Nice piece, Lyman. One interesting thought to add.

The map you chose for the title was of binge drinking rates in the general population. Comparing that map to the map of binge drinking rates among drinkers, I think, shows another piece of the puzzle. Within the US, variation in rates of binge drinking is clearly driven by the prevalence of people who don’t drink at all — rates of binge drinking among drinkers are much less variable and are inconsistently associated with zeal, with zeal in Mormonland and in SouthernBaptistLand seeming to have different effects.

I think this observation leads to the questions: how many people who don’t drink at all are religious T-Totalers? and how important is the social presence of religious T-Totalers in maintaining that behavior in others who may not be religious?