Cash-for-Work Program to Improve Social Cohesion between Syrian Refugees and Host Communities in Jordan

We use a large-scale RCT to evaluate a program aimed to improve cohesion between Syrian refugees and their host communities in Jordan. We build on a cash-for-work intervention, where refugees and Jordanian residents cooperate in upgrading infrastructure or providing social services in Jordanian municipalities exposed to a large influx of refugees. Allport’s (1954) contact hypothesis posits that, under certain conditions, inter-personal contact between groups should lead to a reduction in prejudice. Others, such as Barlow et al. (2012), hypothesize that it can exacerbate divisions. Yet causal evidence on these alternative hypotheses is scant.

We investigate by varying the proportion of Syrian refugees who work in the teams employed by the program, thereby changing exposure to outgroup members. We measure the impact of this on stereotyping, social norms, trust between the groups, and productivity. We also evaluate the effect of disseminating information about refugees amongst residents on these outcomes.