Applied Economics

This article empirically assesses the role of new health information in the decision to quit smoking. Using individual level data from the Swiss Household Panel, health information is proxied by three different types of health events: physical health problems, mental disorders, and accidents. Exploiting retrospectively reported information on smoking behaviour, smoking cessation is modelled using a discrete time hazard model that also accounts for estimation problems arising from almost quasi-complete separation in the data. The empirical results yield robust evidence that general health problems increase the probability of instantaneous smoking cessation. Differentiating between the type of health event reveals that the overall effect is mainly driven by physical health events.