Presented at: Lunch-Time Meetings in Applied Econometrics - University of Innsbruck (November 2023), SASCA Ph.D. Conference in Economics (September 2023); 38th AIEL Conference (September 2023); Workshop on Higher Education and Equality of Opportunities (June 2023); Workshop on Econometric Theory and Applications (Poster - June 2023); 36th Annual Conference of the European Society for Population Economics (ESPE) (June 2023); Ca' Foscari Phd Monday Seminar (June 2023); Economics, Econometrics, and Finance (EEF) seminar series - University of Groningen (March 2023); Ca’ Foscari Internal Seminar (March 2023); Brownbag seminar - St. Gallen University (October 2022).

Abstract: "This work documents the role played by disability insurance, typically part of a wider public pension provision package, on the labour market trajectories and retirement decisions. We will first employ a machine learning approach to estimate a Transition Probability Model able to uncover the most likely labour market histories and then evaluate the effects of policy reforms, including reforms to the eligibility for disability insurance benefits. The main contribution is the introduction of disability insurance programs within a framework, which models the entire life course of older Europeans. This requires the detailed administrative eligibility criteria prevailing in each of the 11 countries from 1970 to 2017. Results show that the disability route and early retirement are substitutes. In addition, tightening eligibility rules of disability programs crowd out disabled workers, whose reductions in working capacities are correctly assessed, towards other compensatory schemes (e.g., unemployment benefits or early pension) in which working is not expected. On the contrary, individuals with over-assessed reductions in working capacities are the most reactive to disability policy restrictions. In conclusion, efficient disability assessment procedures are crucial for incentivising labour market participation without hurting individuals most in need."

Abstract: "This study examines the impact of occupational stressors and tasks throughout an individual's career on their health in older age. Leveraging comprehensive job occupation data from the SHARE dataset, we establish precise connections between stressors and specific jobs at the 4-digit ISCO code level. To ensure accurate measurement of physical exertion, we propose the use of Metabolic Equivalent of Task (MET) based on the metabolic rate consumption associated with each task. Our study makes two key contributions. First, we provide compelling evidence that individuals, especially women, engaged in physically demanding jobs experience significantly worse health in older age. Our results remain valid after conducting several robustness checks and after controlling for a rich set of variables. Secondly, we introduce a novel methodology to identify harmful tasks and measure overall Job Strain Intensity, which also incorporates unobserved occupational stressors. This approach allows us to pinpoint specific harmful tasks and 4-digit ISCO codes, providing valuable insights for targeted retirement schemes and addressing important considerations regarding the fairness of statutory retirement ages. Additionally, policymakers can benefit from our findings to foster healthier work environments and guide investments towards automating high-risk tasks, thereby improving overall workplace safety and well-being."