Employment Agencies and Hiring Frictions in Addis Ababa, Ethiopia

High unemployment and high turnover are the two major issues in the urban context of developing countries, yet little is understood on what prevents firms from employing workers for longer. This project examines one specific matching inefficiency: firms tend to “poach” job applicants with other competing offers. When a vacancy meets many applicants, firms may interview or give an offer to the applicant with the highest productivity at the cost of workers' lower acceptance probability, while decreasing the chance of other firms matching with the same worker, a negative externality that causes too few matches. A new professional hiring service, recruitment agencies, has the potential to address this matching inefficiency. Facing multiple hiring requests at the same time, recruitment agencies maximize the total match surpluses from multiple firms, thus partially internalizing the externality. To understand the effect of recruitment agencies on firms' matching inefficiency, I propose to sample 1,000 small and medium formal firms in Addis Ababa, Ethiopia, who are planning to post a vacancy. For each vacancy, I will first construct a novel dataset on applicants’ details and firms’ action on each applicant. Then, I will randomly select 300 firms and match them with a recruitment agency in Addis Ababa; the agency will recommend a qualified worker for firms' interview directly. To control for the possibility that recruitment agencies may introduce new labor pool, I will randomly select another 300 firms, match them with the same set of recruitment agencies, but the workers recommended by the agencies will still go through firms' existing selection process. Enumerators will conduct follow-up surveys in 1 month and 3 months to collect matching details and hiring outcomes, particularly, whether treated firms interview fewer over-qualified applicants, are more likely to fill in the vacancy, and the new hires are less likely to quit.