Last week I was in Italy teaching a short course and speaking at a conference. At the conference, I attended a session where the author described an an experiment on alcohol warning labels. He had people choose between different bottles of wine that had different warning labels.
I thought this was a bit of a strange experiment because once you've seen one bottle with a warning label, doesn't it tell you something about all the bottles? When I voiced this concern, my friend Maurizio Canavari pointed out that in Italy, different cigarette packages have different warning labels (apparently determined at random).
He sent me this picture yesterday, which reminds me of the joke he told me after the session. A man walks into a tobacco shop and asks for a pack. On his way out, he notices the warning label on the pack says that smoking may cause problems in the bedroom (e.g., see the above label "Il fumo riduce la fertilita"). He goes back in and hands the pack back to the shop owner and says: I'll take the one that just kills you.
Seriously, I wonder about the effectiveness of spreading information out over multiple packs vs. trying to cram it all on one. And, I do wonder if people are more/less likely to pick packs with certain labels despite the fact that the labels warn about smoking in general and not about the effects of one particular pack or brand over another.