Smoking, Expectations, and Health: A Dynamic Stochastic Model of Lifetime Smoking Behavior
构建了一个动态随机模型，分析个体如何通过学习健康风险来调整吸烟行为。模拟发现，重复体检中的心血管生物标志物信息对吸烟行为影响不大，原因包括信号噪声大、对健康影响小以及成瘾性。未建模的选择偏差可能高估吸烟对预期寿命的影响。
I estimate a dynamic, stochastic model of smoking, expectations, and health that makes explicit channels through which an individual may learn about the health risks of smoking. Simulations of the structural model suggest that cardiovascular biomarker information at repeated health exams does not significantly alter smoking behavior because (a) signals of biomarker health are noisy within individuals, (b) the influence of biomarkers on major health outcomes is small, and (c) cigarette smoking is addictive. This paper also presents evidence of selection in smoking that, when not modeled, may cause an overstatement of the effect of smoking on expected longevity.