Give Me Your Tired and Your Poor: Impact of a Large-Scale Amnesty Program for Undocumented Refugees
Dany Bahar, Ana María Ibáñez, Sandra V. Rozo
Key Findings
- A twofold increase in PEP holders per 100,000 inhabitants reduces formal employment by only 0.1 percentage points for Colombian workers—a 0.15% deviation from the mean
- No significant effects on hours worked, wages, or labor force participation for Colombian workers in either formal or informal sectors
- Negative employment effects are concentrated among highly educated and female Colombian workers, consistent with Venezuelan migrants being more educated on average
- The program successfully increased formal employment rates for Venezuelan workers, consistent with its regularization purpose
- Results are not explained by low take-up: 64% of registered undocumented migrants applied for and received a PEP visa
About This Research
What happens when a developing country grants work permits to nearly half a million undocumented migrants at once? This paper studies the labor market impacts of Colombia's Permiso Especial de Permanencia (PEP) program—the largest migratory amnesty offered to undocumented migrants in a developing country in modern history. In July 2018, outgoing President Juan Manuel Santos unexpectedly decreed that all 442,000 undocumented Venezuelans registered in a national census would be eligible for a two-year renewable visa granting the right to work and access public services.
We employ a difference-in-difference methodology comparing labor outcomes in departments with different treatment intensity before and after the program's August 2018 roll-out. To address endogeneity concerns—migrants might self-select into certain areas—we validate our results using three instrumental variables: early Venezuelan settlements from 1993 and 2005 censuses, and the average registration window available to migrants in each department (exogenously determined by registration numbers).
Contrary to fears that often drive resistance to regularization programs, we find only negligible effects on native workers. The effects are concentrated among highly educated and female Colombian workers—likely reflecting that Venezuelan migrants are, on average, more educated than Colombians. The program successfully increased formal employment among Venezuelans themselves. These findings suggest large-scale regularization need not harm host country labor markets, even in developing countries with large informal sectors.