Sunday, September 13, 2015

Biopolitics is Better than Genopolitics: Political Evolution in Six Dimensions

Beginning in the 1970s, a small group of political scientists joined an intellectual movement called "biopolitics" or "politics and the life sciences" (Blank et al. 2014).  They had a grand vision of turning the mainstream of the discipline of political science towards a political science rooted in biological science.  Thirty years later, some of the leaders of this movement (such as Albert Somit and Steven Peterson) lamented that the movement had largely failed, because most political scientists still showed little or no interest in the biological study of politics. 

But then, sometime around 2004, there seemed to be growing interest in biological explanations of political behavior.  One of the most prominent signs of this was the publication in 2005 in the American Political Science Review of Alford, Funk, and Hibbing's article--"Are Political Orientations Genetically Transmitted?"  Their answer to the question in their title was yes, political orientations towards liberalism or conservatism are genetically transmitted. Through the methodology of behavior genetics, using the study of twins, they concluded that about 50% of the variance in ideology is explained by genes.  This article provoked widespread public discussion--including a prominent story in The New York Times--and it become one of the most downloaded articles ever published in the American Political Science Review. This was followed by more articles advancing what was called "genopolitics." One of the most provocative of these articles was by Fowler and Dawes, published in 2008 in the Journal of Politics, which claimed in its title that "Two Genes Predict Voter Turnout."  This was the first time that researchers had identified particular genes as linked to some political behavior.

Those in the biopolitics movement were ambivalent about this new development.  On the one hand, they were happy to see that applying biological science to political behavior had suddenly become a hot research project among political scientists.  And yet, on the other hand, they were offended that those engaged in this new research--Hibbling, Fowler, and their colleagues--gave little or no recognition to the older biopolitics movement.  Hibbing, Fowler, and the others seemed to disdain the older biopolitics as purely speculative theorizing rather than truly empirical science.  Some of those in the biopolitics movement responded by pointing out that there was plenty of empirical research in the older biopolitics (Johnson 2011).

Moreover, as I have argued in many posts over the years, the behavioral genopolitics of Hibbing et al. shows a serious limitation in that it depends upon simplifying models that cannot explain or predict the emergent complexity of political animals, due to the individuality, contingency, and historicity of their behavior.  When Evan Charney published his critique in 2008 of Alford, Funk, and Hibbing's article and his critique in 2012 of Fowler and Dawes's article, I agreed with him.

In May of 2013, the American Political Science Review published three articles on this debate.  Fowler and Dawes wrote a defense of their article against Charney's criticisms.  Deppe et al. wrote a report on their attempt to replicate the results of Fowler and Dawes.  Charney and English wrote a response.  These articles provide substantial empirical confirmation for Charney's critique of genopolitics.  They also support the conclusion that the older biopolitics movement is correct in arguing for a complex interactive biopolitical framework that is superior to the simplifying models of genopolitics.

According to Fowler and Dawes in 2008, the two genes that predict voter turnout are the MAOA gene and the 5HTT gene.  They found that the allele of the MAOA gene promoting efficient metabolism of serotonin was directly associated with voter turnout, and that the allele of the 5HTT gene promoting efficient metabolism of serotonin favored voter turnout only when the individuals with this allele frequently attended religious services.

In their article of 2013, Fowler and Dawes report their attempt to replicate their earlier findings.  This is important because consistent replication of findings is a fundamental standard for good empirical science, and Fowler and Dawes admit that Charney is right in pointing out that the results of candidate gene association (CGA) studies have generally failed to show consistent replication.  Indeed, they admit in this article that they have failed to replicate their earlier association of the MAOA gene with voter turnout, and consequently they conclude that this was a spurious association.  And yet, they also claim that they have replicated the association between turnout and an interaction between the 5HTT gene and church attendance.

Remarkably, Fowler and Dawes now suggest that their 2008 article in the Journal of Politics should never have been published, because they had not yet replicated their results in that article.  Is this their retraction of that article?

Deppe et al. attempted to replicate (through analyzing a new data set) the results of Fowler and Dawes concerning the 5HTT gene.  They found that if they stayed with the original model of Fowler and Dawes, they replicated their conclusion that the combination of the 5HTT gene and church attendance may foster self-reported political participation.  But they also found that if they modified the model to make it better (following the suggestions of Charney), they could not replicate Fowler and Dawes' results.  For example, Fowler and Dawes had originally measured voter turnout by looking at how people answered the question "Did you vote in the presidential election of 2000?"  Charney pointed out that political scientists have noticed that people often report more political participation than is really true.  When Deppe et al. substituted actual voting frequency for self-reported voting, the association between 5HTT and voting disappeared completely.  As Charney and English indicate, this new research confirms all of their original criticisms of the earlier article by Fowler and Dawes.

The fundamental problem, Charney and English indicate, is "the simplistic model of the genome and the genotype-phenotype relationship on which genopolitics relies," because such a simplistic model cannot capture the emergent complexity of political behavior as the product of many interacting causes and levels of analysis (393).  Fowler and Dawes admit that their simple models cannot fully explain the complexity of what they are studying (2012, 363, 369-70).  Deppe et al. admit the same point (376-77, 380-81).

By contrast, the explanatory models for political life proposed by proponents of biopolitics are far more complex than the simple models of genopolitics: genes are there in the biopolitical models, but the genes have no effect on their own, because they interact with many other factors at many different levels of analysis (see, for example, Losco 2014, 41, 60, 63).

I suggest that a biopolitical science of political animals would have to move through at least six dimensions of political evolution:
     1. genetic evolution
     2. epigenetic evolution
     3. the behavioral evolution of culture
     4. the symbolic evolution of culture
     5. ecological evolution
     6. the individual life history and judgment of political agents

As a result of the failure of the genopolitics researchers to replicate their reported findings, the American Political Science Review has added this passage to the "Instructions to Contributors" that appears at the beginning of each issue (see the issue of May, 2015):
"For articles that include candidate gene or candidate gene-by-environment studies, APSR uses the same policy as the journal Behavior Genetics.  In relevant part, that policy states that an article will normally be considered for publication only if it meets one or more of the following criteria:
"It was an exploratory study or test of a novel hypothesis, but with an adequately powered, direct replication study reported in the same paper.
"It was an exploratory analysis or test of a novel hypothesis in the context of an adequately powered study, and the finding meets the statistical criteria for genome wide significance--taking into account all sources of multiple testing (e.g. phenotypes, genotypes, environments, covariates, subgroups).
"It is a rigorously conducted, adequately powered, direct replication study of a previously reported result." 
Some of my posts on the various levels of political evolution can be found here, here, here, here, here, here, here, here, here, here, here, and here.

Some of my posts criticizing genopolitics can be found here, here, here, here, here, here, and here.

REFERENCES

Alford, J. R., C. Funk, J. Hibbing. 2005. "Are Political Orientations Genetically Transmitted?"  American Political Science Review 99:153-167

Blank, R. H., S. M. Hines, O. Funke, J. Losco, and P. Stewart, eds.  2014. Politics and the Life Sciences: The State of the Discipline. Bingley, UK: Emerald Publishing.

Charney, Evan. 2008. "Genes and Ideologies." Perspectives on Politics 6:299-319.

Charney, Evan, and William English. 2012. "Candidate Genes and Political Behavior." American Political Science Review 106:1-34.

Charney, Evan, and William English. 2013. "Genopolitics and the Science of Genetics." American Political Science Review 107:382-395.

Deppe, K. D., S. F. Stoltenberg, K. B. Smith, and J. R. Hibbing. 2013. "Candidate Genes and Voter Turnout: Evidence on the Role of 5-HTTLPR." American Political Science Review 107:375-381.

Fowler, J. H., and C. Dawes. 2008. "Two Genes Predict Voter Turnout." Journal of Politics 70:579-594.

Fowler, J. H., and C. Dawes. 2013. "In Defense of Genopolitics." American Political Science Review 107:362-374.

Johnson, Gary. 2011. "Politics and the Life Sciences." Politics and the Life Sciences 30:43-64.

Losco, Joseph. 2014. "Biology and Political Theory." In Blank, et al., eds., Politics and the Life Sciences, 35-65.

Somit, Albert, and Steven Peterson. 2000. "Review Article: Biopolitics After Three Decades--A Balance Sheet." British Journal of Political Science 28:559-571.

No comments: