What Is Politics About?

What, really, is politics about? Yes, in principle it is about all possible social choices. But in each society people usually seem to line up along one main political axis; just one ideological parameter predicts many opinions well.

One reason is a general tendency for coalitions within a society to support and ally with other coalitions on “their side.” This tends to produce a common one-dimensional axis that people bicker about. But there also seem to be many common themes in the issues different societies and eras choose for their one-dimensional bickering. For example, the World Values Survey found that two factors explain much of the cross-national variation in opinion (more about those soon).

If there are certain innate political dimensions, where could they have come from. And why would political positions be correlated with genes? I can see three general possibilities:

  1. Frequency-dependent fitness – there may have evolved (genetically or culturally) a mixed strategy equilibrium where all political strategies are on average equally effective. Or at least they might have been equal in some past environment. These could either be strategies for dealing with politics in particular, or general personality strategies as they happen to apply to politics.
  2. Context-dependent strategies – we may be seeing different realizations of common context-dependent strategies, different because people grow up in different environments. For example, humans could have evolved to have their political opinions depend on personal or local wealth, health, density, etc.
  3. Evolution in action; in response to recent innovations or environmental changes, some tendencies may be winning out. But if the process is slow enough, we may still see both the old and the new ways of thinking represented in the population.

We have lots of data relevant to these theories. We know many things about the correlates of personal political opinion, and about general trends across nations and across time. But oddly, I don’t know of any general review that summarizes that data for the purpose of choosing between the above categories of possibilities. Does anyone know of such a thing?

GD Star Rating
Tagged as: ,
Trackback URL:
  • T. Bell

    Most reef fish change sex during their lives, evidently in response to environmental cues related to sex ratios in the same species. Perhaps humans benefit from contrasting ideologies in much the same way fish benefit from contrasting sexes: gene/meme mixing helps counter predators, parasites, and habitat disruption.

    On that view, we observe ideological conflict because humans have been selected to create and nurture it. Test case: Do once-ideologically pure populations come to host ideological differences?

    • Miguel Madeira

      “Do once-ideologically pure populations come to host ideological differences?”

      I think that Scotland and Wales were almost unanimous Liberal in 19th century.

  • And why would political positions be correlated with genes?

    They aren’t. Twin studies are extremely weak evidence of anything. Tell us when actual genes correlating with political beliefs (or intelligence, or pretty much anything else) are identified.

    • Are you thinking along the lines of this sort of study, but for political beliefs?

    • So: did you follow the link already?

    • I’m not sure I follow your logic.

      Would you have refused to believe in a genetic component for physical conditions such as diabetes, cancer, or color of eyes, hair, or skin until those specific genes were identified?

      Or do you apply different standards of proof to claims of genetic components to psychological traits than to phenotypes and medical conditions? (And if so, why?)

      I’ll point out that medical conditions such as obesity, diabetes and cancer — and even physical conditions such as skin and hair color — have both a genetic component *and* an environmental one, and also are influenced by behavior.

  • I think we need: 4. People with good quality genes might prefer the weathly keeping hold of their stuff – while people with bad quality genes might prefer wealth redistribution.

    • James

      Similar idea, but here’s how I’d phrase it: “Since generosity and contributing to society are considered good traits and those who do so are regarded highly, people who perceive themselves as able to contribute significant value to others (and therefore raise their status) prefer to be able to create that value and increase their status, while people who perceive themselves as unable to contribute significant value to others prefer that others are less able to so that their status is more equal.” It might explain why my engineering-major friends strongly tend to be libertarians but English and philosophy majors are stereotypically very liberal. (Not that English and philosophy majors actually think they’re worthless, because they aren’t and hopefully don’t, but it’s common to think that academic types don’t “contribute” anything to most people, and that common perception probably influences their own perceptions.)

    • Genes can only be good or bad in relation to a specific context. If we degenerates succeed in wealth redistribution, the environment changes to one we can thrive under.

  • Divalent

    My not very well researched hunch is that there are genetically or mimetically influenced behaviors that predispose (at their extremes) for polar opposite actions (e.g., “stick with what works” vs “try something new”; “venture out to new lands” vs “stay where we are”; “be kind to strangers” vs “be hostile to outsiders”, etc), and at any given moment in time the survival value of one or the other extreme will exceed the other, although which one that is depends on the conditions, and is usually unknown until after the fact. E.g., sometimes it would be better to move on now before deteriorating conditions get worse (giving up advantages in hand for the better conditions are there to be found elsewhere), sometimes better to stick it out and wait for better times (as better conditions now cannot be found elsewhere). Long term, no extreme is always the best, and so the population always has some of each type.

    And (further) my hunch is that a given population is most fit when it is composed of a mix of individuals with various degrees along the continuum, rather than all having some sort of population average (located at any given point along the continuum). Having some individuals at the extremes will produce test cases that can provide information that help the others determined what might be the best course: the friendly individual whose own inclination causes him to establish contact with a neighboring tribe might find that such contact can lead to profitable trade, or he might find that he gets eaten. The aggressive individual who ventures forth to kill them and steal their possessions might actually succeed, or might get eaten. Each individual thereby helps provide information about what the most profitable approach might be for dealing with this neighbor.

    Finally, I suspect our general political orientations reflect these underlying predispositions. (But I don’t think they are one dimensional.)

    But, all this just my hunch.

  • josh

    I would be wary of overgeneralizing based only on recent data.

  • Jayson Virissimo

    What, really, is politics about? Yes, in principle it is about all possible social choices.

    Have political scientists really believed this since Machiavelli?

  • I wrote a simulation, described on LessWrong, of what happens when people decide whose opinions they respect based on how much they agree, and update their opinions based on the opinions of people they respect. (In other words, the society is doing k-means clustering where k=2.)

    The result is that, when the questions are viewed as having only 2 possible answers, this society of agents always converges to an equilibrium of 2 opposed parties who disagree on every issue. But there isn’t a “correct” party and a “wrong” party. The particular pattern of beliefs making up each party is different every time you run the simulation.

    So this model says that people align themselves on a single political axis because they have a set of social issues that they see as having only 2 possible answers to.

    • Oops; the society not doing k-means clustering. K is not defined in the simulation. They’re doing something like k-means clustering; really it’s a bit like gravitation in a high-dimensional space. Each point in the data set attracts other nearby points to it. Also, critical difference: Far-away points repel each other. This is because the model computes the probability that agent X assigns to a proposition, as a function of the beliefs other agents have, and the prior probability agent X assigns to each of those agents being correct on any arbitrary proposition.

  • Anonymous

    Robin, could you please provide data to back up this claim:

    “But in each society people usually seem to line up along one main political axis; just one ideological parameter predicts many opinions well.”

    Do you have data to verify that statement is true across the entire length and breadth of human socio-political history in all of the thousands of cultures and political systems for the last 10,000 or so years?

    Are you saying that people in general do not hold nuanced political views? If so, what data do you have to back that up?


  • Chris

    Haha, so much for anonymity. The system has tied my email address to my gravatar icon automatically. Wonderful!

  • Pingback: The Desert Lamp » Campus National/International » Monday Morning Links- August 30,2010()