Warm Reasoning and Legal Proof of Discrimination
Knowles and Ditto's illuminating analysis of the selective use of principle in decision making is most immediately relevant to employment discrimination law. Their description of how individuals often act inconsistently by weighing multiple criteria in a way that tilts a decision toward a desired outcome has direct application to thousands of routine employment decisions in which supervisors and committees are asked to base their selection of the “best” candidate on a number of merit-related factors with no preassigned weight. Additionally, Knowles and Ditto's treatment of studies demonstrating the willingness of subjects to make racially biased decisions only if they can rationalize their judgments in terms of racially neutral principles well captures the discourse in the contemporary workplace where supervisors and managers have internalized the dictates of antidiscrimination laws.
Most pointedly, the mock hiring experiment cited by the authors, involving a stereotypical male job in the construction industry in which evaluators ranked experience as more important than education when the male candidate had more experience than the female candidate, but reversed the ranking when the applicants were of unknown gender, is the stuff that “disparate treatment” Title VII claims are made of. In employment discrimination law, this lack of consistency—whether uncovered by more favorable treatment of “comparators” or through the results of a regression analysis analyzing a pattern of decisions—is an important marker of discrimination. Indeed, both psychology and law are preoccupied with determining the source of inconsistent reasoning, although the questions asked by the two disciplines diverge quite sharply at this point.
In their intellectual history, Knowles and Ditto tell us that in the past, the focal point of disagreement among social psychologists was over whether judgments were rooted in motivations and emotions (“hot” processes) or in cognitive, rational operations (“cold” processes). As is the fate of most dichotomies, they report that the latest thinking in the field now generally endorses (p.381) a “warm” view of human reasoning that concedes the importance of both processes and is more interested in identifying how motivation and cognition interact—for example, in explaining how a person's attachment to a social group or a particular ideology can skew his weighting of factors in a particular case without at the same time destroying his “logical sensibilities.” In these psychological accounts of “warm” reasoning, it is axiomatic that individuals are seldom fully aware of the factors influencing their judgments and often mistakenly believe they are objective and unbiased. “Motivated reasoning” in this lexicon is a far cry from deliberate or conscious discrimination, and it signals only that individual preferences and emotional commitments can exert a subtle influence on decision making.
In employment discrimination cases, the debate has not centered on motivational versus cognitive theories of discrimination. Instead, the crucial legal controversies have revolved around the meaning of “intent” and “causation.” For antidiscrimination lawyers, discriminatory “intent” is generally regarded as synonymous with discriminatory “motivation,” and is most often assimilated to conscious or deliberate discrimination. The legal definition of causation, in contrast, has not been tied to conscious states of mind; it is a more amorphous concept that encompasses any factor or stimulus that prompts a person to act. With respect to causation, lawyers tend to worry about how much influence a given factor exerts, often trying to discern whether it is a sole cause, a determining factor, or merely one of several motivating factors.
Since the late 1980s, there has been a steady movement to import a more psychologically informed definition of “intent” and “motivation” into Title VII law and to shift the focus in litigation away from the decision maker's conscious state of mind and toward an inquiry into causation. The most contested evidentiary question remains whether a showing of inconsistent or disparate treatment is sufficient to produce liability, or whether plaintiffs should also be required to prove that the decision maker was aware that her decision was discriminatory at the time it was made. The voluminous academic literature urging legal recognition of implicit or unconscious bias has made some headway in the case law but still represents the minority position. (Chamallas, 2001; Krieger, 1995; McGinley, 2000).
The two contending positions are most evident in what are known as “mixed motivation” cases, a very common type of case in which both “legitimate” and “illegitimate” factors are said to enter into an employment decision. Although Congress made it clear in the 1991 Civil Rights Act that plaintiffs were entitled to recover if they proved that race, sex, or another prohibited factor was a “motivating factor” in the employer's decision, the U.S. Supreme Court has balked at this plaintiff-friendly standard, as evidenced by its recent (p.382) refusal to extend the mixed-motivation framework to age discrimination cases (Gross v. FBL Financial Serv., Inc., 2009). Most important, no settled definition of “motivating factor” has yet emerged from the courts. Left unanswered is whether that key phrase will be limited to factors that the actor consciously realizes has affected or motivated his decision or, as interdisciplinary scholars Linda Krieger and Susan Fiske have urged, will be interpreted more broadly to encompass any “internal mental state … that includes cognitive structures like implicit stereotypes or other social schema that influence social perception, judgment and action” (Krieger & Fiske, 2006).
In theory, there is no necessary link between the debate over deliberate versus unconscious discrimination and the mixed-motivation debate. It is possible, for example, for a decision maker to be consciously aware of multiple reasons for her actions or to be totally unaware of the many reasons that prompted her decision. In other words, the number of factors entering into a decision bears no logical connection to whether those reasons are acknowledged by the decision maker. As they align in practice, however, the two debates are closely intertwined. In mixed-motivation cases, the legitimate reason is almost always asserted to have been consciously in the mind of the decision maker at the time he acted, while the illegitimate reason is more likely to be cast as an unconscious force. It is at this juncture that psychological research such as Knowles and Ditto's has the most to offer developing law.
Not surprisingly, liberals tend to favor both legal recognition of unconscious discrimination and the acceptance of mixed-motivation models that allow for recovery whenever discrimination taints the decision-making process, even if there is no proof that bias was the sole or predominant factor contributing to a decision. Conversely, conservatives seek to limit liability to cases of deliberate discrimination and to argue that plaintiffs should be required to prove that discrimination was the “but for” cause or the determinative influence on the outcome. As I see it, these are not just politically compatible stances; they also reflect each side's competing, if largely intuitive, views about human decision making. For liberals, the existence of a conscious legitimate reason does not substantially reduce the likelihood that bias has entered into a particular decision. Liberals believe that there is still a good possibility that bias has also played a role, even if the decision maker was not aware of its influence. In contrast, conservatives tend to assume that the existence of a legitimate reason means that there was no bias at all or that bias played such an insubstantial role in the outcome that it was not even apparent to the decision maker.
Knowles and Ditto's contribution in this volume provides considerable ammunition for the liberal position and puts pressure on disparate treatment law to move away from reliance on conscious state-of-mind determinations (p.383) and toward adoption of a psychologically informed definition of causation. The article also reinforces the notion that cases of “mixed motivation” are ubiquitous and that it is futile to attempt to calibrate the degree of influence that bias exerts in any particular decision.
As Knowles and Ditto portray the decision-making process, principle and preference work together to guide and constrain decisions. According to them, when persons make discretionary judgments involving a weighting of several factors, it is reasonable to suspect that the results will often be skewed, not because decision makers are engaged in conscious cover-ups, but because human beings are so adept at selectively marshaling principles to support a desired result that the process has become almost second nature to them. If this is true, the only way to ferret out inconsistency and bias is to look at a larger pattern of decisions or at very similar “comparator” cases to see whether the variables are given the same weight. In other words, proof of disparate treatment is still the best method to uncover bias. To require additional proof of a deliberately discriminatory mindset is to limit protection of the law to a very small slice of human behavior, exempting the more common forms of bias with the most direct application to employment decisions.
Moreover, Knowles and Ditto also stress that in the post civil-rights era, most individuals will try to check and mask their discriminatory impulses and will not be likely to exhibit bias in extreme cases where there is wide disparity in qualifications between candidates. To this degree, principle exerts a constraining influence on judgments, making bias most likely to surface in those debatable, middling cases where candidates' qualifications are roughly comparable. The strong desire to mask discrimination also generates the production of legitimate and legitimating reasons for the decision such that we can expect that most cases will look like mixed-motivation cases whenever it is plausible that either bias or principle could have produced the decision. In such cases, presumably Knowles and Ditto would not attempt to assess the relative weights of the “legitimate” and “illegitimate” inputs, precisely because of interactive nature of the two forces. Their research suggests that the law would be better directed toward answering the normative question of whether the disparate treatment was justified under the circumstances, rather than trying to determine whether bias constituted a large enough factor to warrant relief.
Finally, it bears mentioning that psychological scholarship such as Knowles and Ditto's does not solve the dilemma of whether to apply any reformulated standard of causation across the board to cases involving bias in favor of women and minorities, as well as to cases involving bias against traditionally excluded groups. Their research indicates that the disposition to mask (p.384) preference can be as robust in situations favoring minorities as it is in antiminority cases, as illustrated by the mock college admissions experiment in which evaluators selectively inflated the factor that favored black applicants and engaged in a subrosa form of affirmative action. Rather than endorsing a broad psychologically informed definition of causation in such cases, however, I suspect that liberals, like me, will be inclined to shift ground and to argue for adoption of an asymmetric, antisubordination approach to antidiscrimination law that enlists the power of law only to dismantle entrenched social hierarchies. In the end, Knowles and Ditto forced me to confront the uncomfortable question of whether it is unethical to wrap up preference in a principle when the underlying preference is not so bad.
Chamallas, M. (2001). Deepening the legal understanding of bias: On devaluation and biased prototypes. Southern California Law Review, 74, 747–806.
Gross v. FBL Financial Services, Inc., 129 S. Ct. 2343 (2009).
Krieger, L. (1995). The content of our categories: A cognitive bias approach to discrimination and equal employment opportunity. Stanford Law Review, 47, 1161–248.
Krieger, L., & Fiske, S. (2006). Behavioral realism in employment discrimination law: Implicit bias and disparate treatment. California Law Review, 94, 997–1062.
McGinley, A. (2000). !Viva la evoluction! Recognizing unconscious motive in Title VII. Cornell Journal of Law & Public Policy, 9, 415–92.