Acknowledging informal power dynamics in the workplace: a proposal for further development of the vicarious liability doctrine in hostile environment sexual harassment cases.

I. INTRODUCTION

In this Article, I evaluate courts' application of the U.S. Supreme Court's affirmative defense doctrine in hostile environment sexual harassment cases. This doctrine provides that employers may avoid being held vicariously liable for hostile environment sexual harassment by supervisors if they can establish that: (1) they have taken reasonable measures to prevent sexual harassment, including setting up adequate complaint procedures, and (2) the employee who suffered sexual harassment unreasonably failed to avail herself of these procedures. The affirmative defense doctrine is U.S. Supreme Court-made law, developed in a series of cases discussed below.

The affirmative defense doctrine has merits. As the Court and commentators have correctly pointed out, the availability of an affirmative defense to employer vicarious liability in supervisor hostile environment sexual harassment cases serves important policy objectives. Chief among those objectives is the creation of incentives for employers to design and implement policies that will deter and punish sexual harassment at the workplace level, avoiding the need to involve the courts. (1) The power of the Court to shape employers' policies through the law is demonstrated by the cottage industry of sexual harassment training that arose after the Supreme Court articulated the affirmative defense doctrine, through which employment lawyers and other consultants have done good business advising employers about implementing sexual harassment policies. (2) Having been involved in some of that work myself, I have no complaints about the incentive-creating dimension of the Court's affirmative defense doctrine. On the other hand, I have increasing reservations about the way the courts have applied the affirmative defense doctrine in subsequent case law, and that is my topic here.

I will argue that the courts' current, extremely confused and contorted articulation of that doctrine contravenes the policy underlying recognition of hostile environment sexual harassment as a form of sex discrimination. I do so by discussing in detail several recent examples of courts' application of the affirmative defense doctrine. I then draw on the excellent, burgeoning literature on sexual harassment law and on the research of experts who study organizational dynamics to argue that courts' application of the affirmative defense writes out of sexual harassment law concern for the operation of informal power dynamics in the workplace. I propose an alternative approach that would call on courts to engage in a more searching inquiry into the ways in which power dynamics in the workplace may prevent persons who have suffered sexual harassment from making effective use of sexual harassment policies. Stated more simply, my argument is that employers should be held vicariously liable for the actions of employees who commit sexual harassment that is sufficiently severe to constitute hostile environment sex discrimination when those employees have abused the power granted to them by their agency relationship to the employer. If employees abuse power granted to them by their employer by carrying out sexual harassment--by credibly threatening retaliation, ordering an employee to carry out particular acts or otherwise exercising credible intimidation tactics--they have been "aided in the agency" in carrying out their harassment by the power conferred on them by their employer. To determine whether a harasser has used employer-granted power in this way, courts should place far greater weight on evidence reflecting the power dynamics in particular workplaces.

To be sure, the approach I advocate demands a far more searching inquiry than that courts typically engage in when considering employers' affirmative defenses. But it conforms to the Supreme Court's test in Ellerth, which asks whether the employer's sexual harassment prevention policies are effective and whether the plaintiff's failure to report or otherwise avoid the harassment is reasonable. Employer sexual harassment policies that allow employees to use power conferred on them by virtue of their agency relationship with their employer to carry out sexual harassment are not effective in deterring sexual harassment within the meaning of Ellerth, and plaintiffs who submit to harassment or do not seek redress for it under conditions that permit harassers to abuse their employer-granted power in such a manner have not acted unreasonably. Finally, I demonstrate that this proposal is feasible by pointing to examples of courts that have engaged in such searching, fact-sensitive analyses.

II. THE ELLERTH AFFIRMATIVE DEFENSE

As already noted, the affirmative defense to supervisor hostile environment sexual harassment arises from a series of cases in which the Supreme Court first recognized the theory of hostile environment sexual harassment, but then sought to develop doctrines to limit employers' vicarious liability for the actions of their supervisory employees.

A. The Beginnings of Hostile Environment Sexual Harassment Law

The key U.S. Supreme Court case recognizing hostile environment sexual harassment as an actionable claim under Title VII (3) was Meritor Savings Bank v. Vinson. (4) It is worth revisiting the facts of this landmark case because, under subsequent doctrinal developments, Vinson might not have a strong case today. Vinson worked at a bank. (5) Bank vice president Sidney Taylor had hired her and was her supervisor. (6) For the four years during which Vinson worked under Taylor, he made repeated requests for sexual favors, leading to forty to fifty incidents of sexual intercourse. (7) Vinson initially resisted the requests for sex but eventually gave in out of what she described as fear of losing her job. (8) In the words of the Court: "[Vinson] testified that because she was afraid of Taylor she never reported the harassment to any of his supervisors and never attempted to use the bank's complaint procedure." (9)

The District Court for the District of Columbia found that the bank could not be held liable for Taylor's actions. (10) The court noted that the bank had an express policy against discrimination, but that neither Vinson nor any of the other employees whom Taylor had sexually harassed had ever lodged a sexual harassment complaint under the bank's complaint procedures. (11) The court thus concluded that "the bank was without notice and cannot be held liable for the alleged actions of Taylor." (12) The D.C. Circuit reversed, holding that employers should be absolutely liable for sexual harassment practiced by supervisory personnel, regardless of whether the employer knew or should have known about it. (13)

The U.S. Supreme Court agreed in part with the D.C. Circuit and ruled on several important issues related to sexual harassment liability. First, it held that, "[w]ithout question, when a supervisor sexually harasses a subordinate because of the subordinate's sex, that supervisor 'discriminate[s]' on the basis of sex." (14) Second, the Court rejected the bank's argument that sexual harassment by a supervisor should be actionable under Title VII only when it involves tangible loss of an economic character. (15) The Court held that Congress's intent was to "strike at the entire spectrum of disparate treatment," including not only "the grant or denial of an economic quid pro quo," but also "so-called hostile environment" harassment, when the harassment was "sufficiently severe or pervasive 'to alter the conditions of the victim's employment and create an abusive working environment.'" (16)

Finally, the Court began the first of what would become a series of increasingly complex discussions about the agency principles that should apply in determining an employer's vicarious liability for a supervisor's sexual harassment in hostile environment cases. The Court noted that Congress had defined the term "employer" in Title VII to include any "agent" of an employer. (17) But the Court reasoned that Congress had not intended to render employers always automatically liable for sexual harassment by supervisors. (18) Conversely, the Court stated, absence of notice to an employer should not necessarily protect the employer from liability. (19) The Court thus rejected the bank's argument that it should be insulated from liability because it had a complaint procedure and a policy against discrimination, which Vinson failed to take advantage of. (20) The complaint procedure and anti-discrimination policy were general and did not alert employees of their employer's interest in stopping the particular form of discrimination involved in sexual harassment. (21) Furthermore, the Court noted, the complaint procedure would have required Vinson to bring her complaint first to the attention of her supervisor, who was the very person committing the harassment. (22) The Court thus affirmed the D.C. Circuit's reversal of the trial court's grant of summary judgment to the employer, (23) and left to another day the project of issuing definitive rules on employers' vicarious liability in supervisor hostile environment cases.

B. Ellerth and its Aftermath

The next set of Supreme Court cases to lay important new ground in defining the scope of employers' vicarious liability in supervisor hostile environment cases are the companion cases of Burlington Industries Inc. v. Ellerth (24) and Faragher v. City of Boca Raton. (25) In Ellerth, a female salesperson alleged that she had been subjected to constant sexual harassment by her supervisor, Ted Slowik. (26) Slowik was a "mid-level" manager, with authority to make hiring and promotion decisions subject to the approval of his supervisor, who signed the paperwork. (27) He was not Ellerth's immediate supervisor. (28) Ellerth answered to her office colleague in the Chicago office, who in turn answered to Slowik in New York. (29)

Slowik made various boorish and sexual comments to Ellerth and told her several times to "loosen up" and dress more sexually if she wanted her career in the company to go well. (30) He also expressed reservations about Ellerth while interviewing her for a promotion, because she was not "loose enough," although she did receive the promotion in the end. (31) After an incident in which Slowik told Ellerth that wearing shorter skirts would make her job a lot easier, Ellerth's immediate supervisor warned her about failing to return phone calls to customers promptly. (32) In response, Ellerth quit, and a short time later alleged that Slowik had been sexually harassing her. (33) During her period of employment at Burlington, however, Ellerth had not informed anyone in authority about Slowik's conduct, despite knowing that Burlington had a policy against sexual harassment. (34) She did not tell her immediate supervisor because she thought he would have to report her complaint to his supervisor, who was Slowik. (35)

On summary judgment, the district court found that Ellerth's allegations met the severe and pervasive standard for hostile environment sexual harassment cases, but ruled against Ellerth on vicarious liability grounds. (36) On appeal, the Seventh Circuit, sitting en banc, produced eight separate opinions and no consensus for a controlling rationale on the agency principles that should govern employer vicarious liability in supervisor hostile environment cases. (37)

The Supreme Court, in its majority opinion drafted by Justice Kennedy, announced an affirmative defense to employer vicarious liability in such cases. The Court recognized that common law agency principles could result in the broad imposition of vicarious liability on an employer for employee wrongdoing, noting that "the concept of scope of employment has not always been construed to require a motive to serve the employer." (38) Nevertheless, the Court chose to follow a line of doctrine that looks to whether an agent's acts are for the purpose of serving the employer or for personal purposes. (39) In the latter case, vicarious liability on the part of the principal may not apply. (40) Since "a supervisor acting out of gender-based animus or a desire to fulfill sexual urges may not be actuated by a purpose to serve the employer" (41) but rather out of personal motives, the Court reasoned that sexual harassment by a supervisor is not conduct within the scope of employment. (42)

The Court also briefly discussed the "aided in the agency" standard for imposing vicarious liability under section 219(2)(d) of the Restatement (Second) of Agency. (43) The Court reasoned that such an analysis arguably could apply to all workplace sexual harassment, since harassers are aided in accomplishing their tortious objective by "[p]roximity and regular contact" with a "pool of potential victims." (44) But the Court reasoned that this alone could not be enough, because it would render employers vicariously liable for all co-worker and supervisor harassment. (45) On the other hand, the Court reasoned, supervisors who had taken a "tangible employment action" against an employee who was being sexually harassed clearly had been aided by the agency relationship with the employer, or principal, since they had used the authority granted to them by the employer to punish the harassee by causing tangible injury related to the harassee's employment. (46)

The Court in Ellerth noted that Slowik had threatened to take negative employment action against Ellerth as a negative quid pro quo for refusing his sexual requests, but he had not carried out those threats. (47) The developing jurisprudence in sexual harassment cases imposed automatic liability on employers in quid pro quo cases, on the theory that in such cases supervisors clearly were acting as agents of the employer and using their authority, vested in them by virtue of their supervisory powers within the company, to extract sex from a subordinate. (48) This doctrine, the Court noted, encouraged plaintiffs to plead their cases as quid pro quo claims, and thus had put expansive pressure on the definition of quid pro quo sexual harassment. (49) The Court resisted further expansion of the quid pro quo concept. (50) The mere threat of negative employment action, the Ellerth Court reasoned, should not be enough to impose automatic liability on the employer, because in such cases the facts do not show that the supervisor did in fact take action as an agent of the employer. (51) On this reasoning, the Court concluded that automatic employer liability should be reserved for cases involving tangible employment action, which "in most cases inflicts direct economic harm" (52) and is "an official company act." (53) These harms, the Court reasoned, are ones that only supervisors can inflict. (54)

I dispute this assumption later in this Article, but even if it were true, it does not as a matter of logic follow that all a supervisor can do to alter the terms and conditions of a harrasee's employment is inflict a tangible employment action. An employee with power in the workplace can make the work life of another employee miserable, and thus alter that employee's "terms and conditions of employment," in many ways other than those encompassed by the concept of tangible employment action. In any event, the Court's introduction of the term "tangible employment action" to define the cases that warrant automatic employer liability did not advance the law very far: Cases involving tangible employment actions usually fall within the quid pro quo category, and the Court's earlier quid pro quo analysis had already established that automatic liability would apply. (55)

Moreover, the tangible employment action test is inconsistent with the theory of why hostile environment sexual harassment constitutes a cause of action under Title VII in the first place. The question of whether a plaintiff suffered a tangible employment action does not address the key issue underlying hostile environment claims. That issue, as Vinson established, is whether the harassment was sufficiently severe as to alter the terms and conditions of the employee's employment such that he or she suffered discrimination on account of sex. (56) Actions other than tangible employment actions--including threats of tangible employment actions, other kinds of significant intimidation, and severe harassment itself--clearly can alter the terms and conditions of employment that a sexually harassed employee has to endure. Employees subject to severe and pervasive sexual harassment have to contend with materially worse conditions of employment than do employees who are not subject to such harassment on the basis of their sex. This is why hostile environment sexual harassment can give rise to an actionable Title VII claim. But the Court in Ellerth gave little attention to this fundamental underpinning of the theory as to why hostile environment sexual harassment constitutes a form of sex discrimination for which employers should be liable.

Instead of ending the scheme of proof for hostile environment cases with the plaintiff's showing of harassment sufficiently severe and pervasive to alter the terms and conditions of employment, the Court adopted an affirmative defense proposed by the Equal Employment Opportunity Commission (EEOC) (57) and approved by some lower courts in prior cases. The affirmative defense allows employers to avoid vicarious liability in cases involving no tangible employment action if they can prove by a preponderance of the evidence the following two elements:

(a) that the employer exercised reasonable care to prevent and correct promptly any sexually harassing behavior, and (b) that the plaintiff employee unreasonably failed to take advantage of any preventive or corrective opportunities provided by the employer or to avoid harm otherwise. (58)

This language appears sufficiently broad to support a searching, context-specific examination of the reasonableness of the employer's care and of the plaintiff's efforts to avoid harm. What "reasonable" steps employers are to take is open to further development, as is the question of when employees act "unreasonably" in failing to avail themselves of preventive or corrective opportunities. Indeed, the Court's companion case, Faragher, easily concluded the same day that a city pool's flimsy sexual harassment procedure did not pass muster under the Ellerth affirmative defense doctrine. (59) But the Court's next major case, Pennsylvania State Police v. Suders (60) presented a less encouraging picture.

The central issue on review in Suders was whether a constructive discharge--that is, an employee's reasonable resignation in response to intolerable working conditions--should be equated with a tangible employment action, so that the affirmative defense doctrine would not be available in constructive discharge cases. (61) The majority opinion, written by Justice Ginsberg, concluded that constructive discharge was not a tangible employment action in itself; rather courts should examine the facts alleged to determine whether the working conditions under which the employee reasonably felt compelled to resign included tangible employment actions such as a demotion. (62) In so concluding, Justice Ginsberg admonished courts not to engage in the formalist logic inherent in the argument that a constructive discharge, being a form of discharge, necessarily and always should be categorized as a tangible employment action. (63)

Many analysts have pondered, both before and after Suders, how best to classify constructive discharge for purposes of sexual harassment law. (64) The constructive discharge issue is a question related to my topic, since it involves, in Justice Ginsberg's apt words, hostile environment harassment "ratcheted up to the breaking point." (65) Rather than becoming sidetracked by the constructive discharge debate, however, I will focus on Suders' general discussion of the vicarious liability principles that apply to supervisor hostile environment sexual harassment. Quite apart from its treatment of the constructive discharge issue, Suders has had the effect of constricting still further plaintiffs' ability to survive employers' motions for summary judgment based on the affirmative defense to vicarious liability.

In some of its language in Suders, the Court appeared to replace, or at least to gloss, the more carefully nuanced language of Ellerth quoted above, which emphasized a reasonableness analysis, with cruder bright line rules. The Court articulated the two prongs of the affirmative defense for cases not involving an "employer-sanctioned adverse action officially changing her employment status," (66) as follows:

   An employer [must show] both (1) that it had installed a readily
   accessible and effective policy for reporting and resolving
   complaints of sexual harassment, and (2) that the plaintiff
   unreasonably failed to avail herself of that employer-provided
   preventative or remedial apparatus. (67)

With this language, the Court, perhaps unintentionally, appeared to signal a willingness to routinize, or make more perfunctory, the analysis it expects of the lower courts in evaluating employers' affirmative defense claims. The focus is on whether the employer has a sex harassment policy in place and whether the plaintiff failed to use it. To be sure, the Court in Suders uses the terms "readily accessible" and "effective" in describing the policies employers must have for reporting and resolving sexual harassment complaints, and also repeats the requirement that courts must inquire into the reasonableness of the plaintiff's decisions about using such procedures. But lower courts have deemphasized these descriptors and have looked simply at whether the employer has a policy and whether the plaintiff used it. This unduly limited inquiry typifies many lower courts' application of the affirmative defense, in cases decided both before and after Suders, as I discuss below. Courts use the affirmative defense to dismiss plaintiffs' cases at the summary judgment stage even when there are important facts in dispute about the adequacy and effectiveness of employer sexual harassment policies.

C. How the Courts Are Getting it Wrong: Cases Following Ellerth

My informal survey of the case law shows that, in numerous cases across the country, plaintiffs are losing sexual harassment cases at the summary judgment stage based on the employer's assertion of the Ellerth affirmative defense, even though they have alleged facts that should allow them to recover if proven at trial. (68) This impressionistic finding is consistent with the findings of other scholars who are critical of Ellerth and of the development of sexual harassment law in general. (69) Under the current state of the law, all an employer needs to do to avoid vicarious liability is to show that it has implemented a standard sexual harassment policy and that the plaintiff did not use it. (70) If courts spend any time at all examining a plaintiff's allegations that it was not reasonable to resort to the sexual harassment policy under the circumstances, which they generally do not, they discuss this question simply to dismiss the plaintiff's assertions. (71)

I will discuss just a few examples in detail that illustrate the general picture of what is occurring in the lower courts' implementation of the Ellerth affirmative defense doctrine. For ease of analysis, I divide my discussion into supervisor and co-worker cases, although an important part of my argument in Part III questions the advisability of continuing this doctrinal distinction.

1. Supervisor Cases

Here I discuss three factual scenarios involving supervisor sexual harassment in which I believe courts have wrongly applied the affirmative defense doctrine. First, consider the facts in Jones v. USA Petroleum Corp. (72) Each of the two female plaintiffs in that case had held the job of night shift cashier at a gas station. (73) One employee worked for three weeks and then quit; the second took her place and worked for approximately three months before quitting. (74) The position required each plaintiff to work alone on the overnight shift in a small booth measuring seven feet by seven feet. (75) Toward the end of their shift, their supervisor, the station manager, would come in to work in the cramped booth with them. (76) The plaintiffs both alleged that this station manager rubbed himself against them and made verbal comments of a sexual nature, including profanity, name calling and sexual insinuations. (77) One plaintiff alleged that he forced open the door and forcibly kissed her while she was using the station's bathroom. (78) As already noted, each plaintiff lasted only a short time on the job, and neither reported the harassment until after she resigned, even though both admitted that they had signed a form acknowledging that they had been informed of the company's sexual harassment procedure, which required them to phone a complaint to the company's personnel manager, located at another site. (79) Neither employee was given a copy of the policy, however, nor was this policy or the phone number of the personnel manager posted in the station. (80) Nevertheless, the court granted summary judgment to the employer on the ground that it had an adequate policy and the plaintiffs had failed to use it. (81)

This is obviously an incorrect result. Applying the reasonableness standard the Court articulated in Ellerth, one must ask, from the standpoint of the objective reasonable person in the circumstances, whether a rational person can be expected to continue to work alone, at night, in a tiny, confined space with a supervisor engaged in escalating, physically threatening sexual harassment, after having reported the supervisor for misconduct to an unknown person at a distant location. Surely in these particular circumstances, a finder of fact could reasonably conclude that quitting and then complaining was reasonable conduct on the part of the plaintiffs. In other words, the court in Jones misapplied the second prong of the Court's affirmative defense: the employer had not established for purposes of summary judgment that a reasonable fact finder would reject plaintiffs' claim that their failure to report their supervisor's misconduct before quitting was reasonable under the circumstances. (82)

As a second example, consider Madray v. Publix Supermarkets, Inc., (83) in which two female store clerks filed suit for hostile environment sexual harassment against Publix Supermarkets after they were subjected to an escalating problem of inappropriate touching, hugging, and kissing by their supervisor, Ronald Selph, the store manager. (84) The plaintiffs were subjected to this conduct for an unreasonable amount of time before management put a stop to it. (85) Both plaintiffs complained about Selph's behavior on numerous occasions to three mid-level managers at the store. (86) One plaintiff told a manager that Selph had grabbed her and kissed her on the neck and that she did not know what to do about his behavior. (87) The manager responded that he was "shocked" by Selph's behavior and "didn't know what to say either." (88) Another manager took the step of warning Selph that his conduct amounted to sexual harassment, but Selph responded that he did not care. (89) Two managers who witnessed incidents told one of the plaintiffs that they would take steps to bring the problem to the attention of higher management. (90) One of the plaintiffs requested that one of these managers set up a meeting with the district manager so that they could complain about Selph's conduct and this meeting later took place. (91) The district manager immediately investigated, told the plaintiffs that he was upset because "the managers knew better and should have let him know what was going on," and gave Selph a demotion, written warning, and transfer to another city. (92)

Affirming the district court's grant of summary judgment to the employer based on the Ellerth affirmative defense, the Eleventh Circuit reasoned that the plaintiffs' earlier complaints to managers in their store about the sexual harassment they were experiencing months before did not suffice to put the employer on notice about the harassment. (93) Plaintiffs provided evidence in the record that another employee had also complained to mid-level managers about Selph's harassing behavior six months prior to their complaints, and that the managers had attempted unsuccessfully to handle the problem "in the store." (94) But the court held that this evidence was not relevant to when Publix should be considered to have been on notice about Selph's behavior, because that prior complaint, too, had not been brought to the attention of Publix's higher management. (95)

At no point does the court consider the possibility that the plaintiffs' attempts to complain to mid-level managers in the store were reasonable under the circumstances. Nor does the court examine, with any real attempt to shift the burden of proof to the defendant, the deficiencies of Publix's anti-harassment policies in failing to train its mid-level managers on what to do when they become aware of sexual harassment. As the case stands, the employer escapes vicarious liability for two rounds of blatant and pervasive sexual harassment by the head manager of one of its stores, despite other on-site managers having been made repeatedly and vividly aware of the problem. In short, in Madray as in other cases, even reasonable attempts by plaintiffs to bring sexual harassment to the attention of management, and failures on employers' part to have adequate policies to deal with such situations, survive summary judgment for the employer under the Ellerth affirmative defense. (96)

In other cases, the facts supporting the plaintiff's claim of hostile environment discrimination are weak for various reasons but, instead of acknowledging this, the court incorrectly grants summary judgment under Ellerth instead. Consider, for example, Thompson v. Naphcare, Inc. (97) The plaintiff in that case was a registered nurse supervisor who worked for the defendant health care services provider, Naphcare, in a correctional institution. (98) On July 3, 2000, Ronald Isaac was appointed as her supervisor. (99) In the period of approximately two months during which Thompson continued on the job after Isaac's appointment, Isaac engaged in five acts that Thompson viewed as sexual harassment. (100) On his first day on the job, he told Thompson that she had a figure that college girls would envy and that she should leave her husband behind and take Isaac on vacation with her. (101) On other occasions, he touched her in an intrusive manner by rubbing her shoulders. (102) Finally, he informed her in a private conference that it was his job to protect her from other personnel and that if she were "not so wrapped up with" her husband, she would see "what Isaac could do for her." (103)

After this initial period of sex talk, which Thompson rebuffed, Isaac's attitude changed. (104) According to Thompson, Isaac began berating her for negligible mistakes, told her that she was unprofessional, and threatened to replace her. (105) He issued written warnings to her for spending too much time with patients and discussed extending her probationary period because of performance problems. (106)

On August 30, Thompson consulted an attorney about Isaac's behavior. (107) The attorney wrote a letter, dated September 5, to her supervisor. (108) Two days later, two corporate officials, Naphcare's director of human resources and its in-house counsel, went to the work site to investigate the charges in Thompson's attorney's letter. (109) They interviewed witnesses Thompson had identified. (110) One witness recalled hearing the conversation in which Isaac stated that he wanted to go on vacation with Thompson, but "did not find Isaac's comments to be inappropriate or offensive." (111) Other witnesses similarly "denied witnessing any inappropriate behavior on the part of Isaac." (112) The two corporate officials concluded that they were "unable to substantiate the charges raised in the letter." (113) They told Isaac not to communicate with Thompson without other employees being present and left the work site, after which Isaac "approached Thompson to apologize and tried to explain his position." (114) Thompson again complained to the director of human resources, who phoned Isaac's supervisor to tell him to tell Isaac not to attempt to speak to Thompson about the matter. (115) Naphcare then officially informed Thompson that it could not substantiate her allegations, and shortly thereafter she resigned from her position. (116)

Affirming the trial court's grant of summary judgment, the Fifth Circuit questioned whether the alleged incidents rose to the level of actionable sexual harassment. (117) Instead of ruling on this ground, however, the court concluded that, even if Isaac's actions had created a hostile working environment, "we would exonerate Naphcare on its affirmative defense." (118) In support of this reasoning, the court pointed to the facts that Naphcare had a complaint procedure in place, the plaintiff did not immediately use it, (119) and Naphcare sent two high-level company officials to the work site after receiving the plaintiff's complaint. (120) Despite the fact that these officials announced that they could not substantiate the complaint, the court concluded that "the employer's response to her complaint was not ineffectual," because Thompson did not allege that the harassment continued after Isaac's last attempt to approach Thompson to talk about the incident. (121)

Missing in the court's discussion is sensitivity to the plaintiff's perspective. The court did not consider whether the two-month delay in reporting Isaac to management was reasonable in the context of what began as fairly mild harassment, nor does it address Thompson's claims that Isaac began to treat her negatively after she rebuffed his overtures. The court's discussion of the investigation by Naphcare's human resources director and in-house counsel similarly fails to convince. As even the court's very short discussion reveals, at least one witness confirmed that one incident alleged by Thompson did in fact occur. (122) Startlingly, the court did not question the employer's termination of the investigation after finding such confirming evidence, but concluded that the employer's finding that the plaintiff's allegations "could not be substantiated," was correct. (123) A plausible alternative explanation, given the facts as stated in the court's opinion, is that the investigation and dismissal of the complaint constituted a whitewash job--one that sent a clear message to the plaintiff, and any other potential victims who learned of the company's handling of the complaint, that allegations of sexual harassment would not be taken seriously even when a witnesses could confirm that an incident occurred.

The point on which the court instead hangs its analysis is that the witnesses did not find Isaac's behavior to be "inappropriate or offensive." (124) But the standard for what constitutes actionable sexual harassment surely should not use workplace norms as the measure. It should be largely irrelevant to the legal analysis whether particular witnesses found the conduct inappropriate, lest workplace norms permitting or encouraging a culture of sexual harassment (125) be permitted to continue because witnesses to harassment found the behavior unobjectionable. The appropriate legal standard for determining whether conduct rises to the level of actionable sexual harassment is that of a reasonable person in the plaintiff's situation. (126) Although one may doubt whether the plaintiff met that threshold--at least in her allegations about the initial harassment--that threshold is not the ground on which the Court disposes of the case. It instead summarily dismisses on affirmative defense grounds, creating another precedent standing for the idea that the mere showing of the existence of a policy and a pro forma employer response under it satisfies the affirmative defense doctrine. (127)

2. Co-worker Cases

Other examples of cases in which courts have failed to appreciate the abuse of power in hostile environment cases involve co-worker harassment. In co-worker harassment cases, the courts do not ask whether the employee committing the harassment was "aided in the agency" relationship in carrying out the harassment, but instead apply a negligence analysis, which seeks to ascertain whether employers knew or should have known about the harassment (usually to find that they did not). (128)

In Baker v. Boeing Helicopters, (129) the plaintiff worked at Boeing as an aircraft assembler. (130) She was sexually harassed by a co-worker, Jack Moser, whom, the plaintiff alleged, made continuous harassing and threatening comments to her and on one occasion touched her on the breast. (131) Moser, however, had "no authority, supervisory power, or opportunity to evaluate [her]." (132) Boeing had implemented and informed its employees about its sexual harassment policy. (133) When Moser continued to subject the plaintiff to harassing behavior and threatening comments, she reported Moser's conduct, first to her union steward, who unsuccessfully tried to get Moser to stop, and then to her senior managers, who did succeed in reprimanding Moser in such a way that he stopped his physically harassing conduct. (134) Her supervisors told her to report the problem to the employer's Equal Employment Opportunities office, but she did not do so. (135) She alleged, however, that this was because Moser had replaced his physically harassing conduct with another, more threatening form of harassment--namely, he threatened to do all he could to cause her to lose her job. (136)

As mentioned, Moser was not a supervisor, but a co-worker, albeit a senior one who held the title of lead assembler and trained less-experienced workers. (137) Moser also had an additional source of power in the workplace: he was a union member and well connected with the union's leadership. (138) It was through these sources of informal power that Moser apparently made his threats credible to the plaintiff about the loss of her job, though the court's discussion of the facts underlying the plaintiff's allegation of further threatening harassment is too sketchy to give much of a picture of what occurred. The court appeared to assume that these facts were completely inconsequential in any event, since it reasoned that "[p]laintiff's co-worker lacked the potential to alter plaintiff's employment," (139) and that the plaintiff had insufficient facts "to prove that management-level employees had actual or constructive knowledge" (140) of the harassment that occurred after the first report.

In short, in this case, and others, (141) courts assume that employees who do not have formal supervisory status lack the ability, or agency, to abuse their position with the employer in order to carry out sexual harassment. But that picture of the workplace belies what experts understand about how power operates in organizational settings, a topic I turn to in Part III below.

III. REFORMING THE TEST FOR VICARIOUS LIABILITY IN HOSTILE ENVIRONMENT CASES

I have argued in Part II that courts are misapplying the Ellerth affirmative defense in hostile environment cases to bar plaintiffs from proceeding to trial. Although the plaintiffs have presented sufficient facts concerning the inadequacy of employers' efforts to prevent and remedy sexual harassment, courts' application of Ellerth precludes their cases from surviving summary judgment. Here, I will articulate how the Ellerth test should be formulated in order to encourage courts to engage in much more rigorous inquiry in assessing employers' affirmative defense claims at the summary judgment stage.

The first element of my proposal is this: courts should follow the Supreme Court's own admonishments and avoid use of unduly formalistic categories of analysis in exploring whether hostile environment sex discrimination has occurred. This is especially necessary in examining the dynamics of workplace settings, a context that calls out for realistic insights into the subtleties of particular situations. Courts should examine the facts without immediately forcing them into dichotomies, such as those between quid pro quo and hostile environment, and between supervisor and co-worker harassment. (142) These categories, while enormously helpful at a certain historical moment (143) to aid the development of sexual harassment doctrine, now frequently obscure rather than shed light on the presence of actionable sexual harassment. The question of whether sexual harassment is severe enough to alter the terms and conditions of employment is answered neither by focusing on whether a harasser is a supervisor or a co-worker, nor on whether tangible employment actions took place or were merely threatened. The key question in vicarious liability cases should instead be whether harassers were aided by their agency relationship with the employer in carrying out sexual harassment that altered the plaintiff's terms and conditions of employment such that it rose to the level of discrimination on the basis of sex. I am willing to accept that not all sexual harassment cases fit this bill--that mere proximity and contact are not enough, as the court noted in Ellerth. (144) But far greater weight should be placed on the informal power dynamics of workplace settings that allow harassers to use the agency vested in them by employers in order to accomplish their objectives.

Although I have offered my own particular take on the inadequacy of sexual harassment law, a host of scholars have described from various perspectives the ways in which sexual harassment doctrine is failing to capture the harm it aims to address. In a brilliant article addressing the conundrum about how to classify constructive discharge for purposes of the tangible employment action doctrine, Martha Chamallas succinctly states, "[t]he rankings denominated in the formal organizational chart may be far less important than the opinion of persons who have real clout in the organization." (145) Theresa Beiner, John Marks, Joanna Grossman, Ann McGinley and others have similarly addressed the inadequacy of the hostile environment doctrine in capturing workplace realities. (146) In Marks' apt words, the Supreme Court's affirmative defense has created "a dubious safe harbor" for employers, based on the courts' superficial examination of the text of employer sexual harassment policies. (147) A comprehensive and persuasive new book by Theresa Beiner analyzes a host of data and makes a series of compelling proposals for reform. (148) Beiner summarizes a large body of empirical work showing that women tend not to report sexual harassment for fear of the career consequences of doing so. (149) Moreover, the data show that women's judgment on this matter is rational: women who do report sexual harassment have worse career outcomes as a result than those who do not. (150) This finding should be of concern to courts and policy makers to the extent that they are sincere in seeking to construct law to further the policy objectives of Title VII. If lodging sexual harassment complaints with employers negatively impacts the careers of those who complain, how is it helping the project of achieving sexual equality in employment to require plaintiffs to lodge such complaints under all circumstances?

On the basis of these and many other sets of empirical data, Beiner makes a series of proposals to modify sexual harassment doctrine in order better to address the harm of sexual harassment. On the question of employers' vicarious liability, Beiner argues that "targets of supervisor harassment should be compensated for the harm they experience regardless of the employer's preventive efforts." (151) Under Beiner's test, the question of employers' efforts to prevent or correct sexual harassment would only arise when deciding punitive damages. (152)

Approaches along the lines that Beiner proposes reflect the law of some states, such as New Jersey. (153) With respect to federal law, however, such an approach may fail to garner sufficient support, even among feminists committed to ending sex discrimination in the workplace. (154) Allowing employers the opportunity to avoid vicarious liability by showing that they have made reasonable efforts to prevent and redress sexual harassment arguably preserves the important policy objectives I pointed to at the beginning of this Article-namely, the creation of incentives to encourage employers to take proactive efforts to prevent and remedy harassment without the involvement of the courts. (155) Thus, although I am sympathetic to Beiner's perspective, my proposal does not go as far as hers. Instead, what I advocate here remains within the existing legal framework that permits an affirmative defense opportunity for employers, but argues that the showing required of employers be made far more rigorous than that which courts often require today.

This approach would require the courts to examine the factual situation presented in particular cases in a far more searching way than many courts do today. (156) Plaintiffs should be permitted to present background evidence concerning institutional culture, including on such matters as the following: Regardless of whether a complaint procedure exists, what were the results of the employee's prior use of that procedure? How many cases resulted in discipline against the harasser? Are low-level employee harassers terminated based on incidents of harassment, but not powerful agents of the institution? And what are the effects of friendships, good ole' boy networks, and political alliances in the workplace at issue?

On a related note, in co-worker harassment cases, plaintiffs should be allowed to present evidence concerning the informal power that persons classified as co-workers possess within the institution. (157) Are they well-respected senior employees? Do they serve important roles in which they have special access to decision-makers' ears? Do they play a role in evaluating or commenting on junior employees' work? Does their degree of willingness to train or mentor junior employees have a significant effect on these employees' career prospects? Do senior co-workers influence access to important job assignments or other career advancing opportunities? Do they have the ability to influence co-workers' attitudes toward the employee in question--for example, to ice him or her out of informal but career-related experiences? Does a supervisor or co-worker have credible authority in making threats in order to bully an employee into granting sexual favors? These are all questions that the courts should ask as aspects of their analysis under Ellerth as to whether employers' policies to deter and punish sexual harassment were effective and adequate and whether plaintiffs' conduct was reasonable under the circumstances. But these important questions all but drop out of many lower courts' analysis, as shown in the examples I discussed in Part II-C. While the affirmative defense doctrine should call for a searching, skeptical inquiry, with the burden of proof on the employer, courts are using the Ellerth doctrine to dispose of cases with as little scrutiny as possible.

A. Work by Organizational Theorists

The work of expert organizational theorists supports the type of inquiry I have outlined above. That work shows that formal organizational hierarchy charts do not capture the real power dynamics of a workplace. These theorists understand the working of power dynamics in organizations as being different from the grant of formal hierarchical authority. As one leading management expert puts it:

   The terms "authority" and "power" are consistently confused
   by students of

   Management.... [A]uthority [is] the right to act, or command
   others to act, toward the attainment of organizational goals.
   Its unique characteristic ... was that this right had legitimacy
   based on the authority figure's position in the organization.
   Authority goes with the job. You leave your managerial job and
   you give up the authority that goes with that position. When
   we use the term "power," we mean an individual's capacity to
   influence decisions. As such, authority is actually part of the
   larger concept of power; that is, the ability to influence based
   on an individual's legitimate position, can affect decisions, but
   one does not require authority to have such influence. (158)

Robbins goes on to explain how an employee may be low in the authority hierarchy but high on the power index by virtue of his or her closeness to what he calls the "power core." Closeness to the "power core" results from one's access to or relationships with key managers or possession of key organizational information and resources. (159) As examples, Robbins explains how executive secretaries, long-time production engineers, and labor union officials may be part of the power core despite their lack of supervisory status. (160) Richard Daft similarly summarizes the extensive literature supporting this understanding of how power in organizations works. (161) Daft breaks down the sources of power individuals may utilize at different levels of an organization's hierarchy, and notes that the result may be that lower level employees have more effective power than do those with more formal authority. (162)

Another classic study of organizational management theory breaks down individuals' power within organizations into six variables, some having to do with the person's position or formal authority, but others with personal characteristics, such as referent power (for example, being a person that occupies a professionally coveted position) and expertise. (163) This study finds that compliance often depends more on legitimacy, expertise and referent power than on ability to invoke formal rewards or punishment. (164)

All of this expert research strongly suggests that the doctrinal distinction between supervisors and co-workers in current sexual harassment law rests on faulty assumptions about who may wield significant power over others in the workplace.

Still other work in organizational management theory looks at how power is deployed and the results of its use in organizations. Robbins describes the operation of power within organizations as a form of coalition politics, through which shifting coalitions fight for influence. (165) Cynthia Cockburn takes the insights of organizational management experts and extends them to examine gender politics within organizations. (166) Cockburn further describes how actors within institutions can use power alliances to block the rise of women within organizations. (167) Insight into this phenomenon was, indeed, important to theorizing hostile environment sexual harassment as a form of sex discrimination; hostile environment sexual harassment provides men lacking in formal authority a means to keep women out of traditionally male jobs, a point Vicki Shultz developed in her ground-breaking work. (168) Researchers have now begun to conduct empirical investigations into how individuals within organizations use informal power to effectuate sexual harassment as well as how they use sexual harassment in order to exert power. (169) This literature, although "still in its infancy," shows that "the relationships among facets of power and types of sexual harassment are underarticulated," and that power concerns arise not only in supervisory harassment but in co-worker and subordinate harassment as well. (170)

Not only organization theory, but also general principles of agency law support an expanded inquiry into the harasser's relationship to the employer for purposes of vicarious liability analysis. As Catherine Fisk and Erwin Chemerinsky have pointed out, courts have taken inconsistent and unduly cramped positions in analyzing the scope of vicarious liability under various federal civil rights statutes. (171) These leading scholars argue that the best approach for courts to take in all these cases is one that follows common law tort principles. (172) Under those principles, as I show below, the vicarious liability analysis for Title VII hostile environment claims should embrace the "aided in the agency" theory recognized in common law.

B. The "Aided in the Agency" Standard

Another way of sharpening the teeth of the test articulated in Ellerth would be to place greater emphasis on common law analysis of agency principles. Courts should ask, not whether a tangible employment action occurred, but whether a harasser was "aided in the agency" in committing wrongdoing by virtue of his or her relationship with the employer. Put otherwise, the affirmative defense should create sufficient incentives to encourage employers to prevent harassers from using the agency or power granted them by the employer to commit harassment. An employer who has not prevented the use of its agency in this way has failed to take sufficient preventative action--or, put otherwise, has failed to act with reasonable care in protecting its employees from hostile environment discrimination.

The "aided in the agency" standard had been applied in diverse areas of law. Courts have held employers liable for the acts of managerial employees under section 219(2)(d) of the Restatement of Agency and related provisions in a wide variety of situations and fact patterns outside the employment context in which managers abused their positions to fulfill personal, rather than official goals. (173) Indeed, a number of courts prior to Ellerth applied this common law "aided in the agency" analysis to sexual harassment claims, even when the employee committing the harassment was not acting within the scope of his or her employment. (174) Some states have similarly opted for this section 219(2)(d) analysis in interpreting their state anti-discrimination statutes.

One such court is the New Jersey Supreme Court, which interpreted the New Jersey Law Against Discrimination (LAD) (175) in Gaines v. Bellino. (176) There, the court granted certification to review the holding of the Appellate Division, which had dismissed a hostile environment workplace harassment claim under LAD because the employer had implemented a sexual harassment policy, trained employees under it, and acted promptly when the plaintiff's complaint was finally brought to its attention. (177) The court followed its earlier, pre-Ellerth precedent in Lehmann v. Toys R Us, Inc., (178) and held that "principles of agency law should control employer liability for compensatory damages in cases of supervisory hostile environment sexual harassment claims." (179) Applying section 219(2)(d), (180) the court explained that the "question whether a supervisor, who creates a hostile environment, was aided by delegated power to control the day-to-day work environment is a fact sensitive inquiry," that requires examination of a number of questions including: (1) whether the employer delegated the authority to the supervisor to control the situation on which the plaintiff's claims are based; (2) whether the supervisor exercised that authority; (3) whether the exercise of authority resulted in a violation of LAD; and (4) whether the delegated authority aided the supervisor in thus injuring the plaintiff. (181) The court further reiterated its earlier precedent that had given teeth to this test, explaining that it should encompass the examination of facts such as "the existence of effective sensing or monitoring mechanisms to check the trustworthiness of the policies and complaint structures, 'evidence of' an unequivocal commitment from the highest levels of the employer that harassment would not be tolerated, 'and' demonstration of that policy commitment by consistent practice." (182)

Applying section 219(2)(d) to the facts of cases I have discussed here, one might add to the New Jersey Supreme Court's list of relevant questions at the affirmative defense stage, factors such as whether the supervisor used his or her agency or authority to keep the plaintiff in the office or other area where harassment was committed, whether he or she used actual or apparent authority to make credible threats that would lead a reasonable person in the plaintiff's situation to succumb to the harassment and/or not to report it, and whether the employer had in place and enforced a policy requiring supervisors to report knowledge of sexual harassment up the chain of command.

At bottom, answering the key question as to how strong to make the Ellerth affirmative defense requires a return to the objectives of recognizing sexual harassment as a Title VII claim in the first place. Leading scholars propose many such objectives on descriptive and normative grounds. (183) Sexual harassment is supposed to be about sex discrimination, after all, but it is also probably about something else--at least in the minds of fact finders, whom, statistics show, are more likely to impose liability for sexual harassment than for other forms of Title VII discrimination. (184) The finding that sexual harassment violates anti-discrimination law stems from an implicit policy against the abuse of power obtained or held by virtue of employment, where such an abuse of power is implemented on the basis of the sex of the plaintiff. (185) If this is indeed an important objective underlying the recognition of hostile environment sexual harassment as a form of sex discrimination, the key to the analysis should be whether power has been abused in this way. (186)

If power obtained or held by virtue of employment has been abused in carrying out sexual harassment, then the harasser has been "aided in the agency"--in other words, he or she has been helped in his or her objective by virtue of the agency relationship with the principal. If that is the case, then under standard principles of agency law, the employer should be held vicariously liable. This test captures the harm of a harasser's abuse of power to commit sexual harassment without leading to the conclusion the Court feared, namely, that all harassers are "aided in the agency" by virtue of mere proximity to their prey. Under the test I propose, which probes for abuse of power granted by the employer, courts should ask whether the harasser used power over the plaintiff held by virtue of the employment relationship--through threats, intimidation, or similar acts--that would cause a reasonable employee to submit to the harassment or not report it. The mere coincidence of co-proximity would not meet this standard.

The Court has made matters enormously and unduly complicated in its cases, but that may be because it is seeking to dance around these basic principles. The matter really is, or should be, much more straightforward. Application of the "aided in the agency" test implements the policy objectives of Title VII, and helps to ensure that merit-worthy cases are not barred by an expansive application of the affirmative defense doctrine.

C. Courts That Get It Right

The argument might be made that the test I propose is unworkable--that no court is going to be sufficiently sensitive to the nuances of informal power dynamics in the workplace as I have proposed. But there is a case that belies that argument, because in that case, the court got it right. At first glance this may seem surprising because the court is the U.S. Court of Appeals for the Fourth Circuit, hardly a sympathetic venue for plaintiffs in employment and civil rights cases. On closer inspection, the sense of surprise goes away--the author of the opinion is Judge Diana Gribbon Motz. (187) The case is Williams v. Spartan Communications, Inc., (188) and it is, unfortunately but not surprisingly, listed as an unpublished opinion. (189)

Williams involved a plaintiff who was sexually assaulted three times during business trips she took with her supervisor. (190) A magistrate judge ruled that the defendant, Spartan, had satisfied both elements of the hostile environment affirmative defense and granted summary judgment in its favor, (191) but the Fourth Circuit reversed. (192) The plaintiff acknowledged that she knew the company had an anti-harassment policy and had attended meetings and seen posted notices about it, which included the identities of the persons with whom she could file a complaint. (193) Judge Motz wrote, however, that

   while the existence of an antiharassment
   policy and prompt corrective action pursuant
   to it provides important evidence that an
   employer has acted to meet conclusion. (194)
   first prong of the affirmative defense, such
   evidence does not compel this

Instead, Motz explained, the employer must also have been reasonably effective in preventing sexual harassment. (195) The magistrate judge had ignored substantial evidence that Spartan's policy was not an effective program. (196) This evidence included the following facts: (1) that Spartan's management tolerated and participated in lewd conversations and publication of sexually explicit jokes and cartoons in the workplace, (2) that complaints to a manager by an employee, other than the plaintiff, about foul language and sexist jokes had produced no corrective action, (3) that the sexual harassment policy threatened to punish employees who falsely reported sexual harassment with discipline up to and including termination, but did not promise to protect complainants from retaliation for pursuing complaints, and (4) that plaintiff's harasser had a close relationship to other senior managers at the company. (197) Judge Motz noted that even though the policy commendably created multiple avenues for complaint, all of those avenues would have required the plaintiff to report to someone who was close to the harasser or who reported to a member of senior management who was a good friend of the harasser. (198) Williams, moreover, had produced evidence that an employee decided not to complain of harassment because of fear of being fired. (199) In short, this judge displayed the requisite sensitivity to how informal power dynamics in the workplace can transform a sexual harassment complaint policy that appears reasonable on its face into an illusory sham for women working within the particular power dynamics of their organization. Only the kind of context-sensitive, nuanced examination Judge Motz was willing to undertake is sufficient to probe these issues.

A second example of an opinion in which a court engages in searching, fact sensitive analysis, also written by a female judge, (200) is Petrosino v. Bell Atlantic. (201) Petrosino was an installation and repair technician at the employer's garage, and for most of her years of employment was the only female technician at that garage. (202) In alleging hostile environment sex discrimination, she complained of constant sexually demeaning conversations conveying a disrespect for women and crude sexual graffiti scrawled by co-workers inside the terminal boxes with which she had to work, including images of headless women with their legs in the air, women's legs spread open, men with their penises out or having sex with animals, and messages that male and female employees performed sex acts with supervisors in order to advance in their careers, including at least one picture of Petrosino engaged in such behavior. (203) Petrosino had also been physically attacked in a parking lot and groped and kissed, endured jokes and terminal box graffiti about this incident, and told by her supervisors that she was "a damn woman," she should "calm her big tits down," her job concerns were attributable to her menstrual cycle, and that women were "too damned thin skinned" to work at the garage. (204)

Petrosino asserted that she complained informally and formally about the hostile environment to co-workers and supervisors. (205) She filed a labor grievance against her direct supervisor for harassment and prevailed, but no discipline was imposed; instead, she and her supervisor were sent to a seminar to help them work out their differences. (206) When she attempted to use the opportunity to voice her concerns, her supervisor told her to "just keep your mouth shut and do what I tell you." (207) When she complained about this comment to a senior manager, he offered to transfer her but she declined. (208) Her supervisor subsequently reprimanded her for going over his head and took away job responsibilities that would have prepared her for a future management position. (209) Soon after, she resigned, without having complained about this last action. (210)

On these facts, the district court, rather incredibly to my mind, granted summary judgment to Bell Atlantic on all claims, which included tangible employment action, hostile environment, and constructive discharge theories. (211) The Second Circuit reversed in an opinion written by Judge Reena Raggi. (212) Declining to reach Petrosino's tangible employment action claims, the court held that Bell Atlantic had failed to meet its burden on its affirmative defense claims against the hostile environment charges. (213) It held that the fact that there was a documented corporate policy against sexual harassment, as well as an ethics hotline through which employees could report incidents of harassment, was not dispositive, especially in light of the fact that Petrosino asserted that she had used the hotline but no one had investigated her complaint or taken any remedial action. (214) Bell Atlantic argued that Petrosino had failed to return follow-up calls after her initial report and had failed to pursue many other incidents of harassment through the hotline procedures. However, the court noted that these disputes about facts and their implications were correctly left to the trier of fact, not to the court at the summary judgment stage. (215)

IV. CONCLUSION

The problem with the lower courts' application of the Ellerth defense is that the key question--namely, whether the supervisor's misconduct has been aided by the agency relationship--is not answered by the affirmative defense doctrine as currently articulated. The affirmative defense doctrine addresses workplace-level deterrence and employer notice objectives, as Susan Sturm and others have pointed out. It thus can play an important role in encouraging policy change at the workplace level and could stand as a positive example of how the law can be used to progressive ends in shaping social relations. But the problem with the Ellerth affirmative defense doctrine, as courts are currently applying it, is that it does not ask the legally relevant question, given that Congress specifically defined "employer" for purposes of employer liability under Title VII as including an "agent." (216) That definition states that employers are liable for the actions of their agents. Under common law agency principles, the touchstone for employer liability is whether an agent was aided in his or her agency--in other words, by his or her power conferred by virtue of an agency relationship with the employer--in carrying out the acts in question. Answering that question in the sexual harassment context requires a searching and fact-sensitive examination of how power dynamics in particular workplaces operate. When a plaintiff has presented evidence that could lead a reasonable fact finder to find such an abuse of employer-granted power, courts should allow plaintiffs to pass the summary judgment hurdle. In these circumstances, indeed, employers have failed to meet their burden of proof under the Court's affirmative defense standard, which requires, after all, that employers exercise "reasonable care" (217) and take measures that are "readily accessible" and "effective" in preventing sexual harassment, (218) and that plaintiffs do not act "unreasonably" (219) under the particular circumstances.

(1.) See, e.g., Susan Sturm, Second Generation Employment Discrimination: A Structural Approach, 101 COLUM. L. REV. 458, 489 (2001) ("The Supreme Court's structural approach encourages the development of a dynamic regulatory regime through ongoing interaction between general legal norms and workplace problem solving.").

(2.) See Lehr, Middlebrooks, Price & Procter, P.C., A Primer on Employee Complaints, ALA. EMP. L. LETTER, July 2002 (noting the cottage industry that developed in sexual harassment consulting); Steptoe & Johnson PLLC, Employment Consultants: Friend or Foe?, W. VA. EMP. L. LETTER, Dec. 2001.

(3.) 42 U.S.C. [section] 2000e-2 (2000).

(4.) 477 U.S. 57 (1986).

(5.) Id. at 59-60.

(6.) Id.

(7.) Id. at 60. The harassment she suffered in the hands of Taylor included being fondled in front of other employees and several incidents of forcible rape. Id.

(8.) Id.

(9.) Id. at 61.

(10.) Id at 57. The court also found that the sexual relationship between Taylor and Vinson was voluntary and thus not actionable under Title VII, a finding the Supreme Court also rejected. See Vinson v. Taylor, 23 Fair Empl. Prac. Cas. (BNA) 37, 42 (D.D.C. 1980).

(11.) Meritor Sav. Bank, 477 U.S. at 62.

(12.) Id. (quoting Vinson, 23 Fair Empl. Prac. Cas. (BNA) at 42).

(13.) See Vinson v. Taylor, 753 F.2d 141, 147-48 (D.C. Cir. 1982). The court of appeals further held that hostile environment sexual harassment constituted a form of sex discrimination under Title VII and that the district court erred in refusing to consider this theory. Id. at 145-46.

(14.) Meritor Sav. Bank, 477 U.S. at 64.

(15.) Id. at 68.

(16.) Id. at 64-67. The Court further corrected the district court on the matter of voluntariness, noting that the correct question was whether the sexual advances were "unwelcome." Id. at 68.

(17.) See 42 U.S.C. [section] 2000e(b) (2000).

(18.) Meritor Sav. Bank, 477 U.S. at 72.

(19.) Id.

(20.) Id.

(21.) Id. at 72-73.

(22.) Id. at 73.

(23.) Id.

(24.) 524 U.S. 742 (1998).

(25.) 524 U.S. 775 (1998).

(26.) Ellerth, 524 U.S. at 747.

(27.) Id.

(28.) Id.

(29.) Id.

(30.) Id. at 748.

(31.) Id.

(32.) Id.

(33.) Id.

(34.) Id.

(35.) Id. at 749.

(36.) Id.

(37.) Id.

(38.) Id. at 757 (citation omitted).

(39.) Id. at 756-59.

(40.) Id.

(41.) Id. at 756.

(42.) Id. at 757.

(43.) See RESTATEMENT (SECOND) OF AGENCY [section] 219(2)(d) (1958) ("A master is not subject to liability for the torts of his servants acting outside the scope of their employment, unless ... he was aided in accomplishing the tort by the existence of the agency relation.").

(44.) Ellerth, 524 U.S. at 760.

(45.) Id.

(46.) Id. at 760-61.

(47.) See id. at 748.

(48.) Id. at 752-53.

(49.) Id. at 753.

(50.) Id. at 753-54.

(51.) See id. at 752 (stating that hostile environment claims require severe or pervasive harassment).

(52.) Id. at 762.

(53.) Id. Note here that the Court switches between the terms, "tangible employment action" and "official company act," which are not, after all, synonymous. In so doing the Court adds still more confusion to an area of law already mired in ambiguity. For this and other reasons, the tangible employment term is another misstep, but I will not go into those other reasons here. Instead, my focus remains on the general vicarious liability analysis and its inconsistency with the purposes of Title VII.

(54.) Id.

(55.) Id. at 762.

(56.) Id. at 752; Meritor Sav. Bank, 477 U.S. at 64-67.

(57.) See EQUAL EMPLOYMENT OPPORTUNITY COMM'N, ENFORCEMENT GUIDANCE: VICARIOUS EMPLOYER LIABILITY FOR UNLAWFUL HARASSMENT BY SUPERVISORS 13-14 (1999).

(58.) Ellerth, 524 U.S. at 765.

(59.) See Faragher, 524 U.S. at 808-09. The policy's inadequacies included the facts that it had not been disseminated among employees, management had made no attempt to keep track of supervisors' conduct under the policy, and the policy did not include any assurance that harassing supervisors could be bypassed in registering complaints. See id.

(60.) 542 U.S. 129 (2004).

(61.) Id. at 134.

(62.) Id.

(63.) Id. at 148-52.

(64.) See, e.g., Martha Chamallas, Title VII's Midlife Crisis: The Case of Constructive Discharge, 77 S. CAL. L. REV. 307, 385 (2004).

(65.) Suders, 524 U.S. at 131.

(66.) Id. at 134.

(67.) Id.

(68.) My survey includes only cases in legal databases easily accessible for legal research. One can only wonder about the many additional cases dismissed on summary judgment at the trial court level that do not turn up in these databases.

(69.) For my discussion, I have chosen to highlight cases other than those other scholars have examined. Surveys that review cases other than the ones I focus on here include John H. Marks, Smoke, Mirrors, and the Disappearance of "Vicarious" Liability: The Emergence of a Dubious Summary Judgment Safe Harbor For Employers Whose Supervisory Personnel Commit Hostile Environment Workplace Harassment, 38 Hous. L. REV. 1401, 1423-36 (2002) (summarizing cases demonstrating weakness of hostile environment doctrine in light of the affirmative defense); David Sherwyn, Michael Heise & Zev J. Eigen, Don't Train Your Employees and Cancel Your "1-800" Harassment Hotline: An Empirical Examination and Correction of the Flaws in the Affirmative Defense to Sexual Harassment Charges, 69 FORDHAM L. REV. 1265, 1265-85 (2001) (statistical analyses of cases decided shortly after Ellerth); Martha S. West, Preventing Sexual Harassment: The Federal Courts' Wake-up Call for Women, 68 BROOK. L. REV. 457, 461-494 (2002). See also articles cited in infra note 146.

(70.) See discussion supra text accompanying notes 57-67.

(71.) Id.

(72.) 20 F. Supp. 2d 1379 (S.D. Ga. 1998).

(73.) Id. at 1381.

(74.) Id. at 1382.

(75.) Id. at 1381.

(76.) Id.

(77.) Id. at 1382.

(78.) Id.

(79.) Id.

(80.) Id. at 1384.

(81.) Id. at 1384, 1386.

(82.) See id. at 1386.

(83.) 208 F.3d 1290 (11th Cir. 2000).

(84.) Id. at 1293.

(85.) Id. at 1295.

(86.) Id.

(87.) Id.

(88.) Id. at 1293.

(89.) Id. at 1294.

(90.) Id. at 1293-94.

(91.) Id. at 1294.

(92.) Id.

(93.) Id.

(94.) Id. at 1294 n.3.

(95.) Id. at 1294.

(96.) Collette v. Stein-Mart, Inc., 126 F. App'x. (6th Cir. 2005) is a similar case in which the Sixth Circuit upheld the grant of summary judgment under the Ellerth defense despite the fact that the plaintiff had complained to mid-level managers.

(97.) 117 F. App'x. 317 (5th Cir. 2004).

(98.) Id. at 319.

(99.) Id.

(100.) Id.

(101.) Id.

(102.) Id.

(103.) Id.

(104.) Id.

(105.) Id.

(106.) Id.

(107.) Id. at 320.

(108.) Id.

(109.) Id.

(110.) Id.

(111.) Id.

(112.) Id.

(113.) Id.

(114.) Id.

(115.) Id.

(116.) Id.

(117.) Here the court may have been right. Not all sex-talk a worker finds uncomfortable does, or indeed, should, rise to the level of actionable sexual harassment, lest sexual harassment law become a means of "sanitizing" work places and extending the state's power to exert social control over its citizens through law. See Vicki Schultz, The Sanitized Workplace, 112 YALE L.J. 2061, 2067 (2003) (arguing that "attempt[s] to banish sexuality from the workplace threaten many important social interests").

(118.) Naphcare, 117 F. App'x. at 323.

(119.) See also McPherson v. City of Waukegan, No. 01 C 9264, 2003 U.S. Dist. LEXIS 9098, at * 22 (N.D. Ill., May 29, 2003) (granting summary judgment for the employer where plaintiff did not report initial incidents of harassment that escalated to sexual assault; although plaintiff reported the assault, the court held that she should have reported earlier incidents), aff'd, 379 F.3d 430 (7th Cir. 2004). This and other similar cases represent yet another misstep in the development of sexual harassment law and place plaintiffs in a Catch-22 situation: They risk being found to have failed to invoke complaint procedures in a timely manner if they wait too long to report escalating harassment, yet if they invoke those procedures prematurely, they risk reporting conduct that is insufficiently severe to constitute sexual harassment.

(120.) Naphcare, 117 F. App'x. at 324.

(121.) Id.

(122.) Id.

(123.) Id.

(124.) Id. at 320.

(125.) Vicki Schultz persuasively argues that these workplaces are precisely where enforcement of sexual harassment law is most important in furthering Title VII's objective of eliminating sex discrimination in employment. See Vicki Schultz, Reconceptualizing Sexual Harassment, 107 YALE L.J. 1683, 1736 (1998) (noting the way in which hostile environment sexual harassment can be used to deter women from working in traditionally sex-segregated workforce sectors).

(126.) See Harris v. Forklift Sys., Inc., 510 U.S. 17, 22-23 (1993) (requiring the plaintiff to show that the environment "would reasonably be perceived, and is perceived, as hostile or abusive").

(127.) Other cases in which perfunctory employer investigations have sufficed to provide an affirmative defense include Holly D. v. Cal. Inst. of Tech., 339 F.3d 1158 (9th Cir. 2003) (dismissing plaintiff's lawsuit on summary judgment despite expert testimony that university's investigation should have been more comprehensive, including an examination of supervisor's computer for pornographic web site bookmarks); Dennis v. Nevada, 282 F. Supp. 2d 1177 (D. Nev. 2003) (granting summary judgment to the employer where employer had dismissed plaintiff's claim of sexual harassment as unsubstantiated and transferred her to the graveyard shift, concluding that plaintiff failed to report the misconduct soon enough; employer's investigation was sufficient, and undesirable transfer did not amount to a tangible job action).

(128.) But an ironic result of courts' misuse of the affirmative defense doctrine in supervisor cases, such as those discussed in Part II.C.1 above, is that the degree of scrutiny applied to the effectiveness of employers' responses to co-worker sexual harassment is sometimes higher than that afforded in supervisor cases, even though the negligence standard for imposing liability on employers applicable to co-worker harassment is intended to be a harder one for plaintiffs to meet. See, e.g., Antonopoulos v. Zitnay, 360 F. Supp. 2d 420, 428 (D. Conn. 2005) (denying summary judgment to employer despite the fact that employer used corrective measures, including a verbal warning and written documentation in the harasser's personnel file).

(129.) No. 01-3565, 2004 U.S. Dist. LEXIS 12620, at * 1 (E.D. Pa. June 30, 2004).

(130.) Id. at * 1.

(131.) Id. at * 2.

(132.) Id. at * 1.

(133.) Id. at * 2.

(134.) ld. at * 2-3.

(135.) Id. at * 3.

(136.) Id. at * 15.

(137.) Id. at * 14.

(138.) Id. at * 1.

(139.) Id. at * 20.

(140.) Id. at * 17.

(141.) See e.g., Wyninger v. New Venture Gear, Inc., 361 F.3d 965, 971-978 (7th Cir. 2004) (upholding grant of summary judgment despite acknowledgement that plaintiff had experienced hostile environment harassment by union representatives).

(142.) For examples of cases in which courts correctly recognize that an employee classified as a co-worker can in fact possess power over another employee for purposes of sexual harassment analysis, see Mack v. Otis Elevator Co., 326 F.3d 116, 125 (2d Cir. 2003) (harassing employee's "authority over Mack, bestowed upon him by Otis, enabled him, or materially augmented his ability, to impose a hostile work environment on her" and he thus should be classified as a supervisor); Entrot v. Base Corp., 819 A.2d 447, 459 (N.J. Super. Ct. App. Div. 2003) ("instead of requiring a litmus test depending on specific factors," analysis should turn "on whether the power the offending employee possessed was reasonably perceived by the victim, accurately or not, as giving that employee the power to adversely affect the victim's working life").

(143.) See CATHERINE MACKINNON, SEXUAL HARASSMENT AND THE WORKING WOMAN 211 (1979) (articulating the difference between supervisor and co-worker harassment).

(144.) Ellerth, 524 U.S. at 742-43.

(145.) Chamallas, supra note 64, at 290.

(146.) See THERESA M. BEINER, GENDER MYTHS V. WORKING REALITIES: USING SOCIAL SCIENCE TO REFORMULATE SEXUAL HARASSMENT LAW 145-79 (2005) (critiquing the Ellerth standard); Joanna L. Grossman, The Culture of Compliance: The Final Triumph of Form Over Substance in Sexual Harassment Law, 26 HARV. WOMEN'S L.J. 3 (2003) (arguing for elimination of the affirmative defense and other reforms to strengthen sexual harassment law); Joanna L. Grossman, The First Bite is Free: Employer Liability for Sexual Harassment, 61 U. PITT. L. REV. 671 (2000) (noting that under the affirmative defense, the employer is not liable for the first instance of a supervisor's sexual harassment); see also Marks, supra note 69; cf. Ann C. McGinley, Functionality or Formalism? Partners and Shareholders as "Employees" Under the Anti-Discrimination Laws, 57 SMU L. REV. 3, 51 (2004) (noting that formalist assumptions that partners possess the political power to protect themselves against discrimination ignores the way in which partners can lack the economic and social power within the organization to avoid hostile environment harassment and sex discrimination).

(147.) Marks, supra note 69, at 1422. Another excellent body of literature critiques the courts' trend toward summary dismissal of cases in general. See, e.g., Stephen B. Burbank, Vanishing Trials and Summary Judgment in Federal Civil Cases: Drifting Toward Bethlehem or Gomorrah?, 1 J. EMPIRICAL LEGAL STUD. 591 (2004). Many of the leading scholars creating this literature, including Elizabeth Schneider, Vivian Berger, Steven Furbank, Desiree Kennedy, and Ann McGinley, gathered for an impressive roundtable discussion at the 2005 Law & Society Conference.

(148.) See generally BEINER, supra note 146.

(149.) Id. at 163-66.

(150.) Id.

(151.) Id. at 173-74.

(152.) Id.

(153.) See discussion infra text accompanying notes 175-182 (discussing New Jersey law); see also State Dep't of Health Servs. v. Superior Court, 79 P.3d 556 (Cal. 2003) (holding that language of state anti-discrimination statute "does not suggest that an employer's liability for sexual harassment by a supervisor is constrained by principles of agency law").

(154.) See, e.g., Sturm, supra note 1, at 489 (noting benefits of the Ellerth affirmative defense).

(155.) Id.

(156.) See discussion of Suders dicta, supra text accompanying notes 65-67.

(157.) See, e.g., Entrot, 819 A.2d at 459 ("Also relevant would be any evidence that the alleged harasser controlled the workplace in subtler and indirect ways, as long as the effect was to restrict the victim-employee's freedom to ignore sexually harassing conduct.").

(158.) STEPHEN P. ROBBINS, ORGANIZATION THEORY: THE STRUCTURE AND DESIGN OF ORGANIZATIONS 173 (1983).

(159.) Id. at 174.

(160.) Id. at 174-75, 176-77.

(161.) See RICHARD L. DAFT, ORGANIZATIONAL THEORY AND DESIGN 387-94 (4th ed. 1992); see also JEFFREY PFEFFER, NEW DIRECTIONS FOR ORGANIZATION THEORY: PROBLEMS AND PROSPECTS 137-38 (1997) (tracing the growth of literature examining the development and exercise of power and influence in organizations).

(162.) See id.; see also John R.P. French Jr. & Bertram Raven, The Bases of Social Power, in CLASSICS OF ORGANIZATION THEORY 311, 313 (Jay M. Shafritz et al. eds., 6th ed. 2004) (describing organization theorists' understanding of various sources of informal power and their use); Jeffrey Pfeffer, Understanding the Role of Power in Decision Making, in CLASSICS OF ORGANIZATION THEORY, supra, at 289, 290-91 (describing the relationship between power, authority, and politics in workplaces); JOHN SCOTT, SOCIAL NETWORK ANALYSIS 82-83, 101,124 (2d ed. 2000) (using social network analysis to analyze relationships and influence of agents in organizations using concepts of centrality, cliques, roles and positions).

(163.) See WILLIAM G. SCOTT, TERENCE R. MITCHELL & PHILIP H. BIRNBAUM, ORGANIZATIONAL THEORY: A STRUCTURAL AND BEHAVIORAL ANALYSIS 134-35 (4thed. 1981).

(164.) Id. at 135. See also RAMON J. ALDAG & LOREN W. KUZUHARA, ORGANIZATIONAL BEHAVIOR AND MANAGEMENT 298-306 (2002); NORMAN JACKSON & PIPPA CARTER, RETHINKING ORGANISATIONAL BEHAVIOUR 79-83 (2000); JOSEPH W. WEISS, ORGANIZATIONAL BEHAVIOR AND CHANGE 233-55 (2001).

(165.) ROBBINS, supra note 158, at 171.

(166.) See CYNTHIA COCKBURN, IN THE WAY OF WOMEN: MEN'S RESISTANCE TO SEX EQUALITY IN ORGANIZATIONS (1991).

(167.) Id.

(168.) See Schultz, supra note 117.

(169.) See YOAV VARDI & ELY WEITZ, MISBEHAVIOR IN ORGANIZATIONS: THEORY, RESEARCH, AND MANAGEMENT (2004) (citations omitted).

(170.) Id.

(171.) Catherine Fisk & Erwin Chemerinsky, Civil Rights Without Remedies: Vicarious Liability Under Title VII, Section 1983, and Title IX, 7 WM. &: MARY BILL RTS. J. 755, 759-81 (1999).

(172.) Id.

(173.) See, e.g., Costos v. Coconut Island Corp., 137 F.3d 46, 47 (1st Cir. 1998) (corporation held liable under section 219(2)(d) for an inn manager's rape of a guest where the manager used a key to the guest room he possessed by virtue of his position); Del Amora v. Metro Ford Sales & Serv., Inc., 206 F. Supp. 2d 947 (N.D. Ill. 2002) (holding a car dealership liable under section 219(2)(d) for its salesman's use of his position to obtain a credit report on his sister's husband for personal reasons related to his sister's pending divorce); Grease Monkey Int'l, Inc. v. Montoya, 904 P.2d 468, 470-75 (Colo. 1995) (holding a company liable for its officer's embezzlement of investors' funds where his position allowed him to commit the fraud); Groob v. Keybank, 801 N.E.2d 919 (Ohio Ct. App. 2003) (applying section 219(2)(d) to conclude that a bank could be held liable where its bank officer had rejected plaintiffs' loan request to buy a business and used their information to buy the business himself); Doe v. Forrest, 853 A.2d 48 (Vt. 2004) (sheriff's department not held vicariously liable where deputy sheriff sexually assaulted a convenience store employee while patrolling his beat). More rarely, isolated courts apply section 219(2)(d) in civil rights-type cases. See, e.g., LaRoche v. Denny's Inc., 62 F. Supp. 2d 1366, 1371-74 (S.D. Fla. 1999) (holding that a restaurant could be held vicariously liable for its manager's refusal to serve black customers, even though he was acting outside the scope of his employment, when the plaintiffs relied on his apparent authority and acquiesced to his demands); Serda v. Hancock, 842 F. Supp. 1315 (D. Kan. 1993) (finding hairdressing school vicariously liable under Title IX for sexual harassment of a student by the school director where the school had turned over all authority to manage the school to the director); but see Gebster v. Lago Vista Indep. Sch. Dist., 524 U.S. 274 (1998) (adopting actual knowledge standard for school liability under Title IX for teacher sexual harassment of students).

(174.) See, e.g., Harrison v. Eddy Potash, 112 F.3d 1437, 1445-46 (10th Cir. 1997) (applying section 219(2)(d) to hold employer vicariously liable), vacated and remanded for decision consistent with Ellerth, 524 U.S. 947 (1998); Karribian v. Columbia Univ., 14 F.3d 773, 780 (2d Cir. 1994) (applying "aided in agency" standard to hold that, in a case in which a high-level supervisor was involved in hostile environment sexual harassment, the employer should be liable regardless of the reasonableness of its complaint procedures); Bouton v. BMW of N. Am. Inc., 29 F.3d 103, 110 (3d Cir. 1994) (noting that applying section 219(2)(d) creates appropriate amount of deterrence by creating incentives "in the amount of the potential judgments to recruit, train, and supervise their managers to prevent hostile environments"); McCoy v. Macon Water Auth., 966 F. Supp. 1209, 1219-20 (D. Ga. 1997) (denying summary judgment to employer despite finding that the employer took prompt and effective remedial action after becoming aware of the plaintiff's complaint where a reasonable fact finder could conclude that the supervisor was aided in his harassing behavior by his status as an agent, based on the fact that he used his authority as a supervisor to require the plaintiff to come into his office and engage in lengthy conversations about sexual matters); Eichenwald v. Krigel's Inc., 908 F. Supp. 1531 (D. Kan. 1995) (finding that a supervisor was "aided in accomplishing his sexual harassment of the plaintiffs by virtue of his high level position with the defendants"); Henry v. Gehl Corp., 867 F. Supp. 960, 969 (D. Kan. 1994) (noting that under section 219(2)(d) mere proximity due to position would not be enough, but under the facts of the case, a reasonable finder of fact could find that the supervisor "capitalized on his authority over the plaintiff to create an intimidating and sexually-charged atmosphere in which the plaintiff realized the almost certain termination facing her if she challenged his offensive conduct"); State v. Schallock, 941 P.2d 1275 (Ariz. 1997) (applying section 219(2)(d) where employer had vested in the executive director the power to run its office); Lehmann v. Toys R Us, 626 A.2d 445 (N.J. 1993) (applying agency principles to analysis of employer liability for supervisor harassment under New Jersey anti-discrimination law).

(175.) N.J. STAT. ANN. [subsection] 10:5-1 to -49 (2002).

(176.) 801 A.2d 322, 329 (N.J. 2002).

(177.) Id. at 328.

(178.) 626 A.2d 445 (N.J. 1993).

(179.) Gaines, 801 A.2d at 328.

(180.) Id. at 313.

(181.) Id. at 329 (citing Lehman, 626 A.2d at 462).

(182.) Id. at 329 (citing Lehman, 626 A.2d at 463); see also Velez v. Jersey City, 817 A.2d 409 (N.J. Super. Ct. App. Div. 2003) (reversing grant of summary judgment to employer where facts cast doubt on adequacy of employer's response, even though no further incidents of harassment occurred).

(183.) See, e.g., Katherine Franke, What's Wrong with Sexual Harassment?, 49 STAN. L. REV. 691 (1997) (arguing that the wrong of sexual harassment is its use as a mechanism to enforce gender norms); Janet Halley, Sexuality Harassment, in DIRECTIONS IN SEXUAL HARASSMENT LAW 182 (Catherine A. MacKinnon & Reva B. Siegal, eds., 2004) [hereinafter DIRECTIONS] (warning against anti-gay uses of sexual harassment law); Vicki Schultz, supra note 117, at 1 (arguing that the chief harm of sexual harassment should be conceived as its use to block women from achieving workplace parity).

(184.) See, e.g., Kevin M. Clermont & Stewart J. Schwab, How Employment Discrimination Plaintiffs Fare in Federal Court, 1 J. EMPIRICAL LEGAL STUD. 429 (2004) (presenting data showing plaintiffs' low chances of winning employment discrimination cases and appeals after success before trial courts); David B. Oppenheimer, Verdicts Matter: An Empirical Study of California Employment Discrimination and Wrongful Discharge Jury Verdicts Reveals Low Success Rates for Women and Minorities, 37 U.C. DAVIS L. REV. 511, 535 (2003) (in a study of California jury verdicts in employment discrimination cases, plaintiffs won 68% of sexual harassment cases but only 41% of other discrimination cases.); Michael J. Zimmer, The New Discrimination Law: Price Waterhouse Is Dead, Whither McDonnell Douglas?, 53 EMORY L.J. 1887, 1943-44 (2004) (summarizing evidence showing that employment discrimination plaintiffs who allege claims other than sex harassment fare much worse than plaintiffs in other types of lawsuits).

(185.) Reva Siegal documents the history of creative turns in the development of anti-discrimination law in Reva B. Siegal, Introduction: A Short History of Sexual Harassment, in DIRECTIONS, supra note 183, at 1, 18-19.

(186.) Of course, to punish or deter the abuse of power in the workplace only when it is implemented according to an impermissible characteristic such as race or sex is wholly inadequate, as Regina Austin and others have pointed out. See Regina Austin, Employer Abuse, Worker Resistance, and the Tort of Intentional Infliction of Emotional Distress, 41 STAN. L. REV. 1 (1988).

(187.) Judge Motz is perhaps best known for her recent, stinging dissent in Hamdi v. Rumsfeld, 316 F.3d 450, 477 (4th Cir. 2003), in which she argued against the indefinite detention of a U.S. citizen without charges or access to a lawyer. See Nat Hentoff, Who Made George W. Bush Our King?, VILLAGE VOICE, July. 25, 2003, available at http://www.villagevoice.com/news/ 0331,hentoff,45847,6.html (praising Judge Motz's opinion).

(188.) No. 99-1566, 2000 U.S. App. LEXIS 5776 (4th Cir. Mar. 30, 2000).

(189.) For an excellent article critiquing the growing practice of classifying opinions as unpublished, see Penelope Pether, Inequitable Injunctions: The Scandal of Private Judging in the U.S. Courts, 56 STAN. L. REV. 1435 (2004).

(190.) Williams, 2000 U.S. App. LEXIS 5776, at *2.

(191.) Id. at *5.

(192.) Id. at *11.

(193.) Id. at *5.

(194.) Id. at *6 (emphasis in original).

(195.) Id.

(196.) Id.

(197.) Id at *7-8.

(198.) Id. at *9.

(199.) Id.

(200.) By pointing out the female identity of two judges, I do not mean to hint at an essentialist implication that women judges necessarily do better in sexual harassment cases, though some empirical evidence suggests that a combination of variables involving race, gender, age, and political affiliation does affect the likelihood that judges will find for plaintiffs in sex discrimination and sex harassment cases. See BEINER, supra note 146, at 7-8 (summarizing studies). Essentialism has been roundly trounced on both theoretical and normative grounds, see, e.g., Margaret Radin, Reply: Please Be Careful With Cultural Feminism, 45 STAN. L. REV. 1567 (1993), but, that noted, experience plainly does affect judgment, in unpredictable and individually mediated ways, cf. Susan Carle, Women in Law, 8 AM. U. J. GENDER SOC. POL'Y & L. 795, 796 (2000), and to throw out this important legal realist insight simply because its vulgar extension into essentialism is discredited seems to me unnecessary.

(201.) Petrosino v. Bell Atlantic, 385 F.3d 210 (2d Cir. 2004).

(202.) Id. at 214.

(203.) Id. at 215.

(204.) Id. at 214-15.

(205.) Id. at 215.

(206.) Id.

(207.) Id. at 216 (citation to the record omitted).

(208.) Id.

(209.) Id.

(210.) Id. at 216.

(211.) Id. at 213.

(212.) Id.

(213.) Id. at 224-25.

(214.) Id. at 226.

(215.) Id. Some other cases within the Second Circuit similarly subjects employers' assertions of the affirmative defense doctrine to more rigorous scrutiny. See, e.g., Presley v. Pepperidge Farm, Inc., 356 F. Supp. 2d 109,129-31 (D. Conn. 2005) (denying summary judgment to employer where there was evidence that a supervisor failed to report to superiors in the company hierarchy his knowledge of allegations of sexual harassment); Miller v. Edward Jones & Co., 355 F. Supp. 2d 629 (D. Conn. 2005) (denying summary judgment to an employer that made a decision not to separate the plaintiff and her alleged harasser, on the ground that a reasonable trier of fact could view this as evidence of an inadequate company response). See also Mack, 326 F.3d at 125 (rejecting employer's claim that an employee committing sexual harassment was a co-worker in light of the power he possessed over plaintiff).

(216.) 42 U.S.C. [section] 2000e(b) (2000).

(217.) Ellerth, 524 U.S. at 765.

(218.) Suders, 542 U.S. at 134.

(219.) Id.; Ellerth, 524 U.S. at 765.

SUSAN D. CARLE, Visiting Professor of Law, Harvard Law School; Professor of Law, American University Washington College of Law. This Article is an extension of a presentation given at the 2005 Law & Society Annual Meeting. It also draws on the author's work as an attorney at the labor law firm of Bredhoff & Kaiser, in Washington, D.C., from 1991-97, where she was involved in sexual harassment consulting and training. She wishes to thank Candace Kovacic-Fleischer and Ann McGinley for extremely helpful comments on earlier drafts, Hilary Dengel and Rebecca Geller for outstanding research assistance, and Aric Elsenheimer for discussions about his Note on a related topic. See Aric G. Elsenheimer, Agency and Liability in Sexual Harassment Law: Toward a Broader Definition of Tangible Employment Actions, 54 AM. U. L. REV. 1635 (2005).