Together with high economic growth, redistribution of public funds has been an important topic for both researchers and policy makers over the last few years. Since 1997 there have been at least six analyses on the incidence of public transfer programs all using the 1992/93 and 1997/98 household surveys (VLSSs). The general conclusion in these studies is that the social transfer system is regressive, and that it does not provide effective social protection to the majority of the poor. With the availability of new data, namely, the 2002 and 2004 household surveys (VHLSSs), the incidence analysis can be updated. We seek answers to the following questions: what is the incidence of the redistribution programs? Are they targeted to the poor? What are the roles of these programs in poverty dynamics? In doing so, we challenge many of the previous studies. We raise concerns about (1) the choice of the counterfactual welfare measure, and (2) the appropriateness of the household survey data for doing the analyses.

In the analysis, we look at two types of transfers: social insurance and social protection, and two types of education fee exemptions: tuition fee exemptions and school contribution fee exemption. The results give a somewhat different picture of the distribution of social transfers than the studies based on the 1992/93 and 1997/98 household surveys. Fist of all, social transfers are not regressive in 2004. Second, the transfers had a sizeable impact on the level of poverty in both 2002 and 2004 and, furthermore, provided protection against falling into poverty in that period.