Using Riots to Measure Place Based Effects: The Consequences of Lower Neighborhood Quality on Adolescent Dropout Rates

Noli Brazil, University of California, Berkeley

I study the short and long run effects of diminished community conditions on city level dropout rates by exploiting plausible exogenous variation in city quality through the occurrence of riots across the United States during the 1960s and in Los Angeles during 1992. Using a basic OLS, a regular difference-in-differences, and a relatively new method, synthetic control matching, on decennial census data, I find that riots depressed enrollment rates between 1960 and 1970, with no rebound in the 1970s. While these results indicate both short and long term effects, using yearly California administrative data, I find only short term effects on dropout rates in Los Angeles after the 1992 riot. The results indicate that community conditions do have an affect on schooling persistence rates and points to a potentially more efficient policy method for changing social conditions within a community, namely place based reform.

  See paper

Presented in Session 22: Demographic Impacts of Social and Economic Crisis