DOI: 10.36347/sjebm.2019.v06i03.003

Pages: 188-197

In this paper, I use national household survey data CHIP2007 (Chinese Household Income Project, 2007) to examine the effect of Social Alms on family’s expenditures. The methodologies I use are propensity score matching and Altonji’s bounding technique. I first adopt propensity score matching and find that people who received the social relief tend to spend statistically significantly more money (333.75 yuan) on the medical insurance and less money on the dwelling expenditure (-653.71yuan) and Durable goods (-493.15 yuan) than their counterparts who didn’t receive the social benefits. Considering there might be selection bias to the estimates, I use Altonji’s bounding technique, which is based on the idea that the amount of selection on the observed explanatory variables in a model provides a guide to the amount of selection on the unobservables to find the selection bias for the effects of social assistance on families’ medical insurance expenditure. Once implement Altonji’s bounding technique based on a non-linear OLS regression, I find that families that receiving social assistance in 2007 tend to spend 393 yuan more on medical insurance than those families who did not receive the social benefits. The policy implications of these findings are also discussed briefly in the paper.