Liberals vs. conservatives, W&M economists say neither play nicer

Popular wisdom may depict liberals as “Santa Claus” and conservatives as “Scrooge” when it comes to contributing to public coffers, but the word from the College’s experimental economics laboratory debunks both misperceptions, says William and Mary economics professor Jennifer Mellor.

Mellor makes her claim based on data gleaned from experiments involving 144 William and Mary undergraduates who participated in two “experimental economics” games between the spring and summer of 2003. The results are discussed in a paper “Do Liberals Play Nice: The Effects of Party and Political Ideology and Public Goods and Trust Games,” which she co-authored with fellow William and Mary economics professor Lisa Anderson and University of Missouri professor Jeff Milyo. The paper recently was the subject of articles in both The Washington Post and the Washington Times.

Results of the games suggest there is no evidence that self-identified liberals and conservatives act differently when asked to make donations to what, in effect, is a public trust, she says. The findings, she notes, differ measurably from results garnered through traditional surveys, which tend to uphold the perception that liberals are more giving.

One of the games students played was called a public goods game. Subjects first were asked to fill out a survey in which they indicated whether they felt their beliefs more closely represented those held by the Democratic Party or the Republican Party. They also were asked to identify themselves as liberals or conservatives. Next a selected group was put into a room and each individual was given a sum of money. Each subject had the option of keeping all of his or her money or contributing some of it to a group account. Money given to the group account would be multiplied by a constant factor and then redistributed among the group members.

“In this kind of setting, economic theory says that it’s in the individual’s best interest to hold onto all of his or her money because they don’t know what the other players are going to do,” Mellor explains. “They wouldn’t want to put money into the group account if nobody else were going to do so” even though if everyone contributed the group as a whole would be better off.

“In the end, the contribution to the group account on average was about 28 percent of each individual’s endowment over the course of the game,” Mellor says. “There was no significant difference in group account contribution by party representation or by political party.”

The other game in which William and Mary students participated was a trust game. Individuals were arranged in pairs, and in each pair one person was designated as a “first mover” and the other as a “second mover.” The first mover was allotted a sum of money and told he or she could keep it all or send a part of it to the second player. Any money sent on was tripled in sum and given to the second person, who could then keep all of it or send some back to the first mover.

“Because the first mover does not know what the second player is going to do, that person is very much in a position of having to trust the second player,” Mellor says. “In that case, the economic prediction is that it is in your own best interest if you are the first mover to send nothing to the second player.”

In fact, whether first movers self-identified themselves as liberal or conservative, on average they sent about 50 percent of what they had been given to the second player.

Mellor and Anderson have been routinely conducting similar tests with students in the Experimental Economics Laboratory located on the third floor of Morton Hall. The lab is supported in part through a grant from the National Science Foundation.

Experimental economics is a field of economics that uses laboratory settings to test economic theories. Says Mellor, “There are a couple of ways you can test a prediction in the real world. You could ask people how they would behave if, for example, taxes went up. Would that affect their consumption of gasoline? They might tell you one thing, but the question is what would they really do. In experimental economics you can create a laboratory setting in which you can watch how people behave when subject to such a change.”

Although the Post and the Times reported on “Do Liberals Play Nice” due to the political timing—during the campaign for the upcoming presidential election, President George Bush has stressed his “compassionate conservatism”—Mellor’s and Anderson’s interests are broader based.

“This paper grew out an interest that didn’t really have much to do with politics,” Mellor says. “It grew out of an interest in trying to see if the way people answer surveys generally matches up with their actual behavior.”

So far, evidence suggests a disconnection, but the research remains inconclusive. Toward finding concrete answers, the economic games continue in Morton Hall.