Tyler Cowen makes an interesting point about duels: they may have served to facilitate conflict resolution. Counterintuitive, but true. Think of nemeses as being engaged in a two-stage game. In the first stage of the game, they try to negotiate to a peaceful settlement of their differences. In the second stage they fight a duel, and this stage is only reached if the first stage results in a stalemate instead of a resolution. If the second stage occurs, both parties can anticipate a high expected cost. (This is true even if you’re the superior gunman. Even if there’s a 90% chance you’ll win, a 10% chance of death or serious injury is nothing to scoff at.) As a result, both parties have a strong incentive to reach a mutually agreeable settlement in the first stage of the game.
The incentive structure is similar to that created by trial by ordeal. In the so-called Dark Ages, some European societies resolved disputes by having the parties engage in an ordeal that created a high likelihood of harm to both – for instance, walking across hot coals and seeing whose burns healed faster, or dipping your right arms in boiling water and seeing who got more blisters. Losing in a trial by ordeal was interpreted as a sign from God that you were in the wrong. This might seem a random and haphazard form of dispute resolution, and it was. But it also served a function akin to that of dueling, as it gave both parties a powerful reason to settle their dispute privately before trial. (I first heard this analysis from John Hasnas at a seminar about 12 years ago.)
The same analysis also throws some light on why we should or shouldn’t have conventions of war, such as the prohibition on using poisoned arrows (in Homer’s time) or torturing prisoners (in modern times). Mark Kleiman, in a blog post I intended to comment on weeks ago, makes a sensible argument in favor of such conventions: that nations engaged in war are trapped in a kind of prisoners’ dilemma. Each nation can gain a slight advantage over the other by choosing to use poisoned arrows; but if both use poisoned arrows, then neither side gains any particular advantage, and both suffer greater losses than if both had abstained from using poison. The conventions of war represent an attempt to escape the prisoners’ dilemma by restricting the allowable set of actions in wartime (perhaps by inculcating a notion of “honor”). Mark’s argument is correct but incomplete, because it treats the existence of a state of war as a forgone conclusion. We might instead treat the war as the second stage of a two-stage game, as with dueling and trial by ordeal. The more costly is war to the nations involved, the greater incentive they have not to fight in the first place.
Just to be clear, I’m not arguing for relaxing the Geneva Convention (any more than I’m arguing for the return of trial by ordeal). There’s a trade-off involved in all three situations just described, and the trade-off can go either way. On the one hand, it’s good for people to have an incentive to resolve disputes without resorting to violence or legal action. On the other hand, some disputes will never be resolved without violence or legal action, and in those cases we’d like to minimize the damage. If the likelihood of a peaceful resolution is highly responsive (in economic terms, elastic) to the expected costs of violence or legal action, then a high-cost second-stage game makes sense. But if the likelihood of a peaceful resolution is not very responsive to the expected costs of violence or legal action, then high-cost second-stage games are just wasteful and destructive. The demise of dueling, end of trial by ordeal, and emergence of the Geneva Convention might reflect a judgment that parties to disputes are insufficiently responsive the expected costs of conflict.