The goal of this project is to help public policy in reducing the social problems that are associated with ethnic diversity. Considerable evidence reveals that ethnic diversity undermines the ability of groups, communities, and societies to act collectively and cooperatively. However, surprisingly little is known about the strategies that might support interethnic cooperation. This research uses experimental methods to evaluate two strategies for reducing intergroup biases and promoting cooperation: a self-categorization (SC) strategy that advocates the construction of unifying superordinate identities to minimize the salience of in-group differences (e.g., based on team, nation), and a self-verification (SV) strategy that proposes the construction of intergroup beliefs that acknowledge the tremendous value of diversity.
Both the SC and SV approaches have been shown empirically and experimentally to be plausible diversity management strategies. But there are three limitations to existing research. The first is that, with few exceptions, virtually all experiments that identify positive effects from SC or SV rely on convenience samples consisting of American students, making it unclear how these strategies would fare in populations with deeper and more intense divisions. Second, most research on SV is qualitative or based on survey research, and does not convincingly control for factors that might affect both SV and group productivity. In other words, there is uncertainty as to whether SV has a causal impact on group productivity. And third, no studies compare directly the relative efficacies of the SC and SV approaches in a well-controlled setting, making it unclear which strategy to prefer in real-world settings.
To make a scientific advance, this research uses a field experiment in group productivity and cooperation to test the individual, relative, and combined efficacies of the SC and SV approaches to diversity. In the experiment, an ethnically diverse sample of participants is recruited to participate in a problem-solving competition for a large cash prize. Participants are divided into groups of three for the competition, and groups are randomly selected to receive an SC prime, an SV prime, both SC and SV primes, or neither (pure control). Under these different treatments, participants attempt a series of incentivized puzzles and cooperative tasks designed by behavioral economists and social psychologists to study various aspects of group productivity and cooperation. Outcomes are then compared across the two cross-cutting treatments to evaluate the individual, relative, and combined efficacies of the SC and SV strategies.
This research has practical implications in a world of increasing economic and cultural integration, where groups and societies are not only getting more diverse, but also fearful of its potential for generating social conflict. This is the reason the experiment will be conducted in Nairobi, Kenya where diversity appears to be undermining a half-century of impressive economic development, with imminent violence that would have been a threat to American interests in the region. Many public and private organizations, in Kenya and elsewhere, have tried to minimize the costs of diversity by becoming "colorblind" and creating strong organizational norms to encourage equal treatment of all individuals regardless of background. Nation-states have similarly resorted to appeals to a collective, national identity as a strategy for alleviating intense interethnic cleavages domestically. Yet it is not clear how effective or even desirable such convergence to homogeneity is. This research explores this tension between the costs and benefits of diversity, and attempts to inform our approaches to ethnically and culturally fractionalized groups, organizations, communities, and societies.
Due to globalization and irreversible technological advances, groups, organizations, cities, and countries are becoming increasingly more diverse ethnically and culturally. Increasing diversity presents both challenges and opportunities. Its challenges are well-known and include a number of socially undesirable outcomes, like an increased likelihood of civil conflict, lower levels of generalized trust, the underprovision of public goods, and democratic instability. But increasing diversity also presents opportunities. Because diverse populations possess a variety of useful skills and perspectives, they are more productive and better at intellective tasks like problem-solving, prediction-generation, and decision-making. Moreover, cross-country comparisons reveal that measures of birthplace diversity have strong positive correlations with economic development, and that the gains from immigration are larger than the gains from trade. This study aims to identify practical strategies and interventions that can enable multicultural societies to minimize the challenges while simultaneously taking advantage of the opportunities created by increasing diversity. Using experimental methods, the study evaluates two widely-used and competing classes of diversity management strategies: (1) assimilationist strategies, which encourage the construction of superordinate social identities (e.g., based on a team, religion, or nation) to increase perceptions of "we-ness," and (2) multiculturalist strategies, which entail the construction of shared intergroup beliefs that acknowledge the value of each group's unique culture (e.g., through diversity training, education, or public policies that celebrate ethnic differences). Assimilationist and multiculturalist strategies are advanced by practitioners, policy-makers, and academics, but their relative effectiveness is a topic of much debate. A large body of experimental work in social psychology and organizational behavior provides support for assimilationist strategies, demonstrating that the construction of superordinate social identities can effectively reduce the tensions commonly associated with ethnic differences in groups and teams. But adherents to multiculturalist approaches caution that such convergence towards homogeneity reduces the productivity gains from heterogeneity by eliminating functional differences. In political science there is evidence that multiculturalism policies—such as those that provide legal protections for minority rights, institute affirmative action, support bilingual education in schools, and permit dual citizenship—are associated with increased ethnic tolerance and social cohesion across developed countries. Critics, however, call for a return to the assimilation model, contending that multiculturalism exacerbates ethnic tensions by highlighting differences and creating a divisive "us" versus "them" mentality. There are thus substantial disagreements about which class of strategies to prefer in real-world settings. One of the primary reasons for why this debate remains unresolved is that there have been, to the best of my knowledge, no systematic, well-controlled tests of the comparative advantages and disadvantages of these competing classes of diversity management strategies. Existing comparisons of assimilationist and multiculturalist approaches rely on qualitative evidence or observational data. While certainly illustrative, such studies cannot disentangle correlations and isolate causal effects, leaving uncertainties as to what the true effects of the two strategies are. To overcome these methodological issues and generate actionable insights, I leverage the analytical power of experimental methods and conduct a field experiment designed to test the individual, relative, and combined effectiveness of assimilationist and multiculturalist approaches on several indicators of group performance. In the experiment, an ethnically diverse sample of participants from Nairobi, Kenya is recruited to participate in a problem-solving competition for a large cash prize. In the controlled competition, participants are randomly assigned to groups of three and asked to complete a combination of incentivized intellective and cooperative tasks that are widely-used by social psychologists, organizational theorists, and behavioral economists to study group problem-solving and cooperation. Saliva samples are also taken to measure participants' cortisol and testosterone levels, hormones commonly associated with stress and aggressive behavior, respectively. To investigate group performance under different diversity management strategies, the experiment features two cross-cutting treatments in which groups are randomly selected to receive an assimilationist prime, a multiculturalist prime, both assimilationist and multiculturalist primes, or neither (pure control) before attempting the experimental tasks. The experiment generates several novel findings, which can be summarized as follows: assimilationist and multiculturalist strategies have divergent effects on intellective task performance and social cohesion. In experimental puzzles that tested group innovation, creativity, and prediction, diversity had a positive causal impact on group performance only in treatment groups that received the multiculturalist prime, implying that assimilationist approaches create an unproductive convergence towards homogeneity. Moreover, in those treatment conditions where diversity generated productivity gains, the effects were non-monotonic, such that moderately fractionalized groups performed significantly better than extremely fractionalized groups and homogenous groups. While the multiculturalist prime enabled moderately diverse groups to realize the gains from heterogeneity, it failed as a strategy for promoting social cohesion. The multiculturalist prime decreased prospects for interethnic cooperation, as indicated by altruistic sharing across ethnic lines, and this negative effect was particularly strong between ethnic groups with a salient history of conflict. The assimilationist prime, on the other hand, increased altruistic sharing across ethnic lines.
