Esprit de corps

Ah, it's Friday. "Think Friday", around here, a day where we try to avoid scheduling meetings, and to use the day for work that needs uninterrupted thinking, or for time to think about new projects or new ideas. Friday's a day when we try to wrap up the week, finish up unfinished things, and prepare ourselves for being away from the office for a couple of days. We get ready to relax for the weekend.

Friday's also the big day for "happy hour" after work. Not so much where I work now — folks around here, at least the ones I know, just go home on Friday afternoon. Maybe I know the wrong people. But when I worked in the D.C. area, lots of people would go out after work on Friday, and spend an hour or two (and, truth be told, some spent more) at the local "spot", enjoying happy hour and schmoozing.

And now I hear on Marketplace that a study sponsored by Reason Magazine shows that happy hour makes your paycheck happy too:

Here's something even more refreshing than a gin and tonic on a hot summer day: drinkers make more money — a lot more money — than teetotalers. Indeed, a study we just sponsored shows male drinkers make 19 percent more than their abstemious full-time counterparts. Among women, drinkers make 23 percent more.

The reason is obvious, and is related to why playing golf or going to the right gym can be useful to one's career: esprit de corps, networking, being "in the know" and being known.

For better or worse, social connections are important to one's career; that's always been true, and nothing's different now. Clearly, if going out for happy hour relates to making those social connections, it will translate into more job opportunities, and better earnings. The question, though, is whether it's actually the drinking that does this, rather than simply being in the social envoronment. The study finds that the drinking itself (though not immoderately) is the primary factor.

But the study is seriously flawed. For one thing, only about 1% of the particpants participate in the social setting but do not drink alcohol. That's well less than the margin of error, so it's impossible to tell whether someone who attends happy hour but drinks club soda, ginger ale, or non-alcoholic beer would see similar benefits. Yet they conclude that drinking gets one 10% more money, and "bar hopping" (going to a bar at least once a month) adds another 7% to that.

Looking at the numbers would make you think you ought to be a married Jewish drinker from New England, if you want to maximize your earnings. Worst would be a single black atheist non-drinker from the "east south central" region. But it's not stated how the participants are distributed with respect to religion, race, marital status, or region, so it's impossible to draw reliable conclusions from the data.

Most disturbing is the fact that, while they include age (and age²) as a variable in their tables, they do not do any correlation between age and habits with respect to drinking or socializing. Age clearly correlates strongly with income — senior workers usually make a good deal more than junior ones (in fact, I'm suspicious of how low the effect of age seems to be in their tables) — and if, for example, age also correlates strongly with whether one goes to happy hour (say, younger workers drink more and socialize more), then it's quite likely that the conclusions drawn here are badly skewed.

I'll point out that this is a commissioned study, and hasn't been subjected to peer review. Further, the Reason Foundation, which sponsored it, has a specific Libertarian agenda, and wants to find reasons to discourage policies against drinking. That's supported by this, in the "conclusions":

Our analysis leads to a number of policy implications. Most importantly, restrictions on drinking are likely to have harmful economic effects.

I want to discourage policies against drinking, too, when it comes to that! But this study doesn't have much credibility to that end. It supports some obvious conclusions — that esprit de corps is important — but the evidence it provides for anything else is spare and misleading.