UNITED STATES DISTRICT COURT NORTHERN DISTRICT OF ILLINOIS EASTERN DIVISION
October 23, 2007
EQUAL EMPLOYMENT OPPORTUNITY COMMISSION, PLAINTIFF,
INTERNATIONAL PROFIT ASSOCIATES, INC., DEFENDANT.
The opinion of the court was delivered by: Judge Joan B. Gottschall
MEMORANDUM OPINION AND ORDER
This is an employment discrimination suit brought by the Equal Employment Opportunity Commission ("EEOC") against International Profit Associates, Inc. ("IPA") pursuant to 42 U.S.C. § 2000e-2 (2006) and 42 U.S.C. § 2000e-5 (2006). The EEOC alleges that IPA fostered a hostile work environment at its business offices in Buffalo Grove, Illinois, because IPA's female employees were routinely subjected to severe and pervasive sexual harassment and IPA failed to take appropriate action to stop the abuse. The EEOC seeks injunctive relief pursuant to the pattern or practice theory of employer liability, see Int'l Brotherhood of Teamsters v. United States, 431 U.S. 324, 335-36 (1977), and also compensatory and punitive damages on behalf of each of the 113 individual claimants that make up the "class"*fn1 pursuant to 42 U.S.C. § 1981a (2006).
The issues currently before the court arise from the court's opinion of March 16, 2007, EEOC v. Int'l Profit Assocs., Inc., No. 01 C 4427, 2007 WL 844555 (N.D. Ill. Mar. 16, 2007). The court's prior opinion dealt with IPA's motion for summary judgment, in which IPA argued that judgment as a matter of law should be entered on many of the individual class members' claims because those claimants cannot establish the requisite elements of a hostile work environment sexual harassment claim. Id. at *4. In response to IPA's motion, the EEOC argued that it is not required to prove each claimant's case individually-even though individual damages are sought-because this is a pattern or practice action and the focus of such an action is the employer's work environment as a whole. Id. at *5. The EEOC contended that certain elements of the individual claimants' cases would be established through proof that IPA engaged in a pattern or practice of unlawful discrimination (i.e., proven as to the entire class), and that the court therefore could not reach IPA's summary judgment motion until the pattern or practice aspect of this case was complete. Id.
Obviously, the parties had very different ideas about the manner in which this case should proceed. Unfortunately, neither the EEOC nor IPA had addressed key issues that required resolution before the court could rule on IPA's summary judgment motion. Specifically, the parties disagreed on numerous prefatory legal issues, but had not directly addressed the method of proof for the EEOC's claims. The court was uncomfortable ruling without giving the parties the opportunity to provide specific, focused input regarding the applicable legal framework, id. at 4, so the prior opinion simply resolved several preliminary disputes, outlined the remaining issues, and directed the parties to submit briefs addressing a number of discrete legal questions, id. at * 6-14. IPA's motion for summary judgment was denied without prejudice, and the court indicated that a new schedule for dispositive motions would be set after the court ruled on the governing legal standards. Id. at *1.
Specifically, the court rejected IPA's suggestion that the EEOC cannot maintain this case as a pattern or practice suit because the EEOC is proceeding under section 706 of Title VII, 42 U.S.C. § 2000e-5 (which sets forth the procedures the EEOC must follow when it brings suit based on a charge of discrimination filed by an employee, as in this case) as opposed to section 707 of Title VII, 42 U.S.C. § 2000e-6 (2006) (which specifically authorizes a the EEOC to bring a "pattern or practice" action on its own initiative). Int'l Profit Assocs., 2007 WL 844555, at *6-9. As the court previously stated, the EEOC is statutorily authorized to bring a pattern or practice action to enjoin systemic unlawful employment practices, as defined by 42 U.S.C. § 2000e-2(a)(1), under either section 706 or section 707. Id. at *9.
The court also determined that the EEOC may bring a pattern or practice suit premised on the hostile work environment theory of sexual harassment, a point that should be obvious since the Supreme Court has long held that sexual harassment that constructively alters the terms and conditions of employment by creating a hostile work environment is an unlawful employment practice. See id. at *9 (citing, inter alia, Burlington Indus., Inc. v. Ellerth, 524 U.S. 742, 752 (1998)). Less obvious, however, is how a pattern or practice case involving the hostile work environment theory of sexual harassment should be tried. Id. at 11.
As the court detailed in its previous opinion, the traditional "pattern or practice" framework was developed by the Supreme Court in Teamsters, a case dealing with racially discriminatory hiring and promotion practices. Id. at *6. In Teamsters, the Court adapted the well-known McDonnell Douglas race discrimination test to the pattern or practice method of proof, holding that the EEOC can establish a prima facie pattern or practice of race discrimination by proving that the employer regularly and systematically discriminated in its employment decisions-i.e., "that 'discrimination was the company's standard operating procedure.'" Id. (quoting Teamsters, 431 U.S. at 336). If the EEOC does so (and the employer does not rebut the EEOC's prima facie case), the employer may be held liable for violating Title VII by discriminating on the basis of race, and the EEOC can obtain "'prospective relief.'" Id. at *7 (quoting Teamsters, 431 U.S. at 361).
Where individual relief is also sought, however, the EEOC must still prove that the individual claimants were victims of the discriminatory policy. Id. For unlawful hiring and promotion claims, the EEOC's initial burden is not a particularly onerous one; the EEOC must show only that the individual claimants sought but did not receive a job or promotion during the time the discriminatory policy was in place. Id. By virtue of the pattern or practice finding, the employer's individual employment decisions are then presumed to be the result of the discriminatory policy. Id. Of course, the employer can rebut this presumption by demonstrating a non-discriminatory reason for a particular decision, and the EEOC is then entitled to offer evidence that employer's proffered reason is pretextual. Id.
As the court explained in its previous decision, in a race discrimination case it is clear why a pattern or practice finding should have an effect on an employer's liability to individual claimants. If an employer has an established policy of making employment decisions with racial animus in violation of Title VII, it is likely that any specific employment decision also violates Title VII, and if a particular decision was not discriminatory, the employer is in the best position to show why. Id. at 11; see also Teamsters, 431 U.S. at 359 n.45 ("[T]he employer was in the best position to show why any individual employee was denied an employment opportunity."). However, the impact of a pattern or practice finding in a hostile work environment sexual harassment case is not so clear. In contrast with a race discrimination case-where the focus is on the employer's basis for making an employment decision that adversely affected the claimant-a sexual harassment case centers on the gravity of the conduct to which a claimant was exposed. Int'l Profit Assocs., 2007 WL 844555, at *11. The sexual harassment suffered by the claimant must have been severe or pervasive enough (measured both objectively and subjectively) to constructively alter the terms or conditions of the claimant's employment by creating a hostile work environment. Id. (citing Harris v. Forklift Sys., Inc., 510 U.S. 17, 21-22 (1992); Whittaker v. N. Ill. Univ., 424 F.3d 640, 645 (7th Cir. 2005)). Otherwise, no Title VII violation has occurred. Id. at *9-10. Therefore, a finding that an employer had a pattern or practice of tolerating sexual harassment in violation of Title VII does not necessarily establish that an individual claimant was exposed to harassment or that the harassment an individual claimant suffered violates Title VII. It is thus unclear what effect a pattern or practice finding should have on an individual claimant's suit for damages.*fn2 Id. at *11.
The Supreme Court has never considered how a hostile work environment sexual harassment case should be tried under the pattern or practice model and, to this court's knowledge, neither has any court of appeals. In the briefing on IPA's motion for summary judgment, the EEOC identified several district court cases that have wrestled with the issue, see, e.g., EEOC v. Dial Corp., 156 F. Supp. 2d 926 (N.D. Ill. 2001) (modified by EEOC v. Dial Corp., 259 F. Supp. 2d 710, 712-16 (N.D. Ill. 2003)); EEOC v. Foster Wheeler Constructors, Inc., No. 98 C 1601, 1999 WL 528200, at *2-3 (N.D. Ill. July 13, 1999) (considering method of proof in hostile work environment pattern or practice case involving individual damages); EEOC v. Mitsubishi Motor Mfg. of Am., Inc., 990 F. Supp. 1059, 1069-82 (C.D. Ill. 1998) (same), but the court had reservations as to whether the lead case, Mitsubishi, properly applied Teamsters to the Supreme Court's hostile work environment jurisprudence and, in any event, the court wanted the parties to provide briefing on the difficult issues presented by this case before making its decision. Int'l Profit Assocs., 2007 WL 844555, at *14 n.20. The court therefore directed the parties to brief the following issues:
(1) what, specifically, the EEOC must prove in order to establish that IPA has violated Title VII by allowing a pattern or practice of employment discrimination; (2) what, specifically, the EEOC must prove in order to recover damages under section 1981a on behalf of the individual claimants; and (3) what effect, if any, a finding in the EEOC's favor as to the pattern or practice aspect of this case should have on the individual aspects of this case.
Id. at *14. The parties have done so, and the court commends them for their efforts. The parties' briefs are well-focused and comprehensive, and the court has found the parties' submissions extremely helpful in considering the full implications of the decision at hand.
In conjunction with its brief regarding the applicable trial framework, the EEOC renewed its motion for bifurcation, as this court invited it to do in the previous order. See id. at *14 n.21. This opinion therefore addresses not only the method of proof for the EEOC's case but also the EEOC's motion. For the reasons set forth in further detail below, this case will be bifurcated into two phases, a pattern or practice phase ("Phase I") and an individual phase ("Phase II"), but will be decided by one jury according to the principles set forth in this opinion.
A. The EEOC's Theory
The EEOC asks the court to bifurcate the trial of this case into two phases pursuant to Rule 42(b) of the Federal Rules of Civil Procedure: a "liability" phase and a "remedial" phase. EEOC's Renewed Mot. to Bifurcate 3. According to the EEOC, the liability phase would focus on the pattern or practice aspect of the case, but would also include a verdict on punitive damages. Thus, punitive damages would be awarded as to the entire class, but not allocated amongst individual class members until the conclusion of the remedial phase.*fn3 A finding that IPA engaged in a pattern or practice of Title VII discrimination at the liability phase would also allow the court to award the EEOC injunctive relief. Under the EEOC's theory, the remedial phase would then focus on the individual claimants' compensatory damages claims.
The EEOC acknowledges that in this case, as with any individual case, it must prove every element of the hostile work environment sexual harassment cause of action in order to prevail. The elements of such a claim were set forth in the court's prior opinion, and are as follows: (1) the alleged misconduct must have been based on sex (i.e., directed at the plaintiff because of her sex); (2) the conduct must have been severe or pervasive enough to render the plaintiff's work environment both objectively and subjectively hostile; and (3) there must be a basis for holding the employer liable for the harassment to which the plaintiff was exposed (i.e., negligence, unless the harassment was perpetrated by a supervisor). See EEOC v. Int'l Profit Assocs., Inc., No. 01 C 4427, 2007 WL 844555, at *9-10 (N.D. Ill. Mar. 16, 2007) (citing, inter alia, Nat'l R.R. Passenger Corp. v. Morgan, 536 U.S. 101, 115 (2002), Harris v. Forklift Sys., Inc., 510 U.S. 17, 21-23 (1992), and Whittaker v. N. Ill. Univ., 424 F.3d 640, 645 (7th Cir. 2005)). The EEOC contends, however, that because this is a pattern or practice sexual harassment case, certain of these elements should be proven with respect to the entire class (i.e., at the liability phase) and others should be proven with respect to the individual claimants (i.e., at the remedial stage). In other words, the EEOC advocates a split method of proof. Specifically, the EEOC argues that every element of the hostile work environment cause of action, with the exception of the subjective component of the severe or pervasive inquiry, should be proven at the pattern or practice stage. Thus, according to the EEOC, the liability phase jury should decide whether the sexual harassment that occurred at IPA during the relevant time period, taken as a whole, was so severe or pervasive that a reasonable woman would find the work environment at IPA to be hostile. The liability phase jury would also consider whether IPA had a "policy of tolerating (and therefore condoning and/or fostering)" such an environment. EEOC Renewed Mot. to Bifurcate 10. The EEOC concedes that this latter inquiry-which goes to the question of IPA's liability-should be resolved under the negligence standard that applies to harassment by co-workers, as opposed to the strict liability standard that applies to harassment by supervisors (subject to the affirmative defenses established in Faragher v. City of Boca Raton, 524 U.S. 775 (1998), and Burlington Indus., Inc. v. Ellerth, 524 U.S. 742 (1998)).
See Int'l Profit Assocs., 2007 WL 844555, at *10. This is because the pattern of harassment that the EEOC accuses IPA of tolerating was perpetrated by the claimants' supervisors and co-workers alike, and it is therefore impossible to distinguish between the actions taken by the two types of perpetrators for purposes of assessing IPA's pattern or practice liability.*fn4 Therefore, according to the EEOC, IPA should be found negligent, and thus liable, if IPA had notice (actual or constructive) that regular or systematic sexual harassment was occurring but did not take adequate steps to address the problem on a company-wide basis. The EEOC asserts that if it establishes these two components- objectively severe or pervasive sexual harassment and a company policy of tolerating it- IPA must be found liable for the pattern or practice. Under the EEOC's theory, this would justify not only injunctive relief but also punitive damages. A finding with respect to the latter is appropriate at the liability stage, the EEOC asserts, because the showing that the EEOC must make to receive punitive damages-that IPA acted with malice or reckless disregard for the federally protected rights of its employees, see 42 U.S.C. § 1981a(b)(1) (2006)-overlaps with the proof it (and IPA, in rebuttal) will offer on the pattern or practice elements.
Once it has established IPA's pattern or practice liability, the EEOC argues, the remedial stage will focus on the individual claimants' subjective perception of the harassment to which they were exposed and the extent of their damages. Thus, according to the EEOC, at the remedial stage it must only prove (through the testimony of the individuals themselves and relevant witnesses) for each claimant that she was exposed to the pattern or practice of a hostile work environment (i.e., that she worked at IPA during the relevant time period and that she witnessed or experienced at least one incident of sexual harassment); that the harassment she experienced was unwelcome; and that she was harmed by it.
EEOC Renewed Mot. to Bifurcate 9-10. The EEOC asserts that all other elements of the cause of action are conclusively established at the liability stage, and that "IPA cannot contend . . . that an individual woman was not subjected to severe or pervasive harassment" or "that it was not negligent as to a particular class member." Id. at 12. The EEOC does allow, however, that IPA could argue "that individual claimants were ill-treated for a reason other than because of their sex, and, of course, that they were not subjectively offended (e.g., by arguing that the alleged conduct was welcome)." Id. If the remedial stage jury determines that an individual claimant was offended by the conduct to which she was exposed, then it must award the claimant compensatory damages. Under the EEOC's theory, the court would then allocate the punitive damages awarded by the liability phase jury according to the percentage of the total compensatory damages awarded to each claimant by the remedial phase jury.
B. Precedent Supporting the EEOC's Position
To support the framework outlined above, the EEOC relies primarily on two district court cases (as the court noted above, no appellate courts have addressed this issue): EEOC v. Mitsubishi Motor Mfg. of Am., Inc., 990 F. Supp. 1059 (C.D. Ill. 1998), and EEOC v. Dial Corp., 156 F. Supp. 2d 926 (N.D. Ill. 2001) (modified by EEOC v. Dial Corp., 259 F. Supp. 2d 710 (N.D. Ill. 2003)). Both cases dealt with the same question that is before this court: how should a pattern or practice case involving the hostile work environment theory of sexual harassment be tried, given that the EEOC seeks both injunctive relief and compensatory and punitive damages on behalf of the individual claimants? In Mitsubishi, the court opted for a bifurcated framework similar (though, as discussed below, different in several key respects) to that sought by the EEOC here. Specifically, the Mitsubishi court held that to establish a pattern or practice of sexual harassment, the EEOC would have to prove by a preponderance of the evidence that (1) an objectively reasonable person would find that severe or pervasive sexual harassment rendered the employer's work environment, taken as a whole, hostile; and (2) the employer knew or should have known that its work environment was permeated with sexual harassment but did not properly address the problem on a company-wide basis. 990 F. Supp. at 1073-76. The court noted that "[t]he purpose and effect of a successful pattern or practice case is to impose liability on a private employer . . . for discriminatory policies created and maintained by an employer that result in a system-wide pattern or practice of disparate treatment against individuals who fall within a protected class." Id. at 1077. Thus, the court decided that an individual employee's subjective perception regarding the conduct to which she was exposed was not relevant to the issue of whether the employer fostered a discriminatory policy toward women in general. Id. at 1076. Accordingly, the Mitsubishi court modified the test established by the Supreme Court in Harris by excluding the subjective element from the pattern or practice phase. Id. at 1076-77. The court stated, however, that a pattern or practice finding would authorize the EEOC to obtain only injunctive relief. Id. at 1077. If the EEOC wished to recover monetary damages, it would also have to present individual proof at an "individual relief" phase. Id.
The individual proof contemplated by Mitsubishi, however, was minimal. Relying on Teamsters, the court adopted what it called "a rebuttable presumption of individual liability to all women in the workplace" based on the pattern or practice finding. Id. at 1079. The court noted that Teamsters did not intuitively fit the hostile work environment scenario. See, e.g., id. at 1078 ("A finding of pattern or practice liability for sexual harassment does not give rise to any obvious presumptions with regard to individual liability for sexual harassment . . . ."). But the court nonetheless constructed a burden-shifting framework because it believed that the individual class members should receive some benefit from a finding that the employer had a pattern or practice of tolerating sexual harassment. Id. Mitsubishi therefore permitted the EEOC to satisfy the objective component of the hostile work environment test for the individual claimants by aggregating the acts of harassment relied on during the pattern or practice phase.*fn5 Id. The individual claimants, even those who experienced only marginal sexual harassment, were thus relieved of their burden to show that the acts they experienced were severe or pervasive. The court based this ruling on its belief that
[a]lthough "isolated" or "sporadic" instances of harassment experienced by a person in an individual action are typically not enough to establish hostile environment sexual harassment, single instances of conduct must be taken together with the continuous pattern or harassment in the workplace, as a whole, which was established at the pattern or practice phase.
Id. at 1081.*fn6 The court also shifted the burden of production on the subjective element- which the court referred to as "virtually a formality," id. at 1078-to the defendant, thus requiring the employer to come forward with evidence that the harassment to which an individual claimant was exposed was welcome. Id. Finally, the court noted that a pattern or practice finding would justify an inference that the employer knew of and tolerated the sexual harassment to which the individual claimants were subjected.*fn7 Id. at 1080. Therefore, under the Mitsubishi framework, the subjective element is the only component of the hostile work environment test which must be established at the individual stage, and the burden of production with respect to this element (and all other elements) is on the employer. If the employer does not present evidence to rebut the individual plaintiffs' claims, it will be held liable and the individual claimants will be entitled to compensatory (and perhaps punitive) damages. Id. at 1081-82.
Dial leaned heavily on Mitsubishi, and the framework adopted by the Dial court mirrors that set forth in Mitsubishi with a few important exceptions. See generally Dial, 156 F. Supp.2d at 946-58 (discussing, inter alia, Mitsubishi and adopting bifurcated trial framework according to Mitsubishi model). First, Dial suggested that a subjective showing should also be required at the pattern or practice phase of the trial. See Dial, 156 F. Supp. 2d at 950 ("Generally, in the pattern-or-practice context, a finding of unwelcomeness will be justified where the plaintiff class shows by a preponderance of the evidence that women, by their conduct, indicated that the acts of sexual harassment were unsolicited and regarded as undesirable or offensive." (citation and internal quotation marks omitted)). Second, Dial decided to divide the claimants into groups at the individual phase of the trial and try their claims using several different juries (as opposed to one jury, as implied by Mitsubishi). Id. at 958 n.18. Finally (and most importantly for present purposes), the Dial court later revisited its initial decision on the trial framework and decided to include punitive damages at the pattern or practice phase. Dial, 259 F. Supp. 2d at 713. In relevant part, the court stated:
Phase I should result in a finding by the jury of whether and when a pattern or practice of tolerating sexual harassment existed. Because evidence to support the claim of such a pattern or practice necessarily will include some episodes that may tend to show whether the pattern or practice was engaged in with malice or reckless indifference to the federally protected rights of aggrieved persons of the class, such state-ofthe-mind evidence may be received at Phase I and the malice-or-reckless-indifference issue should be decided by the same jury as decides the pattern or practice issue.
Id. In all other important respects, Dial followed the framework established by the Mitsubishi court.
It is worth noting that neither Mitsubishi nor Dial ever reached trial. In fact, the EEOC cites only one pattern or practice sexual harassment case that was actually tried: Jenson v. Eveleth Taconite Co., 824 F. Supp. 847 (D. Minn. 1993). Like Teamsters, Jenson involved allegations regarding a pattern or practice of discriminatory employment decisions-specifically, hiring and promotion. Id. at 856. However, the plaintiff class joined hostile work environment claims based on sexual harassment with their Title VII discrimination claims, and the court tried both types of claims using the pattern or practice model. Id. at 879-88. The framework used by the court for the hostile work environment claims was similar to that adopted in Mitsubishi (although the analysis in Jenson was much less comprehensive); indeed, the Mitsubishi court relied heavily on Jenson in formulating its approach. See Mitsubishi, 990 F. Supp. at 1074 (discussing Jenson's articulation of method of proof for pattern or practice hostile work environment sexual harassment action). Jenson, like Mitsubishi, expressed reservations about applying the Teamsters pattern or practice approach to the hostile work environment setting. Jenson, 824 F. Supp. at 875-76. The court noted that a typical pattern or practice discrimination case focuses on the employer's decisions, whereas a hostile work environment case centers on the seriousness of the harassment to which an employee was exposed. Id. However, the court appeared to be concerned only with the treatment of the subjective component of the inquiry, stating that "the proof introduced during the liability phase cannot resolve a disparate issue of fact essential to claims alleging hostile work environment: whether individual members of the plaintiff class were as affected as the reasonable woman would have been." Id. at 876. The court therefore-in contrast with Mitsubishi-refused to shift any burden on the subjective component to the employer during the individual phase of the case.*fn8 Id.
More importantly, Jenson held that the subjective component of the hostile work environment cause of action was the only element at issue during the individual phase. See id. ("[I]ndividual class members need only show that they were at least as affected as the reasonable woman. The other elements of a hostile work environment claim are established by the court's determination in the liability phase of the proceedings."). This is another important distinction from Mitsubishi: the employer in Jenson was not permitted to rebut the impact of the pattern or practice finding on other elements of the individual claimants' hostile work environment claims at the individual phase, for those elements had already been conclusively established based on the court's pattern or practice finding. See id. at 879-88 (finding the defendant had engaged in a pattern or practice of sexual harassment). The court imposed injunctive relief based on the pattern or practice finding, and indicated that it would determine the plaintiffs' damages after considering the subjective component during "the recovery phase of the proceedings." Id. at 888.
C. The Court Rejects the EEOC's Proposed Method of Proof
As the above recitation of the EEOC's theory demonstrates, the EEOC is seeking to blend aspects of Mitsubishi, Dial, and Jenson to create an approach that will dramatically increase its odds of prevailing in this case, particularly with respect to its ability to recover punitive and compensatory damages.*fn9 In essence, the EEOC wants to prove every element (save the subjective showing) of the individual claimants' cases by aggregating its evidence against IPA at the class stage. Under the EEOC's theory, it would then be entitled not only to injunctive relief but also to punitive damages, so long as the jury found IPA reckless or malicious. To recover compensatory damages, the EEOC would then be required only to present evidence that the individual claimants did not welcome the harassment to which they were exposed. IPA would not be permitted to argue that the harassment a specific claimant experienced was not actionable or that it took appropriate remedial action with respect to a particular claimant, for these elements would be established based on the EEOC's pattern or practice showing. Rather, IPA's only defenses to the individual claimants' damages claims would be to argue that the harassment was not based on sex or that the claimants were not subjectively offended by the conduct.
The court rejects the EEOC's proposed method of proof because it is legally flawed in numerous respects. As an initial matter, the EEOC overlooks what is, in this court's opinion, binding Seventh Circuit precedent with regard to the EEOC's position on punitive damages. In Jefferson v. Ingersoll, 195 F.3d 894, 896-97 (7th Cir. 1999), a class action race discrimination suit, the Seventh Circuit considered whether the district court could certify a class under Rule 23(b)(2) of the Federal Rules of Civil Procedure when the plaintiffs sought not only injunctive relief under the pattern or practice theory, but also punitive and compensatory damages under 42 U.S.C. § 1981a (2006). In holding that the district court could do so "only when monetary relief is incidental to the equitable remedy," id. at 898, the Seventh Circuit had the following to say about awarding damages on a class-wide basis: "Money damages under § 1981a(b) are neither injunctive nor declaratory, and they do not affect a class as a whole. It is possible for one applicant for employment to recover substantial damages while another recovers nothing (for example, because the second person would have been rejected under nondiscriminatory conditions, or found a better job elsewhere)." Id. at 897.
The court elaborated on this issue in Lemon v. Int'l Union of Operating Engineers, Local No. 139, AFL-CIO, 216 F.3d 577, 579 (7th Cir. 2000), a case in which a class of plaintiffs sued a union for allegedly administering an employee referral service in a discriminatory fashion. The Seventh Circuit was faced with the same issue as in Jefferson-namely, the applicability of Rule 23's certification requirements to a pattern or practice case in which section 1981a damages were also sought (the district court had decided the certification question before Jefferson was issued). Id. Regarding damages, the Seventh Circuit stated as follows:
Damages can be awarded only after proof of discrimination and injury specific to the individual plaintiff, so deciding the damages claims depends on an individualized analysis of each class member's circumstances and requires additional hearings to resolve the disparate merits of each individual's case. Even if the plaintiffs prove that [the defendant] administered the referral hall in a discriminatory manner and won injunctive and declaratory relief on that ground, each individual plaintiff pursuing damages claims still would need to establish that [the defendant's] discrimination caused her personal injury and would need to show the magnitude of the injury to determine compensatory damages. Similarly, to win punitive damages, an individual plaintiff must establish that the defendant possessed a reckless indifference to the plaintiff's federal rights-a fact-specific inquiry into that plaintiff's circumstances. Id. at 581 (citations omitted) (emphasis added).
Both Jefferson and Lemon relied heavily on Allison v. Citgo Petroleum Corp., 151 F.3d 502 (5th Cir. 1998), yet another race discrimination class action in which equitable relief as well as section 1981a damages were sought. Id. at 407-08. There, the Fifth Circuit rejected the plaintiffs' attempt to seek class-based compensatory and punitive damages. Id. at 417-18. With respect to compensatory damages, the court stated that "by requiring individualized proof of discrimination and actual injury to each class member, compensatory damages introduce new and substantial legal and factual issues" and therefore cannot be awarded based on a pattern or practice finding. Id. at 417.
As to punitive damages, the court assumed arguendo that such damages could be awarded on a class-wide basis-although the court pointed out that the plain language of section 1981a might be read to preclude such relief. See id. (noting that punitive damages are available under section 1981a only when the employer has acted with malice or reckless indifference to the federal rights of an "aggrieved individual"). Nonetheless, the court noted that the possibility of a class-wide punitive damages award would exist only "where the entire class or subclass is subjected to the same discriminatory act or series of acts." Id. Such was not the case in Allison; as the court stated, "The plaintiffs challenge broad policies and practices, but they do not contend that each plaintiff was affected by these policies and practices in the same way." Id. The court further found that "[p]unitive damages cannot be assessed merely on a finding that the defendant engaged in a pattern or practice of discrimination" because "[s]uch a finding establishes only that there has been general harm to the group and that injunctive relief is appropriate." Id.
The Allison court also noted that in BMW v. Gore, 517 U.S. 559, 573-80 (1996), the Supreme Court held that punitive damages must be reasonably related to the reprehensibility of the defendant's conduct and the amount of compensatory damages awarded (the Court has since reiterated its position on this issue in State Farm Mut. Auto. Ins. Co. v. Campbell, 538 U.S. 408 (2003)). Allison, 151 F.3d at 417-18. Because compensatory damages must be awarded on an individual basis, "punitive damages must be determined after proof of liability to individual plaintiffs at the second stage of a pattern or practice case, not upon the mere finding of general liability to the class at the first stage." Id. at 418. The court therefore declined to allow the plaintiff class to seek class-based punitive damages. Id.
Although Jefferson, Lemon, and Allison all dealt with class certification issues, the court concludes that they foreclose the EEOC's request that punitive damages be determined in Phase I. The Supreme Court has stated that the substantive standards applicable to pattern or practice actions brought by the EEOC are the same as those that apply to class actions brought by private plaintiffs, Cooper v. Fed. Reserve Bank of Richmond, 467 U.S. 867, 876 n.9 (1984), and the court cannot discern any reason (the EEOC does not articulate one) to treat this case differently than the class cases considered by the Seventh Circuit. Although the Seventh Circuit has not explicitly passed judgment on the question of whether punitive damages can be awarded on a class basis, it has certainly indicated its leaning-based on Allison-in Jefferson and Lemon. Furthermore, there is no question that the claimants in this case were not subject to the same discriminatory act or series of acts. Rather, as in Allison, each of the EEOC's 113 claimants alleges different facts in support of her claim.
The EEOC's reliance on Palmer v. Combined Ins. Co. of Am., 217 F.R.D. 430, 438-39 (N.D. Ill. 2003) is thus misplaced, for in Palmer the class plaintiffs sought only punitive, not compensatory, damages, and their allegations of discrimination were premised on the same basic series of acts. Based on the potential exception left open by Allison, the court allowed the plaintiffs to attempt to prove class-wide punitive damages because it found that individualized assessments might not be required to determine the punitive damages award and that such an award could thus possibly be distributed on a pro rata basis. Id. at 439. This case, however, is not the "rare exception" in which individualized inquiries are unnecessary, like Palmer. Id. It is entirely possible that one claimant will recover punitive damages while another will not (for instance, because IPA took proper remedial action to correct an individual instance of harassment)-precisely what the Seventh Circuit was concerned with when it refused to allow class-based punitive damages in Jefferson and Lemon. Punitive damages are therefore not a common question that can be resolved as to all 113 claimants, and such damages cannot be premised on a mere pattern or practice finding, which authorizes only prospective relief under Teamsters. Int'l Brotherhood of Teamsters v. United States, 431 U.S. 324, 361 (1977).
Accordingly, the court rejects the EEOC's argument, based on Dial, that punitive damages should be determined in Phase I. In this case, the jury will consider punitive damages at the individual stage, and will make a determination as to whether IPA was reckless or malicious with respect to the federal rights of each individual claimant. See Lemon, 216 F.3d at 581. This will allow the jury to make the fact-specific determinations necessary for each claimant, and will also allow the court to ensure that the Supreme Court's concerns regarding punitive damages articulated in Gore and State Farm are met. The court recognizes that some of the questions pertinent to the determination of punitive damages overlap with the pattern or practice evidence that will be offered at Phase I. See Bruso v. United Airlines, Inc., 239 F.3d 848, 857-59 (7th Cir. 2001) (employer may avoid punitive damages if it proves it made a good faith effort to comply with Title VII by implementing an anti-discrimination policy). However, as discussed below, this case will be tried by one jury, and therefore clear jury instructions in conjunction with effective advocacy should alleviate the EEOC's concerns that the jury will be confused by the overlapping issues.
The court also finds several flaws in the EEOC's proposed adaptation of the Teamsters pattern or practice method of proof to the hostile work environment scenario. Specifically, the court rejects the EEOC's contention that the objective element of the individual plaintiffs' hostile work environment claims is definitively established at the pattern or practice phase. According to the EEOC, "a woman is a victim of a company's pattern or practice of sexual harassment (and entitled to damages if she was subjectively offended), although the incidents of harassment she experienced may not be independently severe or pervasive, because it was the company's practice of tolerating sexual harassment that allowed the harassment to occur." EEOC's Renewed Mot. to Bifurcate 13. In other words, the EEOC believes that it can establish the objective component of the individual plaintiffs' cases-even for claimants who did not themselves experience actionable harassment-by showing that the sexual harassment that occurred company-wide at IPA was sufficiently severe or pervasive that a reasonable person would find the work environment as a whole to be hostile.
The EEOC's belief that it can meet its burden of proof on the objective element by aggregating the acts of harassment to which all the claimants were subjected to derives primarily from Mitsubishi, which eliminated the objective component of the severe or pervasive test at the individual stage.*fn10 As discussed above, the court's rationale for doing so was that "[a]lthough 'isolated' or 'sporadic' instances of harassment . . . are typically not enough to establish hostile environment sexual harassment, single instances of conduct must be taken together with the continuous pattern of harassment in the workplace, which was established at the pattern or practice phase." Mitsibishi, 990 F. Supp. at 1081. The court therefore permitted the individual claimants to rely on the objective showing made as to the entire workplace at the pattern or practice phase rather than requiring them to make individual showings that the specific harassment they suffered was objectively severe. Id. at 1079-81.
The court declines to follow Mitsubishi's approach (and, concurrently, that adopted in Dial and Jenson) because the court does not see how Mitsubishi can be reconciled with the Supreme Court's hostile work environment jurisprudence. In Meritor Sav. Bank, FSB v. Vinson, 477 U.S. 57 (1986), the case in which the Supreme Court first articulated what has been the governing hostile work environment test for more than twenty years, the Court stated that "[f]or sexual harassment to be actionable, it must be sufficiently severe or pervasive to alter the terms or conditions of [the victim's] employment and create an abusive working environment." Id. at 67 (alteration in original) (citation and internal quotation marks omitted). If the conduct at issue does not reach the Meritor threshold, no Title VII violation has occurred. Burlington Indus., Inc. v. Ellerth, 524 U.S. 742, 752 (1998) (citing Meritor, 477 U.S. at 65).
The Court elaborated on the Meritor test in Harris v. Forklift Sys., Inc., 510 U.S. 17, 21 (1992), holding that sexual harassment is actionable only where the conduct at issue is "severe or pervasive enough to create an objectively hostile or abusive work environment-an environment that a reasonable person would find hostile or abusive." Id. The Court also held that the plaintiff must subjectively perceive the environment to be hostile, for otherwise the conditions of the plaintiff's employment have not actually been altered, regardless of whether a reasonable person would believe them to be. Harris, 510 U.S. at 21-22. Thus, the state of the law following Harris can be summarized as follows: "[I]n order to be actionable under the statute, a sexually objectionable environment must be both objectively and subjectively offensive, one that a reasonable person would find hostile or abusive, and one that victim in fact did perceive to be so." Faragher v. City of Boca Raton, 524 U.S. 775, 788 (1998).
In constructing the hostile work environment test, the Court has "take[n] the middle path between making actionable any conduct that is merely offensive and requiring the conduct to cause a tangible psychological injury." Harris, 510 U.S. at 21. The objective component is designed to be "sufficiently demanding to ensure that Title VII does not become a general civility code." Faragher, 524 U.S. at 788 (1998). Thus, courts must "determine whether an environment is sufficiently hostile or abusive by 'looking at all the circumstances,' including the 'frequency of the discriminatory conduct; its severity; whether it is physically threatening or humiliating, or a mere offensive utterance; and whether it unreasonably interferes with an employee's work performance.'" Id. at 787-88 (quoting Harris, 510 U.S. at 23). But the "mere utterance of an . . . epithet which engenders offensive feelings in an employee does not sufficiently affect the conditions of employment to implicate Title VII." Harris, 510 U.S. at 21 (alteration in original) (citation and internal quotation marks omitted). Indeed, the objective aspect of the hostile work environment test is meant to "filter out complaints attacking the ordinary tribulations of the workplace, such as the sporadic use of abusive language, gender-related jokes, and occasional teasing." Faragher, 524 U.S. at 788 (citation and internal quotation marks omitted). "[S]imple teasing, offhand comments, and isolated incidents (unless extremely serious) will not amount to discriminatory changes in the terms and conditions of employment." Id. (citation and internal quotation marks omitted).
The Court's hostile work environment cases demonstrate that it intended to construct a test that would prevent eggshell plaintiffs-those who are offended by comments or actions that would not offend a reasonable person-from recovering under Title VII. The objective inquiry functions as a limitation on suits attacking the "'genuine but innocuous differences in the ways men and women routinely interact with members of the same sex and of the opposite sex.'" Id. (quoting Oncale v. Sundowner Offshore Servs., Inc., 523 U.S. 75, 81 (1998)). But the EEOC's proposed method of proof-and, in the court's opinion, that endorsed by Mitsubishi, et al.-would strip the limitation of its significance in a pattern or practice case. Under the EEOC's theory, if IPA is found to have a policy of tolerating severe or pervasive sexual harassment, then every woman who was exposed to any form of sexual harassment at IPA, however trivial, would be entitled to damages so long as she was subjectively offended by the conduct she was exposed to. In other words, a woman who experienced only an untoward glance or an isolated remark could nonetheless recover compensatory and punitive damages so long as she was subjectively offended. This scenario is inconsistent with the principles announced in Meritor and Harris, and the court does not believe that the Supreme Court would sanction such a result, even in the context of a pattern or practice case.
This is not to say, however, that the terms or conditions of employment are not altered when an employer has a policy of allowing severe or pervasive harassment to occur in its workplace. A policy of tolerating such behavior is itself a Title VII violation, but proof of such a policy establishes only that the employer behaved improperly with respect to a protected group in general, and therefore justifies only injunctive relief under Teamsters. That an employer has such a general policy does not mean that every employee who was harassed has been exposed to harassment meeting the Meritor threshold, and thus does not mean that the terms and conditions of employment have been altered for each employee who experienced harassment. It is entirely possible for an employee to work at a business with a policy of tolerating actionable sexual harassment, yet be exposed to harassment in only the most superficial of ways.
Of course, the extent to which a particular employee was exposed to the conduct relied on by the EEOC to prove the company's hostile work environment (at the pattern or practice phase) is relevant to that employee's individual hostile environment claim. But this is only to say, as Harris instructed, that the totality of each employee's circumstances, not isolated instances, determine whether sexual harassment is actionable. The court therefore agrees with Mitsubishi that the work environment as a whole may be relevant to an individual claimant's experience, see 990 F. Supp. at 1081, but disagrees that an individual claimant can satisfy her burden on the objective element by aggregating all claimants' experiences. Instead, the court will focus on the conduct to which each individual claimant was exposed. This approach is more consistent with the Court's admonition in Meritor and Harris that a Title VII violation occurs only when the harassment to which a particular claimant is exposed is so severe or pervasive that a reasonable person would be offended.
The court therefore agrees with IPA that the claimants cannot be relieved of their burden to make an objective showing at the individual phase of this lawsuit if the EEOC prevails on the pattern or practice aspect of this case in Phase I. An objective showing will be required at both Phase I and Phase II. At Phase I, the EEOC will be required to prove that a reasonable person would find the sexual harassment occurring globally at IPA so severe or pervasive that the work environment as a whole was hostile or abusive toward women. At Phase II, the EEOC must also show that the harassment to which each individual claimant was exposed is severe or pervasive enough to be actionable.
The court also rejects the burden-shifting framework adopted by Mitsubishi and Dial and advocated by the EEOC here. As discussed above, both Mitsubishi and Dial ruled that the burden of production on all elements of the individual plaintiffs' claims would be shifted to the defendant at the individual stage based on the pattern or practice finding (although Dial refused to shift the burden on the subjective element, as in Mitsubishi). However, this approach is inconsistent with Teamsters. As the court noted in its prior opinion, Teamsters was a pattern or practice case in which it was alleged that the employer had a policy of discriminating against minorities when making hiring and promotion decisions. Int'l Profit Assocs., Inc., 2007 WL 844555, at *13 (citing Teamsters, 431 U.S. at 328-29). The court's decision to shift the burden of production to the employer once the case reached the individual stage was based on two considerations:
(1) if the employer had a policy of making employment decisions on the basis of discriminatory criteria (which had already been proven at the pattern or practice stage), it became more likely that specific decisions made during the time the policy was in place were also based on discriminatory criteria; and (2) if not so motivated, the employer was more likely to have access to evidence demonstrating why not. Teamsters, 431 U.S. at 359 n.45. Thus, the Court held that the primary inquiry at the individual stage would be whether the individual minority claimants applied for jobs or promotions while the policy was in place. Id. at 361-62. Unless the employer came forward with evidence demonstrating otherwise, it could be presumed, based on the pattern or practice finding, that the claimants had been injured by the discriminatory policy. Id.
The Teamsters burden-shifting framework is logical in the context of a discrimination-in-hiring case, where the primary question at the individual stage is whether the claimants were subject to the discriminatory policy. Id.; see also Lemon, 216 F.3d at 581 (stating, in class action suit alleging racial discrimination, that a pattern or practice finding justifies injunctive relief, but if individual claimants want to recover damages they must also demonstrate that they were subject to the employer's discriminatory policy). But the Teamsters framework cannot be applied in a straightforward fashion to a hostile work environment case. As discussed above, proof that the employer had a policy of tolerating unlawful sexual harassment does not necessarily make it more likely that any particular claimant was subject to actionable sexual harassment. Rather, a particular claimant could have experienced only mundane comments, even though the employer had a practice of tolerating far more objectionable actions. Thus, to remain consistent with the goals of Title VII, as articulated by the Supreme Court, each claimant must demonstrate that the harassment she experienced meets the objective threshold of Meritor and Harris. Nor would the employer be in a better position than the employee to demonstrate that the harassment experienced by a given plaintiff was not sufficiently severe or pervasive. To the contrary, each individual claimant should have knowledge of the specific harassment she experienced, and the relevant inquiry is whether this conduct, taken as a whole, would render the work environment hostile to a reasonable person. Thus, shifting the burden on the objective element to the employer at the individual phase does not make sense under Teamsters.
The same follows with respect to the subjective element. The subjective inquiry is an inherently personal one, and, as the Mitsubishi court said, it is a relatively minimal burden in the first place. 990 F. Supp. at 1078. Thus, each individual claimant must establish that she found the sexual harassment she experienced to be hostile or abusive. Each claimant should also be required to bear the initial burden of showing that the harassment she experienced was based on sex (which is also not an onerous burden). Of course, the employer is entitled to then offer a legitimate non-discriminatory reason for the harassment, as in any other Title VII case.
It is appropriate, however, to shift the burden on the negligence element of the individual plaintiffs' claims to IPA once this case reaches Phase II. As noted above, the applicable standard for pattern or practice liability in this case is whether IPA knew or should have known that severe or pervasive harassment was occurring at its offices in Buffalo Grove but failed to take adequate steps to address the problem on a company-wide basis. The standard for employer liability in an individual hostile work environment case is typically the same: the plaintiff must prove that the employer was on notice of the conduct but did not take appropriate action to stop it. See Adusumilli v. City of Chicago, 164 F.3d 353, 361 (7th Cir. 1998) (setting forth standard for employer liability when harassment is perpetrated by plaintiff's co-worker). A showing that IPA was negligent with respect to sexual harassment in general justifies an inference that it was negligent in specific instances. As the Seventh Circuit has stated, "If the harassment is pervasive, it can be presumed, subject . . . to rebuttal, to have come to the attention of someone authorized to do something about it." Young v. Bayer Corp., 123 F.3d 672, 674 (7th Cir. 1997).
Pervasive harassment is precisely what the EEOC will try to establish at Phase I. If the EEOC is successful, it will have proved that IPA was on notice that systemic sexual harassment was occurring in its offices but did not take proper measures to prevent it. Consistent with Teamsters, such a finding makes it more likely that IPA knew or should have known of individual instances of harassment; an employer cannot escape liability by remaining willfully ignorant of individual instances of harassment when it knows that harassment is occurring regularly. See, e.g., Wilson v. Chrysler Corp., 172 F.3d 500, 509 (7th Cir. 1999) (when work environment is permeated with pervasive sexual harassment, employer may be presumed to have notice of specific instances). A finding that an employer has a policy of tolerating sexual harassment in general also makes it more likely that the employer tolerated specific instances of harassment. And if IPA was not negligent with respect to a particular claimant-for example, because IPA had no reason to know about a single discrete incident perpetrated by a claimant's co-worker that went unreported by the claimant*fn11 or because IPA took extreme steps to rectify a particular instance of harassment, despite its general policy of tolerance-then IPA is in the best position to prove it.*fn12 Thus, a shift in the burden of production on the negligence element at Phase II is consistent with Teamsters. See 431 U.S. at 359 ("Presumptions . . . are often created to reflect judicial evaluations of probabilities and to conform with a party's superior access to the proof.").
While the court rejects the EEOC's proposed method of proof, the court does agree with the EEOC that bifurcation is necessary in this case. Not only did the Supreme Court endorse a bifurcated approach to pattern or practice cases in Teamsters, 431 U.S. at 361, but bifurcation is proper here because the court will require a jury finding on the pattern or practice aspect of the case before the jury considers the individual claims so that the court can determine whether to shift the burden of production on the negligence element of the individual claimants' cases to IPA. Such an approach will not only be more efficient, but will work no prejudice on IPA because IPA will have a full opportunity to rebut not only the EEOC's pattern or practice case but also each element of the individual class members' claims. See Fed. R. Civ. P. 42(b). Clear jury instructions will prevent IPA from being unfairly prejudiced by the cumulative effect of the Phase I evidence at Phase II.
Unlike Mitsubishi, however, the court will use a single jury at trial. Although the court does not accept IPA's contention that separate juries for Phase I and Phase II would run afoul of the Seventh Amendment, see, e.g., Houseman v. United States Aviation Underwriters, 171 F.3d 1117, 1126 (7th Cir. 1999) ("While both juries can examine overlapping evidence, they may not decide factual issues that are common to both trials and essential to the outcome."), using one jury is a more effective method of adjudicating this case. A single jury will be more capable of considering the full implications of IPA's policy of tolerating sexual harassment, if proven, and will obviate the need for lengthy preliminary instructions before the individual claims are considered.
The court does not accept the EEOC's contention that using a single jury will be unmanageable given that it intends to introduce the testimony of approximately 40 class members during the pattern or practice stage and testimony from all 113 individual claimants at the individual stage. According to the EEOC, a single trial could take as long as three months. In the court's opinion, the trial of this case will be onerous and time-consuming regardless of the method used. But the EEOC's concerns may be partially misplaced. If IPA is correct that many of the 113 claimants do not have viable individual claims, substantially less time will be needed to try Phase II. As IPA will likely be renewing its motions for summary judgment in light of the framework set forth here, it will be far more clear how much time will be needed for trial after the court rules on IPA's motions. Of course, this case could proceed to trial much more quickly if, as IPA suggests, the EEOC simply withdrew its individual damage claims with respect to the claimants who cannot meet the standard set forth in this opinion.
In summary, the court rejects the EEOC's proposed method of proof. This case will be governed by the following legal framework, which is subject to modification if necessary as trial approaches but which will serve as a general method of proof for purposes of addressing dispositive motions and preparing for trial. At Phase I, the EEOC must establish by a preponderance of the evidence that the sexual harassment that occurred at IPA during the relevant time period, taken as a whole, was so severe or pervasive that a reasonable woman would find the work environment at IPA to be hostile or abusive. Phase I will focus on only the objective aspect of Meritor and Harris, as the court agrees with Mitsubishi that the claimants' subjective perception of the conduct to which they were exposed is not relevant to the determination of whether IPA fostered a discriminatory policy toward women in general. 990 F. Supp. at 1076. At Phase I, the EEOC must also demonstrate that IPA knew or should have known that regular or systematic sexual harassment was occurring in its offices but did not take adequate steps to address the problem. A finding in the EEOC's favor at Phase I will allow the court to award prospective relief under Teamsters, and the court will then proceed to Phase II.
At Phase II, the EEOC will be required to prove, by a preponderance of the evidence, that each individual claimant who seeks monetary damages experienced sex-based harassment that an objectively reasonable woman would find severe or pervasive enough to constitute a hostile work environment. The EEOC must also demonstrate that each claimant subjectively perceived the harassment she experienced to be hostile or abusive. The burden of production on the negligence element of the individual class members' claims, however, will be shifted to IPA if the jury returns a verdict in the EEOC's favor at the pattern or practice phase. If IPA comes forward with evidence demonstrating that it was not negligent with respect to a particular class member, the burden will shift back to the EEOC to demonstrate that the steps IPA took were inadequate. As discussed above, if the harassment any individual claimant experienced was perpetrated by a supervisor rather than a co-worker, IPA will bear the burden of establishing an affirmative defense to its liability for the supervisor's harassment, if applicable; if IPA cannot do so, it will be held strictly liable for such harassment so long as the EEOC meets its burden on the other elements. The EEOC must also make individual showings with respect to the amount of compensatory and punitive damages to which each claimant is entitled. If punitive damages are awarded, the court will closely examine each award to ensure that the parameters of Gore and State Farm are met.
In conclusion, the EEOC's renewed motion for bifurcation is granted in part and denied in part. The trial of this case will be bifurcated into two phases, but will be tried by one jury. The court rejects the EEOC's proposed method of proof and holds that the EEOC must establish its case according to the legal framework set forth herein.
JOAN B. GOTTSCHALL United States District Judge