The Island Defense—When a Discriminatory Workplace Includes an “Innocent” Decisionmaker

Courts employ conflicting approaches when addressing evidence involving the bias of people that the employer alleges are non-­decisionmakers. Where a non-decisionmaker has made discriminatory remarks, some courts will disregard that proof, labelling them “stray” comments, while other courts will find them probative. With respect to comparator or “me too” evidence, some courts will find it irrelevant if it involves people outside of the plaintiff’s department, while other courts will consider it as evidence of a biased corporate atmosphere. Some courts will accept the probative value of statistical disparities from across an entire organization, while other courts would deem such evidence inadmissible unless it is linked to a particular decisionmaker that fired the plaintiff. No overarching standard explains or justifies these variations.

The more restrictive judges believe that an inquiring court must focus on the motivations and perceptions of an actual decision­maker. These courts tend to credit the Island Defense—accepting at face value the employer’s testimony that the discriminatory actors in its workplace had no effect on the plaintiff’s termination. Courts sometimes find “island” decisionmakers in a backhanded fashion by requiring plaintiffs to specifically disprove the employer’s claim that the purported decisionmaker acted alone or without being influenced by their biased peers. In the absence of an on-point rebuttal, these courts adopt the employers’ version of events as the default scenario and ignore evidence of bias located outside that narrative.

On the other hand, more permissive judges assert that evidence of a general workplace atmosphere of discrimination, including the biased actions and statements of non-decisionmakers, is helpful in determining an employer’s willingness to consider impermissible factors. According to these courts, the presence of discriminatory non-­decisionmakers add “color” to the decision involving the plaintiff. Such courts reject the Island Defense, for the purpose of dispositive motions such as summary judgment, because the self-serving testimony typically supporting the defense is the sort of evidence juries are free to reject.

Given this divide, a common standard is needed to provide courts with a guidepost for determining whether the bias of a non-­decisionmaker should be considered relevant and admissible. Only with such a standard can courts deal with the Island Defense in a principled manner. This article seeks to provide that standard—one that will be based on objectively verifiable and reliable factors.

This article will first begin with examples of improperly restrictive court decisions, to give context and scope to the problem. Second, it will establish that a successful discrimination plaintiff is not required to identify a formal decisionmaker or describe the employer’s decisional process, and that she may prove her case with circumstantial evidence available to an “outsider” to the decision. Third, given the foregoing, the article will show that evidence of bias can be probative, regardless of whether it is linked with a decisionmaker. Fourth, it will then explain how evidence of a corporate atmosphere of bias can help demonstrate that a particular decision was discriminatory. Fifth, the article will show how the summary judgment standard is inherently at odds with the Island Defense, in that it relies upon evidence that a jury would be free to reject at trial. Sixth, the article will propose a standard for assessing the relevance of the bias of non-decisionmakers. That is, a jury should be permitted to consider the bias of those who are in “a position to influence” the challenged decision. In addition, a jury should be able to consider the bias of others, if the plaintiff’s situation falls within a greater pattern of conduct that indicates an umbrella of common influence. Seventh, the article will survey the case law to determine who should be considered a decisionmaker and who should be considered to be in a position of influence, for the purpose of applying these standards. I have identified over thirty factual scenarios in which the bias of a non-decisionmaker has been held to be probative, and which are consistent with the framework proposed herein.

We need not believe people are islands just because they say they are. We can look to organizational hierarchy and established lines of communication and allow juries to make reasonable inferences to determine whether a biased actor had influence or whether that bias illuminates a discriminatory workplace culture that shaped the decisionmaker’s decision. Indeed, we should be highly skeptical of the Island Defense, given what we know about group dynamics and decision-making in the modern corporation. Most importantly, we need a common standard so that evidentiary rulings in civil rights cases are principled and adequately implement the important policies that underlie the statutes prohibiting discrimination.

I. How the Island Defense Is Applied

Courts often apply the Island Defense to reject the value of evidence of bias when the employer claims that the biased actor was not involved in the adverse action affecting the employee. Judges reject proof of aggregate statistics, discriminatory statements, comparator evidence, discriminatory actions against other employees, and other indicia of a corporate atmosphere of bias.

These decisions may be characterized as having an overriding focus on proof of bias involving a decisionmaker. One court, for example, has stated that, “[t]o defeat summary judgment in a retaliation case, a plaintiff must point to some evidence of retaliation by a pertinent decisionmaker.” Another court has stated that “the discriminatory intent of which a plaintiff complains must be traceable to the person or persons who made the decision to fire him.” Other courts appear to acknowledge the possibility that the conduct of non-decisionmakers may be admissible, but go on to rule that the evidence in a particular case is unusably weak.

A good illustration of the Island Defense is Thompson v. Coca-Cola Co. In that case, the plaintiff, an African-American of Jamaican origin, worked under Gerald Goodsell, the acting Production Manager. In early 2003, according to the plaintiff, Goodsell stated, “I hate Jamaican music and Jamaicans.” In September 2003, Goodsell said to the plaintiff, “I’ll deal with you, you f[***]ing Jamaican.”

In December 2003, Thompson wanted to go to Jamaica for dental surgery, informed a supervisor, and arranged coverage with another supervisor. The supervisor approved the leave, so long as Goodsell was informed, which he was. In January 2004, four members of management, including Goodsell, confronted Thompson based on the claim that Thompson did not adequately inform Goodsell about the vacation. After this meeting, in which Thompson contested Goodsell’s version of events, a recommendation for termination was sent to the employer’s Separation Review Committee. There is no evidence that the Separation Review Committee conducted any additional investigation before it terminated Thompson.

Despite the fact that it was Goodsell who initiated the complaint about Thompson and attended the meeting in which the termination recommendation was issued, the First Circuit ignored his racist comments as “stray remarks.” The court declared that Goodsell’s role in the termination decision was “minimal” and, at summary judgment, concluded that the Separation Review Committee made its “own independent decision to terminate Thompson.” The court did not specify what, if anything, the Review Committee did to exercise independent judgment. The court simply accepted the employer’s version of events, based on the testimony of the employer’s witnesses that claimed the evidence against them was irrelevant.

The controversy involving the Island Defense is based both on competing notions of relevance of the bias of people with influence in the workplace and how that evidence is handled in the context of summary judgment. Both themes will be considered below, with relevance considered first.

II. Broad Notions of Relevance Permit the Introduction of Evidence of Bias by Those Who May Not Have Participated Directly in the Adverse Action

This section examines the relevance and admissibility of bias on the part of those who may not have participated directly in the adverse action. As a general matter, evidence is both relevant and admissible to the extent that it has any tendency to make a fact that is consequential to the outcome of a case more probable than it would be without the evidence. The standard for relevance is permissive. Furthermore, specific pieces of evidence of bias are not viewed in isolation, but instead are considered in the aggregate with other proof offered by the plaintiff. “Evidence must be considered as a whole, rather than asking whether any particular piece of evidence proves the case by itself.” Thus, the touchstone for admitting evidence is whether it is relevant, and not whether it is dispositive standing alone.

With these liberal standards in mind, I will address the relevance of the bias of non-decisionmakers in three subsections. The first subsection will establish that the plaintiff’s ultimate burden is to show that the employer (not individual actors) engaged in discrimination and that the plaintiff need not establish the identity of the decisionmakers. Second, I show that the plaintiff’s ultimate burden of proof may be satisfied with circumstantial evidence available to an ­ex-employee who stands as an outsider to the employer’s otherwise opaque decision-making process. It follows, therefore, that evidence of discrimination need not be tied to an established decisionmaker, and the article surveys the ample case law in which courts have accepted such evidence. Finally, I will examine the various reasons why evidence which focuses on corporate actors other than the formal decisionmaker helps to prove the existence of employment discrimination.

A. The Plaintiff Is Not Required to Identify a Decisionmaker

There is generally no obligation for a plaintiff to identify decisionmakers or to establish the decisional chain of events that led the employer fire the plaintiff. Federal statutes prohibit “employers” from engaging in discrimination based on protected traits such as race, gender, age, religion, and disability. In cases where the employer is a corporate entity, the statutes set forth no requirement for the plaintiff to lay responsibility at the feet of any particular individual. For example, Title VII states, “It shall be an unlawful employment practice for an employer . . . to . . . discharge any individual . . . because of such individual’s race . . . .” The focus is on the “employer” as the responsible party, and not on any individual.

While a corporate entity can only act through its agents, the plaintiff is not required to identify which agents were involved and how the adverse action was decided. By analogy, if a customer falls on a dangerous set of stairs at a store, he/she is under no obligation to establish the identity of the person who designed the stairs or negligently maintained them. The injured party may still sue if the stairs were unsafe because, in general, the corporate entity is responsible for the misconduct on their premises. Likewise, if the overall evidence permits an inference that race was a causal factor in an employment decision, liability attaches to the employer without an additional showing of the particular person(s) at fault or how the decisional process was tainted.

This makes sense, as the employer’s decision-making process often occurs outside the personal knowledge of plaintiffs or their allies. “There will seldom be ‘eyewitness’ testimony as to the employer’s mental processes.” The employer’s description of its decision-making process is rarely subject to direct, informed challenge—especially its verbal components. To the extent a decisionmaker is identified in a discrimination case, that information is often supplied by the employer with self-serving testimony as part of an effort to exonerate itself.

Identification of the biased decisionmaker is unnecessary to establish liability. For example, assume a company makes its employment decisions with a five-person committee. Assume further that when staffing the last fifty positions at a company, there were nine fully qualified Black applicants and one marginally qualified White applicant, for each position. Also assume that in filling the positions, the company selected the White applicant each time. Clearly this scenario provides more than enough evidence to find race discrimination—even though we do not know precisely how bias corrupted the decision-making process. Are all the members of the committee biased? Is a majority of the committee biased? Is one very influential member biased and the rest innocent or indifferent? Are all committee members innocent, but persuaded by a biased actor who is not on the committee? The plaintiff may never know and, indeed, need not care, because she is free to satisfy her burden of proof with evidence of bias available to an outsider to that process.

The burden-shifting framework of McDonnell Douglas Corp. v. Green provides a strong example of how a plaintiff may prevail in a discrimination case without identifying the decisionmaker or those who influenced the decision. It is the most commonly used analysis for assessing the sufficiency of circumstantial evidence in employment discrimination cases.

The McDonnell Douglas framework allows the plaintiff to raise an inference of discrimination with the production of a prima facie case, along with proof of pretext. The prima facie burden may be satisfied with evidence that the plaintiff [1] had a protected trait (e.g., female); [2] was qualified for her position; [3] was fired; and [4] was replaced. It is not an onerous burden. After satisfying the prima facie requirement, the burden shifts to the employer to articulate a legitimate, non-discriminatory reason for the adverse action. The burden then shifts back to the employee to prove that the employer’s asserted reason is pretextual—i.e., not the actual or complete reason for the adverse action. The plaintiff can prevail without any further evidence linking the pretext to the plaintiff’s protected trait. For example, a jury may find pretext when the employer tries to blame an employee for an issue that the employer knew was not the plaintiff’s responsibility. “[T]he trier of fact can reasonably infer from the falsity of the [employer’s] explanation that the employer is dissembling to cover up a discriminatory purpose.”

McDonnell Douglas reminds us that the employer’s decision-making process is opaque from the employee’s point of view. Under its paradigm, the plaintiff is not obligated to establish who made or influenced the employer’s decision. McDonnell Douglas treats the employer as a unitary entity, without focusing on its individual parts, and examines whether the organization can believably explain why it fired a qualified person from a position that it would like to be filled. It is sufficient to rely on suspicious outcomes and incongruous results from the black box of the employer’s decision-making, as opposed to exposing the contents of that black box itself.

McDonnell Douglas illuminates for us that circumstantial proof of discrimination can come from the perspective of an outsider who is unable to examine the workings of an obscured decisional process. The plaintiff can win just by showing qualification, replacement, and that the employer’s announced explanation is pretextual. Thus, neither the text of the statute, nor the ubiquitous McDonnell Douglas paradigm, imposes a duty on the plaintiff to identify either the decisionmaker or the specific circumstances leading up to the decision.

When a plaintiff cannot describe the precise decision-making process that led to a discriminatory termination, it is not a void to be filled with whatever the employer describes. The plaintiffs may satisfy their ultimate burden of proof in the absence of an accepted narrative of the employer’s decision-making process.

B. Circumstantial Evidence Is Probative Even If Not Tied to a Formal Decisionmaker

Given that a plaintiff may establish liability without identifying the decisionmaker or specifying the decision-making process, it stands to reason that evidence of discrimination may likewise be probative without establishing the identities of the decisionmaker or other influencers (e.g., McDonnell Douglas framework). This point can be illustrated by discussing the differences between direct and circumstantial evidence.

Discrimination may be proven with direct evidence, circumstantial evidence, or a combination of both. Direct evidence is evidence that proves the existence of a fact without relying on inferences. In the context of an employment discrimination claim, direct evidence of discrimination may consist of statements by a decisionmaker that express unlawful bias with respect to the adverse action at issue. When relying on direct evidence, a decisionmaker ordinarily must be identified. Discriminatory remarks made by non-decisionmakers generally do not qualify as direct evidence.

Circumstantial evidence can also be used to prove discrimination. It allows a party to prove discrimination indirectly, based on reasonable inferences from facts. Circumstantial evidence can be more probative of bias than direct evidence. Indeed, indirect evidence is sufficient even in criminal cases, which require proof beyond a reasonable doubt. That being said, weaker circumstantial evidence is also probative under the liberal evidentiary rules, which define “relevance” as a fact that it has any persuasive value at all.

Direct evidence must be linked to a specific decisionmaker, but circumstantial evidence has no such requirement. While remarks of a non-decisionmaker might not be the most probative type of evidence, it may well be a scrap of proof that, when considered as part of a whole (and not in isolation), may reflect that, more likely than not, an unlawful action occurred. Consequently, it is not surprising that many courts have acknowledged the value of evidence involving non-decisionmakers.

Remarks: Discriminatory remarks by non-decisionmakers are a well-recognized category of circumstantial evidence that helps prove a case. A biased remark “uttered by a nondecision maker may be relevant, circumstantial evidence of discrimination.” “Discriminatory remarks, even by a nondecisionmaker, can serve as probative evidence of pretext.”

Statistics: Statistical evidence of discrimination is not rendered irrelevant just because it addresses the workplace composition of an entire department, as opposed to focusing only on the job actions of the alleged decisionmaker.

General Practices and Policies: An employer’s past “general practices and policies” concerning the employment of those in protected categories is relevant to determine whether the treatment of the plaintiff conformed to a general pattern at the company to disfavor particular groups. The reference to “general” practices means that such evidence need not be tethered to an individual decisionmaker.

Comparator and “Me Too” evidence: Courts have admitted comparator evidence and “me too” evidence involving employees located in areas outside the plaintiff’s department, or under different supervisors.

Pattern and Practice: Some courts have held that pattern and practice evidence—proof that discrimination was the standard operating procedure in the entire workplace—can be used in individual disparate treatment cases. Considerations of such breadth are inconsistent with a focus on a particular bad actor.

Corporate Atmosphere: Corporate atmosphere evidence reflects bias throughout a workplace. In Conway v. Electro Switch Corp., in which the employer alleged that the president of the company played no part in the termination decision, the court nevertheless considered the president’s biased remarks, despite the fact that they occurred ten to twenty-two months before the plaintiff’s termination and involved a different employee in a different department. The court stated that “[e]vidence of a corporate state-of-mind or a discriminatory atmosphere is not rendered irrelevant by its failure to coincide precisely with the particular actors . . . involved in the specific events that generated a claim of discriminatory treatment.” The First Circuit has elsewhere held that “[s]tray remarks by nondecisionmakers . . . may still be considered evidence of a company’s general atmosphere of discrimination, and thus can be relevant.”

Thus, circumstantial evidence is probative in discrimination cases, and, applying the liberal rules of relevance, many courts have recognized that evidence involving biased non-decisionmakers is relevant and properly considered.

C. Why Culture Matters

The bias of non-decisionmakers matters because culture matters. The motive that created the plaintiff’s work environment may be illuminated by how others were treated in different contexts. This article will survey some of the reasons why it is appropriate to permit a wide range of evidence, even if it does not involve a decisionmaker.

Mechanism of Influence: By viewing an adverse action in the context of an atmosphere of workplace bias, a factfinder may determine that there is a mechanism effectuating prejudice, even if the specific mechanism remains hidden from view. If five people make similar biased remarks, a jury could reasonably infer the existence of an umbrella of influence and that a contributing factor to such bias is a person or entity that has sway over or across departments. Corporate atmosphere evidence telegraphs “the influences behind the actions taken with respect to the individual plaintiff.”

Contagion: While a corporate atmosphere of bias may indicate that attitudes have been communicated by a workplace authority, it can also indicate a contagion of bias that arises among peers. Workplace culture can show that evidence of bias is pervasive and “infects the company’s personnel decisions.” “Stray” comments may reflect workplace norms, as well as the sense of impunity, entitlement, and support for those expressing bias. Courts routinely accept the notion that discriminatory statements made outside of a plaintiff’s earshot can alter the terms and conditions of their employment. Thus, broad evidence can help establish a workplace’s vulnerability to embracing improper motives, demonstrating a “club” among the in-crowd.

Context: Evidence of workplace culture adds “color” to the employer’s actions by providing the context in which it arose. In the case of Ash v. Tyson Foods, Inc., the Supreme Court considered a case in which the plaintiff was addressed in the workplace as “boy.” The employer argued that “boy” was a race-neutral word that did not indicate a racial animus. The Supreme Court held that, although it is true that the word will not always evidence racial animus, a jury might hold that it could, depending on “context, inflection, tone of voice, local custom and historical usage.” Thus, culture and history—how the word was used by non-decisionmakers—may help establish the full context of a workplace incident. Indeed, stereotypes represent a cultural “piling on,” and statements that at first blush are race or gender neutral nevertheless may been seen as reflections of bias when they are consistent with stereotypes.

Knowledge and Lines of Communication: Where there is institutional awareness of a fact, a jury might infer that the purported decisionmaker also has that knowledge. Corporate culture can evidence usual lines of communication and influence within departments and beyond the formal lines of authority. For example, a corporate atmosphere of bias can demonstrate that a corporation understood that it had a pervasive problem with discrimination and ratified it with inaction.

Continuity: Culture can also establish continuity, which may highlight a discriminatory taint to an ostensibly neutral action. For example, in a 2023 gerrymandering decision, the Supreme Court initiated its discussion of voting rights in Alabama by noting, “For the first 115 years following Reconstruction, the State of Alabama elected no black Representatives to Congress,” and that the first Black representative was elected in 1992 only after a lawsuit forced a change in districting. The Court further noted that “Alabama’s extensive history of repugnant racial and voting-related discrimination is undeniable and well documented.” Why is the Court factoring in the conduct of legislators and voters who have long since passed away? Because culture, and the continuity that arises from it, give important perspective as to how decisions are made in the present.

Process: Culture matters because it may shed light on processes that skew in a discriminatory fashion. A discriminatory decision may be based on a variety of factors, procedures, and cascading decisions leading up to a final adverse action, including “recruitment methods, personnel decision-making practices, heterogeneity of work groups, accountability structures, wage structures, and of course, the overall environment or culture.” For example, a job posting with coded language that tracks stereotypical traits can skew later decision-making in the filling of that position. Thus, corporate atmosphere evidence is necessary to demonstrate how bias may have infected interim steps and processes that led up to the final decision.

Plaintiffs’ attorneys should confidently assert that the bias of non-decisionmakers is relevant, but should also carefully explain why it is relevant. When crafting those explanations, attorneys should freely draw from the above categories and others. It is hoped that courts will become more open to the understanding that the environment in which a decision is made gives important context to how and why that decision is made.

D. Our Understanding of Group Dynamics and the Modern Workplace Confirms the Relevance of Non-Decisionmakers

The movement away from exclusive consideration of the bias of final decisionmakers is supported by our understanding of the modern workplace and group dynamics. The idea that a purported decisionmaker is even capable of accurately attesting that she acted alone or of comprehensively identifying all of her influences is deeply problematic. Indeed, a whole industry has evolved around the idea that people can be influenced without realizing that they are being influenced. “The notion that a single decision maker, or even a discrete group of decision makers, is separable from the environment in which he, she, or it operates, works, and forms opinions, is simplistic and one-dimensional.”

Moreover, a great deal of influence occurs nonverbally—with a grimace, a shake of the head, or other social cues that are unlikely to be reported to the plaintiff as factors on which the decisionmaker relied. Thus, juries must be permitted to examine the landscape of objective information to determine whether there was “implicit influence” that they can infer from organizational positions, reporting relationships, communication patterns, and the physical or relational proximity of biased actors.

Requiring plaintiffs to establish a specific bad actor (with the goal of seeking liability against a corporation based on the conduct of that one actor) runs contrary to the contemporary understanding of how decisions are made in an organization.

One of the current problems in employment discrimination law is that courts view discrimination largely as a “problem of errant or rogue individual discriminators acting contrary to organizational policy and interest.”. . . In some cases, the search for the rogue actor is appropriate; however, in others, the search for the rogue actor asks the wrong question about culpability. It ignores the fact that multi-tiered or group decisionmaking processes may make it difficult or impossible to locate intent within a particular person. . . . [It] disregards the ways that both formal and informal processes and policies within an organization shape the intentions and actions of its individual members, and the ways that the actions and intentions of the individual members shape the organization.

Social science confirms the idea that a corporate atmosphere of bias has a tangible effect on individual decision-making. We know that employees’ beliefs regarding upper management’s commitment to ethical conduct affects employee behavior and that the wrong messages can “socialize individuals into evildoing.” Our behavior is very much influenced by that of our neighbors (e.g., standing ovations as socially influenced, yet intentional conduct), and people often make different decisions when acting as a group than when acting individually.

It is improperly reductionist to consider a discriminatory comment solely to determine its effect on a purported decisionmaker. Rather, such a comment also should be considered to determine whether it mirrored or influenced the culture of the workplace in which the decision took place. It should be seen as possibly indicative of what management indicates it would ratify or countenance. For example, it has been found that discriminatory comments made by white people when by themselves, in private spaces, have an effect on their actions taking place outside the “huddle.” Private “racetalk” has the effect or normalizing discriminatory attitudes, is contagious, and strengthens the solidarity with the in-group against others. An employer’s acquiescence to discrimination elsewhere may lead a discriminatory actor to feel free to discriminate, or even to feel pressure to discriminate where no such prior inclination existed. Thus, an organization should be evaluated both by the individuals acting within it, but also by the atmosphere and attitudes that shape the conduct of the people within it.

We know that corporate culture is an important contributor to actual workplace conduct because employers agree. Companies adopt uniform workplace policies and issue corporate statements precisely because they want to establish a cohesive corporate identity. They establish human resource and human relations departments to encourage consistency and reliability across departments. Business schools retain professors to teach the subject of corporate culture precisely because it is consequential.

The Island Defense represents a deeply flawed public policy because it rewards employers who become increasingly deceptive by adopting artificial layers between the purported decisionmakers and actual deciders and influencers. An employer should not be able to “insulate its decision by interposing an intermediate level of persons in the hierarchy of decision.”

In Staub v. Proctor Hospital, the employer argued that there should be no liability when the final decisionmaker is innocent of any unlawful motive. The Supreme Court rejected this argument, as it would allow employers to game the statute by notionally isolating decisionmakers.

Proctor’s view would have the improbable consequence that if an employer isolates a personnel official from an employee’s supervisors, vests the decision to take adverse employment actions in that official, and asks that official to review the employee’s personnel file before taking adverse action, then the employer will be effectively shielded from discriminatory acts and recommendations of supervisors that were designed and intended to produce the adverse action. That seems to us an implausible meaning of the text, and one that is not compelled by its words.

Thus, the Island Defense runs contrary both to the realities of corporate decision-making processes and to the goals of anti-discrimination laws, which is to weed out bias at any level of a multi-tiered process that results in adverse outcomes.

Ultimately, discrimination is a culture-wide problem; indeed, it is a human problem. When courts accept the Island Defense, they are framing the issue of discrimination as one that involves rogue actors who engage in acts and make decisions that are contrary to the interests of their innocent corporate employers. The search for motive leads the courts, improperly, to look for a specific person. But such an assumption ignores the realities of corporate culture and group dynamics and assumes the candor of “decision-makers” who claim to have been unaware of discriminatory misconduct in their midst. It ignores the fact that group practices can shed light on the true motivation of an individual who is proximate to such a group. It also ignores the fact that the identities of actual decisionmakers and influencers can be easily concealed based on biased testimony. Thus, the next section focuses on the proper treatment of evidence that supports the Island Defense.

III. The Summary Judgment Standard Is Hostile Toward the Island Defense

The summary judgment standard is, in important ways, inconsistent with the Island Defense. As shown above, the Island Defense often prevails because courts rely on the employer’s description of the circumstances and individuals contributing to the termination decision, which the plaintiff is unable to directly refute. Courts dismiss evidence of bias in the workplace based on the employer’s assertions that the biased actors were not involved in a particular decision. Taken to its logical extreme, the Island Defense would be successful even for an employer staffed almost entirely with KKK members, if the purported decisionmaker made no outward expressions of racism and claimed to have acted alone. At the summary judgment stage, the Island Defense should not be dispositive, because testimony of the employer’s witnesses, each of whom have a personal interest in seeing their employer prevail, should not be credited.

Summary judgment is a pre-trial mechanism that permits a defendant/employer to challenge the sufficiency of the evidence supporting the plaintiff/employee’s claim. It asks whether, on the evidence presented, there is a scenario in which a reasonable jury could find for the plaintiff. Summary judgment requires putting as harsh a face on the defendants’ actions as the facts reasonably allow. Courts, at this stage, are supposed to examine the evidence in the light most favorable to the plaintiff, with all reasonable inferences resolved in the plaintiff’s favor, to ascertain if it is possible for the plaintiff to win. It assumes the possibility of a jury that is both reasonable and skeptical of the employer’s evidence.

Courts should exercise caution when granting summary judgment. “[A]ll doubt as to the existence of a genuine issue of material fact must be resolved against the party moving for summary judgment.”

With respect to employment discrimination cases, the summary judgment standard was best described by the Supreme Court in the case of Reeves v. Sanderson Plumbing Products. While Reeves notionally focused on the standard for Rule 50 motions for judgment as a matter of law, it asserted that the standard mirrored the summary judgment standard “such that the inquiry under each is the same.” The Court unanimously held that when considering these dispositive motions, “the court should review the record as a whole . . . [but] it must disregard all evidence favorable to the moving party [employer] that a jury is not required to believe.” The opinion further states that “the court should give credence to the evidence favoring the nonmovant [employee] as well as that evidence supporting the moving party [the employer] that is uncontradicted and unimpeached, at least to the extent that that evidence comes from disinterested witnesses.” Thus, the Supreme Court seems to require judges considering summary judgment motions to assume that the jury will believe the plaintiff and will refuse to credit the testimony of the employer’s witnesses who have an interest in the outcome.

Reeves is correct. At trial, generally, juries are free to disregard the testimony of witnesses that they do not believe. A jury may disregard testimony, even if it is unchallenged or not directly contradicted. A fact finder “is not obliged to credit the evidence or testimony offered by the [defendant] simply because it was uncontradicted.” Thus, a court should not rely on an employer’s evidence at summary judgment that a jury might not credit.

A wrinkle arises in discrimination law, where, as discussed above, employers are allowed to explain their employment decisions pursuant to the McDonnell Douglas analysis, and employees are given the opportunity to show that those explanations are pretextual. Reeves’ statement that courts should “disregard” the testimony of an employer’s interested witnesses seems, at first blush, to run afoul of the well-accepted practice of inviting employers to explain their side at the second stage of the McDonnell Douglas framework and evaluating it for pretext.

However, this seeming contradiction reflects an improper understanding of the Reeves standard. Reeves addressed a conflict among the circuit courts. Some circuits had held that that the proper analysis should consider only evidence favorable to the plaintiff, while others had held that the entire record should be considered. The Supreme Court resolved this conflict by recognizing a distinction between the evidence courts should “review,” as opposed to evidence to which courts should “give credence.” The court recognized that, while courts should “review” all evidence in the record, it should not give credence to the testimony of interested witnesses favoring the employer. Ultimately, the Supreme Court harmonized the competing views of the circuit courts, holding that, “although the court should review the record as a whole, it must disregard all evidence favorable to the moving party that a jury is not required to believe.” In other words, both standards are correct, but, to be correct, both standards have to be considered together.

When the Supreme Court asserts that the evidence favoring the defendant should be disregarded, it does not mean the evidence should be completely ignored. Rather, it should be “reviewed” instead of “credited” to determine whether it supports a finding of pretext. Indeed, the Reeves court had no qualms considering the employer’s explanations for its conduct and evaluating that testimony for evidence of pretext. In sum, courts are supposed to review the entire summary judgment record, but only credit evidence that favors the plaintiff and discredit all evidence disfavoring the plaintiff that a jury would not be required to believe. We are not supposed to completely ignore the employer’s testimony for all purposes.

The Island Defense relies precisely on the type of evidence that should be reviewed, and yet not be relied on by the court at summary judgment. The employer’s argument depends on testimony and affidavits of purported decisionmakers and other agents of the employer who have great incentives to “massage” the truth about who was or was not involved in the decision. These witnesses are usually paid by the employer, and they generally hope to remain employed by the defendant/employer for the foreseeable future. They therefore have a direct financial interest in providing testimony that benefits the employer. Moreover, the employer’s witnesses also have an individual, reputational interest in distancing themselves from unlawful bias, as participating in a discriminatory decision can be stigmatizing.

A number of courts have properly rejected the Island Defense based on this mandated skepticism towards the employer’s narrative. In Walton v. Nalco Chemical Co., the plaintiff, a salesman, claimed he was terminated due to his age. The plaintiff supported his claim with evidence that Yankowski, a vice president, had stated, “[W]e can’t have a man in his sixties sitting on his accounts coasting. We need to get a young rep in there selling business.” The employer asked the court to ignore that statement, claiming that Murphy, another vice president, had acted alone and was not influenced by Yankowski. The court rejected the Island Defense, stating that “the jury was entitled to disbelieve the trial testimony that Richard Murphy unilaterally discharged Walton without consultation or input from Yankowski, who was one of Walton’s superiors and the vice president for Nalco’s eastern sales division.”

In Freeman v. Packaging Machinery Co., the First Circuit held that ageist statements of the company president, who was in a position of authority over the purported decisionmakers, must be considered, despite the employer’s insistence that subordinates made the decision. The court held that “[t]he inquiry into a corporation’s motives need not artificially be limited to the particular officer who carried out the action.” The court concluded that “[t]he battle plan of the admiral is a valid datum in assessing the intentions of the captain of a single ship in the flotilla.”

In Travers v. Flight Services & Systems, Inc., the plaintiff sued his ex-employer, alleging that he was fired in retaliation for initiating a Fair Labor Standards Act (FLSA) lawsuit. There was evidence that, after learning about the employee’s protected conduct, the CEO repeatedly yelled that they needed to get rid of the employee. The employer argued that the CEO’s bias was irrelevant, as there was no evidence that the CEO played a role in the decision to terminate the employee or that those who made the decision were aware of the CEO’s views. The First Circuit rejected this argument, stating that “a rational juror could conclude that such strongly held and repeatedly voiced wishes of the king, so to speak, likely came well known to those couriers who might rid him of a bothersome underling.” Significantly, this ruling was based on the concept that the CEO held a position of influence. “A CEO sets the tone and mission for his subordinates, many of whom presumably consider it an important part of their jobs to figure out and deliver what the CEO wants.” The court concluded that “[i]t is neither irrational nor unfair to infer—if a jury is so inclined—that knowledge of [the CEO’s] directive spread to other managers, themselves likely reluctant to frustrate the CEO’s objective.” Many other cases properly reject the Island Defense by recognizing that a jury need not believe the protests of defense witnesses that biased actors did not influence the decision to fire the plaintiff.

Thus, the summary judgment standard, as properly applied, should militate against the acceptance of the Island Defense. The court should assume that a jury will refuse to credit the employer’s description of the decisional process. And even if the plaintiff fails to proffer evidence of the employer’s decision-making process, the plaintiff’s affirmative evidence of bias should be considered against this blank slate, just as a reasonable juror who disbelieves the employer’s interested witnesses is free to do, because an accepted narrative of the decisional process is unnecessary to a finding of liability.

In other words, we do not know exactly how the CEO’s bias in the Travers case was communicated to a decisionmaker; we only know that a reasonable jury could infer that the message was delivered, based on examination of authority, hierarchy, usual lines of communication, proximity, and professional interest of those acting in the vicinity of the relevant decision. At summary judgment, an employer’s narrative should generally be considered only to the extent that it could demonstrate pretext. It should not be considered an unchallengeable veto on the admissibility of the plaintiff’s evidence.

IV. Resolving the Tension—The “Position of Influence” Standard

Having established that [1] the summary judgment standard is generally inimical to the assertion of the Island Defense, [2] it is not necessary to identify a decisionmaker to establish liability, and [3] the Island Defense does not track the liberal standards that define the relevance of evidence, it should nevertheless be acknowledged that, at some point, proof of bias can be too remote to be probative. A biased statement from fifty years ago or by an actor in another country who never interacted with people in the relevant department may simply be too removed from the plaintiff’s situation to shed light upon it. The question then becomes: where do the courts draw the line? And is that line a gut-check, or is there an objective guidepost that can determine admissibility in a principled, reliable fashion?

While courts must make these decisions on a case-by-case basis, the answer proposed by this article is that, at a minimum, a non-­decisionmaker’s bias should be considered relevant if he or she is “in a position to influence” the decisionmaker. At the summary judgment stage, it should be enough to show that the biased actor “may have influenced” the decision.

This is not a new standard. Courts already recognize that one way to prove pretext is to show that “discriminatory comments were made by . . . those in a position to influence the decisionmaker.” “Evidence that discriminatory remarks were made by persons in a position to influence the challenged employment action may suffice to establish pretext.” What is new is the recognition that this guidepost should be elevated from descriptive gloss, to a common default minimum standard for admitting evidence. (An additional, overlapping standard, involving patterns of discriminatory activity, will be discussed in Part V below.)

The “position of influence” standard properly threads the needle between litigants’ competing arguments about relevance—opening the door to evidence involving non-decisionmakers who are “in the mix,” while providing a principled touchstone for excluding evidence that is too remote. The standard also navigates the summary judgment rubric, which requires us to both consider the self-serving protests of employers that its discriminatory agents had nothing to do with the decision but to credit a pro-plaintiff narrative that a reasonable, yet skeptical jury could find based on the plaintiffs’ affirmative evidence supporting their claims. The “position of influence” standard is based on objective, verifiable evidence that the non-decisionmaker “may have influenced” the decision. The standard is consistent with what we know about organizational and group dynamics and provides a necessary tool to uncover motives that employers try to conceal.

The “position of influence” standard has important benefits. Under that standard, the plaintiff need not come up with direct evidence that the biased actor played a part in the decision, other than to establish that the biased actor was in a position to do so. In determining the capacity to influence, it is proper to examine objective information, such as the actor’s authority, access to the decisionmakers, relationship with the decisionmakers, usual lines of communication, and role within the company. Moreover, when it is established that a biased actor has a position of influence, it is the jury that properly determines whether the actor in fact influenced the decision.

For example, in a case where a decisionmaker met with biased managerial officials about the plaintiff’s termination, the court held that “a rational jury could infer that the wishes of senior management were expressed in those meetings.” Or, in the alternative, it could be that instead of directing or instructing the decisionmaker, the non-­decisionmaker exercised influence in a more passive way, and that, by expressing bias in the workplace, this led the decisionmaker to believe that it would be better, more convenient, or create more harmony to prefer a candidate based on an unlawful factor.

Focusing on the biased actor’s objective position with respect to authority, physical location, access, personal or professional interest, or responsibility is a reasonable basis for a finding of actual influence, despite the employer’s denials. For example, where a decisionmaker denies knowledge of an employee’s protected activity, courts have often permitted jurors to infer such knowledge based on the decisionmaker’s proximity and reporting relationship with those who have knowledge. If a jury can infer knowledge (and thus, the existence of a communication that is otherwise denied), then it can just as easily infer the transfer of other information that influenced an employment decision.

Corporate hierarchy is a trustworthy factor in such a consideration. For example, when an employer maintains an organizational chart, jurors should be free to believe that the chart has consequential meaning and be entitled to consider the bias of those whose positions fall under an umbrella of common authority, share a responsibility, or maintain systemic lines of communication.

Focusing on the biased actor’s position is especially appropriate when considering a motion for summary judgment, where the employer’s testimony is not credited and the plaintiff does not have personal knowledge of the decision-making process. Such a standard holds employers properly responsible for placing biased individuals in positions where they can influence personnel decisions. Thus, the proposed standard (along with another standard discussed in Part V) addresses and satisfies the legitimate concerns of all parties, when employers raise the Island Defense.

A. Depending on the Theory of Liability, the Concept of “Decisionmaker” Can Be Broad

Having established that the bias of decisionmakers and those in a position of influence is generally relevant, the next step is to develop a practical understanding of the types of individuals that fall within each of these categories. This article will first focus on the decisionmakers.

There is no precise legal definition of “decisionmaker.” Depending on the theory of liability, the class of people properly identified as a “decisionmaker” can be quite broad. The archetype of an unlawful termination is that a direct manager, vested with authority to make personnel decisions, decides alone, in a spasm of discriminatory animus, to terminate an employee. Under this bare-bones scenario, an employer would argue that the court must focus on the bias of this one actor and ignore everyone else.

However, this picture of a personnel decision is too simplistic to match the realities of the modern workplace. Most corporate decisions involve a chain of process that invites the input of many individuals. The contemporary workplace is a creature of trainings, personnel practices and policies, progressive discipline, documented evaluations, managerial approvals, mentors, performance improvement plans, personnel committees, and human resources oversight. “An employer’s authority to reward, punish, or dismiss is often allocated among multiple agents.” Consider, also, the complication of unionized settings, tenure procedures, and IRS and SEC regulations that impose multiple levels of review and investigation.

Under Title VII, liability is established where bias taints any part of the decision-making process, even when it does not initiate the process. Discriminatory mindsets might not necessarily initiate adverse actions, but they can skew the decision-making process once underway. Thus, a biased actor can be seen as within the “chain of decisionmakers”—perhaps as an investigator, manager, policymaker, or evaluator—even as the employer contends that the actor is a non-decisionmaker.

Where decisions are made by a committee, all biased committee members are deemed to be decisionmakers, even if they do not represent a majority of the committee or cannot themselves determine the outcome. Courts also recognize that a corporate actor without the formal levers of power may be the one actually pulling those levers.

Under the “cat’s paw” theory of liability, the bias of those influencing the decision is imputed to the employer as if they were the decisionmakers. According to this rubric, the employer is liable where an innocent final decisionmakers relied upon the falsely negative recommendations or evaluations of discriminatory actors. The cat’s paw theory can also apply where a biased actor provides the employer with truthful information but with a nefarious motive or where the biased actor withholds exculpatory information supporting the employee. It also applies when an innocent final decisionmaker makes decisions based on ostensibly neutral criteria assigned by a superior when those criteria were developed to effectuate biased goals.

The focus in such cases is not necessarily whether the biased actor discussed or otherwise infected the final decisionmaker with discriminatory animus, but whether the biased actor manipulated the process or the information on which the final decisionmaker relied. Indeed, where such manipulation has occurred, it is not necessarily a viable defense that an innocent decisionmaker exercised their own judgment and performed an independent investigation.

An employee may also recover under a chain-of-causation theory of liability, whereby an initial discriminatory act leads to a string of circumstances that ultimately result in an adverse action, despite the intervening conduct of innocent actors. For example, in Brookline v. Alston, a Black firefighter was subjected to racial harassment and later retaliation to which the employer acquiesced, and, as a result, the firefighter went on medical leave. The employer repeatedly and reasonably attempted to coax the firefighter to return to work, and to try to schedule a fitness for duty examination, but the firefighter refused to respond based on the medical consequences of the original harassment and retaliation. The employer terminated the firefighter for failing to comply with return-to-work procedures. The court held that, since the employer had initiated a chain of causation commencing with harassment and retaliation, it was responsible for wrongful termination when it fired the firefighter for failing to comply with its ostensibly reasonable return-to-work efforts.

Thus, terminations are not seen as a single, isolated decision, but, rather, the result of a process or chain of events, actions, and omissions. At each step in such process is an individual or group of individuals who can be said to have made determinative decisions that led to the adverse action. Consequently, the set of individuals who can be identified as decisionmakers is quite broad.

B. Many Actors Can Be in a Position to Influence a Decision

In addition to decisionmakers, courts have declared that the discriminatory attitudes of people who may have influenced the decision are also considered. At what point can it be said that someone is in a position to influence, such that their bias becomes germane to a wrongful discharge? An influencer is someone who was “involved” in the decision, or who can be said to have “tainted” or “exerted influence” over the ultimate decision, or who “possessed leverage” over the formal decisionmaker.

A broad understanding of the relevance of influencers dovetails with the hearsay exception, admissions of party opponents, which permits the introduction of statements of managerial-level employees who have the ability to influence a personnel decision, even if they are not the direct decisionmakers. In other words, the statement of an individual regarding an issue that falls within the scope of his/her authority is considered probative, even if they did not make the decision themself. “[A] statement may be admissible under [the Federal Rules of Evidence], regardless of whether the declarant was a decisionmaker, so long as the declarant was involved in the process leading to the challenged decision.”

And just as the definition of “decisionmaker” can be broad, so too is the class of individuals who are in a position to influence a decisionmaker. While there is no established definition of an “influencer,” case law provides many examples, some of which are listed below:

• Appointer of Decisionmaker
• Benefited from Job Action Against the Plaintiff
• Close Relationship to Decisionmaker
• Committee Members and Those Who Influence Them
• Consulted in Decision
• Co-Worker
• Designer of Ranking Procedures
• Evaluator of Employee
• Family Member of Decisionmaker
• Former Positions of Power
• Human Resources Personnel
• Informal Authority
• Investigation Initiator
• Investigation Participant
• Leverage over Decisionmaker
• Manager Above Decisionmaker
• Messenger of Employer’s Intent
• Participated in Decision
• Peer of Plaintiff
• Presented the Results of Decision-making Process to Corporate Review Committee
• President or Other Top Officers
• Pretext—Involvement with Employer’s Pretextual Explanation for Termination
• Provided Information Relied upon by the Decisionmaker
• Provided Information Relied upon by the person who Recommended Discharge
• Power to Make Employment Decisions
• Recommended Changes to Qualification Standards
• Recommended the Termination
• Recommended the Layoff/Reorganization That Affected the Plaintiff
• Refused to Work with Plaintiff
• Subordinate to Plaintiff
• Supervisor—Direct
• Trusted by Decisionmaker
• Veto—Power to Veto the Termination

Thus, a plaintiff may support her claims with evidence of bias on the part of anyone surrounding, supporting, or advising the decisionmaker, or those benefiting from the adverse action. By operation of the same notions of relevance, plaintiffs should be able to submit evidence of bias of those who can influence the influencers, as well. Thus, the universe of relevant influencers is broad.

V. Evidence That Reflects Formal or Informal Managerial Attitudes, or Places the Plaintiff Within the Context of a Pattern of Discriminatory Conduct Should Be Considered

The “position of influence” should not be the only touchstone for admissibility, as it is clear that proof from an even wider set of biased actors in the workplace is relevant.

Bias expressed by those who are foreign to the decision may reflect the discriminatory attitudes of those who could have participated in the decision. For example, in the case of Ryder v. Westinghouse Electric Corp., the plaintiff pursued a claim of age discrimination against his former employer. As support, the plaintiff sought to introduce evidence of a memorandum authored by Westinghouse’s CEO, Michael Jordon, which reflected comments made at a meeting occurring a year after the plaintiff’s termination. Jordon was not involved with the decision to fire the plaintiff, became CEO two months after the plaintiff was fired, and wrote the memorandum a year after the plaintiff was fired. Moreover, the speakers at the meeting discussed in the memorandum were likewise not involved in the plaintiff’s termination. Nevertheless, the court found the memorandum to be admissible circumstantial proof of intentional discrimination, as it reflected a cumulative statement of “formal or informal managerial attitudes held by corporate executives.” The court concluded that “if the jury were to believe that these comments accurately reflected a then-existing managerial attitude toward older workers in April 1993, this evidence would make the existence of an improper motive for Ryder’s termination more probable.” Thus, the Ryder case stands for the principle that the bias of corporate actors can be relevant, even if they were not in a position to influence the decision.

Moreover, the actions of corporate actors who are not in a position to influence can be evidence of a pattern of bias, in which the plaintiff’s treatment can be seen as a part. For example, the Supreme Court has stated that “statistics as to [an employer’s] employment policy and practice may be helpful to a determination of whether [the employer’s] refusal to rehire respondent in this case conformed to a general pattern of discrimination against blacks.”

When there is a widespread practice to effectuate biased decisions and the plaintiff is subject to a disadvantage consistent with that biased atmosphere, a jury is free to find that the plaintiff was the victim of an underlying or cultural influence spanning across an employer’s decisional units. A broad pattern of bias may reflect an organization-wide acquiescence to misconduct, reflecting affirmative ratification of discrimination, or, at the least, modelling behavior that the company is willing to abide. Therefore, it is sometimes proper to consider the bias expressed by non-decisionmakers, even when they are not in a position to influence the plaintiff’s circumstances.

Conclusion

It may be true that evidence of bias in those who admit to making a decision can be viewed as stronger than evidence of bias of those whose role is uncertain. However, weaker evidence of discrimination can be described accurately and more pithily as simply “evidence of discrimination.” Evidence is relevant, even if it only slightly makes the ultimate finding more likely. Evidence of discrimination does not lose its value simply because is not the strongest of all possible evidence.

Courts must consider the admission of evidence involving non-­decisionmakers on a case-by-case basis. However, at a minimum, evidence of discriminatory attitudes of others in the workplace should be presented to juries to the extent a biased actor was in the position to make or influence the adverse decision, or a pattern of bias in the workplace is consistent with the plaintiff’s experience. Courts should exercise their discretion to admit other types of workplace bias evidence even when it falls outside these categories.

Plaintiffs must be permitted to use the realities and objective organization of a workplace to define the range of evidence that should be considered. A purported decisionmaker’s testimony concerning his/her subjective state of mind or lack of influence should not hold sway where the objective structure of workplace relationships permits an inference of influence or acquiescence, or where the plaintiff’s experiences appear to resonate across organizational structures.

Plaintiffs are not required to identify the specific process the led to the adverse action, nor the specific people involved. When a plaintiff fails to identify and describe such a process, it is not appropriate to permit the employer’s description of the process to act as the default narrative. At the summary judgment stage, an employer’s denial of discriminatory motive is simply not dispositive, nor should be its description of its decisional process. Instead, the employer’s version of events should not be credited, and the only burden the plaintiff should bear is to identify affirmative evidence of discrimination that satisfies her ultimate burden of proof—to show that is more likely than not that discrimination occurred. It is not the plaintiff’s job to show precisely how the discriminatory decision was made, or how bias was communicated or modelled throughout the workplace. The plaintiff should be permitted to proceed to trial based on evidence available to an outsider to the employer’s decisional process. Evidence of bias throughout the workplace should not be deemed irrelevant just because the employer’s self-interested witnesses testify that it had no influence.