The unintended consequences of the '91 Civil Rights Act: the emphasis on firing-based litigation has had subtle but important effects

Regulation, Summer, 2003 by Paul Oyer, Scott Schaefer

THE EARLY 1990S SAW A REBIRTH OF congressional interest in employment discrimination legislation. After months of national debate, a presidential veto of a 1990 civil rights bill, and the nationally televised Clarence Thomas confirmation hearings, the Civil Rights Act of 1991 was enacted. The Act significantly expanded the rights of plaintiffs in discrimination complaints to the Equal Employment Opportunity Commission (EEOC) and in federal civil court.

Many opponents of the Act warned that the bill would lead to hiring "by quota." Critics focused on its "disparate impact" provisions, whereby the Act made it significantly easier for plaintiffs to use statistical evidence to prove unlawful discrimination even if there was no discriminatory intent on the part of the employer. For example, Roger Clegg, then a deputy assistant attorney general, later wrote that during the negotiations over what became the 1991 Act, the Bush administration was "enormously successful (and accurate) in characterizing the proposed legislation as 'a quota bill' and a 'lawyer's bonanza'." Unsatisfied by the administration's efforts to water down the Act's disparate impact provisions, the Wall Street Journal ran an editorial headlined "It's a Quota Bill" on the day after it was signed into law. Proponents of the legislation argued that such quota fears were overstated and that the Act was necessary in order to open opportunities for women and minorities in fields that had traditionally been unwelcoming.

The two sides of the debate clearly disagreed over how best to achieve the shared aim of racial and gender equality in labor markets. Yet, they appeared to share a set of assumptions regarding how employers would respond to the specific provisions of the Act -- they believed it would cause employers to increase their hiring of minorities and women. Did that belief prove true?

Our research suggests that the quota-based hiring fears were largely unfounded. However, the Act also appears not to have improved employment opportunities in fields with traditionally low representation of women and minorities.

Our research uncovers three main patterns in employment outcomes for protected workers. First, a trend toward greater racial and gender integration in labor markets stopped around 1991. Industries that historically had low minority and female employment shares were hiring more such workers prior to the legislation, but this pattern did not continue into the post-Act period. Second, employers changed their means of dismissing protected workers. Because it is more difficult for workers to win employment discrimination lawsuits after losing their job as part of a mass layoff, it appears employers began to delay dismissal of protected workers until the next round of layoffs. While the overall rate of job loss remained constant for both protected and unprotected workers, the fraction of lost jobs coming as part of a mass layoff increased for protected workers, but not for unprotected, after the passage of the Act. Third, the wage premium on labor market experience (which economists refer to as returns-to-experienc e) increased for some groups of protected workers after the 1991 Act. if employers are concerned about the costs of dismissing potential employees, then increases in those costs will cause firms to reduce their demand for workers who are relatively likely to be dismissed. Because job turnover rates are highest for inexperienced employees, increases in firing costs can increase the returns to experience.

Our primary conclusion is this: While the 1991 Civil Rights Act did not have large beneficial or adverse effects on employment or average wages for protected workers, employers did change their behavior in subtle ways after its passage. The changes in behavior are broadly consistent with the assertion that the Act's main effect was to increase the costs of dismissing protected workers.

THE CIVIL RIGHTS ACT OF 1991

The 1991 Act, which took effect on November 21, 1991, strengthened several prior pieces of employment discrimination legislation, including the Civil Rights Act of 1866, the Civil Rights Act of 1964 (Title VII), the Age Discrimination in Employment Act, and the Americans with Disabilities Act. The broad group of individuals gaining additional protections from the 1991 Act includes racial minorities, women, and those with disabilities.

While many of the law's provisions had the potential to affect employment relationships, we focus our discussion on two main points: changes in the ability of plaintiffs to use statistical evidence to prove discrimination, and increases in damage awards available to plaintiffs who prove discriminatory intent on the part of an employer.

Discrimination and statistics A series of i 970s Supreme Court rulings (most notably the 1971 Griggs v. Duke Power) had allowed plaintiffs to show unlawful discrimination using statistical evidence. Plaintiffs could proceed by demonstrating that an employer's practices led to a disparate impact on protected groups even if there was no discriminatory intent on the part of the employer. A 1989 Court decision, Wards Cove Packing Co. v. Atonio, made it significantly more difficult for plaintiffs to prove disparate impact. After Wards Cove, plaintiffs could not show disparate impact using statistical evidence alone--plaintiffs were required instead to demonstrate that a particular employment practice led to the differing effect on protected groups. The 1991 Civil Rights Act relaxed this standard somewhat by allowing the use of statistical evidence without identification of a particular employment practice, provided the plaintiff can show the employer's decision-making process cannot be separated into specific prac tices.