Current Research

Home Ownership and Structural Changes in the Labour Market

(Jointly with Bernd Fitzenberger)

Despite increased mortgage affordability and declining borrowing costs, home ownership rates among early-career individuals in Germany have fallen since the early 2000s. We hypothesize that this trend is driven by a major shift in the composition of employment, namely, the rise of the service sector and the growing concentration of jobs in urban areas, which has led to greater job-related mobility and higher opportunity costs of home ownership. We document these developments and develop a structural model to estimate counterfactual scenarios, allowing us to assess how the housing market might have evolved under alternative conditions.

(JEL J31, J62, R21, R23)

Play Stupid Games, Win Stupid Prizes

(Jointly with Katharina Hartinger)

We use slot machine games to study strategy choice and learning under high uncertainty. In this natural yet controlled setting, players choose and switch between three slot machine games with different (but unknown) winning chances for a fixed number of rounds. We find that half of players engage in explorative behavior and that the choices of 20 percent of players are best explained by curiosity-based exploration: In the exploration-exploitation trade-off, these players switch significantly more frequently than justified by Bayesian exploration. This behavior is driven by the desire to keep entropy—the element of mystery in the game—alive. When a treatment allows players to satisfy curiosity without sacrificing expected payoff, these players move towards canonical Bayesian game choice. Survey evidence supports curiosity-driven element of decision-making in the presence of real stakes and high complexity.

(JEL C91, D81, D83, D91) [Youtube]

Learning by Problem-Solving

(Jointly with Peer Ederer and Ljubica Nedelkoska)

The standard Mincer model does not account for large differences in earnings across occupations. We present an alternative learning-by-doing model that relates wages and skill development to the level of job complexity of a worker. Using administrative data on German labour market entrants, we find that wage growth is positively related to job complexity and negatively related to the initial level of skill. We calibrate the model and show that workers in highly complex occupations acquire twice as much human capital throughout life compared to workers in simple occupations. The results suggest that learning-by-doing is a far more important earnings determinant than investment in education and training that have been the focus of the literature. (JEL J24, J31, D83) [SSRN]