A variation of the Prisoners’ Dilemma problem may be playing out in the Republican primary campaign. Those elements in the party who are still seeking to prevent Trump from becoming the nominee have apparently abandoned their original tactic of attempting to coalesce support around the strongest non-Trump candidate while encouraging weaker candidates to drop out. Instead, the plan of the Never Trump movement is now to encourage all three remaining candidates to remain in the race, but act more cooperatively toward each other so that each can amass the maximum number of delegates in the states in which each is running most strongly, and thereby prevent Trump from gaining a majority in advance of the convention. Thus, Rubio and Cruz are supposed to reduce their efforts in Ohio to allow John Kasich to win there; while Kasich and Cruz pull back in Florida to allow Marco Rubio to win.
Most likely this strategy is doomed to failure if Marco Rubio, as currently projected, cannot beat Trump in Florida. But even if it remains viable for a while, there will still be strong incentives down the road for all three non-Trump candidates to defect and support a Trump candidacy, rather than to remain cooperative so that one of them (or somebody else) could instead succeed. It is not even clear that any of these three candidates views the prospect of a Trump candidacy as the worst possible outcome as opposed to the candidacy of some other candidate besides themselves. Thus the incentives to cooperate are already weaker than those presented by the classic Prisoners’ Dilemma problem. And the incentive to defect to Trump, instead of to cooperate with the Never Trump forces, becomes more powerful as the process moves toward the convention.
If Trump is short of delegates, naturally he will seek to induce one of the other candidates to drop out and support him, offering whatever inducements he needs to offer (e.g., the vice-presidency or a Cabinet position). And each of those candidates might prefer to support Trump rather than cooperate with the anti-Trump forces, especially if they cannot match Trump’s offer. Trump has already succeeded in getting two former candidates (Christie and Carson) to support him, and it would not take a genius negotiator to get Rubio or Cruz or Kasich to do the same in the right circumstances. (Cruz might be a special case since he has a history of taking positions that do not seem to result in the best outcome either for himself or the group, but for purposes of this discussion, we probably don’t need to delve into each candidate’s idiosyncrasies too deeply.)
In the highly competitive environment of this election, I predict that all of these candidates are likely to defect in some form or other, as each seeks his own best personal outcome. They will not continue to cooperate with the Never Trump movement. At the same time, each of the candidates has an incentive not to rock the boat too much; they should all want a harmonious convention and unity in support of the nominee. These powerful incentives to defect rather than assist with a strategy that might benefit one of them but would result in a bloodbath in Cleveland, make it highly likely that, barring unforeseen events, Trump will end up as the Republican nominee. I’m not saying that is an ideal outcome. But it’s hard to see any scenario in the Republican Party right now that will accomplish what the Republicans have almost always accomplished in past elections, which is to unite the various factions in the party around whomever might be their strongest candidate, whether that’s Trump or someone else.