Mark Fossett, Texas A&M University
We use restricted IPUMS microdata from the 1940 US decennial census to examine residential segregation of select population groups in US urban areas. The main contribution of our research is to establish quantitative estimates of the level and nature of segregation for European immigrant groups in US urban areas in 1940. In all we assess segregation for more than 2,400 pair combinations involving Native-Born Whites, Foreign-Born Whites by country of origin, and Native-Born Blacks. By comparing results on two measures of uneven distribution – the Dissimilarity Index (D) and the Separation Index (S) – we document considerable variation in both the level and the form of segregation. When D and S align at medium and high levels uneven distribution follows a “prototypical” form of “polarized displacement” where the two groups live apart from each other in separate areas (e.g., enclaves and ghettos) where their group predominates. When D and S do not align uneven distribution takes the very different but generally unappreciated form of “dispersed displacement” where D is high but S is low and the two groups live together in neighborhoods that are relatively similar on ethnic composition. We document that patterns of segregation for European immigrant groups are complex. Their segregation from other immigrant groups and from native-born Blacks is “prototypical” in form – that is, it involves clear group separation across areas of the city. But their segregation from native-born Whites takes the form of “dispersed displacement” wherein group differences in attaining parity-level contact with native-born Whites does not involve substantial group separation. Finally, we explore a new option for assessing segregation at a spatial scale below the enumeration district using "pseudo-blocks" based on residential proximate households that are located on the same page of a census enumeration form.
Presented in Session 141. Inequality, Segregation, Mobility