UC San Diego

Recent quantitative studies on the relationship between gentrification and residential displacement have produced inconsistent findings. We examine whether these differences may be attributed to variation in the conceptualization and measurement of gentrification by testing a variety of different operational definitions of gentrification while holding data sources and other methodological decisions fixed. We treat gentrification as a family of related phenomena, estimate a family of operational measures of gentrification from Census data, and, for each measure in the family, test the association between gentrification and displacement in the Panel Study of Income Dynamics. We find that several relationships between gentrification and residential displacement are robust to the choice of measure from the family of gentrification measures we consider. In particular, we find no evidence that gentrification increases the probability of displacement for renters or homeowners, regardless of how gentrification is defined and operationalized. However, consistent with recent studies of particular metro areas, we find evidence that homeowners who live in gentrifying neighborhoods are less likely to be displaced than homeowners in comparable neighborhoods that are not gentrifying.