Contact in the Workplace and Social Cohesion: Experimental Evidence from Uganda

Forced displacement poses a fundamental challenge to social cohesion in host countries: the arrival of large numbers of refugees can disrupt trust and social relations between communities. Understanding how to rebuild cohesion between refugees and host populations is therefore both a policy priority and an open empirical question, particularly in low-income countries that host the majority of the world’s refugees. We conduct a randomized control trial with refugee job seekers and native workers in locally owned and managed firms in Uganda. We study whether contact in the workplace, the most important setting for inter-group interaction in adult life, can improve social cohesion between the two groups. We measure social cohesion through a compound measure incorporating attitudes, implicit bias measured through Implicit Association Tests (IATs), explicit biases, and behaviors in real and hypothetical activities. We find two sets of results. First, explicit bias decreases for both groups, while implicit bias increases only for native workers. Second, both groups of workers improve their behaviors towards the opposite group, but in a slightly dierent way: while local workers want to have more refugee business partners, refugee workers want to be more employed by Ugandan firms. These findings suggest that labor market programs can serve as a vehicle for social cohesion, and that implicit bias measured through IATs may capture perceived occupational competition rather than prejudice – and, importantly, does not translate into discriminatory behavior.