Cigarettes Taxes, Smoking Bans, and Smoking Behavior

Many policies have been implemented to reduce the negative health effects of smoking on both smokers and non-smokers. This dissertation examines the effects taxes, smoking bans, and increased availability of cessation aids on smoking behaviors and health outcomes. First, I estimate the effects of two common policy approaches to reduce smoking: cigarette taxes and smoking bans. I find that taxes and bans increase quit attempts and quits, and decrease overall smoking prevalence. In addition, I find that taxes and bans mitigate each others' effects when they are enacted together, implying they are substitutes. Second, using variation in the timing of local bans, I find that comprehensive local bans lead to decreases in smoking and reported exposure to smoke. In contrast, I find that less restrictive bans do not decrease smoking or exposure, and instead increase the number of cigarettes smoked, overall and at home. Thus, the health effects are ambiguous. To determine whether smoking bans improve health outcomes, I use hospital discharge data from California to estimate the effects on hospitalizations for heart attack and stroke in adults, and asthma in children. I find that bans improve health outcomes for adults, but find conflicting results for children. In addition, I find that the effect of bans differs by population density, and by the type of ban enacted. Third, I analyze another policy that altered the costs and benefits of smoking: the switch of Nicotine Replacement Therapy (NRT) products from prescription to overthe-counter (OTC) status in 1996. I examine how OTC availability, which decreased the cost of obtaining NRT and the cost of quitting, affected NRT product use and quit attempts. I find this change did not increase quit attempts but did increase NRT use, indicating that it altered how, not whether, smokers attempted to quit. I also find that the use of NRT products significantly increases quit success, with and without the use of an instrument for NRT use. This effect is larger than estimates without an instrument, indicating that those who will have more difficulty quitting are more likely to use NRT products.

Note: Abstract extracted from PDF text