Reciprocity Trumps Self-Interest. It Also Explains Why We Disagree About Human Nature.

Some of the more interesting research in behavioral game theory has looked at why people depart from the move that would maximize their self-interest. There are two main possibilities: either they are self-interested but they’ve miscalculated; or they care about the other player’s outcome in addition to their own. (More concisely they are either being dumb or they are being nice).

In Social, Matthew Lieberman reviews one of the more interesting experiments. Players are only playing once (so reputation doesn’t matter) and they are put in a classic prisoner’s dilemma situation. If they both cooperate, they each get $5, if both defect they get $1, and if one defects while they other cooperates the defector gets $10 and the cooperator get’s $0.  The outcomes are summarized in the chart below:

Prisoner’s Dilemma

Player B

Cooperate Defect

Player A

Cooperate 5/5 0/10
Defect 10/0 1/1

The self-interested player will always choose to defect. (To see this, notice that if Player B chooses to cooperate A’s best move is to defect (10>5) and if Player B chooses to defect A’s best move is still to defect (1>0). So A is best off defecting no matter what B chooses to do. Since B faces exactly the same choices as A, the same holds true for B.)

By contrast, actual human beings cooperate about 1/3rd of the time. The interesting part of the experiment, however, is what happens when Player B is told Player A’s choice before deciding. If A defects then B defects. But if A cooperates, then B’s rate of cooperation rises from 36% to 61%. In this scenario, B cooperating essentially amounts to voluntarily choosing to split 10 dollars, 5 bucks a piece, instead of keeping all $10. The initial act of cooperation encourages a reciprocal act of cooperation. This is rather remarkable given that A and B have never met and will never meet. One could expect reciprocity rates among non-strangers to be even higher.

Based on this we have people who initiate cooperation, people who reciprocate cooperation, and people who are jerks. (I suppose technically I should call them people who never cooperate. In the context of the experiment not cooperating doesn’t really seem too much like being a jerk, but think about someone who never, ever returns favors and the term seems justified.)

We all tend to assume that other people are like us, so people who tend to initiate cooperation naturally view human beings as more cooperative and less self-interested. What’s really intriguing about this experiment is that because they initiate cooperation, it really is the case that most other people are willing to reciprocate that cooperation, so their experience bears out their intuition. In a mirror image way, because reciprocating cooperation is much more common than initiating it, people who don’t initiate cooperation will mostly meet people who don’t cooperate with them (but might have if they had initiated the cycle). Thus their initial assumption of self-interested behavior will also be confirmed by experience.

The dominant school of thought in public policy (particularly among economists) has been to design institutions assuming that people will maximize their self-interest and to try to make that work as well as possible. Not only does this sometimes require settling for sub-optimal outcomes, it can also require a great deal of governmental coercion. (Garret Hardin’s famous Tragedy of the Commons argument is essentially that cooperation is impossible and therefore the government must coerce individuals). What if instead we spent some time and effort on investigating how to increase the number of people who initiate cooperation? To be fair, some researchers have already begun working on this, but it isn’t exactly the hottest field in public policy. Psychologists have pointed out that not only does cooperation lead to better outcomes, it also leads to happier people.

We’ve assumed that people are self-interested for so long that it may start to become a self-fulfilling prophecy, as more and more institutions are designed to encourage people to act in a self-interested fashion. We also teach people to think that others are self-interested, which leads them to act in a more self-interested manner, as they fear being chumps. Institutions do need to be designed to hold individuals accountable, but they also need to be designed in a way that encourages voluntary cooperation. This not an easy thing to do (and not a subject for this blog post), but continued efforts in this direction are absolutely necessary. The human brain is wired to be social, and it’s long past time for policy-makers to recognize it.



2 thoughts on “Reciprocity Trumps Self-Interest. It Also Explains Why We Disagree About Human Nature.

  1. This makes me think of the little g in Andreoni. The whole body of literature related to compliance and obedience has always been my favorite.

    • Andreoni never really explains what accounts for little g in people, he just takes it as a given. So yes, I don’t think the approaches are wholly incompatible if you say that another person acting generously first increases the warm glow factor when you choose how to respond. That said, at the top of p. 466 Andreoni does recognize the ‘moral constraints’ or ‘principle of reciprocity’ explanation of giving as separate, and he cites a bunch of that literature. I mainly noticed because I plan to read the Amartya Sen piece that he cites.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s