I’ve been thinking some more about the Bar Flight game. To review, there are two phenomena to be explained: (1) When two roughly equivalent bars are located a short distance from each other, why is one bar crowded while the other is virtually empty? (2) Given the first phenomenon, what would cause a shift in popularity so that the formerly empty bar becomes crowded and vice versa?
The first phenomenon is easily explained. The primary reason people go to bars is they want to be around other people. As a result, bars are subject to network externalities (i.e., a consumer’s satisfaction from a good rises with the number of other consumers using the good). That means consumers are playing a kind of coordination game, in which the object is for everyone to end up at the same bar. The situation is closely akin to the Sides of the Road game, in which drivers decide to drive on the left or the right; it really doesn’t matter which side they choose, as long as they all choose the same side.
The second phenomenon is more difficult to explain, given the “lock in” effect associated with many coordination games. In my first post on the subject, I attributed the shift in crowds to an indeterminacy arising from crowds large enough to fill both bars (on weekend nights). In my second post, I presented another explanation: that some customers, “the hip,” want to socialize with each other but not with another group of customers, “the unhip.” But the latter do wish to socialize with the former. The result is a kind of “chasing” game, which I likened to Dr. Seuss’s tale of the Sneetches.
But now a third explanation has occurred to me, which seems more likely than the other two. Like the Sneetches game, it relies on the existence of two customer groups – not the hip and unhip, but women and men. For the most part, men want to be where there are women, and women want to be where there are men. If that were the whole story, we’d have a simple coordination game. What throws a wrench in the works is that women only want to be around men up to a point. For men, there’s no such thing as too many women. But for women, that’s not the case. When the ratio of men to women gets too high, many women start to feel crowded, perhaps even threatened. Eventually, the gains from coordination are swamped by the losses from fighting off gangs of men, and they start to leave (or go somewhere else next time).
The next step is an unraveling effect. As some women leave, the male-to-female ratio rises for the remaining women, putting pressure on them to leave as well. And then the ratio gets even worse. Eventually, the men find themselves in what is sometimes called a “sausage fest.” They eventually leave as well, hoping to find women at the other bar.
The outcome is a naturally recurring cycle. The cycle may occur within a single night or over a much longer period. In the early stages of the cycle, we have a coordination game, with both men and women hoping to visit the same bar. But in the late stages of the cycle, women gradually desert the bar for better (less threatening) climes. They begin to congregate in another bar, and the cycle begins anew. With this hypothesis, we can explain a third phenomenon as well: the velvet ropes and selective bouncers in front of the more popular bars, which may represent the management’s attempt to short-circuit the cycle by limiting the influx of men.