Democratic elections as iterated prisoners dilemmas

Krugman’s column today points out what is both undeniable and yet still denied completely by the Democratic party as an institution (although not of course by many Democratic politicians and voters:

America’s democratic experiment may well be nearing its end. That’s not hyperbole; it’s obvious to anyone following the political scene. Republicans might take power legitimately; they might win through pervasive voter suppression; G.O.P. legislators might simply refuse to certify Democratic electoral votes and declare Donald Trump or his political heir the winner. However it plays out, the G.O.P. will try to ensure a permanent lock on power and do all it can to suppress dissent.

I am going to quibble here with one aspect of this analysis: the notion that Republicans “might take power legitimately.”

The problem here is essentially that identified by Herbert Marcuse in his critique of liberal tolerance: How do you maintain a social compact with people who simply reject that compact themselves whenever it becomes inconvenient, such as when they lose an election?

I would argue that a party that makes it increasingly clear that, when it wins elections in the future, it will do whatever it can to make sure it is no longer subject to electoral defeat in anything but the emptiest formalistic sense — such as for example in the same way the ruling parties in China and Russia are subject to the possibility of electoral defeat — is a party that cannot win elections “legitimately.” This is because a necessarily axiomatic feature of legitimate elections is that the winners and losers continue to abide going forward with the rules under which those elections are held, the most important being that they aren’t radically rigged in favor of a particular party or individual.

In this sense, a series of legitimate democratic elections could be thought of as an iterated prisoner’s dilemma. Leaving aside all questions of political morality, it’s still from a purely pragmatic perspective in the interest of all parties not to defect, because defection will destroy the possibility of the defecting party enjoying the benefits of future legitimate democratic elections, since defection will lead to a cycle of retaliation that will make such elections impossible.

The problem with this kind of analysis, from a purely game theoretic perspective, has always been what’s called a reverse induction paradox, which works like this:

(1) It doesn’t make sense for a party to defect an iterated game, in which continual cooperation is beneficial to everyone on net.

(2) However, it does make sense for that party to defect when the parties reach the last iteration of the game.

(3) Since it makes sense to defect in the case of (2), the recognition of this makes the iteration prior to the last iteration the same, from a strategic perspective, as the last iteration. That is, to the extent the parties recognize both the total number of iterations in the game — and of course signaling that you won’t abide by the results of any election you lose makes the first election held under these terms the last election for all practical purposes — and the fact that defection is beneficial to whoever defects during the “last” iteration, then they will defect before the last iteration. This recognition makes the next to last iteration for all practical purposes identical to the last iteration from the perspective of parties wishing to gain an advantage. And so on, all the way back to the first iteration.

The solution to this paradox is the solution to all game theoretic conundrums, which is don’t be an amoral sociopath: an option which clearly isn’t available, given the current version of the Republican party.

Since that solution isn’t available, that would seem to consul the Democrats to move to a first strike posture, in regard to not letting your opponent cheat their way to an enduring destruction of the very system that your opponent has all but publicly announced it intends to destroy.

How exactly to do that is shall we say a tricky matter, practically speaking.

And I’m not saying we won’t get our hair mussed, but it’s important to recognize that this is where we are now.