To most observers, it seemed that Craig v. Boren had written the final chapter in the saga of gender-based equal protection: claims of constitutional sex discrimination were to be judged under an intermediate scrutiny, or “middle tier,” standard that fell somewhere between the strict scrutiny and rational basis tests. But in 1996–20 years after Craig-a Supreme Court case involving the all-male Virginia Military Institute (VMI) strongly suggested that the last word on the question had not yet been spoken.
VMI was a state-supported military college with a long and illustrious history. Founded in 1839, VMI prided itself on what it described as its unique “adversative” method of education, which emphasized physical hardship, mental stress, lack of privacy, and exacting regulation of behavior. First-year cadets, called “rats,” had to endure seven months of harsh and demeaning treatment by upperclassmen in a boot-camp atmosphere (the “rat line”). The goal of the system was to mold character and produce leaders, and indeed many of VMI’s alumni, bonded through adversity, had gone on to prominent positions in both military and civilian life.
Of the fifteen public institutions of higher education in Virginia, VMI was the only one that was still limited to one sex. Between 1988 and 1990, VMI received inquiries concerning admission from 347 young women. It responded to none of them. In 1990, a female high school student from Northern Virginia filed a complaint with the attorney general of the United States, alleging that VMI’s male-only admission policy was a violation of the Constitution’s guarantee of equal protection. Finding merit in her complaint, the Department of Justice filed suit in federal district court under the Civil Rights Act of 1964. (Adhering to case law developed under that statute, the Justice Department has kept the identity of the high school student a closely guarded secret, to protect her from retaliation.)
After a six-day trial that included expert testimony on both sides, the district court decided in favor of VMI in June 1991. The district court found the appropriate standard of review in Mississippi University for Women v. Hogan, a 1982 Supreme Court case involving a state-supported nursing school that refused to admit men. “Our decisions . . . establish,” the Court had said in Hogan, “that the party seeking to uphold a statute that classifies individuals on the basis of their gender must carry the burden of showing an ‘exceedingly persuasive justification’ for the classification . . . The burden is met only by showing at least that the classification serves ‘important governmental objectives and that the discriminatory means employed’ are ‘substantially related to the achievement of those objectives.'”
The “important governmental objective” here, the district court held, was to promote diversity in higher education-specifically, to provide the choice of a single-gender educational environment. The evidence showed that such an environment could benefit many students, male or female: students at single-sex colleges of both sexes were more academically involved; interacted more frequently with faculty; and were more likely to be successful in later life, though this evidence was stronger for all-female schools than for all-male schools. And, clearly, there was a substantial relationship between the state’s objective and the means employed to achieve it: the exclusion of women from VMI. Excluding one sex from a school was, in fact, not only substantially related to achieving the goal of providing a single-gender environment-it was the only way to achieve it.
The district court recognized that the state was offering young men a choice that was not available to young women. But, as the district court saw it, the dilemma was that if VMI were required to open its doors to women, the very experience they sought there would no longer exist. “[T]he evidence establishes,” the district court ruled, “that key elements of the adversative VMI educational system, with its focus on barracks life, would be fundamentally altered, and the distinctive ends of the system would be thwarted, if VMI were forced to admit females and to make changes necessary to accommodate their needs and interests.” While some women might in fact thrive in the VMI environment, the vast majority would not-“and educational systems are not designed for the exception but for the mean.” Perhaps the real problem was not that VMI was all male, the district court suggested, but that Virginia did not maintain even one all-female institution of higher education. That issue, however, was not before the court.
When the case reached the court of appeals the following year, that very issue moved to the forefront. The court of appeals agreed with the district court that providing single-sex education was a worthy goal, and that VMI’s system would be substantially changed by the admission of women. But if the state was going to provide a diversity of educational options for men, it needed to figure out some way to provide it for women as well. Therefore, the court ruled that Virginia had violated the Equal Protection Clause by offering VMI’s method of citizen-soldier education to men only. But rather than dictating a specific remedial course of action, the court of appeals suggested three options: the state might choose to admit women to VMI, it might establish a parallel institution for women, or it might turn VMI into a private college. This last option would free the school from the constraints of the Equal Protection Clause, but also deprive it of state funding.
In response, Virginia convened a task force charged with the responsibility of designing a single-sex program for women that would, like VMI, seek to produce “citizen-soldiers.” The result was a plan for the Virginia Women’s Institute for Leadership (VWIL), a state-sponsored undergraduate program to be located at Mary Baldwin College, a private liberal arts school for women. The task force, composed of experts in women’s education, concluded that VMI’s “adversative” model would be inappropriate for most women. VWIL would rely instead on a cooperative method aimed at building up self-esteem rather than destroying it. There would be no rat line, no rigorous barracks life, and no uniforms. Instead, VWIL students would have training in self-defense and self-assertiveness, take courses in leadership, and receive military training through a preexisting ROTC program.
The next step was to return to court to secure approval of the program. In the district court, the Justice Department challenged the VWIL plan as inadequate, pointing out the substantial differences between VWIL and VMI. The average combined SAT score for entrants at Mary Baldwin was about 100 points lower than at VMI, its faculty held significantly fewer Ph.D.s and received lower salaries, and the women’s college did not offer degrees in the sciences and engineering as did VMI. The recreation facilities at Mary Baldwin were minimal (two fields and a gym) compared to VMI’s extensive competition-level fields tailored to different sports and its vast indoor facilities. Nor did Mary Baldwin have the history, reputation, and influential alumni network enjoyed by VMI, which included military generals, members of Congress, and business executives. VMI’s distinguished alumni not only eagerly hired VMI graduates, but supported an endowment of $131 million compared to Mary Baldwin’s $19 million. More fundamentally, VWIL, with its “cooperative” program, was an inadequate substitute for those women who wanted to attend VMI precisely in order to experience its “adversative” educational system.
The district court rejected these arguments. Establishing a new institution that was truly equal to VMI would be an impossible task, the court said, and it would be unrealistic to think that the court of appeals intended to impose such a requirement. The VWIL plan, the court held, “takes into account the differences and the needs of college-age men and women,” based on expert recommendations. “If VMI marches to the beat of a drum,” the court concluded metaphorically, “then Mary Baldwin marches to the melody of a fife and when the march is over, both will have arrived at the same destination.”
The three-judge Court of Appeals affirmed this district court judgment, with one judge dissenting. In the process, the court sharply lowered the Craig v. Boren standard, applying what it called a “special intermediate scrutiny test for classifications based on homogeneity of gender in the context of higher education.” Turning to the first prong of the Craig test-the “important” government objective-the court reasoned that it should take a cautious and deferential approach, so as to avoid simply substituting its own values for those of the government. The government objective itself, the court ruled, need only be legitimate as in the old “anything goes” rational basis test, not “important” as Craig decreed. As long as the state’s purpose was not “pernicious,” the court would defer to the state. The focus of the analysis should instead be entirely on whether the means chosen by the state “substantially and directly”
Here, in the lower court’s view, the government’s objective-providing the option of single-sex education-was legitimate and certainly not pernicious. And the means chosen to achieve that goal-the exclusion of women from VMI and men from VWIL-was not only substantially related, but absolutely essential. The exclusion of women from VMI was also “directly related to achieving the results of an adversative method in a military environment . . .” The court also had other concerns. “If we were to place men and women into the adversative relationship inherent in the VMI program, we would destroy . . . any sense of decency that still permeates the relationship between the sexes.”
But the court also recognized that, in these circumstances, employing the two-pronged equal protection test risks “bypass[ing] any equal protection scrutiny”: if the government’s objective is to provide a single-gender environment, the exclusion of the other gender will always be a “substantially related,” indeed necessary, method of achieving it. To ensure that the challenged gender classification did not entirely escape equal protection scrutiny, the court went on to add a third prong: the alternatives offered to men and women must be “substantively comparable.” Applying the test to the facts of this case, the court concluded that the separate VMI and VWIL programs, while clearly not identical, were sufficiently comparable to pass muster.
When the Justice Department and Virginia sought review in the Supreme Court in 1995, it triggered an avalanche of amicus briefs from women’s rights organizations denouncing the Court of Appeals’ second decision. Not only was the newly devised test less protective than the established intermediate scrutiny standard, it was even more lax than the “separate but equal” doctrine discredited by the Court more than forty years before in the context of school desegregation. Both the Justice Department and the amicus briefs seized on a phrase the Court had used in a case called Mississippi University for Women v. Hogan to describe the intermediate scrutiny test: the government must show an “exceedingly persuasive justification” for the gender classification at issue (see p. 55).
But women’s rights groups, and later the Justice Department, did not just call for a reaffirmation of the Craig v. Boren standard. Drawing on criticism of the intermediate scrutiny test that had been brewing for some time, they invited the Court to announce that gender discrimination, like race discrimination, would henceforth be subject to the strict scrutiny test. The briefs revived the arguments for strict scrutiny that had been offered by the plurality in Frontiero v. Richardson: sex, like race, was an immutable characteristic that bore no relation to ability, and women had been the victims of a long and unfortunate history of sex discrimination. Moreover, the intermediate scrutiny standard had simply proved too vague to be workable. The lower courts were rife with instances of its misapplication-with the case at hand offered as a prime example.
The Supreme Court had adhered to the intermediate scrutiny test in gender discrimination cases that had come before it since Craig v. Boren. But the Justice Department and others who argued for strict scrutiny pointed to footnotes in two Supreme Court decisions-one in the aforementioned Hogan, and another in a 1994 case called J.E.B. v. Alabama ex rel. T.B.-where the Court indicated that strict scrutiny for gender classifications was still a live possibility: it was not necessary to reach the question of strict scrutiny, the Court had said, because the challenged classifications could not even pass intermediate scrutiny. And several commentators concluded that, in the cases since Craig, the Supreme Court had been subtly ratcheting up the level of scrutiny for gender-based classifications-primarily through the phrase “exceedingly persuasive justification.”
Some commentators also argued that the case for applying strict scrutiny to gender classifications had been strengthened by the Supreme Court’s recent affirmative action decisions permitting preference in hiring and college admissions to groups that had previously been subject to discrimination. A traditional justification for heightened scrutiny under the Equal Protection Clause was that “discrete and insular” minorities, who could not defend their rights in the political arena, were in need of special protection from the courts. Since women are not a minority, the argument had gone-comprising, in fact, 53 percent of the population-they were not in need of the kind of heightened scrutiny reserved for racial and ethnic minorities. But in cases decided in 1989 and 1995, the Court had ruled that governmental affirmative action programs giving preferences to minorities were subject to strict scrutiny. While the Court’s application of the strict scrutiny standard in this area was less stringent than usual, it was still exacting. And it was now being applied to discrimination against white males-clearly no one’s idea of a discrete and insular minority.
As a practical matter, it would be difficult to apply the new strict scrutiny to race-based affirmative action programs without also applying it to gender-based programs, because most affirmative action programs used both categories. And retaining the two different standards would appear inequitable. Affirmative action plans for African Americans could be struck down under the stricter race standard while those for women could be upheld under the more lenient gender standard. All these factors gave the Court strong reasons for increasing the gender standard to strict scrutiny.
A number of feminists and some private women’s colleges argued strongly against application of strict scrutiny. Such a ruling might threaten the existence of private single-sex schools, because it could jeopardize federal financial aid and tax breaks. More broadly, they said, strict scrutiny would bar the government from taking into account the legitimate differences between men and women, and-by requiring equal treatment across the board-force women into a male mold.
The advocates of strict scrutiny appeared to suffer a setback at oral argument when Deputy Solicitor General Paul Bender, arguing for the Justice Department, came under sharp questioning from Justice Sandra Day O’Connor (the author of the majority opinion in Hogan). “Couldn’t this case be decided under the intermediate scrutiny standard?” she asked Bender. “Yes, absolutely,” he replied. “Why then,” O’Connor wanted to know, “had the government chosen this case to argue the issue of strict scrutiny?” Bender started to answer that on several occasions the Court had indicated that the standard of scrutiny was still an open question. “Well, it’s not exactly an open question,” O’Connor shot back, “in the sense that the Court has decided a number of cases . . . applying . . . intermediate scrutiny. . . If you look at Mississippi University for Women v. Hogan, the Court certainly tried to articulate a standard.”
“[In] Craig v. Boren [we] said that was the standard,” Chief Justice William H. Rehnquist interjected.
Justice Ruth Bader Ginsburg, who as a litigator had been the prime champion of strict scrutiny for gender classifications, remained silent during this exchange.
But it was Ginsburg who announced the opinion from the bench on June 26, 1996. Writing for a six-person majority, which included Justice O’Connor, Ginsburg ruled that the lower courts had erred in holding that VMI did not have to admit women. She did not squarely address the question of strict scrutiny, but she did leave the door open for further discussion. Ginsburg noted that the Court had not equated gender classifications, “for all purposes,” to classifications based on race or national origin-the classifications for which strict scrutiny had “thus far” been reserved. She also applied the gender standard with the same rigor traditionally seen in strict scrutiny cases, as dissenting justice Antonin Scalia pointed out in great detail.
But while Ginsburg relied on the Supreme Court’s prior intermediate scrutiny cases, the phrase that leapt to the forefront of her opinion was “exceedingly persuasive justification.” The “core instruction” of the Court’s previous cases, she wrote in beginning her analysis, was that “[p]arties who seek to defend gender-based government action must demonstrate an ‘exceedingly persuasive justification’ for that action.” And the justification for the government action “must be genuine, not hypothesized or invented post hoc [afterward] in response to litigation.”
Applying that standard, Ginsburg first viewed with skepticism Virginia’s stated objective of providing educational diversity. No such purpose was evident at the time of VMI’s founding in 1839, when the state excluded women from all higher education because “[h]igher education was considered dangerous for women,” and subsequent history was equally unconvincing. Providing a variety of educational choices was not the state’s true purpose, Ginsburg concluded, but was invented later. The state’s next argument was that admitting women would undermine VMI’s adversative system. It was undisputed, however, that at least some women would be capable of engaging in VMI’s demanding program and would actually prefer it to a methodology such as VWIL’s. And the question was not whether VMI would be suitable for most women; it was probably not suitable for most men. Rather, “the question is whether the commonwealth can constitutionally deny to women who have the will and capacity, the training and attendant opportunities that VMI uniquely affords.”
The fear that the admission of women would destroy the adversative system and the school itself was “a judgment hardly proved, a prediction hardly different from other ‘self-fulfilling prophecies’ . . . once routinely used to deny rights or opportunities.” Ginsburg offered some telling examples: an 1876 state court that ruled women could be prevented from being lawyers in order to “grade up” the profession; medical faculties that barred women from their schools for fear of such evils as women and men jointly displaying “the secrets of the reproduction system”; and police resistance to women on the force because their presence would “undermine male solidarity,” deprive the men of “adequate assistance,” and “lead to sexual misconduct.” She pointed out how successfully women had performed once the federal military academies and services were opened to them. In short, she concluded, Virginia had “‘fallen far short of establishing the “exceedingly persuasive justification”‘. . . that must be the solid base for any gender-defined classification.”
Turning to the adequacy of the state’s remedy-the establishment of VWIL-Ginsburg found it sadly lacking. Virginia’s claim that the dissimilarity between the two schools was justified by “real” differences between the sexes, and not by stereotypes, once again ignored the fact that some exceptional women would prefer the VMI approach. The VWIL program was hardly a match for VMI in many ways: faculty, curriculum, military program, student body, facilities, endowment, and prestige. In applying its newly devised, three-pronged test for examining the state’s purpose, the Court of Appeals had engaged in circular reasoning and had gravely erred. “Women seeking and fit for a VMI-quality education,” Ginsburg concluded, “cannot be offered anything less, under the State’s obligation to afford them genuinely equal protection.”
Chief Justice Rehnquist and Justice Scalia filed their own opinions, one concurring in the judgment and one dissenting. (Justice Clarence Thomas did not participate in deciding the case because his son, Jamal, was attending VMI.) The chief justice, while agreeing with the Court’s conclusions, took issue with its analysis. Having dissented vigorously in Craig v. Boren two decades before, Rehnquist now found himself in the position of defending the precedent. The majority’s reliance on the phrase “exceedingly persuasive justification” had introduced “an element of uncertainty respecting the appropriate test.” The terms of the Craigem> test-“important governmental objective” and “substantially related”-were “hardly models of precision,” but they had more content than the phrase on which the Court had now seized.
Rehnquist agreed that the state’s proffered justification of educational diversity was unconvincing. But-unlike the majority-he based that conclusion only on the state’s actions since the Supreme Court’s decision in Hogan, which he said “placed Virginia on notice that VMI’s admissions policy possibly was unconstitutional.” Even accepting the state’s justification, Virginia’s argument was flawed because it had failed to provide educational diversity for women as well as men. VMI would not necessarily have to admit women to satisfy the demands of the Constitution, but the state would at least have to create two single-sex institutions that “offered the same quality of education and were of the same overall calibre.”
Scalia filed a scathing forty-page dissent. He had no problem, he said, with the Court’s three-tiered system of equal protection standards, although he criticized the Court for applying them “whenever we feel like it.” But those tests could not “supersede-and indeed ought to be crafted so as to reflect-those constant and unbroken national traditions that embody the people’s understanding of ambiguous constitutional texts.” One of those traditions was the all-male, government-supported military college.
But leaving aside the question of tradition, Scalia continued, the Court had not even honestly applied its own well-established intermediate scrutiny test. Like Rehnquist, Scalia viewed the majority’s reliance on the phrase “exceedingly persuasive justification” as a departure from past practice. The Justice Department had argued for strict scrutiny, and the Court, “while making no reference to the Government’s argument, effectively accepts it” (a point Ginsburg’s opinion never denied).
Under the intermediate scrutiny standard, Scalia maintained, the question was not whether some women (or perhaps only one) were interested in attending VMI and capable of participating in its program. Ginsburg’s approach required a “perfect fit” between ends and means; in Scalia’s view, a system that was unfair to a few nonconformists would nevertheless satisfy the “substantial relationship” requirement of the intermediate scrutiny standard. For example, in Rostker v. Goldberg, a 1981 case, the Court had ruled that selective service registration could constitutionally exclude women, even if some women were fit for noncombat roles, because the purpose of registration was to supply a pool of combat troops and Congress had passed laws barring all women from combat. “There is simply no support in our cases,” Scalia said, “for the notion that a sex-based classification is invalid unless it relates to characteristics that hold true in every instance.” In fact, given the choice, he thought the “stronger argument” was to reduce the gender standard to the old rational basis review.
Scalia then identified Virginia’s “important” governmental interest as “providing effective college education for its citizens.” Single-sex instruction was a means substantially related to that interest, because of its proven benefits. While the adversative method was not appropriate for everyone, providing the choice of such a method was also substantially related to the government’s objective. And it had been found as a fact by the district court that if VMI were required to admit women, it “would eventually find it necessary to drop the adversative system altogether.” While the best possible approach would be for the state to provide both adversative schools and traditional ones, in all three possible variations-all male, all female, and coed-Virginia did not have enough money to provide six schools. Given these realities, Virginia’s exclusion of women from VMI was “substantially related to the Commonwealth’s important educational interests.”
Scalia concluded his dissent elegiacally, with a lengthy quotation from a booklet distributed to all first-year VMI students. Listing the characteristics of “a gentleman,” the booklet admonished, among other things, that a gentleman must never discuss “the merits or demerits of a lady,” “hail a lady from a club window,” or “so much as lay a finger on a lady.” It was “powerfully impressive,” Scalia said, that “a public institution of higher education still in existence” sought to hold its students to such standards, and he doubted that “any of us, women included, will be better off for its destruction.”
Outside the Court, some commentators-while agreeing with Rehnquist and Scalia that the VMI opinion signaled a departure from previous gender-discrimination precedents-saw this as a welcome development. “[T]he VMI decision may well be the final stepping stone on the path to a destination envisioned by feminist advocates, including Justice Ginsburg, decades ago,” a law review author exulted. “That destination is strict scrutiny for gender classifications.”
At VMI itself, the mood was grim. “This is everything that is good in our culture,” said Superintendent Josiah Bunting III while the case was pending in the Supreme Court, “and it’s going to change irretrievably if they bring in women.” After the ruling, VMI’s Board of Visitors considered buying the school from the state in order to maintain its all-male status. But ultimately, in September 1996, the board voted 9
The VMI administration made it clear that women would be held to the same standards as men, and did its best to prepare for their arrival: the school hired female staff, required all students and employees to attend seminars on sexual harassment and fraternization, and installed women’s bathrooms and new shades on the barracks windows. In the fall 1997 thirty women enrolled as first-year cadets; seven months later, twenty-three of them were still around to perform the traditional crawl up a muddy twenty-foot hill that marks the end of the rat line. Five of those women made the rank of corporal for their sophomore year, and in May 1999 two women who had transferred to VMI in 1997 became the school’s first female graduates.
The transition to coeducation has not been entirely smooth. Some alumni have complained that standards have been diluted, and feminist groups have remained skeptical of VMI’s commitment to equality. Several students have been disciplined or expelled for sexual misconduct-including the school’s top-ranking cadet, who was thrown out for allegedly using his position to pressure freshman women into having sex. Despite these hurdles, it is clear that coeducation at VMI is there to stay, and even appears to be a success. By August 1999, when the third crop of women cadets took their places on VMI’s rat line, Superintendent Bunting told the Washington Post that standards had actually toughened since women had arrived on campus.
“What we are beginning to see is a certain type of young woman who is attracted to VMI,” he said. “They are young women who have something to prove to themselves and to their friends. They want to take the road less traveled and flourish in a system that measures success in many different ways, not just in academics.”