Play the Game (One Flew Over the Cuckoo’s Nest)

Game Theory is a concept used in many fields including psychology, economics, and yes, political science. 

It’s a way of using models, usually with payoff amounts, to look at cooperation versus conflict scenarios.  A very popular game theory model is the Prisoner’s Dilemma, where two prisoners choose to tell on each other rather than cooperate and stay silent, even though it means more jail time.  They do this because the risks are high that the other party won’t stay silent and they don’t want to be the fool left with big jail time. 


But the concept I want to talk about in Game Theory, is the Repeated Game.  People who study the repeated game, are looking to see if behavior changes when a game is repeated.  So, in the case of the Prisoner’s Dilemma, maybe “prisoners” learn quickly that cooperating by not ratting each other out is better for them.  Now, there is a concept in the repeated game, called a trigger strategy.  The trigger strategy is strategy where some form of punishment will be used for the remainder of the game, that will reduce the benefits to both players for the rest of the game.  Maybe a cycle of, if you don’t cooperate, then I don’t cooperate, repeated over and over.  Or worst case scenario, you don’t cooperate once and I don’t cooperate for the rest of the game.


Now let me use these game theory concepts to share a few thoughts.   First, does our current party wedge politics contribute to a non-cooperative strategy?  In other words, is there so much punishment from their own party for cooperating, that politicians are forced not to cooperate?  Now let’s take a little deeper look here. We are talking about punishing individual candidates here.  You cooperate too much; the party punishes you as an Agent (see my post on Principal-Agent problems).  This is separate from the actual societal punishment for bad policy or a stagnant legislature.  In other words, cooperation may yield benefits to constituents (Principals) in the form of better policy, but yields punishment to our elected officials (Agents).  The bottom line is that there are layers of payoffs once we introduce Principal-Agent problems.


Next, are we stuck in a trigger strategy?  Have the actions by both party’s politicians in recent years poisoned any opportunity for cooperation?  More importantly, does this lack of trust in the game extend to our wider party communities, and independent communities?  If so, how do we break the trigger strategy cycle?  Is the dissatisfaction with both parties enough to raise the punishment for NOT cooperating? Or will wedge politics prevail, and mold this dissatisfaction with both parties into greater punishments for cooperating?


Ok, finally in my Positions post I talked about baseline information that we should ask for from policy positions.  So here I want to talk about the repeated game from a slightly different angle.  The repeated game can be an opportunity to make policy better, if we return to the cooperative game.  We could actually say: ok Round 1, based on our best modeling we believe that changing educational formula funding by Proposal A, will meet the financial needs for W increase in student population over the next X years, with property tax rates remaining between a Y range and we expect this new programmatic approach to increase scores by Z (yes, I know a gross simplification, just a hypothetical example).  More importantly, we can measure the trend of W and Y (meeting financial needs and tax rates) in 12 months, and Z (scores) in 18 months.  Fast forward 18 months to Round 2. Proposal A is either a failure, success, or shows mixed results.  Based on how our policy baseline information measured up, we can modify or go in the direction of our opposition’s Proposal B. 


So, the difficulties of doing this in government are:  1) The stakes are higher, thus mistakes costlier, 2) political parties have a level of hubris over “their” solutions that keeps them in a non-compromising stance, 3) we may not have the appropriate systems to measure results, 4) the timing of results is slow, and 5) the political process is slow.  Let’s put this differently, when your mobile phone maker issues an operating system update, usually they issue several patches in the following weeks that address problems they see.  Their ability to gather the information, and turn around solutions in a timely manner through several rounds of play is impressive.  We can’t begin to move government toward a more responsive repeated game strategy until we move politics back to a cooperative strategy.  The Movie Scene