Previous - Next
This is the old United Nations University website. Visit the new site at http://unu.edu
The discussion above on prevalence and indicators of food shortage has illustrated that its causes are complex. Some hunger indicators, such as production shortfalls, highlight problems that may lead to food shortage. Others, such as DES, directly measure food availability within a country or region. These food-shortage indicators report outcomes of physical and biological factors, sociocultural influences, political-economic forces, and interactions among these elements.
Physical and biological factors
Production is only one determinant of food shortage, but a crucial one. It is obviously critical on the global level but can also be decisive for some countries or communities. Production potential varies across countries, dependent on natural factors (including climate, soils, water, food species, and pests) and cultural factors (including technology and investment strategies).
Climate
Temperature and rainfall are critical elements determining when and how often crops can be sown. While some Asian countries are able to harvest three times in a single year, food production nearly halts during dry seasons in many tropical zones and during winter cold in temperate areas. Extremes or thresholds of heat, increasingly accompanied by high ultraviolet radiation, and of cold, especially early frosts or late thaws, can ruin harvests. They test the limits of growing seasons and moisture-temperature tolerances of particular crop varieties. These extremes will be modified by global climate change, which promises to transform regional cropping patterns. For the present, drought is the most widespread climatic threat to production, and is treated more extensively below.
Seasonality means that there may be food shortage during part of the year in places where total annual production appears to be more than sufficient to meet nutritional needs. Agricultural societies and households adapt to potential seasonal scarcities by planting a variety of early- to later-yielding crops, storing or selling harvests to minimize losses, investing in social feasting when food is plentiful, and drawing on social obligations of reciprocity when food is scarce. They ration, process food staples more coarsely, supplement diets through foraging, and consume less-preferred foods. They also schedule crafts, migrant labour, and other economic activities to diversify and ensure income in "off" agricultural seasons. Despite such adaptive mechanisms, prolonged or multiple years of shortage, as experienced especially in SSA and South Asia, give rise to potential famine conditions (defined as widespread and extreme food shortage leading to elevated mortality and mass movements of population in search of food) that nowadays are addressed relatively successfully by state and international early warning systems (EWS) and response. Even where EWS are well established (as in India), however, seasonal hunger remains a problem addressed neither by state nor by traditional sociocultural mechanisms of food sharing, which tend to be undermined by modernization (Chen 1991).
Although drought is often thought of as the precipitating cause of famine, because so many farmers in a single area experience crop failure simultaneously, drought does not lead to food shortage or its extreme manifestation - famine - if there are adequate carry-over stocks available, or if food is available through market or relief channels (Ravallion 1987).
Creeping disasters, such as drought, are likely to devastate crops but leave infrastructure intact. Both timing and duration of rainfall may be implicated, since, for seed crops, moisture is critical immediately after planting and also at the stages when fertilizer is applied. Drought can also prevent planting if the rains are late, so that soils are too dry to till before the planting season has passed. Failure of rains at any of these stages greatly reduces harvests.
Since the early origins of agriculture, human societies have always tried to extend productive seasons and, especially, available moisture by controlling groundwater evapotranspiration, by various water-storage techniques, and by distribution via irrigation. Water control is closely linked to social power and control, since the ability of irrigation systems to enhance food production is limited by their general state of repair. Hydraulic systems, which initially require large amounts of capital and labour to construct, later require maintenance. Social resources must be mobilized through a sense of common purpose (or coercion) to preserve irrigation ways, which otherwise silt up, leak, and lose effectiveness. The history of political fortunes and social breakdown in the Near East and Middle East has been linked to cycles of "salt and silt," environmental and subsistence crises triggered by failures to control the life-giving waters of irrigation (Jacobsen and Adams 1955). The expansion and sustainability of Green Revolution (GR) agriculture in the modern era is similarly dependent on effective water management, which is necessary to prevent waterlogging and salinization of soils and crops, and silting and disruption of water channels.
At the opposite end of the rainfall spectrum, too much rain can also impair agricultural production, especially where flooding is severe enough to kill crops by uprooting or submergence, but also where it simply slows growths or makes cultivation and harvesting extremely difficult. Even after harvest, flooding can be devastating when it occurs before crops have been safely stored (Good 1986) and leaves crops vulnerable to rot.
Irrigation systems can be overloaded by flooding but also can be one of the mechanisms used to cope with irregular rainfall. Irrigation systems often do little to prevent damage caused by the force of the rains themselves but they do prevent further damage from waterlogging. They also enhance productivity in dry seasons by holding over water from heavy rains; this is accomplished through the use of simple earthen dams in Sri Lanka (Grigg 1985), as well as through more elaborate systems of canals and pumps. Along with terracing, irrigation systems may also limit soil erosion and help sustain soil fertility along flood plains.
Less regularly, food systems can be entirely disrupted by natural disasters, which affect both growing and marketing conditions and sometimes the social fabric. Hurricanes and, to a lesser extent, earthquakes can destroy crops. They can also devastate transport, markets, and other infrastructure, and cause food shortage even where the crops themselves survive. The economic destruction accompanying natural disasters, which can cause scarcities of many materials, often pushes prices upward, increasing the rate of inflation, reducing employment, and more generally contributing to balance-of-payments problems.
Inability of government to deal effectively with natural disaster, such as earthquake or drought, can threaten political stability and overturn fragile regimes. In each of the recent cases of Ethiopia, the Sudan, and Rwanda, their ruling governments' ineffectiveness in dealing with drought and ensuing famine conditions provided an opening for the political opposition to challenge successfully that government's authority and legitimacy. Political leaders were portrayed as prospering while the masses went hungry, and this became the successful rallying point for civil uprisings. Similarly, the inability of Samoza's regime in Nicaragua to alleviate widespread suffering from a hurricane became the trigger cause of the Sandanista's successful ascent to political power. Each case, however, was followed by periods of readjustment, civil disorder, and lowered food productivity.
Although the magnitude of disasters' effects on overall economic development appears to vary (Albala-Bertrand 1993), the short-term economic shocks caused by sudden natural disasters usually decrease marketed food availability in the affected regions, so that food aid may appear to be the only practical means of getting enough food to the affected populations. Inappropriate world response, such as Guatemala's inundation with donated food following the 1976 earthquake, sometimes disrupts markets and income for local farmers, who face plummeting food prices and agricultural income even where their crops have not been destroyed. The question of whether food aid increases food availability at the household level will be revisited in chapter 4.
Politicians and national politics are also important social actors in other ways. They provide the economic, social, and cultural framework (what policy makers increasingly term "the enabling environment") to prevent natural elements from precipitating wider disasters. Despite the seemingly arbitrary nature of sudden natural disasters, they do not affect all areas in the same way, even when their severity is of a similar scale. Healthy economies rapidly bounce back from shocks because they have more internal resources dedicated to mitigating the immediate and longer-term impacts on food production or distribution. Similarly, precautions that minimize the impact of disasters, such as earthquake-resistant housing and roads, are not distributed evenly either across or within countries. Some so-called "natural" disasters might be better thought of as man made. For example, flooding often results as much from deforestation as from excessive rainfall or unusually intense hurricanes. Abuse of land, particularly overgrazing, increases vulnerability to wind and flood erosion (Albala-Bertrand 1993). Sudden natural disasters cannot be prevented, but the effects that they have on food production and food importation are conditioned more by political and economic processes than by the intensity of the calamity.
Soils
Food production also varies according to soil structure and fertility, factors that are less easily measured - but perhaps more easily modified - than temperature or rainfall. The differences in productivity between the dark soils of the US midwest and the sands of the Sahara appear obvious, but it is less clear how much less-dramatic variations in soil conditions matter. Many tropical soils contain less nitrogen and phosphorus, have lower capacity to absorb fertilizers, and therefore have lower conventional productive capacity, but some tropical soils (most notably in the Amazon) have been very intensively farmed and further intensification is possible in other areas (NAS 1986).
Since agricultural methods and inputs vary in areas with different soils, the causes of disparities in productivity are multiple. Grain yields per hectare in SSA are about one-third of those achieved in East Asia, but SSA also struggles with more challenging climate, uses fertilizer at a rate less than 13 per cent of the world average (World Bank 1992), and has made little use of irrigation. Investments in agricultural intensification, including higher-yield-potential seeds, fertilizers, water management, and chemicals for pest control, are costly and make it unlikely that they will be easily or widely available for use by poorer farmers and countries. Especially where imported food is cheaper than domestically produced food, as is the case today in many developing countries, expanding local production may not appear to be economically feasible.
Production may also be limited by low soil fertility and restricted access to fertilizer supplies. Fertilizer is required to realize the full benefits from hybrid GR seeds; quantities of its use often serve as a measure of agricultural improvement or modernization. In contrast to the time when most farmers used local sources of animal or green manure to enhance soil fertility, most developing countries today rely heavily on inorganic fertilizers, which they must import. At both household and country levels, fertilizer constitutes a significant expense, and lack of means to purchase adequate quantities potentially reduces crop yields. Lack of cash, poor credit, or isolation from sources of supply due to underdeveloped infrastructure hamper farmers' access; lack of foreign exchange and balance-of-payments difficulties limit the total supplies imported and available within a country. The cost per unit of fertilizer is also a factor adding to the gap between fertilizer needs and the amount that many developing countries can afford to import (Monteón 1982; UNCTAD and Mukherjee 1985). Upward fluctuations in the price of petroleum, a raw material for inorganic fertilizers, imperilled developing country food production in the mid-1970s, and could happen again. Timely, as well as total, availability constrains output, since optimal fertilizer impact is achieved only by applying it at the most sensitive points in the crop cycle. Only greater domestic production can protect farmers and consumers in developing countries from severe price fluctuations and transport or market bottlenecks.
Biological stressors
Disease, insects, animals, and weeds also damage crops and reduce yields, and are controlled by mechanical, land management, chemical, or biological means, including breeding plants and animals to resist key stressors.
Viruses are controlled by eliminating insect or other vectors and by breeding resistant seeds through conventional or new genetic engineering techniques. Biotechnology also offers new diagnostic techniques for identifying and limiting infections, as well as for producing and multiplying clean seeding stocks for vegetatively propagated crops, such as potatoes and manioc. Diagnostics and planting-material multiplication potentially can be carried out as local-level cottage industries. Bacterial and fungal diseases are also addressed through breeding programmes and sometimes chemical applications. In addition to the breeding and selection of resistant stock, animal diseases are controlled by preventive inoculations and curative remedies. Biological and chemical control of vectors are also common, and sometimes innovative, as where scientists of the Kenya-based International Centre of Insect Physiology and Ecology developed a simple cow urine-baited trap for tsetse flies, to eliminate trypanosomes.
Insect damage similarly is avoided by breeding resistant varieties, mixed-planting strategies that buffer particular host species from their characteristic pests, and insecticidal chemicals that kill (or otherwise interfere with the feeding, maturation, or sexual reproduction of) the target populations. Biological controls involve introducing predators or pathogens of the pest species. Integrated pest management that combines chemical, biological, and some hands-on mechanical strategies - such as removal of insect eggs before they hatch - are increasing worldwide, in response in part to their greater safety and lower costs and in part to the increasing environmental burden and decreasing effectiveness of chemicals. Chemical use soared during the early decades of the GR, as new seeds were accompanied by chemical packages that sometimes indiscriminately wiped out "good" as well as "bad" insects, removing the predators as well as pests, presenting an unprecedented opportunity for pesticide-resistant insects that co-evolved with the plants and chemicals in these GR ecosystems to cause severe crop damage. To restore insect ecology, reduce pesticide poisoning, and re-create a safer balance for plants and humans, Indonesia, as a case in point, banned most pesticides. On Java, in some seasons and places where rice-hopper damage can be anticipated to be enormous, communities have organized massive brigades of farmers and schoolchildren to collect the insects' eggs and so prevent damage (Indonesian National IPM Program 1991).
Other kinds of animal damage, such as that of rodents, birds, or livestock, are conventionally limited by human labour, although the "excess" child labour that traditionally scared birds and livestock from ripening fields is disappearing as schooling competes with agricultural tasks. Reduced child labour availability is one of the factors leading to the selection of maize over sorghum in many parts of Africa, since maize ears are covered and less subject to damage than sorghum seeds, which are exposed. Rodent pests may be chased, physically killed, or poisoned, although the poisons are still not well controlled. Human poisoning is a hazard in cases of insecticides and rodent poisons, since the toxic chemicals tend to be sold in small quantities and not handled with the care specified by their manufacturers. The usual way to prevent larger animals stealing crops is scrupulous tending in areas where such risks are high.
Weed damage also is increasing, as plagues such as Striga spread across the maize and sorghum fields of Africa. Weeds choke out useful plants and also compete with cultivated plantings for soil nutrients and water, reducing their yields and disease/insect resistance. Careful plantings, to give the intended food crop a healthy start, or herbicides are alternative strategies, as less labour is available to hand-weed at very low wages. Seeds of major crops that have been genetically engineered to resist a particular brand of herbicide also are being aggressively developed and marketed by large chemical giants, such as Monsanto, that have purchased seed companies and the molecular-biologist/plant-breeding scientists to advance the match and ensure a continuing market for their chemicals. These developments, although they may increase crop production, also increase the chemical load and the potential for herbicide-resistant genes (traits) to pass from the desirable species to closely related weed species, which would then be equipped to inflict even more damage.
Two modern agricultural factors contributing to greater risk of pest damage are genetic uniformity and uniform or continuous cropping cycles. As most farmers in a region purchase the same single varieties of improved seeds designed to maximize production on the same agricultural cycles, they establish conditions ideal for the explosion of many pests. An overall strategy to reduce plant damage therefore is diversified cropping. Carefully patterned plantings of multiple species and varieties within species is a good hedge and an aspect of traditional agricultural practices that modern cultivators need to review.
Seeds and related technologies
Basic to all production strategies and yields are the seeds, which may be more or less productive; drought and disease resistant; or responsive to moisture, fertilizer, and other chemical inputs. In many regions, "traditional" seed varieties are less and less plentiful because they are unable to keep up with demand for more food and more intensive methods. Modern seeds, bred for shorter growing periods and neutral photoperiod, allow multiple sowings and harvests; they also may be tailored to resist insects and plagues that attack traditional varieties and lower their yields.
Exponential growth in world food production over the past three decades has been made possible largely by new seeds and related technological advances that have greatly increased output per unit of land and labour. The impact of the GR seed-water-chemical technologies on grain yields has greatly reduced food shortage, especially in rice- and wheat-growing regions. But technological advance can prevent food shortage only if (1) it continues to increase yields faster than the population grows, (2) food availability within agriculturally modernizing regions actually increases, and (3) increased caloric availability is not transformed into richer diets.
Controversy rages over whether the dramatic food production increases achieved over 1960-1990 are sustainable and will be able to keep up with projected population growth. Brown and Kane (1994), who think that world population has reached its agricultural limits, argue that prior to the GR there existed a backlog of agricultural technologies waiting to be applied. Today, by contrast, no such "breakthrough" technologies are waiting "on the shelf"; although advancing biotechnologies promise to enhance production, they will not allow the quantum jumps in production necessary to keep pace with growing population. Technological optimists, such as World Bank economists Mitchell and Ingco (1993), differ: they argue that biotechnologies hold very great potential to increase food supplies and that careful pricing of both factors of production and resulting food products also can stimulate more food production. Intermediate positions, such as that of the International Food Policy Research Institute (IFPRI) "2020 Vision for Food, Agriculture, and Environment," argue that there also exists greater potential than Brown and Kane allow, to expand the applications of known technologies to prevent regional shortages, but that the world community must make such agricultural investments their priority. Because of required investments and economies of scale, many poorer farmers in developing countries have not been able to take full advantage of existing technologies. Increasing food production in some areas may depend more on altering social and institutional structures to allow fuller utilization of existing technology; it depends less on radical new breakthroughs.
Other intermediate positions suggest that there exists great potential for significant reductions in expenditures, and increases in available food supplies, by controlling waste all through the food system (Bender 1994). Post-harvest losses in cereals are coming under control (Good 1986; Donahaye and Messer 1992) but those of roots, tubers, fruits, and vegetables remain very high: losses of potatoes are estimated at 5-40 per cent and of bananas at 2080 per cent (Tome et al. 1991). Biotechnology can also contribute to reductions in loss over the entire cultivation cycle and after harvest (Donahaye and Messer 1992).
Such savings and incremental production will not automatically eliminate food shortage for hungry populations, however. Technological advances still need to be directed toward crops and cropping characteristics, and distributed in manners that benefit those vulnerable to hunger. A case in point is Mexico, which has continued to suffer food shortage despite more than doubling its food production over 1960-1985, a period over which the volume of imported grain increased more than 20-fold (Barkin et al. 1990). This massive increase in imports, concurrent with rapid expansion of domestic production, can be accounted for in part by the shift of land out of maize production and into sorghum. Although sorghum is a basic food crop consumed in many areas of the world, in Mexico it is used primarily for animal food and therefore actually contributed to domestic food shortage because of the associated drop in maize production and export of livestock.
In sum, food production, supply, and the productive environment are influenced by a range of physical and biological factors that are not easily separable from sociocultural and political-economic conditions. These illustrate the linkages between production and consumption at local, country, and global levels.
Social and cultural factors
The causes of food shortage are in no way limited to physical and biological factors affecting production. Sociocultural factors include the organization of land and labour use as well as dietary preferences. Political-economic factors involve world markets and government policies to modernize agriculture and increase foreign exchange, which at the local level are translated into incentive structures, especially "getting the prices right." Armed conflict, a major contributor to food shortage, is treated more extensively in chapter 6.
Land and water control
Land tenure greatly influences what and how much is grown. Motivations to invest in land improvements and agricultural technology are tied to anticipated returns from land ownership or rents. Land reform is often cited as the major policy tool to improve food and other agricultural output, but may be most difficult to implement, since it threatens vested interests. Barraclaugh (1989), in his probing analysis of policies to end hunger, found that even allegedly reformist governments, such as the Sandinistas in Nicaragua, are stymied by vested landholding power interests if they seek more equitable land allocations. Communal or private ownership, permanent or usufruct use rights, and tax or labour obligations associated with tenure are three kinds of conditions that define the social and political order and the social relations of production and consumption. Inequality in landholding is an index of social inequality: aggregate landholding (small to large) is used as a measure of household poverty or wealth and a predictor of household agricultural and economic productivity, food self-sufficiency, food insecurity, and technology adoption. Intolerably severe or growing inequality in landholding, experienced by peasant cultivators as denial of basic subsistence rights, has been the cause of most peasant revolutions of the twentieth century (Wolf 1969). Their struggle for land encompasses the larger struggle for individual and communal human dignity, political autonomy and identity, and a decent standard of living. The political violence and discrimination that precede conflict, the destructiveness of active conflict, and the difficulty of restoring communities and reconstructing food systems following conflict, are probably the most significant causes of food shortage and related poverty and deprivation. They are dealt with further below and in chapter 6.
The most important way in which landholding affects food supply is in the choice of crops. Plantation or latifundia systems in Central America historically have focused on cash crops for export, with smallholdings for food to meet subsistence needs. Indigenous cultivators in Guatemala historically were pushed off the most productive lands into mountainous refuge zones; in El Salvador, subsistence holdings were reduced to allow larger holdings for coffee. Struggle for land by former combatants as well as non-combatants in the former war zones of Guatemala, El Salvador, and Nicaragua remains a threat to political stability and future food self-reliance, as indigenous and peasant communities try to regroup, establish rights over land, and produce a mix of subsistence and cash crops for internal and external markets. On the island of Negros in the Philippines, large landholders employed peons to farm sugar. Landlords did not allow workers to convert land-use to subsistence production, even after a precipitous drop in the price of cane: they preferred no product to a possible loss of control to the workers. In Senegal, Franke and Chasin (1980) have documented the ways in which French control over land, and demand for the cash crop peanuts, destroyed subsistence production and the traditional symbiosis between farmers and herders.
All these cases illustrate how the total amount of food grown in any developing country, and global food production overall, are affected by land tenure. The specific impacts, in turn, of land tenure and associated crop choices on household and individual hunger depend on the relative prices of cash versus food crops, and on who benefits from cash-crop revenues. Additional factors are, on large holdings, whether workers are paid enough and food is plentiful and cheap enough to ensure a decent living; and on smaller holdings, whether the net income from cash crops more than compensates for the costs of foods that are no longer home-grown.
Crop choices and hunger impacts also depend on who controls the income from cash crops - control that, in traditional societies, is tied to gender-based land tenure and labour exchange. In the Gambia, women refused to contribute labour to their husbands' potentially highly productive irrigated rice plots (in this case rice was a food crop sold for cash) because they would not control the income, whereas they controlled food that they grew on their own lands (Jones 1983). A 20-year perspective on Nigerian pert-urban agriculture has shown that female cultivators, drawing on kinship-based land rights, have developed an unanticipated niche that supplies greater Ibadan with roots, tubers, fruits, and vegetables (Guyer 1995). In Kenya, non-governmental organizations (NGOs) and the FAO's International Plant Genetic Resources Institute (IPGRI) have designed programmes to promote the production of indigenous food species for home consumption or market sales. Gardening everywhere is being encouraged as a way to increase food and fodder for internal markets and to improve household income and nutrition. Local gardening is also a way of conserving indigenous species in the environment.
Crop choices and production practices in much of Africa are changing in response to greater individualization of land holdings and wage labour but, as will be suggested further below, the impacts of such changes on food supply and household food security are predictably mixed, since the circumstances are so variable.
Whether the context is cash or subsistence crops, an additional supply question is whether larger or smaller holdings are more productive or efficient. Wealthier landholders are more able to invest in agriculture - to assume the risks of new agricultural technologies and the recurrent costs of seeds, fertilizers, machinery, and labour. They may enjoy also superior access to credit and economies of scale in the purchase of inputs and in control over the prices for their products. There is no question that the early adopters of GR technologies were wealthier landowners who were able to absorb the risks and enjoyed major benefits. They became wealthier also by absorbing the smallholdings of neighbours who could not compete in the new productive environment. However, it is questionable whether production per unit land generally is higher on larger holdings, because smallholders, given the means, may invest labour and other inputs more carefully and achieve even higher outputs per unit area than their wealthier neighbours. Equitable or fair access to credit, crop insurance, inputs, and markets may remove most economies of scale. But under modernizing agricultural conditions in much of the world, such conditions are hard to meet. Cooperative ownership and cultivation is one way that the possible disadvantages of smallholdings can be resolved; but, again, the evidence is mixed on whether larger communal or smaller individual holdings are more productive and efficient: so many other sociocultural and political economic factors besides landholding are involved in explaining why productivity is low on a Mexican ejido cotton cooperative or high on an Israeli citrus kibbutz.
In summary, the evidence on the ways in which land tenure influences agricultural investments and agricultural sustainability is mixed. Predictably, throughout the developing world, poverty and insufficient access to land result in unsustainable practices, including reduction of fallow, that lead to soil and water degradation, land exhaustion, and plummeting productivity. Farmers who lack resources to protect or improve land and water supplies have been blamed for soil erosion, desertification, and other ills (especially in SSA) and for deforestation and diminishing ecological resources and biodiversity in Asia and Latin America. Those working but not controlling the land can be expected to select the easiest or cheapest, rather than the most sustainable, cultivation practices. Morvaridi (1995) found that contract farmers in Cyprus used cultivation methods that were optimal only in the short run; because they did not own the land, they would not suffer the costs of long-run degradation. Conversely, extremely plentiful access to land can also discourage conservation: for example, extensive holdings in the Sudan have been tied to unsustainable agricultural methods that mine soil nutrients; after reaping initial profits, large landowners move on to exhaust additional areas. Whether owners deem it economic to maintain soil and water resources depends also on the conditions and costs of labour supply. Communally maintained terracing and waterways are falling into disrepair in countries such as the Philippines, where labour sees greater economic opportunities outside agriculture. Thus, land and labour constraints on productivity are closely tied, particularly under conditions of economic modernization and marginalization.
Labour
The amount of land that can be sown and harvested is, clearly, tied to available and affordable labour supply. Planting and harvesting are both activities that require far more labour than the rest of the agricultural cycle. In communities where these activities are shared, productivity on individual plots may be greater than if families had to provide all the labour that they could not afford to hire. Communal farming, although no longer common, provides some of the same advantages.
Where greater integration into the market economy disrupts traditional labour exchange, production may fall, as shown in Gudeman's classic study of a Panamanian village (Gudeman 1978). The reduction of patronage ties, such as in South Asian villages that have adopted GR technologies, may also produce labour bottlenecks that affect harvests. In many parts of SSA, modern cropping programmes fail where men control most of the land, technology, and proceeds but women are supposed to do much of the work, especially weeding. In Kenya, the increase in crop yield resulting from weeding was 56 per cent in female-headed households, but only 15 per cent in male-headed households; this led researchers to hypothesize that women do a less thorough job of weeding where they do not expect to control the benefits (Gittinger et al. 1990).
High labour costs may discourage extra hand cultivation and marginally lower outputs. But low agricultural wages discourage participation in the agricultural economy, where industrial or other opportunities exist. Very small household plots that cannot provide sufficient food and income push workers off the farm in search of income and can cause bottlenecks at planting and harvest times that lower food production. In Mexico, careful research into maize varieties and associated agricultural packages that would benefit small farmers proved less attractive and raised maize production less than anticipated; workers still could earn more off-farm, so abandoned farming. Scarcity of labour more than land is also a major constraint on production in much of Africa, where larger land areas since colonial times have experienced labour bottlenecks, as men were drawn off to work in the mines or to do other waged work and left women to clear, plant, and weed, with peak agricultural labour demands during the hungry season (Richards 1939). In such contexts, the problem of hunger is linked to underproduction in a vicious cycle.
Dietary preferences
An entire class of sociocultural issues related to food shortage have to do with consumption patterns and preferences. It is easy to think of these as relatively unimportant, because a preference for maize over sorghum in a food-short region is unimportant relative to total calorie availability. However, food preferences are one set of factors that determine what foods are grown and whether farmers, as in drought-prone SSA, sow and harvest drought-tolerant sorghum or sow a riskier maize crop that totally fails in drier seasons. The World Bank has identified overconsumption of wheat and rice as a key contributory factor to the African agricultural crisis: imports of these grains have soared over the last two decades to the detriment of local production of other staple crops. Wheat and rice can both be grown in most African countries but only at costs far greater than the price of imports. Whether tastes for these imports grew in an era of economic prosperity, as was the case during the Nigerian oil boom years (Smith 1991), or as a result of dearth relieved by food aid, dependence on imports persists and has created a mismatch between economically feasible local production possibilities and consumer demand (Nwomonoh 1991). Burkina Faso, which was severely affected by the Sahelian drought of 1968-1974 and again by drought in the early 1980s, imported 94 per cent of its grain as wheat from 1966 to 1970. Wheat still constituted 52 per cent of grain imports during the first half of the 1980s, although the difference was made up by the traditional grains maize and sorghum (Barkin et al. 1990). It is not clear to what extent the continuing preference for imported wheat interferes with the country achieving food self-sufficiency during non-drought years; land, labour, access to technology, and income are probably more important factors.
Political and economic factors
Sociocultural factors influencing tastes and land or labour allocations are closely tied to political-economic forces, especially government policies and market conditions. This section considers other causes of food shortage that are tied to governments and markets - insufficient incentives for production, trade and import constraints, structural adjustment, and food aid. Political and economic factors related to conflict are dealt with in chapter 6.
Insufficient incentives for food production
Government initiatives to change consumption patterns out of economic self-interest are usually ineffective: Nigeria officially stopped rice imports in the mid-1980s in order to save foreign exchange, only to find that, by the end of the decade, rice consumption was still at two-thirds its previous level, owing to illicit re-exports of rice from Benin (Spencer et al. 1989, as cited in Reardon 1993). Efforts to protect local markets from inexpensive foreign grains (including cheaper products from neighbouring countries) are impeded by black market leakage that also distorts the food-availability situations in both donor and recipient countries. Government production and marketing-board policies, by contrast, very much influence what farmers grow. In Kenya, government crop-breeding research, extension, and grain marketing in the 1980s focused on and successfully encouraged the production of maize, not sorghum; and in Tanzania, government crop insurance encouraged farmers in drought-prone areas to assume the greater risk and grow the less drought-tolerant (but, in a good year, more profitable) maize (Louis Putterman, personal communication 1995). Overall, relative crop prices influence what is grown, sold, or later consumed and thus influence food security at the national and household levels.
CASH-CROP VERSUS FOOD-CROP PROMOTIONS. Cash-crop promotions, by governments seeking to increase export earnings, may discourage production of food crops for local consumption through price, tax, and marketing structures. Some governments, as in Kenyan and Tanzanian promotions of tea and coffee, penalize farmers who wish to hedge their bets and maintain some subsistence production alongside cash crops, whose prices are outside the farmer's control. Others, as in the Mexican case cited above, provide incentives in the form of credits or crop insurance to grow particular cash crops, such as sorghum. Commercial export crops also expand relative to food crops where they enjoy the benefits of agricultural research and development, as well as priority access to critical inputs such as fertilizer (Hendry 1988). Where more resources are devoted to coffee than to millet, food availability may decline even while agriculture is modernizing, as was the case in Rwanda. Similar declines are reported where lands that could have produced basic food crops are shifted instead to more profitable livestock, forage, or animal feed (Barkin et al. 1990).
Export production can improve food availability for an individual country, if export earnings purchase more food than might otherwise have been grown. Islam (1994) demonstrated that the majority of countries that had expanded non-food production also experienced an increase in aggregate food supply. But countries that are dependent on raw-materials exports are often severely affected by even moderate shifts in world demand/prices of their products (Sheehan 1987). In addition, diversion of prime rice lands into higher-volume fruits and vegetables may account in part for the plateauing of the Asian GR in recent years.
Production incentives for crops are also limited if surpluses cannot be easily marketed. The US National Academy of Sciences (1986) noted that, in Africa, more intensive production systems tend to be located near railroad lines; in other areas, deteriorating transportation systems have discouraged remote agriculturists from responding to price increases for their products.
PRICE CAPS ON FOOD CROPS. Rural production is, however, more constrained overall by low food prices than by inability to respond to favourable prices. A large number of countries have governmental policies capping the prices of basic foodstuffs. These cheap food policies work primarily to the advantage of urban residents but discourage investments in the agricultural sector. Price ceilings on agricultural goods depress basic food production because rural producers have less incentive to grow food crops, farm larger areas, or intensify their production methods when the rate of return for their efforts is limited. Underpricing of agricultural products on the domestic market has been linked to slow rates of adoption of new agricultural technologies in a number of countries (UNICEF 1988).
Low food prices benefit urban dwellers and manufacturers who desire to keep wages low. Urbanites typically have less access to food through non-market channels than their rural counterparts and, more importantly, they are in a better position to pressure governments into protecting their interests. Even though the majority who are poor lack political power, manufacturers who pay urban wages put pressure on governments to keep urban prices low so that they can keep costs down and compete more effectively in internal and external markets.
In addition, in many countries, peasant producers buy and sell grain on a small scale throughout the year and therefore have conflicting needs as both producers and consumers (de Alcántara 1992). This is particularly true when landholdings are not large enough to meet household food needs, given the expense of agricultural inputs. Cheap food policies may receive political support even from the rural dwellers who would benefit most directly from their elimination.
Even without price ceilings, imported grain may be cheaper than locally produced grain if it is purchased from countries such as the United States that have more efficient, larger-scale production methods and that also subsidize production. Similarly, food aid - even if given only in crisis situations - may remove local incentives to produce surplus because local producers know that emergency carryover stocks are available internationally (Brown and Kane 1994).
Local currency overvaluation also skews internal terms of trade in favour of urban dwellers since it renders imports artificially cheap, and imports drive down prices for domestic produce. The Nigerian naira was so overvalued during the oil boom that imported food was several times cheaper than locally produced food; Smith (1991) described domestic food production under these conditions as "almost irrational."
Higher prices for agriculturists in most cases increase aggregate food supply and contribute to food self-sufficiency. After years meeting its food needs by relying on lower-cost imports, the Belizean government provided incentives to agriculturists that offset the effects of controlled urban food prices; in 1981, Belize became self-sufficient in production of staple foods (Moberg 1992). However, the national marketing board operated at a loss because higher producer prices were not translated into higher consumer prices. Later, when government spending was limited by structural adjustment, the producers' price incentives were removed and the country again became a net importer of food.
OPPORTUNITY COSTS AND FOOD PRODUCTION. An additional consideration is that income from food crops must also be at a high enough level to meet opportunity costs of producing food rather than other cash crops or trade or wage goods. As consumers in the developing world become more integrated into the world economy they need more cash to purchase goods that they now desire but cannot produce themselves. Migration out of rural areas - either of entire families or, more commonly, of individuals - is often a temporary or repetitive strategy to earn cash. The same factors that create one-time need for urban employment often re-create it: patterns of circular migration between cities and the countryside are often repeated year after year. In sparsely populated rural areas, especially in sub-Saharan Africa, this creates a shortage of agricultural labour and a high dependency ratio. Older people and young children often remain in rural areas while those at prime productive economic ages migrate. Households have little incentive to maximize food production if their needs are better met by participation in the cash economy, particularly if some of their needs can only be met by participation in the cash economy. Shifts to less nutritious but less labor-intensive staple crops (especially cassava) have been tied to labour shortages (Benería and Sen 1986; Bukh 1979; FAO 1987; Huffman 1987; Protein-Calorie Advisory Group of the United Nations 1977; Tabatabai 1988; Thaman and Thomas 1985; Ware and Lucas 1988).
Even households that remain oriented toward subsistence agriculture usually need off-farm employment to generate the cash necessary to purchase agricultural inputs and additional food and non-food items. Labour shortage in agriculture has also emerged as a problem in more densely populated rural areas of Latin America, where intensive labour input is necessary to sustain or raise production. De Alcántara (1992) reported that, in Mexico, emigration by some household members in order to generate income for agricultural inputs increases both the need to rely on labour-saving herbicides and the use of hired day-labour. De Alcántara also noted the consequences of labour shortage for the rural community as a whole; households with high dependency ratios have very little labour to contribute to maintaining local irrigation networks and other public works that keep agriculture productive. They also have less incentive to adopt new labour-intensive technologies based on new seeds and chemical packages, if participation in manufacturing and service industries is more remunerative. Especially where farmers do not own land but must pay rent, net revenues after expenses can reduce incentives to grow food.
Producers are also less rewarded for their efforts if a portion of their crop must be paid to the landlord, as is common in sharecropping, and especially if they must repay debts assumed over the cultivation cycle immediately after the harvest, when grain prices are usually low. Even where rents are fixed, cultivators still do not have as much economic incentive to invest in methods that can raise production by retaining soil fertility from year to year. Thus, landholding, access to credit and markets, and relative income all affect production.
TAXATION POLICIES. State taxation policies also reduce incentives for rural agricultural production, since they transfer resources out of the rural sector. Appropriating revenue from the agricultural sector has proven to be one of the most successful short-run methods for increasing state revenue (Hinderink and Sterkenburg 1987). Production is taxed either directly or indirectly, in that many crops must be channelled through government marketing boards which, given low purchase prices, effectively appropriate a share of the produce. As the only legal consumer, marketing boards are in a position to set domestic prices that are typically held artificially low. If produce is sold at world market prices, the marketing board appropriates the difference. The effect of marketing boards is to depress producer prices, sometimes by as much as 85 per cent (Bates 1988).3 Another indirect tax is licensing fees: farmers must pay for the right to cultivate or sell certain crops. All of these practices decrease producers' income. A number of studies conducted in the Philippines and reviewed by UNICEF (1988) concluded that government policies to support the farm sector were insufficient to overcome the negative effects of regulated pricing, government control of trade, export taxes, export quotas, and special levies. Government control of trade in certain Philippine commodities had much the same effect as marketing boards: potential profit was shifted from farmers to government agencies (UNICEF 1988).
Insufficient incentives for food production result from the combination of government policy and market forces. Together, they have created structural problems that interfere with food production and home-generated food supplies.
Trade and import constraints
A country's capacity to import food to compensate for production shortfalls is limited by its export earnings. However, it is overly simplistic to think of export earnings as a direct result of domestic productivity. World terms of trade effectively discriminate against products from developing countries. Although important steps toward trade liberalization have been made in the past decade, tariffs have been reduced to a greater extent on products from developed countries than those from developing countries. Under the General Agreement on Trade and Tariffs (GATT), tariffs are levied in proportion to the level of processing, a convention that exacerbates the inequality between already industrialized nations and those in earlier phases of the industrialization process. The General System of Preferences, which provides for the non-reciprocal reduction of tariffs on developing-country exports, has done little to compensate for this structural inequality because there are multiple individually negotiated preference schemes, many loopholes, and no legal guarantees that the provisions will be followed (Toton 1982).
Within countries, both food availability and revenues from export crops are limited by the ability of producers to market their crops. One of the most positive functions of marketing boards is to facilitate connections with remote rural areas and help develop infrastructure. However, the ability of centralized efforts to improve marketing can be limited by political will, timing of efforts, and available funds. Although the Zambian government in 1985 pledged to haul all maize to safe storage before the onset of rains, the poor repair of roads - as well as the unavailability of grain bags, spare truck parts, and diesel fuel - made this impossible. Good (1986) observed that, even if the problems had been identified in time, government resources would have been insufficient to deal with them. Hence, Zambia was food short in a year with bumper harvests. In other countries, such as Belize, seasonally impassable roads are one of the reasons that urban dwellers rely on imported food (Moberg 1992).
When food is marketed between countries, transportation bottlenecks and lags can become significant issues. Some crops, particularly grains, are more durable, and less is lost if transportation is delayed at some point along the marketing route. For other crops, such as potatoes or vegetables, delays can mean seriously compromised quality or large losses.
Structural adjustment
Many of the political and economic constraints on production are related to heavy debt burdens in developing countries, which exacerbate their marginality and powerlessness to influence terms of trade in the world economy. In addition to the direct costs of financing debt, structural adjustment programmes designed by the World Bank and the International Monetary Fund (IMF) to promote debt repayment alter production relations in ways that can increase food shortage. Structural adjustment can have positive effects on food security in the long run. Restrictions on luxury imports can free foreign exchange for more necessary imports like fertilizer. Currency devaluation, restrictions on food imports, and lifting of food price controls may even increase local production, since some of the disincentives outlined above no longer apply. Food production has increased in some structurally adjusted economies (Meller 1992; Weeks 1995).
However, structural adjustment improves security via mechanisms that require a fairly extended time-frame. In the shorter run, financial austerity programmes associated with structural adjustment more often than not restrict food availability within affected countries. In India, attempts to reduce the government deficit led to decreased expenditures on rural roads, fertilizer subsidies, irrigation, agricultural extension services, agricultural research, and other rural infrastructure (Mukherjee 1994).
Structural adjustment policies that encourage exports of both food and nonfood cash crops to meet external obligations can also create food shortage within regions. A structural adjustment process was implemented in Brazil in 1981; by 1983, per capita production of staple foods had declined by more than 15 per cent while per capita production of sugar cane increased by more than 35 per cent and cultivation of other exportables also expanded (Macedo 1988). Although basic food production rebounded somewhat in subsequent years, it did not grow as fast as export-oriented production, in spite of government programmes that attempted to expand vegetable gardening for local consumption (Macedo 1988). Similar effects have been documented in Peru in response to a more gradual adjustment programme from 1977 to 1985. Annual per capita food production fell by about 26 per cent during that period, and imports only partially compensated for the decline (Figueroa 1988). Thus, adjustment policies can have profound and long-lasting effects on the composition of agricultural production.
Food aid
Aside from these structural or economic issues, imports are also limited by the availability of humanitarian aid. Since there is no global food shortage, food for aid is clearly "available" in an absolute sense, but political agendas of both donor and recipient countries dictate that aid must serve other than strictly humanitarian ends. This finding is no longer shocking, in that US food donations have been made selectively on the basis of military, political, or economic importance at least since the 1970s and 1980s (e.g. Wallerstein 1980), and support to countries that are not of some strategic importance remains limited, although there has been some upsurge in humanitarian emergency assistance in the mid-1990s.
Food aid has successfully expanded commercial markets for producers in donor countries; in addition to fostering heavier dependence on imports, the availability of cheap, non-traditional grains helps change consumption patterns and thereby creates future demand for these grains (Toton 1982). Potentially receiving countries may have very different strategic importance when evaluated by political or military leaders than when evaluated by agribusiness executives. A country is most likely to receive priority for food aid when there is a clear humanitarian imperative, when donor governments are sympathetic to leaders currently in power in the needy country, and when future transfers through market channels are likely. The absence of any one of these factors can be limiting.
Interrelationships between causes of shortage
The environmental, sociocultural, and political/economic causes of food shortage are hardly independent. Even though inadequate fertilizer or water may place biological limitations on production, the main causes of disruption of fertilizer supplies and irrigation systems are economic; even though drought reduces food supply, it rarely leads to food shortage in the absence of armed conflict; even though consumer tastes and preferences are socially and culturally determined, they are heavily influenced by trade and aid.
Food production generally is more and more influenced by international market and policy trends as technology extends even into remote areas and as fewer agriculturists are self-sufficient (Jazairy et al. 1992). The terms of market integration are especially important: smaller producers may be protected against food shortage in cases of local food-crop failure, but they may be more vulnerable to food poverty overall. New agricultural technologies have been key factors allowing food production to keep pace with population growth on the global level, but local levels of food-grain production and incomes have become more variable among those using the technology (Chattopadhyay and Spitz 1987), and affordable food must be available to compensate for shortfalls.
There are some other downsides to increased agricultural productivity. Success in market production does not necessarily mean that production areas earn or produce enough to feed their populations. Vast increases in animal feed production and produce for urban markets in Mexico were accompanied by decreased production of basic foodstuffs. The "success" of these modern agricultural programmes attracted further financial and technical assistance, which reinforced the economic rationality of neglecting basic food production (Barkin et al. 1990). Similarly, favouring of soybeans relative to black beans increased productivity and Brazilian exports, but decreased beans as affordable and essential food for Brazilian producers and consumers. Specializing in export agriculture does not doom a region to food shortage but it does increase local or statewide vulnerability to food shortage and shows that crop choices have important implications for world food supply.