Public Employee Speech, Categorical Balancing and § 1983: A Critique of Garcetti v. Ceballos
Sheldon H. Nahmod *
The Supreme Court’s 2006 decision in Garcetti v. Ceballos is striking. In the course of limiting an important and respected thirty-eight-year-old First Amendment precedent, Garcetti effectively conferred absolute immunity on public employers by excluding employer discipline based on job-required public employee speech from First Amendment scrutiny. It also analogized job-required public employee speech to government speech. It may have thus called into question First Amendment protection for academic freedom because a teacher’s classroom speech is job-required. In ruling as it did, the Court insulated governments and their officials from damages and prospective relief liability under § 1983, the federal statute that is perhaps the most-used vehicle for vindicating public employees’ free speech rights.
I propose to discuss Garcetti‘s First Amendment reasoning as well as the implications of the § 1983 setting in which Garcetti and other public employee free speech cases typically arise. After briefly setting out the Court’s opinion and the three dissenting opinions, I begin by addressing the pros and cons of Garcetti, and in the course of so doing, I discuss the prior Pickering-Connick landscape that Garcetti so significantly altered. I consider the deeper First Amendment implications of Garcetti, including its use of categorical balancing to create an absolute immunity from First Amendment liability for employer discipline based on job-required public employee speech. I also address Garcetti‘s emphasis on citizen status and the functions of public employee speech. The role of the concept of government speech and the possible implications of Garcetti on academic freedom are briefly discussed as well. Finally, I address the relevance of § 1983 to the First Amendment aspects of Garcetti and of public employee speech in general.
I conclude that Garcetti is unsound as a matter of First Amendment policy because it under-protects public employee speech that is vital to self-government. In addition, it creates perverse incentives for public employees to go public and for their employers to broaden job descriptions to capture as much employee speech as possible. Moreover, it is cause for concern that Garcetti calls into question the First Amendment grounding of academic freedom. I maintain that to the extent the Court was worried about over-deterring public employers in the exercise of their managerial authority, the existing § 1983 causation requirement, local government liability, and especially qualified immunity doctrines do an adequate job in minimizing such over-deterrence. For all these reasons, the Court should have retained the Pickering-Connick balancing inquiry and not limited the scope of the First Amendment.
I. The Garcetti Opinions
The Court’s holding in Garcetti, per Justice Kennedy’s majority opinion for five justices, is expressly this: “We hold that when public employees make statements pursuant to their official duties, the employees are not speaking as citizens for First Amendment purposes, and the Constitution does not insulate their communications from employer discipline.” This holding protected supervisory district attorneys sued for damages and injunctive relief under § 1983 by a deputy district attorney who claimed that he was treated adversely because he spoke out on a police officer’s misconduct, a matter of public concern.
The plaintiff, deputy district attorney Richard Ceballos, was a calendar deputy with certain responsibilities over other attorneys. In connection with a pending criminal case, Ceballos was informed about a police officer’s questionable affidavit in support of a search warrant and after investigation, he determined that the affidavit did indeed contain significant misrepresentations. Ceballos informed his supervisors of his findings and also prepared a disposition memorandum recommending dismissal of the relevant criminal charges, followed by another memo a few days later. After meeting with the plaintiff and the police officer who signed the warrant, the plaintiff’s supervisors decided to proceed with the prosecution. The trial court rejected the challenge to the warrant even though Ceballos was called by the defense to testify about his views regarding the affidavit. According to the plaintiff, the defendant supervisors thereafter subjected him to various retaliatory employment actions. Consequently, he sued under § 1983 and the First Amendment, but the district court ruled against him on the ground that the plaintiff’s memo was not protected by the First Amendment because it was written pursuant to his employment responsibilities.
The Ninth Circuit reversed, holding that the memo was protected speech within the meaning of Pickering v. Board of Education and Connick v. Myers because it alleged official wrongdoing and was therefore a matter of public concern. The Ninth Circuit adhered to its precedents rejecting the view that a public employee’s speech loses its First Amendment protection whenever such speech is pursuant to an employment responsibility. The Supreme Court reversed, with four justices dissenting.
A. Justice Kennedy’s Opinion for the Court
Justice Kennedy made several vital points worthy of analysis. First, he asserted that the principle underlying the Pickering-Connick approach was the inquiry into “whether the [public] employee spoke as a citizen on a matter of public concern.” This inquiry, according to the Court, “sought both to promote the individual and societal interests that are served when employees speak as citizens on matters of public concern and to respect the needs of government employers attempting to perform their important public functions.” Consequently, when a public employee speaks pursuant to his official duties (even on a matter of public concern), he is speaking not as a citizen for First Amendment purposes, but rather as an employee subject to employer discipline. This was the proper distinction, not the private nature of the public employee’s speech, or the lack of any relation to the speaker’s job. In this case, Ceballos was not speaking as a citizen but rather as an employee. Interestingly, Justice Kennedy essentially ignored Justice Breyer’s dissenting argument that because Ceballos spoke pursuant to his professional and constitutional obligations as a prosecutor, this should weigh heavily in favor of First Amendment protection.
Second, Justice Kennedy was concerned that a contrary rule of the sort adopted by the Ninth Circuit would displace managerial discretion with judicial supervision, a result incompatible with federalism and separation of powers. He also suggested that current federal and state statutory schemes adequately protected whistle-blowers.
Third, Justice Kennedy acknowledged that the Court’s decision seemed to encourage public employees to go public with their complaints since Garcetti immunity would not apply where the speech was not job-required, but he explained that public employers could institute internal procedures to deal with employee criticism. To the contention that Garcetti immunity provided a strong incentive to public employers to include as much speech as possible in job descriptions, Justice Kennedy warned that the Court would not permit public employers to broaden job descriptions artificially in order to encompass more public employee speech.
Fourth, picking up on an argument in the defendants’ brief, Justice Kennedy analogized job-required public employee speech to government-commissioned or -created speech, with a reference to the Court’s decision in Rosenberger v. Rectors of Visitors of University of Virginia.
Finally, in answer to Justice Souter’s concern in his dissent that Garcetti posed potentially serious problems for First Amendment protection for academic freedom, Justice Kennedy observed that the Court was not deciding whether academic freedom would be covered by Garcetti. Rather, the Court merely noted that “[t]here is some argument that expression related to academic scholarship or classroom instruction implicates additional constitutional interests that are not fully accounted for by this Court’s customary employee-speech jurisprudence.”
B. The Dissents
The dissenting Justices Stevens, Souter, Ginsburg, and Breyer all had a common goal: to reject the majority’s rationale supporting immunity for job-required speech and hence retain the Pickering-Connick ad hoc balancing test. Though each applied the Pickering-Connick test differently, all reached the same conclusion: Ceballos’s speech should be protected by the First Amendment.
Justice Stevens, in his dissent, stated that the majority’s approachâ€”to exclude employer discipline from any First Amendment liability for job-required public employee speechâ€”was “misguided” and went too far. There would be many cases in which public employees could be punished for exposing wrongdoing to supervisors, and “it is senseless to let constitutional protection for exactly the same words hinge on whether they fall within a job description.” According to Justice Stevens, such public employee speech should “sometimes” be protected.
Justice Souter’s more lengthy dissent, joined by Justices Stevens and Ginsburg, argued for a modified Pickering-Connick approach in public employee cases involving job-required speech that would take proper account of the weighty government interests involved. In order to prevail in his suit, the public employee must speak “on a matter of unusual importance and satisf[y] high standards of responsibility in the way he does it . . . . [I]t is fair to say that only comment on official dishonesty, deliberately unconstitutional action, other serious wrongdoing, or threats to health and safety can weigh out in an employee’s favor.” In the course of his dissent, Justice Souter criticized the majority’s distinction between the individual as citizen and as public employee, and its reliance on both the concept of government speech and the existence of federal and state statutes protecting whistle-blowers to support its decision. 
Justice Breyer was alone in his dissent. After pointing out that Garcetti involved an issue the Court had not addressed in the pastâ€”a public employee speaking on a matter of public concern in the course of his duties as a government employeeâ€”he rejected the majority’s position that there is no First Amendment protection whatsoever from employer discipline in such a case. He also argued, however, that Justice Souter’s adjusted Pickering-Connick approach did not give enough weight “to the serious managerial and administrative concerns that the majority describes.” In his view, a properly adjusted Pickering-Connick approach in public employee job-required speech cases should take account of both an “augmented need for constitutional protection” and the “diminished risk of undue judicial interference with governmental management of the public’s affairs.” The Garcetti case itself called for the application of Pickering-Connick because the speech at issue was professional speech that was, at one and the same time, an obligation of the legal profession and an obligation of the Constitution. For Justice Breyer, the need to protect Ceballos’s speech was “augmented” because the speech was required both professionally and constitutionally while the government’s need to control it was “likely diminished.” Thus, the First Amendment mandated “special protection of employee speech in such circumstances” and the Pickering-Connick balancing should have been applied.
II. Garcetti Analyzed
A. Categorical Balancing
In ruling that the job-required speech of public employees is not protected from employer discipline by the First Amendment, the Court used what has evolved into a conventional approach: categorical balancing. Categorical balancing, as distinct from the ad hoc balancing of the Pickering-Connick test, has been used by the Court to deny First Amendment protection to certain categories of speech, e.g., fighting words, obscenity, and child pornography, all of which are considered outside the scope of the First Amendment. On the other hand, categorical balancing has also been used to expand the protection of the First Amendment to include speech previously considered to be outside its scope, e.g., subversive advocacy, defamation, and commercial speech. In addition, categorical balancing was used in Elrod v. Burns to provide First Amendment protection to public employees from employer discipline based on their political affiliation, where party affiliation is not an appropriate requirement for the effective performance of job duties.
What the Court does when it balances categorically is weigh what it considers to be the relevant interests, social and individual, at a fairly high level of generality, and then by balancing those interests, arrive at a generally applicable rule to be applied in later cases without further balancing. Categorical balancing is thought to have the advantages of predictability (through rules), as well as the flexibility inherent in the balancing process.
The recent trend in the Court, with the unique exception of child pornography, has been to use categorical balancing to expand First Amendment-protected speech. Garcetti is an obvious departure from this trend, but it is also a departure in another way. When the Supreme Court determines that certain categories of speech are unprotected by the First Amendment, a crucial aspect of its categorical balancing is its determination that the affected speech has little or no intrinsic First Amendment value. In marked contrast, Garcetti did not question the First Amendment value of job-required public employee speech because the speech was highly valuable and relevant to self-government. The Garcetti Court did, however, assert that the First Amendment value of job-related speech is trumped because of the government’s overwhelming interest in workplace efficiency as employer. To put this another way, even job-required speech that is clearly a matter of public concern and that is required professionallyâ€”perhaps constitutionallyâ€”is not protected from employer discipline by the First Amendment because it is not considered the speech of a citizen.
Garcetti thus resembles Ferber‘s categorical approach to child pornography. Child pornography may indeed have serious social, artistic, or medical value. Nevertheless, this First Amendment value is significantly outweighed by the adverse impact child pornography has on children. Similarly, job-required public employee speech often deals with matters of public concern and has serious First Amendment value because it implicates self-government. Yet the government interest in the efficient operation of the government workplace, which according to Garcetti includes both minimizing the displacement of managerial discretion by judicial intervention that applies ad hoc balancing and promoting federalism and separation of powers, is thought to significantly outweigh any First Amendment value.
B. The Distinction Between a Citizen and a Public Employee
Garcetti described the theoretical underpinning of Pickering-Connick as the distinction between the speech of a citizen (who happens to be a public employee) on a matter of public concern, and all other public employee speech. Pickering itself is a good example of a case in which it was relatively easy to conclude that the public employee spoke as a citizen because he wrote a letter to a newspaper.
On the other hand, sometimes the content, form, and context of a public employee’s speech indicate that it is speech on a matter of private rather than public concern, in which case the First Amendment is inapplicable to employer discipline. Connick is a good example of such a case because the public employee complained about her own employment situation. In Garcetti, the Court used this distinction to reach its conclusion that job-required public employee speech is by definition not the speech of a citizen, and therefore not protected by the First Amendment from employer discipline, even if it addresses a matter of public concern.
Garcetti‘s reasoning thereby turned the Pickering-Connick test on its head by privileging employment status over the subject matter of public employee speech. What was previously crucial for First Amendment protection was not the status of the public employee as a citizen, but rather the subject matter of the speech and its promotion of self-government.  If the subject matter addressed a public concern and promoted self-government, it was highly valuable speech and potentially protected by the First Amendment. Under Pickering-Connick, the public employer’s interest was taken into account through balancing only after this threshold determination was made. Conversely, if the subject matter addressed a matter of lower-value private concern and did not promote self-government, then it was not protected by the First Amendment from employer discipline because there were no individual and social interests worthy of such coverage. In such a case, neither consideration of the public employer’s interests nor balancing was required.
After Garcetti, however, the subject matter of the speech, the promotion of self-government, and the value of the speech are not relevant to the threshold inquiry. Now the function of the speech is determinative in deciding the public employee’s First Amendment status. If the speech is required by the job, the public employee loses his status as a citizen with First Amendment protection against employer discipline, and he is exclusively an employee. Only where the speech is not job-required, such as when that employee goes public with his speech, does the Pickering-Connick test potentially become applicable because the public employee retains his status as a citizen.
Consider also Garcetti‘s disjunctive treatment of the public employee’s status: for employer discipline purposes, a public employee is either a citizen whose speech is covered by the First Amendment, or an employee whose speech is not. This, however, cannot be correct because Ceballos was obviously a citizen when he spoke. As pointed out by Justice Souter in his dissent, with a nod to Robert Frost, this disjunctive treatment does not accord with reality.  A public school teacher who speaks out at a school board meeting about collective bargaining negotiations, or a public school teacher who complains to a superior about racial composition of various school staffs, is at once both a citizen and an employee. Indeed, because we live in a complex and pluralist society, we play various roles simultaneously in our public and private lives with different and sometimes conflicting obligations to groups to which we belong. Even the notion of citizenship itself is sophisticated because one can be both a citizen of the United States and a state, with different rights and responsibilities to each. The Pickering-Connick ad hoc balancing test, particularly if modified to take into account the elevated government interest when public employee speech is job-required, adequately captures this complicated reality. 
There is yet another serious problem with the Court’s disjunctive approach: it is fundamentally inconsistent with the self-government rationale of the First Amendment. By definition, the workplace effected by Garcetti is a government workplace, precisely the place where a knowledgeable public employee’s speech could most effectively educate superiors, and ultimately the public, about government and how it actually operates. Indeed, the First Amendment interest in such cases appears generally to be very weighty. This is not to suggest that this First Amendment interest should invariably trump the government’s interests either in exercising managerial authority and workplace efficiency, or in choosing to adhere to a particular workplace model grounded on hierarchical organizational theory. These are unquestionably valid and weighty interests, but the categorical exclusion of job-required speech from any First Amendment protection with respect to employer discipline, on the ground that it is never the speech of a citizen, is particularly troubling.
Finally, it is instructive to compare Garcetti‘s move from ad hoc balancing to categorical balancing with developments in First Amendment defamation law beginning with New York Times Co. v. Sullivan, and culminating with Gertz v. Robert Welch, Inc. and Dun & Bradstreet, Inc. v. Greenmoss Builders, Inc. In Sullivan, the Court emphasized the political subject-matter of an advertisement and the fact that the advertisement was directed at a public official’s conduct. The Court then used categorical balancing to arrive at the now-famous formulation: a plaintiff who is a public official suing for defamation in connection with his official conduct must allege and prove knowing or reckless falsehood. Thus, it is the plaintiff’s status as a public official that triggers this First Amendment protection. Status serves as a kind of proxy for the inquiry into First Amendment self-government value.
Status subsequently became ever more important when a divided Court once again used categorical balancing to extend the Sullivan rule to public figures. Public figure status also serves, albeit somewhat loosely and less persuasively, as a proxy for the inquiry into First Amendment self-government value. The focus on status continued through Gertz, where the Court used categorical balancing to modify the knowing or reckless falsehood standard, and thereby ruled that private plaintiffs have the lesser burden of proving negligence and actual damages, even where the allegedly defamatory speech involves a matter of public concern.
The emphasis on status and the related use of categorical balancing as determinative of the applicable First Amendment rule in defamation cases came to a halt in Greenmoss Builders. The Court there ruled by plurality that the status of a plaintiff as a private citizen does not automatically mean that the Gertz negligence-and-actual-damages rule applies. Instead, the determination of whether the allegedly defamatory speech is about a matter of public concern is “â€˜at the heart of the First Amendment’s protection.’” The plurality reasoned that where there is both a private plaintiff and a matter that is not of public concern, the common law of defamation applies instead of Gertz.
The Court’s move from Sullivan to Greenmoss Builders partly resembles the reverse of the move from Pickering-Connick to Garcetti. The defamation doctrine shifted from categorical balancing and an inquiry into the plaintiff’s status as public official, public figure, or private person, to an ad hoc inquiry into whether the speech allegedly defaming a private plaintiff deals with a matter of public concern. In contrast, the public employee speech doctrine shifted from an ad hoc inquiry into whether the relevant speech deals with a matter of public concern to a categorical inquiry into the plaintiff’s status as either a citizen or an employee depending on whether the speech at issue is job-required.
Under Gertz and Greenmoss Builders, the determinative inquiry in private plaintiff cases is whether the relevant speech deals with an issue of public or private concern. This inquiry indicates that the Court is confident about judicial competence to distinguish between issues of public and private concern in defamation cases. It also suggests that the costs of unpredictability in the defamation area are worth bearing. The same holds true for the way in which the Court currently applies the First Amendment to the dissemination of confidential or private information. The Court’s high comfort level with distinguishing between issues of public and private concern, with the attendant cost of unpredictability, is thus in stark contrast to its obvious discomfort with Pickering-Connick where job-required speech is involved.
At the same time, the choice of whether to use categorical balancing or ad hoc balancing is neutral with respect to First Amendment outcomes. This neutrality is demonstrated by the fact that categorical balancing is used in Garcetti to limit First Amendment protection for public employees, and an ad hoc inquiry into whether an issue is one of public or private concern is used in Greenmoss Builders to limit First Amendment protection for the media. The criticism of Garcetti is thus not aimed at categorical balancing as such. The criticism is directed only at the way categorical balancing is applied in Garcetti to narrow the scope of public employees’ First Amendment protection because of an overstated, even if legitimate, concern with managerial authority and discretion.
C. Incentives and Disincentives for Public Employees and Employers
The Court’s categorical exclusion of job-required public employee speech, even where professionally required, from First Amendment protection against employer discipline is clearly intended to protect managerial authority and discretion by reducing the specter of judicial intervention in managerial decisions. This exclusion could backfire, however, by increasing judicial intervention. Garcetti creates an incentive for public employees who feel strongly about their concerns to go public with their whistle-blowing and other grievances in an attempt to retain protection under the First Amendment, thereby potentially causing even greater harm to public employers than purely internal employee speech would.  The Court’s responseâ€”that Garcetti encourages public employers to build internal procedures for handling employees’ complaints to forestall their going publicâ€”may turn out to be wishful thinking, particularly where those complaints are directed to official misconduct and the complaining employees expect reprisals. For example, think of police officers who reveal the wrongdoing of fellow officers: they are sometimes disciplined by their superiors and suffer other reprisals for violating the “code of silence.” Public employees in this situation will often have little to lose by going public.
At the same time, Garcetti creates an even more powerful incentive for public employers to broaden their job descriptions so as to capture as much public employee speech as possible and thereby insulate employer discipline from First Amendment scrutiny. The Court’s warning that this will not be tolerated emphasizes the likelihood that there will be many cases in the future dealing with the propriety of public job descriptions. This prospect may also result in considerable judicial interference with managerial decisions and undercut an important justification by the Court for its use of categorical balancing in Garcetti.
Furthermore, and irrespective of employer attempts to broaden job descriptions in response to Garcetti, it is surely not unusual for a public employer to require its employees to report official misbehavior and illegal conduct. After Garcetti, such a job requirement insulates public employee speech reporting such conduct to superiors from First Amendment protection against employer discipline. Finally, some public employers include in employees’ job descriptions the requirement of communicating regularly with the public, including the media, as well as with politically accountable branches of government. After Garcetti, the First Amendment may not protect employees from employer discipline based on such communications, despite their obvious importance for self-government. Consequently, public employers can have it both ways: they can require their employees to report official misbehavior and illegal conduct, and at the same time avoid First Amendment protection for employees who do so.
D. Some Preliminary Observations on Government Speech and Academic Freedom
Extended consideration of government speech is well beyond the scope of this discussion. Still, it is impossible to ignore a potentially significant case citation in Garcetti at the point where Justice Kennedy finished explaining that job-required speech is not protected by the First Amendment because it is not the speech of a citizen. He then states: “Restricting speech that owes its existence to a public employee’s professional responsibilities does not infringe any liberties the employee might have enjoyed as a private citizen. It simply reflects the exercise of employer control over what the employer itself has commissioned or created.” This statement is immediately followed by a reference to Rosenberger v. Rector & Visitors of University of Virginia, and a parenthetical quote from that case: “[W]hen the government appropriates public funds to promote a particular policy of its own it is entitled to say what it wishes.” 
What may be significant about this reference to Rosenberger is that government speech is currently considered to be outside of the scope of the First Amendment. In government speech situations, the government is a kind of market participant in the marketplace of ideas, rather than a regulator of that marketplaceâ€”a role that insulates government speech from any meaningful First Amendment scrutiny. This exception for government speech suggests Garcetti‘s rationale to be that government employees who engage in job-required speech are never protected by the First Amendment from employer discipline, regardless of the content or context of their speech, because the government has bought the speech.
It is this broad implication of the reference to Rosenberger and the Court’s reading of Rust v. Sullivan that may explain Justice Souter’s reaction in his dissent. He emphasizes that unlike the usual government speech case, “Ceballos was not paid to advance one specific policy among those legitimately available, defined by a specific message or limited by a particular message forbidden.” In his view, Garcetti was very different from Rust where employees were required by the terms of the federal grant to their agency to espouse a substantive position the government set out in advance.
This possibility of an overbroad application of the government speech doctrine in the public employee speech setting is intimately connected to the issue of academic freedom. Because public elementary, secondary, and higher education teachers are public employees who are paid to speak, Garcetti could be read for the proposition that little if any speech uttered in the classroom by a public school teacher or professor is protected from employer discipline by the First Amendment. This reading, however, calls into question prior statements by the Supreme Court that academic freedom is protected by the First Amendment, particularly at the university level.
The Court’s unsurprising response to this possible reading is to avoid the subject, because as it correctly observes, the facts in Garcetti do not expressly raise the issue of academic freedom. The Court says only the following: “There is some argument that expression related to academic scholarship or classroom instruction implicates additional constitutional interests that are not fully accounted for by this Court’s customary employee-speech jurisprudence.” However, the very difficult open question, which cannot be avoided by the Court for very long, is how to square Garcetti‘s categorical balancing approach insulating job-required public employee speech with First Amendment protection of academic freedom.
Perhaps one possibility after Garcetti is to focus on the job in a particular educational setting. At the elementary and secondary levels, for example, the primary, albeit not sole, purpose as determined by government is the transmission of basic knowledge and values. Because the students in these environments are a captive audience (and the government typically determines the curriculum), a teacher assigned to teach mathematics should not be protected by the First Amendment from employer discipline if he or she instead teaches a different subject. This appears to be a relatively easy case in which to affirm that the government is hiring the teacher to engage in specified government speech. Also, the government may choose to provide considerable direction to teachers at the elementary and secondary levels to cover, or not cover, a particular topic or sub-topic or to teach by using a specific pedagogical approach. In such cases, if the teacher deviates from that direction, employer discipline for the deviation may be immune from First Amendment scrutiny, although it may be subject to procedural due process protection if inadequate notice was given to the teacher about such a requirement. Insulation from First Amendment scrutiny in such cases makes sense, because government should have the discretion to choose the educational and classroom approach it prefers, just as it should have the discretion to choose the kind of workplace it prefers, regardless of the wisdom of its decision.
In contrast, First Amendment protection may apply more broadly in the university setting than in elementary and secondary public schools, even after Garcetti. The purpose of a university, as determined by government, is not exclusively the transmission of knowledge and values but also the development of critical intellectual faculties. Further, the students are older and not a captive audience. Accordingly, the university classroom could be characterized as a kind of designated educational public forum where both faculty and students have a great deal of speech autonomy. Such autonomy is subject to two uncontroversial educational requirements: first, discussion in the classroom must be related, even if loosely, to the subject matter, and second, the educational process must not be impeded by disruptive tactics. In other words, by analogy to Rosenberger and its use of forum analysis, university classrooms could be viewed as government creations for the enabling of professorial and student speech, with government viewpoint discrimination prohibited by the First Amendment. Moreover, Rosenberger and forum analysis could be used to confer First Amendment protection on the publications of university professors who are required by their jobs to publish; the theory is that universities are in the business of promoting and enhancing the scholarship opportunities of their professors.
Finally, it may be worth noting that, as a practical matter, few public (or private) universities could realistically survive an attempt to treat university professors and their students like elementary and secondary teachers and their students. Not many would choose to teach at or attend institutions of higher education that eliminated academic freedom in teaching and scholarship.
III. The § 1983 Setting
A. The Ever-Present Hurdles for § 1983 First Amendment Public Employee Speech Claims
I have suggested that despite the difficulties attending the application of the Pickering-Connick test, including some lack of predictability, Garcetti‘s categorical absolute immunity approach is questionable as a matter of First Amendment doctrine and policy. But there is more to Garcetti than the First Amendment, and that is the § 1983 context in which most public employee cases, including Garcetti, arise. In my view, Garcetti is in significant part a § 1983 case dressed in First Amendment garb: the Court changed First Amendment doctrine in order to promote not only its conception of First Amendment policy, but § 1983 policies as well. Garcetti, it is true, explicitly addresses the costs of federal judicial intervention into public employer decision-making as a matter of First Amendment policy. Nevertheless, I read Garcetti as also sensitive to the possibility of over-deterrence of managerial decisions resulting from the potential damages (and attorney’s fees) liability arising out of successful § 1983 actions brought against public employers and supervisory officials accused of violating the free speech rights of public employees. Garcetti can also plausibly be read as sensitive to the possibility of over-deterrence resulting from the very threat of being sued. To the extent that these sensitivities to over-deterrence in fact motivate the Court’s approach in Garcetti, I maintain that they are both misplaced and overstated. There is no need to limit the scope of First Amendment protection for public employees because reducing over-deterrence and interference with managerial authority is adequately promoted by current § 1983 doctrines dealing with causation, local government liability, and especially qualified immunity. Because of these doctrines, it is quite difficult for public employees to obtain relief for their First Amendment deprivations.
A § 1983 public employee plaintiff who challenges a retaliatory employment action for protected speech must, of course, allege and prove that the speech is protected under the Pickering-Connick balancing test. But that is only the beginning. The plaintiff then has to prove that the adverse employment action was the causal result of the protected speech, or in other words, that the defendant public employer’s or defendant supervisory official’s adverse employment action was impermissibly motivated by the protected speech. If the plaintiff shows that the sole motive for the adverse employment action was impermissible, then the plaintiff has made a prima facie case and the parties move to the question of immunities. If, however, the plaintiff shows only that the impermissible motive was a substantial factor in the adverse employment action, then the defendant can escape liability altogether through the Mt. Healthy burden-shift affirmative defense. Under this defense, the defendant can prevail by showing by a preponderance of the evidence that there was a permissible motive for the adverse employment action, even though there may have been an impermissible motive as well, and that the plaintiff would have suffered the adverse employment action anyway. In short, the defendant can prevail on the constitutional merits by proving the absence of but-for causation.
If a local governmental entity such as a school district is sued by a § 1983 First Amendment plaintiff, it can be held liable for damages, even in a First Amendment case of first impression, because individual immunities do not apply. The plaintiff, however, must prove that an official policy or custom of the school district caused his First Amendment deprivation. The technical § 1983 requirements for an actionable official policy or custom are quite arcane and difficult to prove, but even where they are satisfied, it is still difficult to prove that an official policy or custom was impermissibly motivated, as required for a § 1983 First Amendment claim.
Moreover, and of special significance here, even if the § 1983 First Amendment plaintiff can make a prima facie case against supervisory officials in their individual capacities and not stumble over the Mt. Healthy burden-shift, such defendants can raise qualified immunity, an affirmative defense particularly intended to minimize over-deterrence. There may be cases in which absolute immunity is implicated, but for the most part it is qualified immunity that is applicable because employer discipline decisions are typically administrative in nature. Qualified immunity, an affirmative defense measured by objective reasonableness, can protect an individual defendant from damages liability, even though the defendant may in fact have violated a plaintiff’s constitutional rights. The defendant must demonstrate that a reasonable official could have believed at the time he acted, in light of then-existing clearly established law and the facts known to the defendant, that his conduct was constitutional.
It turns out that there is a tendency in the circuits to confer qualified immunity protection on § 1983 First Amendment defendants because the Pickering-Connick balancing test is just thatâ€”a balancing test that takes account of various factors. This frequently makes it even more difficult for plaintiffs in such cases to show that the defendants violated then-existing clearly established law and acted in an objectively unreasonable manner.
All of this demonstrates that established § 1983 doctrine, even apart from the applicable First Amendment standard, creates significant hurdles for § 1983 plaintiffs and already goes far to reduce both the costs of liability and of defense in an attempt to minimize over-deterrence and interference with managerial authority. The existence and application of these doctrines suggest that Garcetti overstates the adverse impact on managerial discretion of the Pickering-Connick ad hoc balancing test and undermines a major justification for Garcetti‘s holding that job-required speech is not protected from employer discipline by the First Amendment.
B. The Impact of the § 1983 Setting on the Constitutional Merits
Garcetti, then, is concerned not only with First Amendment doctrine, but also with over-deterrence and the resulting interference under Pickering-Connick, with managerial discretion stemming both from potential § 1983 damages liability and from the very threat of § 1983 actions themselves. Rather than retain the Pickering-Connick balancing test and address this concern through qualified immunity in particular, as it should have, the Court in Garcetti cut back on the scope of the First Amendment and effectively created absolute immunity for public employers sued by their employees for job-required speech. As it turns out, this is not the first time the Court has made this kind of move in the First Amendment public employee setting.
Over a decade before Garcetti was decided, the Court in Waters v. Churchill addressed a § 1983 public employee First Amendment case where the employee was dismissed on the basis of reports of unprotected, insubordinate speech. However, the reports, which had been incompletely investigated by the public employer, were inaccurate and the employee may have actually spoken on protected matters of public concern. The question was whether a First Amendment Pickering-Connick claim was stated. From a straightforward First Amendment perspective, the answer would seem to be yes because the speech that motivated the discharge was indeed of public concern, as Justices Stevens and Blackmun argued in dissent.
Yet, the Justices split on the First Amendment issue and were unable to generate an opinion for the Court. The plurality, in an opinion by Justice O’Connor, adopted a reasonableness requirement under which an employee could not be discharged or otherwise disciplined for speech on an issue of public concern unless the public employer had a reasonable basis for believing that the speech was either disruptive or that it involved a matter of purely private concern. In the plurality’s view, employers have a First Amendment duty to conduct a reasonable investigationâ€”albeit not a full-scale oneâ€”into the nature and content of the speech. Thus, First Amendment protection depended on what the employer reasonably believed the nature and content of the speech to be.
Justice Scalia, however, joined by Justices Kennedy (the author of Garcetti) and Thomas, concurred only in the result. They argued that reasonableness was not the proper standard because it would subject public employers to “intolerable legal uncertainty.” In their view, public employers had no duty to investigate and should only be subject to § 1983 First Amendment liability for retaliating against an employee because he spoke on a matter of public concern.
It is clear from the opinions in Waters that because the Justices were divided about the impact of potential § 1983 liability on public employer decision-making, they were also divided as to the existence of a First Amendment violation in the first place. Consequently, even though the reasonable investigation issue could have been folded comfortably into the qualified immunity objective reasonableness inquiry, the plurality (and those concurring in the judgment) chose not to do so in order to avoid finding a First Amendment violation where an employer’s investigation was reasonable. Instead, the plurality split the difference between the position of the dissenters and those concurring in the judgment, and determined that there is a First Amendment duty to conduct a reasonable investigation, which, if breached by the public employer who discharges or otherwise disciplines an employee for speech that was in fact on a matter of public concern, constitutes a First Amendment violation. If this duty is not breached, however, the public employer does not violate the First Amendment even if the motivating factor for the discharge was in fact speech on a matter of public concern.
It should not be surprising that judicial concerns about interference with independent decision-making, typically arising in § 1983 cases implicating absolute and qualified immunity and thus the costs of both liability and of defending, can and do spill over into the First Amendment merits in § 1983 public employee free speech cases, as in Garcetti and Waters. After all, § 1983 frequently triggers actions for damages for violations of several other constitutional provisions.
For example, procedural due process, substantive due process, and Fourth and Eighth Amendment violations regularly serve as the basis for many § 1983 damages actions. In each case, rather than relying on the qualified immunity doctrine, the Court has manipulated state of mind and other similar requirementsâ€”and therefore the relevant constitutional doctrineâ€”in order to minimize the possibility of over-deterring government officials and employees. Thus, the Court has ruled that: negligence is insufficient for procedural and substantive due process violations; police officers sued under substantive due process (and § 1983) for injuries resulting from high-speed pursuits must be shown to have acted in a “conscience-shocking” manner with a purpose to do harm, and ordinary deliberate indifference is insufficient; police officers sued under the Fourth Amendment for their alleged use of excessive force must have terminated a plaintiff’s freedom of movement “through means intentionally applied” even though the Fourth Amendment speaks of reasonableness, a police officer’s subjective state of mind is irrelevant to the probable cause determination; subjective deliberate indifference, in the sense of the failure to act when prison officials know of a substantial risk of serious harm to inmates, is required for Eighth Amendment violations; and malicious and sadistic infliction of pain is the applicable Eighth Amendment standard for prison officials accused of causing harm to inmates when using force in connection with maintaining prison security.
In all of these cases, the relevant state of mind requirement is determined not primarily through textual interpretation but rather by virtue of the gatekeeper function the relevant state of mind serves in connection with minimizing over-deterrence. This is also true for the gatekeeper function that the Court apparently hopes will be performed by Garcetti‘s exclusion of employer discipline for job-required speech from First Amendment scrutiny.
Garcetti is questionable as a matter of First Amendment policy because it under-protects speech that is vital for self-government. Governments should be discouraged from punishing public employees for speaking truthfully about official misconduct or illegal behavior, even where such speech is job-required. Garcetti unfortunately creates strong incentives for public employers to broaden the category of job-required speech that will insulate them from First Amendment scrutiny for employer discipline. Garcetti will also lead to considerable uncertainty regarding the existence and scope of First Amendment-based academic freedom in classrooms at all educational levels. Significantly, many of the Court’s concerns with over-deterrence in Garcetti are already adequately addressed through § 1983 causation, local government liability, and qualified immunity doctrines. The Court should not have limited the scope of the Pickering-Connick test and thus the First Amendment itself.
Garcetti suggests that government speech is relevant to public employee speech. I agree, although not in the sense the Court has in mind. The Garcetti decision itself is government speechâ€”albeit that of the Supreme Court and not of Congress or the Presidentâ€”that sends several unwise messages. The first message is that public employees must choose between putting their jobs at risk or failing to live up to their perceived moral and legal obligations to report official misbehavior and illegal conduct, particularly where the public is endangered. This is not to suggest that constitutional and moral norms do or should always coincide, but simply that under Pickering-Connick balancing, public employees who report official misbehavior and illegal conduct to their superiors often act in a manner consistent with both applicable First Amendment and moral norms. After Garcetti, though, this First Amendment protection is removed primarily for instrumental reasons, and the opportunity to declare official conduct blameworthy is thereby diminished.
The second message embedded in Garcetti is that there is a First Amendment disconnect between being a public employee and a citizen. The public employee receives no First Amendment protection from employer discipline for highly valuable job-required speech related to self-government, whereas the citizen typically receives wide-ranging protection for such speech. This disconnect is not only false, but it also erodes the character traits required of Americans living in a democracy that promotes liberty.
The third and last message is perhaps the most disquieting. It is Garcetti‘s declaration to public employers that they can escape all First Amendment and § 1983 responsibility for discipline imposed on public employees who report official misconduct or illegal behavior to their superiors. Because it redefined First Amendment harm by unwisely creating absolute immunity for employer discipline arising out of job-required speech, Garcetti permits, and may even encourage, public employers to abdicate their responsibility to act for the public interest.
* Copyright, 2007. Distinguished Professor of Law, Chicago-Kent College of Law. A.B., U. of Chicago; LL.B., LL.M., Harvard Law School; M.A., Religious Studies, U. of Chicago Divinity School.
I want to thank Jack Beerman, Erwin Chemerinsky, Marty Malin, David Rudovsky, and Michael Wells for their helpful comments on an earlier draft. I also want to thank my colleagues at Chicago-Kent for their participation in a roundtable at which I discussed my preliminary views on the subject of this article.
. See generally Pickering v. Bd. of Educ., 391 U.S. 563 (1968). Pickering established the principle that when a public employee speaks as a citizen on a matter of public concern and there is little or no adverse impact on either the employment relationship or the operations of the governmental entity, then the employee is protected by the First Amendment against employer discipline. See id. at 573â€“74. Thereafter, in Connick v. Myers, the Court refined its approach and articulated a balancing test with the “matter of public concern” inquiry serving a gatekeeper function. 461 U.S. 138, 142, 146 (1983). Under Pickering-Connick, the first question is whether the public employee’s speech deals with a matter of public concern. If the answer is no, the First Amendment inquiry is at an end; if the answer is yes, then the court must balance the speech against its impact on the governmental entity’s enterprise. See Connick, 461 U.S. at 150â€“51.
For criticism of the Court’s move in the direction of making the public concern inquiry into a separate First Amendment category, see Cynthia L. Estlund, Speech on Matters of Public Concern: The Perils of an Emerging First Amendment Category, 59 Geo. Wash. L. Rev. 1 (1990).
Every person who, under color of any statute, ordinance, regulation, custom, or usage, of any State or Territory or the District of Columbia, subjects, or causes to be subjected, any citizen of the United States or other person within the jurisdiction thereof to the deprivation of any rights, privileges, or immunities secured by the Constitution and laws, shall be liable to the party injured in an action at law, suit in equity, or other proper proceeding for redress, except that in any action brought against a judicial officer for an act or omission taken in such officer’s judicial capacity, injunctive relief shall not be granted unless a declaratory decree was violated or declaratory relief was unavailable. For the purposes of this section, any Act of Congress applicable exclusively to the District of Columbia shall be considered to be a statute of the District of Columbia.
Id. Section 1983 is the subject of my three volume treatise, Sheldon H. Nahmod, Civil Rights and Civil Liberties Litigation: The Law of Section 1983 (4th ed. 2007) [hereafter Nahmod, Civil Rights and Civil Liberties Litigation].
. 126 S. Ct. at 1960. It probably goes too far to call this holding revisionist, but the factor the Garcetti Court deemed determinativeâ€”whether the public employee’s speech was communicated as part of the employee’s official dutiesâ€”had not previously been considered significant in the Court’s public employee First Amendment jurisprudence.
. 126 S. Ct. at 1962. Justice Stevens dissented. Id. at 1962 (Stevens, J., dissenting). Justice Souter, joined by Justices Stevens and Ginsburg, dissented, id. at 1963 (Souter, J., dissenting), and Justice Breyer dissented. Id. at 1973 (Breyer, J., dissenting). I discuss these dissents later. See discussion infra Part 1.B. For present purposes, it suffices to note that the dissenting justices rejected the categorical approach of the majority. This case was first argued when Justice O’Connor was still on the Court, and it was reargued after she retired and was replaced by Justice Alito, who joined the majority. See id. at 1954 (majority opinion).
. Garcetti, 126 S. Ct. at 1960 (citing Rosenberger v. Rectors & Visitors of Univ. of Va., 515 U.S. 819, 833 (1995)). In Rosenberger, the Court struck down, on First Amendment grounds, a University of Virginia policy which authorized payments of printing costs from a student activities fund for various student publications, but prohibited such payments for any student publication that “primarily promotes or manifests a particular belief in or about a deity or an ultimate reality.” 515 U.S. at 823. According to the Rosenberger Court, this was not a case in which the government spoke or subsidized the trans-mittal of a message which it favored, as in Rust v. Sullivan, 500 U.S. 173 (1991), but rather one in which it expended funds “to encourage a diversity of views from private speakers.” Rosenberger, 515 U.S. at 834. Consequently, under the First Amendment the University could not engage in viewpoint discrimination. See infra note 132 and accompanying text.
. Garcetti was initially argued before Justice O’Connor retired from the Court. See Transcript of Oral Argument, Garcetti, 126 S. Ct. 1951 (2006) (No. 04-473). The Court, however, had not yet handed down the decision when Justice O’Connor retired and was replaced by Justice Alito. Accordingly, Garcetti was reargued. U.S. Supreme Court Docket, Garcetti, 126 S. Ct. 1951 (2006) (No. 04-473). Therefore, the dissenting opinions probably resulted from unsuccessful attempts to persuade Justice Alito to retain Pickering-Connick in cases involving job-required speech.
. See id. at 1966, 1970â€“71. Justice Souter disputed the majority’s claim that such statutes adequately protect whistle-blowers. He pointed out that whistle-blowing statutes typically address exposing an official’s misconduct publicly, not internally. He also suggested that whistle-blower statutes are a “patchwork” under which different statutes protect different employees and different kinds of conduct. See id. at 1970â€“71. Moreover, he observed that federal whistle-blower statutes provide relatively little protection in many circumstances. “[I]ndividuals doing the same sorts of governmental jobs and saying the same sorts of things addressed to civic concerns will get different protection depending on the local, state, or federal jurisdictions that happened to employ them.” Id. at 1971. Justice Souter further observed that the existence of federal and state statutory remedies should be irrelevant to First Amendment analysis.
In my view, however, there is much more to say against the Court’s reliance on whistle-blower statutes. For constitutional purposes, the existence of statutory remedies has been considered relevant to the existence of a constitutional violation only where plaintiffs allege random and unauthorized conduct in violation of procedural due process. See Parratt v. Taylor, 451 U.S. 527 (1981), overruled in part by Daniels v. Williams, 474 U.S. 327 (1986). In contrast, statutory remedies have not ordinarily been considered relevant to the existence of substantive constitutional violations, primarily because the interests protected by substantive constitutional provisions are different from, and often broader than, the interests protected by the statutory remedies. See Bivens v. Six Unknown Named Agents of Fed. Bureau of Narcotics, 403 U.S. 388, 398â€“411 (1971) (Harlan, J., concurring). The only possible exception that comes to mind is the rule that a plaintiff who alleges that a state regulation is so severe that it constitutes an inverse condemnation of his property in violation of due process is not ordinarily permitted to sue under § 1983 for a due process violation. Instead, he must pursue state judicial remedies for a taking of property. See Williamson County Reg’l Planning Comm. v. Hamilton Bank, 473 U.S. 172 (1985). Even under that rule though, the plaintiff is permitted to sue under § 1983 once judicial remedies are exhausted.
In effect, what the Court has done in Garcetti is to improperly import the approach of Bivens, where the Court observed that one of the factors relevant to implying Fourth Amendment (and other constitutional) damages causes of action against federal officials is the unavailability of an effective federal statutory remedy. See Bivens, 403 U.S. at 402 (Harlan, J., concurring). In Bivens, the Court concluded that there were “no special factors counselling [sic] hesitation in the absence of affirmative action by Congress.” Bivens, 403 U.S. at 396. But Bivens never held that the existence of Congressional remedies is relevant to the existence of a Fourth Amendment violation; it held only that such Congressional remedies are relevant to the availability of a constitutional damages remedy. See, e.g., Schweiker v. Chilicky, 487 U.S. 412 (1988) (elaborate Congressional remedial scheme precluded due process damages actions against federal officials administering the Social Security program).
. See 427 U.S. 347, 355â€“56, 360, 364â€“65, 367â€“69, 372â€“73 (1976). The Court categorically balanced the interests in patronage against the free speech interests in political association and concluded that discharging a local government’s non-civil service employees solely because they did not support the Democratic party violated the First Amendment. Id.; see also Rutan v. Republican Party of Ill., 497 U.S. 62, 64, 69, 74â€“75 (1990) (applying Elrod to hiring, promotions and transfers in addition to discharges); Branti v. Finkel, 445 U.S. 507, 513, 517â€“18 (1980) (limiting Elrod to “non-policymaking” public employment).
. See id. at 960â€“61, 979â€“81, (specifically criticizing “definitional balancing,” a term coined by Professor Nimmer that appears to mean substantially the same thing as “categorical balancing”).
There are certain well-defined and narrowly limited classes of speech, the prevention and punishment of which have never been thought to raise any Constitutional problem. These include the lewd and obscene, the profane, the libelous, and the insulting or “fighting” wordsâ€”those which by their very utterance inflict injury or tend to incite an immediate breach of the peace. It has been well observed that such utterances are no essential part of any exposition of ideas, and are of such slight social value as a step to truth that any benefit that may be derived from them is clearly outweighed by the social interest in order and morality.
. Recall Justice Breyer’s argument in dissent that Ceballos was required to disclose his concerns with the police officer’s search warrant affidavit both as a matter of professional ethics and constitutional law. Id. at 1974 (Breyer, J., dissenting).
. What is “employer discipline” within the meaning of Garcetti? See id. at 1960 (majority opinion). Suppose, for example, that Ceballos’s superiors report him to the state’s bar authorities for possible professional disciplinary action. The bar authorities who act on the basis of these reports and impose professional discipline should not be permitted to rely on Garcetti since they are not Ceballos’s employer and their conduct therefore does not constitute employer discipline. Ceballos’s superiors should also not be able to rely on Garcetti if their reports result in professional discipline imposed by the bar authorities; their reports do not constitute employer discipline, even though they arise out of the employment relationship because they do not directly affect Ceballos’s terms and conditions of employment. Cf. Burlington N. & Sante Fe Ry. Co. v. White, 126 S. Ct. 2405, 2409 (2006) (holding that Title VII’s ban on retaliation covers non-job related acts by an employer that “could well dissuade a reasonable worker from making or supporting a charge of discrimination”). In both cases, Ceballos should be able to use conventional First Amendment doctrine as a shield against the bar authorities and his superiors.
Garcetti, therefore, should not insulate governments and their officials who do not act as employers from First Amendment scrutiny when they act adversely to a public employee because of job-required speech. In addition, Garcetti should not insulate public employers from First Amendment scrutiny when they provide information based on job-required speech to another government body that acts adversely to a public employee.
Consider, however, the possibility that the public employer could defend against § 1983 liability on a superseding cause theory because the decision of the other government entity to proceed against the public employee for his job-required speech was made independently of the public employer and is therefore not attributable to the public employer. On the other hand, in Malley v. Briggs, 475 U.S. 335, 337, 339 (1986), a § 1983 damages case involving the plaintiff’s arrest based on a defective arrest warrant prepared by a police officer and judicially approved, the Court rejected the officer’s argument that the judge’s approval of the warrant broke the causal link between the officer’s applying for the warrant and the plaintiff’s subsequent allegedly unconstitutional arrest. The Court emphasized the judge’s reliance on the police officer’s affidavit. Id. at 344 n.7; see generally 1 Nahmod, Civil Rights And Civil Liberties Litigation § § 3:103â€“3:107 (discussing proximate cause in the § 1983 setting).
. For example, Justice O’Connor concurred in Ferber in order to emphasize that the “compelling interests identified in today’s opinion suggest that the Constitution might in fact permit New York to ban knowing distribution of works depicting minors engaged in explicit sexual conduct, regardless of the social value of the depictions.” Ferber, 458 U.S. at 774 (O’Connor, J., concurring) (internal citations omitted). In contrast, Justices Brennan and Marshall concurred in the judgment and argued that “[t]he First Amendment value of depictions of children that are in themselves serious contributions to art, literature, or science, is, by definition, simply not de minimis. At the same time, the State’s interest in suppression of such materials is likely to be far less compelling.” Id. at 776 (Brennan, J., concurring) (internal citations omitted).
. The threshold requirement under Pickering-Connick‘s ad hoc balancing approachâ€”whether the public employee’s speech raises a matter of public concernâ€”is not always easy to apply because it is so very fact-specific, dealing with content, form, and context. Furthermore, even where a matter of public concern is found, balancing must be applied. See supra note 2. The outcome under the Pickering-Connick test is therefore often unpredictable, exposing local governments and individual officials to damages liability under § 1983. See discussion infra Part III (discussing the § 1983 context of Garcetti).
Interestingly, in the Commerce Clause setting, the Supreme Court is explicitly willing to live with uncertainty in another areaâ€”the uncertainty of determining whether a particular activity sought to be regulated by Congress is economic or commercial in nature. See United States v. Lopez, 514 U.S. 549, 566 (1995) (“Admittedly, a determination whether an intrastate activity is commercial or noncommercial may in some cases result in legal uncertainty. But, so long as Congress’ authority is limited to those powers enumerated in the Constitution, and so long as those enumerated powers are interpreted as having judicially enforceable outer limits, congressional legislation under the Commerce Clause always will engender “â€˜legal uncertainty.’”). For a discussion of Garcetti from a different but not necessarily inconsistent perspective, see Charles W. “Rocky” Rhodes, Public Employee Speech Rights Fall Prey to an Emerging Doctrinal Formalism, 15 Wm. & Mary Bill Rts. J. 1173 (2007) (arguing that Garcetti reflects a formalist turn in the Court’s approach to constitutional interpretation).
. Garcetti, 126 S. Ct. at 1961. Strictly speaking, the Court’s mention of separation of powers appears to be misplaced because co-equal branches of government are not involved. The Court is probably referring to federal judicial intervention in executive decision-making generally. Garcetti, because it deals with federal court intervention in local personnel decision-making, more obviously implicates federalism issues.
. In First Nat’l Bank of Boston v. Bellotti, 435 U.S. 765 (1978), the Court invalidated on First Amendment grounds a statute prohibiting corporations from making contributions or expenditures for the purpose of influencing voters on questions submitted to them, unless the question materially affected the property, business, or assets of the corporation. The Court emphasized the public importance of the speech at issue, and rejected the argument that the corporate status of the speaker should be determinative of the First Amendment issue. See id. at 784, 788â€“89, 792. Thus, there appears to be some tension between the reasoning of Bellotti and Garcetti.
. Garcetti‘s First Amendment employer-immunity rule cannot be characterized as content-neutral in its application. Once the content of a public employee’s speech and a public employer’s imposition of discipline for that speech are focused on in a particular caseâ€”as required by Garcetti itselfâ€”then the challenged employer discipline is content-based, and most likely viewpoint-based as well, and hence should be subject to strict scrutiny. See R.A.V. v. City of St. Paul, 505 U.S. 377, 395 (1992). In addition, Garcetti‘s First Amendment employer-immunity rule cannot plausibly be defended as protecting public employers from the secondary effects of public employee job-required speech. See generally City of Renton v. Playtime Theatres, Inc., 475 U.S. 41 (1986) (discussing the First Amendment implications of limiting speech for the purpose of removing its secondary effects). For one thing, the secondary effects doctrine is “something of a [First Amendment] fiction.” City of Los Angeles v. Alameda Books, Inc., 535 U.S. 425, 448 (2002) (Kennedy, J., concurring). For another, even though the Court in Alameda Books upheld the constitutionality of a city ordinance prohibiting more than one adult entertainment business in the same building, Justice Souter, joined by Justices Stevens, Ginsburg, and Breyer, dissented and described the content-based regulation before the Court as “content correlated” rather than content-neutral. Id. at 457 (Souter, J., dissenting). Thus, five Justices appear to have rejected the secondary effects doctrine, a doctrine that is questionable as a matter of First Amendment jurisprudence, and should, in any event, be confined to zoning of adult entertainment businesses.
. Garcetti, 126 S. Ct. at 1966 (Souter, J., dissenting) (“[T]he ranks of public service include those who share the poet’s â€˜object . . . to unite [m]y avocation and my vocation.’”) (quoting Robert Frost, Two Tramps in Mud Time, Collected Poems, Prose and Plays 251â€“52 (Richard Poiriel & Mark Richardson eds., 1995)).
. In his dissent, Justice Stevens observes: “The notion that there is a categorical difference between speaking as a citizen and speaking in the course of one’s employment is quite wrong.” Garcetti, 126 S. Ct. at 1963 (Stevens, J., dissenting).
. See Michael Walzer, Obligations: Essays on Disobedience, War, and Citizenship 10 (1970) (“The duty to disobey . . . arises when obligations incurred in some small group come into conflict with obligations incurred in a larger, more inclusive group, generally the state.”).
. Recall the Supreme Court’s discussion of uncertain political accountability regarding Congressional attempts to commandeer a state’s legislative or executive branches for federal purposes. See Printz v. United States, 521 U.S. 898, 935 (1997) (invalidating Congressional commandeering of a state’s executive branch); New York v. United States, 505 U.S. 144, 149 (1992) (invalidating Congressional commandeering of a state’s legislative branch). These cases clearly implicate the citizenship status of Americans as at least dual: federal and state.
. See Alexander Meiklejohn, Free Speech and its Relation to Self-Government (2000); Alexander Meiklejohn, The First Amendment is an Absolute, 1961 Sup. Ct. Rev. 245 (1961). Other rationales, however, such as the marketplace of ideas, self-fulfillment, and autonomy, can ground the First Amendment. Indeed, I argued long ago that no one rationale satisfactorily explains the Supreme Court’s First Amendment jurisprudence. See generally Sheldon H. Nahmod, Artistic Expression and Aesthetic Theory: The Beautiful, the Sublime and the First Amendment, 1987 Wis. L. Rev. 221 (1987).
. See, for example, Mills v. City of Evansville, 452 F.3d 646 (7th Cir. 2006), in which a police sergeant responsible for supervising crime prevention officers criticized a plan to reduce the number of officers under her command while at a meeting with her superiors. Id. at 647. She asserted that the plan would not work and that community organizations would not let the changes occur. Id. She claimed she was retaliated against as a result of her criticism. Id. The Seventh Circuit applied Garcetti and found that the police sergeant spoke as an employee and not as a citizen. Id. at 648. Thus, she lost on her First Amendment retaliation claim. See id. See also Battle v. Bd. of Regents, 468 F.3d 755 (11th Cir. 2006), in which the Eleventh Circuit applied Garcetti against an employee in a state’s financial aid and veterans affairs office who spoke to university officials about inaccuracies and evidence of fraud in student files submitted in connection with a federal work study program. Id. at 757â€“58. She alleged that she was retaliated against for her speech, id. at 759, but the court determined, and the plaintiff admitted, that she had a clear employment duty to ensure the accuracy of student files and to report any mismanagement or fraud to her superiors. Id. at 761. Hence, she spoke as an employee and not as a citizen. See id.
. See Robert C. Post, Between Governance and Management: The History and Theory of the Public Forum, 34 UCLA L. Rev. 1713 (1987) (arguing with respect to public forum analysis that different kinds of public property should be distinguished not on the basis of its character but rather on the nature of the government authority in question).
. See Toni M. Massaro, Significant Silences: Freedom of Speech in the Public Sector Workplace, 61 S. Cal. L. Rev. 1 (1987) (arguing that hierarchical organizational theory is inconsistent with democratic values of self-government and should not be deferred to by courts in public employee free speech cases). Unlike Massaro, I acknowledge the importance of the interest in allowing government to choose a particular organizational theory for the public sector workplace. Like Massaro, I contend that this interest should not invariably trump the First Amendment.
. In connection with the Garcetti Court’s disjunctive approach to citizenship, recall that the Court previously extended the First Amendment protection of Pickering-Connick to independent contractors in Board of County Commissioners v. Umbehr, 518 U.S. 668, 678 (1996). There, the Court reasoned that for First Amendment purposes independent contractors stand somewhere between public employees and other groups, such as claimants for tax exemptions, users of public facilities, and recipients of small government subsidies, all of whose speech is entitled to First Amendment protection under the unconstitutional conditions doctrine. Id. at 680. After Garcetti, it is likely that independent contractors, by analogy to public employees, will receive no First Amendment protection for speech required by their contracts because such speech is not citizen-speech. On the other hand, even after Garcetti, they should be entitled to First Amendment protection when they go public with their speech.
. Curtis Publ’g Co. v. Butts, 388 U.S. 130 (1967). Chief Justice Warren, joined by Justices Brennan and White, reasoned that many individuals “who do not hold public office at the moment are nevertheless intimately involved in the resolution of important public questions or, by reason of their fame, shape events in areas of concern to society at large.” Id. at 164 (Warren, C.J., concurring).
. Justices Douglas and Brennan argued in dissent that the knowing-or-reckless-falsehood standard should apply where a matter of public concern is at issue. Gertz, 418 at 355â€“60 (Douglas, J., dissenting), 361â€“69 (Brennan, J., dissenting).
. See, e.g., Bartnicki v. Vopper, 532 U.S. 514 (2001) (broadcasting of intercepted private telephone conversation); Landmark Commc’n, Inc. v. Virginia, 435 U.S. 829 (1978) (publishing of confidential information); Cox Broad. Corp. v. Cohn, 420 U.S. 469 (1975) (broadcasting of private information that is newsworthy). It should also not be forgotten that, even after Garcetti, ad hoc balancing under Pickering-Connick remains alive (at least for the time being) when a public employee goes public with speech that is arguably of public concern.
. See infra notes 150â€“54 and accompanying text (arguing that this concern is adequately addressed by § 1983 doctrines of causation, local government liability, and especially qualified immunity).
. There are, of course, always risks to public employees in so doing, including the possibility that there could be more severe discipline imposed if the speech turns out not to be protected by Pickering-Connick. Consider also that if a public employee reports official misconduct both internally and publicly and is thereafter disciplined, the public employer may be able to avoid liability under the First Amendment and § 1983 by accomplishing the Mt. Healthy burden-shift; namely, that the employee would have been disciplined anyway for the unprotected job-required internal speech, regardless of the publicly communicated and protected speech. See infra notes 140â€“42 and accompanying text. Indeed, this could turn out to be the case for Ceballos on remand because he spoke publicly about the alleged police officer misconduct at a local bar association meeting while his internal grievance was pending. See Garcetti, 126 S. Ct. at 1972. A public employee may be able to avoid the Mt. Healthy burden-shift by going public initially. Garcetti may have made this option particularly attractive. On the other hand, the public employee could be discharged for not following required internal procedures.
. This potential broadening of job descriptions could turn out to be a particularly serious problem with regard to academic freedom in public educational institutions of higher learning, particularly since many such institutions require their professors to publish. See infra notes 125â€“32 and accompanying text; see also Garcetti, 126 S. Ct. at 1965 n.2.
. See, e.g., Sigsworth v. City of Aurora, 487 F.3d 506 (7th Cir. 2007). In Sigsworth, a police officer representing his agency on a multi-jurisdictional task force formed to investigate gang activity reported to his supervisors what he believed to be misconductâ€”other task force members allegedly tipping off gang members involved in drug activity. Id. at 507. Ruling that the officer was not protected by the First Amendment from employer discipline and relying on Garcetti, the Seventh Circuit stated that the officer “was merely doing what was expected of him as a member of the task force charged with organizing and overseeing the planning and execution of the arrest warrants.” Id. at 511.
. For example, prosecutors often speak to the press (and therefore the public) about criminal cases. Press secretaries are hired to do just that. And other high-ranking executive officials often report to legislative committees and the press about their behavior and the behavior of their subordinates.
. The literature on government speech is growing larger. See, e.g., Mark G. Yudof, When Government Speaks (1983); David Fagundes, State Actors as First Amendment Speakers, 100 Nw. U. L. Rev. 1637 (2006); Randall P. Bezanson & William G. Buss, The Many Faces of Government Speech, 86 Iowa L. Rev. 1377 (2001); Martin H. Redish & Daryl I. Kessler, Government Subsidies and Free Expression, 80 Minn. L. Rev. 543 (1996); Steven Shiffrin, Government Speech, 27 UCLA L. Rev. 565 (1980); William W. Van Alstyne, The First Amendment and the Suppression of Warmongering Propoganda in the United States: Comments and Footnotes, 31 Law & Contemp. Probs. 530 (1966).
. See Hughes v. Alexandria Scrap Corp., 426 U.S. 794, 809â€“10 (1976) (holding the dormant Commerce Clause does not prohibit a state from participating in economic markets and favoring its own citizens in those markets).
. E.g., Johanns v. Livestock Mktg. Ass’n, 544 U.S. 550 (2005) (finding no First Amendment violation even though viewpoint based); United States v. Am. Library Ass’n, Inc., 539 U.S. 194 (2003) (finding no First Amendment violation even though viewpoint based); Nat’l Endowment for the Arts v. Finley, 524 U.S. 569 (1998) (finding no First Amendment violation even though potentially viewpoint based); Rust v. Sullivan, 500 U.S. 173 (1991) (finding no First Amendment violation where viewpoint discriminatory conditions placed on government subsidies); cf. Legal Servs. Corp. v. Velazquez, 531 U.S. 533 (2001) (holding speech is not government speech and First Amendment therefore applies to prohibit viewpoint based government discrimination); Rosenberger, 515 U.S. 819 (holding speech is not government speech and First Amendment therefore applies to prohibit viewpoint based government discrimination).
. Id. at 1968â€“70. Although too much should not be made of this, perhaps it was Justice Souter’s reaction to the Court’s use of Rust and Rosenberger that led Justice Kennedy to use only a “cf.” rather than a “see.” See id. at 1960 (majority opinion).
. Like government speech, an extensive discussion of academic freedom is well beyond the scope of this article. For more comprehensive pre-Garcetti treatment see, for example, Lee C. Bollinger, Benjamin N. Cardozo Lecture: Academic Freedom and the Scholarly Temperament, 60 Rec. Ass’n B. City of N.Y. 326 (2005); J. Peter Byrne, Academic Freedom: A “Special Concern of the First Amendment”, 99 Yale L.J. 251 (1989); Matthew W. Finkin, On “Institutional” Academic Freedom, 61 Tex. L. Rev. 817 (1983); Walter P. Metzger, Profession and Constitution: Two Definitions of Academic Freedom in America, 66 Tex. L. Rev. 1265 (1988); William W. Van Alstyne, Academic Freedom and the First Amendment in the Supreme Court of the United States: An Unhurried Historical Review, 53 Law & Contemp. Probs. 79 (1990).
. After Garcetti, teacher writing outside of the classroom may also raise difficult First Amendment issues, particularly since institutions of higher education often require their professors to publish.
. See, e.g., Grutter v. Bollinger, 539 U.S. 306, 329 (2003) (noting the importance of “expansive freedoms of speech and thought associated with the university environment”); Keyishian v. Bd. of Regents, 385 U.S. 589 (1967) (academic freedom is a special concern of the First Amendment); Sweezy v. New Hampshire, 354 U.S. 234 (1957) (contents of a professor’s lecture at a state university). Justice Souter cited each of these cases in his Garcetti dissent. 126 S. Ct. at 1969â€“70 (Souter, J., dissenting). Interestingly, Justice Douglas’s opinion for the Court in Griswold v. Connecticut also refers to a First Amendment grounding for academic freedom in the university. 381 U.S. 479, 482 (1965).
Fred Schauer suggests that individual First Amendment-based academic freedom is not really established by Supreme Court precedent, even though there may be normative arguments in its favor. Frederick Schauer, Is There a Right to Academic Freedom?, 77 U. Colo. L. Rev. 907, 908â€“10, 913â€“19 (2006). He goes on to discuss academic freedom as institutional autonomy. Id. at 919â€“26.
. See, for example, Mayer v. Monroe County Cmty. Sch. Corp., in which the Seventh Circuit, post-Garcetti, ruled against a social studies teacher who was disciplined for using a classroom session on current events to take a position on an anti-Iraq War demonstration. 474 F.3d 477, 478, 480 (7th Cir. 2007). The Court stated, “the first amendment does not entitle primary and secondary teachers, when conducting the education of captive audiences, to cover topics, or advocate viewpoints, that depart from the curriculum adopted by the school system.” Id. at 480.
. See, for example, Piggee v. Carl Sandburg Coll., a post-Garcetti case involving student complaints about a cosmetology instructor in a public college’s cosmetology program who, during a clinical segment held in the college’s beauty salon, distributed anti-gay religious pamphlets to students she thought were homosexual and attempted to lecture them to change their sexual orientation. 464 F.3d 667, 668 (7th Cir. 2006). The college thereafter did not renew her contract. Id. at 668â€“69. Ruling against the teacher, the Seventh Circuit emphasized that even though the teacher spoke on an issue of public concern, both the college’s interest in requiring instructors to “stay on message” and the disruptive effect on the students outweighed her First Amendment interest. Id. at 671â€“74.
. Michael Wells maintains that the ad hoc Pickering-Connick test, especially when coupled with the hurdles for plaintiffs imposed by the technical requirements of § 1983, under-protects First Amendment rights because of a lack of predictability. See Michael L. Wells, Section 1983, The First Amendment, and Public Employee Speech: Shaping the Right to Fit the Remedy (and Vice Versa), 35 Ga. L. Rev. 939 (2001). He argues that the Court should set out rules for First Amendment retaliation claims as well as modify § 1983 doctrine regarding causation and damages so as to promote First Amendment values more effectively. See id. I disagree with his position that the Pickering-Connick test under-protects the First Amendment rights of public employees, but I agree that § 1983 doctrines place substantial hurdles before plaintiffs. I take no position in this article on Wells’s argument regarding the need for rules for First Amendment retaliation claims in general, although I do agree with his criticism of the particular rule the Court adopted in Garcetti as “draconian.” Id. at 991.
. See id. at 381, 390 (holding that a deputy constable’s dismissal for saying after hearing that President Reagan had been shot, “I hope they get [The President],” was contrary to the deputy constable’s First Amendment rights).
. Perhaps this is what Justice Souter has in mind in his dissent when he suggests that, on remand, the lower courts should consider the First Amendment status of other speech that Ceballos engaged in. See Garcetti, 126 S. Ct. at 1973 (Souter, J., dissenting). If Ceballos’s other speech is protected by the First Amendment and was the sole motive for the alleged adverse employment actions taken against him, then he will prevail (leaving aside qualified immunity). On the other hand, if the motives for the adverse employment actions taken against Ceballos included both his protected speech and his unprotected job-required speech, then the defendants will win if they show that he would have been so disciplined regardless of his protected speech.
. The resulting consequence is that the plaintiff is entitled neither to damages nor reinstatement and back pay. The argument has been made that so long as a plaintiff shows an impermissible motive, this constitutes a First Amendment violation, despite the fact that a defendant carries the Mt. Healthy burden-shift because that burden-shift goes to remedy and not liability. Under this approach, a plaintiff could recover at least nominal damages, possible punitive damages, and attorney’s fees under 42 U.S.C. § 1988, even if the latter were reduced because of lack of complete success. See 1 Nahmod, Civil Rights and Civil Liberties Litigation § 4:10; Sheldon Nahmod, Mt. Healthy and Causation-in-Fact: The Court Still Doesn’t Get It!, 51 Mercer L. Rev. 603 (2000). However, the Court emphatically rejected this argument in Texas v. Lesage, 528 U.S. 18 (1999) (per curiam).
. See, e.g., Owen v. City of Independence, 445 U.S. 622 (1980) (local governments have no absolute or qualified immunity from § 1983 liability); cf. City of Newport v. Fact Concerts, Inc., 453 U.S. 247 (1981) (local governments are absolutely immune from punitive damages liability).
. Local government liability can be based on the formal acts of local governments, on their customs, on the acts of high-ranking officials who make policy for local governments, and on failures to train. Characterizing the requirements for local government liability as arcane is actually an understatement. See generally 2 Nahmod, Civil Rights And Civil Liberties Litigation § § 6.1â€“6.71. See also Bd. of County Comm’rs v. Brown, 520 U.S. 397, 435â€“37 (1997) (Breyer, J., dissenting), in which Justice Breyer, joined by Justices Stevens and Ginsburg, argued in dissent that the § 1983 local government liability doctrine had become so complicated that it was time to reexamine the Court’s rejection of respondeat superior liability in Monell v. Department of Social Services of New York, 436 U.S. 658 (1978).
. Absolute immunity from damages liability protects state and local legislators for their legislative acts, judges for their judicial acts, and prosecutors for their acts as advocates. See Bogan v. Scott-Harris, 523 U.S. 44 (1998) (local legislators); Stump v. Sparkman, 435 U.S. 349 (1978) (judges); Imbler v. Pachtman, 424 U.S. 409 (1976) (prosecutors); Tenney v. Brandhove, 341 U.S. 367 (1951) (state legislators). It does not, however, protect them for their administrative acts, including decisions to fire or otherwise impose employer discipline. See Forrester v. White, 484 U.S. 219 (1988) (adopting functional approach to individual immunities); see generally 2 Nahmod, Civil Rights and Civil Liberties Litigation § § 7:1â€“7:65.
. See Saucier v. Katz, 533 U.S. 194 (2001); Anderson v. Creighton, 483 U.S. 635 (1987) (requiring some fact specificity); Harlow v. Fitzgerald, 457 U.S. 800 (1982) (setting out qualified immunity test); see generally 2 Nahmod, Civil Rights and Civil Liberties Litigation § § 8:1â€“8:99.
. See, e.g., Martin v. Baugh, 141 F.3d 1417 (11th Cir. 1998); Bartlett v. Fisher, 972 F.2d 911 (8th Cir. 1992); Melton v. City of Oklahoma City, 879 F.2d 706 (10th Cir. 1989); Benson v. Allphin, 786 F.2d 268 (7th Cir. 1986). But see, e.g., Roth v. Veteran’s Admin., 856 F.2d 1401, 1408 (9th Cir. 1988) (refusing to adopt this approach to qualified immunity in balancing cases because the “rule . . . would effectively eviscerate whistleblower protection for public employees”). See generally 2 Nahmod, Civil Rights and Civil Liberties Litigation § 8:19 (collecting and discussing cases).
. The Court’s move to categorical balancing in Garcetti is analogous to its conversion of qualified immunity into a kind of absolute immunity in the last two decades because of the Court’s concern not only with the costs of liabilityâ€”previously a qualified immunity policy considerationâ€”but also with the costs of defending, which was previously an absolute immunity policy consideration. As a result, qualified immunity has become a much more defendant-protective affirmative defense under which discovery is ordinarily deferred until a defendant’s qualified immunity motion to dismiss or for summary judgment is decided. See Harlow v. Fitzgerald, 457 U.S. 800 (1982) (a defendant is not limited to one such qualified immunity motion); see also Behrens v. Pelletier, 516 U.S. 299 (1996) (the denial of such a defense motion is immediately appealable as to the legal issues raised); Mitchell v. Forsyth, 472 U.S. 511 (1985). See generally 2 Nahmod, Civil Rights and Civil Liberties Litigation § § 8:1â€“8:99.
. In this connection it is interesting to compare Garcetti with Hartman v. Moore, 547 U.S. 250 (2006), a First Amendment Bivens damages case decided in the same term as Garcetti. The question in Hartman, answered affirmatively by the Court, was whether a plaintiff who alleges a retaliatory prosecution in violation of the First Amendment must further allege and prove the absence of probable cause by a preponderance of the evidence. See Hartman, 547 U.S. at 252. In contrast to Garcetti, which modifies First Amendment doctrine in part because of § 1983 damages liability and related concerns, Hartman does not modify First Amendment doctrine because of Bivens damages liability concerns. See id. at 258â€“59. Instead, the Hartman Court, as a matter of federal common law interpretation, imposed this additional burden on plaintiffs alleging retaliatory prosecutions on the ground that these cases are not ordinary retaliation cases because there is a presumption of prosecutorial regularity that plaintiffs should overcome. See id. at 263. Whatever one thinks of the outcome in Hartman, at least the case correctly does not conflate the federal common law and constitutional issues.
. Of course, other considerations may also be in play, including federalism, because the absence of a federal constitutional violation, and hence the unavailability of a § 1983 damages remedy, means that a remedy must be sought in state court under state law. See generally Parratt v. Taylor, 451 U.S. 527 (1981), overruled in part by Daniels v. Williams, 474 U.S. 327 (1986). In Parratt, the Court addressed a prisoner’s procedural due process damages claim against prison officials arising out of the loss of his hobby kit materials. See 451 U.S. at 529. The Court held that although negligent conduct could serve as the basis for a procedural due process violation, the plaintiff lost on the merits because he alleged random and unauthorized conduct and there was an adequate state post-deprivation remedy for such conduct. See id. at 535â€“37, 543â€“44. Thereafter, the Court overruled Parratt in part in Daniels and ruled that negligence is not enough for any due process violation, whether procedural or substantive. See Daniels, 474 U.S. at 330â€“31.
For extensive theoretical discussions of the relation between constitutional rights and remedies, see generally John C. Jeffries, Jr., Disaggregating Constitutional Torts, 110 Yale L.J. 259 (2000); John C. Jeffries, Jr., The Right-Remedy Gap in Constitutional Law, 109 Yale L.J. 87 (1999); Daryl J. Levinson, Making Governments Pay: Markets, Politics and the Allocation of Constitutional Costs, 67 U. Chi. L. Rev. 345 (2000); Daryl J. Levinson, Rights Essentialism and Remedial Equilibration, 99 Colum. L. Rev. 857 (1999).
. See Daniels, 474 U.S. at 330â€“31. The Court, pointing out that negligence is too low a threshold as a matter of constitutional interpretation, emphasized that the Due Process Clause is intended to prevent abuses of government power. See id. at 331â€“32.
. See Whitley v. Albers, 475 U.S. 312, 320â€“21 (1986) (underscoring “the appropriate hesitancy to critique in hindsight decisions necessarily made in haste, under pressure, and frequently without the luxury of a second chance”).
. See Vincent Blasi, Free Speech and Good Character: From Milton to Brandeis to the Present, in Eternally Vigilant: Free Speech in the Modern Era 61, 62 (Lee C. Bollinger & Geoffrey R. Stone eds., 2002) (“[A] culture that prizes and protects expressive liberty nurtures in its members certain character traits such as inquisitiveness, distrust of authority, willingness to take initiative, and the courage to confront evil. Such character traits are valuable . . . for their instrumental contribution to the collective well-being, social as well as political.”). See also Justice Brandeis’s concurrence in Whitney v. California, 274 U.S. 357, 375, 377 (1927):
Those who won our independence believed . . . that freedom to think as you will and to speak as you think are means indispensable to the discovery and spread of political truth . . . that the greatest menace to freedom is an inert people; that public discussion is a political duty; and that this should be a fundamental principle of the American government.
Those who won our independence by revolution were not cowards. They did not fear political change. They did not exalt order at the cost of liberty.