A new article (ungated, pdf) by Evan Charney and William English in the American Political Science Review argues that two genes do not predict turnout. The article responds to an earlier article (ungated, pdf) by Christoper Dawes and James Fowler, which found evidence for such a link. More interestingly, the authors make a much more general sobering claim about the possibility of identifying genes that predict social and political behavior:
The cogency of the search for single main-effect genes in complex human behavior must be reconsidered. Proteins encoded by at least 266 genes are involved in variation in aggression in fruit flies, yet at the same time, the heritability of aggression is less than 0.1 because of the high level of environmental variance (even though the researchers assumed the environments were identical). If such is the level of genetic complexity and the importance of environmental interaction implicated in behavioral variation in fruit flies, why should we assume that,when it comes to human behavior, things are any simpler?
I am unqualified to assess the validity of this claim or their critique of Dawes and Fowler so I am going to leave it to more qualified commenters to respond.