MitIgating the Consequences of Job Loss: Experimental Evidence from Ethiopia after a Tariff Shock

Lukas Hensel, Girum Abebe, Francois Gerard, Stefano Caria

Job loss is an understudied risk for formal workers in low-income countries, and it is unclear how to optimally insure workers against it. 

We provide evidence on the impacts of job loss among female factory workers in Ethiopia and on how these impacts can be mitigated. We leverage quasi-experimental variation in job loss, experimental variation in job-loss support payments, and high-frequency data spanning a period of 13 months after displacement. 

We find that job loss is a persistent shock that reduces employment and consumption spending for longer than one year, and almost doubles the rate of poverty. An additional lump-sum payment encourages early spending and reduces both overall and manufacturing employment. In contrast, providing an equivalent amount in monthly tranches -- a payment modality preferred by a majority of workers -- enables workers to better smooth consumption expenditures without negative employment effects. We show that workers have high willingness to pay for additional job-loss insurance, but also heterogeneous preferences over the payment modality.  This generates a key trade-off between workers' private welfare and the government industrialization objectives: allowing workers to choose their preferred insurance product maximises their surplus, but at the cost of reducing future manufacturing employment.