The problematic status of statistics on race and ethnicity: An "imperfect representation of reality."

Abstract

This article extends Stratford's brief observations about the problematic status of racial and ethnic group statistics to a discussion of the relationship among these statistics, public policy, and the conceptual status of race and ethnicity. Federal statistics are organizational products that are socially constructed. They represent the implementation of public policies that govern political, social, and economic life. It is the interaction between politics and the subjective meaning of race and ethnicity that is responsible for the continual modification of racial and ethnic group statistics. The article discusses the premises on which racial and ethnic group statistics have been based and illustrates how they were implemented in the instructions of the decennial censuses for classifying the race and ethnicity of the population. The article then summarizes some of the empirical evidence from recent research conducted by federal agencies and social scientists to show that racial and ethnic group statistics produced by government record keeping systems have no objective status. The meaning of race and ethnicity is contextual, situational, and subjective, and, thus, how respondents and observers define these concepts has significant consequences for the quality of federal statistics.