Evolution and evaluation of poverty alleviation programs in Sri Lanka

Since independence Sri Lanka has made impressive progress in reducing poverty. This reflects in terms of improvement in human development indicators as well as in consumption terms. Even though the progress in reducing consumption poverty has been impressive, substantial poverty still remains in Sri Lanka. The purpose of the research is to examine and evaluate the performance of welfare programs in Sri Lanka since 1979. Within this broad goal our specific objectives are to formulate important aspects of welfare programs in measurable terms and then estimate them using two alternative sources of Sri Lankan data. We compare the evolution of these attributes over time and across the three welfare programs that were in place during our period of study. We also examine if there is any correlation between increased government expenditure on welfare, and the efficiency of the programs, estimate the relative contribution of efficiency and spending, and suggest possible measures for improving the present system. The use of two sources of data where neither is expected to provide compelling conclusions has been motivated by the consideration that they would indicate the ranges over which the actual magnitudes may lie. The qualitative conclusions from the two exercises are mostly similar. The decline in the quantum of transfer needed for poverty alleviation over the period is common to both exercises. On the other hand the expenditure on transfer programs as a ratio of existing poverty gap is found to have increased steadily in both exercises. However, success of the programs, as measured by the reduction in poverty gap as a percentage of pre-transfer poverty gap, failed to match the decline in need and the progress in spending - as shown in both studies. This obviously points to the necessity of tightening the efficiency of the programs. Of particular concern is the estimated negative correlation between efficiency of the programs and an index of government transfers. It implies that parts of the transfer fail to produce results, not only because of the existing inefficiency, but also by somehow generating additional inefficiency in the system.