A 1997 World Bank Study has rekindled the debate on whether economic growth is central to poverty alleviation in India. Invoking the main findings of a research project based on household surveys over a 40-year period, the study, India: Achievements and Challenges in Reducing Poverty,(1) claims that "overall growth accounted for the lion's share of poverty reduction: 80 percent of the decline in the number of households below the poverty line between 1951 and 1970, and almost 100 percent since 1970."(2) Further, the findings "clearly refute any presumption of immiserizing growth--that is growth that tends to marginalize or impoverish significant sections of the population--in India over this 40-year period."(3) The message therefore is that "it is through rapid growth that India will be able to reduce poverty and generate the resources to invest in the health and education of its people--who will in turn sustain this growth."(4) Although the overall antipoverty strategy is similar to that advocated in the 1990 World Development Report--in as much as the main components are labor-intensive growth, human capital development and social safety nets--the former is distinguishable in its greater emphasis on growth.
In the present paper, we contest this claim. While growth matters, its contribution to poverty alleviation is much less than the World Bank claims. The related claim that growth did not immiserize or impoverish certain sections is inconsistent with panel evidence. In fact, there is a large core of poverty that persists despite rapid growth. The issue then is whether there are specific factors that tend to exclude the poor from participating in growth. In this paper we will attempt to identify some key factors through an analysis of a panel survey. As noted, the primacy of growth in the World Bank strategy rests not just on its direct poverty-reducing effect, but on the generation of additional resources for poverty alleviation (interpreted broadly to include human capital development). The latter claim is examined with state data on two major anti-poverty programs, the Jawahar Rozgar Yojana (JRY) and the Integrated Rural Development Program (IRDP). Finally, based on the analysis of factors that impeded participation of sections of poor households in growth, we elaborate some major components of an anti-poverty strategy. Although there is some overlap with the World Bank strategy (specifically regarding human capital formation), the emphases differ. The present study focuses more on the design and implementation of this growth focused strategy. In this context, community initiatives have a potentially significant role. But such initiatives seldom work, as the impediments are not easy to overcome. A review of some recent field evidence on the functioning of the Panchayats (an elected body), following the implementation of the 73rd Constitutional Amendment Act in 1993, yields useful insights. Although a strong coalition of the poor would help overcome some impediments, it is not obvious how the coalition could be strengthened.
EXTENT AND SEVERITY OF POVERTY
Despite over four decades of planned development in India, a large section of the population continues to live in abject poverty.(5) According to a National Sample Survey for 1993 and 1994, about 36.7 percent of the rural population and 30.5 percent of the urban population live in poverty--implying a national average of 35 percent.(6) Although there has been some reduction in the proportion of the poor over time--especially between the mid-1970s and late 1980s--their aggregate number has risen. From about 164 million in 1951, it rose to 312 million in 1993 to 1994. There are also sharp regional disparities. In Bihar, for example, the proportion of poor in the rural population (64 percent) in 1993 and 1994 was two and a half times higher than that of the combined states of Punjab and Haryana (25 percent). Similarly, there are wide variations within states.(7) Of the 10 poorest regions in 1987 and 1988, five were either in Bihar or Orissa. Poverty was evenly distributed across the five regions in these two states. But in Uttar Pradesh (UP)--another poor state (with about 42 percent of the rural population classified as poor)--barely 8.4 percent of the rural population in the Himalayan region was poor, the lowest incidence in all of India. Classifying all of India's rural population into various groups, 68 percent of the landless wage earners, 51 percent of scheduled tribes/castes (STs/SCs) and 45 percent of illiterates were poor in 1993 and 1994.(8)
Using the poverty gap index to measure depth of poverty in seven out of the 14 states for which estimates are available, the gap in rural areas ranged from a low of about 5 percent in West Bengal to a high of well over 17 percent in Bihar in 1993 and 1994. Given the low poverty threshold, these estimates point to the extreme poverty of large segments of the rural population in several states.
As noted earlier, the 1997 World Bank Study asserts the centrality of growth in poverty alleviation. The rate of reduction in poverty was indeed faster during periods of rapid growth (i.e., between the mid-1970s and 1980s); the average annual reduction (2.4 percent) was three times as fast as during a period of slow growth (i.e., from 1951 to 1975). These patterns hold for both rural and urban areas. At the state level, while the centrality of growth in poverty alleviation is maintained, the study notes that human capital and infrastructure have important roles as well.(9) However, as argued below, an effective anti-poverty strategy must have broader concerns.
IMPACT OF POLICY REFORMS
Of some concern is the impact of policy reform, initiated in India in mid-1991, on the poor. These reforms were a response to the macroeconomic crisis that erupted in early 1991.(10) Following policy reforms, the head-count ratio for rural areas rose sharply--from 36.4 percent in 1991 to 43.5 percent in 1992--and dropped to 36.7 percent in 1993 and 1994. While the 1997 World Bank Study reports that no more than one-third of the initial increase in rural poverty is attributable to policy reforms,(11) how much of the increase is attributable to such reforms remains a debatable issue. It is plausible that some impoverishment was a result of these reforms, even if an allowance is made for (random) variability of incomes from cultivation and associated variability of wages.(12) Equally, the subsequent reduction in rural poverty in 1993 and 1994 should not be viewed as a result of policy reforms leading to a higher overall growth rate. In general, whatever the contribution of the policy reforms in accelerating growth in recent years, we argue below that growth alone will have a limited impact, as there are large segments of rural poverty that persist due to the inability of some sections of the population to participate in the growth process.(13)
M. S. Ahluwalia's influential contribution to the debate on growth and poverty provoked a lively discussion on whether agricultural growth `trickles down' to the poor.(14) Several critiques drew attention to the impoverishment of some sections as a direct consequence of growth.(15) Given the inequality in endowments of both physical and human capital, as well as a rigidly hierarchical social structure, it is not surprising that some sections benefit while others lose when economic opportunities expand. Under certain extreme conditions, when inequality in the distribution of land manifests itself in an oligopsonistic labor market with few large landholders dominating through wage-setting, shifts in labor demand may have a small effect on agricultural wages relative to that in a competitive labor market. Consequently, poverty persists despite growth.(16) Indeed, if dominance of large landholders also precludes some remunerative supplementary options (i.e., working part-time on another farm), poverty could worsen through a dampening effect on the earnings of agricultural laborers. A recent analysis confirms that while agricultural growth reduces rural poverty, a measure of oligopsony exacerbates poverty.(17)
The 1997 World Bank Study relies on a decomposition of change in rural poverty into three components: growth, redistribution and the residual. If poverty falls over time, part of the reduction may be due to a higher average income without any change in its distribution (i.e., the growth component); some reduction may be due to a change in the distribution itself without any change in the average income (i.e., the redistributional component); while the remaining reason for reduction may be due to the interdependence between growth and distributional change (i.e., the residual). The issue is whether the favorable effect of growth is neutralized or reinforced by income distributional change. The study found that growth accounted for 80 percent of the decline in the head-count index over the period 1951 to 1994 and almost 100 percent since 1978. This, however, is suspect.
Two problems with this decomposition procedure suffice. One is the presumption that the growth and distributional components are separable. In case they are not, the decomposition is ambiguous. Figure 1, based on 1997 World Bank estimates, illustrates vividly that changes in mean consumption and changes in the Gini coefficient in Uttar Pradesh--one of the poorest Indian states--moved together over the period 1959 to 1994 (i.e., higher values of change in consumption were associated with higher values of change in the Gini coefficient).(18) Another problem is that changes in consumption distribution understate income distributional changes when there is consumption smoothing.(19) To that extent, the redistributional component would therefore be lower.
[Figure 1 ILLUSTRATION OMITTED]
A surprising claim in the 1997 World Bank Study is that the decomposition refutes any presumption of immiserising growth over the...