Implicit Bias: Moving from Theory to the Courthouse

Jerry Kang, Mark Bennett, Devon Carbado, Pam Casey, Nilanjana Dasgupta, David Faigman, Rachel Godsil, Anthony G. Greenwald, Justin Levinson & Jennifer Mnookin, Implicit Bias in the Courtroom59 UCLA L. Rev. 1124 (2012).

Having a total of ten authors for one article would make this rather exceptional even without regard to the topic. That these authors participated together in a symposium on implicit bias is not a surprise. But what is unusual, if not exactly surprising, is that they together wrote this one article. This is not the typical scenario for the papers delivered at a conference. The ten include legal academics, scientists, researchers, and a sitting federal judge. Six are law professors, though two of them hold joint appointments. One is a research consultant for the National Center for State Courts, two are psychology professors and one is a federal district court judge. They all come to the study of implicit bias from their respective points of perspective but the article is a fully integrated article.

The question the article begins to answer is: What, if anything, should we do about implicit bias in the courtroom? The scientific literature on implicit bias and the role implicit bias might play has been the subject of considerable legal literature. Part I provides a clear, straightforward introduction to the science involved in understanding that implicit bias exists, what it is, and how it works. The article defines implicit attitudes and stereotypes as biases “not consciously accessible through introspection.” Accordingly, their impact on a person’s decision-making and behaviors does not depend on that person’s awareness of possessing these attitudes or stereotypes. Consequently, they can function automatically, including in ways that the person would not endorse as appropriate if he or she did have conscious awareness. If you remember, Ross Perot in the 1992 Presidential campaign gave a vivid example of implicit bias that gets disclosed unintentionally. While speaking at an NAACP meeting,  Perot addressed his audience several times as “you people” or “your people.” Especially since the topic of the speech was crime and illegal drugs, many perceived Perot as expressing a stereotypical view of African Americans as being identified with crime and drugs. He had no incentive to be avowedly racist and there is no indication that he was conscious of the obvious inference listeners drew based on the language he used.

Part I develops a full description of the key research upon which the concept of implicit bias has been based — the Implicit Association Test (IAT). The test shows those who take it pairings of pictures of either whites or African Americans and pictures presenting either negative and positive images. The test measures the time differences that test-takers take to deal with these different combinations of pictures.  Thousands of people have taken the test (if you haven’t, you can find it at The huge amount of data that has been generated has proven to the satisfaction of the researchers that implicit bias is pervasive, large in magnitude, disassociated from explicit biases, and predictive of certain real-world behavior.  A meta-analysis of these studies found that implicit attitudes such as discrimination against African Americans predict certain behaviors better than measures of explicit bias.  In order to defuse emotionally loaded responses, the article uses examples, such as the presence or absence of implicit biases of a vegetarian.

Part II then traces two “trajectories,” with the first treating implicit bias in criminal trials and the second, which this post will deal with, giving “an empirical account of how implicit bias may potentially influence” employment discrimination litigation. First, in dealing with implicit discrimination in discrimination cases, the article digs much more deeply into the studies that take different approaches and show different facets of the existence and extent of implicit bias in the employment setting. Tester studies in which applications that are identical except for race or gender are sent to employers. The results reveal that, in 20 to 40 percent of the cases, the employers treat the subordinated group worse than the privileged one in terms of who get callbacks. Field experiments with actual applicants confirm that bias. Further studies correlate discriminatory employment evaluations with implicit bias. The next step is to show that additional studies have demonstrated that the mechanism by which this discrimination occurs is that evaluators engage in motivated reasoning – changing merit criteria “on the fly” to the advantage of the privileged group applicant but without realizing that the criteria have been changed, much less that they have been changed in order to advantage the privileged person.  In sum, these varied studies demonstrate that implicit motivations influence behavior which is then rationalized after the fact to avoid recognition that bias is at play.

In the next step of Part II the article describes how implicit bias operates at several steps in the litigation of discrimination cases. After Twombly and Iqbal, judges are now able to use their “judicial experience and common sense” to evaluate the “plausibility” of plaintiff’s claim of discrimination in the context of Rule 12(b)(6) motions to dismiss based on the pleadings. Judges, no less than the rest of us, are subject to the influence of implicit bias, which shapes their experience and common sense. Implicit bias has been shown by numerous studies to be particularly powerful when, as in the situation of a judge deciding a motion to dismiss without the benefit of the parties’ discovery, the decisionmaker lacks “sufficient individuating information.”  The lack of information about the actual case increases the risk that implicit bias will affect the decision. Recent statistical studies show that there has been a significant increase in the dismissal rate of discrimination cases at the pleading stage since Iqbal was decided. That is no surprise, but what these studies also show is that there is a significantly higher increase in the rate for discrimination cases compared with other types of civil litigation. Based on experiments replicating jury trials, studies show that juries are, like judges, likely to engage in motivated reasoning as well as “performance preference.” “Performance preference” involves the risk that implicit bias influences how the jurors evaluate the judges, attorneys and witnesses in the trial.

Part III – entitled “Interventions” – describes some strategies that can be used to decrease implicit bias in the courtroom. This is a daunting challenge since, by definition, we are not consciously aware of these biases that nevertheless influence our behavior. The article describes studies that show that increasing diversity, “the direct contact with countertypical people,” decreases the operation of implicit bias. Even increasing the manifestation of diversity in the courts through the kinds of pictures, posters, etc., that are on display can help.  Judges present a special challenge since a survey showed that 97 percent of those surveyed thought they were in the top quarter in avoiding “racial prejudice in decisionmaking.” Just as all the children can’t be above average, all the judges can’t be in the top 20 percent in terms of their objectivity. This sense of certainty that one is not subject to bias is particularly dangerous since studies demonstrate that, when someone “believes himself to be objective, such belief licenses him to act on his biases.” So, judges as a group are likely to be particularly vulnerable to implicit bias. A way to increase motivation to be fair and to avoid letting implicit bias influence decisionmaking is to increase a person’s scientific knowledge about implicit bias.  That can be done through individual study as well as part of the organized judicial education that helps train judges. As part of that education, judges can learn how to engage “in effortful, deliberative processing,” which includes reducing the level of one’s emotional state of mind when making decisions. Judges can also learn to “count,” or to consciously keep track of the number of situations in which implicit bias could play a role and to analyze whether or not implicit bias could have crept into the decisionmaking process.

A point not made in the article is that plaintiffs’ counsel might find it useful to include in their discrimination complaints significant information about implicit bias to help a judge minimize her implicit bias when deciding motions to dismiss under the new Iqbal plausibility test. In fact, I am surprised not to have seen a group that represents discrimination plaintiffs, such as the National Employment Lawyers Association, provide clear, straightforward factual information about the operation of implicit bias to be used to bolster plausibility arguments in motions to dismiss. Perhaps, this article can provide the framework for plaintiffs’ lawyers to educate judges about the plausibility of bias and of discrimination.

The article makes an interesting point by showing that studies do not support giving prospective jurors individualized screening tests such as the IAT. While the test-retest reliability of the IAT as to people generally is very well established, it is less so at the level of the individual test-takers. So, that leaves jury diversity and the education of jurors about implicit bias as the primary tools to try to debias the operation of the jury system.

This article takes us another step down the road toward integrating what science has shown us about implicit bias into law. There are more steps to be taken, including attempts to be able to figure out how IAT tests can be shown to be reliable at the individual level. Even without taking that step, articles such as this one are building a context for a fundamental legal question that has yet to be definitively answered. That question is whether action based on implicit bias is intentional discrimination for purposes of deciding disparate treatment discrimination cases. In the 19th Century, signs on the doors of Boston employers that “Irish need not apply,” were explicitly biased. Since then we have learned that bias influences the behavior of all of us in much less obvious ways. Yet its influence is discrimination.

Four of the authors of this article, along with many other experts in the field, also participated in a conference sponsored by the Charles Hamilton Houston Institute for Race & Justice at Harvard Law School on June 14, just before this article was published. That conference launched a book of edited essays about implicit bias and the law entitled Implicit Racial Bias Across the Law. So, the academic study of the impact of implicit bias on the law continues to grow, expand, and increase in credibility. Knowledge is necessary, but so is action.


1 comment
  1. 1
    On December 29, 2012 at 5:43 pm, Chris Budgell said:

    I found this website very recently and have posted one comment already to the latest WORKLAW article.

    This subject is something of a revelation to me as I’ve spent over ten years pursuing an evolving legal case without counsel, and almost every action has been summarily dismissed using what I now call collectively “gatekeeper” devices.

    I have not yet thoroughly read the UCLA Law Review Article, but I expect to find no evidence that the authors considered that judges, and other members of the legal establishment, have an implicit bias against self-represented persons who appear before our tribunals and in our courts, and especially when they are facing parties represented by counsel.

    Looking at my own now fairly extensive experience I can say that for this bias there is compelling evidence, the most notable example of which for me was a hearing conducted (but apparently not recorded) by three justices of the British Columbia Court of Appeal in 2003:

    I subsequently discovered in another law journal (and I note it is still not readily accessible to the public) the text of a speech given by the author of that judgment to the members of the B.C. Council of Administrative Tribunals in 1999. The title was “Know Thyself: Some Thoughts About Impartiality of Individual Decision-makers From an Interested Observer”. I note she said “Interested”, not “Disinterested”, and in my view the speech amply demonstrates that indeed she was very “interested” – in the sense of biased.