This is a story about greed, distrust, and fear. But it starts with people
playing a simple game. In some sparse little rooms in Davie Hall, students are divided into groups of three, learn a few rules, then play against other groups. The task is monotonous—each group chooses either option A or option B.

What makes the game interesting is that each group’s choice determines how much money they take home. Option A means cooperating with the other group, while option B means competing. If both groups cooperate, each makes an equal amount of money, say, $12. But if only one group cooperates, the competitive one gets $18, while the cooperators get nothing. If both groups compete, each makes $6.

Sounds simple. But the psychologists who conduct these trials have found that the groups, small and temporary as they are, reveal conditions that foster competition and greed, or cooperation and trust.

“Groups are greedy in ways that individuals aren’t,” says John Schopler, professor of psychology.

Schopler has worked with Chet Insko, professor of psychology, for ten years, studying the “discontinuity effect”—decent people, when banded together into groups, will behave indecently. Insko named this the discontinuity effect because behavior in groups seemed discontinuous with the characteristics of individuals.

Though Insko and Schopler have documented the effect, it’s an age-old problem, Schopler says. Plato acknowledged it when he wrote, “Had every Athenian citizen been a Socrates, every Athenian assembly would still have been a mob.”

The researchers have studied this behavior by having individuals and groups play trials of the “Prisoner’s dilemma game.” The “dilemma” seems like a no-brainer: everyone comes out best if both sides choose to cooperate. But what if you agree to play along, and then your opponent turns on you? “By being cooperative, I leave myself vulnerable,” Schopler says.

Despite that risk, “If just you and I are doing this, and we can communicate ahead of time, ninety to ninety-five percent of the time, we’ll cooperate,” Schopler says. But adding just two people to each side, which forms a group, reduces the rate of cooperation to less than 50 percent.

Schopler and Insko have established the persistence of the effect by running trials with individuals and groups of varying sexes, and with already-intact groups, such as army platoons at Fort Bragg. But each time, the cooperation rate for groups stayed at just under 50 percent, while individuals consistently had a cooperation rate of 90 to 95 percent. The rates also stayed the same no matter how many people were in a group, as long as there were at least three. So in their trials the researchers use the smallest—and least expensive and time-consuming—group size.

Insko and Schopler have tried to find some of the reasons behind the effect.
One might be good old-fashioned fear. “Our studies have shown that people expect their interactions with groups to be more competitive, unfriendly, distrustful, and aggressive than their interactions with individuals,” Schopler says. In one study, they asked participants to carry for one week a diary in which they recorded and rated all interactions lasting more than two minutes. The subjects generally rated their group interactions as more competitive and abrasive than individual ones. This trend held true even when the researchers removed sports activities from their analysis.

Distrust of groups can also be seen in the trials. Often, letting individuals communicate beforehand increases cooperation. But with groups, such contact did little or nothing to increase cooperation. “Groups don’t trust each other anyway, so they don’t believe the other group when they promise to cooperate,” Schopler says.

“So, part of the discontinuity effect is a function of fear that the other side will be competitive,” Schopler says. “If my group expects your group to be competitive, then we have to pick the competitive choice too, unless we’re into martyrdom. Because if we cooperate, we’re getting zilch.”

Another factor is greed. The anonymity of a group lets people act selfishly without their opponent knowing exactly who to blame. “If we’re playing one-on-one, and I say I’m going to cooperate, and then you exploit that, I know who’s done me in.” Schopler says.

Now the researchers are trying to figure out what might negate the discontinuity effect. It has been a common notion among psychologists that, to increase harmony among groups, you assign them to cooperate with each other on a project. Schopler and Insko have tried that, but with little effect. Participants played ten trials of the prisoner’s dilemma game, then were asked to dissolve their groups and work together to solve a problem. Then they were divided into the same two groups again.

“They did another ten trials, after having this wonderful cooperative experience,” Schopler says. “And the last ten trials turned out to be just as competitive as the first ten.”

Another possibility for reducing competition is to increase “future orientation,” in which people focus more on the long-term effects of their actions than on immediate gain, Schopler says. This idea may have played a part in one real-life instance of inter-group cooperation. In The Evolution of Cooperation, Robert Axelrod recounts a story of soldiers fighting in the trenches during World War II. “Apparently it wasn’t that rare for noncommissioned soldiers on both sides to work out a business where they would shoot to miss rather than shoot to kill each other,” Schopler says. “They had to shoot to keep their officers happy.” But they worked out this mutual cooperation so that both sides came out better.

“Evidently this never happened in any subsequent war,” Schopler says. “In wondering why, Axelrod thinks that the stationary nature of the trench warfare meant that if one side cooperated, the other had a chance to reciprocate. If you didn’t kill the guy on the other side, he’d still be there tomorrow, and it’d be his turn to cooperate.

“If we can get the groups to not be so focused on the immediate trial, then cooperation levels will rise,” Schopler says. One possibility is to have participants make their choices successively, and with knowledge of the other group’s choice, rather than simultaneously and blindly. Playing multiple trials using a tit-for-tat strategy may also help—if you know your opponent cooperated on the last trial, maybe you’ll cooperate on the next. The group may also be swayed if enough of the individual members have a strong orientation toward the future and toward trusting others.

The researchers are optimistic. Shopler says, “We believe that intergroup hostility and competition can be replaced by intergroup cooperation and trust.”