Via the magic of an alert for inbound links, I find an artblogger riffing on a recent post here who, oddly enough, brings up that old game theory classic the Snowdrift Game:

The situation of the Snowdrift game involves two drivers who are trapped on opposite sides of a snowdrift. Each has the option of staying in the car or shoveling snow to clear a path. Letting the opponent do all the work is the best option (with a pay-off of 300 used in this study), but being exploited by shoveling while the opponent sits in the car still results in a pay-off of 100. (The other two possibilities, both shoveling and both sitting, have pay-offs of 200 and 0, respectively.)

If the cost of shoveling is low compared to the benefit of getting out of the drift, it will be in your interest to shovel by yourself. Sure, the other passenger is a freeloader who shares the benefit undeservedly, but so what? If the cost of shoveling was too high for you to bear, you’d have refused to do it, letting both of you freeze there. That would be the Prisoner’s Dilemma. But if the cost of shoveling is low compared to the costs of doing nothing, then a mixed strategy will be optimal. As long as freeloaders aren’t too common, that strategy will pay off. So a population engaged in the Snowdrift game will come to a mixed proportion of shovelers and freeloaders.

Now, it struck me here that, except for the “both-sit” outcome, this is set up to be a constant-sum game: The total of utilities is 400 for every other mix of strategies. But under the circumstances described, if we bracket for the moment whatever feelings of resentment a unilateral shoveler might feel about being exploited, this doesn’t seem quite right. Rather, I’d expect the big dropoff in utility to come from having to leave the comfort in the car and shovel in the cold. Given that you’re going to shovel, obviously you’d rather it only take half as long because the other person is chipping in, but it seems plausible to think there’d be diminishing disutility to extra time spent shoveling, at least up to a point. And more generally, I’d expect there are a lot of situations like this, where the subjective costs of contributing to a collective good are front-loaded, such that having some subset of the beneficiary group bear the costs while the rest free-ride. Hence, if some friends are having drinks on a cold day, they pick one or two people make a beer run, even if sending the whole group into the cold would lighten the load for each on the way back.

Now, in the real world, we have this nice granular instrument called “money” that often renders this tractable in practice. We don’t all take turns volunteering to provide public goods; we get taxed and then particular people do the providing—the loafers chip in more for  drinks than the guy who makes the beer run. But as the relevant community scales up, so does the overhead for any kind of compensation scheme—and often there are other complications. My friend Tim Lee, following legal scholar Mark Lemley, has questioned the horror of free riding in the context of intellectual property in particular. Bracketing the question of how bad this is, it seems clear that (1) problems of enforcement obviously lead to a good deal of free riding, and (2)  there are a range of pretty overwhelming objections to a more centralized tax-and-transfer model for funding most information goods. (I assume these are obvious enough that I need not run through the litany.) If we didn’t have to worry about incentives for future creation, the overhead of vainly trying to internalize all the value of an information good and eliminate free-riding would pretty clearly diminish aggregate welfare massively.  At least in one-off interactions, there are almost certainly  lots of real world cases of efficient free-riding. Even when you add the dynamic/incentive considerations back in, the practical obstacles to charging individual consumers different prices, there’s a lot of deadweight loss to people who’d get $200-worth of value out of Photoshop but aren’t prepared to shell out the prevailing market price. Whether the dynamic considerations of undersupply of the collective good trump the greater static efficiency of the free-riding in these cases will be an empirical question: Blockbuster movies, quite probably; snowdrifts, maybe not.

The idea of “efficient free-riding” will often bump up against our intuitions about fairness. In the limiting case, an absolute prohibition on free-riding is susceptible to the familiar “leveling down” objection against strong egalitarian views. In most cases, though, there’s a fair question of whether it’s better to  have a more equal distribution of burdens, even if this means the total burden is heavier. By now this sounds familiar to the political philosophy geeks, because it’s the question of maximizing aggregate utility versus maximin—except in the context of a specific type of interaction, rather than over the whole lives of large groups of people. That distinction actually matters, though, because whatever plausibility maximin has comes from Rawls’ insistence that each person has only one life to live, and so it’s unreasonable to demand that members of the worst off group forego improvements in their position for the sake of—admittedly larger—benefits to those who are already better off on the whole. It’s not at all clear that you can apply the same kind of argument in more specific cases: The difference principle is not a fractal.

This is especially the case if occasions for efficient free-riding are sufficently commonplace that the person who bears an unfair burden in one case is the benificiary of free-riding in another. People who routinely “defect” in the context of a specific opportunity for cooperation can still “cooperate” in the aggregate if they’re playing the right mix of cooperative and free-riding strategies. What’s unfair at the microlevel might be better for everyone on net. Think, by analogy, of Kant’s “imperfect duties,” such as charity: They’re binding on everyone, but not all the time.

Obviously, this still requires people to be quasi-cooperators and not total free-riders. But if this kind of situation is sufficiently endemic, it may be a practical point in favor of sometimes thinking in the language of virtue ethics rather than looking to the rightness or wrongness of individual acts. Because an individual act of free-riding may be part of a pattern of behavior that is, on net, both efficient and fair, for certain types of cases it may be that we want to ask, not: “Has this person acted correctly in this case?” but rather “Is this person, in general, a sufficiently helpful sort?”

Possibly related: In a TED talk on or “buggy moral code,” behavioral economist Dan Ariely talks about the malleable boundaries of our moral “fudge factor”. He’s primarily interested in the idea that fairly trivial structural nudges can profoundly affect the rate at which students cheat on a certain type of test, given the opportunity. What’s also interesting, though, is that for most students, the cheating still remains quite bounded even when they’re basically given free reign to lie about their scores with no possibility of detection. Lots of people will inflate their scores a bit (and gain for themselves a few extra dollars), but most won’t take the opportunity to go all the way. Ariely suggests that we’ve got a degree of leeway in our moral strictures, within which we can deviate from the rules to benefit ourselves without seriously undermining our moral self-image. I’m loath to meta-rationalize people’s rationalizations, but maybe we’ve got this dispositions because there are a range of contexts (academic tests, in fairness, almost certainly not among them) where we’re actually all better off in the aggregate if most people “cheat” (free-ride) a little bit some of the time.