Americans, no doubt about it, are lousy when it comes to going to the polls. Since 1945, U.S. turnout has hovered around 48.3 percent, according to the Stockholm-based International Institute for Democracy and Electoral Assistance, which tracks such figures. Those statistics have had political scientists-and politicians for that matter-grappling for years about what sends people to the ballot box. Now a bunch of California researchers have a new theory: genes.

People with the right variant of a key gene could be 10 percent more likely to vote under certain circumstances, according to James H. Fowler and Christopher T. Dawes, two political scientists at the University of California, San Diego, and Laura A. Baker, a psychology professor at the University of Southern California. Their study says two key genes, MAOA and 5HTT, have been shown to have a strong influence on the serotonin system, which regulates fear, trust, and social interaction. Those with the most "efficient" variants of these genes-or the variants able to best regulate the metabolism of serotonin-are likely be more socially oriented, according to the authors. "Voting is an act of pro-social behavior," says Fowler, lead author of the study, which appeared in the July issue of the Journal of Politics.

In particular, the most efficient variant of the MAOA gene raises the likelihood of voting by about 5 percent, the study says. The study also finds people with the most efficient version of the 5HTT variant or allele and who are active in religious organizations are about 10 percent more likely to vote. The religious activity is key as these people aren't any more likely to vote if they only have one of the two characteristics-the right variant or are active in their church. The thinking is that these people must be strongly exposed to the religious group's sense of community as well as better equipped to handle the potential pain associated with social risks thanks to their ability to better metabolize serotonin.

To arrive at their findings, Fowler, an associate professor of political science at UCSD, and his fellow researchers examined genetic information as well as the voting patterns for thousands of young adults in the eight-year National Longitudinal Study of Adolescent Health. The researchers concluded that based on the voting behavior data on identical twins, 72 percent of the variance in self-reported voter turnout can be attributed to genes. "The strength of the effect is pretty surprising," says Fowler.

The study adds that the findings on the role genes play could explain two constants in voter behavior. Parental turnout is a strong predictor of turnout, the authors say, but previous studies concluded it was because of social influence. (We do it because our parents do it.) The other constant, according to the authors, is that turnout is habitual-people either always go out to vote or always refrain.

But the study has some major skeptics. Among them is Evan Charney, assistant professor of public policy and political science at Duke University. Charney says he appreciates that Fowler acknowledges some of the study's limitations-it hasn't yet been replicated, for example-but says it has many more. For starters, some 20 to 30 percent of people say they vote when they really don't, says Charney. What's more, another 20 percent of American voters only vote once in their lives. "Both of these facts alone reduces the correlation [between genes and turnout] to statistical insignificance," says Charney, who has studied genetics and political ideology.

Another failing, according to Charney: the study doesn't take into account factors known to influence voting behavior including things like the cultural background of voters as well as the distance to the place of voting. "You have to take into account that voting behavior is going to be profoundly constricted and shaped by a number of environmental variables," Charney says. Marcy Darnovsky, associate executive director at the Center for Genetics and Society, has similar criticisms. And even if the study were to hold up through further replication, she's concerned about the implications. "The worry is we are going to focus less on people understanding the issues, and campaigns are going to focus less on trying to engage voters, if we start to believe it is all in the genes," says Darnovsky.

For his part, Fowler admits the study has shortcomings, including the fact the authors can't test the potential causal pathways they suggest. "The goal of this study is to show association rather than causality," the study says. Fowler is also clear to state there is no single "voter gene," saying humans have about 25,000 genes, which are moderated by environmental factors as well as other genes. "There is likely going to be hundreds of genes that are associated with these behaviors," he says. Sounds like the debate over voter turnout is not over.