Academic journal article Contemporary Economic Policy

Perceived Arsenic-Related Mortality Risks for Smokers and Non-Smokers

Academic journal article Contemporary Economic Policy

Perceived Arsenic-Related Mortality Risks for Smokers and Non-Smokers

Article excerpt


Arsenic in drinking water can cause lung and bladder cancer and increase the risk of dying from either disease. The exact relationship between arsenic mortality risk and concentration levels is difficult to estimate with certainty, but the best current estimate places the risk for non-smokers at approximately 1 in 100 for U.S. households with high arsenic concentrations (50 parts per billion [ppb] (1) or more) in their drinking water (NRC 2001). It is well known that smoking greatly increases the risk of dying from lung cancer (Samet 2001), but the mortality risk of lung or bladder cancer associated with exposure to high concentrations of arsenic is believed to double when the subject also smokes (NRC 2001; U.S. Environmental Protection Agency 2000). (2) In this paper, we examine perceptions and averting behavior for the conditional mortality risk of drinking arsenic-contaminated water given that a subject smokes. In particular, we (1) explore the connection between perceived arsenic risk perceptions and smoking behavior and (2) examine the link between smoking status, exposure to arsenic in drinking water, perceived risk, and the decision to remove arsenic from drinking water.

Smokers' risk preferences and perceptions have been a topic of considerable interest for at least 20 years (see, e.g., Slovic 2001; Viscusi 1990, 2002). Past research shows that smokers and non-smokers tend to display different risk preferences as evidenced by choices made in the face of mortality risk (Hersch and Viscusi 1998; Khwaja, Sloan, and Salm 2006; Viscusi and Hersch 2001). These studies find that although smokers may differ from non-smokers in their preferences and/or estimated probabilities for risk, these authors caution against concluding that smokers have lower estimates of perceived risks than non-smokers. Our study makes two important contributions to this literature. First, we examine how smokers and non-smokers perceive risks of exposure to arsenic, which inflates the risk of lung and bladder cancer in smokers. Studies that address the perception of conditional mortality risks are uncommon and we are not aware of any studies that elicit the conditional lung-cancer related risks associated with ingesting arsenic, given that a subject smokes. We find that on average, subjects underestimate risks from ingesting arsenic, even after reading an information booklet about the science of these risks. Still, we show that smokers, ex-smokers, and people who have never smoked understand the relative risks associated with arsenic exposure and smoking.

Second, perceived arsenic/smoking risks are used to explain differential water-treatment behavior on the part of smokers and non-smokers. We find that although smokers perceive greater risks from arsenic contamination than non-smokers, smokers are less likely to remove arsenic from their drinking water than non-smokers. When smokers--with greater perceived risk--fail to treat their drinking water, this may indeed provide support for the theories that higher rates of time preferences, or less health-risk aversion, help explain individuals' decisions to smoke. To proceed, the paper provides some background information on smoking, and risk perception in general, followed by a simple presentation of the underlying theory for our estimating equations. The fourth section briefly discusses the data, and the fifth and sixth split the discussion of the results into the perceived risk models, followed by the water-treatment model results, respectively. The final section offers some conclusions.


A. Is Smoking Rational? Time Discounting and Risk Preferences

Rational addiction models of smoking counter the contention of Weinstein (1999) and other psychologists that smoking is inherently irrational and/or that smokers do not accurately perceive smoking-related risks. We note that a "rational" decision does not mean the decision results in something that is always inherently good for an individual, merely that she or he took in the best available information at the time, weighed costs and benefits of various actions, and chose an action that provided the highest perceived net benefit. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.