Tuesday, January 3, 2012

Abandonment of methodological individualism in aggregate analysis is not the same as the attribution of agency to non-individuals

Demographers abandon methodological individualism by studying whole populations. But they clearly don't think there's some kind of Borg-brain that drives reproduction decisions. They just think that individuals make reproductive decisions, but they make those decisions in consistent enough patterns that it is amenable and useful to study it at an aggregate level.

This is the mistake I think Don Boudreaux is making in this post. You can't think of the community as an economic actor just because it's an important unit of analysis, as Don interprets Krugman's claim here. It's just like (moving away from methodological individualism in the micro direction) a neuroscientist doesn't think a small section of the brain "acts" just because it is an important unit of analysis for studying human action.

Methodological individualism is wrong because it confuses the nature of human action with the imperatives of social science. Methodological individualists should not project their mistakes onto non-methodological individualists. Just because we often think in terms in aggregates does not mean agency is being ascribed to aggregates. You get muddled when you think about it that way. Aggregates don't have agency. But who the hell said we only had to study things with agency?

btw - here's a little Kenneth Arrow on methodological individualism.


  1. But Krugman does talk about 'us' as having not just agency, but moral culpability for that agency, and this 'us' doesn't just extend to, well, us, but instead represents an intemporal and eternal 'us' in which 'we' maybe be responsible for things 'we' did before 'we' were ever born. All Bourdreaux's doing is having a little fun by teasing out some of the implications of Krugman's chosen language and I read it as more a critique of rhetoric than economics. Phrases like 'we owe it to ourselves' are not meant to communicate economic ideas but to sell them by appeal to vague patriotic feelings.

  2. Synergy requires that the whole is more than just the sum of its parts. Of course, Don Boudreaux here is incorrect to assume that a community is an economic actor, as you point out.

    Keynes knew it best to have a restricted holistic approach, in fact...


  3. "Methodological individualism is wrong because it confuses the nature of human action with the imperatives of social science."

    It seems like it is "right" or "wrong" depending on the circumstance.

    Anywho, in the bit of research I did on the concept this is what I found to be the most interesting short read as far as criticisms go: http://plato.stanford.edu/entries/methodological-individualism/#6

  4. If it is NOT the attribution to non-individuals, you are bordering on methodological instrumentalism, which is very dangerous.

    Or is there a "none-of-the-above" response to,

    a) methodological individualism
    b) attributing of agency to non-individuals
    c) arguing that it doesn't matter what is really true because our theories are just tools

  5. Yes, increasingmu: work at whatever level of analysis helps solve the problem you are studying.

    Jeez, pick up any book of, say linguistics, and you'll get a "none of the above": No linguist thinks the English language or universal grammar are agents! But they analyze at that level.

  6. Gene's response is the best one. I really have clue zero why this is such a big deal amongst economists.


All anonymous comments will be deleted. Consistent pseudonyms are fine.