Re-examining Neighborhood Eects on Local Political Participation - An application of the generalization of propensity score matching.

This paper reevaluates socioeconomic neighborhood effects on political participation taking account for the effects from sorting. Past literature of political science has identified that poverty and socioeconomic homogeneity are the two major neighborhood-level inhibitors of political participation. Those studies, however, ignore the effects of sorting either theoretically or statistically. Given the magnitude of sorting having occurred after WWII, there is enough room to suspect that neighborhood effects simply represent the difference in the dispositions of residents between neighborhoods. Moreover, randomized social experiment is not necessarily the best method in this case because it tends to pick up the applicants who have high level of voice. Using Social Capital Benchmark Survey matched with tract level census and methodologically relying on the generalization of propensity score by Imai and van Dyk(2004), this paper attempts to elucidate the heterogeneity of neighborhood effects. The results show that consideration of the propensity score makes most statistically significant neighborhood effects under parametric model disappear. Breakdown coefficients also suggest that sorting is not necessarily the inhibitor of political participation, but mismatched residency is. ..PAT.-Unpublished Manuscript [ABSTRACT FROM AUTHOR]