In the Eye of the Beholder: Observed Race and Observer Characteristics
PSC Research Report
Population Studies Center at the Institute for Social Research at the University of Michgan
Report No. 02-522
August 2002
36 pages
David R. Harris, Deputy Assistant Secretary for Human Services Policy
U.S. Department of Health and Human Services
Deputy Provost and Vice Provost for Social Sciences
Cornell University
Over the past decade there has been intense debate about racial categories, but surprisingly little discussion of racial categorization. Specifically, there has been little attention devoted to if, and how, characteristics of observers affect observed race. This is troubling because racial classification data are the foundation upon which scholarly studies of race, and policy initiatives to fight discrimination. In this paper I present findings from a web-based survey that was designed to address gaps in our understanding of racial categorization. As part of the Study of Observed Race, 1,672 college freshmen were presented with photographs of individuals, and asked to identify each person’s race. Results indicate that observers’ race, sex, and familiarity with racial groups each influence how they classify by race, and that there are important interactions between observers’ characteristics. I close by discussing the implications of my findings for the 2000 census.
On April 1, 2000, the U.S. Census Bureau did something that was both revolutionary and controversial. Rather than ask individuals to identify with a single racial group, the 2000 Census allowed people to identify with as many as six racial groups. Even casual observers of American racial classification realized that this was revolutionary, as it directly challenged the precedent that individuals identify with one, and only one, racial group. What was less obvious was just how controversial this change in racial classification procedures would be. When the race counts were released in early 2001, few were sure how to interpret these new data. Many wondered what it meant that people had identified with multiple racial groups? For others, there were very real concerns about how to use the data. Should those who identified with more than one racial group be reassigned to a single-race group? If so, how?
In order to make sense of the new census race data, it is imperative that one first understand that these data implicitly assume that the race people select for themselves, or that is selected for them by someone in their household, is the same race that would have been selected by any other observer in any other context. Previous work has challenged this assumption by arguing that self-identified race can vary by context (Harris and Sim 2002), and that self-identified and observer-identified race need not agree (Hahn, Mulinare, and Teutsch 1992; Harris and Sim 2000). In this paper I pursue a further challenge to the assumption of fixed racial classifications by examining if, and how, observed race varies by observer characteristics. If it is true that an individual’s race can be adequately described through self-reports, then observed race should not vary by observer. If observed race does vary by observer, and especially if it does so systematically, then there may be further reasons to question the census race counts.
This work is important for at least three reasons. First, it provides basic information about processes of racial classification. Second, the new census race question is part of a major change in how the federal government measures race. By 2003, all federal race data will be collected using the “one or more races” approach (Office of Management and Budget 1997). Furthermore, because organizations outside the federal government are also likely to adopt this new racial classification scheme, work that affects our understanding of the new census race data will have a broad impact. Third, because a central impetus for collecting race data is to enforce civil rights laws (Office of Management and Budget 1997), and much discrimination is based on observed race (e.g., racial profiling), it is imperative that we understand if, and how, race varies by observer. Failing to do so will limit our ability to effectively enforce civil rights laws…
Read the entire report here.