14.3. The Problem of Nice

Since the Prisoner’s Dilemma was first discussed in the 1950s, it has been a popular topic of study in social psychology. Based on the analysis in the previous section, we can say what a perfectly rational agent should do; it is harder to predict what real people actually do. Fortunately, the experiment has been done.

If we assume that people are smart enough to do the analysis (or understand it when explained), and that they generally act in their own interest, we would expect them to defect pretty much all the time. But they don’t. In most experiments, subjects cooperate much more than the rational agent model predicts.

The most obvious explanation of this result is that people are not rational agents, which should not be a surprise to anyone. But why not? Is it because they are not smart enough to understand the scenario or because they are knowingly acting contrary to their own interest?

Based on experimental results, it seems that at least part of the explanation is plain altruism: many people are willing to incur a cost to themselves in order to benefit another person. Now, before you nominate that conclusion for publication in the Journal of Obvious Results, let’s keep asking why:

This apparent contradiction is the “problem of altruism”: why haven’t the genes for altruism died out?

Among biologists, there are many possible explanations, including reciprocal altruism, sexual selection, kin selection, and group selection. Among non-scientists, there are even more explanations. We leave it to you to explore the alternatives; for now we will to focus on just one explanation, arguably the simplest one: maybe altruism is adaptive. In other words, maybe genes for altruism make people more likely to survive and reproduce.

It turns out that the Prisoner’s Dilemma, which raises the problem of altruism, might also help resolve it.

You have attempted of activities on this page