WHILE writing the article to which Professors Mitchell and Bielby have published responses,1 I was mindful of the many ways in which the article could be misinterpreted. In taking issue with the assumption that legal controls work in a direct, linear manner to deter discrimination, I thought I might be misunderstood to say that people are not responsive to incentives. In worrying about how legal sanctions exert external pressure that may crowd out the inclination of well-intentioned people to self-monitor for bias, I feared that the article would be read mistakenly to oppose strong and appropriate legal rules against discrimination. In arguing that we should take people's good intentions not to discriminate as a useful starting point for better workplace policies, rather than as the cynical exhibition of people's self-delusion, I anticipated that the article would be dismissed as a fanciful and naïve denial of the existence of race and gender bias. In arguing that well-intentioned people can overcome their natural tendencies to discriminate, I was concerned about appearing to claim that good intentions are sufficient to end discrimination.
In the case of the responses by Professor Mitchell and Professor Bielby, these fears were unwarranted. The responses engage the article's actual objectives—to clarify the state of empirical knowledge about unconscious workplace bias and to evaluate proposed approaches to reducing it. Professors Mitchell and Bielby, both leading figures in research about workplace bias, bring their considerable expertise to bear on evaluating and expanding the themes of the article. They each agree that legal scholars have ignored some of the relevant psychological literature and interpreted the research on which they have drawn in a shallow way. They also correctly identify my concerns about the overuse of legal coercion to reduce workplace bias, and concur that more legal controls are not likely be effective in reducing implicit bias.
Beyond these areas of agreement, there are differences in emphasis and enthusiasm. Professor Mitchell warns against too much reliance on research findings that are based on the responses of people (typically college students) to hypothetical situations. Professor Mitchell is well-known for his cautions against the application of experimental research findings to real-world settings. Some of his work criticizes social psychologists and legal scholars for exaggerating the prevalence of implicit bias.2 He does not direct this criticism against me here, perhaps because I make no claim about the frequency with which implicit bias occurs and nothing in my article turns on its exact magnitude.3 He does draw further attention to the need for better field research and offers one helpful explanation for why it is so hard to get: managers in a position to adopt testable workplace practices are reluctant to adopt these practices in the absence of good evidence that they might actually work.4 No one could dispute the need for further applied research. Fortunately, as Professor Mitchell notes, some scholars are busy at it.5
While noting the dangers of overreading the laboratory research on which the article cautiously relies, Professor Mitchell proposes a "scal[ing] up" of the article's thesis from company managers to the company itself.6 At the firm level, he observes, the legal system sometimes discourages companies from adopting workplace practices that might ameliorate bias. His example is a well-intentioned firm that undertakes in-house audits and organizational culture reviews to determine which of its workplace structures might obscure or enable workplace bias, only to have these materials used against the company in subsequent adversarial litigation.7 Whether and to what extent a firm would be penalized by such audits and reviews depends in part, even under existing law, on the way the company responds to them.8 The example, though, is a useful one, in showing that companies, as well as individuals, can have good intentions that might be thwarted by policies that are too eager to find fault and impose liability.9
Professor Bielby's principal response to the article is that it focuses on the cognitive aspects of discrimination rather than the structural forces that support it. While Professor Bielby agrees that more law to combat implicit bias is a "bad idea,"10 he believes that the case I have made for more attention to the motivational consequences of workplace diversity measures is wrongly grounded in the psychological processes of implicit bias rather than in the "organizational policies, practices, and structures that create and sustain unlawful racial and gender inequality at work."11
Professor Bielby is correct that the focus of the article is on the psychology of discrimination. The article's purpose is to respond to legal scholars who assume that bias is a pervasive psychological phenomenon to which the solution is stronger legal controls. The article provides a fuller account of the psychology of bias by differentiating between the motivations that might lead people to avoid it. It demonstrates that measures that motivate people internally are more effective in reducing bias than external constraints, and that the law, insofar as it functions as an external constraint that may crowd out people's internal motivation to be unbiased, may actually exacerbate bias rather than reduce it. The article does not ignore the role of workplace structures. In fact, it endorses various structural measures that work to strengthen people's internal motivation to act in unbiased ways, including collaborative structures that facilitate positive intergroup interactions, job rotation systems, upper management role models who demonstrate a genuine commitment to diversity, and accountability systems that further participation and a sense of ownership by those involved.
Professor Bielby's criticism is that, in making the case for these and other measures, I remain too focused on what goes on in people's heads, rather than on what they do. Susan Sturm, in conversations with me, has made a similar observation. I take the point—to a point. Our shared objective is the reduction of bias, not moral salvation; people's good intentions are valuable as a means of reducing bias, but is not itself the goal. The same thing can be said, however, about workplace structures; they are a means to an end, not the end itself. Both good intentions and workplace structures that promote unbiased decision-making are important, independently and in interaction with one another. As the research in the article suggests, attitudes and beliefs affect the impact that various workplace structures might have, and vice versa. I accept Professor Bielby's point that discrimination operates in institutions. I hope he will accept mine that discrimination operates in people's minds as well. Both matter.
Professor Bielby notes the vagueness of the line I draw between organizational strategies that people find coercive and threatening and those that promote people's intrinsic commitment to nondiscrimination.12 I concede that ascertaining whether a particular policy will promote or discourage the internalization of desired nondiscrimination norms requires judgment and finesse, to which knowledge of the basic principles of human motivation and the dynamics within a particular work force are both highly relevant. Professor Bielby notes that management plays a role in constructing the message as positive or negative.13 Indeed. That is the point. Good leadership knows its people and figures out how best to communicate with them, aware that the internal constraints it stimulates can have an even more favorable effect in how they behave than the external constraints it imposes.
What happens when the organizations do not get it, or fail to transmit positive, nondiscriminatory values throughout the workforce? Professor Bielby notes that it is "risky" to rely on "nondiscrimination efforts that require internally motivated well-intentioned acts in order to be effective."14 I could not agree more, which is why Title VII remains an essential piece of an integrated nondiscrimination policy. Title VII is the legal backstop—a necessary tool, especially for workplaces that are not well-intentioned and well-run, and do not have the institutional commitment or structures to prevent discrimination. Title VII does a reasonably good job of defining what the law can define. But Title VII, no matter how well it is drafted, cannot do the job itself. No law can identify or apprehend all instances of discrimination, which is why it is so important to engage people's positive attitudes as well as their fear of legal sanctions.
Mitchell and Bielby are right to emphasize, in their different ways, the importance of workplace structures and institutions. Positive norms and good intentions alone will not do the trick. In my defense, however, the article does not claim that they will. To the extent that Professors Mitchell and Bielby understand the article to say that good intentions are enough,15 they misread the piece. The choice is not between internal norms or external structures. It is, rather, between continuing to focus on laws and workplace structures as the only meaningful constraints on discrimination, or figuring out how to trigger the internal constraints that might bear some of the load.
Preferred citation: Katharine T. Bartlett, Good Intentions Matter, 96 Va. L. Rev. In Brief 35 (2010), http://www.virginialawreview.org/inbrief/2010/06/15/bartlett.pdf.
[Back to this issue]