Are Any Human Cognitive Biases Genetic?

This is  important for Overcoming Bias, because overcoming genetic biases may be much more difficult than overcoming learned biases. But it is highly controversial.

Last week, economics Professor Paul Rubin proposed the hypothesis that humans have a genetic bias opposing Free Trade.

But earlier, Matt Ridley (former US editor of the Economist) proposed a genetic bias favoring Free Trade.

In Foreign Policy (March 2007) Robin Hanson proposed that Overconfidence Bias and the Fundamental Attribution Error are genetic biases. But Daniel Kahneman objected.

Is there any evidence from genetics on these hypotheses?

The only direct evidence would be finding genes for a bias. Identifying specific genes for human traits has  recently become possible, and human genes currently evolving have been found for at least 45 traits (here, in Types of Genes Under Selection, paragraphs 4-11): But not for cognitive biases.

Two sources of indirect evidence:

If the genes are fixated, then the trait will be universal in the species ( though not all universal traits are genetic). But no one claims universality for  biases about free trade or immigration, nor does Hanson claim universality for  Overconfidence Bias or the Fundamental Attribution Error, so this doesn’t apply.

If the genes for the bias are not yet fixated but are evolving, then the bias should run in families: Biological relations should have similar biases on free trade, etc., more so the closer the genetic relations. But no such evidence has been found.

So is there no scientific evidence from genetics for the hypothesis that any cognitive biases are genetic?

Robin Hanson says:

… it is fine to spin hypotheses, and evaluate them on the basis of how well they fit with preconceptions and other hypotheses ( personal communication, 5/15/07)

Let’s spin the hypothesis that human cognitive biases are genetic: how well does this fit with our preconceptions? And how well does it fit with what other hypotheses? If it fits well with them, then are we justified in concluding that human cognitive biases are genetic?

GD Star Rating
Tagged as:
Trackback URL:
  • I thought psychologists did claim that overconfidence and the fundamental attribution error were relatively universal biases. And it seems Rubin and Ridly are claiming universal free trade biases. “Universal” means that important underlying tendencies are substantially based in our common (probably genetic, but perhaps cultural) heritage, producing a tendency in a certain direction that is displayed in almost all cultures. It doesn’t mean that such errors are always and extremely expressed in every situation where you might imagine them, or that there are no other influences on such expressions.

    I did post before on genetic influences on beliefs.

  • I don’t see that being either for free trade or against it would need to be universal in order to be genetic. People vary a lot and some of the variation in inborn. It seems plausible that people will be pro-free trade to the extent that they believe it will improve their situation, and anti-free trade to the extent that they believe it will protect them from losses.

    It wouldn’t surprise me a bit if there’s some evolutionary pressure towards mixed neophile/neophobe groups.

  • Degree of religiosity has a large genetic component. I imagine there are some biases involved there. I wouldn’t be surprised to see other genetic factors influencing biases, for example risk-taking, self-confidence, optimism, etc.

  • LP

    Are we interested in biases that are not themselves genetically coded, but result from the basic configuration of the brain (which is)? If so, this introduces a host of biases, such as poor risk assessment when dealing with large populations. In fact, an awful lot of cognitive bias could be fundamentially adaptive and configured into the brain, as a mechanism for making decisions quickly in an informationally complex world. If this is the case, it seems like “overcoming bias” is a useful heuristic for trying to make better decisions, but probably impossible in the end.

  • For those who believe universal genes “for X” imply universal presence of trait X, think of hair color. All humans appear to have what most would call genes for non-blue hair. Yet we occasionally see humans who have overcome those genes and dyed their hair blue.

    I believe that humans have a genetic bias toward overconfidence, but that doesn’t mean I believe in anything resembling what critics imagine when they attack a “gene for overconfidence”. I imagine that overconfidence results from much more general purpose genes. For instance, a genetic bias toward wanting high status might, when combined with a few other ordinary limits on intelligence that cause others to award status in response to confidence, cause humans to learn to be overconfident.

    Bruce, you say (about genetic variation in free trade biases) “no such evidence has been found”. Has anyone looked carefully enough that the absence of good evidence should indicate much?

  • This question seems ignorant of the basic distinction between evolutionary psychology and behavior genetics.

    From Evolutionary Psychology Primer:

    Evolutionary psychology is not behavior genetics. Behavior geneticists are interested in the extent to which differences between people in a given environment can be accounted for by differences in their genes. EPs are interested in individual differences only insofar as these are the manifestation of an underlying architecture shared by all human beings. Because their genetic basis is universal and species-typical, the heritability of complex adaptations (of the eye, for example) is usually low, not high. Moreover, sexual recombination constrains the design of genetic systems, such that the genetic basis of any complex adaptation (such as a cognitive mechanism) must be universal and species-typical (Tooby and Cosmides, 1990b). This means the genetic basis for the human cognitive architecture is universal, creating what is sometimes called the psychic unity of humankind. The genetic shuffle of meiosis and sexual recombination can cause individuals to differ slightly in quantitative properties that do not disrupt the functioning of complex adaptations. But two individuals do not differ in personality or morphology because one has the genetic basis for a complex adaptation that the other lacks. The same principle applies to human populations: from this perspective, there is no such thing as “race”.

    Unless experiments in different cultures have uncovered the qualitative presence or absence of a bias, there is no reason to think that the bias is learned, nor does it seem probable that such fundamental mechanisms as the availability heuristic are culturally taught. Unless experiments uncover heritable quantitative variation in a bias, there is no reason to think that there is any current genetic variation in the mechanisms of that bias.

    The notion of a “gene for” a bias is very naive. Biases arise from the functioning of genetically standard human brains; any number of genes, if disabled, will result in an anencephalic baby that doesn’t have much of a brain and thereby lacks the bias. Doesn’t mean the gene is “for” the bias.

  • Douglas Knight

    The idea of fixated genetic causes of bias is meaningless. We don’t start with rational thought and add bias; bias is probably an epiphenomenon of heuristics we use to approximate rational thought.

    Genetic variation in bias would be interesting, but the claim that it would be harder to overcome than learned bias is just wrong.

    I’m repeating what others have said, but I don’t think they said it explicitly enough.


    To Robin Hanson, Eliezer Yudkowsky, Nancy Lebovitz, JewishAtheist, and Peter McCluskey:

    It is easy to show that none of the five cognitive biases you mention is universal, because empirical studies of cognitive biases have been done:

    1. Overconfidence Effect: “…over 70% of respondents classified themselves as ‘better than average’ drivers…” That means 30% did not, which is far from universality.
    2. Fundamental Attribution Error: “ Persons in a state of cognitive load are more likely to commit the fundamental attribution error.” So the error must originally have been found in less than 100% of the persons, otherwise it could not have been increased by cognitive load. Less than 100% is not universal.
    3. Availability bias: “…people TEND to rate ‘newsworthy’ events as more likely…OFTEN rate the chance of death by plane crash higher after car crashes…” ‘Tend to,’ ‘often,’ ‘significantly more’ mean not universal.

    When I look at the data for the empirical studies used to demonstrate the 67 standard cognitive biases, none have unanimity, so none have universality.

    Not listed among the 67:
    4. Rubin’s bias opposing free trade: Not universal because of Ridley’s followers.
    5. Ridley’s bias favoring free trade: Not universal because of Rubin’s followers.