Race Versus Ethnic Heritage in Models of Family Economic Decisions

Race is an important dimension which survey researchers use to examine a number of social phenomena. Despite its importance, few researchers realize the measurement implications of using race as a proxy for experience and culture in statistical modeling. Particularly problematic is the tendency to use race and ethnic heritage interchangeably. This article proposes that one cannot use race and ethnic heritage interchangeably without impacting the results and interpretation. Through a case study, measurement errors in models that use race and ethnic heritage interchangeably to examine family decisions are explored. Results using race are different from results when ethnic heritage is used. This article concludes with a proposed framework for research that contrasts the utility of race and ethnic heritage in statistical models.