IS ADDICTION RATIONAL THEORY AND EVIDENCe成瘾是合理的理论和证据精品.pdfVIP

IS ADDICTION “RATIONAL”? THEORY AND EVIDENCE* ¨ JONATHAN GRUBER AND BOTOND KOSZEGI This paper makes two contributions to the modeling of addiction. First, we provide new and convincing evidence that smokers are forward-looking in their smoking decisions, using state excise tax increases that have been legislatively enacted but are not yet effective, and monthly data on consumption. Second, we recognize the strong evidence that preferences with respect to smoking are time inconsistent, with individuals both not recognizing the true difficulty of quitting and searching for self-control devices to help them quit. We develop a new model of addictive behavior that takes as its starting point the standard “rational addiction” model, but incorporates time-inconsistent preferences. This model also exhibits forward-looking behavior, but it has strikingly different normative im- plications; in this case optimal government policy should depend not only on the externalities that smokers impose on others but also on the “internalities” im- posed by smokers on themselves. We estimate that the optimal tax per pack of cigarettes should be at least one dollar higher under our formulation than in the rational addiction case. I. INTRODUCTION Many of the habits that pervade everyday life can be properly described as addictive. While the degree of addictiveness varies from activity to activity and person to person, habits such as smoking, drinking, eating, and a host of others often meet the two conditions required for addiction: reinforcement, in that the more you partake of the activity, the more you want to partake; and tolerance, in that the more that you partake of the activity, the lower your future utility given the amount of future consumption [Becker and Murphy 1988]. The importance of ad