Implicit Bias: Does It Have Any Relationship to Biased Behavior?

In recent years, social scientists have demonstrated that all individuals likely harbor implicit, or unconscious, biases. Additionally, based on empirical research, some scholars contend that laws or policies that disparately impact marginalized groups result, at least in part, from implicit biases. Other studies suggest that certain behaviors, such as statements reflecting subtle prejudice against marginalized groups (e.g., microaggressions) result from implicit biases. As a result, many organizations in the public and private sector have instituted training programs that focus on implicit bias, its allegedly deleterious effects, and the methods by which to alleviate such bias in, for example, the hiring and promotion of employees or admission of applicants to universities throughout the United States. And researchers at Harvard University have developed the Implicit Association Test (IAT), which purportedly measures the degree to which an individual harbors implicit biases in a variety of contexts, including those affecting traditionally marginalized groups.

Certainly, striving to eradicate biases that produce discriminatory or disparate impacts on individuals or groups is a moral and legal imperative; discrimination in any form is intolerable and contravenes the guarantee that citizens of all backgrounds enjoy liberty, equality, and due process of law.

But does implicit bias actually – and directly – correlate with biased behavior?

Recent research in the social sciences suggests that the answer to this question remains elusive and that the effect of implicit bias on biased behavior may not be as significant as previously believed.

To begin with, there is a general consensus among scholars that implicit bias exists. Put simply, all individuals, regardless of background, arguably harbor implicit biases or prejudices. Importantly, however, the distinction between implicit and explicit bias is difficult to ascertain and operationalize. In other words, how can researchers claim with any degree of confidence that discriminatory behaviors or policies that, for example, disparately impact marginalized groups are the product of implicit rather than explicit bias? Currently, there exists no reliable and objective criteria to make this distinction.

Furthermore, if, as some researchers contend, implicit bias resides outside of consciousness, it would seem difficult, if not impossible, to remedy the effects of such bias. After all, if we cannot be aware of these biases, how can we regulate their manifestation in particular contexts? Also, how can researchers reliably claim that implicit bias predicts biased behavior if not a single person, including researchers, can be aware of its presence and influence? This is not to say, of course, that individuals are unable to develop an increased awareness of the explicit biases that they harbor and take steps to minimize the effect of such biases on their behaviors. It is to say, though, that the relationship between implicit bias and biased behavior remains uncertain, and that there is no method by which to quantify the effect of implicit bias on biased behavior given the presence of other relevant factors (e.g., explicit bias).

Moreover, recent research suggests that the correlation between implicit bias and biased behavior is dubious:

Researchers from the University of Wisconsin at Madison, Harvard, and the University of Virginia examined 499 studies over 20 years involving 80,859 participants that used the IAT and other, similar measures. They discovered two things: One is that the correlation between implicit bias and discriminatory behavior appears weaker than previously thought. They also conclude that there is very little evidence that changes in implicit bias have anything to do with changes in a person’s behavior.[1]

These findings, the researchers state, “produce a challenge for this area of research.”[2]

Additionally, the IAT, which is a popular assessment of implicit bias, has faced significant criticism concerning its methodology and practical value. For example, the IAT sets arbitrary cutoff scores to determine whether an individual’s responses reveal implicit biases, yet fails to provide any assessments of the differences, if any, between the many individuals who score above or below those cutoffs.[3] Also, scores on the IAT are arguably context-dependent and thus produce different results for individuals who take the test multiple times.[4] Consequently, although results on the IAT are “not as malleable as mood,” they are “not as reliable as a personality trait.”[5] Likewise, it is difficult to assess whether the IAT is measuring unconscious attitudes of mere associations that result from environmental influences.[6]

In fact, researchers have conceded that the IAT is flawed, stating that, although the IAT “can predict things in the aggregate … it cannot predict behavior at the level of an individual.”[7] In fact, one of the IAT’s creators acknowledged that the IAT is only effective “for predicting individual behavior in the aggregate, and the correlations are small.”[8] Perhaps most surprisingly, one researcher explained that “what we don’t know is whether the IAT and measures like the IAT can predict behavior over and above corresponding questionnaires of what we could call explicit measures or explicit attitudes.”[9] As a social psychologist explains:

Almost everything about implicit bias is controversial in scientific circles. It is not clear, for instance, what most implicit bias methods actually measure; their ability to predict discrimination is modest at best; their reliability is low; early claims about their power and immutability have proven unjustified.[10]

Of course, this does not mean that implicit bias bears no relationship to biased behavior. It simply means that more research is necessary to determine whether, and to what extent, implicit bias predicts biased behavior. After all, given that eradicating all forms of discrimination is a moral imperative, researchers and policymakers should ensure that society is using the most effective measures to do so. This includes assessing whether implicit bias is a credible predictor of biased behavior.

[1] Bartlett, T. (2017). Can We Really Measure Implicit Bias? Maybe Not. Retrieved from: https://www.chronicle.com/article/Can-We-Really-Measure-Implicit/238807 (emphasis added).

[2] Id.

[3] Azar, B. (2008). IAT: Fad or Fabulous? American Psychological Association. Retrieved from: https://www.apa.org/monitor/2008/07-08/psychometric.

[4] Id.

[5] Id.

[6] Id.

[7] German Lopez, For Years, This Popular Test Measured Anyone’s Racial Bias. But It Might Not Work After All. (March 7, 2017), available at: https://www.vox.com/identities/2017/3/7/14637626/implicit-association-test-racism; see also Heather MacDonald, The False Science of Implicit Bias, (Oct. 9, 2017), available at: https://www.wsj.com/articles/the-false-science-of-implicit-bias-1507590908.

[8] Id.

[9] Id. (emphasis added).

[10] Lee Jussim, Mandatory Implicit Bias Training Is a Bad Idea (Dec. 2, 2017), available at: https://www.psychologytoday.com/us/blog/rabble-rouser/201712/mandatory-implicit-bias-training-is-bad-idea.