Game theory is an area of microeconomics which studies interactive behaviour, i.e. situations where players’ actions are interdependent. Players could be firms, competing on the same market, politicians, hoping to attract the same voters or a husband and wife, playing on each other’s nerves. As such, game theory can be applied to any situation, involving strategic interaction, and it can help analyse and predict behaviour.
Prisoner’s Dilemma is a simple game with only two players, but it’s a brilliant tool to understand economic, political or social phenomena, where two parties fail to cooperate despite the fact that it would be in their best interest to do so. The original game considered a situation where two prisoners, partners in crime, were being questioned in separate rooms. Each prisoner (A and B in the figure below) had a choice of either confessing to a crime, and thereby implicating the other, or denying his participation.
If only one prisoner confessed, then he would go free, and the police would charge the other prisoner and put him into prison for 5 years. If both prisoners denied being involved, then both would be given 6 months for a minor offence. If both prisoners confessed, they would be locked in for 1 year.
The figure below is a payoff matrix for each prisoner. For simplicity, the payoffs are indicated with numbers: the best outcome is 0 years in prison and the worst is 5 years, represented as a negative number.
Since both players choose to confess, they get a year each (-1, -1). You don’t need to be an economist to see that both players could do better, if they hang tight. They would both get 6 months (-1/2, -1/2), which is a better outcome in aggregate. It’s also a Pareto efficient outcome, meaning that you cannot make anyone better off, without making someone worse off. It would be better for the society (comprising two prisoners in the game) if they both chose "Deny, Deny" (speaking in economic, not ethical terms!).
The problem is that the prisoners cannot coordinate their actions, and they probably don’t trust each other enough to sit tight. They both confess and get a year each.
How can we apply this game to real life? Ukrainian government and separatists fight in a civil war. Interpret “Confess” as “keep fighting” and “Deny” as “stop fighting”. Obviously, the society would be better off if both players stop fighting, but because it is tempting to crash the opposition when one player stops fighting, the war goes on when two sides cannot agree on the future of their country.
In business, consider Virgin Media and BT setting the price for fibre broadband in the UK. They both would be better off keeping the retail price at £15 month - “Deny” in our example. But if one competitor keeps the price at £15, another would be better off undercutting its rival slightly by offering a deal, e.g. 6 months at £7.50, thereafter at £15, thereby attracting more customers and making more profit through the higher volume of sales. This is a “Confess” scenario, both Virgin Media and BT offer deals and end up shaving off their own profit margins.
Finally, brothers Mads and Kristian agree to call their mother regularly. If they both do that, she’ll be best off. But if Kristian sticks to this agreement, Mads can skip calling his mother because she is already being cheered up by one son. Unfortunately, Kristian applies the same logic, and the poor woman does not hear from her sons for weeks, thinking them ungrateful brats.
Can you think of other commonplace applications of Prisoner’s Dilemma, where two players end up making decisions which could be bettered, had they cooperated? Jot them down in comments below.