An empirical assessment of effectiveness of the US tobacco control policies: a smoothed instrumental variables quantile regression approach

A sound understanding of the potency of tobacco control policies is key to tobacco prevention. This study exploits a Smoothed Instrumental Variables Quantile Regression estimator to gauge the effectiveness of these policies while addressing major methodological and data limitations plaguing the previous literature. Specifically, smoke-free indoor air laws and tobacco control expenditures are examined in a single framework, which has the promise of accounting for potential complementarities thereof. Further, endogeneity of price (a proxy for tax policy) and other tobacco control policies is addressed through a unique set of instruments while allowing for differential impacts across the conditional distribution of cigarette consumption. Finally, our use of the nationally representative individual-level price and consumption data is essential to precise estimation of price elasticities and policy effects. Results indicate that ignoring price and policy endogeneity leads to inconsistent estimates. Further, tobacco control expenditures appear to be effective only for relatively more addicted smokers. Meanwhile, state-level smoke-free indoor laws, whose primary goal is to reduce exposure to second-hand smoke, do not affect cigarette use among smokers. In contrast, tax policy appears to be most potent for less addicted individuals. Therefore, optimal policy responses should combine tobacco control expenditures with sin taxes.