The Racial Frontier: Machine Learning, Biracials, and Racial Group Boundaries

• In Event: Multidimensionality in Measurement and the Study of Racial and Ethnic Boundaries

Abstract

A major, cross-disciplinary thread of research has grown this decade which uses Biracials—those who exist at the intersection of our major social cleavages (racial groups)—to reveal the current nature and future trajectory of racial group boundaries. Specifically, researchers explore whether Minority-White Biracials (individuals with one White parent and one Minority parent) tend to be more similar to either Whites or to their Minority counterparts—with the former circumstance auguring the decreased salience of racial group boundaries, prejudice, and minority discrimination. However, one limitation to existing studies is that they tend to analyze only one or a handful of different benchmarks at a time to compare Biracials to their single-race counterparts. In this study, I offer a machine learning approach which can use hundreds of variables at a time in order to compare Biracials to their single-race counterparts. Through an assessment of over 500 political attitudinal variables from the 2020 Collaborative Multiracial Post-Election Survey and Nationscape, I show that Asian-White and Latino-White Biracials often shirk the expectations of hypodescent and resemble Whites, while Black-White Biracials are resolute in their resemblance of Blacks. Results presented here pose major implications for race scholarship both classic and modern in nature, and suggest a renewed salience of the Black/non-Black color line in the near future.

Author

• Gregory John Leslie, The Ohio State University