Are Big Cities More Segregated? Neighbourhood Scale and the Measurement of Segregation

Abstract

Many studies show that larger metropolitan areas are more segregated than smaller ones. To some extent, this tendency is part of the conventional wisdom. However, the reason for this tendency is not apparent. This paper suggests that the correlation between segregation and metropolitan scale is spurious. Segregation measures based on census data will tend to rank larger cities higher because larger cities have more neighbourhoods that are big enough to 'fill up' entire census tracts, while smaller cities with equally homogeneous, but smaller, neighbourhoods have to pair neighbourhoods to fill up a census tract. This bias will be reduced at smaller levels of spatial aggregation. This prediction is tested by comparing segregation measures computed at several levels of spatial aggregation and with American Housing Survey data. The results suggest that spatial aggregation effects are important: the correlation between city size and measured segregation appears to be at least partly spurious.