How Congress could reduce job discrimination by promoting anonymous hiring.

• Author: Hausman, David

INTRODUCTION I. THE PROBLEM OF IMPLICIT BIAS II. WHY ANONYMOUS HIRING? III. THE ANONYMOUS HIRING DEFENSE AGAINST DISPARATE TREATMENT CLAIMS A. A Defense That Prevents Discovery B. EEOC Guidelines: Practical Details C. A Fraud Exception D. Incorporating Affirmative Action E. Positive Externalities: Reducing Weight and Attractiveness Discrimination IV. INCENTIVES TO PARTICIPATE V. OBJECTIONS A. Adverse Selection B. Cost C. Substitution Effects CONCLUSION: POLITICAL FEASIBILITY INTRODUCTION

After the Supreme Court's decision in Wal-Mart Stores, Inc. v. Dukes, class action plaintiffs cannot succeed in Title VII disparate treatment claims without evidence of individual intent to discriminate. (1) Unfortunately, most job discrimination is not intentional. Title VII, designed to end the open exclusion of women and minorities from the workplace of the 1960s, is ill-suited to the subtle task of preventing unconscious bias. Yet proposals to modify the law tend to be ambitious and politically infeasible, often suggesting a looser intent requirement or broader rulemaking and injunctive power for the EEOC. (2) This Note offers a more modest proposal. In one area of employment discrimination--hiring--the law could take account of implicit bias without a large legislative or administrative project.

Employers can only hire on the basis of race or sex if they know the race or sex of the applicants they are considering. Title VII should therefore encourage employers to hire anonymously. Obscuring the identity of applicants, by removing names and addresses from applications and eliminating interviews, would make implicit biases irrelevant. Employers may protest that interviews are necessary for good hiring decisions, but a long, often-ignored line of psychology research has demonstrated the opposite: in-person appraisal is usually worse than statistical or actuarial evaluation at predicting performance. Eliminating interviews would help many firms select better employees. (3)

Recognizing that a law mandating anonymous hiring would be both overbroad and politically infeasible, I propose a voluntary scheme in which employers could hire anonymously in exchange for immunity from disparate treatment hiring claims. In order to survive a motion to dismiss, a disparate treatment hiring claim against an employer with an anonymous hiring system would have to allege, under a heightened pleading standard, that the decision was not in fact anonymous. Adopting anonymous hiring would therefore reduce firms' liability risk, lowering their premiums on insurance against employment practices liability. At the same time, the continued threat of disparate impact claims would prevent employers from adopting anonymous hiring procedures that systematically exclude minorities. If many hiring discrimination suits are frivolous, as some claim, (4) an anonymous hiring defense would protect firms from such suits at little cost. And the defense would also protect firms from meritorious hiring discrimination suits--by preventing the discrimination that gives rise to them. Firms should therefore be eager to adopt the defense, regardless of whether anonymous hiring would end unintentional discrimination or merely prevent frivolous litigation. Similarly, workers' rights advocates should support an anonymous hiring defense because it would effectively end unconscious bias in hiring at the firms that adopt it.

In Part I of this Note, I describe the problem of second-generation employment discrimination and briefly outline existing legal remedies for such discrimination, which are more limited than ever after Dukes. In Part II, I explain why anonymous hiring would almost certainly reduce hiring discrimination, and summarize the psychology literature demonstrating that job interviews are unnecessary at best. In Part III, I offer a statutory proposal for an anonymous hiring defense to disparate treatment claims, including a fraud exception, and defend the statute as a novel but incremental measure to counter implicit biases without a new and politically contentious regulatory regime. I also describe how the scheme could incorporate affirmative action to break ties in the hiring process. Part IV argues that the proposed immunity could persuade a significant number of firms to adopt an anonymous hiring policy. Part V addresses several objections to the scheme, and contends that it would be less costly than traditional hiring and that adverse selection effects would increase the incentive to participate. Finally, in the Conclusion, I argue that the scheme is politically feasible--or at least more feasible than the broad structural changes that legal scholars have often advocated.

1. THE PROBLEM OF IMPLICIT BIAS

   Title VII of the Civil Rights Act of 1964, (5) as interpreted by the Supreme Court, allows both disparate treatment and disparate impact claims. (6) A successful disparate treatment claim requires showing intentional discrimination on the basis of race, color, religion, sex, or national origin. (7) Although plaintiffs can successfully bring intentional discrimination claims where an employer acted with "mixed motives"--both a discriminatory and nondiscriminatory purpose--the basic showing of purpose remains necessary. (8)

   Disparate impact claims, by contrast, require no showing of intent, but they are also less likely to be lucrative. Only injunctive relief (including back pay) is available, and courts have generally limited disparate impact claims to the narrow area of standardized employment tests. (9)

   Although disparate treatment claims are more common and more lucrative, most employment discrimination is likely unintentional. A large literature in sociology and psychology documents the influence of unconscious bias in interpersonal decisions. (10) For example, when experimental subjects are randomly placed in groups, they begin to favor members of their own group and become more inclined to remember negative attributes of members of the out-group. (11) Social cognition experiments have also found that subjects evaluate the performance of a person from a minority group more harshly when he or she is the only member of that group present in an interaction. (12) Similarly, subjects more frequently ascribe unusual characteristics to a minority group, even when that group is only identified as Group B. (13) The cumulative effect of in-group bias leads people to adopt inaccurate stereotypes that affect their interaction with others. (14) Such unconscious bias influences hiring, promotion, and dismissal decisions, particularly when employers have discretionary personnel policies. (15)

   Legal responses to the problem of unconscious discrimination have been few and, by many accounts, inadequate. (16) Since the Civil Rights Act of 1991, which increased the damages available for intentional discrimination and introduced jury trials in Title VII actions, (17) Congress has not altered antidiscrimination law to account for the increasingly well-known consequences of unconscious bias. Although plaintiffs' lawyers have often argued--sometimes with success (18)--that evidence of widespread disparities demonstrates discriminatory intent, that argument is now unavailing.

   The Supreme Court made clear in Dukes that plaintiffs can no longer use disparate treatment claims to attack discretionary employment practices that enable unconscious bias, even if management is aware of the problem. In Dukes, the Court reversed the trial court's certification of a class of nearly 1.5 million women who claimed that Wal-Mart's pay and promotion policy intentionally discriminated against women. (19) That policy delegated pay and promotion decisions to local managers, who were generally not required to base their decisions on objective criteria. (20) The plaintiffs supported the claim of intentional discrimination by assembling expert testimony on the discriminatory effects of subjective decisionmaking, statistical evidence of women's disadvantage in pay and promotion, and anecdotal evidence of discrimination. (21) The Court held that the class failed to satisfy the commonality requirement of Federal Rule of Civil Procedure 23(a) because the record, which depended on region-by-region statistical evidence, did not demonstrate that discrimination had occurred in particular stores or had affected all class members in the same manner. (22)

   Although the Court considered only the question of class certification, the opinion acknowledged that "proof of commonality necessarily overlaps with respondents' merits contention that Wal-Mart engages in a pattern or practice of discrimination." (23) Addressing that contention, the Court squarely rejected the legal theory that discretionary employment practices, even if linked to discriminatory outcomes, can be the basis of a disparate treatment claim under Title VII: "Merely showing that Wal-Mart's policy of discretion has produced an overall sex-based disparity does not suffice." (24)

   Title VII, with its intent requirement, was always an unlikely vehicle for attacking unconscious discrimination, and Dukes made clear that what was unlikely is now virtually impossible. David Freeman Engstrom drew out this implication in his commentary on the Dukes decision:

   Damages suits, some have argued, are particularly ill-suited to remedying "second generation" discrimination in which patterns of inequality result from "implicit" or "subconscious" forms of bias and often cannot be traced to particular decision-making nodes. Indeed, the Dukes plaintiffs, by linking local managerial discretion to broader statistical patterns, were part of a wider effort to adapt Title VII to these changing workplace realities. (25) The Supreme Court has now decided that this effort will not succeed.

   Dukes makes the present a particularly good moment to consider other solutions to the problem of unconscious discrimination. There is no shortage of proposed solutions, but most...