For more than a decade, employment discrimination causation law has been a confusing, often overly restrictive quagmire that has contributed substantially to the paltry success rate of plaintiffs in employment discrimination cases. Most of these cases are dismissed pretrial, all too often based on a failure of causation. A key reason traces back to loose and misleading language—centered on a single word—in a 2009 Supreme Court opinion involving the but-for causation standard that applies in most discrimination cases. The Court said that the discriminatory motive must be “the” but-for cause of the employer’s action when it should have said “a” but-for cause. This language incorrectly implies that the discriminatory motive must be the sole cause—“the” cause—of the employer’s action, and though the sole-causation standard is demonstrably wrong, many courts nevertheless have required such a showing.
In 2020, the Supreme Court held in Bostock that gay and transgender workers are protected from employment discrimination. But it did more than that. The Court discussed but-for causation in broad terms, making it clear that but-for causation does not require proof of sole causation. That discussion will likely change how causation impacts not only gay and transgender workers but employment discrimination litigation generally.
This Article explains why. It studies the root of the causation problem and shows its real-world consequences with extensive case law analysis. Then, it analyzes how Bostock’s broad causation language, properly interpreted, should apply to most employment discrimination claims and details why Bostock should remedy each of the identified causation issues. Based on its study of post-Bostock litigation outcomes and the promise of Bostock’s broad application, the Article predicts a revolution in employment discrimination litigation.
“[A] lot here turns on a small word.”1Niz-Chavez v. Garland, 141 S. Ct. 1474, 1480 (2021).
Introduction
A nurse lost her retaliation suit at the pleadings stage, even though she was fired a mere two weeks after complaining of race discrimination, because her file showed documented performance problems.2 She also could not sue for both retaliation and race discrimination in the same case because asserting any nonretaliatory reason for the employer’s actions—including race discrimination—negated her retaliation claim.3 A milk delivery driver’s supervisor allegedly telling him “[Y]ou are too old to be here and I’m going to get rid of you” was not enough to save his age discrimination suit because the driver had also been disciplined at work.4 A science teacher who alleged she was mistreated for being the oldest female teacher in the department could not even argue to the jury that the employer’s actions resulted from both her age and her sex.5
These cases all turned on causation. Judges denied (incorrectly, I would contend) these plaintiffs the chance to have their day in court based on judicial interpretation and application of the causation standard. Cases like these are all too common in American courtrooms. Causation plays an outsized role in employment discrimination litigation.6 Employment discrimination claims are notoriously unsuccessful,7 and causation is often one of the reasons why.8
Which standard should apply in any particular case has been a recurring issue since Congress first passed protections against employment discrimination in 1964. In a key 2009 case, the Supreme Court in Gross v. FLB Financial Services held that claims under the Age Discrimination in Employment Act (ADEA), which prohibits discrimination “because of . . . age,”9 must be evaluated using but-for causation, meaning that the employer would not have acted as it did if it had not considered the plaintiff’s age, rather than a motivating-factor standard, which requires only that age play a role in the decision, even if not an outcome-determinative one.10 It concluded that the “ordinary meaning” of “because of” necessarily mandated a but-for standard.11 The Court did not elaborate on the meaning of the but-for causation standard, stating only that these plaintiffs must prove “that age was the ‘but-for’ cause” of the alleged discrimination.12
This language is quite restrictive. Even if the Court did not intend to do so, stating that the illegal motive must be “the” but-for cause of the treatment suggested there can be only one such cause. Gross set off a firestorm. Scholars criticized Gross on many grounds, not the least of which is its wording of the but-for causation standard.13
Given the Supreme Court’s analysis that the ordinary meaning of terms like “because of” mandates but-for causation, lower courts began applying the Gross standard in other employment discrimination settings and beyond.14 The Court then added fuel to the fire in 2013 when it held that Title VII’s retaliation protections were also governed by the but-for standard.15 It characterized but-for causation as the traditional common-law rule of causation-in-fact that applies by default unless Congress chooses to implement a different standard.16 Mirroring its holding from Gross, the Court stated that “Title VII retaliation claims require proof that the desire to retaliate was the but-for cause” of the employer’s decision.17
Gross and its progeny have wreaked havoc on the employment discrimination landscape. The Court’s overly restrictive wording of the but-for standard has often caused litigants and courts to equate but-for cause with sole cause, and that strict standard has directly caused many dismissals.18 This is what happened to the nurse, the milk delivery driver, and the science teacher. This occurs frequently in cases where an employer merely points to an unrebutted legitimate reason for its action, such as a performance issue, or where an employee alleges discrimination based on multiple or combined illegitimate motives, such as both retaliation and race. Cases have been dismissed at the pleadings stage, at summary judgment, and at trial. Sorting out the proper causation standard, figuring out the right language to describe it, and determining at what stage of the proceedings it applies is complex and has caused uncertainty. Confusion abounds—with scholars, litigants, judges, and juries.19
The Supreme Court continued to endorse but-for causation as the traditional, default causation rule20 but had not again addressed causation significantly in the discrimination context until 2020, when it issued three decisions involving causation.21 All three discussed but-for causation and reaffirmed its status as the ordinary, traditional, default causation standard.22
The most significant of these decisions is Bostock v. Clayton County, where the Court held that Title VII’s prohibition of discrimination “because of . . . sex” prohibits discrimination based on gender identity and sexual orientation.23 While that holding is momentous in and of itself, the Court took an unexpected route in reaching that conclusion with a deep dive into the meaning of but-for causation. Justice Gorsuch, writing for the 6–3 majority, explained that traditional but-for causation is a “sweeping standard” that does not require the illegal motive to be the sole or even primary motive as long as, without considering the prohibited factor—here, sex—the outcome would not have been the same.24 He also emphasized that a single event can have multiple but-for causes.25 This is a far cry from what the Court said—and did not say—in Gross.
Bostock’s statements regarding traditional but-for causation law should shake the core of employment discrimination litigation. Properly understood, the Bostock language on but-for causation should apply expansively across employment discrimination litigation and should solve many of the problems Gross left in its wake.
Gross has been widely criticized.26 This Article adds to the discussion by comprehensively explaining and documenting the practical consequences of Gross’s but-for causation language. And though many commentators and scholars have predicted wide-ranging consequences from Bostock,27 they have often assumed that Bostock will apply outside the Title VII context without analyzing that assumption.28 This Article contributes by thoroughly making the case for why Bostock’s causation language should apply broadly to the employment discrimination spectrum. It also extensively documents the problematic Gross case law and details Bostock’s language to show how Bostock’s vision of but-for causation should solve the problems Gross created. It also analyzes the litigation trends after Bostock to show how courts are responding, both in applying Bostock’s but-for causation language outside of the Title VII context and how they actually use that language.
This Article is not about the substantive debate, raging for decades, over which causation standard is most appropriate in antidiscrimination litigation, either from a policy perspective or as a matter of statutory interpretation.29 That ground is well trod, and in some ways, is now beside the point. The Supreme Court has dug in deep on but-for causation, and, barring congressional action,30 it is the standard that will apply in much of employment discrimination litigation. The practical issue, then, is not whether but-for causation is proper but how but-for causation applies in light of Bostock.31
This Article predicts that Bostock’s expansive but-for causation language will apply broadly and will transform employment discrimination litigation over the coming years. Part I provides necessary background regarding causation in employment discrimination litigation. Part II details the Gross decision and the Supreme Court’s subsequent entrenchment of the but-for causation standard. Then in Part III, the Article explores Gross’s disastrous consequences for employment discrimination plaintiffs based on its restrictive description of but-for causation. Finally, Part IV explains the analysis behind the author’s prediction that Bostock will remedy many of Gross’s problems and revolutionize employment discrimination–causation litigation.
I. Causation in Employment Discrimination Litigation
Causation is at the core of employment discrimination litigation.32 To understand its significance and fully grasp why Bostock will likely lead to revolutionary changes in employment discrimination cases, some background information on causation standards in employment discrimination litigation is useful.
At their base, employment discrimination statutes prohibit adverse employment actions caused by a plaintiff’s protected status, such as disability or race. The two causation standards most often at issue in these cases are but-for causation and motivating-factor causation.33
The but-for standard is a traditional causation standard rooted in tort and criminal law.34 As applied in the employment discrimination context, the but-for standard provides that causation is established when the employer’s action would have been different if it had not considered the protected classification, such as sex.35 In other words, if that one factor was changed—so if the employee was a man instead of a woman—would the employer’s decision have been the same? If not, then but-for causation is established.36 The prohibited factor must play a necessary or essential role in the employment decision.37
But-for cause, however, is not the same as sole cause.38 Sole cause is the most restrictive form of causation and precludes a causation finding if any other cause contributes to the injury.39 A single event can have many but-for causes, and a but-for cause does not need to be the primary cause; as long as changing a factor would change the outcome, then that factor is a but-for cause, even if other causes also exist.40
The motivating-factor standard, by contrast, examines whether the protected classification was one of the employer’s motives in acting.41 This standard is easier to show than but-for causation because even if a protected classification was not a determinative component—so even if the employer would have fired the woman anyway based on, for example, performance issues—the employee would prevail if her gender played a role in the employer’s decision-making process.42 At its most fundamental level, the difference between but-for causation and the motivating-factor standard is not whether multiple causes exist but how big a role the protected status must play in the employment decision.43 If the protected characteristic is a necessary step in the causal chain, but-for causation is established; if it does not play a necessary role because the outcome would be the same regardless, then it is not a but-for cause but can still be a motivating factor.
Scholars have long debated which standard is most desirable from a policy perspective.44 But-for causation seems like a straightforward concept at its core, simply asking whether something such as age changed the outcome.45 It is, however, not always so easy to apply in the real world. But-for causation usually works well for physical causation, such as the cause of a fire or a death, but not as well for human thought processes, such as motive.46 Our brains are complex. Human decisions, such as whether to discipline or discharge an employee, are often multifaceted.47 Determining exactly what motivated an employer’s action or what might have happened if the employer had not considered a factor such as race is often an impossible task.48 The but-for standard also allows employers to get away with some level of discrimination so long as the discrimination is not enough to have changed the outcome.49 Motivating-factor causation does not try to parse how much of a role the protected classification played; it looks at whether it merely played a role in the actual decision.50 This standard thus promotes a policy of keeping consideration of a protected classification completely out of the workplace decision-making process, but some see it as unfair because it allows liability even if the protected classification made no difference in the ultimate outcome.51
How these standards play out in any case depends in part on the specific wording used, but traditional wisdom is that the but-for standard favors employers and the motivating-factor standard is pro-employee.52 Indeed, empirical evidence suggests that employers win significantly more often with but-for jury instructions than with motivating-factor instructions.53 Not surprisingly, antidiscrimination advocates have historically championed the motivating-factor standard.54
How do these standards fit into employment discrimination statutes? Congress passed Title VII of the Civil Rights Act in 1964, prohibiting employment discrimination “because of” sex, race, national origin, color, or religion or in retaliation for engaging in some forms of protected activity.55 Other than stating that employers could not act “because of” these factors, Congress did not specify a causation standard. The same held true in 1967, when Congress enacted the ADEA, prohibiting employment discrimination “because of” age.56 It is broadly accepted that Congress did not create a sole-causation standard here, based on the statutory language and legislative history.57 The question left open was whether liability under these statutes depends on showing that something such as race or age was a motivating factor, or if but-for causation was required.
In 1989, a splintered Supreme Court in Price Waterhouse v. Hopkins held that a Title VII plaintiff can meet her burden of proving the employer acted “because of” a protected classification by showing that it was a “motivating” or “substantial” factor.58 Once the plaintiff makes that showing, the burden then shifts to the defendant to prove it would have taken the same action even without considering that factor.59 In other words, but-for causation became an affirmative defense, called the same-action defense, allowing the plaintiff to win on a showing of motivating-factor causation if the defendant could not prove the affirmative defense that it would have taken the same action anyway.60 This decision, though universally viewed as pro-employee,61 did little to clear up the causation debate because no one opinion gained a majority and because the Court used fairly imprecise language in describing the standard62—Professor Martin Katz counted more than twenty different causation formulations throughout the various opinions.63
In the Civil Rights Act of 1991, Congress amended Title VII and codified the motivating-factor standard.64 It did so, however, with a significant modification. A plaintiff could still establish causation if a protected characteristic “was a motivating factor for any employment practice, even though other factors also motivated the practice,”65 but the same-action defense was no longer (as in Price Waterhouse) an affirmative defense that could completely eliminate an employer’s liability. Rather, if the employer could prove it would have made the same decision regardless of the prohibited characteristic, the plaintiff would still win, but the defendant would have a defense to damages.66 Obviously, this is viewed as even more pro-plaintiff than Price Waterhouse.67 The motivating-factor standard did not replace but-for causation but supplemented it, allowing plaintiffs an alternative route for establishing liability under Title VII.68
Congress had enacted the Americans with Disabilities Act (ADA) in 1990, which prohibits disability-based employment discrimination.69 In the wake of Price Waterhouse and the Civil Rights Act of 1991, lower courts and scholars were left to figure out how to apply the motivating-factor standard and to determine what standard to apply in age and disability cases or in cases alleging retaliation under Title VII or other statutes.70 It is against this backdrop that the Supreme Court took up the case of Gross v. FBL Financial Services.71
II. Gross and the Entrenchment of But-For Causation
When Jack Gross’s employer restructured positions and assigned many of his responsibilities to a younger employee, Gross sued for age discrimination under the ADEA.72 A key trial issue centered on the type of evidence Gross needed to obtain a jury instruction regarding age as a motivating factor in the demotion decision.73 The trial court instructed the jury on motivating factor, including the same-decision defense, and Gross prevailed.74 The Eighth Circuit reversed, holding Gross was not entitled to that instruction.75 The Supreme Court granted certiorari on the issue of the type of evidence necessary to obtain a motivating-factor instruction in a non-Title VII case.76
But the Court did not address that issue. Instead, it considered what it characterized as the “threshold” issue of whether an ADEA plaintiff is ever entitled to a motivating-factor instruction.77 Writing for the 5–4 majority, Justice Thomas held that an ADEA plaintiff is never so entitled.78 Though Congress modeled the ADEA on Title VII and the Court previously had interpreted the two statutes fairly uniformly,79 Justice Thomas stated that Title VII cases such as Price Waterhouse did not control because Congress amended Title VII after Price Waterhouse (via the Civil Rights Act of 1991) to add a motivating-factor provision but did not simultaneously amend the ADEA.80
Freed from the constraints of Title VII cases and statutory law, the Court focused on the ADEA’s text.81 The ADEA prohibits employment discrimination “because of such individual’s age.”82 Citing dictionary definitions and a torts treatise, the Court held that the “ordinary meaning” of the “because of” age provision meant that the plaintiff must show that “age was the ‘reason’ that the employer decided to act.”83 The Court explicitly labeled this as but-for causation: “To establish a disparate-treatment claim under the plain language of the ADEA, therefore, a plaintiff must prove that age was the ‘but-for’ cause of the employer’s adverse decision.”84 The burden never shifts to the employer—as it does with a motivating-factor standard—to prove it would have made the same decision without considering age.85 Thus, it is “never proper” for an ADEA plaintiff to receive a motivating-factor instruction.86
Gross caused an immediate stir. As Justice Stevens stressed in his dissent, the Court ruled on an issue not briefed by the parties or amicus curiae.87 And the Court did not explain the causation standard, other than to say that age must be “the ‘but-for’ cause” of the employer’s decision.88 More on that crucial wording later. Further, the Court broke with prior practice of interpreting Title VII and the ADEA consistently, thereby creating different causation standards for claims under those two statutes, despite their similar language, history, and structure.89 What did that mean for interpreting other statutes with “because of”–type language?
Right away, courts began applying Gross to these other statutes.90 For example, before Gross, most courts allowed liability under the ADA based on the motivating-factor standard.91 Since Gross, the trend is in the other direction, requiring plaintiffs to show but-for causation.92 Professor Deborah Widiss has documented how courts have similarly applied Gross to at least ten other statutes, including the Family and Medical Leave Act and the Labor Management Reporting and Disclosure Act.93
There was also the question of what standard should be applied in Title VII retaliation cases because the Civil Rights Act of 1991 did not directly address the antiretaliation provision.94 Before Gross, courts had gone both ways on the issue, but after Gross, the “tide really turned,” with the majority of courts requiring but-for causation.95
The Supreme Court weighed in on that precise question four years later in University of Texas Southwestern Medical Center v. Nassar, holding that the motivating-factor provisions in Title VII apply only to status-based discrimination claims and not to Title VII retaliation claims.96 Title VII’s antiretaliation provision prohibits discrimination “because” an employee engaged in certain protected activity.97 The Court said the motivating-factor standard from the Civil Rights Act of 1991 did not cover the retaliation provision because of its location in the statute.98 It then characterized but-for causation as “textbook tort law” and “the background against which Congress legislated in enacting Title VII.”99 Thus, but-for causation, according to the Court, is the “default rule[],” and it is “presumed” that Congress incorporates that standard into a statute unless the statute itself indicates otherwise.100 Noting the similarity between the ADEA’s language and the antiretaliation provision in Title VII, the Court then relied on Gross to hold that the statutory language requires but-for causation:
Given the lack of any meaningful textual difference between the text in this statute and the one in Gross, the proper conclusion here, as in Gross, is that Title VII retaliation claims require proof that the desire to retaliate was the but-for cause of the challenged employment action.101
Justice Ginsburg, writing for the four dissenters, castigated the majority for creating two separate causation standards in Title VII cases and for affording retaliation claims less protection by requiring plaintiffs to meet a more difficult causation burden.102
As Gross and Nassar make clear, the Supreme Court is deeply entrenched in but-for causation as the baseline causation standard, particularly in employment discrimination cases.103 The Court considered but-for as the default causation rule in two criminal cases in 2014,104 and in 2015, it cited Nassar in stating that “typically” in antidiscrimination laws, “because of” language signifies “the traditional standard of but-for causation.”105 Otherwise, the Court remained fairly silent on the topic, with lower courts continuing to apply but-for causation in the employment discrimination and other contexts.106 In 2020, however, the Court, in three late-term decisions, reaffirmed the primacy of the but-for principle.107 In the final case, Bostock, the Court substantially clarified the scope and application of the but-for standard.108 But, to fully comprehend the significance of this development, it is important to first understand in detail the problems Gross and its progeny have caused.
III. Gross Has Been a Disaster
Gross caused immediate confusion and upheaval in the antidiscrimination causation world, and those concerns have persisted. Scholars have lambasted Gross on many grounds, including that but-for causation does not, as the Court stated, naturally flow from “because of,”109 is based on a misreading of the statute and its history,110 and is otherwise the wrong standard for employment discrimination cases,111 particularly because the Court made no attempt to justify importing tort law into the employment discrimination context.112 Though sharing many of these concerns, this Article focuses elsewhere. Rather than debating whether the Supreme Court should have adopted but-for causation, this Article concentrates on what the Gross Court said about but-for causation—and the problems that flow directly from that.
One small word. Much of the negative fallout from Gross comes down to the word “the.” The Gross Court said that age must be “the” but-for reason for the employer’s action.113 Not “a” reason—“the” reason. This wording strongly suggests that under the but-for standard, the protected characteristic must be the one-and-only reason for the adverse employment decision.114 But that is incorrect. One-and-only is the sole-causation standard, not the but-for causation standard.115
The Supreme Court surely knew this when it decided Gross. The distinction between but-for and sole causation is a fundamental tenet of tort law.116 The Price Waterhouse Court debated the merits of but-for cause, sole cause, and motivating factor.117 Moreover, in his Gross dissent, Justice Stevens noted that the dictionary definitions the majority relied on did not define “because of” to mean “solely by reason of.”118 So what explains the Court’s markedly unsharp wording?119
As many scholars have noted, the Court sloppily and skimpily imported tort causation law into the employment discrimination context and did not pay attention to subtleties, in particular those involved when multiple causal factors are at issue.120 A single injury can have many but-for causes.121 Moreover, the common law of torts on which the Court purportedly relied includes not only but-for causation but a “bundle of causal standards” that accompany it to account for other situations when many causal factors are at play.122 But the Court overlooked all of this nuance by stating simply that age must be “the” but-for cause of the employment decision.
As demonstrated earlier, Gross’s “because of equals but-for” rationale spread like wildfire, and with it, mounting confusion over what the but-for standard requires.123 Many litigants124 and courts125 have explicitly equated sole and but-for causation, and others have used wording essentially leading to the same result.126 Incorrect, but who can blame them? As Professor Brian Clarke explained,
The Court’s word choice in Gross, combined with its limited explanation of the parameters of the cause-in-fact standard it adopted, opened the door for interpreting factual causation in the disparate treatment context as sole cause. It is even the most natural interpretation of the Court’s plain language.127
Moreover, some scholars magnified the problem; in their zest for advocating for the motivating-factor standard, they used hyperbole that perhaps overplayed the restrictiveness of the but-for standard.128 Even though the concept of sole causation in the employment discrimination context has been characterized as “too stupid to take seriously,”129 Gross’s sloppy language created a sole-cause trend.130
Not all courts have jumped onto the “but-for equals sole” bandwagon,131 but the problem is widespread and has caused direct, concrete injuries to discrimination plaintiffs. Recall the nurse who sued for race discrimination and retaliation.132 She alleged that she complained to her employer about thirteen unfair employment conditions, stating that she had been singled out and treated differently.133 Her employer fired her two weeks later.134 She sued for race discrimination and retaliatory discharge based on her complaints about her treatment.135 The court granted summary judgment on her retaliation claim,136 emphasizing Nassar’s requirement that retaliation be “the” but-for cause of the employer’s action.137 Before Nassar, the court reasoned, two weeks between protected activity and termination would have been sufficient to survive summary judgment, but temporal proximity alone, even when this close, “does not meet the new ‘but for’ causation standard.”138 Thus, the Gross standard, as applied to Title VII retaliation claims in Nassar, directly led to this summary judgment that previously would have been denied.
Most employment discrimination cases involve many potential motives for the employer’s conduct139—after all, employers typically do not readily admit that they acted based on something like the plaintiff’s race.140 The interpretation of Gross that but-for causation requires sole cause typically plays out against the plaintiff in these cases.
Take, for example, a scenario where a plaintiff has alleged discrimination based on an illegal motive and the employer counters that it acted for a legitimate reason, such as a job performance problem.141 Many courts have held that if the plaintiff cannot negate the employer’s proffered reason, the plaintiff loses as a matter of law, despite evidence of an illegal motive.142 Under the strict interpretation of Gross’s language, the existence of a legitimate motive means it is impossible that the employer also acted for an illegal reason because, according to this theory, but-for causation requires a showing that the illegal factor is the one and only motive. It is a binary choice: either the employer acted based on the illegal factor or it did not.
This was the downfall of the milk delivery driver discussed earlier.143 He had undisputedly been disciplined at work, and that alone justified summary judgment under Gross.144 Even though the Fourth Circuit gave lip service to the idea that but-for cause need not be sole cause, it still said that age must be “the reason” for the employer’s decision, and because the plaintiff had documented performance issues, it was impossible to show that age was “the but-for cause” of his termination.145 The same result has played out in case after case, especially those involving age claims,146 as in Gross, or retaliation claims,147 following on the heels of Gross in Nassar.
Similarly, plaintiffs face an uphill battle when they allege alternative impermissible motives and at least one of those motives is governed by a but-for standard, such as being discriminated against based on age or race. In the nurse’s case, she sued for both race discrimination and retaliation.148 In addition to rejecting her retaliation claim because fourteen days of temporal proximity was insufficient as a matter of law to meet the new but-for standard, the court said her allegation that race motivated her termination defeated her retaliation claim because she must plead and prove that retaliation was “the ‘only’” motive for termination.149 Court after court has ruled in the same fashion, rejecting age and retaliation claims—which require but-for causation—when combined with Title VII status claims,150 disability claims,151 or claims based on any other illegal motive.152
These rulings mean that these plaintiffs have no right to present their cases to a jury, regardless of the merits of their claim that a factor such as age played a role in their termination, because the existence of a second motive—whether legal or illegal—makes proving Gross but-for causation impossible. Some courts have gone so far as to apply this standard at the pleading stage, holding that plaintiffs cannot even plead alternative discrimination theories when one of those theories requires but-for causation, because pleading a fact in support of the alternative theory (such as race) negates any possibility of but-for causation for the theory that requires it (such as age).153 Fortunately for plaintiffs, most courts have rejected this theory.154 They say that Gross did not address pleading and that alternative, inconsistent pleading is a staple of federal litigation.155 They do, however, acknowledge that the Gross standard might mean that these plaintiffs will lose at summary judgment or will have to elect before trial, even if not at the pleading stage.156
A distinct but related problem arises when plaintiffs allege discrimination based on the intersection of protected characteristics—such as being an older woman—as opposed to multiple protected characteristics in the alternative—such as being older or being a woman.157 Courts have struggled with these intersectional claims on various grounds, but the primary objection centers on causation when the intersecting bases would have different causation standards if brought as separate claims or when one basis is governed by but-for causation.158 The science teacher who alleged she was mistreated for being the oldest female teacher faced this problem.159 The court noted the “conflicting standards of proof” for age and sex claims and cited Gross for the proposition that age must be “the only reason for the conduct” in an age claim.160 Because of this, the teacher could not combine her claims, either under Title VII or the ADEA.161 Either combination would allow the plaintiff to argue age as a basis for liability without showing that age was the only motive for the decision, and, according to the court, that is inconsistent with Gross.162 Thus, the plaintiff could not argue to the jury that the employer acted both because of her age and her sex.163 The court forced her to contend either that the employer acted because she was older, or because she was a woman, but not an older woman—even though it is well established that older women face discrimination distinct from that of younger women or older men.164 Many other courts have held similarly.165
The causation confusion in these cases flows directly from Gross. A Supreme Court opinion generated the problem. Perhaps another Supreme Court opinion can repair it.
IV. Bostock Will Revolutionize Employment Discrimination Litigation
Bostock has the potential to revolutionize the causation landscape in antidiscrimination litigation. This Part lays out the basis for predicting that it will do so. After explaining the Bostock decision, this Part will show why Bostock’s but-for causation language should apply beyond Title VII status claims to all federal antidiscrimination statutes requiring but-for causation. From there, a detailed comparison between the language in the Gross line of cases and Bostock makes abundantly clear that, properly applied, Bostock’s vision of but-for causation will undo much of Gross’s damage. Though predictions are inherently uncertain, especially regarding employment discrimination litigation trends, there are good reasons to be hopeful, even in the face of a cautionary tale from an earlier attempted course-correction. This Part ends with thoughts on how the motivating-factor standard fits into this Bostock-dominated future.
A. Bostock: Not Gross’s But-For
In Bostock, the Supreme Court considered three consolidated cases involving one transgender and two gay workers.166 Their employers had fired each of them allegedly for no reason other than their sexual orientation or gender identity.167 The issue before the Court was whether Title VII’s prohibition of discrimination “because of . . . sex” includes discrimination based on sexual orientation or gender identity.168 The Court held that it did,169 but the route it took to reach that conclusion came as a surprise to many.170
A linchpin of Justice Gorsuch’s analysis was causation law. Hearkening back to cases such as Gross and Nassar, Justice Gorsuch stated that “the ordinary meaning of ‘because of’ is ‘by reason of’ or ‘on account of,’” and that indicates the “‘traditional’ standard of but-for causation.”171 That standard “change[s] one thing at a time [to] see if the outcome changes,” and if so, that thing is a but-for cause.172 Describing this standard as “sweeping,”173 Justice Gorsuch emphasized that there can be multiple but-for causes and took great pains to distinguish but-for causation from other standards such as sole causation.174 Playing a necessary role is sufficient, he explained, even if the causal factor at issue was not the primary or main cause.175 He noted that the Civil Rights Act of 1991 had added a motivating-factor standard to allow liability in some instances when discrimination is not a but-for cause, but that new standard had merely supplemented, rather than displaced, the but-for standard that flows directly from the statutory “because of” language.176
So what does all this have to do with whether firing a gay or transgender worker violates Title VII?177 Justice Gorsuch said that “sex plays an unmistakable and impermissible role” in any employment decision based on gender identity or sexual orientation because if the employee’s sex is changed, that would yield a different outcome.178 In other words, for example, if an employer “fires [a] male employee for no reason other than the fact he is attracted to men, the employer discriminates against him for traits or actions it tolerates” in a female employee.179 Bottom line: “[I]t is impossible to discriminate against a person for being homosexual or transgender without discriminating against that individual based on sex.”180 That makes the employee’s sex a but-for cause of the termination.181 Even if sex is not the employer’s primary or main motivation, because sex “plays a necessary and undisguisable role in the decision,” Title VII prohibits that conduct under a but-for causation analysis.182
As should be immediately apparent by this point, Justice Gorsuch’s description of but-for causation in Bostock is markedly different from the Gross Court’s description and how lower courts have interpreted it. As in Bostock, Gross said that “because of” means “by reason of” or “on account of” and thus its ordinary meaning indicates but-for causation.183 From that starting point, the Gross Court then stated that “a plaintiff must prove that age was the ‘but-for’ cause” of the employment decision.184 Other than a few parentheticals,185 the Gross Court did not further explain how the but-for causation standard should work in general or in employment cases. As demonstrated earlier, many lower courts interpreted Gross’s “the” cause language to require sole causation.186 For example:
- “‘But-for’ causation is ‘sole’ causation.”187
- The plaintiff must show “that age—not age and other motives—was the only impetus” for the employment decision.188
- Under Gross, “only the age motive truly matters.”189
- The plaintiff must prove that the discriminatory factor “was the only reason for the conduct.”190
Bostock’s language is striking by comparison. The most immediate difference is the article—liability attaches if discrimination based on the protected status is “a” but-for cause,191 not “the” but-for cause. And rather than stopping there, as Gross did, Bostock provides detailed examples and explains, repeatedly, several key aspects of but-for causation, including:
- “Often, events have multiple but-for causes.”192
- If sex “was one but-for cause” of an adverse employment decision, “that is enough to trigger” Title VII.193
- The employee’s protected status need not be “the sole . . . cause” of the employer’s action.194
- The employee’s protected status also need not be the “primary cause”195 or “main cause”196 of the employer’s action.
- “[A] defendant cannot avoid liability by just citing some other factor that contributed to its challenged employment decision.”197
- “It doesn’t matter if other factors besides the plaintiff’s sex contributed to the decision.”198
- But-for causation can be established even if another factor “play[ed] a more important role in the employer’s decision.”199
- “An employer violates Title VII when it intentionally fires an individual employee based in part on sex.”200
All told, Justice Gorsuch emphasized the breadth of the but-for standard, that it is not the same as sole cause, and how it accounts for multiple causal factors—even ones that are more important than the protected status—at least thirty-five times.
B. Bostock Should Apply Broadly Across Employment Discrimination Statutes
Bostock, of course, was written in the context of Title VII sex discrimination claims. For Bostock’s vision of but-for causation to stand a chance of cleaning up the post-Gross causation chaos, it cannot be limited to the Title VII context in which it arose. Bostock’s causation language must apply to all employment discrimination statutes allowing liability based on but-for causation. Justice Gorsuch acknowledged employers’ fears that Bostock “will sweep beyond Title VII to other federal or state laws that prohibit sex discrimination” but noted that “none of these other laws are before us.”201 Even so, the Court’s rationale that discrimination “because of . . . sex” includes discrimination based on gender identity and sexual orientation has already been applied to many other federal statutes and regulations.202 Indeed, one of President Biden’s first-day executive orders directed that Bostock’s reasoning be applied to other federal laws “that prohibit sex discrimination,” including Title IX of the Education Amendments of 1972 and the Fair Housing Act.203 While tremendously important, these developments are beside the point for this analysis. Here, the issue is whether Bostock’s explanation of the scope and contours of traditional but-for causation will apply to antidiscrimination statutes other than Title VII’s status-based protections.
A proper reading of Bostock and the cases it builds on unmistakably demonstrates that Bostock’s but-for language should apply to all antidiscrimination statutes that require a but-for analysis. Justice Gorsuch purported to apply the “ordinary” meaning of “because of” to indicate the “traditional” but-for test.204 “Ordinary” and “traditional” are concepts that span statutes.205 This is why Nassar, in considering Title VII’s antiretaliation provision, relied on Gross’s interpretation of the ADEA—because they shared the same language and thus shared a common, ordinary meaning.206 And Bostock was not written on a clean slate; it came on the heels of Comcast v. National Association of African American-Owned Media, issued only three months earlier and also written by Justice Gorsuch, where he described the but-for standard in § 1981 race discrimination cases using these same “ordinary” and “traditional” principles.207 These traditional but-for principles are the default rules, the background law that Congress is presumed to have incorporated into statutes unless the text indicates otherwise.208
Bostock relied on cases such as Gross and Nassar—cases that used dictionary definitions and basic tort law to discuss the “ordinary” meaning of phrases like “because of” to apply “traditional” but-for causation principles in the age and Title VII retaliation contexts.209 Justice Gorsuch appropriately drew on these sources involving other statutes to show the ordinary and traditional meaning of these default but-for causation concepts. There is no justification for any interpretation of Bostock that categorically limits its understanding of basic but-for causation law to the Title VII context.
Most courts that have considered the issue have applied Bostock’s but-for causation language outside the Title VII status-based discrimination context.210 This includes cases involving age discrimination claims,211 an intersectional claim based on age and sex,212 Title VII retaliation claims,213 § 1981 race claims,214 a § 1983 Equal Protection sexual harassment claim,215 retaliation under the False Claims Act,216 and various other federal statutes217 and state-law torts.218
The Sixth Circuit, however, has gone the other way. In Pelcha v. MW Bancorp, Inc., the court held that Bostock applies only to Title VII cases.219 Pelcha sued her employer under the ADEA, and her employer said it fired her not based on her age but for insubordination.220 The Sixth Circuit affirmed summary judgment for the employer after going through the now-familiar process of quoting Gross for the principle that age must be “the ‘but-for’ cause” of her termination.221 Summary judgment was proper, said the Sixth Circuit, because Pelcha could not show that the “true motive” for her termination was age rather than insubordination.222 Pelcha argued that Bostock altered her burden under Gross and thus she need not prove that age “was the only cause of the termination.”223 The court rejected her argument, stating simply that Justice Gorsuch’s comment that the Court was not at that time considering other laws meant “the rule in Bostock extends no further than Title VII and does not stretch to the ADEA.”224 Thus, the Sixth Circuit applied Gross because it “directly controls” age discrimination claims.225 Even after briefing on rehearing, including an amicus curiae brief from the Equal Employment Opportunity Commission explaining the court’s error in endorsing the idea that age and Title VII claims are governed by differing but-for causation standards,226 the Sixth Circuit refused to budge.227
This is nonsense. Bostock relied on Gross to support its initial conclusions and then expanded beyond Gross’s minimal explanation of but-for causation.228 The whole point of a default rule is that it applies routinely across cases. Nothing in Bostock suggests it was differentiating the ordinary, traditional, default but-for causation standard applied to age claims from that same standard as applied to Title VII status claims. There is only one but-for causation standard, not separate but-for standards for different statutes.229 And there would be no reason for Justice Gorsuch to broaden the standard for Title VII status claim (as compared to age claims under the ADEA in Gross) because Title VII already has the less restrictive motivating-factor standard available.230 No rational statutory interpretation supports the outcome that “because of age” means age must be “the” but-for cause of discrimination while “because of . . . sex” means that sex must be “a” but-for cause.
Fortunately, one district court in Arizona seems to be the only court outside of the Sixth Circuit going down this erroneous path.231 Though the defense bar continues to assert that Bostock is limited to Title VII,232 most courts have properly rejected this argument. The Sixth Circuit’s position is indefensible in the context of the ADEA and any other statute that applies traditional but-for principles.233 Pelcha has drawn the most attention, but it is an outlier.234
C. Bostock’s But-For Language Will Remedy Many of Gross’s Problems
The law of but-for causation as Bostock described would, if properly applied, ameliorate many of the problems Gross has caused, helping plaintiffs both have their day in court and prevail when they get in front of the jury.
The root of many Gross-based dismissals is the belief held by many defense lawyers and courts that but-for causation is sole causation, based on Gross’s “the ‘but-for’ cause” language.235 Bostock puts that flawed argument to bed. Post-Bostock, it is simply no longer legally viable to argue or hold that traditional but-for causation is equivalent to sole cause.236 As Justice Gorsuch stated, “[A] defendant cannot avoid liability just by citing some other factor that contributed to its challenged employment decision.”237 That shield is now gone.
Eliminating this defense should have profound consequences in many types of employment discrimination cases. In cases where the employer points to a legal reason for firing a plaintiff, that alone cannot form the basis for dismissal because, according to Bostock, “[i]t doesn’t matter if other factors besides the plaintiff’s [protected status] contributed to the decision.”238 Similarly, a worker’s claim alleging two illegal motives for termination, such as both race and retaliation, cannot be rejected on the bare ground that asserting two illegal motives cancels out one or the other because, as Bostock makes clear, both illegal causes can be but-for causes: “Often, events have multiple but-for causes.”239 This should have a significant impact on age and retaliation claims, where the Supreme Court has held that the but-for standard applies,240 and on disability claims, where most post-Gross courts have applied the but-for standard and likely will continue to do so.241 So too for intersectional claims, where employees allege discrimination based on a combination of protected characteristics, such as age and sex. The primary ground for dismissing such claims has been the supposed incompatibility of their strict causation standards,242 but Bostock shows that multiple bases of discrimination fit comfortably within the but-for causation framework.243 Thus, it is no longer defensible to reject intersectional claims on the basis that the intersecting categories have strict causation standards that allow for only one motive.
Bostock’s language can also help these plaintiffs when their cases, having avoided dismissal, actually get to the jury. Empirical evidence from mock jury studies has shown that employers win more often with but-for causation jury instructions than with those based on motivating factor.244 The motivating-factor instructions in these studies, however, were more elaborate and nuanced than the but-for instructions. The but-for instructions included language that (1) “age played a role” and the plaintiff “would not have [been] demoted . . . if [the employer] had not considered age”245 or (2) the defendant “would have promoted Plaintiff had he not been of Mexican national origin but everything else was the same.”246 By contrast, the motivating-factor instructions (1) stated that the plaintiff was “not required to prove that his age was the sole motivation or even the primary motivation,”247 (2) provided that liability could be established “even though other factors allowable under the law may also have motivated [the employer],”248 or (3) specifically mentioned factors that “contributed to the Defendant’s decision.”249 This explanatory language in the motivating-factor instructions is 100% consistent with Bostock’s description of but-for causation.250 Thus, using Bostock’s language in but-for jury instructions—language that tracks the successful motivating-factor instructions in these empirical studies—should improve employees’ chances of succeeding at trial.
This is especially true considering the results of Professor James Macleod’s study about the meaning of ordinary causation principles. Because the Supreme Court has repeatedly relied on what it terms the “ordinary” meaning of causation phrases like “because of” to mandate but-for causation,251 Professor Macleod decided to study what people actually think about these terms.252 Professor Macleod surveyed nearly 1,500 people nationally and found that they do not view causation words in the way the Supreme Court thinks they do.253 In the employment context, participants indicated that they thought an employer who fired a worker based in part on a prohibited factor considered that firing to be “because of” that factor, even if it did not rise to the level of a but-for cause.254 “These results demonstrate that the courts have been incorrect in claiming that but-for causation tracks the ordinary, plain meaning of the statutory causation language at issue” in Gross and other Supreme Court cases.255 Professor Macleod also noted that the study showed “participants were responsive to small differences in causal language.”256 It stands to reason, then, that jury instructions using Bostock’s language—which matches the plaintiff-friendly motivating-factor language from the empirical jury studies and comports more closely with the actual common understanding of “because of”—would produce more favorable jury outcomes for employment discrimination plaintiffs.
D. This Optimism Is Justified
It is by no means guaranteed that Bostock will in fact be applied broadly and correctly to remedy much of the mayhem in employment discrimination causation law. Employers are worried.257 Antidiscrimination scholars and advocates are hopeful.258 Is this optimism justified?
So far, the signs are mostly favorable. Barely a month after Bostock, the Tenth Circuit became the first federal appellate court to recognize an intersectional claim based on age and sex under Title VII, and it explicitly relied on Bostock to reach this conclusion.259 The Ninth Circuit reversed summary judgment on a Title VII retaliation claim, rejecting the argument that retaliation must be “the,” rather than “one,” but-for cause of the termination under Bostock.260 In an age discrimination case, a district court in Alabama slapped down a similar sole-cause argument in style, stating, “This argument crashes, Wile E. Coyote-esque, into veritable mountains of contrary precedent,” including Bostock.261 Other courts have followed suit.262
Of course, as Pelcha exemplifies, not all courts have taken the right path. Pelcha seemed to understand exactly what Bostock said and refused to follow it, but other courts appear perplexed about the meaning of Bostock’s but-for causation language. For example, the Seventh Circuit suggested that Bostock’s “one but-for cause” language is the standard for motivating-factor liability under Title VII rather than a description of how multiple causes can support liability under a but-for causation theory.263 Other courts have simply stated, without elaboration, that Bostock cannot apply outside of the Title VII sex context because the Supreme Court said it was not deciding those issues in Bostock.264 Practitioners have also gotten it wrong.265 And, of course, advocates will advocate—some lawyers continue to press sole-cause arguments, despite Bostock’s clear command to the contrary,266 and this might lead to further court and jury confusion.
Even with these missteps, the forceful and unequivocal language in Bostock should course-correct Gross’s wreckage better than the Supreme Court’s apparent previous attempt to do so in 2014 in Burrage v. United States.267571 U.S. 204 (2014). Burrage involved a criminal statute that imposed a mandatory sentence on a defendant who unlawfully distributes certain drugs when “death or serious bodily injury results from the use of such substance.”268 The statute does not define “results from,” so the Court gave the term its ordinary meaning and determined that this phrase, like “because of” in the employment statutes, required but-for causation.269 In doing so, the Court relied on Gross and Nassar, in addition to many criminal law authorities.270 Burrage involved the role of multiple drugs in the victim’s system, and the Court analyzed how those drugs (only one of which the defendant provided) implicated the defendant’s guilt.271 So with multiple causal factors at issue in the case, when Justice Scalia, writing for the majority, quoted Gross and Nassar, he did it with a twist. He altered the key phrase—“the ‘but-for’ cause”—by bracketing “the” and changing it to “a.”272 Thus, Justice Scalia quoted both Gross and Nassar as stating that to establish but-for causation, the illegal factor must be “[a] ‘but-for’ cause” of the employer’s conduct.273
These alterations would seem to indicate that the Court recognized the implications of its poor wording in Gross and Nassar and was attempting to correct its mistake. But Burrage has not had much impact in the employment discrimination realm. Being a criminal case, Burrage probably did not cross the radar of most employment-law attorneys or courts analyzing discrimination claims. And when it was noticed, courts were reluctant to read too much into the alterations. A couple of courts noted the bracket trick but did not really do much with it.274 One court called it dicta—which it clearly was not.275 A few cited Burrage in employment cases without seeming to notice the changed language.276 And several courts have flat-out rejected plaintiffs’ arguments that Burrage signaled a change in the Gross standard or in any way clarified that but-for cause is not sole cause.277 If Justice Scalia meant Burrage to be an errata sheet for Gross and Nassar, he was too subtle. As the Fourth Circuit stated, “Although we are unsure how to regard the Supreme Court’s alteration of this excerpt from Gross, we suspect that if the Court desired to make a radical change to recent precedent, it would not do so quietly in a case having nothing to do with employment discrimination.”278
Bostock is not nearly so subtle. It is an employment discrimination case, and it unambiguously states, at least thirty-five times, that but-for cause is not sole cause or uses similar language showing the breadth and scope of the but-for standard, especially where multiple causal factors are at issue.279 No reading between the lines is necessary to see that sole cause is off the table. Granted, the Court did not expressly disavow the troublesome Gross and Nassar language. It would have been better to take Justice Kagan’s recent approach where she specifically acknowledged some inelegant and potentially incorrect language from a prior opinion.280 Nonetheless, given Bostock’s clarity and repetition of key language showing that equating but-for causation with sole causation is not legally viable, it is difficult to see a rash of courts misinterpreting it, Pelcha notwithstanding.
E. What About Motivating Factor?
This Article has focused mostly on the promise of a robust but-for causation standard of liability in employment discrimination litigation. But what about the motivating-factor standard? It provides an alternative basis for liability for Title VII status-based discrimination claims281 and is traditionally considered the most plaintiff-friendly causation standard.282 So how does motivating-factor liability fit into the post-Bostock world?
Despite its facial potential and the strong advocacy of antidiscrimination scholars, motivating-factor liability does not seem to have changed the overall success rate for employment discrimination plaintiffs.283 Many explanations have been offered, but there seems to be widespread agreement that one issue comes down to incentives—plaintiffs and their attorneys often do not use the motivating-factor approach because, through the same-decision defense discussed above,284 it can lead the jury to split the baby by handing plaintiffs nominal victories but shutting them out of damages awards.285 With but-for causation seeming much more attractive post-Bostock, plaintiffs are even less likely to choose the motivating-factor route.286
Congress has attempted to broaden motivating-factor liability in employment discrimination statutes. Immediately after Gross, both the House and the Senate introduced bills called the Protecting Older Workers Against Discrimination Act (POWADA).287 They explicitly criticized Gross and would have amended the ADEA to incorporate the same motivating-factor standard as found in Title VII.288 They also would have applied that same standard to “any Federal law forbidding employment discrimination.”289 The bills did not pass.290 Other versions of POWADA have been introduced in every congressional session since,291 all unsuccessful (so far).292 Over time, the bill has kept the same name (focusing on older workers) but expanded to explicitly amend not only the ADEA but also Title VII’s retaliation provision and the Americans with Disabilities Act.293 The latest versions were introduced in March 2021,294 and the House passed its bill in June 2021.295 President Biden intends to sign the law if it makes it through the Senate.296
Even if Congress enacts POWADA, it is unlikely to eclipse the use of but-for causation. By firmly rejecting sole causation and clarifying that multiple but-for causes can exist, Bostock makes clear that but-for causation offers many of the benefits of motivating-factor liability without the burden of its limited remedies.297 Thus, even with POWADA, many plaintiffs will likely find the robust but-for standard Bostock described much more appealing.
Conclusion
As Professor Sullivan has observed, “[F]orecasts of the transformation of Title VII have a habit of being proved wrong.”298 He documented that phenomenon in the context of motivating-factor liability after the Civil Rights Act of 1991, calling it “the revolution that wasn’t.”299 Perhaps the predicted Bostock causation revolution will suffer the same fate.
But that seems unlikely. The Supreme Court unequivocally eviscerated a key defense theory that has hobbled employment discrimination plaintiffs for more than a decade. This theory—that but-for causation is sole causation—has blocked scores of claims based on various combinations of legal and illegal motives. Gross, a Supreme Court decision, caused much of that problem. Surely a later Supreme Court decision that so clearly rejects that theory will have some positive impact. The Supreme Court caused this mess; hopefully now it has fixed it.
Difficulty in proving but-for causation post-Gross is not, of course, solely responsible for plaintiffs’ dismal success rate in employment discrimination cases. Scholars have written about many problems, including unfair evidentiary standards300 and technicalities of the McDonnell Douglas burden-shifting framework that has dominated employment discrimination litigation for more than forty years.301 Bostock cannot, alone, solve all of these problems. But it is certainly a step in the right direction. And it might open the door to less direct but still profound changes.302
Antidiscrimination proponents must do their part to help Bostock change the landscape of employment discrimination litigation. Advocates must carefully and thoroughly brief these arguments for the courts and push for expanded jury instructions incorporating Bostock’s key language. Scholars are thoughtfully and abundantly writing about Bostock’s potential and should continue to do so. I add my voice to theirs.