The assumptions made in program evaluations are often dictated by data availability. In assessing the gains from workfare programs it is common to only have access to a single cross-sectional survey. However, the assumption that the forgone income of participants is zero can still be tested with such data. Propensity-score matching methods of evaluation combine a single cross-sectional survey of program participants with a comparable larger cross-sectional survey from which a comparison group is chosen. With sufficiently detailed cross-sectional data on both participants and non-participants, these methods can allow an assessment of behavioral responses without pre-intervention baseline data or randomization. The accuracy of this method will depend on how well one can assure that treatment and comparison groups come from the same economic environment and were given the same survey instrument. The method cannot rule out the possibility of selection bias due to unobserved differences between participants and even a well-matched comparison group, though there is some evidence this may well be an overrated problem (Heckman, et al., 1998). This paper applies recent advances in matching methods to Argentina's Trabajar Program. While neither a baseline survey nor randomization were feasible options in this case, the program is well suited to matching methods. The authors find that program participants are more likely to be poor than non-participants by a variety of both objective and subjective indicators. The participants tend to be less well educated, they tend to live in poorer neighborhoods, and they tend to be members of neighborhood associations and political parties. The relatively low wage rate clearly makes the program unattractive to the non-poor. The authors estimate the net income gain from the program and find that ignoring foregone incomes greatly overstates the average gains from the program, though sizable gains of about half the gross wage are still found. Even allowing for forgone incomes, the distribution of the gains is decidedly pro-poor, reflecting the self-targeting feature of the programs' design. Average gains are very similar between men and women, but are higher for younger workers. Higher female participation would not enhance average income gains, and the distribution of the gains would worsen. Higher participation by the young would raise average gains, but also worsen the distribution.
Bibliography: Jalan, Jyotsna and Martin Ravallion. 1999. "Income Gains to the Poor from Workfare: Estimates for Argentina's Trabajar Program." Policy Research Working Paper 2149, World Bank, Washington, DC.
