This headline from the peer-reviewed Journal of Politics:
Two Genes Predict Voter Turnout
is pretty bold.
After all, two lonely genes are rarely responsible for physical characteristics, much less a complex social behavior like voting. Previously, James Fowler and Christoher Dawes have used twin studies to find the heritability quotient for partisan attachment, always being cautious not to assign causality to particular genes. Indeed, there does appear to be a heritable component to party attachment, although it is not clear whether one's home environment plays an ever bigger role.
Are you a Democrat because your biological parents were Democrats? Because they raised you to be a Democrat? Or because there was something in their genes that made it more likely that they woulld become Democrats in the first place?
The thought process goes like this: the genes influence behavior, and certain behaviors are associated with partisan inclinations. Figuring how which genes influence behavior, which behaviors influence partisanship, the pathaways of causality -- that's hard stuff.
This new study finds that Americans who possess a specific variant of the monoamine oxidase-A gene (MAO-A) were much more likely to have voted in the 2004 election.
Additionally, the 5HTT gene that codes for the seratonin neurotransmitter is implicated in religious attendence, which very much predicts voter turnout. Here, religious attendence is an environmental modifier that associates the gene with a particular behavior.
As the authors put it, "In this article, we hypothesize that people with more transcriptionally efficient alleles of the MAOA and 5HTT genes are more likely to vote."
Transcriptionally efficient? It's easier for specific variants of these genes to transcribe proteins.
Both genes are implicated in modulating anti-social behavior, which itself has correlates to voting. Using regression, they find that by segregating all other factors that influence turnout, Americans with a variant of the MAO-A gene called "high" are 5% more likely to vote. Among regular churchgoers, those who a variant of the 5HTT gene called "long" are 10% more likely to vote.
In other words, Dawes and Fowler contend, after isolating every other known influencer on turnout, humans who possess these gene polymorphisms are more likely to vote.
We're back to some chicken and eggs questions: parents who vote are more likely to have kids who vote.
But what makes parents more likely to vote?
The next step here is to study the effect of monozygotic twins who were reared apart -- something that, unless I read wrong, the study doesn't do. That is -- twin A is raised apart from twin B. if twins with those gene variants are more likely to vote, then the case for heritability is even stronger. (Of course, we cannot isolate the effect of a genetically-influenced environment -- kids with certain innate behaviors change the way their parents rear them, thus entangling their genetic inheritence with what their genes do to their environment.
But more practically: is blood-testing the next trick of microtargeters? Finding out populations where these folks tend to concentrate?
| 
