Research

In my research program I explore pathways to reducing conflict in the context of diversity. Specifically, I measure judgments of and responses to contemporary group-based disparities and discrimination and test interventions to address threats to social cohesion. I take a multi-method, multi-level, collaborative approach to investigate:

1. how racial bias affects and is affected by interracial interactions

2. how attributions and beliefs about bias shape evaluations of perpetrators of discrimination and decisions about punishment

3. how race- and gender-based economic disparities are misperceived

More recently, I have begun to write pieces directed toward organizational leaders, policy makers, and practitioners about how organizations can apply insights from social psychology and organizational behavior to address inequality.

Below, I detail selected findings from each line of research.

Research Line #1: Racial Bias Affects and is Affected by Interracial Interactions

One major threat to social cohesion in the United States is the racial segregation of social networks. The social network of the average White American is over 90% White, which likely hinders their ability, willingness, and motivation to effectively confront inequality. However, the American workforce is swiftly diversifying, increasing the likelihood that White people will come into contact with people of color. In work in Psychological Science, I examined the effect of reported interracial contact on explicit and implicit bias two years after the reported contact (Onyeador et al., 2020). Controlling for initial contact with Black people and initial anti-Black bias, non-Black physicians’ contact with Black people during medical school was related to less explicit and implicit racial bias during residency. Our findings indicate that more frequent and, especially more favorable contact with Black people can contribute to reducing bias and have implications for disparities in health and healthcare more broadly.  

Future directions
In the future I plan to explore whether organizational (e.g., number of Black medical students or faculty; average racial bias at the medical school level) and structural factors (e.g., average racial bias at the county or state level) influence individual physicians’ change in bias over the course of medical school.

In this work, we explored but did not find evidence that self-reported hours of diversity training were related to reduced explicit or implicit anti-Black bias—a concerning discovery because of the cost and popularity of these trainings. To follow up on this, I launched studies evaluating diversity trainings for residents at the Yale School of Medicine and managers in the Yale-New Haven Health system.

Research Line #2: Information and Beliefs about Bias Shape Evaluations of and Decisions to Punish Perpetrators of Discrimination

A second threat to social cohesion is that allegations of discrimination often spark defensiveness from majority group members. In my second research line, I examine two factors that shape how people evaluate and punish those who commit acts of discrimination: whether the discrimination is attributed to implicit or explicit bias and whether people believe that one’s bias can or cannot change.

Implicit bias is increasingly used to explain discrimination, especially in organizations. It is not clear, however, what effect attributing discrimination to implicit (vs. explicit) bias has on decisions to punish perpetrators of discrimination. In several studies, my colleagues and I have found that attributing discrimination — such as racial bias in police violence or gender bias in employment — to implicit bias reduces how intentional the discrimination is perceived to be, and thus reduces evaluations of the harm done to the victim and the tendency to blame and punish the perpetrators (Daumeyer, Onyeador, Brown, & Richeson, 2019, Journal of Experimental Social Psychology; Onyeador, Shapiro, Daumeyer, Richeson, & Henderson, in prep, based on dissertation 2017).

Future directions
I secured a multi-year postdoctoral research grant from the National Science Foundation to further explore the effects of framing discrimination in terms of implicit bias. To build on my initial work, I plan to examine the effect of implicit bias framing on vigilance for bias as incidents unfold on video (e.g., police violence against Black citizens). If implicit bias framing also undermines punishment for mistreatment captured on video, that would indicate that we should exercise caution in attributing police violence to implicit bias. It may also help explain some of the reticence to punish officers accused of excessive force against Black citizens.

I am also exploring whether a publicly available implicit bias training might also undermine, rather than increase, willingness to punish perpetrators of discrimination. Given the money spent by organizations on bias trainings, it is clear that an empirical approach is needed to guide their design, implementation, and evaluation.

Beliefs about whether bias itself is malleable can also affect observers’ decisions to punish those who discriminate. In recent years, various public figures, including business leaders and politicians, have been accused of group-based harassment and discrimination. Even after the accused apologized, observers’ reactions varied. I conducted a series of studies to better understand when an observer would condemn rather than forgive a person who has been accused of discrimination but who has apologized. Although most people forgive perpetrators after they apologized, those who believe or who have been led to believe that bias is fixed and feel personally responsible for addressing bias continue to condemn both biased actions and those who committed them (Onyeador, Lassetter, Wang, Todd, Shapiro, & Neel, in prep).

Future directions
Beliefs about the malleability of traits are not only relevant with respect to deciding to punish those who discriminate. They can also play a role in deciding to reconcile after conflict. Given the polarization and conflict in the US political system leading up to and following the 2016 election, I extended this research by testing whether beliefs about the malleability of political views would predict support for political reconciliation; Onyeador, Littman, Lassetter, Neel, & Rand, in prep).

Research Line #3: Estimates of Group-Based Economic Disparities are Inaccurate and Difficult to Change 

A third threat to social cohesion in the United States is economic inequality, which is growing rapidly, is present between social groups, and requires complex, large-scale interventions at multiple levels to properly address. Many racial disparities, especially economic disparities between Black and White Americans, have exhibited little change over the last 50 years. In contrast, economic disparities in the domain of gender have shown a marked reduction over approximately this same period. The downstream costs of economic inequality include reduced social mobility, increased crime, and poorer health, all of which negatively affect American society. My third research line examines Americans’ perceptions of group-based economic disparities and tests interventions to improve the accuracy of these perceptions.

Across a series of studies, we have found that Americans overestimate contemporary economic equality between Black and White people (Kraus, Onyeador, Daumeyer, Rucker, & Richeson, 2019, Perspectives on Psychological Science). Americans also overestimate contemporary economic equality between men and women, although misperceptions in the gender domain are smaller in magnitude than those in the racial domain (Onyeador, Zendell, Albrecht, Daumeyer, Rucker, Kraus, & Richeson, in prep). In an attempt to reduce misperceptions of racial economic equality, we exposed White Americans to information about the role past racism has played in producing contemporary racial disparities (Bonam et al., 2018; Nelson et al., 2013). White participants read about the persistence of racial discrimination or control information prior to estimating economic equality between Black and White Americans in the past and present.  Unexpectedly, Whites who read about discrimination estimated more racial economic equality in the past than those who did not read about discrimination, perhaps because, if discrimination persists into the present, past inequality must not have been so bad. Further, reading about discrimination did not affect estimates of present racial equality (Onyeador, Daumeyer, Rucker, Duker, Kraus & Richeson, 2020). This work highlights the difficulty of improving the accuracy of White Americans’ perceptions of contemporary racial economic equality and, relatedly, their resistance to evidence that the nation may not be as fair and just as they think.

Future directions
In my future work in this research line, I plan to consider the role of more socio-structural factors, like the extent of economic inequality at the local county level.

One major consequence of misperceiving economic disparities is that these misperceptions can dampen the support needed for policies that earnestly and effectively address them. Thus, another future direction is to examine the relationship between perceptions of group-based economic equality and support for equity-enhancing policies.

What can organizations do about inequality?

It’s not enough to understand the myriad challenges that we face in producing equity in contemporary society. Individuals and organizations will need to apply these insights in their own efforts and processes. More recently, my colleagues and I have written pieces that synthesize research and make recommendations.

A common response when diversity, equity, and inclusion challenges arise is to call for diversity training. By some estimates, companies spend $8 billion on diversity training (Lipman, 2018), often focusing on unconscious or implicit bias. However, diversity trainings in general, have limited, if any, utility for increasing the underrepresentation of people of color (Kalev et al., 2006) and can result in defensiveness and feelings of exclusion amongst Whites (Plaut et al., 2011). In Behavioral Science & Policy, we offer recommendations for organizations that want to develop and deliver effective anti-bias training (Carter, Onyeador & Lewis, 2020). We highlight 5 challenges that arise when attempting to develop anti-bias training, and offer 5 recommendations for addressing those challenges. At core, it is important that trainings are recast in light of what they actually can do: educate and raise awareness about bias and offer attendees strategies for behavioral change. Further, trainings are more effective when implemented alongside other diversity initiatives, like mentoring programs (Bezrukova et al., 2016).

The diversity trainings that organizations provide are often focused or framed as addressing “implicit bias.” However, recent findings from the field of social psychology raise challenges for this approach, because implicit bias is resistant to change (Lai et al., 2014; Lai et al., 2016). Further, change in implicit bias is not associated with change in explicit bias or discriminatory behavior (Forscher et al., 2019) and diversity trainings generally do not affect implicit (or explicit) bias in the long term (Forscher et al., 2017; Onyeador et al., 2020). In fact, as mentioned above, when incidents of discrimination are framed in terms of implicit bias rather than explicit bias, observers believe discrimination is less intentional and are less willing to punish perpetrators of discrimination (Daumeyer, Onyeador et al., 2019, 2020). These findings raise the question, how central should implicit bias be in our approaches to addressing diversity, equity, and inclusion challenges? In Policy Insights from the Behavioral and Brain Sciences, we argue that we may want to move beyond framing these challenges in terms of implicit bias, and we offer several insights from social psychology and organizational behavior for implementing interventions (Onyeador, Hudson, and Lewis, 2021).