The United States Supreme Court
The University of Michigan Admissions Lawsuits

Gratz v. Bollinger

As summarized by District Judge Patrick Duggan

Plaintiff Jennifer Gratz sued Lee Bollinger, then president of the University of Michigan, alleging unlawful admissions preference based on race at the University's College of Literature, Science and the Arts (LSA). In federal district court, Dec. 13, 2000, Judge Patrick Duggan held the College's admissions policies to be unlawful in the years 1995-1998, but lawful since that time. Beginning at page 32 of his opinion, Judge Duggan gives the following description of the pertinent facts.
(The complete Opinion is online here.)

The instant action involves the LSA's admissions programs from 1995 through the present. In 1995 and 1996, admission decisions were based primarily on a set of guideline tables referred to as grids, with GPA 2 ranges represented on the vertical axis, and ACT/SAT scores represented on the horizontal axis.15 In 1995, four grids were used: (1) in-state non-minority applicants, (2) out-of-state non-minority applicants, (3) in-state minority applicants, and (4) out-of-state minority applicants. In 1996 only two grids were [33] used- (1) in-state and legacy applicants and (2) out-of-state applicants - with non-minority applicant action codes listed in the top row of the grid's cells, and minority action codes listed in the bottom row. In 1997, the same grids as in 1996 were used. However, in 1997, the LSA also added .5 to under-represented minority applicants' GPA 2 scores.

From 1998 through the present, the LSA has used a 150 point system, under which admission decisions were generally determined by the applicant's rank on the 150 point scale. Under-represented minority applicants automatically receive 20 points based upon their membership in one of the identified under-represented minority categories. In 1999 and 2000, the LSA also added a system whereby certain applicants, including under-represented minority applicants, could be "flagged," thereby keeping such applicants in the review pool for further consideration.

Beyond the fact that rigid quotas are impermissible, Justice Powell's opinion in Bakke fails to set forth any bright line regarding what constitutes a permissible consideration of race in admissions decisions. Furthermore, in situations such as this, it is often a thin line that divides the permissible from the impermissible. Applying the principles set forth by Justice Powell in Bakke, this Court is satisfied that when examined in its entirety, the LSA's current admissions program (1999-present) represents a permissible use of race. At the same time, however, the Court is satisfied that although the LSA views its current system as "chang[ing] only the mechanics, not the substance" of its prior systems, its prior systems, when examined in their entirety, cross that thin line from [34] the permissible to the impermissible. 16

A. The LSA's Current System (1999-Present)

[35] Foremost in the Court's decision that the LSA's current admissions program is constitutional is the fact that the LSA's current program does not utilize rigid quotas or seek to admit a predetermined number of minority students. Therefore, the LSA's current program does not contain the fatal flaw identified by Justice Powell in Bakke. Instead, race is taken into account in two ways under the LSA's current program. First, admissions counselors may assign each under-represented minority applicant twenty points in calculating their selection index score on account of their race. Second, under the LSA's current program, counselors may "flag" applicants that possess certain qualities or characteristics the LSA deems important to the composition of its freshman class, one of which is "under-represented race," thereby keeping an applicant who may not necessarily pass the LSA's initial admit threshold in the review pool for further consideration. 17 Such uses of race, however, operate as nothing more than the "plus" spoken of with approval by Justice Powell in Bakke.

In response, Plaintiffs contend that the LSA's practice of adding twenty points to under-represented minority applicants' selection index scores really operates as the functional equivalent of a quota. Justice Powell, however, rejected essentially the same argument in Bakke explaining: [36]

 It has been suggested that an admissions program which considers race only as one factor is simply a subtle and more sophisticated--but no less effective-- means of according racial preference than the Davis program. A facial intent to discriminate, however, is evident in petitioner's preference program and not denied in this case. No such facial infirmity exists in an admissions program where race or ethnic background is simply one element--to be weighed fairly against other elements--in the selection process. "A boundary line," as Mr. Justice Frankfurter remarked in another connection, "is none the worse for being narrow." McLeod v. Dilworth, 322 U.S. 327, 329, 64 S. Ct. 1023, 1025, 88 L. Ed. 1304 (1944). And a court would not assume that a university, professing to employ a facially nondiscriminatory admissions policy, would operate it as a cover for the functional equivalent of a quota system. In short, good faith would be presumed in the absence of a showing to the contrary in the manner permitted by our cases. See e. g., Arlington Heights v. Metropolitan Housing Dev. Corp., 429 U.S. 252, 97 S. Ct. 555, 50 L. Ed.2d 450 (1977); Washington v. Davis, 426 U.S. 229, 96 S. Ct. 2040, 48 L. Ed. 2d 597 (1976); Swain v. Alabama, 380 U.S. 202, 85 S. Ct. 824, 13 L. Ed. 2d 759 (1965).

Id. at 318-19, 98 S. Ct. at 2762-63.

Minority applicants are not insulated from review by virtue of these twenty points any more than other applicants are insulated from review by virtue of the six points awarded for geographic factors, four points awarded for alumni relationship, three points awarded for an outstanding essay, five points awarded for leadership and service skills, twenty points awarded for socioeconomic status, or twenty points awarded for athletes. In fact, the Court notes that in certain circumstances, these points may be combined for a total of up to forty. The fact that these points may "tip the balance" in favor of a particular applicant, however, does not necessarily lead to a conclusion that such applicants have been insulated from competition in the sense that Justice Powell spoke of [37] in Bakke.

The Court agrees that in certain situations, a pre-existing commitment to a fixed preference may translate into an exact proportion of the favored group being selected. This is most clearly illustrated by programs in which a preference is only available for one factor, for example race. However, under the LSA's admissions program, there are many factors that may entitle an applicant to a preference, thereby making the results of any one factor less predictable.

Plaintiffs also contend that the LSA's program, under which twenty points may be added to the selection index score of under-represented minority applicants, operates as the same type of "dual" or "two-track" system prohibited by Justice Powell in Bakke. (Pls.' 8/11/00 Br. at 15). The "two-track" system Justice Powell spoke of in Bakke, however, was not a two-track system that employed lower thresholds for minority applicants vis a vis majority applicants. Such a conclusion is evident from the fact that not once in his narrowly tailored analysis did Justice Powell ever discuss the fact that under the University of California's admissions program, majority students were subject to a 2.5 minimum GPA requirement, whereas there was no minimum GPA requirement for minority students. See Id. 274-75, 98 S. Ct. at 2739-40.

Instead, the "two-track" system of which Justice Powell spoke was the university's system under which one group of students, i.e. minority students, competed for one set of seats, and another group of students, i.e. majority students, competed for another set [38] of seats, thereby in effect creating two separate admissions systems. Justice Powell's statement that the university's interest in genuine diversity would not be served by "expanding [the university's] two-track system into a multitrack program with a prescribed number of seats set aside for each identifiable category of applicants" further confirms that the two-track system Justice Powell spoke of related to the university's practice of setting seats aside for minority applicants, not the university's practice of employing different GPA requirements for minority vis a vis majority applicants. Id. at 314, 98 S. Ct. 2761.

Furthermore, unlike the University of California, the LSA does not utilize a separate admissions review committee for under-represented minority applications. Instead, admission counselors are assigned to specific geographic areas. Each counselor reviews all applications from his or her territory, under-represented minority and non-minority alike. There is no separate review or assignment of under-represented minority applicants as there was in Bakke. (Spencer Dep. at 273; Gauss Dep. at 31, Vanhecke Dep. at 99).

Moreover, whenever a point-based system such as the LSA's is used in making admissions decisions, any factor that may result in points being added to an applicant's total score will necessarily have the effect of lowering that applicant's admission threshold. For example, any time an applicant receives three points for an outstanding essay, it can just as easily be said that the threshold for admission for that applicant has now been reduced by three points. The same is true for each of the other factors used during the admissions process, e.g., geographic location, alumni relationship, socioeconomic status, [39] etc. Such a result is the natural byproduct of using such a point system.

What Plaintiffs really appear to contest is the fact that race is accorded twenty points, while other factors that may more consistently favor non-minority students are not typically accorded the same weight. However, as Justice Powell recognized in Bakke, universities may accord an applicant's race some weight in the admissions process and, in doing so, universities are not required to accord the same weight to race as they do other factors. Bakke, 438 U.S. at 317-18, 98 S. Ct. at 2762. As long as the admissions program does not work to isolate the applicants from review, it withstands constitutional muster, despite the fact that it may provide individuals with a "plus" on account of their race.

Plaintiffs also attack the LSA's "flagging" system. According to Plaintiffs, the files of under-represented minorities remain "protected" insofar as the LSA's flagging procedure ensures that the files of such students remain in the "review pool" because of their race.

Under the LSA's current admissions system, a counselor may "flag" an applicant if (1) in the counselor's estimation, the applicant is academically prepared for the University, (2) has a selection index score of at least 80 for residents, or 75 for non-residents, and (3) possesses a quality or characteristic the LSA deems important to the composition of its incoming class, including under-represented minority status. Counselors, however, are not required to flag every under-represented minority applicant. Furthermore, applicants other than under-represented minority applicants may also be [40] flagged. For example, other factors a counselor may consider in flagging applications include whether the applicant was in the top of his class, resided in a preferred county of Michigan, exhibited any "unique life experiences, challenges, circumstances, interests or talents," exhibited a disadvantaged background, had an important connection to the University community, or was a recruited athlete. (Defs.' Exs., Ex. Z at 3-4). Given that any number of applicants, including applicants other than under-represented minority applicants, may be "flagged" under the LSA's current system, the Court rejects Plaintiffs' contention that under-represented minorities remain "protected" by virtue of such system.

Plaintiffs also contend that the University Defendants have failed to sustain their burden of showing that it considered race-neutral alternatives to its current program. For example, according to Plaintiffs, one obvious race-neutral mechanism would be to randomly select all or a portion of the class from the entire pool of 'qualified' applicants, regardless of race. Plaintiffs, however, fail to explain how randomly selecting all or a portion of the class from the entire pool of applicants would produce a sufficiently diverse student body. As discussed supra, to achieve the educational benefits associated with a racially and ethnically diverse student body, more than a token number of under-represented minority students is required. Given the small size of the applicant pool, it is highly unlikely that a random selection process would result in a sufficiently diverse student body. Furthermore, the University Defendants have presented evidence that a race-neutral [41] admission program would substantially reduce the number of under-represented minority students in the LSA's incoming student body. (See Raudenbush 2/24/00 Supp. Rep. at 4-5; Raudenbush 3/3/99 Supp. Rep. at 9-11). If race were not taken into account, the probability of acceptance for minority applicants would be cut dramatically, while non-minority students would see only a very small positive effect on their probability of admission, due largely in part to the size of their respective applicant pools. (See Raudenbush 3/3/00 Supp. Rep. at 11).

The University Defendants have also presented evidence that a system that relied entirely on test scores would also lead to the rejection of a number of qualified minority applicants. (Bowen Rep. at 10). This is due to the fact that nationally, minorities are very under-represented at the higher level of standardized test scores, and over-represented at the lower level. (Id. at 10).

The University Defendants have also presented the expert opinion of William G. Bowen that the race-neutral admissions program currently used by the University of Texas, under which all students who finish in the top ten percent of their high school class are guaranteed admission, and any system based purely on an applicant's income level, would not be as effective in enrolling "an academically well prepared and diverse student body." (Bowen Rep. at 12). According to Bowen, the Texas approach would have the effect of "admit[ting] some students from weaker high schools while turning down better-prepared applicants who happen not to finish in the top tenth of their class in academically stronger [42] schools." (Id.). Bowen further hypothesizes that "[s]o long as high schools differ so substantially in the academic abilities of their students and the level of difficulty of their courses, treating all applicants alike if they finished above a given high school class rank provides a spurious form of equality that is likely to damage the academic profile of the overall class of students admitted to selective institutions." (Id.).

According to Bowen, income-based strategies are just as ineffective as there "are simply too few blacks and Latinos from poor families who have strong enough academic records to qualify for admission to highly selective institutions." (Id. at 13). Bowen also reports that if universities were totally eliminated from considering race during the admissions process, "over half of the black students in selective colleges today would have been rejected." (Id.).

Furthermore, the University has attempted to enlarge its pool of under-represented minority applicants through vigorous minority recruitment programs, which have all proved to be unavailing. For example, the University's efforts have included personal contact with minority students at symposia, attendance at recruiting fairs, direct mailings, campus visits, and offices in Detroit. (Vanhecke Dep. at 11-12; Spencer Dep. at 29, 196-97). Nevertheless, according to the University, its pool of qualified minority applicants for the LSA has remained small, most of whom are also highly recruited by other selective institutions and universities. (Defs.' 7/17/00 Br. at 27).

The Court also rejects Plaintiffs' contention that the LSA's current system amounts [43] to nothing more than racial balancing. Unlike the cases relied upon by Plaintiffs, the LSA does not seek to achieve a certain proportion of minority students, let alone a proportion that represents the community. To that end, the Court agrees with the Fourth Circuit in Tuttle v. Arlington County Sch. Bd., 195 F.3d 687, 707 (4th Cir. 1999), that an educational institution's goal of achieving the educational benefits associated with a diverse student body "do[es] not require racial balancing." However, the University's interest does require a sufficiently diverse student body and therefore, requires that some attention be given to the composition of the student body. Although fixed racial quotas and racial balancing are not necessary to achieving that goal, the consideration of an applicant's race during the admissions process necessarily is.

In summary, the Court is satisfied that the LSA's current admissions program, under which certain minority applicants receive a "plus" on account of their race but are not insulated from all competition with other applicants, meets the requirements set forth by Justice Powell in Bakke and is therefore constitutional. Accordingly, the University Defendants' motion for summary judgment shall be granted with respect to the LSA's current admissions program, and Plaintiffs' motion for injunctive relief shall be denied.

B. The LSA's Prior Programs (1995-1998)

Although the Court finds the LSA's current admissions program to be permissible under the principles set forth by Justice Powell in Bakke, the Court is similarly satisfied that the LSA's prior admissions programs, when examined in their entirety, represent an [44] impermissible use of race. One of the most significant factors that transforms the permissible into the impermissible in this case is the LSA's prior practice of "protecting" or "reserving" seats for under-represented minority applicants. As explained by the LSA itself:

 Because the class is selected on a rolling basis, rather than at one point in time, a certain number of seats is designated during the admissions cycle for in-state students and for certain other groups of students, including, for example, athletes, foreign applicants, underrepresented minority candidates, and ROTC candidates (sometimes referred to as "protected" space). This space is "protected" to enable OUA to achieve the enrollment targets of the University and of the individual units while using a rolling admissions system. If this space is not filled by qualified candidates from the designated groups toward the end of the season, it used to admit students from the postponed pool or the extended waiting list, applicants to other units, etc.

(Defs.' Answer Interrog. No. 1) (emphasis added); (See also Pls.' 4/9/99 Exs., Vol. I, Exs. P, Q, & R) (showing number of protected seats for particular academic years). The LSA has also explained this practice in the following manner:

 LSA admissions occur on a "rolling" basis: rather than waiting for one date in late winter to notify all applicants of the disposition of their applications, LSA admits applicants throughout the admissions season. Under such a process, offers of admissions must be carefully monitored and managed to ensure that sufficient spaces are reserved, or protected, for attractive applicants who apply later in the cycle. The number of protected spaces is determined by the expected pool size of various groups of applications. As applicants from a particular group are admitted over the course of the admissions season, the protected spaces reserved for that group are used. If the pool of qualified applicants never reaches the number of protected spaces, those slots are filled with qualified applicants off the wait list.

(Pls.' 4/9/99 Br. at 7 n.4) (quoting Defendants' Motion for Reassignment or Designating [45]Actions as Companion Cases in Grutter v. Bollinger, 97-CV-5928-DT).

Other memoranda refer to these "protected" seats as being "reserved" for particular groups of applicants. (See Pls.' 4/9/99 Exs., Vol. I, Ex. M). In fact, one memorandum specifically states that the number of "protected groups" for Fall 1997 would be decreased, thereby "opening up slots for non-protected applicants." (Id., Ex. O). This evidence clearly indicates that these slots were not merely protected from the admissions process itself, but from competition by non-protected applicants. It is clear that the LSA's system operated as the functional equivalent of a quota and therefore, ran afoul of Justice Powell's opinion in Bakke.

In this Court's opinion, there is no significant difference between the LSA's prior practice of "protecting" or "reserving" seats and the University of California's quota system. The fact that non-minority applicants may have had a chance at any "leftover" spots at the end of the admissions cycle does not change this conclusion. Under both systems, preferred minority applicants were insulated from competition from non-preferred applicants for a given number of seats. The fact that this number may have changed from year to year, or even during any one particular admissions cycle, does not change this fact. It can hardly be said that those non-preferred applicants who, by pure chance, may have had an opportunity to compete for one of the leftover seats, were allowed to compete on equal footing for every seat in the class.

Furthermore, it is undisputed that from 1995 through 1997, the LSA used facially [46] different grids and action codes based solely upon an applicant's race. Under these differing grids, a certain group of non-preferred applicants were automatically excluded from competing for a seat in the class without any type of individualized counselor review solely on account of their race, 18 whereas, preferred minority applicants were never automatically rejected, regardless of their grades and test scores. Rather, all minority applicants received some type of individualized counselor review. (McKinney Dep. at 41, Spencer Dep. at 105). This practice of "automatic exclusion" continued into the 1998 academic year despite the LSA's switch to the selection index system. (McKinney Dep. at 97-98, 150-52).

The Court agrees with the University Defendants' assertion that all who are ultimately admitted to the LSA are "qualified" academically, and neither Plaintiffs nor this Court seek to imply that those minorities who are admitted under the lower admissions standards are not academically qualified for admission. It cannot be seriously disputed, however, that the effect of the LSA's differing standards was to systematically exclude a certain group of non-minority applicants from participating in the admissions process based solely on account of their race.

[47] For example, in 1995 and 1996, the LSA used two grids for instate applicants, 19 one for "non-minority" applicants and another for "minority" applicants. The non-minority grid indicated that an applicant with a GPA 2 of 3.2-3.3 and ACT of 18-20/SAT of 400-500 would be automatically rejected, whereas a minority applicant with the same grade/score would have most likely been admitted. 20

In so doing, the LSA's policies foreclosed each of the automatically rejected applicants from further consideration for a seat simply because they were not of a favored race or ethnicity. These applicants were never accorded any further individualized review as envisioned by Justice Powell in Bakke. In this Court's opinion, the LSA's different treatment of preferred minority applicants vis a vis non-preferred applicants at this stage of the admissions process further adds to the infirmity of the LSA's prior programs.

Moreover, the only distinguishing factor between the grids used by the LSA during these years was the applicant's race. Therefore, to the extent admissions counselors used such grids in making admissions decisions, it is clear from the face of the grids themselves [48] that in some cases, the only defining factor was race. Although the LSA's use of facially different grids/action codes based upon an applicants' race, in and of itself, may not have been constitutionally impermissible, when combined with the other components previously discussed by the Court, i.e. the LSA's use of protected seats and the LSA's system of automatic rejection, the Court is convinced that the LSA's prior programs, when examined in their entirety, fall within the impermissible under the principles enunciated by Justice Powell in Bakke. Accordingly, Plaintiffs' motion for summary judgment shall be granted with respect to the LSA's admissions programs employed from 1995 through 1998.


15. An applicant's GPA 2 was calculated by adjusting the applicant's high school GPA based upon several factors, including the quality of the applicant's high school, the strength of the applicant's high school curriculum, any unusual circumstances, the applicant's geographical residence, and the applicant's alumni relations, if any. [return to text]

16. It is understandable that the LSA does not contend that there is a "substantive" difference between its current and prior admissions programs. To acknowledge a substantive difference in such programs would give some support to Plaintiffs' claims that the admissions programs in effect from 1995 through 1998 were not "narrowly tailored." Furthermore, the University Defendants' contention that "only the current system can form the basis for [injunctive relief]," and their arguments in support of this contention, (Defs.' 8/11/00 Br. at 92), suggest that there are some significant differences between the LSA's current admissions programs and those used from 1995 through 1998. [return to text]

17. Other qualities or characteristics include high school class rank, unique life experiences, challenges, circumstances, interests or talents, socioeconomic disadvantage, and geography. [return to text]

18. The Court notes that the LSA's practive of automatically rejecting applicants also ended with the 1999 academic year. [return to text]

19. The Court notes that another set of grids was used for out-of-state applicants, for a total of four grids. [return to text]

20. The minority grid also contained the option of delaying such a minority applicant "for senior year SAT's or ACT's." [return to text]

Copyright by the Regents of the University of Michigan


Return to