Over the past decade and a half, the number and proportion of college students opting not to reveal their race when asked have shot up,  to 5.9 percent of all students in 2001 from 3.2 percent a decade earlier. The increases have raised two major questions: Who are these students, and why are they declining to identify themselves? The answers have implications for college officials and policy makers on a wide range of issues, including affirmative action and student life.
A new study  by the James Irvine Foundation's Campus Diversity Initiative takes a first step at answering the "who they are" question. The report, " 'Unknown' Students on College Campuses: An Exploratory Analysis," examines undergraduates at three private colleges in California and finds that a "sizable" number of those who declined to identify their race were white. That means that on the three campuses, the study found, the proportion of enrolled students who were white rose by anywhere from 10 to nearly 30 percent.
"As many campuses report progress toward compositional diversity by touting the presence of either underrepresented minority students or students of color generally, our findings suggest that the racial/ethnic composition can be distorted when there is a large unknown population," the authors write. "Even if a relatively small portion of this group is white, it will change the demographic diversity of a campus and have repercussions in terms of the handling of this category in data reporting."
The study, which was a collaboration between the Irvine Foundation, the Association of American Colleges and Universities, and Claremont Graduate University, compared information gathered from student applications during the admissions process at the three institutions to data collected during the Cooperative Institutional Research Program Freshman Survey done annually by the University of California at Los Angeles's Higher Education Research Institute. (The authors emphasize that the two sets of numbers did not match up perfectly, which explains in part why they use the word "exploratory" in the report's subtitle.)
Many more students tended to identify themselves by race (often more than one) in the UCLA survey than they did during the admissions process. The authors speculate that that may be because the information is collected after students have enrolled, so any perceived impact of students' racial classification on their admissions chances -- which is one commonly held theory for why students may withhold their racial identities -- may be mitigated.
The comparison worked this way on one of the colleges -- "Campus A" -- in the study. Of students who entered in 2000, 32 percent were "unknown" because they did not choose a racial category. Of admitted students who did identify themselves, 42 percent were white, 12 percent Asian American/Pacific Islander, 11 percent underrepresented minority (presumably black and Hispanic), 3 percent other.
Seventy percent of respondents to the UCLA survey on that campus said they were white, but since the students were permitted to choose more than one category, some of them may be multiracial. So the Irvine Foundation study estimates that between 57 and 70 percent of Campus A's students are white.
The authors are careful to say that the results may not be broadly applicable to other institutions, since private institutions like those that the Irvine Foundation studied represent about 10 percent of all American colleges and universities, and educated a relatively small proportion of all students. Still, they say that their findings "may have dramatic implications for higher education institutions."
"Over the course of several decades of working with campuses to establish diversity initiatives," the authors write, "we have encountered many people who assume that most, if not all, students in the unknown category are multiracial. This study directly challenges that assumption."
The report calls for better and more accurate data collection about the racial makeup of students, both to "eliminate our reliance on assumptions about unknown students and establish a way of collecting more accurate official enrollment data on all students. With this more accurate data, we will have not only a better sense of the true racial/ethnic composition of our colleges and universities, but also a better gauge of the access various students have to, and the success they have through, higher education."