From Average Effects to Targeted Assignment: A Causal Machine Learning Analysis of Swiss Active Labor Market Policies

2024-10-30

AI Summary

This study evaluates the heterogeneous causal effects of Switzerland’s 2014–2015 active labor market policies (ALMPs) on employment and earnings, focusing on immigrant status (Swiss citizens, non-EU/EU permanent residents) and educational attainment as effect moderators. Leveraging administrative big data, we innovatively integrate causal forests with shallow policy trees—the first such application in Swiss ALMP evaluation—enabling a methodological shift from average treatment effects to individualized program assignment recommendations. We enhance identification precision via R-learner estimation, doubly robust inference, and high-dimensional matching. Results show that temporary wage subsidies significantly increase employment and earnings in Year 3, especially among non-EU immigrants and low-educated individuals; by contrast, basic job-search training exhibits negative effects. The policy tree substantially improves targeting accuracy, delivering empirically grounded, actionable guidance for optimizing ALMP resource allocation.

Technology Category

• Machine Learning: Active Learning
• Search and Optimization: Algorithm Configuration
• Reasoning under Uncertainty: Causality

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work
• Responsible Web: Human-perceived consequences of algorithmic deployment on the web

Abstract

Active labor market policies are widely used by the Swiss government, enrolling over half of all unemployed individuals. This paper evaluates the effectiveness of Swiss programs in improving employment and earnings outcomes using causal machine learning and rich administrative data on unemployed individuals in 2014 and 2015, including detailed labor market histories and other covariates. The findings for Swiss citizens and immigrants with permanent residency indicate a small positive average effect of a Temporary Wage Subsidy program on employment and earnings in the third year after program start. In contrast, Basic Courses, such as job application training, exhibit negative effects on both outcomes over the same period. No significant impacts are found for Employment Programs conducted outside the regular labor market or for Training Courses such as language or computer classes. The programs are most effective for individuals with a non-EU migration background, while Temporary Wage Subsidies also benefit those with lower educational attainment. Finally, shallow policy trees provide practical guidance for improving the targeting of program assignments.

Problem

• Evaluating effectiveness of Swiss labor market policies using causal ML
• Assessing impact of Temporary Wage Subsidy vs Basic Courses
• Improving targeted program assignments for unemployed individuals

Innovation

• Uses causal machine learning for policy evaluation
• Analyzes rich administrative data on unemployed
• Employs shallow policy trees for targeted assignments