Adopting an empirical strategy developed by David Cutler and Edward Glaeser in their 1997 QJE paper, “Are Ghettos Good or Bad?”, my research employs public use micro data from the 2000 Census and broadens their original question to examine the impact of residential segregation in the case of Hispanics and Hispanic subgroups as well as for African Americans. This approach involves using residential dissimilarity indices measured across census tracts at the metropolitan statistical area (MSA) level to determine whether minorities living in MSAs with higher levels of such dissimilarity fare worse with respect to certain outcomes than those who live in less segregated areas; the outcomes considered include educational attainment, employment and earnings, and single parenthood among (non-immigrant) young people aged 20-30. Results are obtained using PROBIT, OLS and 2SLS.
As Cutler and Glaeser did using 1990 Census data, I find that blacks in more segregated areas continue 10 years later to experience significantly worse outcomes than those living in less segregated areas. The same general pattern is observed for Hispanics as a whole, although Hispanic vs. White, non-Hispanic dissimilarity scores tend to be lower than those seen in the case of Black vs. White, non-Hispanic segregation.
When I break down the all-Hispanic group into its major subgroups, an interesting set of results emerges. For Hispanics of Mexican heritage, who comprise the largest such subgroup, the general pattern persists, but the size of the coefficients of interest are considerably smaller than those observed for the aggregative group. For Hispanics of Puerto Rican heritage, the opposite occurs: larger coefficients along with a general pattern of results most similar to that seen for African Americans. Larger still are the apparent effects of residential segregation among Hispanics of Dominican ancestry. I find no significant relationship between segregation and socio-economic outcomes among those of Cuban descent. For Hispanics with Central and/or South American ancestry, I find only marginal evidence in support of an interpretation that ‘barrios are bad.’
This research forms the basis for an ongoing project that will extend and deepen the analysis using data from the 2010 Census.
Sean Flaherty
Professor of Economics
Franklin and Marshall College
Lancaster, PA 17604
sean.flaherty@fandm.edu