We introduce the package ddml for Double/Debiased Machine Learning (DDML) in Stata. Estimators of causal parameters for five different econometric models are supported, allowing for flexible estimation of causal effects of endogenous variables in settings with unknown functional forms and/or many exogenous variables. ddml is compatible with many existing supervised machine learning programs in Stata. We recommend using DDML in combination with stacking estimation which combines multiple machine learners into a final predictor. We provide Monte Carlo evidence to support our recommendation.
This paper investigates whether employment restrictions contribute to refugees having poorer labor market outcomes than citizens. Utilizing linked register data from Switzerland and within-canton policy variation between 1999-2015, we find substantial negative effects on employment and earnings when refugees are barred from working upon arrival, excluded from specific sectors or regions, or face resident prioritization. Removing 10% of refugees' outside options reduces job-to-job mobility by 7.5% and wages by 3.0%, widening the wage gap to citizens in similar jobs. The restrictions depress refugees' labor market outcomes even after they apply, but do not spur emigration nor benefit other immigrants.
Since 2015, over 7 million Venezuelans have been forced to leave their homes, seeking refuge predominantly in neighboring countries across Latin America and the Caribbean. The displacement is typically accompanied by vulnerability and marginalization, yet there is a scarcity of actionable evidence on how to alleviate poverty among immigrants and refugees and facilitate their economic, political, and social integration. This study evaluates the impact of a cash-based intervention (CBI) on multidimensional integration outcomes of highly vulnerable Venezuelan immigrants in Peru. Utilizing an original panel survey of beneficiaries and the staggered rollout of the program, which provided a one-time payment of 760 soles (approximately 190 USD or 74% of the monthly minimum wage), we estimate that the CBI increased the IPL-24 index---an overall measure of immigrant integration capturing several dimensions---by at least 0.12 standard deviations. This increase is mainly driven by improvements in the navigational, social and economic components of the IPL-24 index. Moreover, the CBI boosted self-employment by 2 percentage points and raised the intention to emigrate from Peru by 1.2 percentage points. Additionally, our heterogeneity analysis reveals that the benefits of the fixed-amount cash payment diminish significantly with the size of the household. We discuss how these findings inform the design of future CBI programs aimed at supporting vulnerable immigrant and refugee families.