An empirical model of health care demand under non-linear pricing

In 2004, the German Social Health Insurance introduced a co-payment for the first doctor visit in a calendar quarter. I combine a structural model of health care demand and a difference-in-differences strategy to estimate the effect of that reform on the number of visits. In the model, the implied incentive to delay a first visit also affects subsequent visits, as the expected remaining time to the end of quarter is reduced. This effect has been ignored by the prior literature using standard hurdle count models. Data are from the German Socio-Economic Panel. Results show no statistically significant reduction in visits due to the reform.