Cooperation is a fundamental feature of social life, but encouraging cooperation and enforcing cooperation can be surprisingly difficult. Let me illustrate this with the story.
The cooperation dilemma
It’s a story about pairs of workers who collaborate on cleaning jobs. Several times a day each pair will perform a cleaning job, each time for a different client. When they are at their job, each of the two workers can choose to do their job well cleaning like they’re supposed to or each worker can choose to shortchange the client not doing the job that the client expects.
Let’s represent each cleaning job as a simple game: when each worker does the job they are supposed to their pay off in the game is the difference between what they received for the job and the effort they expend. If they shortchange, the client the near net payoff actually increases since they receive the same amount, but they are not expending as much effort, and since each workers payoff doesn’t depend on what the other worker does each not only has significant incentive to cheat, but also has no incentive to monitor the partners cheating.
In a game like this, one you might expect that the workers would systematically shortchange their clients, they are better off individually for doing so but that’s not what happens. Let me tell you a little more about these workers because they aren’t who you might expect. The workers are actually small fish called cleaner wrasse. Their clients visit them on cleaning stations on coral reefs where pairs of cleaner fish work together to remove dead skin and external parasites, sometimes even by swimming into the mouth of their client.
When the pair of fish work together to clean the client each get a little bit of nutrition from the dead skin and the client fish come back to the cleaning station over and over, but some of the time, one of the cleaner fish doesn’t do his job. Instead of removing parasites it might take a small bite out of the client trying to get a little extra food that could cause the client to swim away if it notices.
Why we and other animals work together
So here’s the amazing thing: when one of the cleaner fish cheats the client the other cleaner fish chases the cheating cleaner fish away. That fish stops eating expends energy unnecessarily and potentially risks a dangerous swim through open water just so that it can punish its partner. And, what happens next? The cheating fish cooperates the next time, it works together with his partner to do its job and they both feed equal. Cooperation was enforced by punishment.
In today’s article all talk about cooperation, why we and other animals work together for mutual gain. Let’s begin by walking through a simple game that allows us to measure cooperation in the laboratory, this time among humans instead of fish, it’s called the trust game. The trust game is played by two people at a time and investor in the trustee. For now, imagine that these people are sitting our computers in two different rooms. They don’t know each other’s identity, won’t meet each other during the game and can only communicate through their decisions in the game.
When the game begins, the investor is allocated a starting endowment, say $20, then the investor must decide how much of that money to invest with the trustee. Any money invested is multiplied by some factor, for now let’s see if the money is triples. So if $10 were invested then the trustee would have $30. The trustee then decides how much money to give back to the investor and how much to keep. This is a powerful game because it distills trust down to its essence, the belief that another person will reciprocate a prosocial action evening against their self-interest.
Let’s think about the equilibrium solution to this game. What would rational self-interested players do if they play this game once? What should you do if you are playing? Suppose you’re the trustee, a random anonymous person has just given you $30 and there’s nothing forcing you to send any money back. So, if you are purely self interested you should keep the money, a self interested trustee should keep any money given to them.
Knowing that a self interested investor shouldn’t invest anything, there should be no trust in the trust game, but that equilibrium solution no investment no repayment only rarely happens, most people invest at least some money. Across many different studies the investors typically send back about 50% of their initial endowment to the trustee and most trustees return a significant amount of money, but a substantial minority do not. On average trustee is returned slightly less what the investors send.
I want to emphasize that there is considerable variation here. For example, if the investor sent $10 and that was multiplied to a total of $30 some trustees would return $15 that seems like a fair split, but some would only return a few dollars or nothing at all. So some of the time investors benefit from being trusting but often they don’t and there’s also evidence for cultural variation. This game has been used in experiments in many different countries.
In some countries like Germany trustees returned relatively little of what was invested, while in other countries like Bulgaria trustees returned much more. That doesn’t necessarily mean the Bulgarians are more trustworthy or cooperative than Germans since many other factors and contributed to the differences in these experiments, but it does point to the potential for quantitative measurement of something that seems so subjective: trust.