Q & A on the Personal Costs of Smoking

Last week, I wrote up a new NBER paper on the personal costs of smoking which has generated a bit of interest around the blogosphere.

W. Kip Viscusi, a professor of law and economics at Vanderbilt University, who along with Joni Hersch, coauthored the paper, put the personal cost of smoking at $222-per-pack for male smokers and $94-per-pack for female smokers.

Viscusi has been one of leading researchers in attaching a dollar figure on the value of a human life. (Go here for a recent Richmond Federal Reserve interview with Viscusi.)

I spoke with him recently about the smoking paper. Here is a slightly Deborah Solomon-ized version of the interview:

Q. What is the best way to think about the concept of mortality cost and your finding that a pack of cigarette costs $222 for a male smoker?

In this case, mortality cost is the personal risk to smokers as opposed to the ramifications for society. This is actually the main consequence of smoking. By far the largest cost of smoking is to smokers themselves. If smokers were completely ignorant of the risks they face from smoking, then you would have to pay male smokers just over $200-per-pack for them to be willing to incur risks just as big.

In your paper you write that other attempts to put a dollar amount on the personal costs to smokers put the figure in the $20 range. What is different about your approach?

In the fairness to the previous attempts, they were just really back-of-the-envelope calculations. Our estimates differ in two main ways from what came before.

First, the previous studies all assumed that when smokers lose years off their life, that occurs at the end of a normal life-span, so that if you're a male smoker who would have lived to 75 without smoking, now you'll only live to 71.

Under our approach, we take into account that smoking doesn't necessarily kill you at the end of life. It could kill you in your 40's, 50's, 60's, and so on. And if it kills you before what would normally be the end of your life, then you're going to lose a whole lot more of your life. And these deaths happen more immediately, and immediate losses are more valued than losses further down the pike.

The second way we differ is in terms of the downward value we put on the deaths. The past estimates had used my value of statistical life estimates which they then divided by the average number of years of life to get a value of each year of life of $100,000.

What we did instead was say 'if people valued their life at a constant value per year then the value of your life would peak at birth and decline steadily thereafter as your life expectancy declined.' But it may be that as you become richer as you get older, the value of your life could go up. So we analyzed the value of life for workers who were smokers and nonsmokers by age to come up with a value and it turned out, surprisingly, that people's value of life remains quite high even in their 50's. If you're 60 years old, you may value your life in terms of your willingness to incur risks more than a 20 year old. That's not surprising because 20-year-olds are more likely to take on risky jobs than 60-year-olds.

What are the differences between smokers' cost to themselves and smokers' cost to society?

The other study I've done is looking at the financial ramifications to smoking for the rest of us. These include higher medical costs on the one hand, but lower social security, pension, and nursing home costs on the other hand because smokers die sooner. On balance if you put those together, smokers don't cost us money, but save society $0.32 per pack.

Are there any policy implications for your findings?
I'm sure a lot of people would like to run with these in various different directions. I've also found that smokers overestimate the life-expectancy loss associated with smoking. So, if that's the case, then they would've already taken [our findings] into account.