Job Search Assistance for Syrian Refugees in Jordan
- Cash transfers
- Unconditional cash transfers
- Psychosocial support
The labor market in Jordan has low employment rates, and Syrian refugees face myriad challenges in finding work. Researchers conducted a randomized evaluation that introduced an adaptive targeted treatment assignment methodology to measure the impact of providing different types of job search assistance on job search rates and labor outcomes for Syrian refugees and local jobseekers in Jordan. While none of the interventions had an impact on whether participants were in wage employment after six weeks, a small cash transfer had large impacts on refugee employment and earnings two and four months after the intervention.
With 26 million refugees displaced globally, and displacement lasting for ten years on average, generating employment for refugees is a pressing policy issue. The Syrian crisis alone has displaced more than 13.1 million people, with almost 1.3 million settling/arriving in Jordan. Syrian refugees in Jordan have significant needs for basic amenities, humanitarian aid, and economic security. However, employment rates for Syrian refugees in Jordan are low: in 2017, the employment rate was 14 percent for all refugees and only 2 percent for women refugees. This employment was mostly informal and below the national minimum wage.
Syrian refugees face challenges securing employment in Jordan. For example, firms report that it is hard to process work permits for Syrians, and Syrian refugees compete for work with Jordanian nationals and other migrants. Refugees could also be at a disadvantage in searching for employment due to a lack of experience in and information on the local labor market. Can employing an adaptive treatment assignment algorithm to distribute carefully designed labeled cash transfers, information, or psychological support improve the job search rates and labor market outcomes of Syrian refugees and local jobseekers in Jordan?
Context of the evaluation
The Jordanian labor market has low employment rates. During the last three months of 2016, for example, the employment rate was only 30 percent. Furthermore, employment rates among refugees are even lower as they face additional constraints to participating in the labor force such as permit issuing, loss of networks, and restrictive labor policies.
This evaluation took place in three cities: Amman, Irbid, and Mafraq. For Syrian refugees or Jordanian nationals to be eligible for the study, participants had to have a valid government identification, be 18-45 years old, and be interested in participating in the near future in low-skilled, formal wage work that paid around the minimum wage.
Study participants included 1,663 Syrians and 2,107 Jordanians. Both the Syrian and Jordanian participants had similar characteristics, with 60 percent being women with an average age of 29. Two percent of participants worked in wage employment, with the average person having 4-5 years of work experience. Syrian participants had lower average educational attainment of 7 years in comparison to 12 years for Jordanian participants.
Details of the intervention
In partnership with the IRC, researchers conducted a randomized evaluation to measure the impact of providing different types of job search assistance on job search rates and labor outcomes for Syrian refugees and local jobseekers in Jordan.
The IRC enrolled participants for a program called “Project Match” on a rolling basis between February and November 2019. IRC employment service officers contacted potential program participants based on referrals and conducted door-to-door home visits in neighborhoods with a high number of refugees. In addition, people who had not been contacted by the IRC could also apply for the program.
Researchers developed a novel methodology for adaptive targeted treatment assignment to run the evaluation for “Project Match,” motivated by the fact that the evaluation had two objectives. The primary objective was to employ as many Syrian refugee and Jordanian program participants in formal jobs as possible, and the secondary objective was to test the effectiveness of alternative interventions.
The algorithm implemented by the researchers had two key features. First, the algorithm was adaptive, meaning that it changed the probabilities of being assigned to a particular intervention over time by incorporating information about the employment successes of interventions for existing program participants. Second, the algorithm was targeted, meaning that it used information about the success rates of interventions for sixteen heterogenous groups of participants (defined by nationality, sex, education level, and employment history) in order to target interventions for each individual group.
The algorithm then assigned people willing to take up formal wage work to one of three job search interventions, designed in collaboration with the IRC, or a comparison group:
Labeled cash transfer: Participants received a labeled cash transfer of 65 JOD (equivalent to around US$92) to cover the cost of a job search (e.g., transport, childcare, etc.) for about ten weeks. Study participants were given suggestions on how to spend the cash (i.e., for job search expenses), but they were informed that they were free to use it as they wished.
Information: Participants received detailed information on how to prepare for a job interview in Jordan and the legal rights of employees in the formal sector. The information was provided through face-to-face interactions, descriptive videos, and paper tools.
Nudge: Participants received a package of psychological support through a four-week job-search planning calendar, a video tutorial on how to use the calendar, a face-to-face demonstration, and SMS reminders.
Comparison: Participants did not receive cash, information, or psychological support.
All participants received 4 JOD (equivalent to around US$5.60) and a flyer with information about job interviews upon registration.
To measure the impacts of the job search assistance interventions, researchers collected information on participants during intake registration interviews, as well as follow-up phone surveys six weeks, two months, and four months after the initial interview.
Results and policy lessons
While the job search assistance interventions had no impact on the six-week employment outcomes of jobseekers, the cash intervention had large impacts on refugee employment and earnings two and four months after the intervention.
Job search: Six weeks after the program began, none of the interventions raised employment for program participants. However, after two months, the cash transfer intervention led to an increase in the proportion of Syrians who searched for work by 5.6 percentage points (a 13 percent increase from the comparison group job search rate of 44 percent) and resulted in Syrians submitting 0.5 more job applications (a 44 percent increase from the comparison group average of 1.2 applications). The information and nudge interventions had no impact on the job search rates for Syrians, but they led Syrians to place more job applications.
Among Jordanians, the cash and information interventions had no impact on the job search rate after two months. However, the nudge intervention increased the proportion of Jordanians who looked for work by 6.5 percentage points (an 11 percent increase over the comparison group search rate of 58 percent). No intervention had an impact on the number of applications placed or the number of hours of job search among Jordanians.
Employment and earnings: Among Syrians, the cash intervention led to a 5.2 percentage point increase in the employment rate after two months (a 57 percent increase over the comparison group employment rate of 9.1 percent) and a 3.8 percentage point increase after four months (a 73 percent increase over the comparison group employment rate of 5.2 percent). Earnings among Syrians also increased by about 40 percent after two months and 65 percent after four months. For Syrian refugees, after two months the information intervention increased employment and earnings by nearly the same amount as the cash intervention; however, after four months the information intervention had no impact on employment or earnings. Similarly, the nudge intervention had no impact on employment or earnings of Syrian refugees after four months. For Jordanians, none of the interventions had an impact on employment or earnings after two or four months.
Overall, cash grants had the largest impacts on job search and employment for Syrian refugees. The results suggest that cash grants were effective among refugees because they targeted liquidity constraints, such as refugees’ inability to pay for job search and application costs, that were key barriers in the labor market.