This book is concerned with rural poverty, not poverty in general. It might be thought that this is an unreasonable narrowing of the subject matter-that rural poverty is fated to decline in importance as urbanization and industrialization occur. And in fact when we see images of third-world poverty they are more often than not drawn from the urban poor-the slums, the malnutrition, the childhood diseases of the millions of poor people crowded into the major cities of the developing world. However, several facts are salient. First, the majority of the poor today-and for the foreseeable future-are found in rural areas. [data]. Most developing countries are still largely rural in their population and agricultural in their economies. Low-income countries had an average urbanization rate of 35 percent in 1988 (25% if we exclude China and India). [1] So two-thirds to three-quarters of the population of developing countries are rural. The agricultural structure of developing economies may be quantified in different ways: as a percentage of the labor force or a percentage of value-added. As figure 2.0 shows, the percentage of labor in agriculture remains well above other forms of employment for most developing countries. (Because agricultural products have lower prices on world markets than industrial products, however, the prominence of agriculture slips when we look at value-added data.) These two dimensions might look as though they coincide, but they do not. Not everyone in rural society subsists on agriculture, and not everyone in industrial employment lives in towns and cities. At the same time, it is generally true that the largest component of the rural economy is agricultural production. [data]
Second, the pang of poverty is generally more severe in rural areas. States are compelled to give some attention to the material interests of the urban poor, or face the destabilizing bread riots that can dislodge a government from power. As a result, most developing societies have some level of socially-funded urban amenities-food subsidies, clean water, education. The rural poor, by contrast, are more easily ignored. They lack the clear threat advantage of the urban poor, and governments can easily be tempted to slight their interests. [data]
Finally, it is true that virtually all developing societies are facing a process of structural transformation from agricultural and rural societies to industrial and urban. But this is a lengthy process, and one that does not move all in one direction. (Indeed, a powerful factor for rural economic development is the emergence of rural industry-the creation of non-urban forms of manufacture and industry.) So even though there is a longterm dynamic leading to the diminution of the problem of rural poverty (replaced, perhaps, by the problem of urban poverty), for the foreseeable future the numbers of the rural poor will be virtually overwhelming, and it will be impossible to imagine a solution to their plight that depends on taking them out of the countryside. The problem of rural poverty, in short, must be addressed directly.
Let us turn now to a consideration of the causal background of rural poverty. What are the social and economic processes that produce poverty within a rural economy? Any economy is a system of production and distribution. Goods are produced and income streams are assigned to persons entitling them to gain access to some quantity of these goods in virtue of their position within the property system and income-generating institutions. In order to understand the causes of poverty, therefore, we need to have a disaggregated view of the institutional structures through which incomes are generated and distributed within a given developing society. [2]
So far we have examined the symptoms of poverty: low income, insecure access to food, and poor performance on a variety of welfare criteria. Let us turn now to a consideration of the causal background of rural poverty. What are the social and economic processes that produce poverty within the rural economy?
How do the rural poor earn their incomes?
Any economy is a system of production and distribution. Goods are produced and income streams are assigned to persons entitling them to gain access to some quantity of these goods. A chief determinant of the distribution and character of poverty, then, is the system of entitlements that a given economy creates for its population: the means through which persons gain income through wages, interest and rent, sales of products, state-funded subsidies, and the like. So we can get an initial view of the physiology of poverty by examining some of the sources of income in typical developing economies that produce chronic and acute poverty.
A substantial shortcoming of neoclassical approaches to development theory is insufficient attention to the institutional determinants of income distribution; but analysis of these institutional arrangements is mandatory if we are to have an informed basis for designing poverty-alleviating strategies of development. Local institutional arrangements-the property system, the institutions of credit, the characteristics of labor markets, and the circumstances of political power-decisively influence the distribution of the benefits of economic growth in existing rural economies. In his major study of the rice economy of Asia, Randolph Barker argues that the Green Revolution technologies themselves do not create greater inequalities, but that unequal ownership of land and capital leads to greater inequalities of income through technical change [Barker, 1985, Rice, : 157]. Barker comments that the decisive factor determining distribution is the set of property relations and institutional arrangements present.
If ownership of these resources is concentrated in a few hands, then their earnings will likewise be concentrated. . . . The effect of resource ownership on the distribution of earnings is so great that any effect caused by technological change is marginal. . . . That does not say that when incomes are increased because of a technological change, all participants benefit equally. On the contrary, they benefit in proportion to their ownership of resources and the earnings of the resources. . . . The important factor determining who receives the direct income benefits is the ownership of resources. [Barker, 1985, Rice, : 157]
Let us turn, then, to the institutional framework that determines the generation of income. A chief determinant of the distribution and character of poverty is the system of entitlements that a given economy creates for its population: the means through which persons gain income through wages, interest and rent, sales of products, state-funded subsidies, and the like, as well as the distribution of ownership rights in productive assets. So we can get an initial view of the physiology of poverty by examining some of the sources of income in typical developing economies that produce chronic and acute poverty.
This category includes small peasant farmers who own enough land for subsistence. The incomes flowing to member of this category depend on farm product prices, the prices of inputs, and the terms of credit available to the farmer.
Tenant farmers include the land-poor and the landless: those farming households that own too little land for self-sufficiency and consequently rent or lease land. This situation leads to the result that a portion of the agricultural surplus flows from the producer to the landlord. The poverty and well-being of the tenant depends, on the one hand, on absolute agricultural productivity, and on the other on farm product prices and the terms of tenancy. There is another important dimension of welfare for tenant farmers, however, that has less to do with absolute income levels; this is the legal and institutional facts that determine the level of tenancy security available to the tenant. To the extent that tenure security is weak, the tenant is exposed both to the risk of dispossession and to chronic upward pressure on rents, reflecting the weak bargaining position of the tenant [Kohli, 1987].
The chief asset in the countryside in developing societies is land. Those who lack land and lack as well the right, opportunity, or resources to rent land are then forced to find other sources of income; chief among these is agricultural labor. The Rudolphs report that about 27 percent of Indian agricultural households were landless between 1954 and 1972 [Rudolph, 1987, #831, : 336]. The well-being of the landless class depends almost entirely on the circumstances of local labor markets: if demand for labor is strong and regular and wages are high, then this group will do fairly well, whereas if demand for labor fluctuates widely and wages are low, it will do poorly. And the latter is the case more commonly than the former. Farm labor is unavoidably cyclical, with peaks of demand around soil preparation, planting, weeding, and harvesting. Farm laborers may be hired by the year or casually; if the latter, their income is unavoidably intermittent.
The average daily farm wage for agricultural labor in West Bengal is 00 rupees. Farm laborers average 00 days of work a year, yielding an average income of 00 dollars a year. This wage leaves the farm laborer's family at the poverty line-if the wage is evenly distributed throughout the year. But, as pointed out above, this is unlikely to be true; as a result, the farm laborer is almost certain to go through extended times of dearth throughout the year, separated by periods of higher income.
The labor market is further disrupted by processes of agricultural modernization and the institutional reforms that often accompany it. Some innovations are labor-intensive; thus multiple cropping increases the demand for labor by increasing the number of crops and tasks that must be attended to. And increased use of fertilizers may increase the demand for labor by increasing the return on such tasks as weeding. Other innovations, however, are labor-displacing. Mechanization of various parts of the cultivation process generally has the effect of drastically reducing the demand for labor; thus one tractor harvester can replace several hundred man-days of harvesting labor in the Malaysian rice economy (Scott 1984:??).
In his careful study of the agrarian economy of Tamil Nadu V. K. Ramachandran arrives at the startling conclusion that agricultural labor has increased as a percentage of the total workforce-a direction of change that belies the common view of economic development as a process of structural transformation from agriculture to industrial production. Consider an illustrative case drawn from Ramachandran's study of Tamil Nadu. The average daily farm wage for a male agricultural laborer in Tamil Nadu in 1976 was 6 rupees. Farm laborers average 109 days of work a year, yielding an average income of 652 rupees per year [Ramachandran, 1990, #795, : 152]. For 1975 the poverty line in Tamil Nadu was 536 rs. per capita per year. This wage leaves the single agricultural worker at the poverty line-if the wage is evenly distributed throughout the year. But, as pointed out above, this is unlikely to be true; as a result, the farm laborer is almost certain to go through extended times of dearth throughout the year, separated by periods of higher income. If the worker has a family, then other sources of income are needed-female and child labor, sideline activities, etc.-in order to bring the household's income up to the poverty line. And-as Ramachandran emphasizes-female labor is substantially less well paid than male labor in the Indian economy.
The labor market is further disrupted by processes of agricultural modernization and the institutional reforms that often accompany it. Some innovations are labor-intensive; thus multiple cropping increases the demand for labor by increasing the number of crops and tasks that must be attended to. And increased use of fertilizers may increase the demand for labor by increasing the return on such tasks as weeding. Other innovations, however, are labor-displacing. Mechanization of various parts of the cultivation process generally has the effect of drastically reducing the demand for labor; thus one tractor harvester can replace several hundred man-days of harvesting labor in the Malaysian rice economy [Scott, 1985, #860].
An important characteristic of rural development is the emergence of rural industries: relatively small-scale facilities producing common commodities (cement, food processing, iron products). These enterprises are typically labor-hiring, and wages are generally higher than those in the farm economy. Rural economies in which there is robust development of rural enterprises generally witness sharp increases in the demand for labor, with increasing trends in wages as well. (See [Vogel, 1989] for an extensive description of this process in Guangdong Province in China in the 1980s.)
There are many intersticial positions in which the rural poor may find themselves within an agricultural economy: street peddlers, petty craftsmen, handicraft producers, magicians, itinerant teachers, and necromancers. The incomes flowing to these positions depend almost entirely on the economic security of the rest of the rural economy; in times of relative boom there will be demand for these services, whereas in times of hardship people will do without them. This means, among other things, that these intersticial positions will be among the first to suffer from entitlement shocks: if incomes to farmers have suddenly decreased, we may expect the shock to be transmitted almost immediately to the martial arts instructor and barber as well. (This observation follows Sen's analysis of famine as the consequence of entitlement shock; [Sen, 1981, Poverty].)
A final response to rural poverty is migration, either permanent or seasonal. And in fact remittances play a large role in many developing economies: for example, hundreds of thousands of Indian workers have migrated to the Persian Gulf states and support their families through wages they send home. More common than long-distance migration are urban migration (in pursuit of unskilled labor in towns and cities) and seasonal agricultural migration (following the peak of agricultural labor demand around harvest and planting seasons).
Here, then, we have a more segmented analysis of the micro-sociology of rural poverty: the economic relations and categories that determine that particular groups will have low incomes. An important lesson follows from this taxonomy: different groups of the poor may be very differently affected by different kinds of economic policies. This is true because the income and security of a given group depends on the stability of the economic relations within which it finds itself, and measures designed to improve one group's lot may actually harm another group. A classic case is the use of state-enforced grain prices to keep food prices low. Policies of this sort may have some immediate benefit for some parts of the rural poor; but the policy has a depressing effect on output (since farmers, large and small, have less of an incentive to increase production). Moreover, contraction of grain production will lead to a contraction in the demand for labor as well, and a reduction in the incomes flowing to small farmers. (See [Varshney, 1989] for analysis of these sorts of effects in India's food price policies.)
So far we have looked at the institutional framework of poverty. But there is also an important geographical dimension to poverty: some regions are inherently poorer than others. There is an uneven distribution of resources across any national economy. Some regions have good cropland, whereas others have poor soil. Some regions contain extensive natural resources-coal, oil, or mineral deposits. Some regions are advantaged within the transportation system (ports, rail and road hubs, etc.), thereby making investment and economic activity more attractive to outsiders. And it emerges frequently that these patterns of unequal distribution of assets tend to coincide; so that poor farming areas are also poorly served by the transportation system, have low levels of social investment in health and education, and have low levels of non-agricultural economic activity. Disadvantaged and peripheral areas will tend to be poorer across the board than more advantaged areas. Thus China's economy since 1949 has largely succeeded in reducing intra-regional inequalities of wealth and income through land reform and other social policies. But it has been much less successful in evening out inter-regional inequalities [Lyons, 1991]. Coastal cities and their hinterlands have gained substantially in past decades, whereas interior provinces have lagged behind. The most acutely disadvantaged provinces are in the southwest and northwest of China [Perkins, 1984, #748]. [3]
There is a similar geographical pattern within regions as well. G. William Skinner has documented for China a pronounced core-periphery structure of the economy, in which resources are concentrated in urban cores and thinly spread on peripheries [Skinner, 1977, Urbanization; Skinner, 1977, Hierarchy; #899]. This pattern is paralleled by the distribution and character of poverty; there is typically a lower average income in peripheral areas, with a concomitantly higher rate of poverty. This pattern also imposes an administrative handicap on poverty alleviation. Poverty occurs most intensively in peripheral areas that are generally on administrative borders. So effective development and poverty alleviation requires complex and unlikely levels of coordination between administrative units (counties, provinces).
A clear result of this analysis of the diverse social positions of the poor has to do with the structure of stratification within a developing economy. We may think of the institutions and economic relations that define a given economy as a distributive system, channeling flows of income to various groups. And it is apparent that there is substantial inequality in most such systems in the developing world, with large streams of income flowing to some social groups and irregular trickles of income flowing to others.
A second lesson that we may draw is the centrality of ownership of assets in the distribution of income. Land is often very unequally distributed; access to credit is very uneven; ownership of capital is very narrowly concentrated. The poor are poor, in large part, because they control few assets beyond their labor power. It is a familiar truth that productive activity requires proper tools; in economic terms, value-added is a function of the amount of capital set into motion by a quantity of labor. The poor have very little access to capital goods; consequently, the products of their labor have relatively little value, and their incomes remain low.
The low productivity of the labor of the poor is exacerbated by the small degree to which the poor have access to the services that would enhance the value of that single asset (education, health care). Illiterate, innumerate workers are less productive than their better educated counterparts; with the result that their incomes are lower as well. So a reasonably direct way of improving the welfare of the poor is to increase the productivity of their labor through education and training.
Third, the economic condition of the poor depends a great deal on the character and quality of government programs for social welfare: food subsidy programs, rural health and education programs, credit regulations and provision, and the like. The state is a substantial player (often by inaction rather than action). But equally important are negative effects of state policy: anti-rural bias in agricultural policy, restrictive migration policies, anti-agricultural bias in national credit programs, and the like.
Finally, preceeding analysis shows that the particular institutional arrangements of the rural economy have a profound effect on the character and distribution of poverty. The most general economic institution is the market: the institutional setting in which buyers and sellers, producers and workers, meet and exchange products at market-determined prices. We have seen that market relations conjoined with sparse assets and low-productivity labor skills guarantees low income to the poor; if we are to overturn this outcome then specific steps must be taken to offset the workings of the market.
But other more specific institutions are pertinent as well. The details of the land tenure system determine the relative shares of farm income that flow to tenant and landlord-thus profoundly affecting the ability of the tenant to survive hard times. The terms of labor hiring likewise have substantial effects on the well-being of workers: where employers are required to meet minimal conditions of wage, security, health, and safety, workers who find employment will be better off than otherwise. Economic institutions have the effect of channeling income into the hands of various groups; to the degree to which these institutions are skewed in the direction of the interests of the rich, the poor will suffer. (And, as will be argued in chapter 6, there is every reason to expect just such a skew, since the rich have greater access to political power than the poor.)
A clear result of this analysis of the diverse social positions of the poor has to do with the structure of stratification within a developing economy. We may think of the institutions and economic relations that define a given economy as a distributive system, channeling flows of income to various groups. And it is apparent that there is substantial inequality in most such systems in the developing world, with large streams of income flowing to some social groups and irregular trickles of income flowing to others.
A second lesson that we may draw is the centrality of ownership of assets in the distribution of income. Land is often very unequally distributed; access to credit is very uneven; ownership of capital is very narrowly concentrated. The poor are poor, in large part, because they control few assets beyond their labor power. It is a familiar truth that productive activity requires proper tools; in economic terms, value-added is a function of the amount of capital set into motion by a quantity of labor. The poor have very little access to capital goods; consequently, the products of their labor have relatively little value, and their incomes remain low.
The low productivity of the labor of the poor is exacerbated by the low degree to which the poor have access to the services that would enhance the value of that single asset (education, health care). Illiterate, innumerate workers are less productive than their better educated counterparts; with the result that their incomes are lower as well. So a reasonably direct way of improving the welfare of the poor is to increase the productivity of their labor through education and training.
Third, the economic condition of the poor depends a great deal on the character and quality of government programs for social welfare: food subsidy programs, rural health and education programs, credit regulations and provision, and the like. The state is a substantial player (often by inaction rather than action). But equally important are negative effects of state policy: anti-rural bias in agricultural policy, restrictive migration policies, anti-agricultural bias in national credit programs, and the like.
Finally, preceding analysis shows that the particular institutional arrangements of a rural economy have profound effects on the character and distribution of poverty. The most general economic institution is the market: the institutional setting in which buyers and sellers, producers and workers, meet and exchange products at market-determined prices. We have seen that market relations conjoined with sparse assets and low-productivity labor skills guarantees low income to the poor; if we are to overturn this outcome then specific steps must be taken to offset the workings of the market.
But other more specific institutions are pertinent as well. The details of the land tenure system determine the relative shares of farm income that flow to tenant and landlord-thus profoundly affecting the ability of the tenant to survive hard times. The terms of labor hiring likewise have substantial effects on the well-being of workers: where employers are required to meet minimal conditions of wage, security, health, and safety, workers who find employment will be better off than otherwise. Economic institutions have the effect of channeling income into the hands of various groups; to the degree to which these institutions are skewed in the direction of the interests of the rich, the poor will suffer. (And, as will be argued in the final section, there is every reason to expect just such a skew, since the rich have greater access to political power than the poor.)
[1] World Development Report 1990:238.
[2] This approach reflects two important streams of thought in development theory: Marxist economic theory and A. K. Sen's theory of entitlements. Marxist theory emphasizes the relationship between incomes and ownership of assets (Bardhan 1988); while Sen's approach emphasizes the diverse opportunities for exchange that a given individual possesses (Sen 1981).
[3] Walther Aschmoneit has constructed a Quality of Life index based on the Chinese 1982 census that bears out this pattern (Delman et al 1990).