Dan Lakeland writes:
Apropos your recent posting of the Churchill/Roosevelt poster, there has been a bit of a controversy over the effect of smoking bans in terms of heart attack rates. Recent bans in the UK have given researchers some plausible “experiments” to study the effect on a larger scale than the famous “Helena Montana” study. For example, this.
On the other hand, when looking for info about this to follow up your poster I found a variety of usually rather obviously biased articles such as this one. But that’s no reason to ignore a point of view if it can be backed up by data. The second link at least attempts (poorly) to display some data which suggests that an existing downward trend could be responsible for the reductions, and if poorly done the statistical research could have missed this.
Have you looked at the statistical methodology of any smoking ban studies? it seems like an area ripe for Bayesian modeling, and could be a subject along the lines of the fertility and beauty more girls/more boys research that you recently meta-analyzed.
Yes, I imagine that some people have looked into this. I would guess that a smoking ban would reduce smoking and thus save lives, but of course it would be good to see some evidence.
Smoking behavior is a funny thing: It can be really hard to quit, and I’ve been told that the various anti-smoking programs out there really don’t work. It’s really hard to make a dent in smoking rates by working with smokers one at a time. On the other hand, rates of smoking vary a huge amount between countries and even between U.S. states:
And smoking bans might work too. Thus, smoking appears to be an individual behavior that is best altered through societal changes.