I consider the problem of estimating the effect of a health care reform on the frequency of individual
doctor visits when the reform effect is potentially different in different parts of the outcome distribution.
Quantile regression is a powerful method for studying such heterogeneous treatment effects. Only
recently has this method been extended to situations where the dependent variable is a (non-negative
integer) count. An analysis of a 1997 health care reform in Germany shows that lower quantiles, such
as the first quartile, fell by substantially larger amounts than what would have been predicted based
on Poisson or negative binomial models.