clock menu more-arrow no yes

Filed under:

In an unequal society, feeling richer than others makes you a jerk

Virtual game demonstrates how wealth visibility affects human behavior

Scott Olson/Getty Images

In unequal societies, rich people cooperate less with their poor neighbors if they know they're wealthier, according to a game-based study published today in Nature. But that's not that case when rich people don't know how good they have it; in that case, they're a lot more likely to help their neighbors out.

In terms of wealth inequality, the US stands out like a sore thumb; the wealthiest 10 percent of US households actually control 76 percent of all the wealth in the country. That's a problem for everyone because more inequality leads to less economic growth. But figuring out how economic inequality affects human behavior in the US isn't easy; it's hard to separate economic inequality from other factors like geography, education, race, etc. That's why the researchers who conducted today's study turned to virtual worlds; doing that might give them a better chance of understanding the relationship between economic inequality and human behavior.

Economic inequality, wealth visibility, and human behavior

In the study, about 1,460 online participants were put in different "virtual societies," each with their own economic structures. Some societies were equal, whereas others were very unequal. In the later case, that meant that some participants started out rich while others started out poor. The researchers also manipulated whether participants could see how wealthy their neighbors were.

During gameplay, participants could decide to either cooperate or defect. Cooperating meant reducing their wealth by 50 "units" per neighbor — the currency used in the game — but also increasing the wealth of all of their neighbors by 100 units. That means that if everyone cooperates, everyone gets richer. But that’s a gamble, because players can defect — refusing to give money and taking advantage of their neighbors’ possible generosity. In some cases, participants were also given the opportunity to change their neighbors by making or breaking ties. But before they could make a decision, players were told whether or not the person they could connect with had previously defected or cooperated. Each game went for 10 rounds, and at the end of the game, the units were converted into real money.

In an unequal society, wealth visibility perpetuates inequality

The researchers found that inequality alone didn't really affect the level of cooperation seen in a society — but visibility of wealth certainly did. For instance, when wealth was visible and inequality was high, people who were richer than their neighbors were less likely to cooperate and give up some of their wealth. This is what the researchers refer to as the "exploitation scenario" — a scenario in which rich individuals who defect take advantage of their poorer neighbors, who are more likely to cooperate and invest in their social network. But wealth visibility only had this effect in very unequal societies. In those were wealth was distributed more evenly, visibility actually led to more cooperation.

"In other words, when you randomly assign people to unequal or equal worlds, visibility has different effects on people's probability of cooperating," says Nicholas Christakis, a sociologist at Yale University and a co-author of the study. And in an unequal society, knowing how wealthy one's neighbors are perpetuates inequality.

"Fear of being in last place."

It's unclear why visibility causes so many problems. The researchers speculate that in unequal societies, it might cause participants to perceive the game as a competition, where wealth signals a player's social position. It might also make participants "fear being in last place," the authors write. If that's the case, then all of those effects could reduce cooperation.

"I think the main findings are intriguing," says Brian Uzzi, a sociologist at Northwestern University who didn't participate in the study. The results "should matter to people trying to understand the origins and consequences of economic inequality." But David Coburn, a sociologist at the University of Toronto, isn't convinced that the behaviors observed in the game reflect how humans normally act. "When you play games, do you think your game playing carries over, or reflects your real life choices?" he says. "The authors of the study note that there is no direct relationship shown between the study and reality — some people like to play games."

"When you play games, do you think your game playing carries over?"

Christakis responds to that criticism by stating that although the study wasn't about "a real society of people out in the world playing for large stakes," the participants were "real people playing for real money." He also points out that conducting any experiment — as opposed to an observational study — means dealing with trade-offs. Experiments can help scientists get to the bottom of what's causing a given behavior, but that means sacrificing some verisimilitude, he says. In contrast, "when you go and observe the world, you can get realism, but you aren't sure what's going on causally." The experiment only looked at economic inequality, which means all the other factors that come into play when people think about inequality in general — like educational inequality, health inequality, or racism — can't be accounted for. But that's the nature of experimentation, Christakis says. "There's always a trade-off."

Still, given that the experiment played out in a virtual game with specific parameters, it's hard to see how this might apply to a real society. "According to our experiment, one way to retard the impact of inequality would be to make it less visible, but that's not very realistic," Christakis admits. But societies that are already fairly equal — societies like those found in some Scandanavian countries — might benefit from highlighting that equality. Doing so could make members of those societies feel more connected, and perhaps more cooperative.