Stumbling and Mumbling: Methodological individualism, again

« Chebyshev and chickens | Main | Nothing unusual »

August 16, 2007

Methodological individualism, again

Norm and Danny aren't happy with my efforts to defend methodological individualism. Norm says:

We can only properly understand individuals, what they do and why they do it, by understanding their place within different social structures and institutions, different sets of rules, different cultures.

And Danny says:

Understanding an individual's disposition does not lead directly to an understanding of how they will behave. For that you need to understand their situation. Behaviour is a result of all sorts of pressures and incentives derived from your relationship with others. If you want to change behaviour you might (might) be better tackling institutions and group norms than trying to alter the disposition of each individual.

I agree. Indeed, I'm often the first to accuse those who talk of the importance of individuals' disposition of committing the fundamental attribution error.
However, both these claims are consistent with methodological individualism. All this says is that the individual is the basic unit of analysis of society.
Methodological individualism doesn't deny a role for situation, environment and social structure in explaining individual behaviour. It merely asks: what's the mechanism?
Take the claim "poverty causes crime." In one sense this is silly. Only individuals cause crime: no-one ever got mugged by a Gini coefficient. The question is: what are the routes through which poverty causes crime? How does poverty affect individuals' incentives? How does it shape their outlook on life?
Methodological individualism, then, recognises that individuals make choices - a point Norm has often made so well - albeit ones that are influenced by their environment. How much so is an open question requiring study. That's all.
What worries me about methodological individualism is not that it's wrong, but that it might be trivially true.
But I don't think it is. It does two things. First, it focuses on the question: how, exactly, do "social factors" - poverty, institutions etc - affect behaviour? It tells us that glib lines like "poverty breeds terrorism" are inadequate.
Second, it tells us to be wary of any claims about collective agency. Nothing is done by the working class, jews, Arabs, homosexuals, or whatever. Only by people.

August 16, 2007 in Philosophy | Permalink

Comments

It's generally called fundamental attribution "bias", rather than "error", error implies there is a "correct" way of explaining behaviour. Hugely overblown explanation in my view. Criminals rarely admit they committ crime because they are bad people who can't be bothered to work, so they blame drugs, poverty, family, peer pressure i.e the situation rather than themselves although strangely the opposite is true of sucesfull people. I've never heard Bill Gates directly attribute his success to those around him.
The most that I think can be said about social context is it normalises aberrant behaviour, but as the same environemnt doesn't have the same effect on all people, there must be other variables (i.e individual drives) at play. It's perfectly possible to predict individual behaviour, independent of situation, based on past behaviour, we all do it constantly.

Posted by: Matt Munro | August 16, 2007 at 05:34 PM

"poverty causes crime"?

This could be a post hoc fallacy.

Has anyone considered the fact that a criminal predisposition - often as an inability to comprehend consequence - is one likely to result in poverty?

How about: "The criminally inclined are likely to become poor". I prefer that to "the poor are inclined to become criminal".

Posted by: Roger Thornhill | August 16, 2007 at 06:32 PM

Roger Thornhill: The problem with "the criminally inclined are likely to become poor" is that it conflicts with empirical data. One notes that many of the richest people in the world are criminally inclined, to say the least.

Posted by: rootless2 | August 16, 2007 at 07:46 PM

Do we need a basic unit of analysis? What I mean is, might not the appropriate unit of analysis vary with the question being asked?

Although, I suppose that the point of methodological individualism - it says that even if you are talking about culture etc. you always have to ground your analysis in individual behavior.

Posted by: Luis Enrique | August 16, 2007 at 09:21 PM

"Nothing is done by the working class, jews, Arabs, homosexuals, or whatever. Only by people." Aye, much what Mrs Thatcher famously observed.

Posted by: dearieme | August 16, 2007 at 10:53 PM

Which, come to think of it, you pointed out yourself a day or two ago. Sorry.

Posted by: dearieme | August 16, 2007 at 10:54 PM

"One notes that many of the richest people in the world are criminally inclined, to say the least."

Fascinating stuff. Will Warren Buffett, who criticised Bush's tax cuts, shortly feature on the FBI's most wanted list?
http://en.wikipedia.org/wiki/Warren_Buffett

And as for Lord Sainsbury, a leading donor to the Labour Party . .
http://www.forbes.com/lists/2006/10/9CUI.html

Posted by: Bob B | August 16, 2007 at 11:11 PM

Just browsing the internet, your blog is very, very interesting.

Posted by: Freddie Sirmans | August 17, 2007 at 04:38 AM

I think the crime leads to poverty argument holds, except that the outcome is also a function of intelligence (in the sense of sucesfull adapatation to environment).
I've never understood people who shoplift, or deal drugs at a retail level, or nick car radios. You will never get rich doing that.
You need to be in the wholesale drugs business, or commit a decent size fraud, or a bullion roberry to get to the level where crime pays, and intelligent criminals know this. Big risk, big reward. Intelligence also gives a better chance of evading detection/capture.
So:
Intelligence plus criminal tent = Rich
Low intelligence plus criminal intent = Poor

Posted by: Matt Munro | August 17, 2007 at 01:07 PM