Abstract
This study focuses on segregation as it plays out at the micro-level of housing unit transition. Employing a unique sample that places housing units into micro-neighborhoods and census tracts, this study tests whether the characteristics of the previous residents of the unit, the local micro-neighborhood, or the broader tract best explain the race/ethnicity of the new residents in a housing unit. The results show that the racial/ethnic composition of the local micro-neighborhood has even stronger effects on the race/ethnicity of the new residents than does the racial/ethnic composition of the broader census tract. The results also reveal that even when the racial/ethnic composition of these two contexts are accounted for, the race/ethnicity of the prior residents has a very strong effect on the race/ethnicity of the new residents. I consider possible explanations for this household-level effect. One new theoretical explanation I put forward is that prospective residents use the race/ethnicity of the prior residents as a signal regarding the neighborhood’s appropriateness for them; I test and find that this hypothesized signaling effect is even stronger in certain micro-neighborhood, neighborhood, and county contexts.