Name-Based Measures of Neighborhood Composition: How Telling Are Neighbors' Names

Abstract

Name-based ethnicity classification is a common tool in the sampling of minority populations. In recent years, however, it has become a popular technique to construct measures of neighborhood composition if more objective data are unavailable. In this article, we test the accuracy of such name-based measures of neighborhood composition, relying on the example of German neighborhoods. Drawing upon previous research, we assert that ethnic groups differ as to how well they are identifiable via name-based classification. Moreover, the ethnic mix in neighborhoods varies systematically, the ethnicities of immigrants residing in majority-dominated neighborhoods differing from those residing in minority-dominated neighborhoods. Taken together, these two notions imply that a name-based classification bias should be neighborhood-specific. Results indicate a tendency to overestimate majority shares in minority-dominated neighborhoods and slightly underestimate them in majority-dominated neighborhoods. All analyses rely on data from the "Children of Immigrants Longitudinal Survey in Four European Countries (CILS4EU)" as well as on neighborhood compositional data from local statistics of two German cities. The article closes with a discussion of potential strategies to cope with the name-based classification bias.