From a reader:

Jonah,
Pretty silly study actually. The whole basis is an argument about causality, and yet in our field of study, causality is very tricky to find with good data, impossible with bad. Do the researchers have data on the parent’s ideological stances? Do the researchers know if the families talked politics at home or were partisan. Are the parents of matching partisanship? These are basic questions that should be answered before we can talk about transmission of ideology meaningfully, regardless of whether we are discussing genetics as the transmitter.
Probably the most damning part however is that the researchers would have to know the answer to a simple question before they even began- and I bet they don’t know it. Are fraternal twins more likely than identical twins to differentiate themselves attitudinally? You have to know this baseline to draw ANY conclusions.
As for the bipartisan gene pool dying out, that conclusion is silly. For one thing, this assumes that people have ideology to begin with, which is a big assumption since more than half of people nowadays are weak partisans or independents. Second, it assumes that these people do not change over time- they do. We know from good longitudinal studies that intergenerational transmission of partisanship occurs much more when both parents are strong partisans of the same party. The evidence of partisan deallignment in this country would suggest that we are much more likely have a growing pool of independents than partisans- that should lead to more bipartisanship- or centrism, than less.