Additional Working Paper:
Temporal-Difference Estimation of Dynamic Discrete Choice Models (with K. Adusumilli)
Abstract: We propose a new algorithm to estimate the structural parameters in dynamic discrete choice models. The algorithm is based on the conditional choice probability approach, but uses the idea of Temporal-Difference learning from the Reinforcement Learning literature to estimate the different terms in the value functions. In estimating these terms with functional approximations using basis functions, our approach has the advantage of naturally allowing for continuous state spaces. Furthermore, it does not require specification of transition probabilities, and even estimation of choice probabilities can be avoided using a recursive procedure. Computationally, our algorithm only requires a low-dimensional matrix inversion. We find that it is substantially faster than existing approaches when the finite dependence property does not hold, and comparable in speed to approaches that exploit this property. For the estimation of dynamic games, our procedure does not require integrating over the actions of other players, which further heightens the computational advantage. We show that our estimator is consistent, and efficient under discrete state spaces. In settings with continuous states, we propose locally robust corrections in order to achieve parametric rates of convergence. Preliminary Monte Carlo simulations confirm the workings of our algorithm.
Research In Progress:
Training, Occupations, and the Effects of Graduating in a Recession
Abstract: A large body of evidence shows that labor market entry conditions have persistent effects on careers of graduates in a variety of settings. This project studies the effects of adverse entry conditions on apprentices. To this end, I use an administrative employment panel which contains the apprenticeship training for a large sample of workers in Germany, a context in which 70% of individuals with at least upper-secondary education hold apprenticeships. I then use variation in the local unemployment rate at time of graduation to study the effects of entry conditions on wages. Preliminary results suggest that apprentices suffer large initial wage losses when graduating in a recession which gradually fall and fully disappear after ten years of work experience. These effects display important heterogeneity, with higher-paying trainings generally suffering less. One underlying mechanism appears to be the selection of workers into occupations. Apprentices graduating in a recession, and particularly those trained in lower-paying occupations, are significantly less likely to start work in their training occupation, which in turn leads to substantial negative effects on wages. These results have important implications for policies aimed at mitigating the effects of adverse labor market entry conditions for young workers.
Worker Retraining Programs (with E. Ilzetzki)
Abstract: Skilled labor shortages during World War II were filled largely by unexperienced workers coming from unemployment, entrants into the workforce (women, youths), or workers transitioning from other occupations (e.g. agriculture). The U.S. War Manpower Commission (WMC) had a number of programs to train workers and help workers relocate to other regions to address these shortages. We are currently digitizing archival resources to investigate the effectiveness of these public programs in addressing mismatches between employers’ needs and the skills of the extant workforce. Modern sectoral transformation due to technology and trade has spurred renewed policy interest in worker retraining. In this regard, World War II can be viewed as a large, sudden, (albeit partly transient) sectoral transformation that created a mismatch between labor demand and the human capital of the extant labor force. Investigating what worked (and what didn’t) in the government’s attempts to address this mismatch may shed light on these modern policy questions, and help gain a better understanding of the labor market’s matching function more generally. There was some local variation in exposure to these job training programs that we expect will allow identification of their effects during World War II.