Modeling the Structure of Multi-Racial Identity: A Latent Class Approach

Ross Macmillan, University of Minnesota
Carolyn A. Liebler, University of Minnesota

New data on race and Hispanic origin provide a vast array of information about the racial and ethnic composition of the United States. While data collection is well-advanced, theoretical and methodological developments are only beginning. This paper articulates and applies latent class analytic techniques to data from Census 2000. Specifically, we analyze the underlying structure of racial identity based on multiple racial indicators. The first aspect of our work maps out the latent structure of racial identity in the country as a whole, while the second aspect makes systematic comparisons across regions. We use these analyzes as a springboard for theorizing the social structuring of racial identity in the United States and as a foundation for consideration of social, economic, political and demographic factors that shape the structure of racial identity.

  See extended abstract

Presented in Session 3: The Measurement of Race and Ethnic Origin