By Robert Lerner, Ph.D. and
Althea K. Nagai, Ph.D.

Executive Summary

For nearly 30 years, racial and ethnic preferences have played a key role in how admissions officers at the nation's public and private colleges and universities have chosen their undergraduate classes. A system of racial and ethnic preferences in admissions operates by establishing different standards of admission for individuals based on their racial or ethnic background, with some students held to a higher standard and others admitted to a lower standard.

Earlier in this century, some colleges and universities denied admission to Jews, blacks, women, and members of other racial and ethnic groups even when their grades, test scores, and other measures of academic achievement surpassed those of white males who were offered an opportunity to enroll. The passage of new civil rights legislation in the 1960s made this kind of blatant discrimination illegal.

Since then, however, many colleges and universities have created programs meant to boost the enrollment of students whose backgrounds previously had excluded them from pursuing a higher education-especially blacks and, to a lesser extent, Hispanics-by granting them preferences during the admissions process. These policies, when their existence was made public, became immediately controversial, and they remain so today.

Defenders of preferences claim that these policies are not discriminatory and help administrators choose between equally or almost equally qualified students, giving a slight edge to applicants who likely have faced discrimination or have come from disadvantaged backgrounds. Critics of preferences say that these policies are no better than the discriminatory ones they replaced and that, in any event, the advantages they confer upon certain applicants are much greater than supporters are willing to admit.

Public colleges and universities have seen their ability to use Affirmative Action increasingly restricted. Court decisions have generally been hostile to such preferences. The enactment of California's Proposition 209 (also known as the California Civil Rights Initiative) forbids discrimination against or granting special treatment to any applicant on the bases of race, ethnicity, or sex in the public programs of the country's most populous state. Grassroots activists elsewhere are trying to place similar proposals on their own state ballots. Eventually legislators, both in Congress and in state capitals, may enact legislation modeled on the new California law.

This study examines the extent to which racial and ethnic preferences are used in the admissions policies of two public universities in the state of Washington. It submits admissions data supplied by these universities to a rigorous statistical analysis.

This report is the latest in a series published by the Center for Equal Opportunity (CEO), a public policy research organization. Earlier CEO studies have focused on the public colleges and universities of Colorado, Michigan, and North Carolina, the U.S. service academies at West Point and Annapolis, and the branches of the University of California at Berkeley, Irvine, and San Diego. Previous reports have shown that blacks and Hispanics often receive an enormous degree of preference in undergraduate admissions.

Methodology

The two universities studied here are the University of Washington (UW) and Washington State University (WSU). Each provided data on the fall 1995 class of incoming freshmen. The 1997 edition of Barron's Profiles of American Colleges ranks UW as "very competitive" and WSU as "competitive."

We requested data in the form of individual records (with all personal identifiers removed) on rejectees, enrollees, and admittees who did not attend UW and WSU as to their racial or ethnic group membership, verbal and math SAT scores, and high school grade point average (GPA). Our efforts were not entirely successful. UW provided individual records on rejectees and admittees who did not enroll at UW.

But unlike all the other colleges and universities we have previously studied, UW refused to provide us with individual records for enrollees. Instead they provided grouped summary data for student GPAs and percentile charts for SAT scores. WSU, which did provide individual records for all admissions statuses, provided information in the form of separate files for test scores (versus grades) for rejectees, admittees who did not enroll, and enrollees.

These data restrictions made it impossible to carry out certain kinds of analyses, such as those predicting the odds of admissions, a technique which we have used in our earlier studies. The restrictions also make other analyses more difficult to carry out than would otherwise be the case. Nonetheless, a clear picture of the situation at these two institutions can be drawn using the data made available.

As in our previous work, we omit from our data analyses those cases for which race or ethnicity is listed as "other," "missing," or "unknown." We also omit American Indians because of their small numbers in this context. Lastly, we omit cases with missing academic data.

Also as before, we do not report group means for test scores or GPAs. Using group means places greater weight on extreme values than is warranted. A few unusually high or low scores can have a substantial effect on the value of the mean. Standard deviations, which are based on squared deviations from the mean, are even less useful for describing the spread of cases for asymmetrical, badly skewed distributions. This is because standard deviations reflect the mathematical square of these extreme values.

The median, however, and related statistics are far less affected by the values of extreme cases. The median represents the middle of the distribution so that 50 percent of all students have higher scores, and 50 percent have lower scores.

We also report scores at the 25th and 75th percentile, again to deal with the problem of extreme cases. While the median represents the middle of the distribution, the 25th and 75th percentile scores taken together represent the actual spread of scores. For example, a 25th percentile score of 650 means that 25 percent of the scores were below 650, while 75 percent of the scores were above it. A 75th percentile score of 700 means that 75 percent of the scores were below 700, while 25 percent were above it.

Racial and Ethnic Differences in Enrollee Qualifications

Applicants, Admissions, and Enrollments

Figure 1 shows the raw admission rates at the schools studied. At UW, blacks and Hispanics are admitted at much higher rates than Asians and whites. At WSU, Hispanics, Asians, and whites are admitted at roughly the same rates, and blacks are admitted at a lower rate than the others are.

We examine three pairs of differences in qualifications: white-black, white-Hispanic, and white-Asian. Treating each pair of comparisons separately makes it easier to see whether substantial differences in racial and ethnic qualifications exist, for which school they are greatest, and between which groups they are the greatest.1

Differences between Whites and Blacks

White enrollees on average have considerably better credentials than do black enrollees, regardless of whether we examine SATs or GPAs. Figure 2 shows the white-black gaps in verbal SATs, math SATs, and GPAs. The white-black gap in verbal scores at UW is 80 points; at WSU, it is 70. The math gap is larger: at UW the white-black gap in median scores is 140 points, and at WSU it is 110. There is also a gap in GPAs: at UW, the white-black gap is almost a half a grade point (0.47), while at WSU it is 0.37 of a grade point.

This is strong evidence of the operation of Affirmative Action at both institutions in favor of blacks.2

Differences between Whites and Hispanics

White enrollees also have better credentials on average than Hispanic enrollees, although the white-Hispanic differences in scores and grades are smaller than are the white-black gaps.

There is a gap between whites and Hispanics in verbal SAT scores. At UW it is 30 points. It is 40 points at WSU. At UW, the math gap is 70 points. It is 25 points at WSU.

There is a modest difference between white and Hispanic GPAs. At UW it is 0.18 of a grade point, and at WSU it is 0.12 of a grade point.

These gaps between whites and Hispanics provide substantial evidence of racial preference in favor of Hispanics at both UW and WSU. The evidence is not as strong as evidence for preferences favoring blacks over whites, however.3

Differences between Whites and Asians

There is little evidence of racial preference in favor of Asians over whites. The white-Asian gap in verbal scores is moderate. White enrollees at UW outscore Asian enrollees by 60 points at UW and 30 points at WSU.

The math gap is mixed. At UW, white enrollees score somewhat higher than Asians-20 points-but the reverse is true at WSU, where Asians outscore whites by 20 points.

Lastly, there is virtually no gap in GPAs between whites and Asians at either school. The differences in favor of whites are 0.02 and 0.08 of a grade point, respectively.4

How Preferences Affect Graduation Rates

If students gain admission to colleges and universities for reasons other than their academic preparation, it is likely that they will face greater burdens in school than their peers who have met a higher academic standard of admission. They may in fact not complete their degrees. So it makes sense that racial and ethnic preferences will have a negative effect on the graduation rates of students who supposedly "benefit" from them. Figure 5 shows the percentage of students that have graduated from UW and WSU over a six-year period.

Differences in rates of graduation parallel the differences in test scores and grades. For both UW and WSU, whites have the highest graduation rates, while blacks have the lowest. Hispanics and Asians fall in between.

The negative correlation between preference and graduation rate is confirmed by inter-university comparison, too. At UW the white- black gaps in scores are greater than those at WSU. Thus, blacks at UW graduate at a lower rate (29 percent) compared to WSU (44 percent ). Moreover, the white-black gap in graduation rates at UW is 41 percentage points; it is 17 percentage points at WSU.

Black Enrollees vs. White, Hispanic, and Asian Rejectees

The use of Affirmative Action inevitably leads to less qualified black students taking the places of more qualified white, Hispanic, and Asian students.

We computed the number of white, Hispanic, and Asian rejectees with median test scores-math and verbal SATs-equal to or better than the black enrollee median and also compared the white, Hispanic, and Asian rejectee scores to the black median score for those who were accepted but did not enroll (ANE median). Second, we computed the number of these individuals who also had GPAs equal to or greater than the black enrollee median and also compared the white, Hispanic, and Asian rejectee GPAs to the black ANE median. At UW we were able to combine these results, but at WSU we were not, so they are reported separately.

We find that 912 whites, 164 Asians, and 6 Hispanics were rejected by UW despite having equal or higher verbal scores, math scores, and GPAs than the black enrollee medians. Similarly, we find that 266 whites, 36 Asians, and 2 Hispanics were rejected by UW despite having equal or higher verbal scores, math scores, and GPAs than the black ANE medians. This means that at least 266 and as many as 912 whites, at least 36 and as many as 164 Asians, and at least 2 and as many as 6 Hispanics were rejected by UW despite their superior academic qualifications.

At WSU, 24 Asian, 8 Hispanic, and 163 white rejectees have both higher or equal verbal SATs and math SATs than the black enrollee verbal and math SAT medians. Similarly, 15 Asian, 2 Hispanic, and 77 white rejectees had higher or equal verbal and math SATs than the black ANE verbal and math medians. In addition, 7 whites rejected by WSU have GPAs higher than or equal to the median GPA for black enrollees, while 4 whites rejected by WSU have GPAs higher than or equal to the black ANE median GPA.

University of Washington

Applicants, Admittees, and Enrollees

The University of Washington provided data on 9,935 individuals who applied for admission in 1995. Of these, 7,566 were admitted and 3,444 enrolled. The overwhelming proportion of applicants, admittees, and enrollees was white.

University of Washington applicants

• 2 percent black
• 3 percent Hispanic
• 18 percent Asian
• 77 percent white

University of Washington admittees

• 3 percent black
• 4 percent Hispanic
• 18 percent Asian
• 75 percent white

University of Washington admission rates

• 97 percent of black applicants
• 90 percent of Hispanic applicants
• 78 percent of Asian applicants
• 74 percent of white applicants

University of Washington enrollees

• 3 percent black
• 4 percent Hispanic
• 25 percent Asian
• 68 percent white

Differences in Verbal SAT Scores

Figure 6 shows the group differences in verbal SAT scores. There is a white-black gap of 80 points. The score of black enrollees at the 75th percentile (460) is still 40 points lower than the white median.

Figure 6: Verbal SAT Scores, Univ. of Washington
The white-Asian gap is 60 points. The Asian median (440) is the same as the white score at the 25th percentile. This means that half the Asians enter UW with verbal scores equal to or less than all but 25 percent of white enrollees.

There are modest differences in scores between white and Hispanic enrollees. The white-Hispanic gap in median scores is 30 points, as is the gap at the 25th percentile. The white-Hispanic gap at the 75th percentile is 20 points.

Differences in Math SAT Scores

Figure 7 shows the group differences in math SAT scores. The gaps in median scores are greatest between whites and blacks (140 points). The 75th percentile score of black enrollees (540) is still 50 points lower than the white median.

The white-Hispanic gap is somewhat smaller (70 points). The Hispanic median (520) is equal to the white score at the 25th percentile. In other words, half the Hispanic enrollees have the same or a lower score compared to 75 percent of white enrollees.

The white-Asian gap is the smallest, although here, too, whites on average have higher scores. The white-Asian gap in median math scores is 20 points; the gap is 30 points at the 25th percentile, but white and Asian scores at the 75th percentile are the same.

Differences in GPAs

There are also group differences in high school GPAs (see Figure 8). Whites have the highest GPAs at all percentiles reported. Black GPAs are generally lower than those of Hispanics, Asians, and whites. The black median GPA is almost a half grade-point lower than the white median GPA. The black 75th percentile GPA is still lower than the GPAs for whites and Asians at the 50th percentile.

Hispanic GPAs are somewhat lower compared to Asians and whites. The difference in median scores is less than a quarter point.

Asian and white GPAs are nearly the same. The differences in median scores is only 0.02 points.

Rejectees vs. Enrollees

UW rejected 381 Asians, 8 blacks, 29 Hispanics, and 1,951 whites. Some whites and Asians denied admissions had superior academic credentials compared to many black enrollees.

1,650 whites, 24 Hispanics, and 346 Asians were rejected by UW with equal or higher verbal scores compared to the average black enrollee.

1,582 white, 23 Hispanic, and 264 Asian rejectees have math SAT scores equal to or higher than the black enrollee median.

Many rejectees also have higher GPAs compared to the average black enrolled at UW. 1,137 whites, 14 Hispanics, and 233 Asians rejected by UW have better grades.

Finally, 912 whites, 6 Hispanics, and 164 Asians were rejected by UW despite having equal or higher verbal scores, math scores, and GPAs compared to the black enrollee medians.

Six-Year Graduation Rates

• 29 percent of blacks
• 48 percent of Hispanics
• 65 percent of Asians
• 70 percent of whites

The graduation rate for whites is nearly two-and-a-half times the graduation rate for blacks.

Washington State University

Applicants, Admittees, and Enrollees

In 1995, 6,238 persons applied for undergraduate admission to WSU. Of these, 5,711 were admitted and 2,309 subsequently enrolled. The overwhelming proportion of applicants, admittees, and enrollees was white.

Washington State University applicants

• 3 percent black
• 4 percent Hispanic
• 8 percent Asian
• 85 percent white

Washington State University admittees

• 2 percent black
• 4 percent Hispanic
• 8 percent Asian
• 86 percent white

Washington State University admission rates

• 76 percent of black applicants
• 90 percent of Hispanic applicants
• 88 percent of Asian applicants
• 92 percent of white applicants

Washington State University enrollees

• 3 percent black
• 4 percent Hispanic
• 6 percent Asian
• 88 percent white

Differences in Verbal SAT Scores

Figure 9 shows the range of verbal SAT scores by racial and ethnic group. Whites have the highest scores at all reported percentiles. The black median score is lower than the white score at the 25th percentile, and the black verbal score at the 75th percentile is only 10 points higher than the white median.

Hispanic and Asian verbal scores also are generally lower than white scores. The median Asian and Hispanic scores are only slightly higher than the white score at the 25th percentile.

Differences in Math SAT Scores

Figure 10 shows the group differences in math SAT scores. Asian scores are slightly higher than whites at the median and the 25th percentiles, but identical at the 75th percentile.

The white-black gap is substantial. The black median is 110 points lower than the white median. The black score at the 75th percentile is still 35 points lower than the white median.

The white-Hispanic gap in median scores is smaller (25 points), with the Hispanic median closer to the white median than it is to the white 25th percentile.

Differences in GPAs

Figure 11 shows the group differences in high school GPAs. White and Asian GPAs are roughly the same, with the white median slightly higher.

There is a modest gap in scores between whites and Hispanics. The white-Hispanic gaps at the 25th, 50th , and 75th percentiles are all less than a quarter-point.

The white-black gaps are more substantial, but less than a half-point at the same percentiles. The white-black gap in median GPAs is 0.37 points. The black median is also lower than the white GPA at the 25th percentile, and the black 75th percentile score is still less than the white median.

Rejectees vs. Enrollees

Many Asian, Hispanic, and white applicants were rejected by WSU despite the superiority of their academic credentials to black enrollees. 26 Asian, 13 Hispanic, and 205 white rejectees have verbal SAT scores higher than or equal to the black enrollee median.

43 Asian, 13 Hispanic, and 227 white rejectees have math SAT scores higher than or equal to the black enrollee median.

24 Asian, 8 Hispanic, and 163 white rejectees have both higher or equal verbal SATs and math SATs than the black enrollee verbal and math SAT medians.

No Asians or Hispanics but 7 whites rejected by WSU have GPAs higher than or equal to the median GPA for black enrollees.

Six-Year Graduation Rates

• 44 percent of blacks
• 53 percent of Hispanics
• 58 percent of Asians
• 61 percent of whites

Graduation rates reflect the general differences in qualifications among racial and ethnic groups.

1. Although we would have preferred to use admittee data, which combines information on enrollees with information on those who were admitted but did not enroll, creating combined scores using UW's SAT data did not prove feasible. In the interest of simplicity, we decided to report enrollee and not admitted-but-not-enrolled scores. In the endnotes 2-4, we provide data for those who were admitted but did not enroll so that the reader may compare the two sets of results. One can assume that the combined median scores--the admittee medians--are somewhere between the enrollee median scores and the median scores of those who were admitted but not enrolled.

2. The patern of these findings did not change when considering data from those who were accepted but who did not enroll (ANE). (1) For the SAT verbal, at UW the white ANE median is 600 and the black ANE median is 540, a difference of 60, while at WSU the white ANE median id 450 and the black ANE median is 400, a difference of 50 points. (2) For the SAT math, at UW the white ANE median is 615 and the black ANE median is 530, a difference of 85 points, while at WSU the white ANE median is 520 and the black ANE median is 460, a difference of 60 points. (3) For GPAs, at UW the median for black admittees is 3.32 and for whites is 3.74, a difference of 0.42; the white ANE median GPA is 3.76 and the black ANE median is 3.40, a difference of 0.36. At WSU, the white admittee median is 3.39 and the black admitte median is 3.05, a difference of 0.34; the white ANE median GPA is 3.43 and the black ANE median is 3.05, a difference of 0.38. At WSU, where individual records were provided, the admittee median for whites was 440 on the SAT verbal and 500 on the SAT math, while the median for blacks was 390 on the SAT verbal and 360 on the SAT math.

3. The patern does not change when ANE data are considered. (1) For the SAT verbal, at UW the white ANE median is 600 and the Hispanic ANE median is 580, a difference of 20 points, while at WSU the Hispanic ANE median id 425 and the white ANE median is 450, a difference of 25 points. (2) For the SAT math, at UW the white ANE median is 615 and the Hispanic ANE median is 580, a difference of 35 points, while at WSU the white ANE median is 520 and the Hispanic ANE median is 460, a difference of 60 points. (3) For GPAs, at UW the white admittee median is 3.74 and the Hispanic admittee median is 3.58, a difference of 0.16; the white ANE median GPA is 3.76 and the Hispanic ANE median is 3.74, a difference of 0.02. At WSU, the white admittee median is 3.39 and the Hispanic admittee median is 3.17, a difference of 0.22; the white ANE median GPA is 3.43 and the Hispanic ANE median is 3.29, a difference of 0.14. At WSU, where individual records were provided, the admittee median for whites was 440 on the SAT verbal and 500 on the SAT math, while the median for Hispanics was 410 on the SAT verbal and 460 on the SAT math.

4. The patern of findings does not change when ANE data are considered. (1) For the SAT verbal, at UW the white ANE median is 600 and the Asian ANE median is 580, a difference of 20 points, while at WSU the white ANE median id 450 and the Asian ANE median is 420, a difference of 30 points. (2) For the SAT math, at UW the white ANE median is 615 and the Asian ANE median is 640, a difference of -25 points, while at WSU the white ANE median is 520 and the Asian ANE median is 515, a difference of 5 points. (3) For GPAs, at UW the white admittee median is 3.74 and the Asian admittee median is 3.70, a difference of 0.04; the white ANE median GPA is 3.76 and the Asian ANE median is also 3.76. At WSU, the white admittee median is 3.39 and the Asian admittee median is 3.33, a difference of 0.06; the white ANE median GPA is 3.43 and the Asian ANE median is 3.35, a difference of 0.08. At WSU, where individual records were provided, the admittee median for whites was 440 on the SAT verbal and 500 on the SAT math, while the median for Asians was 420 on the SAT verbal and 510 on the SAT math.