smosa/adam
/delights
/game-theory-and-seemingly-bad-decisions
/


In politics, we follow the happenings in our government in sound bites and articles. These media are carefully calibrated to appeal to cognitive ease and often carry with them the political bias of the institution or author.

On the one hand, we are pretty good at understanding that politics are very complicated. On the other hand, we latch on to these sound bites and slogans. We embrace oversimplification.

When we try to move in the other direction to say the issue is more complicated, its very easy for us to appear as if we are justifying something we are merely looking to explore with more clarity.

Take the 2008 financial crisis in the United States. We bailed out the banks but not the home owners. The decision was criticized across the two party system and was oddly the policy of a democratic administration. Morally, it's probably not home owners and not the banks who deserve the bail out, and the politicians stated their reasons for this citing protecting a major collapse of the economy among other factors. This rhetoric only infuriated those concerned because pushing people out of their homes and into massive debt is probably not good for the economy either.

We can say for simplicity that we should have bailed out the homeowners, not the banks. That was the morally right thing to do. If you don't agree, that's fine too because what we end up deciding was morally right will now be separated from what we in the shoes of a politician would have acutally done to uphold our moral code. But are those different things?

Yes. Or at least they can be for the same reason giving up a pawn in Chess does not work in your favor unless it's a strategic means to gain some greater tactical advantage. If this sounds like we're making a moral argument about means and ends, you don't have to worry about that either. This is not going to be a moral argument at all. We're only interested in what actions one could take in the game of politics.

An explanation of the Nash equilibrium that is more exhaustive than the one I will provide here is in the Nash equilibrium article on Wikipedia. For a working definition, imagine we have two players that are playing the game with a given strategy. That strategy is contingent on their knowledge of what the best strategy is for their opponent, knowing their opponent will use it. Each player's strategy became their best strategy because of the equilibrium they find themselves in the moment they are mutually unable to improve their strategy given their opponent's existing strategy. If this is true for both players, you're in a Nash equilibrium. The prisoner's dilemma is a game where the players are in a Nash equilibrium.

I can't say if the Nash equilibrium is specifically at play in the 2008 crisis, but isn't it something to think about that each player is doing some thing that both gives them the most optimal chance of success in their goals while at the same time, not necessarily being a direct action of the goal itself.

These nuances should be somewhat familiar. Can we guess why someone would amicably quit a job they need and love if they are faced with the strategy of their employer to fire them? This probably isn't a Nash equilibrium but the dynamics of it all are simple enough for us to see why the action in someone's best interests ends up not being the result that is in their best interests. They wanted to stay employed, so self-terminating their own employment was the best way for them to continue being employable.

So would you actually bail out the home owners and not the banks? We sure want to, but how does this work as a function of competing strategies? There is an impact on letting the banks fail. We have to remember that is a fact even if it runs against what we think homeowners and banks each deserver. The answer could be that bailing out the homeowners is the best thing for the economy despite the damage the collapse of the banks would cause. If that is true, we're playing to our best strategy and getting what we identify is the ideal solution.

But what if it objectively is not? Again, we're just doing a thought experiment here. What if the damage to the economy by not bailing out the banks paradoxically puts the average American in a worse position than the alternative?

What if there are other strategies we aren't even considering? Is there an angle that is political but very real in its consequences? As little as we want to care about a corporation's feeling of how they are being treated in a country, the actions they take based on their need to optimize their strategy could cause great damage to the economy. A damaged economy is ultimately not going to be good for everyday people. That is an uncomfortable fact to wrangle with if rewarding bad actors is a principle means of supporting that relationship. So what do you do?