The concept of adolescent “invulnerability” remains pervasive in both scientific and lay circles, is used to explain adolescents’ decisions to engage in potentially harmful behavior, and is incorporated into many tobacco-related prevention and intervention programs (Weinstein 1983; Weisenberg et al. 1980; see also Reyna and Farley 2006). We turn to a review and critique of the empirical literature on the relationship between risk judgments and adolescent tobacco use.

The majority of studies testing the relationship between perceptions of risk and tobacco use have compared perceptions of tobacco-related risks between those who have and have not smoked. Although some studies find that adolescents who have smoked perceive greater smoking-related risks than those who have not smoked, others found that smokers perceive less risk. For example, Halpern-Felsher and colleagues (2004) found that adolescent smokers and those who intend to smoke estimated their chance of experiencing a smoking-related negative outcome as less likely that did nonsmokers and non-intenders. A study by Johnson and colleagues (2002) also revealed that for both high school (aged 16–18 years) and college students (aged 18–22 years), smokers saw their outcome risk as higher than that of nonsmokers.

Similar findings in a study conducted by Weinstein and colleagues (2005) concluded that adult smokers underestimated their relative risk compared to other smokers and to nonsmokers. A national survey of 6,369 people (1,245 current smokers) examined beliefs about the risks of smoking. Key questions separated samples of smokers who were asked either about their own risk or about the risk of the average smoker. More than half of current smokers thought that their own risk was only twice as high or less as that of nonsmokers. The data clearly indicate that smokers underestimate the extent to which smoking elevates lung cancer risk above that of nonsmokers (Weinstein et al. 2005). Arnett (2000) found greater optimistic bias among smokers than among nonsmokers for both adolescents and adults. Another set of studies has found that smokers, both adolescents and adults, believe that the health risks of smoking are lower for themselves than for other, same-age smokers (Weinstein 1998; Hansen and Malotte 1986; McCoy et al. 1992), but higher compared to nonsmokers (Arnett 2000).

There are six salient problems with the literature base on the relationship between risk perceptions and tobacco use. First, Slovic and colleagues (2001; 2004) have argued that studies on tobacco-related risk perception fail to consider the affective components that surround decision making (e.g., Slovic 2001; Slovic et al. 2004). In brief, the affect heuristic is thought to play a role in decisions in part through its influence on perceptions of risks and benefits. For example, if one feels good about engaging in a particular behavior, one might judge risks to be lower and in turn be more inclined toward engagement. With regard to cigarette smoking, Slovic (2004) extends this argument by stating that adolescents might not be weighing the risks and benefits in their decisions to smoke, but instead are driven by affective impulses such as enjoying the new experience or having fun with friends (Slovic et al. 2004). We discuss more about the affect heuristic later in this appendix.

A second problem with the literature base is the lack of consideration of whether smokers have or have not experienced a related positive or negative outcome. Studies focusing on the role of outcome experience in risk judgments have shown that individuals who have personally experienced a negative outcome linked to an event or risk behavior perceive the same or similar outcome as more likely to happen than do individuals without such outcome experience (e.g., Gochman 1997; Greening et al. 1996; Roe-Berning and Straker 1997; Vaughan 1993; Weinstein