One further note on Foucault, concerning methodological individualism

In my previous post, I neglected one point.  Reading Foucault is one useful path out of extreme positions of methodological individualism.  By methodological individualism I mean the view that "method aimed at explaining and understanding broad society-wide developments as the aggregation of decisions by individuals," as Wikipedia puts it.

Foucault understood how actual historical explanation relies on the use of broad categories, classes, and exemplars, and in a manner which is not logically reducible to statements about individual beliefs and desires.  The writer (theorist) has nothing close to a complete mental model of how the interacting categories reduce to component individual parts, and so some or most of the moving parts of the explanation retain their autonomy at a partially macro level.  The Austrians will kick and scream on this one, but if you combine imperfect information and the sense/reference distinction, methodological individualism ends up as more of a slogan than anything else.  There is a reflective equilibrium to the explanatory process, and micro relies on some macro foundations, not just vice versa, and individuals rely on the social for some of their cues.  Atomistic reduction to the level of the individual in general will not succeed.

The denier of strict MI is not committed to extreme Hegelian views about the autonomous existence of collectivities and it is debatable how much even Hegel himself made that mistake.

I grant that Foucault takes his own method too far in the anti-individualist direction, as did Hegel.

Foucault is by no means the only or even the best path out of extreme methodological individualism.  See this article by David Levy or late Wittgenstein or William James on pluralism, for instance, or more recently Geoffrey Hodgson, perhaps the best place to start.  Here is a quick overview of some of the debates, though it does not cover the best criticisms.  Neuroeconomics, and modular models of the mind, also can be read as critiques of MI, suggesting, as did Nozick, there is no particular reason to stop at the level of the individual in doing the explanation.

Oddly, for all their talk about methodological individualism, economists hardly ever engage in the medium for which it is most appropriate: biography.

A while ago I wrote a review essay on biography and economics.  Here's a challenge: if economics is so powerful, and MI is so persuasive, try writing a biography of a person, using economic tools, and see how much of that person's life you can explain.  It is a humbling and instructive experience and you can read my attempt here.


Economists should use biography. They don't. Therefore they shouldn't. ?

The Stanford Philosophy article claims Max Weber was the father of methodological individualism. I’m not sure. But it is right to say his was not purely a self-interest theory. It was an action theory that starts with the individual as first step in the direction of modeling benchmarks of rationality. Weber was not saying that you could analyze individual action without reference to social context, group dynamics, etc. And of all the social scientists he was arguably the one who most emphasized or most grappled with the complex “multi-causal” nature of human action. However he did say about the state that “for sociological purposes there is no such thing as a collective personality which acts”.

Now, Hegel, though I don’t know his work well, did prioritize collectives like family, community, and state. The state acts ethically much as family does. They have equivalence. As Tyler suggests, Hegel was not dogmatic in doing this. These guys were too clever to take a strong position either way. Instead seems he was putting forward idea that individuals realize their goals and their interests through collectivities such as family, community, or state. So the state harmonizes the many individual interests. And don’t forget that Hegel was also pretty passionate about the autonomous role of the exceptional individual as a world shaping *leader* escaping history pursuing higher ends (not totally unlike Schumpeter’s view of the entrepreneur). In fact there’s a new translation of Hegel’s Lectures on the Philosophy of World History coming out in April, and the Amazon blurb says it treats "the core of human history as the inexorable advance towards the establishment of a political state with just institutions -- a state that consists of individuals with a free and fully-developed self-consciousness”. So there!!

Perhaps the paradox is that methodological individualism is a more effective mode of thought when considering collectives than individuals.

The record isn't ambiguous. The Stanford Dictionary is wrong.

The term "methodological individualism" was introduced by Schumpeter and brought to the modern English academic audience by Hayek in the early 1940s.

Both were attempted to explicate and vindicate the methodology first consciously articulated by Carl Menger in his classic treatise on scientific explanation in the domain of the social, which explicitly pointed to the example of Menger's own explanation for the origin of money as _the_ exemplar of sound scientific explanation in that domain.

After Menger and Hayek, you might say it was all down hill from there ...

Michael Heller writes:

"The Stanford Philosophy article claims Max Weber was the father of methodological individualism."

Let me say it again.

The explanatory paradigm found in Menger & Hayek focuses on learning in a social context, it does not focus on "action" or "behavior" or "rationality". And they both start with empirical problems of design like order which arise in our empirical experience.

Hayek is explicit about this, see his "Economics and Knowledge", his _The Pure Theory of Capital_, and his "The Use of Knowledge in Society", and his "Degrees of Explanation".

Mises & Robbins & Popper et al failed to grasp this, and they turned attention towards explaining "action" and "behavior" and "rationality", and confining methodological attention on those elements, mostly under the influence of demands from the philosophical tradition.

Later philosophers and economists continued to work under the influence of Mises, Robbins & Popper -- adding to it the fixation on math constructs and formal logic, giving us so much of the pathological literature on "methodological individualism" we find today.

This is a confusing blog post. Of course collectives can be reduced to the aggregate decisions of individuals because collectives are aggregates of individuals. In the same way, individuals can be reduced to the interaction of body systems, systems can be reduced to the interaction of types of cells, cells can be reduced to chemistry, chemistry can be reduced to statistical mechanics, statistical mechanics can be reduced to quantum mechanics, and quantum mechanics - well - is just the fundamental nature of the universe.

So I take issue with the notion that the macro is logically irreducible to the micro. On the contrary, for them to be compatible models, the macro has to be logically reducible to the micro: we must be able to describe macroeconomics as some suitable aggregate of microeconomic quantities. (Once we can do that in principle, we have satisfied ourselves as to the compatibility of the fields and we can go about purely micro/macro business as we please; such a link between the fields is going to be very clunky.)

Michael -- the central and original methodological writings are by Menger.

Schumpeter was attempting to vindicate Menger's methodological position and the economics of the Austrian school with more "modern" ideas, including those of Mach.

But of course there was a larger German language literature, which included importantly Weber, as well as Richert, Windelband, and various neo-Ricardians and historical school economists.

Weber was important, but I'd suggest Menger is is the origin.

"Of course collectives can be reduced to the aggregate decisions of individuals because collectives are aggregates of individuals. In the same way, individuals can be reduced to the interaction of body systems, systems can be reduced to the interaction of types of cells, cells can be reduced to chemistry, chemistry can be reduced to statistical mechanics, statistical mechanics can be reduced to quantum mechanics, and quantum mechanics..."

Well, we have been HEARING from reductionists for a couple of hundred years that OF COURSE these things are all reducible to physics.

We just have never, ever seen anyone come remotely close to doing it.

So, excuse me if I regard this as mere bluster.

You've given us a patently false parody of Hayek and Menger's explanatory strategy -- and H and M and the two persons most responsible for advancing the modern notion of MI in the English language.

The explanatory exemplar of Hayek and Menger's explanatory strategy stands as a refudiation of what in the literature a dozens of alternative "explications" of the notion of MI -- none of which are true to the explanatory work of Hayek and Menger.

Your own bogus account of MI is in the same boat -- e.g. Hayek's account of socially acquired rule governed traditions and behavior is incompatible with your bogus stipulation of the ideas of MI. Menger's learning model (e.g. the spontaneous process in which money emerges) is incompatible with your false account of MI as well

Sorry, make that neo-Kantians, not neo-Ricardians!

MI holds that purposeful action is a characteristic of individuals, not of collectives themselves. That is, collective entities, organizations, or sets of relationships certainly exist, but do not themselves act. This is what I've always understood MI to mean. It may have been interpreted otherwise in other contexts, but based on the historical development of the idea, I believe the above understanding is the "standard".

So, if Tyler is indeed arguing that "collectives"/"non-individual entities" do engage in purposeful action, pursuing their own ends, then I'm having trouble seeing the substance of his argument here. I see quite a bit of hinting at explanations, topics clearly related to this issue, but nothing is tied together into what would constitute a rebuttal. Sorry if this seems like nitpicking, but I feel the nub of the issue was not really addressed.

In practice, in economics, arguments against methodological individualism are made to justify quite preposterous "macroeconomic" methodology.

The subtle criticisms of methodological individualism presented usually do not come even close to justifying the extremely crude non-individualist methodology of said macroeconomic models.

That Tyler Cowen can not explain biography with current economic tools says more about current economic tools than than it says about methodological individualism.

"Of course collectives can be reduced to the aggregate decisions of individuals because collectives are aggregates of individuals."

MI, the way it is used by its supporters, says more than this, I think - it says that collective behaviors can be fully explained in terms of the self-interest of the individuals making up the collective. (Thus, the Wikipedia definition is too broad.)

What convinced me that this more strict understanding of MI is wrong is the literature on cultural evolution (e.g. Pascal Boyer explains pretty well why religious behaviors cannot be reduced to self-interest). Also, I think Ortega y Gasset's 'Man & People' was relatively successful in providing the middle ground between MI and Durkheimian-style talk about collectives having an independent existence. For instance, he has a discussion about hand shakes - while it may be a useful practice to greet each other, this utility does not fully determine the exact form that the greeting must take (hand shakes); so, in order to explain all the details of collective interactions you need at least one an additional phenomenon to self-interest.

Now you are criticizing biographers.

Gene is right about the reductionists. They don't take into consideration the fact that the interactions among parts result in the emergence of new traits at a higher level of complexity. So in a real sense, things are not and can never be merely reduced to their least complex elements. We must take into consideration the interactions of those elements, which give rise to more complex entities which themselves interact, etc.

This brings us to complex humans interacting in complex ways to give rise to complex social systems. But of couse, as Hayek pointed out, a complex system cannot understand anything as or more complex than itself -- thus, it is impossible to fully understand the complex social systems we are a part of (or even outselves, fully). So the best we are left with is methodological individualism for understanding humans in society.


I'm agreeing with you.

Comments for this post are closed