Why Smokers Still Smoke

GIVEN how hazardous it is to their health, why do smokers continue to smoke?

It’s not that they are all hopeless addicts. Many smokers are capable of quitting.

It’s not that they are ignorant. Studies show that smokers are at least as informed as nonsmokers about the risks of smoking — and possibly more informed.

You might suspect, then, that smokers tend to be risk takers by nature. And some evidence suggests that smokers do take more risks than nonsmokers: they are more often involved in traffic accidents, less likely to wear seat belts and more likely to engage in risky sexual behavior. Women who smoke even have mammograms less frequently than their nonsmoking counterparts.

But we don’t believe that smokers have a greater tolerance for risk. As we argue in a study published this month in the journal PLoS One, the personality trait that distinguishes smokers from nonsmokers is their relative inability to delay satisfaction and respect long-term considerations (like their health). In other words: it’s their poor self-control.

Key to our study is a card game known as the Iowa gambling task, commonly used to measure risk taking. You are presented with four decks of cards. Each card reveals a financial outcome: “You won $100,” or “You lost $250,” for example. Your task is to keep picking cards from whichever decks you would like, trying to make as much money as you can. The twist is that the decks have different payoff distributions: two of them offer higher risk (cards with larger gains and larger losses) but long-term overall losses; the other two decks offer lower risk (smaller gains and smaller loses) but long-term gains.

Most participants, after selecting several dozen cards from the various decks, learn to stick with the two “good” decks and end up faring well. Studies have shown, however, that people with brain lesions affecting decision making tend to favor the “bad,” riskier decks and fare poorly. The same is true of chronic cocaine and cannabis users. But when smokers and nonsmokers perform the Iowa gambling task there are no significant differences between the two groups, as the psychologist Carl Lejuez and his colleagues have demonstrated. This strongly suggests that smokers are not, in fact, especially tolerant of risk.

So what accounts for smokers’ risky-looking behavior? Our contention is that smokers exhibit poor self-control in the face of immediate temptation — which can look like a willingness to assume risk. (For instance, you might choose to have sex without a condom not because you are comfortable with the risk but because you are too weak-willed to bother with the inconvenience.)

To test our hypothesis we took 100 research participants (smokers and nonsmokers) and had them perform a modified version of the Iowa gambling task. We focused on a subtle difference between the two “bad” decks. One offers payoffs that commonly result in immediate satisfaction — it produces a gain of $100 in 9 out of 10 draws — but in 1 of 10 draws it produces a disaster: losing $1,250. The other bad deck is less immediately satisfying — the losses are smaller ($250) but more frequent, occurring about every second draw. Thus both decks are bad overall but the first is more satisfying in the short run.

In addition, in our modified version, every time a participant selected a card from one of the four decks, we showed the outcomes that would have resulted from selecting from the other three decks. This increased the salience of the immediately satisfying deck, since when choosing any other deck, 9 times out of 10 the player could see that this deck produced the best outcome.

We found that as the game progressed, smokers selected this deck about 1.5 times more than nonsmokers. They could not resist the short-term pleasure despite the long-term disaster.

(We also had our research subjects play another version of the game in which the good and bad decks differed in terms of risk but not in the immediacy of their satisfying outcomes. This change wiped out the differences between smokers and nonsmokers, confirming the Lejuez team’s finding that smokers are not more tolerant of risk.)

Our finding is admittedly nuanced: it is not risk taking per se that drives smokers’ risky behavior; it’s a weakness for activities that are profitable most of the time yet hazardous eventually.

This distinction may suggest strategies for fighting smoking. For example, hospitals and universities have started to ban smoking not only inside buildings but also at their perimeter; while originally proposed to address secondary smoking hazards, this may also have the benefit of imposing an additional short-term inconvenience on smoking. If we are right that smokers tend to have poorer self-control, such manipulations may be used to help sustain their willpower.

Eyal Ert is a senior lecturer of behavioral and management sciences at the Hebrew University of Jerusalem. Eldad Yechiam is an associate professor of behavioral sciences at the Technion-Israel Institute of Technology.