Opinionator: Are Our Political Beliefs Encoded in Our DNA?

Tom Edsall on politics inside and outside of Washington.

In December, 2008, The New York Times Magazine reported on the emerging science of "genopolitics."

"For years," Emily Biuso wrote,

scholars have assumed that a voter pulls the lever because she grew up in a voting household or perhaps sat through a lot of civics classes. But this year two political scientists published studies claiming that in addition to environment, genes may be a primary influence on political engagement.

Biuso was among the first to talk about what has become a bitter dispute within the study of politics, a dispute in which the stakes are high and the potential consequences significant.

On one side of the debate are those like James Fowler, a professor of medical genetics and political science at the University of California-San Diego, and Darren Schreiber, a political scientist at U.C.L.A who point out, in a 2008 article in Science, that

In the past 50 years, biologists have learned a tremendous amount about human biology and its genetic basis. At the same time, political scientists have been intensively studying the effect of the social and institutional environment on political attitudes and behaviors. However, biologists and political scientists have been working largely in isolation of one another. Little cross-disciplinary work has been done.

The authors go on to argue:

This must change for two important reasons. First, recent evidence is making it increasingly clear that genetic variation plays an important role in explaining variation in human political behavior. Second, additional evidence in neuroscience indicates that the human brain may be adapted particularly to solve social problems that are explicitly political. Much of this evidence is associational, and we therefore should be cautious in using it to build causal theories. However, if the need for sophisticated social cognition drove the evolution of the human brain then a new science of human nature will require comprehending human biology in a sociopolitical context.

Representing the other side of the debate are Evan Charney at Duke and William English at Harvard, who contend that "genopolitical" analyses produce "absurdly high estimates of heritability of behavior."

In an email, Charney wrote:

"Genopolitics," uninformed by some of the most important developments in molecular genetics and evolutionary biology over the past fifty years, is the modern day equivalent of phrenology.

Working somewhere between the two extremes are political scientists who are engaged in innovative, if more traditional, research. In the spirit of inquiry, I emailed a few more mainstream professors of government to get their opinion about this new line of scholarship.

Alan Abramowitz of Emory responded cautiously:

I'm sure genetics is a factor in shaping our political and social views but I think that it's very difficult to separate the influence of genes from that of the family, the community and the social environment in general.

Gary Jacobson, of the University of California, San Diego, wrote in an email:

These are serious, competent, careful researchers, and they are engaged in a valid research enterprise. The evidence that there is a genetic influence on some attitudes and behaviors seems quite strong.

Peter K. Hatemi, a professor of political science, microbiology and biochemistry at Penn State, is a leader in the field of genopolitics. Writing with sixteen colleagues in a forthcoming paper, Hatemi summarizes results from "analyses of a combined sample of over 12,000 twins pairs, ascertained from nine different studies conducted in five western democracies (Australia, Hungary, Denmark, Sweden, and the U.S.A.), sampled over the course of four decades":

We provide definitive evidence that heritability plays a role in the formation of political ideology, regardless of how ideology is measured, the time period or population sampled.

In a carefully worded conclusion, Hatemi and his collaborators make it clear that their findings should not be over-interpreted or used to draw causal linkages.

The findings suggest that while genes undoubtedly matter in the aggregate development of political attitudes, individual common variants will have small effects on ideology. Hunting for a single "political gene" is a fruitless endeavor, and we make no such attempt here. Rather, by attempting to identify the countless specific genetic variants of small effects, related to ideology, we hope to uncover and unite larger neurobiological pathways which account for a substantial portion of how ideologies are formed and maintained in a world where both genes and environment interact and remain in continuous dialogue to guide human behavior.

Some of the outcomes of genopolitical research are intriguing. Late last year Hatemi and Rose McDermott of Brown University published "The Genetics of Politics: Discovery, Challenges and Progress." Their paper includes a chart —"Summary of relative genetic and environmental influences on political traits," reproduced here in Figure 1— which is based on "findings from all reported twin and kinship studies which provided estimates of genetic and environmental influences on political traits from 1974-2012." The chart illustrates the authors' estimate of the relative proportion of genetic (purple) and environmental (green) influences, and the level of combined (brown) genetic and environmental influences.

How much does the question of whether our political leanings might be underpinned by our biology really matter?

In an email to The Times, John Alford of Rice and John Hibbing of the University of Nebraska make the argument that genetically inherited political predispositions are reinforcing contemporary polarization between Republicans and Democrats, liberals and conservatives.

"Fissures in the polity now match divisions in people's biology," Alford and Hibbing write, noting that the shift

from New Deal economic issues to race, sex, and cultural issues means many central political divisions in society follow the differences in people's biologically based predispositions.

Alford and Hibbing argue that

it is not so much that people are becoming more in tune with their genetics as that the nature of modern American politics, with ideologically based news outlets and open discussion of hot button social issues, now is free to reflect bedrock divisions. Fissures in the polity now match divisions in people's biology.

Hatemi and McDermott, working with Lindon Eaves of the departments of human genetics and psychiatry at Virginia Commonwealth, see a substantially greater degree of complexity. In their 2012 paper, "It's the End of Ideology as We Know It," they write:

Typically scholars, politicians and the media identify a cluster of attitudes that combine opposition to such issues as gay rights and abortion. It is also assumed that these positions accompany opposition to logically unrelated topics such as immigration and taxation and relative support for capital punishment and military strength. Together, these attitudes help constitute a "conservative" ideology. However, it seems that genes may "have other ideas". It appears that individuals who are genetically inclined toward attitudes typically considered liberal on matters of sexual behavior and reproduction tend, if anything, toward attitudes understood to be more conservative in relation to a broad range of other issues including the draft, capital punishment, taxation, immigration and segregation.

While genopolitics is in its infancy, proponents have succeeded in getting articles published in highly respected, peer-reviewed journals including the American Journal of Political Science, Science, Political Behavior and the Journal of Theoretical Politics. At least five journals have devoted entire issues to the subject.

Much of the debate over the legitimacy of genopolitics has been conducted in the pages of the American Political Science Review, which as early as 2005 published one of the first articles on the subject: "Are Political Orientations Genetically Transmitted?" by Alford, Hibbing and Carolyn L. Funk of Virginia Commonwealth University.

In seeking to answer questions regarding the balance of environmental versus gene-based influences, a bit of eye-glazing prose is in order. The 2005 Alford-Hibbing-Funk paper used twin studies to estimate the heritability of political traits. They describe their methodology as follows:

MZ twins develop from a single egg, fertilized by a single sperm, and share an identical genetic inheritance. DZ twins develop from two separate eggs, fertilized by two separate sperm, and are in effect simply two siblings that happen to be born simultaneously. As such, DZ twins share the same average of 50% of genetic material as do any two biological siblings. It is this ﬁxed ratio (two to one) of genetic similarity between MZ and DZ twins, and the contrasting average equivalence of environment inﬂuence, that provides most of the power of twin designs

The American Psychological Association separately describes twin studies in more user-friendly language:

The classical twin study design relies on studying twins raised in the same family environments. Monozygotic (identical) twins share all of their genes, while dizygotic (fraternal) twins share only about 50 percent of them. So, if a researcher compares the similarity between sets of identical twins to the similarity between sets of fraternal twins for a particular trait, then any excess likeness between the identical twins should be due to genes rather than environment.

Twin studies notwithstanding, biologically-underpinned approaches to the study of politics are hardly going unchallenged. Charney and English published a major critique of genopolitics, "Candidate Genes and Political Behavior," in the February 2012 issue of the American Political Science Review. In an email, Charney listed some of the weaknesses he sees in this particular experimental arena.

First, he writes, the high heritability of political traits found in twin studies are disputed by a new analytic technique — "genome wide complex trait analysis (G.C.T.A.)" — which derives "heritability estimates by utilizing scans of large sections of the genomes of thousands of unrelated persons." When the G.C.T.A.-based estimates and twin studies are compared, the "G.C.T.A. estimates are consistently lower and in many cases statistically insignificant."

Charney goes on to argue:

It is not our claim that genes do not affect behavior. Of course they do. It is, rather, that there are not genes, whether one gene or 10,000, that "predispose" to being a 21^st century American liberal or conservative, or to voting in an American political election, and that these behaviors are not "heritable." A gene (whether 1 or 10,000) can only play a causal role because "the environment" – cells, tissues, organisms – uses genes to create proteins and ranks (which are then used to create more cells, tissues, and organisms). DNA is one resource among others used by cells. Development is not the running of a preexisting genetic program.

Hatemi, one of the objects of Charney's criticism, is aware of the issues English and Charney raise, and continues to wrestle with them. In his email to me, Hatemi wrote:

We know genes and experience continually interact, and they do not operate independently of our social world. Parenting style, resources, education, friends, weather, and all the things we experience growing up, or in adulthood, have a phenomenal role in why we differ. But everyone reacts slightly different even to the most extreme experiences, and this difference in reaction is partly due to our differences in DNA, partly to how we grew up and partly to our experiences in adult life and the infinite interactions between those forces. The problem is science is never that nuanced, and we pretty much have caveman tools to explain a phenomenally complex process, so we use statistics that result in point estimates, knowing the real world is much more complex.

Jacobson of the University of California, who believes that the work of Hatemi, McDermott, Fowler, Alford, Hibbing and others is producing valid results, poses the key question:

What we do with this knowledge is another matter. How do we look at public opinion differently knowing that some of what we measure has a genetic basis? I am not sure what the answer is.

If genopolitical analysis holds up under continued scrutiny, its explanatory potential is enormous. Not only can the field add to the understanding of polarization and the sources of conflicts that have now shut down the federal government, but analyses like these might also shed light on the logic of, say, supporting abortion rights while opposing the death penalty, or opposing food stamps "giveaways" while supporting subsidies to agribusiness.

With so much riding on political outcomes — from default on the national debt to an attack on Syria to attitudes toward climate change — understanding key factors contributing to the thinking of elected officials and voters becomes crucial. Every avenue for understanding human behavior should be on the table: How do we evaluate our goals? How should we judge trade offs? And just how do we actually make decisions?