Are politics in your DNA?

Twenty-one years ago, a young Australian geneticist named Nick Martin published a paper in Proceedings of the National Academy of Science (83:4364-8, 1986) that described a curious sideline to his regular work on the epidemiology of disease in twins. The study, which Martin coauthored with his mentor Lyndon Eaves, probed the transmission of social attitudes among more than 4,500 pairs of fraternal and identical twins. The results suggested that genetic factors, rather than cultural ones, were mostly responsible for family resemblance in social attitudes.

The potential implications of those results were remarkable, but for two decades the paper languished. It was a frustrating experience for Martin, now head of genetic epidemiology at the Queensland Institute of Medical Research. "It really irritated me that the work was ignored for so long," he recalls. So last year, when he got a call from a group of US political scientists who wanted to follow up the research, he was happy to help.

The leaders of the group, John Alford from Rice University in Houston, and John Hibbing from the University of Nebraska, had begun to feel that standard political science models, which focus on environmental factors, were missing something big. Prompted in part by reading Steven Pinker's The Blank Slate, they began scouring the behavioral genetics literature for relevant papers and soon came across Martin's 1986 study.

The data were tantalizing, but didn't quite fit their needs, so they decided to do some analyses of their own, making use of old survey data from Martin and Eaves. The survey questions that most interested them were those that made up the Wilson-Patterson Attitude Inventory, which asks participants to mark their agreement or otherwise to a list of words or phrases such as death penalty, striptease shows, socialism, and apartheid.

Alford and Hibbing reanalyzed this data with an eye to political orientation, calculating a simple index of conservatism or liberalism based on the spread of yes or no responses, and constructing a measure of political opinion by looking at how many neutral responses were given. They calculated that between 40% and 50% of variation in political orientation was genetic, and almost none of it resulting from parental socialization. On the other hand, when they examined a specific question about political party affiliation, the results were nearly the reverse: Heritability had little to do with it, while shared environment was key.

When the paper appeared in a top political science journal, American Political Science Review (99:153-67, 2005), it generated a mixed response. Many political scientists were intrigued, and a few were downright positive. "We got a more favorable response than we expected," says Hibbing.

Others were vehemently opposed. Political scientist Evan Charney from Duke University, for example, called the paper "incoherent" and "historically inaccurate." Charney believes that twin studies are incapable of dissecting the genetic and environmental effects of being identical twins. "All twin studies face a very basic problem," he writes via E-mail. "It is well known that identical twins share ?environments' that are much more alike than nonidentical twins." Also, he thinks it is historically spurious to say that political ideology can be genetic. "Could a member of a Neolithic Amazon tribe (who is genetically identical to the rest of us) be ?born,' genetically, a communist? A liberal?"

It is no surprise that Alford and Hibbing disagree. They argue that the methodology of extended twin studies is well validated, and they say they're looking for underlying tendencies, not predestination into political ideologies. "This is not about determinism," Alford says. "The point is that the environment is not everything here."

In any case, they're pursuing their line of research. In a late 2006 visit to Martin's lab they threw around a few ideas for potential collaborations, including a search for candidate alleles, functional imaging to see where in the brain politics takes place, and more twin studies. "This isn't a revolution in political science, but it does require a fundamental change in the way social scientists think," Hibbing says. "I think we'll see a gradual incorporation of our ideas into the field."

For his part, Martin is also hopeful. "I'd like to see a whole research program develop in this area," he says. "What I'd like to see is political science starting to take biology seriously."