Origin-Country Social Protection and Migration Intentions: Causal Evidence from Tunisia*

This replication package reproduces all results reported in “Origin-Country Social Protection and Migration Intentions: Causal Evidence from Tunisia” (submitted to Economic Modelling). The analysis uses microdata from the Tunisia Household International Migration Survey (HIMS) 2021, focusing on non-migrant individuals currently residing in Tunisia. The study estimates the effect of two social protection regimes—contributory (employment-based) and state-assisted (means-tested)—on migration intentions among employed workers. The main estimation strategy relies on Inverse Probability Weighted Regression Adjustment (IPWRA) with survey probability weights. Additional robustness checks include Double Machine Learning (DML), propensity score trimming, and Oster (2019) sensitivity bounds. All results (tables and figures) are fully reproducible from the raw data using the provided master do-file, which constructs all variables and runs all estimations sequentially without manual intervention. The accompanying README file provides complete documentation, including data access instructions, variable definitions, and step-by-step guidance for full replication of all results.