, food supplementation, food stamps, and subsidies are all programmes designed to raise the real incomes and food consumption in households (Rogers 1995). Giving food directly to households improves intake, if money that would have been spent on food in the absence of the transfer is not spent instead on other goods. Providing households with more food than they were purchasing in the absence of the programme guarantees an improvement in income but is too costly to be practical. Instead, smaller transfers are targeted toward poor households in ways that will be most likely to increase food consumption, even though there may be some leakage.
Government safety nets function most effectively when they utilize existing markets to distribute food; common mechanisms include food stamps, ration shops, and specially subsidized foods (Kates 1996). It is difficult to reach households on a sustained basis with food-welfare programmes (particularly in rural areas), but changing social, economic, and environmental conditions continually re-create the need for short-term assistance as well as for programmes that create and diversify entitlements. Even for households whose entitlements are usually sufficient to avoid food poverty, government programmes may be necessary to cope with additional stress introduced by disease (Kates 1996).
Given the difficulty of designing and implementing interventions that will absolutely increase household food intake, such programmes are arguably most important as a relatively stable source of household entitlement to food. Even if children consume less at home if fed through school lunch or other direct-feeding programmes, their intake is less vulnerable to changes in family income than it is in the absence of such a programme. Even if direct feeding and other similar programmes do not have dramatic effects on intake in "normal" times, they may shield households from food poverty in times of crisis. Supplementary feeding has been shown to decrease the prevalence of both moderate and severe malnutrition in children; it was an important component of the Iringa nutrition programme in Tanzania, which succeeded despite being implemented during early stages of structural adjustment (Serageldin 1990). Supplementary-feeding programmes have also been highly successful in parts of India, including Tamil Nadu. Such programmes are often administered through existing health and education systems.
Contents - Previous - Next