Game Theory in the Decision Making of Donald Trump
Throughout his presidency, Donald Trump has made many unpredictable decisions. Many people wonder whether or not he is intentionally unpredictable, since, from a game theory standpoint, it is beneficial to have a level of randomness.
The article, “Donald Trump and game theory: Expert says no brilliant 3-D chess is involved” by Chauncey Devega does a question and answer about Trumps tactics with NYU professor Steven Brams. He says that Trump’s unpredictability is based on the game theory idea of mixed strategies. In politics, there is almost never one dominant strategy, so there is always some sense of randomness that must be accounted for. Trump does not know what his opponents are going to do, so he must choose a strategy and hope it works out best.
Another issue with Trump’s decision-making is that he plays a zero-sum game. Either he wins completely and his opponent gets nothing, or it is the exact opposite, as in the attacker-defender game that we learned about in class. In this game, either the attacker or the defender wins, but not both. There can not be a split in the winnings, which is why with Trump we saw situations such as the talks with Kim Jong Un and North Korea and the current crisis in Iran.
Brams goes on to say that this strategy is very successful in a game such as poker, but it does not bode well in politics. Trump’s unpredictability causes even our own allies not to be able to trust him. He only looks into the short-term outcomes of his decisions instead of looking ahead to the impacts his decisions could make. An example of this is him pulling the United States out of the Paris Agreement. In game theory, there are sometimes very different outcomes to a game depending on whether you look at a single iteration or multiple. Trump’s decisions can be seen as being the one iteration version. A great example of this disparity is with the prisoner’s dilemma. In this game, there are two prisoners in separate rooms with different payoffs that depend on whether neither, one, or both prisoners cooperate with the police. If both prisoners confess, then they both serve an average amount of jail time. If both prisoners do not confess, then both of them serve very little jail time. However, if one prisoner confesses and the other does not, the prisoner that does not confess serves a very long time, while the prisoner who does confess serves no time (gets a deal). If you look at one iteration, the dominant strategy and Nash Equilibrium is to confess. However, if you repeat this many times, the best strategy is to not confess, as both prisoners will serve little time. Trump using the one iteration method can either be greatly beneficial or greatly catastrophic similar to one prisoner confessing while the other does not. However, if he looked at the long-term, he will most likely be able to make more equal non zero-sum decisions and be seen as less unpredictable.
Prisoner’s Dilemma Payoffs (https://policonomics.com/prisoners-dilemma/)