Research

One out of every 200 people in the world is a refugee. In most host countries, refugees face legal barriers to work, confining them to informal work or unemployment. This paper studies how granting refugees the right to work reshapes the allocation of both refugee and host labor across occupations. I leverage a unique natural experiment – a large-scale work permit scheme for Syrian refugees in Jordan – and assemble a novel dataset to study how the policy impacted the labor market outcomes of both refugees and hosts. Using a shift-share measure of exposure to refugee competition, I document three effects on Jordanian workers. First, Jordanians exit occupations highly exposed to refugees, re-sorting elsewhere. Second, consistent with a standard sorting model, exit coincides with an increase in the average wage of the Jordanians who remain in exposed occupations. Third, re-sorting leads to occupational upgrading, as college-educated Jordanians enter less exposed, higher-paying jobs. To formalize sorting as a mechanism, I then build a model of occupational choice nested in general equilibrium. The estimated model implies Jordanians experience modest wage gains and a small rise in unemployment from the policy. Distributionally, the poorest Jordanian workers benefit the most from the work permit scheme, despite being the ones to lose in a benchmark without re-sorting. Unsurprisingly, refugees see large wage gains from the work permit scheme and the corresponding improvement in refugee labor utilization increases total output by nearly 11%. Work permits unlock aggregate efficiency gains and, through re-sorting, reduce host country income inequality. 

Do central slums provide essential economic and social benefits to the poor? We collected bespoke data for 5,000 households to study mass forced clearances in Addis Ababa. Evictees were offered alternative subsidized housing further from the center. Exploiting sharp clearance zone boundaries, regression-discontinuity estimates show negative impacts on social networks, but positive impacts on work, earnings, housing quality and environmental amenity. Relocating households close to their ex-ante neighbors eliminates social costs. Slums are not essential: relocation policies can be designed to fully compensate residents, and the sale value of cleared land more than covers the cost.

While macro models estimate high migration flows in response to climate change, little is known about why individual households migrate. This project studies the climate-driven migration decisions of ultra-poor rural households, who are among the most climate-vulnerable, and the role that place-based poverty-alleviation programs play in these decisions. We develop a discrete-choice model of the decision to migrate and apply it to a hallmark poverty-alleviation program: the Ultra-Poor Graduation (UPG) program. The project aims to answer whether UPG programs insure households against climate stress in the short-run, and whether, and how, this deters or enables optimally timed migration to climate-resilient urban areas. To discipline the model, we will build on an ongoing J-PAL evaluation of UPG programs in Upper Egypt, by collecting a five-year endline to identify key model parameters.

Most refugees live in long-term displacement. Given these time horizons, it is critical to improve not just access to jobs but to the right jobs—those that match individuals’ existing skills and training. This project studies the misallocation of refugee skills in the context of Syrian refugees in Jordan. Despite the 2016 Jordan Compact granting Syrian refugees the legal right to work, there is still a high concentration of Syrian workers in sectors that were easiest to access prior to the policy. One potential reason for this occupational persistence is Syrian refugees’ limited access to personal connections, or wasta, which are central to job search and matching in Jordan. This project aims to alleviate this barrier by experimentally connecting Syrian refugees to employed Jordanian volunteers, through NGO-facilitated one-on-one “coffee meetings.” During these informal meetings, pairs discuss job opportunities and strategies for navigating the labor market. To evaluate the impact of these new connections, I estimate the treatment effects on refugees’ knowledge of job opportunities – the number and quality of job opportunities learned about – as well as concrete outcomes such as the job applications submitted, interviews received, and job offers. The project aims to provide evidence on whether improving access to host-community networks can reduce mismatch and improve refugee labor market outcomes. 

74% of refugees live in protracted displacement, with the majority facing work restrictions. Governmental policies limiting refugee employment could be detrimental to both refugees and host communities, who have to support refugee populations while missing out on their contributions. This study aims to provide the first experimental evidence on refugee labor market integration. We partner with the Government of Ethiopia's Refugees and Returnees Service to randomize an intervention, which provides administrative support to refugees to obtain work permits and business licenses, across 16,000 refugees in four camps. We measure impacts across multiple domains: economic outcomes including employment status, hours worked, earnings, and consumption; labor market outcomes such as job search behavior, skills utilization, and formality; psychosocial outcomes including mental health, autonomy, and dignity; and integration outcomes spanning social networks, mobility, and interactions with host communities.

This project examines how refugees’ legal access to work shapes the educational investments of refugee youths. I exploit a natural experiment in Jordan, where Syrian refugees were granted the right to obtain work permits after years of no legal access to employment. The reform relaxed binding labor-market constraints, shifting the expected returns to education for refugees and locals. To study the impact of the work permit scheme, I compile a harmonized dataset of over 350,000 individuals spanning 2014–2021, covering both Syrian refugee and Jordanian youths. Using a cohort event-study design that exploits variation in refugees’ year of arrival and age at the time of the reform, I estimate how exposure to the policy at different developmental stages affected schooling and early labor outcomes. The design isolates changes in educational investments driven by shifts in labor market returns from those driven by longer exposure to host-country norms, as Jordanians have more education than their Syrian counterparts on average. Preliminary results show that the introduction of work permits narrowed school-enrollment gaps between Syrian and Jordanian youth—especially among secondary-school–aged girls—and modestly increased attainment for younger cohorts. These findings reveal that easing legal barriers to work can foster human-capital accumulation in protracted displacement and highlight a new channel through which inclusive labor-market policies can promote long-run integration.