The paper examines 49 empirical studies that estimate the impact of forced displacement on host communities. The reviewed studies cover 17 major forced displacement crises occurring between 1922 and 2015, to host countries at different levels of economic development and different types of forced migrants. The authors undertake a comparative analysis of the empirical models used in this literature and identify the specific modeling and econometric challenges that this type of work entails. They find that all studies can be classified as ex-post quasi-natural experiments. The unexpected nature of the crisis and the randomness of the allocation of displaced persons are used to defend the natural experiment assumption. However, all papers address the question of endogeneity, mainly using an instrumental variable approach focusing on either distance from the shock or previous location of migrants. There is a preference for partial equilibrium modeling, differences-in-differences methods, and cross-section econometrics, depending on the type of data available. Employment is the outcome most studied followed by wages, prices and wellbeing. The authors also conduct a meta-analysis of 762 empirical results emerging from this literature, and find:
- About half of the results on income, consumption or wealth indicate net improvements in household wellbeing. 45-52 percent of the results are positive and significant, 34-42 percent of the results are non-significant, and 6-20 percent are negative and significant. Overall, between 80 and 94 percent of results are either positive or non-significant indicating that the likelihood of a forced displacement shock resulting in a negative outcome for host communities is less than 20 percent.
- The majority of results on employment and wages are not significant. 12-20 percent of the results are positive and significant; 63 percent are non-significant, and 22-25 percent are negative and significant. Overall, the likelihood of a forced displacement crisis resulting in a negative impact on employment for the host population is below 25 percent; negative and significant results are associated with young and informal workers. The overall evidence for wages shows that only about one in four results are negative and negative results seem to turn positive in the long-term.
- The results on prices show asymmetric behavior across types of products. 37-46 percent of results are positive and significant, 18-23 percent are non-significant, and 36-39 percent are negative and significant. Overall, a forced displacement crisis generates significant changes in prices. The probability of these changes being positive or negative is approximately similar but associated with the type of product or service observed. Prices for food and rents tend to increase whereas prices for labor intensive products and services tend to decline.
- Overall, less than one in five results can be considered negative for host communities.