Discrete choice models are usually derived from the assumption of random utility maximization We consider the reverse problem whether choice probablities are consistent with maximization of random utilities This leads to tests that consider the variation in these choice probabilities with the...

We investigate the presence of moral hazard and advantageous or adverse selection in a market for supplementary health insurance. For this we specify and estimate dynamic models for health insurance decisions and health care utilization. Estimates of the health care utilization models indicate...

In a non-stationary job search model we allow unemployed workers to have a permanent option to leave the labour force. Transitions into nonparticipation occur when reservation wages drop below the utility of being nonparticipant. Taking account of these transitions allows the identification of...

We investigate the presence of moral hazard and advantageous or adverse selection in a market for supplementary health insurance. For this we specify and estimate dynamic models for health insurance decisions and health care utilization. Estimates of the health care utilization models indicate...

We provide evidence from a large-scale field experiment on the causal effects of audit rules on compliance in a market for long-term care. In this setting care should be provided quickly and, therefore, the gatekeeper introduced ex-post auditing. Our results do not show significant effects of...

Randomized experiments provide policy relevant treatment effects if there are no spillovers between participants and nonparticipants. We show that this assumption is violated for a Danish activation program for unemployed workers. Using a difference-in-difference model we show that the...

This paper reports on a field experiment testing for sunk-cost effects in an education setting. Students signing up for extra-curricular tutorial sessions randomly received a discount on the tuition fee. The sunk-cost effect predicts that students who receive larger discounts will attend fewer...

We structurally estimate a novel job search model with endogenous job search effort, job quality dispersion, and effort monitoring, taking into account that monitoring effects may be mitigated by on-the-job search and search channel substitution. The data are from a randomized experiment...

Randomized experiments provide policy relevant treatment effects if there are no spillovers between participants and nonparticipants. We show that this assumption is violated for a Danish activation program for unemployed workers. Using a difference-in-difference model we show that the...