Bias and Diversity Training for Police Works—Temporarily

iStock image by Antonio Diaz licensed to Art Markman

All people have biases, and most of the time that is a good thing. When faced with a choice for a dessert to eat, I am biased to choose peanut butter pretzels over Jell-O, and that bias has consistently led to good dessert experiences. This works well for me, because (in my view), just about any peanut butter pretzel is better than any Jell-O. So, the category something belongs to (peanut butter pretzel versus Jell-O) is a nearly perfect predictor of my preference.

But, bias plays a negative role in many situations when these biases involve categories of people. Favoring one group of people over another based on gender, race, or ethnicity, for example, creates problems because there are many cases in which we believe there is a relationship between these categories and some behavior that is not actually there. Even when there is a real difference on average between groups, the amount of variation inside of each group makes it difficult to predict much about any individual.

One domain in which bias based on categories of people creates problems is policing. For example, there is ample evidence that police in the United States are much more likely to arrest Black people than White people. This can lead to a bias that Black people are more dangerous than White people, which can then increase the likelihood that an armed police officer may injure or kill a Black person relative to a White person.

In recognition of these statistics, police departments have engaged in bias training that aims to reduce these biases. The focus of this training is on helping officers to recognize that may have explicit biases (that they may knowingly discriminate), but may also have implicit biases in which they react differently to different types of people without being aware of it.

This training also focuses on a variety of mechanisms to reduce the impact of these biases on the actions of police officers. Participants are taught to be mindful of their biases, to focus on the positive characteristics of individuals rather than negative stereotypes, to try to take the perspective of other people and understand their situation and motivation, to get to know people as individuals rather than treating them as members of a group, and to seek out opportunities to get to know people from different backgrounds. This training has been shown to be effective in academic settings. The question is whether it works with police officers.

This question was addressed in a 2023 paper by Calvin Lai and Jaclyn Lisnek in Psychological Science. They gave a variety of tests to over 3,500 police officers given this anti-bias training as part of their work. Participants were given a pre-test before the training to assess their attitudes about bias and their tendency to use the strategies being taught in the training. Most participants took only a post-test immediately after the training. A smaller number (about 150) also got a follow-up test a month after the training.

The results were mixed.

Immediately after the training, it had an impact on both attitudes about bias and intention to use the techniques discussed. Officers believed more strongly that bias exists and is important. They also indicated that they were interested in using the techniques in the next seven days.

When assessed a month after the training, though, the results were more muted. Officers continued to believe in the existence of bias a month after the training, but their concern about the impact of bias returned to the level it had prior to the training. In addition, participants indicated that their use of the techniques described in the training was about the same as it was prior to the training. So, the training did not appear to have a lasting impact on the behavior of officers.

Being a police officer in the United States is a difficult job. The job is dangerous and involves a lot of stressful situations. It is not surprising that a brief training presented once to officers did not have a long-term impact on behavior even though similar manipulations have worked in other contexts. It is valuable to know that this training did affect beliefs about bias. That is an important first step. Ultimately, a big takeaway from this study is that it is important to have studies to determine both the short-term and long-term impact of this kind of training.