On the other hand, most epidemiological and medical studies, if not all, are statistical in one way or another, so it shouldn't be an indicator of anything. I recommend Bad Pharma [1], a very interesting book whose primary goal is to get rid of the many kinds of bias present in the current medical science. Among many interesting things, it describes a few different ways to perform medical studies, all of which are in some way statistical.
[1] http://www.amazon.co.uk/s/ref=nb_sb_noss?url=search-alias%3D...
Another way is that alcohol competes with other neurotropes, so maybe pharmaceutical companies and drug dealers would benefit. Or even makers of psychedelic music and images.
A third way to view it is that alcohol competes with other stress-removers, so massage therapy and calming music vendors would benefit.