Two justices want to revisit McDonnell Douglas in Title VII cases

Two justices on the U.S. Supreme Court want to alter the 52-year-old framework for resolving employment discrimination claims, arguing that the so-called McDonnell Douglas model is unworkable and confusing.

The case is Hittle v. City of Stockton. On March 10, Justices Thomas and Gorsuch issued a decision dissenting from the Court's decision not to hear the case. Having lost their case in the lower courts, Hittle's lawyers filed a petition for Supreme Court review. But even the other conservative justices on the Court evidently do not want to revisit McDonnell Douglas. Still, the dissent provides insight into how two of the Justices approach this issue. When even one Justice wants to revisit or overrule a major precedent, it's important news.

Every employment discrimination lawyer knows about McDonnell Douglas v. Green, 411 U.S. 792 (1973). In cases involving circumstantial evidence, the plaintiff has to make out a minimal prima facie case, usually through proof they were qualified for the position and were terminated or demoted under circumstances creating an inference of discrimination, i.e., a Black employee was replaced by a white employee, or there was something fishy about the adverse action. If the prima facie case is made out, the employer has to articulate a reason for termination or demotion (or other adverse action). Employers almost always satisfy that minimal requirement, as well. The burden then shifts to the plaintiff to prove the real reason was discrimination, usually proven through evidence that the employer's articulated justification was a knowing falsehood. The pretext fight is usually where the action is.

Thomas and Gorsuch write that this framework "has been widely criticized" over the years as confusing and inefficient, having "taken on a life of its own" as it was originally designed for bench trials but now guides summary judgment motions and jury trials. They note that the Supreme Court devised this evidentiary framework "out of whole cloth," though to be fair to the McDonnell Douglas model, most Supreme Court doctrine is devised by the justices to help apply statutory and constitutional principles. Title VII, when enacted in 1964, said nothing about how courts and litigants were supposed to apply it in the real world. Maybe the thinking back then was that discrimination was often so blatant that you did not need a complex evidentiary model to determine if someone was really the victim of discrimination. By 1973, however, employers got wise and knew that direct evidence of discriminatory intent would doom their chances in court, and that more subtle means of discrimination were in order. McDonnell Douglas takes the latter approach into account.

Here are the problems with McDonnell Douglas, as Thomas and Gorsuch see it:

Because the McDonnell Douglas framework was designed for use in a bench trial, the language this Court has used to describe the framework does not neatly track the plaintiff ’s summary-judgment task. or example, the first step requires a plaintiff to “establis[h] a prima facie case.” And, the third step requires the plaintiff to show, “by a preponderance of the evidence,”  that the employer’s stated reason “was in fact pretext” for discrimination. But, a plaintiff need not establish or prove any elements—by a preponderance or otherwise—to survive summary judgment. At that stage, he need only offer enough evidence to create a genuine dispute of material fact.

In addition, 

Our precedent makes clear that the framework is, at most, a “procedural device, designed only to establish an order of proof and production” when evaluating circumstantial evidence. Put another way, it is “‘merely’” a “‘way to evaluate the evidence’” that bears on the ultimate finding of liability. Yet, some courts treat McDonnell Douglas as a substantive legal standard that a plaintiff must establish to survive summary judgment or to ultimately prove a claim.

There is more:

Some courts also fail to appreciate that McDonnell Douglas is necessarily underinclusive. The framework sets forth criteria that, if satisfied, will allow a plaintiff to prove a Title VII violation. But, satisfying McDonnell Douglas is not the only way or even the best way to prove a claim. The McDonnell Douglas Court itself explained that “[t]he facts necessarily will vary in Title VII cases” and the prima facie proof required in one case “is not necessarily applicable in every respect to differing factual situations.” Yet, some courts treat the McDonnell Douglas framework as the exclusive method for evaluating evidence at summary judgment.

Thomas and Gorsuch rely heavily on a Ninth Circuit concurring opinion from December 2023 that questions the McDonnell Douglas framework. Concurrences like this will often find their way into Supreme Court rulings and dissents, and the judges write them as a message to the Supreme Court and other circuit courts to reevaluate some legal principle. 

Another issue that Thomas does not reference, which I have observed, is that the prima facie portion of the three-part framework is "not onerous" and "de minimus" under settled case law. But courts do not so instruct the jury, which may rule against the plaintiff for not having a preponderance of the evidence in support of one of the prima facie elements even though the cases hold the plaintiff to a lower standard. Relatedly, the Second Circuit has said that you should not charge the jury with the prima facie elements, but that rule is not settled; some cases say you have to do it. Yet another quandary is this: the circuit courts have not settled upon how to resolve when pretext alone is enough to prove discrimination. Some circuits, including the Second Circuit want more than just pretext; we call that pretext-plus. Other circuits say evidence of pretext alone is enough to support a jury verdict.

In the 1990's, some federal judges began to question to McDonnell Douglas framework, abandoning the prima facie analysis and turning directly to examining the evidence as a whole to see if there is enough evidence to support the plaintiff's case. That is how juries do it; trial judges charge them on pretext, but I wonder if jurors go beyond the step-by-step instructions and just ask themselves if there is enough to prove discrimination. I have not seen trial judges lately abandon the prima facie test.

I don't know what alternatives the Court may adopt in place of McDonnell Douglas. While the ultimate question for the plaintiff is whether the defendant discriminated against her, the jury probably needs some kind of framework to facilitate its deliberations. For one, thing resolving cases involving circumstantial evidence are not easy and forcing the employer to articulate a reason, and allowing the plaintiff an opportunity to prove that articulated reason is a knowing falsehood certainly focuses the jury's attention on the main issues in the case. Juries may rationally find that if the employer's reason for the plaintiff's termination is a knowing falsehood, that deception may very well prove discrimination. Why else would the employer lie in court?

If the Supreme Court abandons the McDonnell Douglas framework, that will be a major event in the lives of employment discrimination attorneys, sort of like when the Court revised the employer liability standards in sexual harassment cases in 1998, when it issued the the Faragher and Ellerth rulings that upended the case law through the circuit courts. That only two Justices want to revisit McDonnell Douglas means the framework is safe for now. Both Thomas and Gorsuch are outside-the-box thinkers. But Supreme Court case law has been changing very quickly over the last 10 years. Litigants will continue to ask the Court to review this issue, I am sure.

Employer cannot publicly attack plaintiff who won her prior lawsuit

The Court of Appeals holds that an employer, in the course of a long-running lawsuit, must be enjoined from further publishing identifying information about a wage-and-hour plaintiff after she won her case. Publishing such information is unlawful retaliation.

The case is Williams v. Harry's Nurses Registry, Inc., a summary order published on March 18. After plaintiff won her initial FLSA case, the defendant repeatedly challenged the judgment by, among other things, trying to reopen the case, taking up a frivolous appeal, and trying to get the Judicial Panel on Multidistrict Litigation to transfer the case to a federal court in Missouri. The employer also attacked the plaintiff, her lawyer, and the judge online, calling the plaintiff a scammer and a criminal. Defendant also posted plaintiff's date of birth and Social Security number online. The Court refers to this as personal identifying information, or PII. Hence, this retaliation lawsuit.

Retaliation may include adverse actions that post-date the plaintiff's employment. The district court ordered the defendant to remove the identifying information. The Court of Appeals finds the posting of such information was a retaliatory act, as “An employment action disadvantages an employee if it well might have dissuaded a reasonable worker from making or supporting similar charges.” That is black-letter retaliation law under the FLSA and other employment laws. While the employer said he did not post plaintiff's identifying information, plaintiff was able to prove otherwise in the district court.

The trial court also found, and the Second Circuit (Bianco, Park and Nardini) agrees, that defendant intended to retaliate against plaintiff. "The sequence, timing and nature of events surrounding the publication of the information reinforces that its predominant purpose was to get back at Williams. More specifically, the district court found that 'sufficient evidence based on defendant[s’] own words in the postings, [their] rantings about having lost at the district court and Court of Appeals level, and [their] railing against all of the participants in that process, even the judges . . . makes it so clear that . . . at least the primary purpose was to get back at everybody involved in this,' including Williams." While defendant argued that he posted this information to prevent other people from committing identity fraud to obtain employment, the court disagrees and finds :there is no reason [defendant] has to use this forum in this way to police who applies to [defendant] and whether they provide [defendant] with truthful information," as the employer can privately verity the employment information it receives.

Plaintiff also has to prove irreparable harm to win the injunction. She can do so, and therefore recover relief right now and not at the end of the case, the Court of Appeals holds, because "so long as Williams’s PII remains publicly displayed on HNR’s website, she suffers  an ongoing harm that cannot be adequately remedied by a retrospective award of monetary damages, especially when it exposes Williams to 'identity theft and misuse of personal information.'”

There is now way any court is going to put up with this kind of retaliation against a plaintiff who won her lawsuit against a prior employer. For this reason, the Court holds, the balance of hardships and the public interest favor plaintiff's position and not defendant's position. These factors are also relevant to the preliminary injunction analysis. As the trial court held, "the district court found that HNR has no 'real or genuine or . . . non-delusional reason' to post Williams’s PII on its website."

Complex search and seizure case will go to trial in Syracuse

This is a police misconduct case against the Syracuse Police Department. The police raided the plaintiff's house and searched the place up and down, all without a warrant. Plaintiff was ultimately arrested. So we have a claim for unlawful search and seizure, false arrest, and malicious prosecution under the Fourteenth Amendment. The district court granted summary judgment on all claims, but the Court of Appeals reverses on most of the claims, which will go to trial.

The case is Alexander v. City of Syracuse, issued on March 14. This is a complex case: the ruling is 85 pages long, and the case was argued in October 2023. The facts are ugly. According the decision, a young woman suffered a brutal sexual assault at the plaintiff's house the night before. The police barged into the house and stayed there for 12.5 hours. Then after later getting a warrant, the police returned and found drugs in the house, prompting plaintiff's arrest. 

Plaintiff has a case because the police had a chance to obtain a warrant before they entered the house but did not do so. The Court of Appeals (Lee and Robinson, with Judge Newman in dissent) holds there were no exigent circumstances that permit police entry into the house without obtaining a warrant. There may have been a brutal sexual assault the night before, but that does not mean violent activity was taking place when the police decided to enter the house, or that someone was in the process of destroying evidence. The warrants rules, especially under settled law, hold police to a strict framework when they want to enter your house with a warrant. Under the law, a man's home is his castle. The Second Circuit says the police never explained why they did not get a warrant that night.

We have an interesting dissent. Judge Newman views the case from an entirely different angle, stating that plaintiff was operating a prostitution business from his house and sold narcotics, and a 19 year-old was brutally raped the night before the police showed up. Plaintiff was apparently her pimp. You do not want to read the details of what happened to this women, but the dissent's point is that the police reasonably believed they had to enter the house right away after what had happened to this woman. Here is how Judge Newman wraps up his dissent: 

Alexander’s house was a den of prostitution, a base for selling narcotics, and the scene of a brutal rape committed against a teenager, for whom Alexander was the pimp, when he was in the living room while the rape was being committed in the basement. The prospect that Alexander will have an opportunity to require a conscientious police detective to pay him money damages is a bizarre result that even Kafka could not have imagined.

Dissents like this -- decrying the possibility that an undeserving plaintiff may recover money from a public defendant -- are rare, but they are not uncommon. I recall a decision from 1999 when another Second Circuit judge bemoaned how a Section 1983 plaintiff who suffered little damages was able to force a municipality to a costly trial. That case was Amato v. City of Saratoga Springs. 

The concurrence in Amato was written by a conservative judge, but a liberal judge wrote the dissent in Alexander. These judges are not always predictable. What I draw from the dissent is that there is no official way to resolve any case, particularly one brought under the Constitution, where the legal standards are judge-made and therefore draw from case law, which may have different interpretations. Think about it: a different judicial panel could have ruled for the City in this appeal. Even a liberal judge who might ordinarily sympathize with the plaintiff will rule against the plaintiff if they feel strongly about it. 

Inmate wins First Amendment retaliation appeal

You'd be surprised how often inmates win their appeals in the Second Circuit. This case alleges that correction officers threatened and then assaulted the plaintiff after they learned he wanted to file a grievance challenging prison conditions. The case was dismissed under Rule 12, and other portions of the case were dismissed on summary judgment, but the Court of Appeals finds plaintiff has a case, in part.

The case is Walker v Senecal, issued on March 6. Senecal is the CO. According to plaintiff, he ripped up plaintiff's grievance and said plaintiff would wind up "dead" or "in the box" if plaintiff pursued the grievance. The box is the special housing unit and, really, the last place any inmate wants to find himself. The next day, two other CO's slapped plaintiff around and roughed him up and asked plaintiff "if you saw how easily he could get killed for filing grievances against Officer Senecal" and said that any such grievance would land plaintiff in the box. Plaintiff also alleged that Senecal recruited another CO, Benware to retaliate against plaintiff by filing a false misbehavior report against him and firing plaintiff from his job in the prison law library.

Here are the holdings:

1. Plaintiff cannot show a connection between his First Amendment speech (filing a grievance) and Benware's actions. It is not enough for plaintiff to note that these two CO's hung out in the same area at the jail, and plaintiff's other attempts to prove this connection are conclusory.

2. But plaintiff has a case against Senecal, The Court of Appeals (Livingston, Jacobs and Menashi) rejects the trial court's finding that destroying plaintiff's grievance and threatening to throw him into the box were not adverse actions. They were. What also factors into this holding are the officers who slapped and threatened plaintiff over the grievance, making reference to Senecal in doing so. (There was also a retaliatory pat frisk against plaintiff; while cases hold that such pat frisks can also constitute an adverse action in the prison context, the pat frisk in this case does not rise to that level because it took place more than five months after the last retaliatory action, and pat frisks are common in the prison system).

 This does not seem a remarkable holding. The standard is whether the CO's actions would dissuade an inmate from again asserting his speech rights. In addition, the Court of Appeals has held that "a combination of seemingly minor incidents [may] form the basis of a constitutional retaliation claim once they reach a critical mass." The case for that is Tripathy v. McKoy, 103 F.4th 106 (2d Cir. 2024). Ripping up the grievance and threatening the box would convince the inmate that it was not in his interests to speak out again.

The lesson here is that the district court has to aggregate the minor actions to determine if the plaintiff has a retaliation case. That is also the rule in employment discrimination cases. Such an approach provides a realistic view of what the plaintiff really endured. For now, due to the extended discussion on what constitutes an adverse action, this is the definitive cases involving retaliation against prison inmates who seek relief under the First Amendment. 

Vermont's lawsuit screening process violated the First Amendment

The Court of Appeals holds that state court system in Vermont violated the First Amendment in delaying public and media review of new lawsuits. The reason for the delay was that the clerk's office needed time to review the complaints to ensure they did not include Social Security numbers and identifying information.

The case is Courthouse New Service v. Corsones, issued on March 11. I guess the clerk's office in Vermont meant well; the pre-screening was intended to prevent fraud associated with the inadvertent release of personal information. Once that information was removed or redacted, then the media could review the new lawsuits. The trial court held, following an evidentiary hearing, that this process violated the First Amendment right of access to judicial documents. However, in affirming this ruling, the Court of Appeals adds a nuance: the State of Vermont is still legally able to forbid some review of complaints before they are released to the public, depending on the sensitivity of the information in the complaint.

The Court of Appeals (Leval and Chin, with Sullivan in dissent) reminds us that the First Amendment protects the right of access to complaints filed in court. While the Court holds that guarding against the release of sensitive information protects a "higher value" under the First Amendment, the problem is how Vermont went about doing things. The process in Vermont is not narrowly tailored, the Second Circuit holds. Narrow tailoring is a First Amendment concept. It ensures that governmental practices that have potential to infringe First Amendment rights requires careful application, or what we call "narrowly tailoring." 

The Second Circuit holds that, on the evidentiary record developed at trial, that "Vermont has failed to demonstrate that the delays it imposed . . . could not have been reasonably shortened to a significant degree without impairing the higher value sought to to be protected." For one thing, some complaints were not made available to the public right away not because they contained Social Security numbers but because they did not comply with the technical requirements imposed by the clerk's office. Another reason for some of the delays is the clerk's office was looking for attorney notes that somehow found their way into the filed complaints; these notes would not contain confidential information but just careless notes the lawyer forgot to omit from the final version. Put another way, the delay in publicly releasing the complaints sometimes flowed from considerations that do not rise to the level of the "higher values" that might justify a brief delay.

Federal abstention is no romp in the park

A slew of rules guide the relationship between the state and federal courts, specifically, when you can sue in federal court on state law matters, and when you can sue in federal court on issues that remain under review in state court. We call this the "abstention" doctrine, created by the courts and not the Constitution. These are not the most exciting rules in the world, but they exist to ensure that the federal courts are not stepping on the state courts.

The case is Gristina v. Merchan, issued on March 12. This case was argued in January 2023, demonstrating the complicated nature of federal abstention. Having handled abstention issues, I can tell you that abstention is not a romp in the park. Very few cases take more than two years to decide, but one reason for this is the lengthy dissent that challenges the majority's reasoning.

Plaintiff was convicted in state court in 2012 on prostitution-related offenses. Nearly 10 years after she pleaded guilty, she wanted the state court to unseal several transcripts and other materials related to her prosecution. She said the Sixth Amendment allowed her to make such a motion. The state court, for the most part, ruled against plaintiff, stating that some transcripts were already available (but should not have been made available in the first instance) and others had to remain under seal. 

Plaintiff then challenged the criminal court's ruling in an Article 78 proceeding, which is unique to state practice and allows for an expedited challenge to arbitrary and capricious rulings. The Article 78 challenge was denied, but plaintiff took an appeal to the Appellate Division and the New York State Court of Appeals. While those appeals were pending, however, she sued the criminal court judge in federal court, seeking an order that would release the transcripts and related materials. So plaintiff actually had simultaneous state and federal proceedings.

In the end, plaintiff lost her state appeals. But the federal case was dismissed as well, under the so-called Younger-abstention, named after a U.S. Supreme Court case that prohibits federal court interference with state court proceedings. Younger v. Harris was decided in 1971. While Younger is more than 50 years old, courts are still applying it in new situations. This is one such situation.

First, the Second Circuit (Lee and Merriam, with Menashi in dissent) says that the Article 78 petition qualifies as a state court proceeding under the Younger doctrine. This is so because the Article 78 deals with the state court's ability to govern itself. 

The other main issue is whether Younger abstention warrants dismissal of the federal case even if the state proceeding is over by the time the federal judge rules on the federal case. The Second Circuit holds that "our Court's case law . . . clearly indicates that the Younger abstention issue is evaluated at the time of filing, and it is not continuously re-evaluated throughout the pendency of the proceeding." The Court of Appeals stands behind these prior cases and says they serve to "create an incentive for plaintiffs to file duplicative proceedings in federal court before the end of their state court proceedings, hoping that by the time the district or circuit court ruled on the merits, the state proceedings would have ended." That would undermine one of the policies in Younger: to avoid duplicative state and federal court proceedings.

Administrative Law Judge is legally denied promotion over mask-related argument on public train

The plaintiff was an administrative law judge for the Workers' Compensation Board. He sued the WCB under the First Amendment after it rescinded his promotion for Senior Administrative Law Judge. The promotion rescission took place after the WCB saw a TikTok video depicting plaintiff of arguing with a New Jersey Transit conductor about wearing a mask on the train. Does plaintiff have a case?

The case is Cestaro v. Rodriguez, a summary order issued on March 12. Public employees have the right to free speech, but those rights are limited because the government has an interest in ensuring that the public workplace operates efficiently. The government can also win the case if it proves it would have disciplined the plaintiff for other reasons, separate and apart from the speech. The latter concern draws from the Supreme Court's Mt. Healthy ruling, issued in 1977.

Plaintiff says the speech was his objection to wearing a mask on the train. But the Court of Appeals (Park, Chin and Merriam) finds that "Defendants were entitled to summary judgment because they established as a matter of law that they revoked Cestaro’s promotion based on his unprotected conduct, rather than on any protected speech." One judge at oral argument noted the plaintiff's behavior was "rude and belligerent." Plaintiff's counsel noted that, during the argument, plaintiff was not wearing a suit and tie but "shorts and flip-flops." The mask requirements during COVID certainly raised passions, and generally, disagreement with public safety mandates might raise a matter of public concern. You can imagine what this argument looked and sounded like to the other passengers on the train, or to the people who viewed the argument on TikTok.

But it was not plaintiff's objection to the mask mandate that cost him the promotion. Defendants proved on the summary judgment motion that plaintiff's conduct "was a poor way to treat workers," he was not "fair to the transit staff," and he was "unprofessional and aggressive" towards the conductor. Defendants did not discuss the constitutionality of masking requirements on public transit, or plaintiff's views on that subject. Instead, the record shows that defendants expressed concern about "having a supervisor at the state who behaves in this manner, because he cannot be trusted to be fair with the staff or the public."