In an old post, Chris Auld attacked the "zombie argument" that healthy lifestyles substantially decrease demand on the health care system. As he put it:
All people -- and I do not mean to shock anyone -- die some time, even including people who live very healthy lifestyles. Preventing someone from dying of a smoking-related illness only means that they will die of a non-smoking related illness. The effect of smoking on lifecycle health care costs is the difference between costs which are incurred if the person smokes and the costs which would be incurred if the person doesn't smoke. Whether improvements in lifestyle increase or decrease lifetime health care costs depends in a complicated manner on how a healthy lifestyle affects length of life and health care costs at any given age. Whether smoking or other unhealthy behaviors increase or decrease health care costs is an empirical question.
One answer to that empirical question is given below the fold:
There is a quadratic relationship between the percentage of males in a country who smoke and per capita health care spending. The relationship is statistically significant, as is shown in the table on the right.
The moral of the story? Don't conclude anything from regressions on country-level data.
The problem with analyses like this one is shown in the picture below. There is a very strong positive relationship between a country's income and the amount of money that is spent on health care: as income increases, health care spending increases. There is also a relationship - though a more complex one - between income and smoking. The poorest of the poor cannot afford cigarettes. As income increases, smoking rates increase. But in the most affluent countries, rates of smoking are lower again.
Even if smoking has only a fairly small impact on health care expenditures, the fact that both smoking and health care spending are so strongly related to income means that the two will appear to be strongly related in cross-country data. Desperately poor countries such as Eritrea (ERI), the Democratic Republic of the Congo (COD) and Niger (NER) have low smoking rates (because people can't afford tobacco) and low rates of health care spending (because people can't afford health care). People in middle income countries like Indonesia (IND), Papua New Guinea (PNG), Somoa (WSM) and Russia (RUS) are rich enough to afford a lot of cigarettes and a moderate amount of health care (other confounding factors: a climate suitable for growing tobacco, and cultural prohibitions on other intoxicants). Finally, the very rich countries, like the US (USA), Italy (ITA) and Canada (CAN), can afford to spend a great deal of money on health care, but have lower rates of smoking.
It's called spurious correlation. It's a good reason to stay away from cross-country regressions like the one in this post.
Note: I've closed comments to this entry as it was attracting too much spam.