Implicit Racial Bias Affects Risk Preferences

Discrimination is difficult to notice, as it often takes place in everyday situations, such as during job applications or apartment searches. To test this, we ran an online experiment that transformed decision-making into a game-like experience to capture the behavior of ordinary people. What we found striking was that people’s decisions about others change depending on their skin color.

In the social sciences, discrimination describes unfair treatment, exclusion, unequal opportunity, or the marginalization of people based on characteristics irrelevant to a social interaction or exchange. This can be their assumed ethnicity, language, religion, gender, country of origin, or skin color. Discrimination can manifest itself in all aspects of life, including the labor market, education, housing, everyday interactions, or legal rights.

While much research has been devoted to describing patterns of discrimination, understanding the circumstances under which discriminatory behavior is more common has proven more challenging. Economists never tire of pointing out that discrimination is costly, such as when an employer passes over the most qualified candidate if their stereotypes lead them to refuse an interview. The implication is that the more is at stake, the less discriminatory behavior we should observe. However, we observe little evidence of declining patterns of discrimination in the labor or housing markets, where such evidence is available.

Game-Like Survey

To better understand the relationship between risk-taking and discriminatory behavior, we developed a game-like survey. The survey is presented as a soccer-themed game to make it more accessible to participants (who know they are being surveyed). They were asked to select players for a soccer team, and for some participants, the amount they earned for participating depended on how successful they were in selecting strong players. So they had a real stake in the online experiment.

Now, we’re social scientists, not game designers, so we had participants select players in different ways. Each time they were shown players in a way that is done in computer games, with a picture and an indication of objective skills. The conditions under which the participants selected players varied a bit throughout the survey. This allowed us to test different patterns of decision-making — an application of game theory.

Costs of Discrimination

With everything in place, we simply observed the behavior of participants under certain conditions. As economists would predict, we found that, in general, higher costs of discrimination can reduce disparities. Participants who were given an extra financial incentive to select more skilled soccer players did better than other participants who did not have this incentive. In these cases, people spent more time looking at the objective information we provided about the players and were less likely to be guided by skin color.

Risk and Decisions

But there was an unexpected twist to this situation. We were able to show that higher costs of discrimination — through financial incentives to do better — can also trigger implicit racial bias. This happened in one part of the game, as participants had to make several rounds of risky decisions in a lottery, although they could stop early to avoid the risk of losing too much.

Surprisingly, participants who were given an additional financial incentive to select more skilled soccer players were more likely to keep taking risks, but only if it allowed them to avoid a player with darker skin. Put differently, the fact that their final payout depended more on their choice did not encourage them to make decisions that actually increased the chances of picking the strongest players. Instead, it rather encouraged them to take risky options, and continue playing in the lottery, just to avoid ending up with a darker-skinned player.

Rethinking Risk and Bias in Decision-Making

In other words, with the online experiment, we can show that the presence of racial markers — skin color — changes people’s risk preferences when their decisions have costly consequences, that is participants assess the risks differently.

This is remarkable because our behavior seems to depend on who we interact with. We call this implicit racial bias, and this implicit racial bias may be part of the reason why members of visible minority groups often face discrimination in competitive markets, such as getting a job or finding a place to live.

It appears that we simply see the world and its risks differently when racial minorities are present. This calls for even greater care in making decisions that may affect members of minority groups, a conscious effort to follow objective criteria, and the use of institutionalized processes that minimize the risk of discrimination.

Didier Ruedin is a Senior Lecturer at the Swiss Forum for Migration and Population Studies at the University of Neuchâtel and a Project Leader of the nccr – on the move project” Narratives of Crisis and Their Influence in Shaping Discourses and Policies of Migration and Mobility.”

Reference:

–Auer, Daniel, and Didier Ruedin. 2024. ‘Experimental Evidence on How Implicit Racial Bias Affects Risk Preferences’. *Journal of Ethnic and Migration Studies.* 50(20): 5250-5269. DOI: 10.1080/1369183X.2023.2278396