Does Team Diversity Reduce Loss Aversion? Evidence From COVID-19 Decisions

Most major decisions happen in team contexts, but relatively little is known about how the composition of a team might impact individual team members' internal decision-making processes. We explore whether diversity could be a “nudge” (Thaler & Sunstein, 2009) that improves individual decision-making in team contexts by reducing loss aversion. Loss aversion is a common and powerful decision bias (Kahneman & Tversky, 1979) and it can be influenced by the social context (Polman, 2012). Previous work suggests three contrasting views about how diversity could influence loss aversion. First, an individual-cognition perspective might suggest that diversity could “prime” individuals with a more deliberative mindset (Lount & Phillips, 2007), making people more reflective and thus less loss averse when making decisions in either team or solo contexts. Second, a social-motivation perspective might argue that individual’s attitudes towards team members will change in diverse teams, such that individuals (or at least non-minority individuals) might feel less socially connected to the team (Polman, 2012) and may ultimately care less about the team, potentially reducing how much individuals care about the team’s losses and thus reducing loss aversion. On the other hand, being in a diverse team could trigger self-protection motives, which could potentially increase loss aversion (Li et al., 2012). Finally, a social-perception perspective might argue that being in a diverse team will shift individuals’ perceptions of their teammates and of how team dynamics might play out. Such social perceptions could potentially increase loss aversion by increasing psychological distance and triggering high-level or abstract construals of decisions (Irmak et al., 2013). On the other hand, individuals might anticipate more team conflict in diverse teams (Lount and Phillips, 2007) and thus may think more deliberatively and act less loss averse when making decisions on behalf of the team (but not in solo contexts).

Method. We examined the impact of team diversity on loss aversion within the context of COVID-19 policy decisions. We pre-registered hypotheses that when making decisions on behalf of a team, individuals in diverse teams will show less loss aversion than individuals in homogenous teams. This study was pre-registered on Aspredicted.com (#57459; https://aspredicted.org/blind.php?x=d8qk5y). We requested 800 U.S. adult participants on Prolific and received responses from 771 participants (Female: 52.6%; Mean Age: 32.6 SD=11.0). We paid participants $1.50 for a 10-minute study.

Participants were asked to imagine that they were assigned to a team of six scientists and had to make decisions about whether to approve or reject different public health policies to manage COVID-19. To manipulate team diversity, participants were either assigned to a diverse team (3 male, 3 female; 2 White, 2 Black, 2 Hispanic) or to one of three types of homogenous teams (6 males; 6 White OR 6 Hispanic OR 6 Black). We displayed images of all team members’ faces (using images from the Chicago Face Database that were matched for attractiveness ratings).

We used a measure of loss aversion adapted from Gächter et al. (2007). We asked participants to make seven decisions about whether to implement a new risky health policy initiative to tackle COVID-19. Decisions varied in terms of how many people would die if the policy “failed.” Participants’ degree of loss aversion is revealed by the point where participants switched from implementing to not implementing the progressively riskier programs. Each decision was explained thusly:

“For City [A/B/C/D/E/F/G], your team has generated estimates for the consequences of the new public health policy as follows:

If you choose to approve the policy, there is a 25% chance that it will cause [1000/2000/3000/4000/5000/6000/7000] additional people to die, and there is a 75% chance that it will cause 6000 additional people to be saved.

If you choose to reject the policy nothing will happen. No additional people will die, and no additional people will be saved.”

Results. Consistent with our pre-registered hypothesis, we found that individuals who were assigned to a diverse team showed less loss aversion in their decisions than individuals who were assigned to a homogenous team (b=.43, SE=.16, p=.006). In addition, these results held when controlling for participants’ age, gender, and ethnicity (b=.41, SE=.15, p=.008).

Which of the three aforementioned theoretical perspectives best explains this main effect? First, we examined the individual-cognition account by testing whether being in a diverse team primed individuals to engage in more deliberative ( “System 2”) information processing or altered individuals’ risk preferences. However, we found that the diversity manipulation did not lead to better performance on a deliberative decision task (CRT; b=.05, SE=.09, p=.592), nor did it impact individuals’ generalized risk taking propensity (b=.16, SE=.10, p=.124).

Next, we examined the social-motivation account – namely, that individuals may care less about diverse teams because they feel less socially connected to them. Instead, we found that individuals in diverse teams in fact identified more with their team (b=1.04, SE=.10, p<.001), making it implausible that they also cared less for their individual team members.

Finally, we examined the social-perception account. We found that while team diversity did not lead participants to believe their team was more “warm” (b=-.09, SE=.11, p=.416), it did lead participants to believe their team was more competent (b=0.44, SE=.08, p<.001). In addition, participants in diverse teams perceived their team members as putting significantly more “effort into reducing the spread of COVID-19” (b=.25, SE=.06, p<.001), suggesting individuals perceived diverse teams to be more active in tackling COVID-19. Together, this indicates that diversity shifted participant’s social perceptions of their team suggesting this to be a potential mechanism driving the effect.

Discussion. In the uncertain times of the COVID-19 pandemic, teams of scientists, businesspeople, and policy makers must come together to make difficult decisions involving the potential loss of lives and livelihoods. Our project identifies how team diversity can help individuals overcome their intuitive biases towards loss aversion and, in so doing, make better decisions that could save more lives. Future research should continue exploring the potential of diversity as a “team nudge” to combat biases in individual and team decisions.