The Precedent That Wasn't:
College Hate Speech Codes

Jon B. Gould
U. Mass. Law and Society Review (2001)

Copyright (c) 2001 Law and Society Association, University of Massachusetts 
Law and Society Review


35 Law & Soc'y Rev. 345

LENGTH: 22648 words

ARTICLE: The Precedent That Wasn't: College Hate Speech Codes and the Two Faces of Legal Compliance

Jon B. Gould

...   On March 30, 1995, newspaper headlines declared that hate speech regulations were dead. ... That each of these five institutions rescinded its offending speech policy between 1992 and 1997 does not mean that they all made a compliance decision. ... A high-level official went so far as to say, "We don't have a hate speech policy." ... To comprehend the retention of Plains' hate speech policy, one must also understand its development. ... That he did not remove the policy, however, reflects a campus consensus that the hate speech code has symbolic value. ... Indeed, one of the policy's prime proponents, a women's studies professor, says she is "surprised" that the policy would be labeled a hate speech policy. ... The case strengthened the view of the university's attorneys that Agrarian's hate speech policy was unconstitutional. ... Although neither official was responsible for the ultimate decision to settle, they both recognize the pressures now on them when administering the college's hate speech policy. ... No other administrator with whom I spoke would call the rule a "hate speech code," labeling it, instead, a "policy of inclusion" or "diversity," but the dean (who has since retired) allows that it is "substantially similar" to hate speech policies at other institutions. ...  

This article undertakes an empirical investigation into the development and persistence of college hate speech codes, asking why so many elite institutions of higher learning either retained or created speech policies that contradicted a national series of court cases. The method is both quantitative and qualitative, looking for broad patterns of response among the schools and also explaining why individual institutions did or did not comply with the court decisions. In the end, the article not only teases out the why of compliance decisions but also provides a greater understanding of the relationship between legal compliance and judicial impact.

On March 30, 1995, newspaper headlines declared that hate speech regulations were dead. After six years of litigating over university hate speech codes, a California superior court confirmed what four other courts before it had already ruled - that college hate speech codes, and in this case Stanford's policy, were constitutionally suspect (Corry v. Stanford, No. 740309 [Cal. Super. Feb. 27, 1995]). In the wake of the Stanford decision, many commentators rejoiced. Nat Hentoff, a longtime critic of hate speech codes, hailed the end of a doctrine that had suffered from "fundamental weaknesses" (1995:18). In an editorial titled "Thought Police Disarmed: Campus Speech Codes R.I.P.," the Arizona Republic delighted that "the First Amendment has been reinstated on America's college campuses" (30 March 1995), and the Rocky Mountain News opined that hate speech codes were now "dead letters - unenforced law" (7 April 1995).

But such coverage aside, college hate speech codes are far from dead. As this article demonstrates, hate speech policies not only persist, but they have actually increased in number following a series of court decisions that ostensibly found many to be unconstitutional. This apparent contradiction - between judicial precedent on one hand and collegiate action on the other - may not be surprising to those who study judicial impact, or even to those who understand collegiate policymaking. But such concerted and widespread noncompliance provides an excellent opportunity  [*346]  to examine the process by which institutions respond to a change in the legal environment. Much of the literature to date has focused on the overall impact of Supreme Court case law or on the decisions of individuals or government bodies in responding to new cases. Less is known about the process of organizational compliance or about the connection between individual compliance decisions and aggregate judicial impact.

This article, then, undertakes an empirical investigation into the development and persistence of college hate speech codes, asking why so many elite institutions of higher learning either retained or created speech policies that contradicted a national series of court cases. The method is both quantitative and qualitative, looking for broad patterns of response among the schools and also explaining why individual institutions did or did not comply with the court decisions. In the end I seek not only to tease out the why of compliance decisions but also to provide a greater understanding of the relationship between legal compliance and judicial impact.

In so doing this article is divided into five parts: Section one provides an overview of prior research in judicial impact and compliance, identifying what we already know about the process and suggesting the likely gaps in our understanding. Parts two and three quantify the speech codes, creating a stratified, random sample of one hundred four-year colleges and universities and gathering and coding faculty, staff, or student hate speech policies from each of the schools. In a first, I estimate the prevalence of college hate speech codes, calculating the percentage - and types - that were created between 1987 and 1992 and then measuring them again in 1997 following five court cases that rejected many of the speech policies.

Section four is the heart of the article, which reports on qualitative research at several schools from the sample of one hundred. Having found that many schools failed to comply with the court decisions, the study probed to find out why institutions behaved as they did. The research uncovers distinct motives and behaviors for the schools, including those that complied with the court decisions, those that passively resisted, those that sought to evade the court rulings, and those that ignored them completely in favor of other institutional goals. Finally, with these findings as a base, I return to the distinction between judicial compliance and impact to argue that, rather than being completely separate, the two are more like different sides of the same coin.

I. Theoretical Background
For almost four decades now, public law scholars have studied the impact of judicial decisions and the behavior of individuals and agencies to comply (or not) with those rulings (Patric  [*347]  1957; Sorauf 1959; Murphy 1959; Wasby 1970; Canon & Johnson 1999). There is a tendency at times to treat impact and compliance as similar phenomena. Both seek to estimate the effects of court decisions - whether judicial decisions are followed and what political, social, or cultural impact they produce. But in many respects impact and compliance are the inverse of each other. Impact analysis is a macrolevel question, asking what the societal effects are of a given decision or series of holdings. Compliance studies, though, operate at the microlevel, trying to determine why individuals, political bodies, or organizations do or do not follow court decisions.

The bulk of research on judicial impact or compliance is more than two decades old and together shares a common message: Court decisions do not in themselves command compliance or create a larger impact (Wasby 1970; Canon & Johnson 1999). But if this tenet seems settled, the process of compliance and impact are hardly so clear. To be sure, scholars recognize that law on the books is not "what law is," but this principle appears "settled precisely because it is not talked about much, not because there is nothing to figure out" (Brigham 1996:6-7). What we lack is a fuller understanding of why and under which circumstances individuals, political entities, or organizations comply, or rather, do not, with the law. Why, for example, are some court decisions followed and others not? Why do some decisions have greater impact than others?

Scholars have devoted significant attention to the second question, with Rosenberg (1995, 1991), Kluger (1975), Wasby (1970), Canon and Johnson (1999), and Johnson and Canon (1984), among others, offering explanations for how and when court decisions take on larger, societal meanings. Canon and Johnson (1999), too, have moved impact research closer to the line of compliance analysis; their model of interpreting, implementing, consumer and secondary populations provides a useful guide for understanding the process by which judicial decisions are implemented. But even here much of the theory is most relevant to understanding the impact of court decisions rather than the decisions to comply (or not) with those rulings.

At the compliance level, research has successfully documented what happened in a particular situation, but because this research relies heavily on case studies it has not been as effective at providing theoretical connections to other findings. The result is a collection of competing explanations not that different from the many chronicled by Wasby (1970) thirty years ago. In some cases the explanation for compliance is based on cost-benefit analysis (Giles & Gatlin 1980), while in other studies compliance follows from individuals' policy preferences (Bowen 1995). Still other examples represent as much complementary as they do contradictory explanations (Spriggs 1996; Canon 1991). Nor do  [*348]  many distinguish the amount of deliberation that precedes a compliance decision, let alone distinguish between those court decisions that reflect a "mandatory choice situation" (where action is required by the court case) and those that provide a "voluntary consumption situation" (where consumers of the decision may, but are not required, to act) (Bowen 1995).

It is also worth noting that many of the studies on compliance investigate individuals (Giles & Gatlin 1980; Bowen 1995) or political bodies, including legislatures (Lindgren 1983) and agencies (Spriggs 1996; Johnson 1979; Wasby 1970; Milner, 1971). With a few notable exceptions (Edelman 1990, 1992), there has been less attention given to compliance decisions by nongovernmental organizations. This seems to be a theoretical gap, one that is all the greater considering the concurrent rise in new institutional theory over the past several decades. If scholars now recognize that institutional response is not an exogenous action (Olson 1969; Smith 1988), then organizational compliance may as well be dependent on background pressures, enduring social structures, and a sustained legal environment (Edelman 1990). In fact, private organizations face special pressures. Not only are they subject to formal law but they also operate as mediating institutions in the shadow of such mechanisms, helping to construct the social meaning of legal rules and court decisions and reconciling the impact of law on society outside of government's purview (Edelman 1992).

The upshot of these questions and theoretical gaps is that Wasby's (1970) conclusion of thirty years ago is still very much true today: "We are not ready, it seems, for a broad 'theory of impact'" (245). Indeed, as other scholars have suggested (Lindgren 1983; Spriggs 1996), the move toward a broad theory of impact and compliance requires additional study. In this article I join that fray, using a methodology traditionally employed by others, the case study. I examine the development and persistence of college hate speech codes through a combination of quantitative analysis, archival research, and qualitative interviews. My goal is to understand which - if any - of the competing theories of compliance explains the large-scale decision of American colleges and universities to maintain their hate speech policies in the face of contradictory legal precedent. This is not merely a descriptive study, but rather one that seeks to understand why certain schools removed their speech policies and others did not, to determine which hypotheses best explain organizational compliance.

In this regard I focus on four hypotheses previously offered by scholars to explain the compliance decision: (1) utilitarianism, (2) policy preferences of decisionmakers, (3) threats of judicial enforcement, and (4) organizational inertia. The first two are perhaps the most-cited explanations for judicial impact or  [*349]  compliance (Canon & Johnson 1999; Giles & Gatlin 1980, Bowen 1995), although they are generally mentioned in the area of individual compliance decisions. My question here is whether it is possible to extend them to the organizational level; whether, for example, organizational leaders act on utilitarian or preferential motives in reacting to legal decisions, and then whether the utility or preferences they regard are the organization's or their own. The third theory is one of the most basic hypotheses among compliance scholars, suggesting that agencies, governmental bodies, or organizations are unlikely to follow judicial policies unless the courts bring special enforcement powers to bear (Canon & Johnson 1999). In this case, then, the hypothesis would help to explain noncompliance; except for a small number of schools, the courts did not have jurisdiction over the institutions' policies, and those that did wielded few enforcement powers. Finally, the fourth hypothesis comes from Canon and Johnson (1999) and is echoed by Edelman (1992) and Wasby (1970), suggesting that noncompliance is likely among bureaucratic agencies or organizations where compliance would require changes in organizational processes or forms of governance.

My sense is that other compliance theories will prove less helpful. These include explanations based on cognitive dissonance of organizational leaders, their respect for the judicial process, and their (or in this case the organization's) available resources. The first is really duplicative of the policy preferences theory. Because the views of an organization and its leaders are often so closely linked, understanding the latter's preferences allows one to assess whether the organization's response is consistent with those views. Judicial respect relates to legitimacy theory, a subject of study that has occasionally found different estimates between elites and the general public (Lawrence 1976; Caldeira & Gibson 1992). But since the college speech codes uniformly involve educational elites, it is unlikely that we would find etiological differences in perceived legitimacy. Finally, an organization's concern about available resources would likely be factored into the larger utilitarian calculus of whether or not to comply.

That I do not emphasize these other theories does not mean that I am closed to them should the research suggest their presence and power. For example, while I have excluded communication theory (Canon & Johnson 1999) from among the likely hypotheses, I do so not because it is irrelevant but because it is unlikely to explain differences in organizational response. n1 That point is explained in detail later, but the general message here is  [*350]  that the case study is both informed by causal hypotheses and open to competing and complementary explanations.

On one methodological point, however, I plead guilty. This is a case study about a civil liberties issue, one that, like many others, appears to flow from restrictive judicial decisions (Bond & Johnson 1982). n2 However, while the judicial holdings left little discretion for some schools, they also represented a voluntary consumption choice for others (Epstein 1995), allowing schools the space to decide whether, and in which way, to respond. In this article, then, I try to tease out the decision to comply, seeking to understand why organizations would comply (or not) with a law when their obedience is only voluntary. In the end, the study links judicial impact analysis with compliance studies while providing a fuller and more-nuanced explanation for the organizational pressures that explain compliance.

II. Hate Speech Codes
The topic of college hate speech codes should hardly be unfamiliar to scholars, for the subject has generated great controversy and received considerable attention over the past decade (ACLU 1989; Altman 1993; D'Souza 1992; Hentoff 1995; Lederer & Delgado 1994). Although many have written about the codes' merits, only one other study has attempted to track their numbers empirically. Even that one, conducted at Vanderbilt University's First Amendment Center in 1994, simply counted speech policies at public universities (Korwar 1994).

To estimate the number of schools that adopted hate speech codes I have conducted a stratified random sample of one hundred four-year institutions drawn from the 1987 Classification of Institutions of Higher Education produced by the Carnegie Foundation. While the Carnegie list also includes two-year institutions (generally community colleges) and specialized institutions (music schools and the like), this study is limited to those schools whose mission included a traditional baccalaureate program. n3 The sample was stratified to account for the higher frequency of speech codes among large research universities and top liberal arts colleges. The Appendix lists the one hundred schools.

 [*351]  Schools were studied initially for the years 1987 to 1992, reflecting the period from the codes' initial appearance on college campuses to their potential death knell following the Supreme Court's decision in R.A.V. v. City of St. Paul (1992). The research was only concerned with speech policies adopted during this time. If a school already had a speech policy on the books before 1987 (one school), it was coded as null. Similarly, if a school adopted a policy after 1987 but rescinded it before 1992 (two schools), it was coded as reflecting the more restrictive policy. These decisions mean that the data may slightly overstate the number of speech codes up to 1992 (because two had been rescinded) but understate those that were still valid later. (A school coded as zero for adopting a policy prior to 1987 remained as that in the later data unless it created a new policy.)

I borrowed categories from the Vanderbilt study to code the speech policies. Vanderbilt's researchers used six ordered categories to code university speech policies. The rankings were on a sliding scale, with each group intended to reflect decreasing comportment with First Amendment norms. Eliminating only their code for obscenity policies, n4 the categories I employed are

0. did not adopt speech policies

1. Punished fighting words

2. Banned verbal abuse or harassment

3. Forbade verbal abuse or harassment against minorities

4. Proscribed offensive, demeaning, or stigmatizing speech

Starting from the bottom, it is self-evident that a school averts First Amendment concerns by avoiding speech policies. The next step, fighting words, recognizes that courts have carved out limited areas of expression that state actors may proscribe. According to Chaplinsky v. New Hampshire (1942), public entities may prohibit words "which by their very utterance inflict injury or tend to incite an immediate breach of the peace." Over time the courts have cut the definition of fighting words in half, forbidding only those words which "incite an immediate breach of the peace" (Walker 1994), but the prohibition presumably still exists.

The next category skirts the line of constitutionality, as verbal abuse falls into the common law category of harassment, which itself is a narrow, subjective basis to restrict expression. However, when codes distinguish verbal abuse by its racial/sexual/ethnic message, these policies almost certainly contradict current law. Finally, offensive expression, to the extent that it does not reach fighting words, is constitutionally protected. Any measures that restrict such expression are unconstitutional on the grounds of overbreadth and/or vagueness.

 [*352]  Applying these categories to the schools in the sample (and adjusting for the study's stratification), the data in Table 1 present the percentages of American colleges and universities that adopted various forms of speech policies between 1987 and 1992.

[br n:tab1,l(.10,.10)] [tdm'Verbal harassment against minorities',ql,vu1 [tcgp1,m'Frequency',qc,vu1] [tn1,2]Table 1. Percentage of Schools That Adopted Speech Codes, 1987-1992
[tn1,2][lrw,5d,8] [lc2,4]Type of Speech Policy
Frequency [tu4;2] No speech policy 65 Fighting words 1 Verbal harassment 15 Verbal harassment against minorities 14 Offensive speech 4 [tu4;2] [tn1,2]N = 100. Due to rounding, percentages do not sum to 100. Computing these percentages to raw numbers, Table 2 presents the number of four-year colleges and universities that adopted these policies, as well as the confidence intervals around the results.

[br n:tab2,l(.10,.10)] [tdm'Verbal harassment against minorities',ql [tcgp1,m'Number of Schools',qc,vu1 [tcgp1,m'Confidence Interval',qc,vu1] [tn1,3]Table 2. Number of Four-Year Schools That Adopted Speech Policies, 1987-1992 (range at 95% confidence level)
[tn1,3][lrw,5d,8] [lc2,4]Type of Speech Policy
Estimate for Number of Schools Confidence Interval [tu4;3] No speech policy 897 767-1027 Fighting words 14 0-40 Verbal harassment 207 110-304 Verbal harassment against minorities 193 98-288 Offensive speech 55 3-108 [tu4;3] According to these data the vast majority of college hate speech policies were consistent with constitutional doctrine as late as 1992. I say consistent, for as a matter of law private colleges are generally not bound by the First Amendment (U.S. Const.) and may choose to ignore court decisions on these questions. However, as I discuss shortly, administrators from many of the private colleges in the study represented that they nonetheless try to track First Amendment law in creating or enforcing their policies. Thus the central issue is what the courts would require under the First Amendment. On one end of the spectrum are those schools that did not adopt any speech policies, for by definition they did not challenge First Amendment doctrine. The next group of schools are those that adopted rules for either fighting words or verbal harassment. Although these institutions may have been trying to redirect campus speech or behavior, their policies broke no new constitutional ground. That leaves us with schools that enacted policies against verbal harassment of minorities or offensive speech, codes that were on increasingly shaky legal grounds. In fact, this third division of speech policies may be further divided, with offensive speech prohibitions more  [*353]  suspect than rules banning verbal harassment of minorities. At least until 1992 the 14%, or 193 schools, that adopted policies against verbal harassment of minorities teetered on the line of constitutionality. On one hand, these policies appeared to offer a broad challenge to constitutional law, since the rules singled out particular kinds of verbal harassment as worse than others (R.A.V.). However, given such cases as Beauharnais v. Illinois (1952) and its progeny, the constitutionality of these policies was very much in flux at the time they were adopted. The fact that the Supreme Court even had to take the R.A.V. case says that it was hardly clear whether institutions could punish certain verbal abuse as being particularly worse than others.

By contrast, the 4%, or 55 schools, that adopted the most-restrictive policies - those prohibiting offensive expression - were taking clear constitutional risks at that time. The courts had historically been loathe to prohibit expression that is demeaning, stigmatizing, or offensive, yet many of these policies proscribed offensive speech that stigmatized or victimized another because of race, gender, or ethnicity. These rules seemed to be a significant departure from established constitutional norms, and as the following shows, a number of these policies were struck down by the courts.

III. Court Decisions and Aggregate Response
The hate speech codes attracted considerable attention and opposition, with commentators and politicians alike calling for their repeal. n5 Not surprisingly, several of these policies were challenged in court. At the University of Michigan, University of Wisconsin, Stanford University, and Central Michigan University, student and faculty litigants contested these policies as unconstitutionally overbroad and vague. The first case to go to court was Doe v. University of Michigan in 1988, in which a federal district court ruled that the university could not

establish an anti-discrimination policy which had the effect of prohibiting certain speech because it disagreed with ideas or messages sought to be conveyed... . Looking at the plain language of the Policy, it was simply impossible to discern any limitation on its scope or any conceptual distinction between protected and unprotected conduct... . The terms "stigmatize" and "victimize" are not self defining [and are thus ambiguous]. These words can only be understood with reference to some exogenous value system. What one individual might find victimizing or stigmatizing, another individual might not.

Two years later a federal judge in Wisconsin adopted much of this reasoning when, in the case of UWM Post v. University of Wisconsin (1991), he held that the University of Wisconsin's speech code was also unconstitutional. Although the judge was more forgiving in determining that most terms of the code were unambiguous, he still ruled that Wisconsin's rule was overbroad. Said the court, the "UW Rule has over breadth difficulties because it is a content-based rule which regulates a substantial amount of protected speech... . Content-based prohibitions such as that in the UW Rule, however well intended, simply cannot survive the screening which our Constitution demands" (UWM Post at 1181).

The Michigan and Wisconsin decisions may well have stood as a model for the two later courts that considered the speech codes of Central Michigan University and Stanford University. In Dambrot et al. v. Central Michigan University (1993), a federal district judge overturned a campus rule that prohibited any "verbal ... behavior that subjects an individual to an intimidating, hostile or offensive educational ... environment by demeaning or slurring individuals ... because of their racial or ethnic affiliation" (Dambrot at 481). Borrowing from Doe v. Michigan, the court found that CMU's policy was overbroad, vague, and an impermissible viewpoint restriction. This decision was upheld on appeal by the U.S. Court of Appeals for the Sixth Circuit.

Similarly, in Corry et al. v. Stanford University (1995) a Santa Clara Circuit Court threw out Stanford's speech code on the grounds that the University "prohibited certain expressions based on the underlying message" and thus violated the State's "Leonard Law," which held private postsecondary institutions to the terms of the First Amendment. n6 Although Stanford's code appeared initially to be a little different - in that it professed to prohibit only "fighting words or non-verbal symbols" - the court's decision followed on that of Doe and CMU. Because the speech code proscribed only those fighting words "based on sex, race, color, and the like," the court held it to be an "impermissible content-based regulation prohibited by the First Amendment."

It is important to note that not all types of speech policies were challenged. Those that went to court fell generally into one of two categories: verbal harassment of minorities and offensive  [*355]  speech policies. Theoretically, we might also have expected challenges to generic verbal harassment policies, since, as some scholars claim, prohibitions against any kind of verbal harassment necessarily threaten the speaker's First Amendment rights (Browne 1991). However, what primarily concerned litigants - and undoubtedly troubled the courts - were those policies that sought to proscribe particular viewpoints and/or instituted what they considered to be vague terms to punish particular expression.

In the midst of these cases the United States Supreme Court also took up the question of hate speech prohibitions in R.A.V. v. City of St. Paul, Minnesota (1992). Although involving a municipal statute, R.A.V. had clear overtones to the college speech codes. In its decision the High Court ruled it unconstitutional to punish particular kinds of hateful messages as being worse than others. Even when these restrictions covered a class of expression that was already unprotected (i.e., fighting words), public bodies could not choose to proscribe particular epithets - e.g., racist - while leaving others, perhaps anti-Catholic, untouched.

To some observers R.A.V. and the lower court decisions may appear to be limited precedents. As a matter of strict constitutional doctrine, of course, the cases apply only to public bodies, and there may remain questions about R.A.V.'s clarity and the prominence of the decisions. I suppose there will even be those who claim that each of the lower court decisions applies only to its judicial district and that R.A.V., because it involved a municipal ordinance and not collegiate policies, has no relevance to the hate speech codes. These constructions, however, are overly narrow. Taking the arguments in reverse, it is a matter of basic jurisprudence that the courts will look to other jurisdictions when hearing issues of first impression. Corry and Dambrot relied in part on the reasoning in UWM Post, and the Wisconsin decision borrowed from the reasoning in Doe. Moreover, any ruling of the U.S. Supreme Court, especially on an issue that appears integrally related to later cases, will have precedential if not persuasive authority.

But we need not rely exclusively on the legal rules of stare decisis for this point, as impact analysis itself suggests that the power of court decisions is explained in part by the manner in which they are communicated to interested audiences (Canon & Johnson 1999). In these cases not only were the courts speaking in a single voice, but their message was also read and understood as such outside of the courtroom. R.A.V., in particular, was immediately recognized as speaking directly to the speech codes. Said Justice Blackmun in dissent, "I fear that the Court has been distracted from its proper mission by the temptation to decide the issue over 'politically correct' speech and 'cultural diversity,' neither of which is presented here" (R.A.V. at 416).

 [*356]  Blackmun was hardly alone in suggesting that R.A.V. was directed against campus hate speech codes. The first commentators on the case offered a similar analysis. As the St. Louis Post-Dispatch reported, political correctness "never appeared in Justice Antonin Scalia's decision on Monday striking down a hate-speech law from St. Paul ... but legal scholars said Tuesday that Scalia's opinion was clearly aimed at the proliferating state laws, municipal ordinances and campus codes aimed at racist, sexist and anti-Semitic speech" (Freivogel 1992). Indeed, scholars and observers on various sides of the spectrum saw R.A.V. rooted in concerns about restricting speech on college campuses. Appearing on the MacNeil/Lehrer Newshour, Professors Charles Fried and Lawrence Tribe of Harvard Law School (two ideological opposites) both agreed that the R.A.V. decision was "going to have an effect on campuses which have passed these so-called 'politically correct restrictions'" (26 June 1992). Said Fried, "It seems to me that Scalia was putting his finger" on Stanford, "Michigan[,] and Wisconsin, and a number of other places [where] certain kinds of speech are banned because they upset people who belong to certain categories."

On campus, administrators received communiques from their professional associations summarizing the court decisions and urging them to take note of the evolving legal rules. The American Council on Education, the National Association of College and University Attorneys, and the Association for Student Judicial Affairs, among others, sent circulars to their members about the decisions, and at least one of these organizations held a conference in which members were briefed about the court holdings and warned that their speech policies required a new review. These communiques were monitored on campus by the same attorneys, judicial affairs officers, and law professors who had helped to write the original speech policies (Gould 1999). Given their initial involvement, it is improbable that they would later "tune out" to court cases involving the policies.

But were the holdings so clear? Again, there will be those who say that R.A.V. was far from a clear stroke against the speech codes and that an ambiguous footnote left open the possibility that the speech policies might still be constitutional. That footnote created an exception for situations where "a particular content-based subcategory of a proscribable class of speech can be swept up incidentally within the reach of a statute directed at conduct rather than speech" (R.A.V. at 389). According to Justice White, the Court intended this exception to cover sexual harassment claims that, although potentially involving expression, were based on acts. Although there are doubts about whether sexual harassment law is premised on acts (Browne 1991), this cannot be the case for college hate speech codes, where expression proves the very basis of the claim. Nor can one  [*357]  avoid the message from the Doe, UWM Post, Dambrot, or Corry decisions that offensive or content-based speech policies were constitutionally suspect. Even a leading proponent of the hate speech policies acknowledges that "efforts to regulate ... hate speech not only will fail, but also should fail, to the extent that they trivialize or subvert ... the longstanding, and newly revivified, principle of viewpoint neutrality" (Kagan 1994:202-3).

We are left, then, with a traditional argument from judicial implementation theory, that to the extent that these cases speak to compliance decisions they are relevant only to the actions of public schools. Because the First Amendment does apply to private institutions, the argument goes, private schools were free to ignore the constitutional holdings in Doe, UWM Post, Dambrot, Cory, and R.A.V. But the key word here is "apply," for individuals and institutions may apply legal norms to their conduct even when they are not required to do so by law. Drivers wear seatbelts even when the state in which they are traveling does not require it. Companies offer family leave even if they are exempt from the statute. Organizations offer smoke-free workplaces when statutes only cover restaurants. The common denominator in these examples is that those who are not covered by a governing legal rule may nonetheless choose to apply the norm to themselves.

The same, surprisingly, is true for many private colleges and universities and their acceptance of First Amendment doctrine. Throughout the qualitative research I was startled to find administrators at private schools who readily acknowledged that they created and enforced college policy with a careful eye on the First Amendment. In fact, of the four private schools selected for on-site research, officials at three institutions made this very point, and often with the same terms. Administrators said they wanted their policies to be "legally sound." Initially, their answers sounded like a defensive strategy, that, rather than tracking First Amendment doctrine, schools would create hate speech policies to minimize the possibility of discrimination claims. But as I probed further, the concern seemed to be with the First Amendment. Officials talked about their desire to stay "within accepted legal bounds," of ensuring that their policies are "consistent with free speech norms." In fact, Stanford University, while declaring that it "is not bound by the First Amendment," has reiterated its longstanding policy to "treat itself as so bound" (Grey 1996b: 956). n7

Though their explanations varied, collegiate officials collectively described a First Amendment jurisprudence that extended past its jurisdictional limits to become larger, accepted legal norms. Some took a philosophical approach, talking about the  [*358]  theoretical bases of the First Amendment and arguing that free speech provides the basis of American cultural life and academic freedom. In a similar vein, others spoke of higher education's mission to encourage discourse, saying that, by definition, their institutions would want their policies to conform with First Amendment doctrine. Perhaps the most candid response came from an administrator who worried about the public's perception of a college that did not follow the First Amendment. As he said, "Can you imagine the headline - [College] restricts speech?! ... It is one thing to tell a student privately that 'you don't have First Amendment rights here,' and quite another thing to proclaim publicly that [the College] will not follow the First Amendment."

My claim is not that private schools have voluntarily applied the First Amendment to their activities. Even if that were absolutely true, a case study by its very methodology could not reach that far. But I want to suggest that the traditional public/private distinction in constitutional law neither explains the development of college hate speech codes nor deciphers schools' compliance decisions after the five court cases. Apart from the qualitative results that reveal a preoccupation with the First Amendment at private schools, quantitative analysis suggests no appreciable difference in the behavior of public and private schools. As the data in Table 3 demonstrate, private schools were no more likely than public schools to adopt speech codes in the first place. n8 For that matter, a Logit regression reveals no statistically significant difference between public or private schools in choosing the most restrictive - and thus constitutionally suspect - speech policies. (No table shown.)

[br n:tab3,l(.10,.10)] [tdm'Verbal harassment of minorities',ql [tcgp1,m'Public Sschools (%)',qc,vu1 [tcgp1,m'Private Schools (%)',qc,vu1] [tn1,3]Table 3. Cross-Tabular Analysis: Frequency of Hate Speech Policies in the Years 1987-1992, by Schools' Public/Private Status
[tn1,3][lrw,5d,8] [lc2,4]Type of Speech Policy
Public Schools (%) Private Schools (%) [tu4;3] No speech policy 56 57 Fighting words 8 2 Verbal harassment 5 11 Verbal harassment of minorities 18 18 Offensive speech 13 12 [tu4;3] [tn1,3] [chi] [su'2'] = 0.51, N = 100 Furthermore, examining the number and type of hate speech policies in 1997, there remains little difference between public and private schools. According to the figures in Table 4, speech codes were no more likely to exist at private than at public schools; if anything, the trend is slightly the opposite. Even  [*359]  when examining those schools that had constitutionally suspect policies prior to 1992, a separate Logit regression failed to identify a statistically significant difference between public or private schools in removing or amending their speech policies in light of the court decisions. (No table shown.) Whatever dynamic that explains compliance decisions, then, it is not the fact that public schools were bound by the court decisions and private schools were not.

[br n:tab4,l(.10,.10)] [tdm'Verbal harassment of minorities',ql [tcgp1,m'Public Sschools (%)',qc,vu1 [tcgp1,m'Private Schools (%)',qc,vu1] [tn1,3]Table 4. Cross-Tabular Analysis: Frequency of Hate Speech Policies in 1997, by Schools' Public/Private Status
[tn1,3][lrw,5d,8] [lc2,4]Type of Speech Policy
Public Schools (%) Private Schools (%) [tu4;3] No speech policy 51 56 Fighting words 13 3 Verbal harassment 19 13 Verbal harassment of minorities 13 15 Offensive speech 13 13 [tu4;3] [tn1,3] [chi] [su'2'] = 0.49, N = 100. Due to rounding, percentages do not sum to 100.

A. The Compliance Decisions
The fact that private schools were no more likely than public schools to adopt speech codes, and conversely that public schools were no more likely than private schools to amend or rescind them, brings us to the actual question of compliance. We know that there is no significant difference between public and private schools, but what were their compliance decisions?

If we were to subscribe to a broad model of judicial impact - that judicial decisions command public action and affect public opinion - then R.A.V. and the four lower court cases should have convinced schools with policies prohibiting offensive speech or verbal harassment of minorities to amend or rescind their rules. Here we have a case where not only did the Supreme Court intend to send a message about public behavior but also its meaning was understood as such. Coupled with the decisions in Doe, UWM Post, Dambrot, and Corry, restrictive speech codes should have been a dead letter at public colleges and universities. For that matter, given the courts' influence beyond public bodies, and given the importance that many Americans ascribe to the First Amendment, we might also have expected many private schools to follow suit.

However, as data in Tables 5 and 6 indicate, the trend was just the opposite. By 1997 the percentage of schools with speech policies had jumped 11%, and, while policies against verbal harassment of minorities had dropped 3%, those covering offensive speech codes had tripled.

[br n:tab5,l(.10,.6),tx][pp tpn16] [tdm'Verbal harassment against minorities',ql [tcgp1,m'Frequency',qc,vu1 [tcgp1,m'Range of Estimate',qc,vu1] [tn1,3]Table 5. Speech Codes in 1997 (intervals at 95% confidence level)
[tn1,3][lrw,5d,8] [lc2,4]Type of Speech Policy
Frequency Range of Estimate [tu4;3] No speech policy 54 45-63 Fighting words 4 1-5 Verbal harassment 19 12-26 Verbal harassment against minorities 11 5-17 Offensive speech 12 6-18 [tu4;3] [tn1,3] N = 100 [tdm'Verbal harassment against minorities',ql [tcgp1,m'Net Change in Frequency',qc,vu1] [tn1,2]Table 6. Change in Speech Codes from the Years 1987-1992 to 1997
[tn1,2][lrw,5d,8] [lc2,4]Type of Speech Policy
Net Change in Frequency [tu4;2] No speech policy -11 Fighting words +3 Verbal harassment +4 Verbal harassment against minorities -3 Offensive speech +8 [tu4;2] [tn1,2] N = 100 Admittedly, the change may not be so dramatic when taking into account the confidence intervals, but the number of speech  [*360]  policies clearly rose following the court decisions, with the largest percentage jump coming from the most-restrictive speech policies. Moreover, as the percentages in Table 7 indicate, the vast majority of schools with constitutionally suspect speech policies kept theirs on the books in the face of contrary legal precedent. Table 8, too, provides a closer look at the various strategies that schools followed. There, "offending policies" reflect those speech restrictions considered unconstitutional by the five court cases - verbal harassment of minorities and offensive speech - while "nonoffending policies" cover fighting words and generic verbal harassment, restrictions that were still permitted after the decisions. Although a majority of schools maintained speech policies neither before nor after the court cases, almost a quarter of institutions either retained offending policies or adopted new ones following these decisions.

[br n:tab7,l(.10,.10)] [tdm'Verbal harassment of minorities',ql [tcgp1,m'Percentage in Sample',qc,vu1] [tn1,2]Table 7. Percentage of Questionable Speech Codes That Went Unchanged
[tn1,2][lrw,5d,8] [lc2,4]Category
Percentage in Sample [tu4;2] Verbal harassment of minorities 78 Offensive speech 66 [tu4;2] [tn1,2] N = 100 [br n:tab8,l(.10,.10)] [tdm'No policy before and after court casesx',ql [tcgp1,m'Estimated Percentage',qc,vu1 [tcgp1,m'Estimated Number',qc,vu1] [tn1,3]Table 8. Estimation of Schools' Actions Following Court Cases
[tn1,3][lrw,5d,8] [lc2,4]Schools' Actions
Estimated Percentage of Schools (%) Estimated Number of Schools (N) [tu4;3] Kept offending policy 14 193 Adopted offending policy 9 124 Removed offending policy 2 28 Kept nonoffending policy 17 235 Adopted nonoffending policy 6 83 Removed nonoffending policy 0 0 No policy before and after court cases 51 704 [tu4;3] [tn1,3] Due to rounding, percentages do not sum to 100. That the courts' decisions had neither a powerful impact nor compelled widespread compliance is consistent with prior research in the field (Rosenberg 1991; Canon & Johnson 1999). The question is why this happened. Initially, it is important to define what it means for a school to comply or not comply with the courts' decisions. Returning a moment to Table 8, not all of the schools represented there made a compliance decision. To comply with judicial holdings is to bring a school's policies into line with the courts' rules. Noncompliance, by contrast, means  [*361]  permitting speech policies that conflict with the cases. Thus, the data in Table 8 distinguish between "offending policies" - those whose terms conflict with the courts' holdings - and "nonoffending policies," those that were not touched by the cases.

Given these terms, a school that complied with the courts' rulings would have removed an offending policy, replacing it either with a nonoffending policy or none at all. By contrast, noncompliance reflected two possibilities. Certainly, a school failed to comply with the decisions when it adopted an offending policy even after the cases, but schools that kept offending policies on the books were also in noncompliance. Put another way, noncompliance includes acts of both commission and omission.

In addition, there are several schools that were not faced with a compliance decision. Those schools that regulated fighting words or generic verbal harassment - institutions with "nonoffending policies" in Table 8 - were not affected by the court cases. Putting aside jurisdictional issues of a school's private/public status, the holdings did not reach speech codes less restrictive than those covering verbal harassment of minorities. As a result, schools with nonoffending policies did not confront the question of whether to comply with the courts' decisions.

The preceding discussion notwithstanding, I do not mean to put too fine a point on the distinction between compliance and noncompliance. To be sure, under technical rules of judicial precedent, only those parties to a lawsuit are required formally to comply with the holding. In the case of hate speech codes, then, just four schools - Michigan, Wisconsin, Central Michigan, and Stanford - would have been subject to judicial sanctions had they not amended their policies in light of the court decisions against them. But compliance also has a broader meaning, encompassing an individual's or entity's response to a new and contradictory legal rule. That is the case here, where a host of elite institutions were faced with the issue of whether to bring their own policies or administrative processes into line with judicial precedent that seemed to invalidate their approach. This is an attenuated process, more so than even the school prayer (Muir 1967; Dolbeare & Hammond 1971) or civil rights (Edelman 1990,  [*362]  1992) case studies chronicled in years past. Unlike those examples, just a handful of colleges and universities would have been subject to an adverse declaratory judgment had they not amended or removed their speech policies in light of the court decisions. n9 Yet what makes this a case of legal compliance is the fact that the issue at hand concerns the First Amendment. Even if many of the schools were not technically bound by the court decisions, collegiate leaders still sought to keep their institutions within the bounds of free speech norms, and as a whole they recognized the persuasive authority of the speech code decisions. Indeed, the fascination of the speech codes is the way in which individuals - and in this instance nongovernmental institutions - may seek to comply as much with the spirit of a new legal rule as with the explicit precedent itself. Put another way, this may be a case where the shadow cast by formal law (Mnookin & Kornhauser 1979) has more symbolic influence than the underlying case law or the formal legal order (Edelman 1990).

IV. Case Studies
To understand how schools responded to the hate speech court cases, whether or not they complied with the holdings, the remainder of this article focuses on those institutions that removed, maintained, or added an offending speech policy after the series of contrary holdings. Although it would certainly be interesting to examine schools with fighting words or other nonoffending speech policies, the courts did not condemn these policies, and, as a result, they do not raise as clear an issue of compliance. Thus I examine three types of schools: those that removed an offending hate speech policy, those that maintained an offending hate speech policy, and those that developed a new offending hate speech policy. I estimate these categories as consisting of, respectively, 2%, or 28 schools; 14%, or 193 schools; and 9%, or 124 schools nationwide (Table 8). Within the stratified sample of 100 schools, there were 5 institutions that removed an offending speech policy, 21 that kept an offending speech policy, and 5 that created a new offending speech policy.

A. Case Selection
Thirty-one schools was too large a set to create detailed case analyses for each institution, nor was the process of compliance as complex across the three categories of schools. As a result, I  [*363]  studied the schools through different means. I began by utilizing Lexis/Nexis and other on-line search tools to investigate news coverage of the speech codes and their deliberations at each of the 31 schools. Having recognized that certain schools were more likely to receive extensive coverage in major publications than others, I was careful to sample from regional and local media as well. I was also fortunate to receive access to the archives of the National Institute Against Prejudice and Violence, which subscribed to three clipping services in the 1980s and 1990s and had a plethora of news stories about the speech codes at both noted and less-known schools.

Among these news stories and other published chronologies of the speech codes, there was generally sufficient material to deconstruct the decisions of those five schools in the study sample that had removed their offending speech policies. Many of these institutions had been fully covered by both the national and the local press, and three of the five schools had already been dissected by earlier research (D'Souza 1992; Grey 1996(a) and 1996(b); Arthur & Shapiro 1995). Nonetheless, I also utilized a series of phone interviews at two of the five schools to fill in chronological gaps from the secondary coverage. I spoke with five to eight individuals at each institution, interviewing those identified by the secondary coverage as either ultimate decisionmakers for the speech codes or their chief critics. Because some faculty or administrators had since left the schools, I had to track them down at other institutions, or even in retirement. In only one case did a potential subject decline to be interviewed, citing a desire to "close the door on that controversial chapter," since he had since left the school. Here, I worked around the official's refusal by interviewing others with whom he had worked.

I conducted similar secondary research at the 22 schools in the study sample that had retained offending speech policies in the face of contrary precedent. But, here, organizational decisionmaking was much more complex, necessitating campus visits to delve deeper into the institutions' compliance decisions. Given the demands of time and resources, I selected a pair of schools to investigate at this level, trying to match institutions that shared characteristics previously linked to the development of speech codes.

As earlier research has shown, restrictive hate speech policies were most prevalent at schools that, between 1987 and 1992, had up to six common attributes. These institutions (1) were considered prestigious; (2) had enrolled a small percentage of part-time students; (3) had experienced past protests over South African divestment; (4) had sponsored a gay or lesbian student group; (5) had maintained a minority studies department; and (6) had had a conservative student newspaper on campus (Gould  [*364]  1999). The two schools chosen for campus visits shared four of these six variables (a match that was as close as any between the schools in this category) n10 while also having both initially adopted and subsequently retained an offending speech policy.

I began each campus visit with additional secondary research, reviewing copies of past campus newspapers, local coverage of campus events, as well as archived deliberations from campus committees, the faculty senate, and relevant administrative offices. Having steeped myself in the chronology of events (and in several cases uncovering memoranda that revealed internal deliberations), I then turned to a series of semistructured snowball-style interviews with ten to twenty faculty, staff, and administrators who were involved in, or familiar with, their campus' consideration of hate speech policies. At each school I was able to interview the president or chief academic officer of the institution, the vice president or dean for student affairs, affirmative action officers, and campus attorneys. I was also careful to interview front line student services staff, who were familiar with the application and effect of speech codes on the campus community.

Faculty were key participants in the research, including professors who served on committees responsible for the speech codes and individuals who were noted in secondary materials or mentioned by others as notable proponents or opponents of the speech codes. In addition to these advocates, I repeatedly asked interviewees for recommendations of faculty who were traditionally considered impartial and respected in campus discussions. I thus turned to these professors for a more-detached and less-impassioned interpretation of campus events surrounding the speech codes. I did not, however, interview students. Since the secondary research was conducted between 1997 and 2000, most of the students who had been on campus at the time of speech code deliberations were gone and inaccessible. The omission of students excludes the perspectives of an essential part of any school's community, but, here, I must note the potential bias and move on. In the study's defense, I was still able to reach the other three bulwarks of a campus - its faculty, staff, and administrators.

I used similar methods to investigate schools that adopted offending speech policies in the face of contrary legal precedent. Because of their apparent and direct challenge to legal compliance theory, I sought to visit all five from the study sample. In the end I was able to visit four of five schools, the fifth excluded because the institution was undergoing a change of administration at the time of study, and staff and top officials were unable to meet. None of the intended subjects refused an interview, but  [*365]  academic administrators at two of the four schools were extremely reluctant to disclose much, if any, information in the interviews. Fortunately, I was able to find faculty at each of these schools who were willing to explain the stories their administrators balked at discussing.

One of the strict conditions of my research was that I would not name interviewees or their institutions. Even when I asked subjects to comment on the conclusions of other interviewees I never quoted from the original source. As a result, I use pseudonyms throughout the remainder of this article for both institutions and individuals. In only two cases do I make an exception, where, as with the University of Michigan and Stanford University, institutional decisionmaking was meticulously detailed in open court (Doe; Corry). My goal in using pseudonyms is not to be obtuse but rather to ensure that an enterprising reader cannot link an official's description or quotation with the name of his or her school, thereby revealing the individual's identity and breaking the terms of confidentiality.

B. Qualitative Explanations
The secondary research and campus visits were quite successful at illuminating the various approaches identified in the quantitative results. Here, schools that complied with the court decisions fell into three distinct groupings: those whose policies had been challenged in court, those that decided on their own that the speech codes were unworkable, and one that abided by the court decisions by removing its hate speech policy in light of the adverse legal precedent. By contrast, schools that maintained their offending speech policies - an approach I label "passive noncompliance" - did so generally out of a calculation by collegiate officials that the symbolic advantages of keeping the policies on the books outweighed the scant chance that they would be challenged in court. Finally, schools that adopted new offending-speech policies - a decision I call "active noncompliance" - fell into two categories. Some consciously tried to evade the court decisions, drafting policies they thought would just skirt the court decisions. Others, though, were delayed diffusers, adopting policies in the face of contrary precedent to catch up with what they believed had become the norm in higher education.

C. Compliant Schools
The data in Table 8 suggests that 2% of schools nationwide removed offending speech policies between 1992 and 1997; in the stratified sample there were five such cases. The five schools include a mix of private and public institutions, both large and small, and all well known to most observers of American higher  [*366]  education. But of the five institutions, only three made the decision in light of the court cases, and in fact, two were directly subject to court challenges over their speech codes. Because the machinations of these cases have been well publicized, I can identify the two institutions by name: the University of Michigan and Stanford University. Both are world-class institutions with large and diverse undergraduate and graduate enrollments. Each was also politically charged, boasting alternative, conservative student newspapers and having experienced past protests over South African divestment (Gould 1999). The two schools that acted independently of the court cases are both elite, private institutions, with student bodies that are predominantly white. Although each is primarily a liberal arts college, the two also maintain graduate programs. Like both Michigan and Stanford, the two schools are known for battles over identity politics, having sponsored minority studies departments, gay/lesbian groups, and alternative, conservative student newspapers, while also having experienced considerable student protest over divestment (Gould 1999). The fifth school, which neither acted independently of the courts nor was compelled by a court order to rescind its speech policy, is a public institution. Predominantly white and undergraduate, it nonetheless has a long history of liberal political activism. While few would call it elite - the school accepts over 80% of all applicants - the institution maintains a good academic reputation.

That each of these five institutions rescinded its offending speech policy between 1992 and 1997 does not mean that they all made a compliance decision. Compliance, again, suggests a reaction to judicial precedent - whether explicit or indirect - in which an institution changes its policies or procedures to make them isomorphic with the prevailing legal order (Edelman 1990). However, the two elite schools were not guided by the court decisions in rescinding their speech codes. To the contrary, they responded to considerable backlash, both on campus and in the press, over the enforcement of their policies. At the smaller of the schools a student was charged with racial and sexual harassment following a drunken rage of epithets. Student services staff sought to discipline him for his behavior, but the student responded with an orchestrated media campaign accusing school officials of selective enforcement in punishing him for the content of his speech. Coverage became so great that the school's president felt obliged to respond, claiming that the institution was holding the student accountable for his behavior, not his speech. However, the student was able to marshal allies from conservative members of the faculty, who took pen in hand to chastise the school and its president for adopting a "politically correct" attitude in restricting undesirable messages on campus. The fact that the student also faced possible suspension for his  [*367]  slurs (and past behavioral problems) gave some credence to the charge that the school was overreacting.

The ensuing firestorm seems to have had little influence on the handling of the initial case, but school administrators certainly took a lesson from the situation in considering future discipline. College policy was changed to ensure that students' actions, and not their speech, were open to sanction, and there has not been a case since in which speech has been targeted. A high-level official went so far as to say, "We don't have a hate speech policy." What drove this change is more difficult to determine, although the court decisions do not appear to have played a role. At the highest level, officials maintained that college policy had been "recast" in light of the school's "longstanding commitment to free speech norms." But those who implement student discipline acknowledged that practices were changed to focus exclusively on student behavior. Similarly, while the college's former president disclaimed any influence from the media firestorm, faculty and student services personnel noted the severe beating the school took in the press, and from alumni, during the prior case and acknowledged that top administrators were under tremendous pressure to change policy.

At the other institution, administrators proposed a new speech policy designating zones about campus in which different standards for speech applied. For example, free speech would reign in the student commons, but greater restrictions would be allowed in residence halls. On paper the policy may have had merit, but students reacted negatively to the new plan, drawing chalk lines across campus and openly deriding college officials for "treating [them] like children." Their protests quickly caught the attention of the New York Times, a dynamic that has much to do with the substance of the controversy as it does with the fact that the school is an elite institution, well known by the paper's readership. In fact, students took full advantage of their window of news coverage, painting administration officials as paranoid foes of open discourse and their own followers as champions of free speech and civil liberties. In the end their strategy was successful. The policy had been championed by middle-level administrators at the school, who saw the proposal as an opportunity to set clear rules for an issue - hate speech and violence - that had vexed other institutions. But with few such incidents on campus, it was difficult to overcome the growing exasperation of top officials who feared the negative press attention and continuing protests on campus. About a month after its inception, the policy was retired.

It would be a mistake to count these schools as reaching compliance decisions. Although they did rescind their policies after the bulk of the court cases, the holdings were not part of their calculus. Instead, their decisions reflected a cost-benefit calculation,  [*368]  with the costs of internal strife and negative press attention outweighing any benefits administrators may have anticipated in the quality of campus life or the expectations for racial, gender, or ethnic relations at the school. It is also instructive that the ultimate decisionmakers were top administrators, specifically the institutions' presidents, who overruled middle-level officials who had pushed for the policy initiatives.

This is a dynamic unique to large organizations (be they private or public) and one that reflects the different constituencies of officials at varying level of responsibilities within the organization. At lower levels - in this case in student services - administrators were concerned about the quality of student life and multicultural relations on campus. Officials in the upper echelons of collegiate administration shared some of these interests, but they had other constituencies to accommodate, including alumni, prospective students and their families, and national opinionmakers. They were also up against a wall of advocates - either the suspected student who marshaled conservative supporters or the student demonstrators who "chalked" the campus - who were determined to provoke press coverage and attempt to force the administrators' hands. As scholars have noted in other contexts (Walker 1994), advocates can be crucial in affecting policy decisions, especially when, as here, top officials did not confront protests from the speech codes' supporters.

On one level the speech policies may not have been that popular on these campuses, but it is also instructive that their prime proponents were lower-level administrators, who may have recognized the danger in opposing ultimate supervisors who fretted over the speech policies. I was unable to uncover data that elucidate this question, and as an academician I am well aware that faculty and students often show little trepidation in challenging top administrators. But when the proponents of a policy are nonunionized staff, a group that is neither served by the institution nor protected from recrimination by tenure, it is plausible that some would mute their opposition when convinced that top administrators intend to rule the other way.

If the actions of these two schools do not represent compliance decisions, the reactions of the University of Michigan and Stanford University are the most obvious form of compliance. Both schools were defendants in hate speech cases, and each faced a judicial order to rescind or revise its speech policy. That they would choose to comply with a direct ruling is not that surprising, since as parties to the suits they were subject to direct "legal coercion" (Giles & Gatlin 1980:725). Still, it is interesting that neither institution chose to appeal the trial judge's verdict, a determination that, although similar to that of the University of Wisconsin in UWM Post, contrasts with the strategy of Central Michigan University in Dambrot.

 [*369]  Here, their decisions appear to be a combination of a utility calculus coupled with the policy preferences of senior administrators. Like both of the elite institutions that rescinded their speech policies during this time, Michigan and Stanford had suffered through a plethora of negative, national news coverage over their speech codes. Michigan took the brunt of the criticism since its policy was the first to be challenged and one of the most-restrictive codes, but Stanford also took considerable heat for its speech code. The fact that the author of Stanford's policy, Professor Thomas Grey, wrote extensively about the rule's rationale, only fueled attention of the policy (Grey 1996a, 1996b). In some sense, then, by the time that the Doe and Corry decisions came down, senior administrators at the two schools were tired of the notoriety that their schools had engendered. Moreover, at least at Michigan, university leadership was in transition at the time of Doe. The search committee had successfully recruited a new president, and the interim president, who had presided over the development of the contested speech policy, was able to move on. There were also personnel changes in the university's general counsel's office, the result of which had new administrators looking at the policy in a different light. Some came to agree with the judge in Doe that the explanatory guide for its policy was overreaching, and others simply wanted the controversy to "go away." As a result, the university's response was to amend its policy, to narrow the code's scope so that it was similar to a general harassment policy. This change reflected the policy preferences of new administrators as well as their cost-benefit calculation that whatever influence the existing policy would have in creating a communitarian environment on campus would be overshadowed by the bad press and hard feelings of continued litigation. There had been no groundswell of support from faculty or students for the contested speech policy, and any fears that removing the code would "send the wrong signal" to women, minorities, and others likely to be protected by its terms were relieved by the university's decision to maintain a narrower semblance of a speech policy.

In this sense, then, officials at both Michigan and Stanford had a compliance decision to make - whether to comply with a court order or to challenge the ruling by either appealing its terms or seeking to evade its implications. However, the fifth school to comply had an even more difficult decision to make - whether to respond to a growing body of legal precedent by removing or amending its offending speech policy. This university, a large public institution, has a history of activism but sits in an increasingly conservative area. It is also a school that relies on a state agency for its legal representation. While faculty and administrators were already beginning to question the propriety of the school's speech policy at the time of the court cases, upper-level  [*370]  administrators were pushed to rescind the policy by the state's legal counsel. Senior officials in the attorney general's office had little sympathy for hate speech policies - seeing them as offshoots of a "political correctness" movement - and the growing number of court cases gave them the ammunition to declare to university administrators that the policy had to be revised or removed to comport with "developing constitutional doctrine." The university, in turn, substituted a fighting words policy for the earlier version to satisfy the attorney general's office.

It might have been possible for university administrators to contest the legal advice, but bureaucratic pressures or interorganizational politics explain the school's decision to comply. Campus leaders recognized that lawyers at the attorney general's office were tightly tied to state legal and political leaders, and realizing that the institution depended upon the continued goodwill of those figures for its appropriations, university officials were unlikely to evade the legal advice even if they disagreed. The fact that faculty, staff, and students were already questioning the policy's appropriateness made the decision to comply all the easier.

D. Noncompliant Schools
At first blush it may seem surprising that only 11% of schools that had an offending speech policy would remove it, n11 and more shocking still that even a smaller percentage would act in response to the court decisions. Such intransigence contradicts the traditional model of judicial authority and begs the question why so many schools would ignore the spirit and terms of the court rulings. To some observers the answer undoubtedly lies in a broad coalition of support for the speech codes - that a virtual cacophony of liberal and activist groups had lined up behind the hate speech policies (Walker 1994). Even if identity politics explains the speech codes, a conclusion that other research challenges (Gould 1999), student and faculty activists did not have legal responsibility for the codes, nor would they be called to court to defend the policies. That task fell to university administrators, whose commitment to the speech codes was balanced with other organizational considerations that varied by campus.

The variety of such concerns also explains the diversity of approaches among schools. As Table 8 data indicate, almost 25% of schools nationwide failed to comply at some level with the court decisions. But there is a distinction between schools that kept an offending policy on the books and those that added new ones. I call this dichotomy the difference between active and passive noncompliance, for it takes more administrative action and consumes  [*371]  more organizational energy and human capital to adopt a new policy than simply to keep an existing one on the books. Moreover, as the research shows, the decision to retain an offending speech code reflects different motives than the determination to adopt new policies. This distinction revolves around several competing pressures, including the policy preferences of administrators, their perception of support and opposition for the speech codes, the likelihood of a court challenge, the level of campus activism, alumni reactions, and the relative prestige of each school.

1. Passive Noncompliance
By far, the majority of schools that did not comply chose simply to keep their offending speech policies on the books. Together they total 14% of the sample, but as the data in Table 7 point out, between two-thirds and three-quarters of schools with offending policies in the period 1987-1992 still had them as of 1997. The overwhelming reason for their noncompliance was utilitarianism. As earlier work has found, college hate speech codes were originally adopted for instrumental or organizational goals (Gould 1999). Developed primarily by deans, presidents, or other high-level administrators, the speech policies were designed to counteract unpleasant racial incidents on campus - allowing administrators to assure on-and off-campus audiences that things were "under control" - or as a reaction by less-prestigious institutions to keep up with their more renowned peers in higher education. In only a few cases did deans or administrators promote speech policies because they personally believed in the codes' merits (Gould 1999). Nor were many policies advanced by faculty or students. The policymaking process at most schools was a top-down operation, with administrators acting to serve organizational interests (Gould 1999).

So too, the decision to continue the speech policies reflected a combination of utilitarian calculus and policy preferences by institutional leaders. In this case the benefits of hate speech codes were symbolic, for few if any schools have actively enforced their hate speech policies (Gould 1999). Nevertheless, as several administrators reminded me, one should not mistake symbolism for impotence. Although her speech policy has rarely been used, one dean of students at a northeastern liberal arts college told me that the policy "sets a standard on campus. It gives us something we can point our finger to in the catalog to remind students of the expectations and rights we all have in the community." This sentiment was repeated by a former college president, who claimed that "we didn't set out to enforce the policy punitively but to use it as the basis for our educational efforts at respecting individuality."

 [*372]  If there were symbolic benefits to the hate speech policies, administrators also feared the symbolic costs of removing the speech codes. Here, the operative word was "signal," in that collegiate officials worried that rescinding any type of antidiscrimination policy would send a message to ethnic and racial minorities that they were no longer as welcome on campus. When coupled with the scant possibility of legal action to challenge the speech codes, administrators were content to leave their policies in place.

This calculation was no more pronounced than at the two schools I visited in this category. The first, which I call "Plains University," is a large public university in the central United States. Pulling mainly from its region, 90% of Plains' students are undergraduates, less than 10% of whom are minorities. The other school I call "Middleberg College," a quaint, private liberal arts institution in the eastern half of the United States. Middleberg's enrollment is strictly undergraduate, a large majority of whom are women. Like Plains, less than 10% of Middleberg's students are minorities, although almost three-quarters of the total enrollment comes from out of state.

To comprehend the retention of Plains' hate speech policy, one must also understand its development. The code dates to the late 1980s, when the State Board of Trustees adopted a statement on racial and sexual harassment calling on Plains to develop policies that "identify prohibited conduct in these areas ... [and] eliminate such behaviors... ." Participants in the process disagree over the Trustees' motivation, but any confusion over the policy's genesis belies the symbolic meaning it achieved at Plains University. Shortly after the Trustees' directive a racial incident shook the Plains campus in which a black female student was accosted by a drunken white fraternity brother who proceeded to make racist slurs against her. What otherwise might have been considered an isolated bad act by an inebriated student became a rallying cry for black students on campus, many of whom said the incident was only representative of the racist treatment they received at Plains. Within a week a student group came forward with twelve demands to improve the climate for minority students on campus, and as part of its protest rallied 300 to 500 students demanding university action on a number of concerns.

The university's response was swift, including five directives by the university president "to work [on] problems of racial and cultural harassment and intimidation." Although the eventual hate speech policy was not part of these original mandates, it was issued shortly after the incident and was seen by many as a response to student concerns. Indeed, campus coverage reported the policy as "an answer to intolerant acts on campus." More significantly, the policy has retained that connotation, with senior administrators now in agreement that the speech policy actually  [*373]  helps to attract and retain minority students. Of the four senior administrators I interviewed, each agreed that the hate speech policy is important because it sends a signal to minority students that "they have a place at [Plains]."

Decisions about the speech policy have consistently been the province of upper-level administrators at Plains, the issue having never aroused sufficient faculty or student ardor. Part of the reason may be that student protestors never sought speech policies, but it is instructive that the speech code has never been invoked in the decade it has been in effect. Without any cases to attract attention, the policy has fallen off the campus radar screen.

So, why then would Plains administrators retain their hate speech policy? When asked, both the university's attorneys and top officials in student services say they were aware of the contrary legal precedent and considered it persuasive. Were top officials willfully blind to the decisions, did they personally favor the speech policy, or were there other more pressing matters? The answer must come somewhat secondhand, for the university's chancellor throughout this period - the individual who was ultimately responsible for the decision - has moved on to another position outside of academe and was unavailable for the research. Nonetheless, interviews with university counsel, top administrators in student services, former leaders in Faculty Senate, and a former aide to the Trustees suggest that the chancellor was concerned about "stirring up protest" if the policy were removed. Said a dean, "We had just gone through three or four years in which minority students regularly questioned whether the university was committed to their presence on campus." Regardless of the chancellor's position on the policy, whether students had even pushed for the policy, hate speech codes have been interpreted elsewhere as protection for minorities. As a result, "There is a risk in the signal you send when you change policies."

Still, what about the threat of a declaratory judgment against the policy? Since the state's attorney general openly opposed the policy, might an enterprising student or conservative group challenge the policy in court? Yes, say respondents, but again net utility counseled against a policy change. On one hand, university counsel reports that Plains has rarely, if ever, been sued for a declaratory judgment. Moreover, since hate speech codes had fallen off the campus agenda - if they had ever been a priority - counteractivism was unlikely. Put simply, said a student services staff member, "There was little legal exposure from the policy ... but folks worried about opening a can of worms if we tinkered with it."

The experience at Middleberg College was a little different, for while the school still retained its hate speech policy, the college's dean changed the enforcement strategy out of personal concerns for the policy's reach. That he did not remove the policy,  [*374]  however, reflects a campus consensus that the hate speech code has symbolic value.

Middleberg adopted its hate speech policy in August 1989, prohibiting sexual harassment as well as "harassment based on race, national or ethnic origin, religion, age, and/or disability." Originally styled as a sexual harassment policy, other bases were included late in the policymaking process after a series of racist incidents at neighboring colleges raised the salience of racial and ethnic discrimination. Middleberg had also hired an ombudsperson, whose past experience at a larger school suggested a single, "all-encompassing" antidiscrimination policy for Middleberg.

Adoption of the policy was quite controversial at Middleberg, and unlike the experience at Plains, the code was hotly debated in the college senate before the president eventually adopted it. However, opposition has been surprisingly muted since the policy's adoption, with deans, faculty, and student services staff all crediting the code with "articulating community standards" so that "people can measure their behavior and the behavior of others against" it. Enforcement actions have been rare. A prior ombudsperson recalls three cases from the 1997-98 academic year, but these all turned on action, not expression, and concerned alleged age discrimination, reverse racism, and sexual harassment. Still, with the exception of one faculty member (who calls himself "a strict libertarian") all of those interviewed agreed that the policy has been a success. As one person explains the few formal charges, "You have to understand that 'enforced' is a funny word to use at [Middleberg]. 'Discipline' is informal here. The goal of policy is to set standards of behavior," which are often resolved through informal networks.

Nevertheless, the policy's acceptance lies in more than its enforcement record, for the speech rules are often eclipsed by the policy's prohibitions on sexual harassment. To be sure, the code includes expansive language about verbal harassment against racial and ethnic minorities, but the policy was originally seen as being about sexual harassment, not hate speech. Indeed, one of the policy's prime proponents, a women's studies professor, says she is "surprised" that the policy would be labeled a hate speech policy. Her reaction is particularly confusing when she reports that the college worked closely with its outside counsel in drafting the policy. While their specific advice is not available, most recall the attorneys' admonition that the policy was "too far reaching" and ought simply to have tracked Title VII's language against sexual harassment.

That the college did not adopt this advice does not mean that similar concerns went unraised later. In the year following the policy's adoption the college recruited a new dean of the college, a scholar who calls himself a "free speech advocate" and who believes the college's speech policy was adopted "without elaborate  [*375]  consideration of what it would mean to implement it." The dean reports that in the mid-1990s he, the college president, and the college as a whole were trying to decide "what did this policy mean, what kind of statements are covered, what should we do when incidents arise?" His own concern - as well as one he ascribes to the president and some faculty - was that the code was "excessively cushioning debate." In turn, he proposed that they "back away from the more egregious restrictions on speech" with "subtle changes in tonality." The policy was never rewritten, nor were there proposals for "legislative or judicial changes" in college policy. Instead, in keeping with much of Middleberg's culture, the dean initiated informal discussions among himself, the president, and other members of the college community to redefine collective understandings of the policy's reach. Practically, this meant that administrators would be less likely to invoke the policy (and faculty, staff, and perhaps students would be less likely to file charges) when verbal taunts were not especially persistent or severe.

Still, if the dean were so opposed to speech limitations, why did he not push to amend or remove the policy's terms? He reports that he was aware of the various court cases against college speech codes, and he agreed with their overall message. Instead, he and the college president made a utility calculation. Like the chancellor at Plains University, they were concerned about the message sent to students, staff, and faculty if they minimized or removed the harassment policy so soon after it had been adopted. As a private school they did not fear legal action over the policy's constitutionality, and with so few enforcement actions the threat of any lawsuit about the code was minimal; the potential cost was campus protest and the loss of goodwill. Since the policy was seen as largely covering sexual harassment, any change to the policy - even if on a different issue - might be misinterpreted as condoning sexist behavior. Given the campus uproar five years earlier, both officials were determined not to repeat the debate and protest over sexual harassment; so, they took a middle ground. Moved by a reexamination of the policy's merits, they crafted a new and informal understanding for enforcement practices. But to stave off "collateral protest" from this change, they retained the policy's terms in the college's handbook.

2. Active Noncompliance
There is an important difference between an individual or organization's passive decision not to alter its behavior in light of contrary legal precedent and its active choice to undertake new actions or policies that contradict judicial authority. As the percentages in Table 8 pointed out, 9% of schools engaged in active  [*376]  noncompliance, having adopted new or expanded speech policies that conflict with the five court cases. Of these institutions there were 5 in my sample of 100 schools, and I visited 4 of them for qualitative research.

Among the four schools I found two distinct motives for new speech policies, explanations that I call "evader schools" and "delayed diffusers." Evaders sought to finesse the cases, believing they were binding, or at least persuasive, but still trying to find a way to create an expansive speech policy. Delayed diffusers, by contrast, were not trying to sidestep the decisions so much as they were motivated by a lagged desire to bring their schools "into the mainstream of higher education policy." Administrators at the evaders, institutions that included both public and private schools, had a sophisticated understanding of constitutional law. They knew that their schools were either bound by the First Amendment, or had voluntarily chosen to follow First Amendment doctrine, and not only were they aware of the relevant court decisions but they also sought to draft policies that did not overtly contradict those norms. Even though they wanted to minimize the contradictions between their policies and the court decisions, they were also willing to "push the envelope," so to speak, to achieve other organizational interests.

One of these schools is a large public research institution with agricultural roots, which I call Agrarian University. Agrarian sits in a conservative region and state, a fact that may explain the dearth of political activism on campus. Nearly half of the student body is female, but only 5% of students are minority. The school does not maintain a women's studies or minority studies department, nor did it experience divestment protest.

Agrarian adopted its speech code in 1989 for many of the same reasons as did Plains University. The state Board of Regents had begun to survey other state systems, and believing that racial harassment was the next potential wave of litigation under Title VII, ordered institutions to create policies to protect against such incidents and the liability they presented. At Agrarian, drafting went to a collective of an affirmative action officer, a university attorney, and representatives of the provost's office. Together, they modeled their policy on Title VII's protection against sexual harassment. But by extending the rule to speech and expression while also limiting the bases on which a violation would be found (e.g., race and ethnicity), officials were arguably doing that which R.A.V. eventually prohibited - punishing certain messages as being worse than others.

Participants recall few incidents in which the policy was ever invoked, and archived documents do not identify any others, save one. In the 1991-92 academic year a faculty member was charged under the policy for racist speech in the classroom. Many of the facts remain confidential, but it is clear that the  [*377]  faculty member threatened litigation over the denial of his First Amendment rights, and the university privately settled the case with him. The case strengthened the view of the university's attorneys that Agrarian's hate speech policy was unconstitutional. The attorneys had been tracking the Doe and UWM Post decisions in 1989 and 1991 and increasingly came to believe that their own policy "had constitutional weaknesses." Having analyzed the faculty member's potential court challenge, they concluded that he probably had standing to challenge the entire policy, and that the policy might very well be overturned by the courts.

The attorneys recommended to the president and provost that the university substantially rewrite, if not totally rescind, the policy, and in fact, the revision that followed was prompted, said a dean, "by the recently developing case law regarding First Amendment limitations on universities' regulation of speech... ." Although the same administrator claimed that the resulting policy modification "should be capable of withstanding constitutional scrutiny," an attorney who was closely involved says that the changes were at best a Band-Aid, designed to give the appearance that the university was "still affirmatively and aggressively addressing the issues of racial and ethnic harassment in the university community," while at the same time reducing the chances that the policy would be challenged. As this person acknowledges, the policy that emerged in 1994 "was still unconstitutional," if not more so. Where before the policy covered expression based on "race, ethnicity, or racial affiliation," the revised edition not only engaged in content discrimination, but it used vague and potentially overbroad terms such as "degrade, demean, or stigmatize." The difference, however, was that the revised policy was limited to slander, libel, or obscenity. Thus, while the policy violated R.A.V. for its content discrimination and Doe and UWM Post for its vague terms, the types of potential cases had been narrowed, meaning that the policy would only rarely be invoked.

Those close to the policy deliberations say this crafting was deliberate and that it was done "above the deans' level." Essentially, officials were making a cost-benefit determination similar to those at schools that engaged in passive noncompliance. Since the university had been under investigation by the U.S. Department of Education in the late 1980s for cases of race discrimination, administrators were hypervigilant about signs that the university might retreat from its equal opportunity policies. As the transmittal memorandum for the policy stated, "we have taken the position that the University can and will continue to enforce a policy prohibiting racial and ethnic harassment to the fullest extent the law allows, even if it might be simpler and less legally risky to completely eliminate our policy." But at the same time, the policy quite clearly took a legal risk; it was still unconstitutional.  [*378]  Administrators had attempted to evade the court decisions with legal craftsmanship. While still standing against messages of racial and ethnic hatred, the policy's reach had been sufficiently limited so that few, if any, cases would arise to create a legal challenge.

The other school to evade the courts' rulings is an elite private institution with a religious affiliation, a school I call St. Ann's College. St. Ann's is a small, liberal arts college, split roughly in half between men and women, and containing about 5% of minority students. Although St. Ann's has none of the traditional signs of progressive politics - there being no women's studies, minority studies, or gay/lesbian organizations on campus - both faculty and admissions personnel describe the college as "unusually activist and progressive" for a parochial institution.

As a private, religious school, St. Ann's is not subject to First Amendment doctrine. Nonetheless, administrators expressed a strong allegiance to the First Amendment, each offering several bases for college policy to track First Amendment norms. The school's existing policy was written as late as 1997, but its evolution traces to a change in administrative philosophy for student services. Through the mid-1990s, the dean of student's office had been run by a minister, a former high school principal who had the title of "Dean of Discipline."

The arrival of a new president for the institution spawned a change in student affairs, and the institution moved to what the current dean of students calls a "professional student services model." As part of that turn, staff were hired with graduate training in student services, many of whom belonged to professional associations for student affairs personnel. Their recruitment also invigorated veteran administrators at the school, as officials sought to apply the "best professional practices" to student services. One of these "practices," the associate dean says, was antidiscrimination policies that maintain "an environment of civility that is free from disparagement, intimidation, harassment, and violence of any kind." However, it was not until 1996 that administrators sought to enforce such norms through a punitive policy. Their goal was not only to address the cultural changes on campus that would lead to harassment claims but to move the college, and especially student services, from a parochial to a professional model of administration. To do so, officials contracted with three consultants in higher education law and policy to draft a policy with sufficient breadth but one that remained within constitutional bound. One dean who was involved is quite clear that the school sought a policy that would be consistent with the First Amendment.

What the school got was a policy that seemed to evade the court decisions, a result that an administrator confirms in his description of the consultants. The experts "knew the law well"  [*379]  and tried to "go as far as possible" in preventing harassing speech while still not running afoul of the cases. For example, in its description of characteristics that may invite harassment, the policy says these "attributes include, but are not limited to: race or ethnic origin, gender, physical or mental disability, age, religion, economic class, and sexual orientation." The use of inclusive but not limiting language, of course, avoids content discrimination. But the policy trips up in its definition of harassment, which it describes as physical contact or verbal comments that "degrade the status of another human being." Given the holdings in Doe, UWM Post, Dambrot, and Corry, "degrading speech" would likely be considered vague and overbroad. For that matter, the policy also covers "offensive pictures or 'jokes,'" a restriction that would assuredly be found unconstitutional if challenged in court.

When asked about the consultants' recommendations, an administrator who worked with them admits that the experts have "gotten slammed" in court on similar policies written for other institutions, and officials at the school are still debating how, and under what circumstances, they should enforce their own policy. Nevertheless, three veteran student service officers acknowledge that the college endorsed, if not encouraged, the consultants to finesse the law by pushing the policy's reach as far as practical given the case law. Even more, the college's enforcement pattern suggests that administrators are willing to apply the policy beyond now-established First Amendment limits. When asked if the hate speech policy would apply to a student who claimed in a single classroom debate that "women belong in the home," the dean of students suggested the message would be covered as discriminatory speech. Nor is this a hypothetical case. In the late 1990s officials charged a student with hate speech for distributing a flyer decrying the "sin of homosexuality." Whatever the goodness of their intentions, such restrictions are inconsistent with First Amendment law. n12 In fact, the student filed suit against St. Ann's charging the school with censorship. College attorneys were quick to discount the case as frivolous since St. Ann's is private and therefore is not bound by the First Amendment. But rather than defend the policy in court, the school settled with the student. Their reason, according to both the affirmative action officer and the assistant dean of students, was the president's concern for public relations - his sense that the school would court untold bad press for "being on the wrong side of the First Amendment." Although neither official was responsible for the ultimate decision to settle, they both recognize the pressures now on them when administering the college's hate speech policy. On one hand, the college is committed to rooting out discrimination  [*380]  and has given them broad discretion to apply, or threaten to apply, the policy widely and vigorously. Yet, on the other hand they are wary of the president's support should they actually approach or exceed the boundaries of the First Amendment. In a sense, they say, St. Ann's experience reflects a strategy of evasion - looking for a balance that allows them to cast the policy widely but enforce it in ways that do not court legal or public complaint. n13

The remaining group of schools to actively disregard the court decisions involves what I call "delayed diffusers." In this case they are not technically in noncompliance, for as private schools they are immune from First Amendment doctrine. Moreover, unlike the private institution chronicled as an "evader," these schools have neither considered themselves bound by the First Amendment nor have sought in the past to follow First Amendment doctrine. Within the study sample two schools fit this category, and I visited them both. The two are conservative religious institutions, located in different parts of the country but similar in appearances. The first, which I call Cherrydale College, is a small liberal arts institution tucked away in a rural hamlet. Without a national reputation, Cherrydale recruits students primarily from a three-state region or from members of its denomination. The school is over 95% white, with a 2:1 ratio of female to male students. Alcohol is forbidden on campus, which, the dean of students says, makes the campus atmosphere "respectful." But even faculty who disagree with the school's (and the region's) conservatism do not consider Cherrydale "repressive." The dean of students acknowledges that Cherrydale students have been known to take the one-hour drive to a big state school and "party" for the weekend.

The second school, which I call Ezekiel College, is virtually the same size as Cherrydale and is also located in a remote area about one hour from the nearest metropolitan area. More than 95% of the students are white, although male and female students are evenly matched. Almost all of Ezekiel's recruitment comes from within its denomination, with the academic dean acknowledging that one of Ezekiel's (relatively) unstated goals is to help "proper Christian young people" find suitable mates. Unlike the other schools visited, Ezekiel's rules for student behavior are extremely strict. Dancing was only recently permitted on campus, with the current rule discouraging entertainment that "detracts from spiritual growth and breaks down proper moral inhibitions and reserve." Other rules follow in the same spirit, resembling  [*381]  those that were in place at secular institutions nearly fifty years ago.

To put these schools in the category of noncompliance may seem odd, since as conservative religious institutions the top administrators at each school felt allegiance to a higher power than judicial precedent. But here their actions bring together notions of judicial impact and compliance decisions. In both cases the schools adopted hate speech policies as a way to bring themselves into the mainstream of higher education administration. Each had recently hired student affairs deans with secular backgrounds, and, much like policy entrepreneurs, these new administrators "promoted policy ideas," "articulated policy innovations," and "energized the diffusion process" by which their new schools sought to become more professional and less provincial (Mintrom 1997:739). At Cherrydale the president recruited a new assistant dean of students with an eye toward grooming him for the deanship when the incumbent retired. The new hire came from a prestigious public institution in the East, where he had become active in one of the professional associations for student services. He took the job, he says, "because it was clear that [Cherrydale] was looking to upgrade student services and I could make a real difference in merging the academic and extracurricular sides of campus."

At the time he was hired, Cherrydale's outside lawyers had noticed that the school did not have an antidiscrimination policy, and they began work on a rule that would apply to the college's employees. The assistant dean of students says that he, the then-dean of students, and the school's counselor worked to expand the policy and apply it to Cherrydale's students too. By the fall of 1994 they had succeeded in crafting a policy that, among other things, forbade jokes that "belittle or demean an individual's or group's sex, race, color, religion, or national origin... ." They did not face objection from the college's outside counsel, which coincidentally had changed from a local to a regional law firm during the time and which "passed on the policy" because Cherrydale is a private school. The then-assistant dean of students says he was aware of the Doe, UWM Post, and R.A.V. cases at the time, but in his view, "We had civility goals to achieve" and the cases were not dispositive.

The process was surprisingly similar at Ezekiel College, which over a seven-year period had appointed three presidents. The result was to consolidate institutional power in the dean of the college, the vice president for business, and the dean of students. Two of these men, the academic and student deans, were both hired at about the same time and from secular institutions. Although Ezekiel already had many rules for student conduct, the dean of students undertook to include one more - a provision that would punish students for racial harassment or discrimination.  [*382]  Again, Ezekiel's student handbook was already a model in content restrictions - students being subject to prohibitions on vulgarity, for example - but the dean himself championed the racial harassment provision. No other administrator with whom I spoke would call the rule a "hate speech code," labeling it, instead, a "policy of inclusion" or "diversity," but the dean (who has since retired) allows that it is "substantially similar" to hate speech policies at other institutions. Several faculty members agree that the policy may be termed a hate speech code, but as one cautioned, "It's a misnomer to think of [Ezekiel] policy as 'anti-hate.' This is a conservative, Christian school. Policy here is written with the goal of promoting Christian conduct."

Maybe so, but the author of the policy, the now-retired dean of students, says that "diversity and acceptance" were behind his initiative. Was Ezekiel experiencing a rash of racial violence or harassment? No, say all interviewed. Was the college's purpose even to become more "tolerant" of campus members who did not already subscribe to the code of student conduct? No, too, admit those most closely involved. Instead, says the former dean of students, the racial harassment policy was designed to put the college on record as castigating that which had troubled other institutions - racial intolerance - as well as moving the college "into the mainstream" of student services by adopting a policy which seemed to be "advancing" at other well-known schools.

The dean of students found support for the policy with the dean of the faculty, and together they advanced the policy to Ezekiel's president. Presenting it as a "professional, almost legal requirement," they also managed to bypass the faculty, who, all without tenure, were reluctant to tangle with an issue that was clearly backed by top administrators. Like his counterpart at Cherrydale, Ezekiel's president was supportive of the policy, reflecting as it did the administration's interests in diversifying the institution. In the officials' view, hate speech codes were an important ingredient in this process. Not only would the policies make clear that racial and ethnic minorities were welcome on campus, but with new administrators itching to apply the professional standards they had learned elsewhere, antiharassment codes would bring the schools into line with the "best practices" of other, more-distinguished institutions.

At first the creation of these policies does not seem to make sense. If officials were trying to adopt the best professional practices, why would they embrace policies that had already been found to be unconstitutional? The answer is not simple, but their actions bring us closer to the other side of the impact/compliance coin. While these schools are not technically in noncompliance with the court decisions - being neither bound by the First Amendment nor historically having followed it - they have declined to accept the larger meaning or impact of those holdings.  [*383]  Considering that the five hate speech cases were publicized widely, that their holdings were understood within academe and were publicized by the popular press, their reach presumably went beyond the technical limits of jurisdictional application. These became decisions about the proper limits of hate speech, not simply anonymous disputes involving a few schools or a midwestern city. To disregard these cases, then, was to ignore the implications, nay, the impact, of the courts' decisions. It is a process that is more attenuated than the failure to comply with a verdict. Administrators at these schools did not see themselves as directly flaunting the authority of the courts; instead, their policy preferences and the aspirations for their schools outweighed any deference they would give to the courts' rulings.

One of the features that makes this dynamic so interesting is its relation to the converse situation found in the area of equal opportunity and affirmative action. As Edelman (1992) has argued, the "institutionalization process" of equal employment and affirmative action practices within private organizations makes them "somewhat immune from changes in the political environment" (1568). Yet this process is just the converse for college hate speech codes. Here, the institutionalization process makes policies immune to the changing legal environment. As schools developed these policies in the late 1980s to the early 1990s there arose an administrative constituency for the policies' preservation and enforcement. Whether reflecting the policy preferences of administrators or simply the utilitarian calculus that more was to be gained from the codes than lost, American academe seems to have accepted that the policies have value. At the very least, this norm sets up interests on one side of the scale that are resistant to removing the policies but for direct judicial enforcement. When those enforcement actions did not materialize, the persuasive impact of the courts' hate speech decisions were not enough to overcome the interests that schools and their officials have in the speech policies.

But it may be more than simply organizational interests that explain the speech codes' persistence. Although few administrators offered this explanation in the qualitative research, it is worth considering whether the public has been increasingly accepting of hate speech restrictions. Again, there is not room in an article of this scope for a lengthy discussion of the process of mass attitude change, but one cannot ignore the change in public presentations of the hate speech codes.

[br n:tab9,l(.10,.10)] [tdm'0000',ql [tcgp1,m'Magazines',qc,vu1 [tcgp1,m'Newspapers',qc,vu1 [tcgp1,m'Newspapers',qc,vu1 [tcgp1,m'Television',qc,vu1 [tcgp1,m'Service',qc,vu1 [tcgp1,m'Total',qc,vu1] [tn1,7]Table 9. Annual Press Coverage of U.S. College Hate Speech Codes, 1988-1997
[tn1,7][lrw,5d,8] [lc2,4]Year
Magazines Major Newspapers Other Newspapers Television Wire Service Total [tu4;7] 1988 0 0 0 0 0 0 1989 0 13 14 0 6 33 1990 10 25 27 0 4 66 1991 17 159 221 6 20 423 1992 20 121 192 7 24 364 1993 40 197 254 16 25 532 1994 29 151 256 47 25 508 1995 36 126 207 17 12 398 1996 16 51 97 5 1 170 1997 11 35 87 6 4 143 [tu4;7] As indicated in Table 9, media coverage of the speech policies has dropped precipitously since 1994. With the hate speech codes no longer novel, it may be that the national press corps simply lost interest in collegiate speech policies and chose not to cover them. Or perhaps these policies were no longer as controversial on campus, their terms having been accepted by many students  [*384]  and faculty as part of campus regulations. Since controversy spurs media coverage (Graber 1980), quiet acceptance of the policies would not generate news stories.

At the same time, the decline in media reports has been accompanied by a softening of coverage in at least one medium. As can be seen in Table 10, the speech codes have generated more positive magazine coverage since the court decisions. n14 In 1990, when only Doe had ruled on college hate speech codes, 80% of magazine stories on the speech codes were unfavorable, with only 20% neutral. Three years later, shortly following R.A.V., the stories continued to speak unfavorably of the speech codes. But while none of the stories was favorable, the ratio of unfavorable to neutral had moved to 45%:55%. Remarkably, by 1996 a quarter of the stories portrayed the speech codes favorably. A full majority of the stories were still unfavorable, but for the first time in six years the codes received some positive coverage. This change runs counter to traditional explanations of judicial impact. With the Supreme Court and four lower courts having already found the speech codes unconstitutional by 1995, we might well expect that news coverage of the codes would continue negatively (Wasby 1970); that is, unless public attitudes about the codes were moving away from the court decisions. Although recent national polls have not tested this issue, it is interesting to note that two surveys of incoming college freshmen have found more than 60% favoring the prohibition of racist and sexist speech (Sax et al. 1996; Chronicle of Higher Education 2001). Unfortunately, there are no earlier surveys against which to compare these results, but the data may well suggest that the costs of ignoring the court decisions were minimal. Not only were litigants unlikely to challenge the continuing speech policies in court, but overall attitudes about the codes were becoming more positive, or at least accepting.

[br n:tab10,l(.10,.10)] [tdm'0000',ql [tcgp1,m'Favorable (%)',qc,vu1 [tcgp1,m'Unfavorable (%)',qc,vu1 [tcgp1,m'No Viewpoint (%)',qc,vu1] [tn1,4]Table 10. Magazine Coverage of College Hate Speech Codes, by Selected Year and Viewpoint
[tn1,4][lrw,5d,8] [lc2,4]Year
Favorable (%) Unfavorable (%) No Viewpoint (%) [tu4;4] 1990 0 80 20 1993 0 45 55 1996 25 56 19 [tu4;4]  [*385] 

V. Where Do We Go from Here?
In this article I have examined the mechanics of an organization's compliance decision, but the results have implications for the distinction between judicial compliance and impact. As an initial matter, the study confirms the postulates of Wasby (1970) and Giles and Gatlin (1980) that compliance sits on a continuum. It is a misnomer to think of the compliance decision as a dichotomous variable; instead, as the study showed, schools varied between active and passive noncompliance; and within active noncompliance, institutions differed between evasion and outright nonconformity. The study also provides a fuller picture of organizational compliance. Like others who have examined judicial impact (Canon & Johnson 1999; Giles & Gatlin 1980), compliance was almost nonexistent without the threat of judicial enforcement; of the five schools that removed speech policies following the relevant court cases, only three did so as a result of the decisions, and two were already parties to the litigation. However, unlike in past work (Bowen 1995), in this study schools did not differ by whether they faced mandatory or voluntary precedent; the more significant variation was in their level of noncompliance. As others have suggested at the individual level (Giles & Gatlin 1980), schools differed both in how much they considered their decision to comply as well as in how actively or passively they challenged the case law.

Still, the question is why an institution would comply or not comply with judicial precedent, particularly when the holdings came from five different courts across the nation, including the U.S. Supreme Court. Earlier research has attempted to answer similar questions, with one scholar (Edelman 1990) positing five hypotheses that may explain an organization's response to changing legal norms:

1. organizations comply - or appear to comply - with a new law to maintain their legitimacy;

2. institutions close to "the public sphere" are most likely to amend their policies to mirror new legal requirements;

3. organizations that respond early set a trend that others may follow;

 [*386]  4. within organizations, personnel administrators play a crucial intermediary role in crafting policy; and

5. organizations are influenced by the symbolic value of new measures.

These postulates have some application to the college hate speech codes, but ultimately they do not explain the institutions' behavior. If we define "public sphere" to mean schools in the public's eye, then, certainly, those institutions that had been sued or that had received extensive publicity over their speech codes were the most likely to remove policies that conflicted with the court decisions. So too, the speech codes' symbolic value was a paramount concern for school administrators. Here, however, the symbolic trade-offs worked in the opposite direction of Edelman's study. Administrators were less concerned with appearing to ignore the court decisions so much as they were fearful of campus protest if largely symbolic measures were eliminated from college handbooks. What makes this decision so interesting is that officials were balancing anticipated objections, since the hate speech policies had few open and ardent proponents on campus.

The same is true for Edelman's first postulate, as the speech codes' scant enforcement may suggest the appearance of judicial compliance. Indeed, the response at Plains University and Middleberg College may reflect this approach, for both institutions retained hate speech policies for their symbolic value but avoided a courtroom confrontation by rarely if ever enforcing their policies. But lax enforcement did not begin as a result of the court cases. Moreover, even if nonenforcement reflects a defensive legal strategy, n15 its genesis seems rooted more in the fear of any adverse litigation than a desire by officials to comply with the spirit of new precedent.

Finally, the speech codes diverge from Edelman's other two postulates. Contrary to predictions, delayed diffusers failed to copy schools that removed speech policies, instead of following the example of institutions that years before had adopted them. In addition, personnel professionals - in this case primarily student services staff - were instrumental in advancing or defending hate speech codes, not removing them.

If these studies differ in explaining organizational behavior, the difference is probably in how we define the public legal order. Edelman is right to suggest that institutions will change policies or procedures to become isomorphic with the prevailing legal order (1990). But the legal order goes beyond formal mechanisms of law - those being statutes, regulations, and court cases - to include collective understandings of what those legal  [*387]  norms should and do include. For example, it would be absurd to think that the public legal order includes statutory prohibitions on jaywalking. Certainly, such laws exist on the books, and an individual who sought to remain attentive to formal legal mechanisms would routinely abide by these statutes. But the larger, informal understanding of this norm suggests the opposite consensus, that jaywalking is regularly accepted around the country.

There is not room in this article to examine the process by which informal legal understandings develop, although, fortunately, a pair of other scholars have recently shined a light on this mechanism (Ewick & Silbey 1998). Nor am I prepared to define the moment at which common understandings of legal norms become part of the legal order. My point is to suggest that organizations, like individuals, may comply as much with their perceptions of informal legal norms as they do when acting on the formal representation of law. Certainly, that is the case with college hate speech codes, where delayed diffusers acted on their understanding of what hate speech codes had become rather than on the court cases that seemed to order their removal. So too, many of the student services staff who defended the policies were networked through professional societies such as the Association for Student Judicial Affairs, which failed to concede the institutional goal of stamping out racial violence and expression.

Nonetheless, if the prevailing legal order includes informal understandings as well as formal representations of law, it is a balanced enterprise. Few would argue that common understandings of law ignore formal legal mechanisms, for formal law undoubtedly casts a shadow over policy development and implementation (Mnookin & Kornhauser 1979). Moreover, its shadow is compelling, at least when institutional behavior is held up to public scrutiny. Just as Edelman found a link between an organization's policies and its proximity to the public sphere, so too a college's national renown and press coverage affected its willingness to stray from the formal court decisions. Among other reasons, that is why Michigan, Stanford, and the elite liberal arts colleges were more likely than others to remove, amend, or narrow the application of their speech codes: They fretted over potential negative publicity if opinion leaders were aware that they had enforced their speech codes beyond First Amendment norms.

But formal legal structures are only the starting point for interpreting and negotiating over the meaning of law (Ewick & Silbey 1998). Even if college officials were aware of the contrary judicial decisions, there was an understanding on many campuses that hate speech policies had value, or even if they did not, that others across the institution might be gravely concerned if these codes were removed. Indeed, the polls of college students suggest that, whatever legal precedent the court decisions may  [*388]  have created, they have not persuaded students that hate speech restrictions are improper or unconstitutional. When measured against the retention or creation of new speech codes, we might even say that the courts' interpretations are not accepted within higher education - a fact that the delayed diffusers clearly understood. Their reaction, like those of the other institutions that retained their speech codes, illustrates a constant tension in the creation of legal meaning - how to respond to formal law while still finding a way to rebel against it when one disagrees. As the speech codes indicate, noncompliance likely exists on a continuum, from passive noncompliance, through evasion, up to outright confrontation.

The upshot of this conclusion is that the distinction between compliance and judicial impact is not as clear as past studies have suggested. To be sure, impact analysis is a macrolevel concept, representing an amalgam of discrete compliance decisions; but, as the study suggests, even organizations that are not technically bound by a court case may make decisions about whether to comply with the larger, socially interpreted meaning of the judicial precedent. In a sense, then, these institutions make a decision about whether to comply with the understood impact of court decisions, a process that ties compliance and impact analysis together so that they become different sides of the same coin.

This study, having been a case analysis, is necessarily limited in its reach. Nonetheless, the research should serve as a call to reexamine what we think we know about compliance and judicial impact. Apart from the short shrift that organizational compliance has received, our relatively broad understanding of judicial impact does not necessarily translate to compliance decisions. In this article I have attempted to expand that knowledge base, but there is more that can be done, using both quantitative and qualitative empirical methods. If, as this study suggests, judicial impact and compliance are closely linked, we would do well to devote more attention to their connections and distinctions.


Appendix: Schools in Study Sample
Where a school's state is not clear from its name, the state's abbreviation is added.

[vc2,12.6,1.] American University (DC)

Auburn University (AL)

Beloit College (WI)

Bethany College (WV)

Brandeis University (MA)

Brown University (RI)

California State University, Long Beach

Case Western Reserve University (OH)

Central State University (OK)

Chapman College (CA)

Claremont McKenna College (CA)

Clark College (GA)

College of Great Falls (MT)

College of the Holy Cross (MA)

College of Mt. St. Joseph (OH)

Davis Elkins College (WV)

Denison University (OH)

Drury College (MO)

Duke University (NC)

D'Youville College (NY)

Emmanuel College (MA)

Emory University (GA)

Fitchburg State University (MA)

Florida State University

Georgetown College (KY)

Georgetown University (DC)

Goddard College (VT)

Johns Hopkins University (MD)

Hamilton College (NY)

Hamline University (MN)

Haverford College (PA)

Idaho State University

Illinois Institute of Technology

Indiana/Purdue University at Indianapolis

Indiana State University

Kansas State University

Kent State University (OH)

Kings College (PA)

Knox College (IL)

Linfield College (OR)

Luther College (IA)

Massachusetts Institute of Technology

Messiah College (PA)

Middlebury College (VT)

Mississippi State University

Monmouth College (IL)

Moorehead State University (MN)

Mt. Holyoke College (MA)

New York University

North Carolina Central University

North Carolina State University

North Carolina Wesleyan

Northwest Nazarene College (ID)

Northwestern University (IL)

Nova University (FL)

Oglethorpe College (GA)

Old Dominion University (VA)

Park College (MO)

Queens College (NC)

Regis College (MA)

Rice University (TX)

Ripon College (WI)

Roanoke College (VA)

St. Ambrose College (IA)

Smith College (MA)

Southern Illinois University

Southern Methodist University (TX)

Stanford University (CA)

Sul Ross University (TX)

SUNY, Binghamton

SUNY, Stony Brook

Texas Woman's University

Thomas More College (KY)

Tufts University (MA)

University of Arizona

University of California, Berkeley

University of California, Santa Barbara

University of California, Santa Cruz

University of Florida

University of Hawaii

University of Kansas

University of Maryland, Baltimore County

University of Massachusetts, Amherst

University of Michigan

University of New Hampshire

University of New Mexico

 [*390]  University of Northern Colorado

University of Oklahoma

University of Oregon

University of Puget Sound

University of San Francisco

University of Tampa

University of Texas, Arlington

University of Tulsa

Wagner College (NY)

Wesley College (DE)

West Virginia University

Western Michigan University

Western Washington University

Williams College (MA)


Altman, Andrew (1993) "Liberalism and Campus Hate Speech." In J. Arthur & A. Shapiro, eds., Campus Wars: Multiculturalism and the Politics of Difference. Boulder, CO: Westview Press.

American Civil Liberties Union (1989) "Policy Statement on Free Speech and Bias on College Campuses." In S. Coliver, ed., Striking a Balance: Hate Speech Freedom of Expression and Non-Discrimination. London: Article 19.

Arthur, John, & Amy Shapiro (1995) Campus Wars: Multi-Culturalism and the Politics of Difference. Boulder, CO: Westview Press.

Baum, Lawrence (1977) "Policy Goals in Judicial Gatekeeping: A Proximity Model of Discretionary Jurisdiction," 21 American J. of Political Science 13-35.

Bond, Jon R., & Charles A. Johnson (1982) "Implementing a Permissive Policy: Hospital Abortion Services after Roe v. Wade," 26 American J. of Political Science 1-24.

Bowen, Lauren (1995) "Do Court Decisions Matter?" In Lee Epstein, ed., Contemplating Courts. Washington, DC: CQ Press.

Brigham, John (1996) The Constitution of Interests: Beyond the Politics of Rights. New York: New York Univ. Press.

Browne, Kingsley R. (1991) "Title VII as Censorship: Hostile-Environment Harassment and the First Amendment," 53 Ohio State Law J. 481-551.

Caldeira, Gregory A., & James Gibson (1992) "The Etiology of Public Support for the Supreme Court," 36 American J. of Political Science 635-64.

Canon, Bradley C. (1992) "The Supreme Court as a Cheerleader in Politico-Moral Disputes," 54 J. of Politics 637-53.

Canon, Bradley C., & Charles A. Johnson (1999) Judicial Policies: Implementation and Impact, 2d ed., Washington, DC: CQ Press.

Chronicle of Higher Education (2001) "Fact File. This Year's Freshmen at 4-Year Colleges: A Statistical Profile," January 26: 48-49.

Dolbeare, Kenneth M., & Philip E. Hammond (1971) The School Prayer Decisions: From Court Policy to Local Practice. Chicago: Univ. of Chicago Press.

D'Souza, Dinesh (1992) Illiberal Education: The Politics of Race and Sex on Campus. New York: Vintage Books.

Edelman, Lauren (1990) "Legal Environments and Organizational Governance: The Expansion of Due Process Rights," 95 American J. of Sociology 1401-40.

- - - (1992) "Legal Ambiguity and Symbolic Structures: Organizational Mediation of Civil Rights Law," 97 American J. of Sociology 1531-76.

Epstein, Lee, ed. (1995) Contemplating Courts. Washington, DC: CQ Press.

Ewick, Patricia, & Susan S. Silbey (1998) The Common Place of Law: Stories from Everyday Life. Chicago: Univ. of Chicago Press.

Freivogel, William H. (1992) "Ruling Aimed at Hate-Speech Laws," St. Louis Post-Dispatch, June 24, p. 1C.

Giles, Michael W., & Douglas S. Gatlin (1980) "Mass-Level Compliance with Public Policy: The Case of School Desegregation," 42 J. of Politics 42:722-46.

Gould, Jonathan B. (1999) Symbolic Speech: Legal Mobilization and the Rise of Collegiate Hate Speech Codes. Ann Arbor: UMI Dissertation Services.

 [*391]  Graber, Doris A. (1980) Mass Media and American Politics. Washington, DC: CQ Press.

Grey, Thomas C. (1996a) "Civil Rights Versus Civil Liberties: The Case of Discriminatory Verbal Harassment." In S. J. Heyman, ed., Hate Speech and the Constitution. New York: Garland Publishing.

- - - (1996b) "How to Write a Speech Code Without Really Trying: Reflections on the Stanford Experience," 29 U.C. Davis Law Rev. 891-957.

Hentoff, Nat (1995) "Against the Odds: A Historic Free Speech Victory," Village Voice, May 2.

Johnson, Charles A. (1979) "Lower Court Reactions to Supreme Court Decisions: A Quantitative Examination," 23 American J. of Political Science 792-804.

Johnson, Charles A., & Bradley C. Canon (1984) Judicial Policies: Implementation & Impact. Washington, DC: CQ Press.

Kagan, Elena (1994) "Regulation of Hate Speech and Pornography after RAV." In L. Lederer & R. Delgado, eds., The Price We Pay: The Case Against Racist Speech, Hate Propaganda, and Pornography. New York: Hill & Wang.

Kluger, Richard (1975) Simple Justice. New York: Knopf.

Korwar, Arati R. (1994) War of Words: Speech Codes at Public Colleges and Universities. Nashville, TN: Freedom Forum First Amendment Center.

Lawrence, David G. (1976) "Procedural Norms and Tolerance: A Reassessment," 70 American Political Science Rev. 80-100.

Lederer, Laura, & Richard Delgado, eds. (1994) The Price We Pay: The Case Against Racist Speech, Hate Propaganda, and Pornography. New York: Hill & Wang.

Lindgren, Janet S. (1983) "Beyond Cases: Reconsidering Judicial Review," Wisconsin Law Rev. 583-639.

The MacNeil/Lehrer NewsHour (1992) June 26.

Milner, Neal (1971) The Court and Local Law Enforcement. Beverly Hills: Sage.

Mintrom, Michael (1997) "Policy Entrepreneurs and the Diffusion of Policy," 41 American J. of Political Science 738-70.

Mnookin, Robert, & Lewis Kornhauser (1979) "Bargaining in the Shadow of Law: The Case of Divorce," 88 Yale Law J. 950.

MSNBC (2001) "Court Strikes Down Harassment Policy,", February 15.

Muir, William K. (1967) Prayer in Public Schools: Law and Attitudinal Change. Chicago: Univ. of Chicago Press.

Murphy, Walter (1959) "Lower Court Checks on Supreme Court Power," 53 American Political Science Rev. 1017-31.

Nightline (1991) "'Political Correctness' on U.S. Campuses," May 31.

Olson, Mancer (1969) "Towards a Mature Social Science," 27 International Studies Quart. 29-37.

Patric, Gordon (1957) "The Impact of Court Decisions: Aftermath of the McCullom Case," 6 J. of Public Law 455-64.

Rosenberg, Gerald N. (1988) "Protecting Fundamental Political Liberties: The Constitution in Context." Paper presented at the Annual Meeting of the American Political Science Association, Washington, DC.

- - - (1991) The Hollow Hope: Can Courts Bring about Social Change? Chicago: Univ. of Chicago Press.

- - - (1995) "The Real World of Constitutional Rights: The Supreme Court and the Implementation of the Abortion Decisions." In L. Epstein, ed., Contemplating Courts. Washington, DC: CQ Press.

Sax, Linda J., Alexander W. Astin, William S. Korn & Kathryn M. Mahoney (1996) The American Freshman: National Norms for Fall 1996. Los Angeles: Higher Education Research Institute.

Smith. Rogers M. (1988) "Political Jurisprudence, The 'New Institutionalism,' and the Future of Public Law," 82 American Political Science Rev. 89-108.

 [*392]  Sorauf, Frank (1959) "Zorach v. Clauson: The Impact of a Supreme Court Decision," 53 American Political Science Rev. 777-91.

Spriggs, James F., II (1996) "The Supreme Court and Federal Administrative Agencies: A Resource-Based Theory and Analysis of Judicial Impact," 40 American J. of Political Science 1122-51.

U.S. Senate, Committee on Labor and Human Resources (1992) Hearing on university responses to racial and sexual harassment on campuses. 102nd Congress (2d sess.) September 10.

Walker, Samuel (1994) Hate Speech: The History of an American Controversy. Lincoln, NE: Univ. of Nebraska Press.

Wasby, Stephen L. (1970) The Impact of the United States Supreme Court: Some Perspectives. Homewood, IL: Dorsey Press.

Cases Cited
Beauharnais v. Illinois, 343 U.S. 250 (1952).

Chaplinsky v. New Hampshire, 315 U.S. 358 (1942).

Corry et al. v. Stanford University, no. 740309 (Cal. Super. filed Feb. 27, 1995).

Dambrot et al. v. Central Michigan University, 839 F. Supp. 477 (E.D. Mich. 1993).

Doe v. University of Michigan, 721 F. Supp. 852 (E.D. Mich. 1989).

R.A.V. v. City of St. Paul, Minnesota, 112 S.Ct. 2538 (1992).

The UWM Post et al. v. Board of Regents of the University of Wisconsin System, 774 F. Supp. 1163 (E.D. Wis. 1991).

n1. Communication theory suggests that implementation of judicial decisions depends in part on how such rulings are communicated to those affected by them (Canon & Johnson 1999). As described later, most schools found out about the hate speech rulings through the same means - -by media reports and circulars from educational membership organizations.

n2. According to Bond and Johnson (1992), many court decisions are restrictive - -requiring a party to refrain from particular behavior. These cases are said to confront the party with a mandatory choice situation (Epstein 1995), an oxymoron of sorts that suggests the party has little choice but to follow the court's holding.

n3. This cut reflects both theoretical and practical grounds. On one level the cultural debate over hate speech codes played out almost exclusively at four-year colleges and universities. Because any study depends on variation in the dependent variable, the present research examines the class of schools that both created and eschewed hate speech policies. In addition, because two-year schools are so numerous, the sample was limited to four-year institutions to make data collection feasible.

n4. Although I can imagine several instances in which obscene expression is speech (without also being fighting words), most instances of obscenity involve actions, drawings, or exhibitions rather than speech. As a result, if a college had an expression policy prohibiting only obscenity, I coded it as zero.

n5. While Congress did not act on proposed legislation to prohibit colleges from restricting offensive speech, the Senate Labor and Human Resources Committee did hold a hearing in September 1992 to examine "how American colleges and universities have responded to racial and sexual harassment on campuses" (Senate Committee Report 1992).

n6. The Leonard Law, named for its sponsor Bill Leonard, is unusual among states and was adopted in direct response to the rise of college hate speech codes. The statute forbids "private postsecondary educational institutions" from "subjecting any student to disciplinary sanctions solely on the basis of ... speech or other communication that, when engaged in outside the campus ... is protected from governmental restriction by the First Amendment to the United States Constitution or Section 2 of Article 1 of the California Constitution" (Calif. Educ. Code 94367 [1992]). The Leonard Law does, however, exempt schools controlled by a religious organization to the extent these terms "would not be consistent with the religious tenets of the organization."

n7. This pronouncement came before California's passage of the Leonard Law, which legally bound Stanford and other private institutions in the state to the First Amendment.

n8. To be sure, there is a slight difference in the division of speech codes between those covering fighting words and those prohibiting generic verbal harassment. Nevertheless, public and private schools shared virtually the same rate for creating any speech code, not to mention having the same division over the most restrictive policies.

n9. In the case of school prayer, a decision of the U.S. Supreme Court subjected every public school in America to prohibitions on state-sponsored prayer. So too, the federal Civil Rights Acts prohibited a wide class of public accommodations from discriminating. However, because most of the speech code cases came from lower courts - -and since the Supreme Court's decision in R.A.V. was not precisely on point - -only a subset of colleges and universities were absolutely bound by the holdings.

n10. In this context, "share" means within a single standard deviation of scaled variables.

n11. Of the 18% of schools that originally had an offending speech policy, only 2% were removed (Table 8).

n12. In fact, the U.S. Court of Appeals for the Third Circuit recently struck down such a restriction at the secondary school level (MSNBC 2001).

n13. Given the pressures they face, it is surprising that neither St. Ann's officials nor the other schools visited considered speech codes without penalties - -perhaps simply a statement of college philosophy. Respondents said they were concerned about sending a clear signal that "the policy had no teeth."

n14. Magazines were chosen not simply because the number of stories was more manageable but because their periodic coverage may represent more reflective reporting.

n15. Anytime an institution enforces a disciplinary policy it leaves itself open to a potential lawsuit from the aggrieved party. Thus, institutions may be reluctant to initiate enforcement cases, especially in regard to controversial policies.

Prepared: March 26, 2003 - 5:02:29 PM
Edited and Updated, March 27, 2003

Back to
Home Page