the sociological approach to politics

When I started graduate school in the last century, my approach to political analysis was very close to an old school rational choice model. People had interests and ideological tastes. Then they asked government to defend their interests or enforce their tastes.  In the last 15 years that I’ve been working on institutions, movements, and related issues, my views have changed. With respect to politicians, I still adopt a somewhat standard rational choice model. Elected leaders have fairly intuitive utility functions, it’s just that the political environment is stochastic in nature and suffused with ambiguity.

However, my approach to voters and “retail” politics is completely different. For example, I no longer believe that people (even fairly educated people) have consistent ideological beliefs. Public opinion research and everyday observation shows that people hold contradictory views on policies, when they even have any knowledge at all. I also don’t believe that many people have terribly stable material interests that are expressed at the voting booth.

So what’s left? The big drivers of politics are group identity and individual self-image.  Basically, my current position is that a lot of mass politics is some version of group identity writ large. For example, a great deal of partisan identity in the US is driven by being pro or anti-black. Foreign policy makes little sense until you understand that a lot of it has to do with fighting outsiders (e.g., Islamists, communists).  In many nations, party coalitions are defined along class lines, linguistic lines, and ethnic lines. In fact, Lipset and Rokkan have an old book that succinctly argues that multi-party politics is really easy to understand once you take all these social categories into account.

While most sociologists appreciate group identity, they tend to under appreciate the role of self-identity, which is really appreciated by psychologists. For example, it is certainly true that the Democratic/Republican cleavage rests on racial attitudes. But that doesn’t explain why Democrats would be less into the military. Theoretically, you might imagine a party that combines pro-black and pro-military attitudes. Once you accept that unrelated identities  can be bundled, it is easy to see that attitudes toward defense probably reflect an individual’s desire to be seen as tough, which through historical accident can be bundled with racial attitudes.

Now, when I try to understand polls or parties or policies, I do consider interests, but I also use the lens of group identity and self-image. It clears up a lot of things.

50+ chapters of grad skool advice goodness: Grad Skool Rulz/From Black Power

Written by fabiorojas

December 2, 2014 at 12:02 am

8 Responses

Subscribe to comments with RSS.

  1. I don’t think attitudes toward defense necessarily indicate a desire to be seen as tough, in terms of toughness as an end in itself. What Ronnie Janoff-Bulman’s research shows is that that the conservative-liberal difference is captured pretty well in the truism, “A conservative is a liberal who’s been mugged; a liberal is a conservative who’s been indicted.”

    Conservatives tend to be vigilant about crimes committed by people at the bottom of the social hierarchy–these tend to be crimes of violence or moral non-conformity. Toughness reduces one’s chance of being victimized here.

    Liberals tend to be crimes committed by the people at the top of the social hierarchy–these tend to be crimes likes abuse of power and economic exploitation. Forms of care (like welfare policies and human-rights declarations) reduce one’s chance of being victimized.

    Just as toughness isn’t quite relevant to liberals, care isn’t quite relevant to conservatives.

    Of course, the problem with this simplified model is that it glosses over the distinction between right-wing authoritarianism, which is vigilance against unconventional people, and social dominance orientation, which is vigilance against subordinate people. The two tend to be correlated, so perhaps that’s why in the U.S. people are satisfied to have one generally conservative party and one generally liberal-centrist party. If you’re interested in reading more about this, though, have a look at Jarret Crawford’s article “Right-wing authoritarianism and social dominance orientation differentially predict biased evaluations of media reports.”



    December 2, 2014 at 4:32 pm

  2. “For example, a great deal of partisan identity in the US is driven by being pro or anti-black.”

    I wonder if attitudes towards African-Americans are the most salient divider. Would it be more or less accurate to say that a great deal of partisan identity in the US is driven by being pro or anti-white? Or pro or anti-WASP?

    In Paul Kleppner’s book about 19th century American politics, for instance, there seemed to be a similar division between, in general, those who identified with the culturally dominant majority (vaguely under the umbrella of the Republican party), and those who placed themselves in opposition to this core group, while at the same time attitudes towards African-Americans (or towards Southern whites, for that matter) played the opposite role they do today.



    December 2, 2014 at 5:09 pm

  3. Having been taught the beauties of a rational choice approach to politics and how much we can learn about how parties, candidates, and voters behave, I have perhaps gone whole-hog in the direction of group identity or collective identity as an explanation of how people actually vote and express support or opposition to a given issue. The people/groups that someone hangs out with and considers herself a part of explains far more about where they fall on the liberal-conservative spectrum than anything having to do with philosophy, interests, or ideology. Although that is something of an overstatement, research in political psychology clearly shows that people care more about supporting their “team” than actually being right; so people take ideological positions on the basis of which team they see themselves on rather than on the basis of fact, logic, or philosophical principles. Jonathan Haidt’s book, The Righteous Mind, is an accessible and entertaining summary of the research that supports that (although it’s not quite sociological enough for my tastes and I disagree with his claim that authority is not a moral principle for liberals). From political science, I like Diana Mutz’s book, Hearing the Other Side, which shows that people with more politically heterogeneous networks are less likely to participate politically in various ways. I interpret this to mean that they don’t see themselves as on one “team” or another, and political participation risks being seen as alienating one group of people you identify/associate with. The more politically homogeneous our networks, the more politically active we are, and the less we actually encounter the facts, logic, and philosophy of other side.



    December 3, 2014 at 6:53 pm

  4. thinkb4post, Haidt’s research is based on empirical data. So when you say you disagree that authority isn’t important for liberals, what exactly do you think is missing?



    December 3, 2014 at 7:09 pm

  5. chrismartin76, I will confess to having not examined the empirical data to mount an equally empirical/methodological critique. However, I have two concerns. One is that there are many types of authority and many sources of authority, so the question isn’t about authority per se, but whose authority. For example, the authority of scientists, professors, and medical professionals (the anti-vaccine crowd notwithstanding) would probably be highly respected among liberals. Because legitimate authority is so contextual and variable, I suspect that Haidt’s findings represent more of a sociological-contextual process by which contemporary liberals have become openly skeptical of certain types of authority (religious, military, police, etc.). Gordon Gauchat’s 2012 ASR article on conservatives’ declining trust in science illustrates the kind of process I have in mind. My other concern is that if the data are based on self-reports of how much people think authority is a good moral justification for something, then they are measuring discursive claims, not implicit cognitive beliefs. Since I haven’t looked into it, that concern might be baseless in this case. If you (or anyone) knows this off the top of their head, please educate me!



    December 4, 2014 at 7:29 pm

  6. thinkb4post, one of the better known questions in Haidt’s hierarchy inventory is “Would you be willing to slap your father in the face, with his permission, as part of a comedy skit?” For both liberals and conservatives, parents are authority figures, yet liberals are more likely to answer yes to this.

    As for implicit cognition, I don’t think Haidt is going to head in that direction. Explicit results consistently seem to have some validity, so I don’t think Haidt’s reliance on them is necessarily a weakness.

    I don’t think scientists are perceived as people who are higher in a hierarchical arrangement. They have expertise, but I’m fairly sure neither liberals nor conservatives see them as authority figures. Of course, I could be wrong.

    Thanks for the pointer to the ASR article.

    Liked by 1 person


    December 5, 2014 at 2:07 am

  7. Hmmm, I missed this when you first posted. I personally found Klandermans’ discussions of alliance and conflict structures (originally published way back when, in the 1990s? — I have to dig out the reference) to be the key. People form relations of “these are my friends” and “these are my enemies” and the attitudes/issues fall into place from there. That said, it is my view (supported more my informed informal observation than hard data) that the race lines laid down in the civil rights movement are the major axis in US politics for who one’s friends and enemies are.

    I do think identities matter, a lot, but part of the identity is not only who you are, but who you are not.



    December 5, 2014 at 4:04 am

  8. Unless I’m missing something, thinkb4post’s point is basically the classic social psych insight that people whose networks are mixed are “crosspressured” and inclined therefore to nonvoting. Loads of evidence from 60s and 70s.



    December 5, 2014 at 2:53 pm

Comments are closed.

%d bloggers like this: