Socioeconomic dimensions of poverty of Sri Lanka and policy implications

in that country. The study examines the nature and the extent of poverty in Sri Lanka and suggests alternative policies to address the problem. No great effort was devoted to define a poverty line or to discuss methodology in any detail, since these aspects of the problem are well known. In Sri Lanka, in particular, numerous authors have concentrated on methodological considerations to such an extent that poverty alleviation policies have played second fiddle to such activities. This paper tries to readdress this balance by concentrating on mapping the socioeconomic characteristics of the poor and examining policy alternatives. Has poverty increased over the past decade? The conflicting evidence available gives no clear indication. The most recent estimates done for the World Bank suggest that from 1978-87 all-island poverty rose from 22.3% to 27.4%. There is some evidence to suggest that income distribution worsened over this period, especially with the elimination of food subsidies for poor groups. This conclusion must be tempered by uncertainty over data, price series, and methodological problems. There is general agreement that the poor are concentrated among landless laborers; smallholders; urban, unskilled, informal-sector workers; and unskilled estate workers. However, few studies in the literature identify vulnerable groups by these sorts of socioeconomic characteristics and most concentrate on identifying the poor by decile and geographic location (estate, urban, or rural). The poor are a heterogeneous group. What helps one may hurt another. Thus an increase in the price of rice may help the marginal farmer with a small marketable surplus, but hurt the casual laborer, who must purchase rice. Such heterogeneity makes targeting of the poor difficult if not impossible. We need to know much more about those who fall below the poverty line to target with precision policy measures toward the poor. The poor have smaller than average household sizes and contain slightly more young people than the island average. Ethnicity does not appear to be a major determinant of poverty, with the exception of the Indian Tamils on the estates. The poor have lower levels of education than the nonpoor to the extent that anyone who has achieved `O" level or more is almost certain to be above the poverty level. Most authors argue that unemployment and poverty are closely linked, but this study finds that conclusion unproven. Poverty has been identified across the socioeconomic spectrum in Sri Lanka, with the largest group being the landless laborers. However, if a very restricted definition of poverty is used -- that by Sahn for the ultra-poor -- then pockets of extreme poverty are more prevalent in the urban than the rural areas. We conclude that no amount of provocation of the ultra-poor to drive themselves out of poverty will work, since their circumstances are too desperate for them to have the energy and vigor to do this. But the ultra-poor are a small group (5-8% of the population) compared with the large numbers of poor identified when poverty lines are set at higher levels. Direct grants are probably the main solution for the ultra-poor and should not be too onerous on the exchequer since there are relatively few ultra-poor. The next group of poor can be helped more to help themselves. For example, the provision of small-scale training facilities at the village level plus modest payments for attendance could start this process, and aspects of the revised Janasaviya program seem to appreciate this point. But, no amount of training will help on the supply side if the demand for skills is lacking. Thus, a reexamination of Sri Lanka"s macro development strategy is also required. (Author abstract)