The paper empirically studies whether pharmaceutical firms uses advertising as a signal for high quality drugs. A nested random effects count data hurdle model is introduced to handle the excess number of zero observations in the sample as well as nested random drug, firm and substance specific...

The paper empirically studies whether pharmaceutical firms use advertising as a signal for high quality drugs. A multiple random effects count data hurdle model is specified and a simulated maximum likelihood estimation approach is developed and utilized in the empirical analysis.