The relationship of scientific research, industry advocacy, government action, and public behavior on smoking is frequently cited in discussion and debate, but I have actually seen little empirical research that backs up the various claims that are often made on this issue. Two Harvard economists, David Cutler and Edward Glaser, have a paper available at the National Bureau of Economic Research that raises some interesting questions about the role of science and advocacy on the smoking issue.

A few points it seems can be taken as fact:

1. smoking is a health hazard
2. The smoking industry for many years engaged in an active campaign to suggest that smoking is not a health hazard
3. Smoking is addictive
4. Smoking rates, and the change in those rates, are different in different parts of the work

Cutler and Glaeser seek to untangle some aspects of this last point by asking “Why do Europeans Smoke More Than Americans?” The answer that Cutler and Glaeser provide it turns out is complicated, and in my view fundamentally flawed. Here is what they claim in their abstract:

While Americans are less healthy than Europeans along some dimensions (like obesity), Americans are significantly less likely to smoke than their European counterparts. This difference emerged in the 1970s and it is biggest among the most educated. The puzzle becomes larger once we account for cigarette prices and anti-smoking regulations, which are both higher in Europe. There is a nonmonotonic relationship between smoking and income; among richer countries and people, higher incomes are associated with less smoking. This can account for about one-fifth of the U.S./Europe difference. Almost one-half of the smoking difference appears to be the result of differences in beliefs about the health effects of smoking; Europeans are generally less likely to think that cigarette smoking is harmful.

Their findings are interesting, among them:

. . . price differences cannot explain why Americans smoke less than Europeans.

While our conclusions thus need to be interpreted with some care, the U.S. does not appear to tax or regulate tobacco consumption particularly highly, making these explanations unlikely to account for the lower smoking rate in the US.

Given that our model suggests a possible non-monotonic relationship between income and cigarette consumption, the lack of a clear consensus on the income elasticity of smoking is not so surprising.

So price, regulation, and income cannot explain the differences in smoking between the U.S. and Europe. Cutler and Glaeser next turn to beliefs about smoking, for which they begin by noting:

The U.S. has one of the highest rates of believing that smoking is harmful; 91 percent of Americans report believing that smoking causes cancer. Given the high proportion of Americans that believe in UFOs and the literal truth of the bible, this must represent one of the most remarkable instances of the penetration of scientific results in the country. Beliefs about the cancer-causing role of cigarettes in some European countries, like Finland, Greece, Norway, and Portugal, are almost identical to those in the U.S., but in other places beliefs are far weaker. For example in Germany only 73 percent of respondents said that they believed that smoking causes cancer.

An interesting aside, if this recent poll is to be believed more people in Europe believe that global warming is a serious problem than believe that smoking causes cancer. By contrast, in the U.S., the opposite is the case. I’m not sure what to make of this, but the answer is surely tied up in the question Cutler and Glaeser are trying to unravel.

Cutler and Glaser then go out on what I think is thin ice when they assemble an argument based on several lines of reasoning to conclude:

On the whole, our evidence suggests that differences in beliefs are the most important factor explaining the differences in smoking between the U.S. and Europe.

One of the most important lines of evidence is presented in their Figure 11, which I reproduce below.

View image

Figure 11 purports to show a relationship between beliefs about smoking and smoking rates. However, the relationship is entirely a function of excluding Greece from the analysis. Cutler and Glaeser argue earlier in their paper why Greece is excluded from the regressions:

When we examine bivariate relationships between smoking and other factors (prices, regulations, or beliefs), it is important to have a relatively homogeneous sample of countries by income. Within Europe, the major income outlier is Greece, with a per capita income that is 60 percent below the European average ($10,607 in Greece versus $25,858 in Europe in 2000) and 25 percent below the next lowest country (Spain, at $14,138). For this reason, we omit Greece from many of our regressions, though we present raw data for Greece in the tables and show the country in the figures.

I find this justification wholly unsatisfactory. Greece, despite its lower income has the highest rate of belief that smoking causes cancer. Greece is also highly regulated, and does not have unusually cheap cigarettes. In short, its inclusion would certainly muddy the analysis. Excluding it, somewhat arbitrarily in my view, does not add any confidence that Cutler and Glaeser have an adequate explanation for what is going on.

What I get from this paper is that in fact, we don’t yet have an adequate explanation for why we see large differences in smoking between the U.S. and Europe. Consequently, if we can’t explain the role of scientific information in smoking, it is safe to say that our understanding of the role of scientific information in societal outcomes on other more complex issues remains frustratingly undeveloped.

In part II I will take a look at trends in U.S. smoking presented by Cutler and Glaeser and suggest a similar lack of understanding.