Sisters and partisanship

In summary, we find that having sisters makes males more politically conservative in terms of their gender role attitudes and their partisanship. Particularly for gender role attitudes, we find that these political socialization effects persist until respondents are well into adulthood. Since sibling gender is randomly assigned, we can interpret our results as causal evidence that the household environment (cleaned of genetics, the external environment, social forces, and other such omitted variables) influences political attitudes. 

Source: Healy, Andrew and Neil Malhotra. 2012. “Childhood Socialization and Political Attitudes: Evidence from a Natural Experiment,” (revise and resubmit, Journal of Politics). You can find a link to the paper here. See other papers by Andrew Healy here. 

Imagine the implications for the political future of China . . . 

Where does political orientation come from?

Twin studies within political science consistently indicate that 40–60% of the variation in adult political orientations is heritable. The notion that this variance being attributed to genes is in reality entirely due to a combination of nonpolitical early childhood environmental similarity (i.e., being dressed alike or sharing a bedroom), the mutual influence of a sibling, prepartum in utero environment, physical appearance, and the seemingly inherent tendency to treat others differently based on these (biologically based) appearances strikes us as extremely unlikely. But even if it is, key elements of this argument—in utero environment, the genetics of physical appearance—automatically bring its adherents into agreement with our central point: that political science must increase the absorption of biological variation into its epistemology. Critics and proponents of twin studies are in general agreement that a wholly nonbiological paradigm for the acquisition of political attitudes will unnecessarily limit progress. Our analysis here certainly reinforces that conclusion while also advancing empirical support for a model of gene-environment interaction that gives both nature and nurture their due.

That is from the article "Biology, Ideology, and Epistemology: How Do We Know Political Attitudes Are Inherited and Why Should We Care?" by Smith et al. (American Journal of Poltical Science, January 2012).
More on the topic is here.