Measuring ethnicity: challenges and opportunities for survey research

Abstract

Measuring ethnic identity in social surveys has traditionally been problematic, often using a single question and allowing the respondent to choose one category from a pre-defined list. In this paper we discuss the rationale for and limitations of measuring a complex and multi-dimensional concept with a simple, uni-dimensional question. We propose that operationalising ethnicity as multi-dimensional requires multiple questions to capture the complexity of the concept. Giving researchers a number of different measures enables them to focus on the dimensions of interest to them, and has the potential to open up the rich resources of theoretically-robust survey research to researchers from a range of disciplines concerned with questions of ethnic identification.