by Gene Callahan
Many Austrian economists embrace the doctrine of ‘methodological individualism,’ as I myself did, for instance, in my book Economics for Real People. But subsequent study on my part, most significantly of the work of Tony Lawson in his philosophy of economics project he calls ‘Critical Realism,’ as well as my readings of the social theories of the British Idealists, has led me to question the soundness of that position.
I will begin by noting that I think the adoption, by Mises, Hayek, and other Austrians of their era, of methodological individualism is understandable, given that the chief alternatives available when they wrote were some sort of ‘social holism,’ a la Durkheim, or Marxist historical and class determinism. If one’s choice is restricted to those three options, I believe methodological individualism is indeed the preferable stance to take.
However, the very either/or framing of these options was a symptom of the reductionist view of science that was prevalent at that time. Social phenomena, so it was thought, ought to be explained using a single theoretical framework: either explanations should be couched in terms of, say, individual choices, or society as a whole, or in terms of material, historical ‘forces.’ This view was in line with the ‘unity of science’ project that dominated early twentieth-century philosophy of science. But today, even in regards to the physical sciences, the assumptions of that project have been widely rejected; for instance, LSE philosopher of science Nancy Cartwright contends that we live in a ‘dappled world,’ and that the dream that we can reduce the plurality of explanatory schemes to a single, ‘fundamental’ mode of explanation is, in fact, nothing more than a dream.
That reductionist project was tied to the notion that there is some ‘basic entity’ that should be the foundation of all ‘fundamental’ scientific explanations, and that reference to higher-level, composite entities is merely a convenience springing from human limitations in getting at the ‘real’ story. But nothing in the modern history of science suggests that all scientific theories are converging on such ‘fundamental’ explanations: to the contrary, much progress has been made by ignoring the reductionist impulse. The entire science of thermodynamics rests on the aggregation of particles into a statistical ensemble; as Werner Heisenberg noted, it is hardly sensible to ask what the ‘temperature’ of a single atom is. Evolutionary theory explains the traits of individual organisms by looking to the history of the species of which they are a member, a distinctly anti-reductionist enterprise. Ecology arose as a science when it was recognized that the behavior and survival of individual organisms can not be fully understood without recognizing the larger ‘web of life’ in which those organisms are situated. While a hurricane surely consists of a multitude of molecules in motion, the very phenomenon of hurricanes would vanish if our view was limited to looking at each molecule in the atmosphere individually. And quantum mechanics, which is, at present, the culmination of the quest to find the ‘ultimate’ constituents of nature, contends that the behavior of a single particle is intrinsically conditioned by its previous interaction with other particles; in other words, the behavior of the ‘atomic’ components of reality can only be explained by referencing a larger context than individual particles.
So, if physical science has progressed by such pluralistic methods, is there any reason to think that social science cannot do likewise? Surely, Mises and Hayek were correct to argue that the action of an individual is not determined by social wholes or historical forces, as Lawson recognizes when he writes, for instance, that ‘given the open nature of human action—the fact that each person could always have acted otherwise—it follows that social structures can only ever be presented in an open system.’
But why isn’t acknowledging the reality of individual, human choice as compatible with the validity of social explanations couched in terms of higher-level entities as is acknowledging that storms are composed of molecules compatible with a meteorological understandings of the weather? Without denying the reality of individual choice, can we not employ other levels of explanation whenever they are more efficacious? For instance, does it really advance our understanding of social phenomena to insist that ‘the German army invaded Poland’ is somehow an unsatisfactory explanation of the events of September, 1939, and to demand that a ‘real’ explanation must be put in terms of why a vast number of German-speaking individuals wearing very similar clothing just happened, at the same time, to rush eastward and begin shooting at a large number of Polish-speaking individuals who all were wearing a different sort of clothing? Indeed, doesn’t the Austrian explanation of a phenomenon like inflation rely on the existence of a supra-individual institution such as money? And isn’t Menger’s theory of the origin of money reliant on supra-individual social facts that determine the most saleable commodity in any particular society?
Now, it may be objected, to the above, that Mises, Hayek, and other Austrians were well aware that individuals are socially embedded, and that their choices are ineluctably situated within a social context. Just so, I respond! My contention is that Mises and Hayek were led to advocate methodological individualism as a philosophical stance because it seemed preferable to the other options they saw as available for grounding social theory, but, when they turned to explaining real social phenomena, they were far too capable as social theorists to actually restrict themselves to the sort of explanation their purported philosophy ostensibly required. But contemporary philosophy of science has made room for methodological pluralism, and, I suggest, that is the proper attitude for today’s Austrians to adopt: social explanations should be framed in terms of whatever level of reality makes the phenomenon in question most comprehensible, and not restricted by an injunction that declares all but individualist explanations to be invalid.
ThinkMarkets 
It is not clear what is meant for methodological non-individualism. What are the short-comings of individualism? What are the alternatives? (and, considering that there are many possible meanings for the term: what do you mean for individualism?)
Let’s analyse the Germany invaded Poland case. Germany and Poland don’t exist without german and polish individuals, and this is obvious.
The level of aggregation in the analyis depends on a judgement of relevance: in macroeconomics it is common practice to disregard differences among individuals (i.e., the representative agent), but this is not an argument against individualism, because what a representative agent model implies is that the theoretical benefits of further disaggregation are supposed to be negligible.
The same applies for the “Germany invades Poland” problem. At a certain level of description it is sufficient to treat nations as homogeneous entities: once the internal politics of Germany is to be taken into account, however, the collective term Germany hampers understanding because it suggests homogeneity where there is none (like the representative agent).
The more details are required in the analysis (German domestic politics, society vs the state, single relevant individuals, economic constraints….) the higher the level of disaggregation.
The bottom line is that as long as society is interpreted in terms of goals and meanings – and I see no available alternatives – the only source, and the fundamental building block, of these things is the individual. “Only the individual thinks. Only the individual reasons. Only the individual acts.”
“social explanations should be framed in terms of whatever level of reality makes the phenomenon in question most comprehensible, and not restricted by an injunction that declares all but individualist explanations to be invalid.” – Gene
I, for one, entirely agree with you (for a change), though I would probably word it differently: explanations should be framed in terms of whatever level of reality is appropriate for the problem under consideration. It seems to me that some economic problems can be adequately explained at quite a high level–the wisdom is in knowing the limits of such explanations (i.e. when to switch to a lower level, or what qualifications to include).
Having said that, should all high level social explanations be–in principle, if impracticable–reducible to the lowest level? I think so, and I think we should search for such reductions, but it seems to me that sensible arguments can be made for irreducibility in other fields, and so we should be open to such a possibility in the realm of the social sciences.
Gene writes, “But why isn’t acknowledging the reality of individual, human choice as compatible with the validity of social explanations couched in terms of higher-level entities as is acknowledging that storms are composed of molecules compatible with a meteorological understandings of the weather?”
Yes, I see your point. Austrians don’t talk about the unemployment rate of the United States. That is way too aggregated. They like to talk about the molecules that make up the weather, so to speak. Austrians talk about each individual’s personal unemployment rate. Bob’s unemployment rate is 100%. Mary’s unemployment rate is 0%. Steve’s is not in the labor force.
Despite Austrians’ aversion to indexation, you never hear any Austrian say the US Consumer Price Index for All Urban Consumers year on year change is 2.1%. No, Austrians always talk about the individuals that make up the CPI. They say my price inflation over the year was 3.9%. Or the price inflation of John was -0.4% over the year.
And I have never heard any Austrian say that US GDP increased in the 4th quarter by 5.9% at an annualized rate. It is always Fred’s personal output increase 3.1% last quarter at an annualized rate.
Hmmm, on second thought I guess I don’t get your point. Actually, I have never heard Austrians talk about all the molecules that make up the weather when referring to unemployment, price inflation, or GDP. Maybe I am just listening to the wrong molecules that make up the higher level entity called Austrian economists.
I’m with Gene.
I really think we should see, with Smith, that individuals are sympathizing with imaginary minds who represent, as it were, society.
So when we talk about individual action, whether as explanandum or explanation, we are at the same time speaking of the imaginary mind that the individual associates with society.
“Hmmm, on second thought I guess I don’t get your point. Actually, I have never heard Austrians talk about all the molecules that make up the weather when referring to unemployment, price inflation, or GDP.”
Yes, Tom, I know. My point is that, per methodological individualism, talk about unemployment, inflation, and GDP is invalid. I believe I said right in my post that Mises and Hayek did not actually follow their own precept, didn’t I?
I think Gene is right. The particular foil for Mises’s methodological individualism was Othmar Spann whose “universalism” really held that there is some sort of entity called “society” which is somehow above individuals and prior to them. He truly reified “society.” It is true, I think, that Mises relatively early statements were fuzzy on some important points such as the difference between ontological and methodological individualism. But he’s got a pretty sophisticated version worked out in Human Action, one that suffers little or no damage from Gene’s insights’, I think. There is a long literature contrasting sophisticated methodological individualism with more naive and crude versions. This literature goes back at least to Joseph Aggasi’s “Methodological Individualism” paper of 1960 in the British Journal of Sociology. Dick Langlois has written great stuff on this theme, too.
Gene,
That’s an interesting post. I would follow Roger in my comments. The Misesian view of methodological individualism is not a crude one. It is an extremely rich view that draws on economic sociology a la Weber. Perhaps your point is to emphasize that there is more than one form of methodological individualism. The most basic view is impervious to the richness of the social world and thus gives a very crude and basic explanation of human action. Richer approaches see human action as taking place in a social space with all its complexity in terms of rules, history, and relationships. This second view is present in some modern scholars such as Doug North (his work on path dependence) and Vernon Smith (his work on ecological rationality). Another scholar to read on this is Virgil Storr (following Don Lavoie) and his work on the “market as a social space” to use the title of one of his papers (that I criticized at the time, but I like it a lot).
Yet, the crude version of methodological individualism gives us supply and demand curves and the fundamental tools of economic science. It is crude but helps us understand a lot. It is reductionism, but has a lot of predictive power (in terms of patterns). So I am not sure we need to ditch the crude view necessarily. The two are complementary it seems to me, rather than substitutes. The molecules in a hurricane do not decide to go somewhere else than where they must be by necessity. Soldiers in the German army can leave even if they are threatened (and some did). I think there is still basic fundamental differences that put limits on the analogies with the physical sciences and still give relevance to the crude view of methodological individualism.
Gene, I want to make sure I understand your position before I proceed. Are you saying you object to anything that Mises himself wrote on these methodological issues, or are you just saying Mises’ disciples blindly take up his slogans and apply them inappropriately? Basically, I’m wondering if it’s even worthwhile for me to look through Human Action and see if Mises is saying exactly the kinds of nuanced things that you are bringing up here. If I found such passages would you say, “Oh, I guess Mises wasn’t saying the bad thing I originally thought,” or would you say, “I know Mises was nuanced, but his modern followers aren’t. They say silly things on this point.” ?
Frederic said:
“Yet, the crude version of methodological individualism gives us supply and demand curves and the fundamental tools of economic science.”
I’m sorry I don’t get what we’re talking about now. If there is a “fashion craze” and “mob psychology” takes over, that can cause the demand curve for some new type of clothing to shift to the right. I don’t know what you guys are talking about at this point.
E.g. a standard Misesian would say something like, “The Keynesian thinks of the macroeconomy as a few aggregate statistics which interact with each other through an accounting tautology, whereas we Austrians ultimately want to couch an explanation of price inflation or the business cycle in terms of acting individuals.”
So are you guys saying that maybe it’s wrong to insist that all good economic explanations at least in principle be traceable back to individual choices? In other words, that we shouldn’t categorically rule out such economic theories, but we should evaluate each one on its merits, where “merits” doesn’t include the ability to be traced to individual actions?
I fail to grasp any substantive point in what Gene writes. If you are going to question what Mises believed in relation to Methodological Individualism, why not have issue with specific writings? Human Action Chapter II Section 4 might be a place to start.
Similarly, if you are going to make analogy to the physical sciences, address the literature. I understand this is a blog, but what is written is unclear (at least to me).
One final point, appeals to aping the method of physical science seem quite ill conceived, due to the obvious distinctions between studying action and non-acting things. Further, the standards of the physical science community as to what constitutes good science are often as low, and often lower than those of the mainstream academic economist.
“Similarly, if you are going to make analogy to the physical sciences, address the literature.”
What are you talking about?
“One final point, appeals to aping the method of physical science seem quite ill conceived, due to the obvious distinctions between studying action and non-acting things.”
Who said anything about “aping” the methods of the physical sciences. I fail to grasp any substantive point in this comment.
“So are you guys saying that maybe it’s wrong to insist that all good economic explanations at least in principle be traceable back to individual choices?”
Certainly not all social explanation needs to go to the individual to be a “rock bottom” (Watson’s phrase) explanation. Here’s a perfectly good explanation that proceeds in the other direction:
“Why is he painting his face like that?”
“It’s the custom of his people.”
Gene,
I don’t think any Misesian would say that that explanation was a violation of methodological individualism. After all we are trying to explain why *he* painted the face in a certain way.
So if that’s the kind of thing you have in mind, then I don’t think you have a problem with methodological individualism.
Oh sorry I get what you are saying now Gene (I think) with that example. You’re skipping the intermediate step and not saying that the guy “chose” to paint his face because he wants to fit in or because he was taught to etc., but you are explaining it by saying, “Of course that’s what he’s doing, that’s what creatures like him have always done”?
If that’s what you’re saying, OK I agree that it’s not praxeological, but by the same token I don’t think it’s a good explanation at all. If you want to say that the guy has volition and wants to fit in, doesn’t want to violate customs and be an outcast, etc., then I think that is a good explanation. But of course then we’re back in standard Austrian territory.
Considering that methodoligical individualism as actually practiced by the Austrians resulted in Hayek’s spontaneous order theory and his defense of the incremental evolution of tradition (by individuals’ choices, by the way), it seems that this objection is a bit of a characature of the methodology. Of course, I don’t see any actual criticism of the method, excent in the broadest, most generic, and thus least meaningful terms. Nor do I see a proposal for a real alternative here to which to respond.
Let us take what Mises says in “Liberalism,” under “Private Property and Ethics,” for example, to see that he most certainly sees methodological individualism as being precisely social:
“Morality consists in the regard for the necessary requirements of social existence that must be demanded of each individual member of society. A man living in isolation has no mroal rules to follow. He need have no qualms about doing anything he finds it to his advantage to do, for he does not have to consider whether he is not thereby injuring others. But as a member of society, a man must take into consideration, in everything he does, not only his own immediate advantage, but also the necessity, in every action, of affirming society as such. For the life of the individual in society is possible only by virtue of social cooperation, and every individual would be most seriously harmed if the social organziation of life and of production were to break down” (14).
Please note that for Mises — and certainly for Hayek as well — the actual human individual is a member of a social species living in a social environment. Methodological individualism is thus social — it is just not collectivist. And there is a huge difference between the two. Also, as Mises points out, his methodological individualism is designed to tell you what to do if you want to succeed at achieving certain things in an economic environment. It is not designed to help us to understand how people actually behave in a wide range of social and cultural environments. If that is what you are looking to do, then one does have to have a certain amount of methodological pluralism. But if you want to engage in practical praxeology, then you have to do methodological individualism.
“I don’t think any Misesian would say that that explanation was a violation of methodological individualism. ”
Yes. The practice is better than the theory! In any case, as pointed out by Stephen Lukes, once one allows explanation of this type, then “it is hard to see why they insist on calling their doctrine ‘methodological individualism.'”
“Considering that methodoligical individualism as actually practiced by the Austrians resulted in Hayek’s spontaneous order theory and his defense of the incremental evolution of tradition…”
I guess you didn’t read as far as the part where I mentioned the actual practice belied the theory?
MI as expounded by Mises, Hayek and Popper is not challenged by Gene’s account. Is is, after all, methodological individualism, not ontological or atomistic individualism. It does not deny the existence of the situations where individuals find themseles. It may help to talk in terms of situational analysis to point up the need to identify relevant factors outside the individual that need to be considered. These include the lawlike regularities which put parameters on human action, the laws of the land, traditions and institutions, also macro factors like interest rates etc etc. The point of MI is to insist that where deliberative human action and valuation are concerned, the buck stops with the individual and cannot be shifted off to the situation or a collective.
As an aside on Nancy Cartwright’s take on natural laws, do you know if she is aware of Popper’s shift from the Popper-Hempel model of explanation to a theory of propensities which makes a connection between his probability theory and his cosmology? There is an uncanny similarity between the revised Popper theory and Cartwright’s locutions on the topic but I am not aware that she has acknowledged the parallel.
I did read that far. My point is that you misunderstand the theory itself. Methodological individualism deals with human individuals, who are social, and not nonsocial individuals, like sociopaths and nonsocial mammals. Society/culture/etc. emerges from the interactions of various individuals, and the society/culture/etc. then influences those individuals. So the theory, according the Mises, assumes a social individual. Again, Mises points out that methodological individualism is prescriptive, not descriptive. In other words, if you want to have a certain outcome, you have to act in a certain way.
“It does not deny the existence of the situations where individuals find themseles. It may help to talk in terms of situational analysis to point up the need to identify relevant factors outside the individual that need to be considered.”
Right. It is at this point that Lukes asks, “So why are they calling this methodological individualism?” (His paper is collected in a nice book called _The Philosophy of Social Explanation_, by the way.)
Re Cartwright: I don’t know.
I recently read a paper by Ahdieh against methodological individualism on SSRN:
http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1518836
I fear we are discussing definitions… so let’s take Human Action and see if there really are theoretical inconsistencies, maybe solved only by pragmatic eclectism, of the “do as I do not as I say” variety.
I’m pretty sure that Ahdieh’s and of course Callaghan’s arguments are valid for some methodological individualism as really practised by many economists, and for views about it expressed by some Austrians, at least in loose speaking.
In order to prove his points, Ahdieh was forced to say that Hayek was not a methodological individualist. At that point, I would, in his place, have stopped writing the paper and looked better in Austrian writings.
The problem is that his definition (never explicitly given) was a straw-man. It was something like “individualism treats individuals as given atoms which interact in an institutional vacuum”. Probably straw-man is overstatement: he focused on neoclassical economics and probably in that field he was right. But he shouldn’t have quoted Austrians, in my opinion.
This post is surely better, and I’m not surprised sbouy iy, as we are within the Austrian tent in which some philosophical issues have always been treated with sophistication.
However, if the thesis is that Mises and Hayek practised methodology better than they preached in theory, let’s quote them and see if it’s true. It’s a couple of years that I don’t read Mises’s methodological writings, but I remember nothing that can be criticized…
Focusing on Human Action, with the caveat that I may have forgotten something, I agree with Prof. Koppl. However, if, as he says, “{Mises} got a pretty sophisticated version worked out in Human Action, one that suffers little or no damage from Gene’s insights”, I don’t see where’s the point to discuss.
Gene, another comment on the Critical Realists, there is a lot of overlap with another CR, Critical Rationalism especially on the topic of realism. How come Lawson gives the impression that they have invented realism when it looks like recycled Popperism?
http://www.criticalrationalism.net/2010/03/06/two-schools-of-cr/
Rafe, I have never seen or heard Ton claim to have invented realism! And the term itself, and the philosophy being “implemented, come from Roy Bhaskar.
Oh, I meant to write “Tony” not “Ton.” In any case, Bhaskar’s work is highly Kantian, something I don’t really see in Popper. And Lawson cites Popper 14 times in _Economics and Reality_, so you’re claim that the CR people don’t mention Popper is just false.
Readers might be interested to take a look the following two papers, where two prominent critical realist economists engage – critically but constructively, I think – with Popper:
(1) Lawson, T. ‘Social Explanation and Popper’, in Tom Boylan and Paschal O’Gorman (eds.), Popper and Economic Methodology: Contemporary Challenges, London and New York: Routledge. Available online at http://www.econ.cam.ac.uk/faculty/lawson/PDFS/Popper.pdf ; and
(2) Runde, J. (1996) “On Popper, probabilities, and propensities.” Review of Social Economy, 54(4): 465-485. Reprinted in S. Fleetwood (ed., 1999). Critical Realism in Economics: Development and Debate. London and New York: Routledge.
I think that a reasonable reading of these papers ought to dispel the notion that ‘gives the impression that they have invented realism when it looks like recycled Popperism’, which seems to me at least to be a little harsh.
Gene asks:
“Right. It is at this point that Lukes asks, “So why are they calling this methodological individualism?””
Because there were economists (and arguably are economists) who don’t satisfy this criterion, such as old-school hydraulic Keynesians.
To take your face painting example, they wouldn’t even try to answer the question, “Why did he paint his face like that?” Instead they would say something like, “The number of painted faces bears a strict relationship to the amount of paint in the community, divided by the average surface area of an adult face. We can write F=alpha*S. In our paper, we have estimated alpha using a data set from Jane Goodall.”
And you can’t even say, Gene, that current Austrians are attacking long-dead bogeys. Remember the econo-physics guys that we took on in our Mises.org article lo these many years ago? They weren’t methodological individualists.
Basically Gene, I think you are wrong when you say the Austrian theory violates the practice. You only get that result, when you narrowly restrict what the theory says.
Are you claiming that correctly understood Austrian theory–which can easily handle a guy painting his face because of custom–is still a pointless restriction on the set of possible social theories? If so, maybe you’re right, but so far I’m not sure exactly what you’re claiming.
No, I’m claiming that the “sophisticated” version is not, in any serious way, methodological individualism! Once you allow things like social practices, traditions, language structures, etc., etc. into your explanations, calling them “individualist” is odd — as Lukes points out. And, he makes the point that one needn’t embrace something called “methodological individualism” to reject holism and historicism.
By the way, here is Lawson stating a view that I believe is:
a) correct; and
b) not methodological individualism:
“Human agency and social structure then presuppose each other. Neither can be reduced to, identified with, or explained completely in terms of the other, for each requires the other.”
Gene is not alone in claiming that Austrians’ practices are not reasonably described as exemplofying ‘methodological individualism.’. For similar arguments, see for example Geoff Hodgson (2007) ‘Meanings of Methodological Individualism.’ Journal of Economic Methodology, 14: 211-26, and even yours truly, namely Lewis, P.A. (2005)‘Agency, Structure and Causality in Austrian Economics: Tensions and Resolutions.’ Review of Political Economy, 17: 291-316.
I think Bob Murphy, Fred Sautet and Roger Koppl are right here. I think Gene is confused about what methodological individualism is.
I think Gene’s comment about natural sciences is also wrong. Gene wrote:
“The entire science of thermodynamics rests on the aggregation of particles into a statistical ensemble; as Werner Heisenberg noted, it is hardly sensible to ask what the ‘temperature’ of a single atom is. Evolutionary theory explains the traits of individual organisms by looking to the history of the species of which they are a member, a distinctly anti-reductionist enterprise.”
Both Thermodynamic and Evolutionary theory are *reductionist* enterprises.
In Thermodynamic the activities at the level of atoms are *not* ignored. Rather the theories of thermodynamics explain how temperature, pressure and so on arise from the behaviour of atoms. Similarly, evolutionary theory explains in terms of genes and the fate of individual creatures how species evolve. So, these are both examples of reductionist scientific theories.
An example of a non-reductionist theory would be the work done by NASA on stress fractures. This work isn’t grounded in a molecular understanding of how material behave. Instead the scientists took many materials and induced metal fatigue in them. They then plotted many thousands of graphs to correlate one behaviour against another. A set of “curve fitted” equations were then constructed.
“exemplofying”
Folks, before you criticize Paul here, remember, he is a foreigner, and is doing the best he can with English.
BTW, I should have cited Paul earlier as well, as his paper was one of the first to make me start questioning MI — although at first I rejected his conclusions, with many of the arguments being made here against my post!
I am a proud Welshman!!!! 😉
“I think Bob Murphy, Fred Sautet and Roger Koppl are right here. I think Gene is confused…”
So when Roger said, “Gene is right,” he meant “confused”?
And I guess Heisenberg was confused about physics as well. And when the great biologist Ernst Mayr wrote “To equate these reductionist views with the theories of the evolutionary synthesis is unjustified, however, as I pointed out in a critical review of similar statements published by M. W. Ho and P. T. Saunders (Mayr, 1984b)” he was confused as well.
At least I’m in good company.
Are the comments here showing up in a weird order for other people as well? I’m seeing my responses now to comments I can’t see!
I share Gene’s esteem of Paul Lewis’s paper. It becomes a matter of personal preference whether “sophisticated methodological individualism” is “really” a properly labeled “methodological individualism” at all. In chapter 2 of my Big Players book I say, “I use the term ‘methodological individualism’ with some misgivings. It has come to hurt, not help communication.”
I think sophisticated methodological individualism should count as methodological individualism. But the discussion can quickly sink into squabble over words. Sometimes words are worth fighting for as with IMHO “liberal” and “hermeneutics.” I think this is such a case, but we should avoid too much passion over word usage, at least in this case. The same guys who came up with it in the first place eventually gave us a pretty sophisticated version, so we should probably use the label for certain views Paul and Gene think are *not* MI. Eh!
Leland Yeager taught me the nub of MI. It’s a check on your model You can use any “aggregate” or “whole” you please. But ask what your model is saying about who does what. Does your model impute implausible behaviors to individual actors? If so, it is probably wrong; if not, it might be right. Cost push theories of price inflation turn out to imply that money holders are indifferent to the purchasing power of their money holdings. That’s not plausible, which is why demand push theories are better. You don’t have to be very smart about social ontology to make such a plausibility check, and making such a check is the real point of MI.
Ack, I meant “demand pull” of course.
Yeah, weird — I now see Roger’s comment correcting what he wrote, but not the comment he is correcting!
“I am a proud Welshman!!!!”
See what I’m getting at? In Paul’s language there are words like,
“Llanfairpwllgwyngyllgogerychwyrndrobwllllantysiliogogogoch,” so this English business has to be a challenge.
If I understand Roger correctly, I think that his position is as follows. The term ‘methodological individualism’ may have had considerable strategic value in the middle of the twentieth century, because at that time Mises and Hayek were attempting to counter people who really did believe that there existed over-arching social forces that shaped society indepdenent of human agency. Now, however, there are few if any advocates of that kind of crude holism, so that the strategic advantages to be had from differentiating oneself from them are more limited. There is more scope, therefore, for employing terms to describe one’s methodological practice that accurately portray what one does. And, as Gene has argued, because Austrians like Hayek invoke social structures as causally efficacious, explanatorily irreducible and emergent features of the social world – none of which it seems to me, are typically associated with ‘methodological individualism’ as that term is currently usually understood – then Austrians might be better off using a new term that leaves them less open to misinterpretation. Dear Roger makes just this point on p. 35 of his Big Players book, where he writes that, ‘I use the term ‘methodological individualism’ with some misgivings. It has come to hurt, not to help, communication.’ Bruce Caldwell adopts a similar position in Hayek’s challange, writing that, ‘His [i.e. Hayek’s] agent is a real human being who inhabits a specific social space [which] … contains institutions that are both the product of and a constraint on the agent’s and other agents’ behaviour. For these reasons, Hayek is perhaps best viewed as endorsing what Agassi (1975) has called institutional individualism or what Pettit (1993) refers to as hoilistic individualism’ (p. 286).
Most of the comments seem to be missing Gene’s legitimate point: If you explain an individual actor’s motivation in terms of some collective phenomena, e.g. social custom, your methodological stance ceases to be purely “individualist.” There is a mutual interdependency at work; neither “pure” individualism nor “pure” collectivism can explain all social phenomena. This is one of Chris Sciabarra’s points of emphasis in Total Freedom: Toward A Dialectical Libertarianism.
‘If you explain an individual actor’s motivation in terms of some collective phenomena, e.g. social custom, your methodological stance ceases to be purely “individualist.” There is a mutual interdependency at work; neither “pure” individualism nor “pure” collectivism can explain all social phenomena.’
Thank you! Now, as I said, Mises and Hayek were too good as social theorists not to grasp this. But then why stick yourself with a label that tends to block that understanding?
The posts are coming in a strange order for me too.
> So when Roger said, “Gene is right,” he meant “confused”?
Well, I didn’t see that post at the time. Let me revise that to saying that I think Bob Murphy and Fred Sautet are right.
I think your all making this far too complex.
Jeffrey M. Gough wrote above: ‘If you explain an individual actor’s motivation in terms of some collective phenomena, e.g. social custom, your methodological stance ceases to be purely “individualist.” There is a mutual interdependency at work; neither “pure” individualism nor “pure” collectivism can explain all social phenomena.’
If I explain an individual actors motivation in terms of social phenomena then I have an incomplete explanation. Since why the social phenomena provoke the actor to behave that way hasn’t been explained. The point of methodological individualism is that this issue cannot be ducked. A writer who supports methodological individualism won’t necessarily justify an individual action so carefully in every case. However, he will recognise that requesting justification is a valid.
The key difference between those who follow methodological individualism and those who don’t is that the later dismiss the need for an individualistic explanation.
For example, in his upcoming book “Zombie Economics” John Quiggin explicitly rejects methodological individualism. He claims that there are too many difficulties in understanding microeconomic phenonmena. So, the only valid approach is to examine the macro level variables.
I would consider this a methodologically individualist story, albeit a highly incomplete one. The individual painter makes a choice, and that choice is to follow a custom.
Now, you might ask where the custom came from. And in that case, you would have to ask about the choices of individuals stretching back generations. E.g., one influential painter in this culture’s history painted in a particular way, and other painters liked it and copied it, and other painters didn’t like it but copied it anyway because they wanted to sell their work, and then they exaggerated the original painter’s style in order to attract more attention from potential buyers, etc., etc.
Methodological individualism does not deny the existence of social outcomes. Nor does it forbid referring to them, at least as a form of shorthand. It says social outcomes, if they are to be fully understood, must be explained in terms of the preferences and actions of individuals.
Gene: “And I guess Heisenberg was confused about physics as well.”
As you noted earlier Heisenberg said “it is hardly sensible to ask what the ‘temperature’ of a single atom is.”
Certainly not. However, the theory of temperature is a reductionist theory. I’m an engineer, and I understand that you are too, so I’m sure you know how the energy of atoms and molecules relate to temperature.
Thermodynamics wasn’t always like that, it took many decade for it to become so. In earlier times the molecular understanding of temperature and other properties was lacking, but we have them now.
Glen – exactly.
“However, the theory of temperature is a reductionist theory.”
Classical thermodynamics was reduced (at least partially) reduced to statistical mechanics. But statistical mechanics uses concepts (like entropy) that occur nowhere in the theory of single particles. The former reduction (as far as it actually was one) has nothing to do with my point — temperature has no meaning at the atomic level, and we must look at a different level to understand it.
“I would consider this a methodologically individualist story, albeit a highly incomplete one.”
And “He preferred it” offers a complete explanation? Isn’t every explanation intrinsically incomplete, in that there is always more that could be added in?
“Methodological individualism does not deny the existence of social outcomes. Nor does it forbid referring to them, at least as a form of shorthand. It says social outcomes, if they are to be fully understood, must be explained in terms of the preferences and actions of individuals.”
Yes, this is like Watson’s “rock bottom” description of MI. But it is arbitrary. Why *must* explanation flow in that direction? Why can’t I explain an individual choice in terms of a custom? What have I gained when I add, “And, oh yeah, the guy chose to follow the custom.” I somehow now understand the situation better than in my formulation? Why is this any less arbitrary than saying, “the preferences and actions of individuals must be understood in terms of social facts”? Isn’t, in fact, Lawson’s formulation more accurate than this one-sided picture that choices are fundamental and society secondary: in fact, you can’t have one without the other.
Choice and culture, choice and culture
Go together like a corpse and vulture
This I tell you brother
You can’t have one without the other
Try, try, try to separate them
It’s an illusion
Try, try, try, and you will only come
To this conclusion
“Classical thermodynamics was reduced (at least partially) reduced to statistical mechanics. But statistical mechanics uses concepts (like entropy) that occur nowhere in the theory of single particles. The former reduction (as far as it actually was one) has nothing to do with my point — temperature has no meaning at the atomic level, and we must look at a different level to understand it.”
Yes, but concepts like entropy and temperature have a relationship to energy and spatial distribution of single particles.
You can take the physics of single particles and derive from it the concepts of temperature and entropy and their meanings.
Gene writes, “Why *must* explanation flow in that direction? Why can’t I explain an individual choice in terms of a custom? What have I gained when I add, “And, oh yeah, the guy chose to follow the custom.” I somehow now understand the situation better than in my formulation?”
And why did he choose to follow the custom? Maybe he painted the face in the custom of his people because he wanted to sell the painting to make a profit? Maybe it was because he enjoys making traditional style paintings? There should be some individualistic motivation at rock bottom, not just society made him do it that way.
“There should be some individualistic motivation at rock bottom, not just society made him do it that way.”
Yes, that’s exactly the arbitrary rule at work that I am criticizing. Why did he want to sell a painting? Because he grew up in a society that commercializes art? Why is it less arbitrary to insist on explanations flowing in this direction than if someone insisted “Preferences must be given a social explanation.”
And by the way, folks, I do understand MI. I wrote a whole book chapter on it, in which I put forward every argument defending it that has been put forward here. When I first encountered Paul’s paper, I countered it in exactly these ways. I both get MI and disagree with it.
Well, Current, I think you are incorrect about thermodynamics, and so do at least some other people deeply learned on this topic, but I realize that there are others who agree with you, and I don’t think we want to de-rail an economics blog on the philosophy of statistical mechanics.
Evolution is much easier — an evolutionary story involves not only genes and individual creatures, but all other creatures of that species, plus all other creatures in the environment, plus the inorganic contents of the environment — so we have a perfectly good scientific theory that explains the traits of an individual by citing a much larger picture. Simply invoking genes an individual one even get a Darwinian story off the ground — remember how much Darwin stressed competition for resources, which means other creatures and the whole environment are essential parts of the picture. And my only reason for invoking these is to show these sorts of explanations (upward pointing) are fine in science.
Hmm, that came out oddly — should have been “Simply invoking genes and an individual one can’t even get a Darwinian story off the ground…”
OK I am totally lost now. Gene, do you think the face-painting story as due to custom is a contradiction of Misesian methodological individualism or don’t you? It seems you are saying, “I know perfectly well Mises would agree with this explanation, so now why doesn’t Mises think this is a good explanation?”
You seem to be flipping back and forth between saying:
(A) “I think Mises is using a misleading label for his methodological views, which are perfectly fine views by the way,”
versus
(B) “Mises was too narrow when he talked about the scope for good explanations in the social sciences.”
To repeat, the face painting example is a terrible one for motivating your discussion, Gene. You and I both agree that Mises would include this as part of “methodological individualism.”
So can you give me an example of what you think is a perfectly fine social theory, that doesn’t obey what Mises himself would classify as methodological individualist?
If not, I think you are basically saying, “Austrians should label their views something else, because otherwise it confuses people as to what they actually think regarding method.”
Why did I get up in the morning and drive to work? Because everyone in society gets up in the morning and drives to work. If I ask myself if that the reason, I will say no. I got up and drove to work because I want to earn money to pay my bills.
Why did I drive on the right hand side of the road? Because society drives on the right hand side of the road. Now it is true that society is shaping my decision to drive on the right hand side of the road. But there is a deeper individualistic motivation than just conforming to society and that is I don’t want to die by driving on the wrong side of the road.
There are rules that society has that shape my decision, but it is individualistic motivation at rock bottom that make me act a certain way.
You can say this or that happens because that’s the local culture. But if the local culture as you describe it requires mothers to prefer other children to her own, then I will question your description of the culture. That’s pretty much what MI is all about.
There is a sense (as I personally learned from Paul Lewis) in which we might say that language has causal efficacy for human action. When in Rome, speak Italian. You didn’t choose Italian. It’s more like Italian chose you! This point should not be taken to deny the humble notion that the linguistic choices of agents in your model should be intelligible. It would not be intelligible, for example, to assume that Tuvan is spontaneously adopted as the linga franca of tourists. Saying they each “chose” Tuvan would explain nothing. On the contrary, you explain the choice of English as linga franca in part by the independently existing role of English in the world, a role not chosen by any of the agents in your model. Thus, your “methodologically individualistic” explanation of language choice “reduces” in part to a social structure.
Yeah but if we want to explain that social structure, won’t we “reduce” it to individual actions? I don’t think so. At every point in the story of the evolution of English and it’s role in the world, the individuals in you model will be responding the English, or Middle English, or whatever, as an existing social structure. No “final” reduction to asocial atoms happens.
If you think my comments go against MI, then you are not giving MI the meaning I think it properly has. Whatever meaning we stipulate, however, the point about social structures remains true. If that’s not “really” MI, so much the worse for MI.
“To repeat, the face painting example is a terrible one for motivating your discussion, Gene.”
Quite to the contrary, the discussion makes it obvious to me it was an excellent choice, because it has made it clear that there are two positions, MI1 and MI2, one of which I disagree with and the other I find oddly named.
MI1, represented by Glen, for instance, holds that this example fits within MI, but not really: “I would consider this a methodologically individualist story, albeit a highly incomplete one.”
Explanations invoking society factors are inferior: “[MI] says social outcomes, if they are to be fully understood, must be explained in terms of the preferences and actions of individuals.”
To this position, I respond “All explanations are partial, and nothing is ever ‘fully understood’ — but the priority given to individual preferences here is arbitrary — why is it not equally valid to demand that preferences be explained by culture as vice-versa? Do you know anyone who formed their preferences outside of a cultural context? Did I at some point choose to be American?”
Look at the Lawson quote above again and the difference between his position and MI1 should be obvious.
If Mises holds MI1, he is mistaken.
MI2, represented by Roger, does not try to force all explanations to reduce in this fashion to be ‘full’ explanations. As Roger said, the only significant difference between the position Paul and I are arguing here and Roger’s seems to be one of terminology.
If Mises holds MI2, I say, “Hmm, I wouldn’t call it that, but OK.”
Without doing a lot of analysis and just saying what I think given what I’ve read in the past, I would say that Mises preaches MI1 and practices MI2. But nothing in my argument hinges on that at all.
Tom, I know you think that is true. I just don’t see why you think that is true.
Tom Dougherty: “There should be some individualistic motivation at rock bottom, not just society made him do it that way.”
Gene: “Yes, that’s exactly the arbitrary rule at work that I am criticizing. Why did he want to sell a painting? Because he grew up in a society that commercializes art? Why is it less arbitrary to insist on explanations flowing in this direction than if someone insisted ‘Preferences must be given a social explanation.'”
As Tom Dougherty points out preferences need more than just a social explanation. They need an individual explanation too.
Let’s take a bit of Rosa Luxembourg Socialism, for example. The working classes prefer to be paid more rather than less. So, the working classes form unions to demand higher wages from the capitalists by withholding their labour.
This is superficially reasonable because it is accurate to say that the working classes prefer to be paid more. But, as we all know this isn’t the full story. Groups of workers are offered pay rises separate from each other by different businesses. The simple fact that a group has a preference doesn’t tell us if that preference will ever be acted upon. That’s why we have to attempt to see things through the eyes of the actors themselves.
Gene said:
“Without doing a lot of analysis and just saying what I think given what I’ve read in the past, I would say that Mises preaches MI1 and practices MI2.”
This is what I was getting at in my very first question to you. I am positive that Mises specifically addresses at least some, and possibly all, of the “objections” to MI you brought up in your original post. And I don’t merely mean that Mises in practice does those sorts of things in the later “applied” sections of Human Action.
I am saying that within the methodological section, probably within two pages of his first statement of MI, Mises takes pains to make sure nobody would buy into the strawman objections that have been leveled against MI, some (possibly all) of which you are reproducing in your blog post here.
So if you want to say, “Look at all these pointless arguments, Mises actually agrees with me perfectly, and yet I thought he didn’t because of the silly label he gives to his methodological position,” then OK fine.
But you seem to be doing a lot more than that.
Again, I think focusing on the face painting example is wrong. A clear cut example of a non-MI explanation would be the one I gave, where we explain face painting by reference to aggregates and variables that have nothing to do with individual action at all.
In contrast, your “counterexample” flows through an individual action, and we are quibbling over the psychologizing of *why* the guy acted as he did. But nobody is disputing that the way to explain the face painting, first and foremost, is that the guy painted his face.
Wow, Bob, I myself answered the strawman objections in EFRP. Look, Glen does NOT think I offered a good MI explanation — mine was partial, and a good explanation to Glen has to resolve into individual preferences and action. So MI1 DOES NOT include my explanation as a ‘good’ MI explanation. Tom D. makes much the same point. And this was exactly how Watson phrased his version of MI, real, ‘rock bottom’ explanations have to be at the level of the individual.
Bob, I get the theory. I really, really do get the theory. Really, really, really.
And I think it is wrong.
1) “preferences need more than just a social explanation. They need an individual explanation too.”
Ok, try this out:
2) “Preferences need more than just an individual explanation. They need a social explanation too.”
Is there some reason 1) is ‘more true’ than 2)? If not, then Glen and Tom’s version of MI is mistaken.
Bob, just look at what Glen, Tom, and current are arguing here, and you’ll see what I’m arguing against: They all feel explanations at the level of the individual are somehow ‘better’, ‘fuller’, ‘more complete’ or something like that than are explanations citing, say, a custom or practice.
It’s clear that they see just where the disagreement lies. I think their preference is arbitrary. And I haven’t seen any justification for it, just a repeated insistence that I really ought to agree!
Thanks to Paul Lewis for demonstrating that I am well behind the play on Lawson and developments in CR. Sydney Uni library does not have the proceedings of the Galway conference yet and I have not found anything on line. The paper that Paul linked gets to grips with the work that Popper did in the social sciences and it indicates some of the difficulties that arose when he gave up on economic laws and opted for a rather confused model lifted from physics. Because this is strictly off topic I will not pursue it beyond thanking Paul again.
Gene, I don’t want to pursue this either but you can think of Popper as a modified Kantian, a “fallible apriorist” in the language of Barry Smith.
OK last post and then I must retire from this debate; I have some face painting to do.
Gene, I’m sure you are getting frustrated for the mirror image reasons, but it seems that this happens a lot. At the risk of annoying you, here is my summary:
==========
GENE says, “Man Mises was a fool. He thinks X.”
MISESIANS say, “What are you talking about Gene? Mises never said X. Mises said Y.”
GENE says, “I know Mises never said X, I have been reading Mises since I was 10. And anyway, Y is stupid.”
===========
I submit Gene that the reason everyone is getting all tripped up on your face painting example is because you didn’t really spell out exactly what your “explanation” is. (That’s not your fault, it’s a quick blog comment.) So if you elaborated and it became clear that you were actually offering a theory of the face painting being due to individual action, which itself could have been influenced by other things, then the Misesian has no problem with it. On the other hand, if upon your elaboration it becomes clear that you are sidestepping the fact that custom flows *through* the individual agent in order to “cause” face painting, then the Misesian doesn’t like it.
I really really really get a lot of your points on this thread, I really do. But it’s a pain to argue with you because you often seem to be holding contradictory objections, so that if I get you on one point you act as if you were never saying it.
(And I concede that every person in a blog argument probably does this, including me. I’m just saying I pay special attention to your objections since you are smart and have read a ton of Austrian literature, so I go nuts when it seems you are misrepresenting Mises’ views in order to knock them down.)
I will try to keep this comment short, but that won’t be easy.
Gene, you say that I represent “MI1” and Roger represents “MI2”. I would be willing to bet that Roger’s and my positions are 95% the same. What do we have in common? A realization that methodological individualism is a relatively minimal constraint. It doesn’t require nearly as much as a lot of people seem to think it does.
Among other things, it doesn’t require that analysis stop once we’ve reduced things to individual actions and preferences. It’s perfectly acceptable to try to explain how actions are chosen and preferences formed. The MI requirement is that the explanation must go through the individual, not that it must stop there.
To take the painting example again, the key element is that you want to know why the individual painter chose to paint as he did. Custom didn’t lead directly to the painting; the line of causation goes through the individual to the painting.
Moreover, what is this “custom” of which you speak? Turns out it consists of lots of other individuals painting in a particular way, other individuals admiring those paintings, etc. You cannot define a “custom” without reference to individuals.
This, it turns out, is the primary justification for MI: it prevents us from asserting the existence of social wholes that cannot be directly identified. Societies, nations, customs, and languages cannot be apprehended directly; they must be defined, and that definition will be in terms of individual actions and preferences. (This is also the reason your analogies from the physical sciences don’t work. In the physical sciences, it is often possible to identify a whole directly; a mitochondrion, for instance, can be identified under lab condition without breaking it down all the way to level of quarks. The same cannot be said of, say, “the general will.”)
MI is not exhaustive of all the things we want from good social science. An explanation is not satisfying merely by virtue of its methodological individualism. We could easily have an MI-consistent theory that is based on incredibly unrealistic assumptions about preferences (for instance, that everyone loves work and despises leisure). MI is a necessary, not a sufficient, condition.
Lastly, I will disagree with Roger on one point. He said, “‘Yeah but if we want to explain that social structure, won’t we “reduce’ it to individual actions? I don’t think so. At every point in the story of the evolution of English and it’s role in the world, the individuals in you model will be responding the English, or Middle English, or whatever, as an existing social structure. No ‘final’ reduction to asocial atoms happens.” Leaving out the ‘asocial’ part, I think we can, in principle, trace language to its origins in the behavior of our pre-linguistic hominid ancestors.
But Roger’s right that there will always be a social context; before language, for instance, there were grunts and facial expressions that tended to reflect emotional states, and these were the context in which language began to form. Each individual responds to the social context into which he is born — a context created by the actions of other members of his species. This is perfectly consistent with MI.
I agree with Glen.
Whew!
I think Glen is right. He wrote a great paper on the compatibility of group selection and methodological individualism. (He’s posted it here: http://www.csun.edu/%7Edgw61315/GSMIfinal.pdf.) It appeared in vol. 7 of Advances in Austrian Economics, along with comments on the paper by Dick Langlois, Adam Gifford, Todd Zywicki, and Sober & Wilson, plus his reply. (http://www.csun.edu/%7Edgw61315/GSMIrejfinal.pdf)
IMHO Glen completely destroyed Sober & Wilson on MI. I think that episode illustrates the continuing importance of MI. After Aggassi (1960), properly informed opinion on MI doesn’t vary much from one scholar to the next. And yet we have wonderful, important, open minded scholars like David Sloan Wilson and Elliott Sober getting it wrong on the substance and not just the label. Not all relevant opinion is “properly informed.” I mean, these are serious guys. Wilson rehabilitated group selection and thereby earned an enduring (likely permanent) place in the history of biology. Nevertheless, on this issue they came up seriously short. Whatever judgement we make about the term, the substance of the issue matters and we need to explaining the meaning and importance of MI.
BTW: If language is in part a product of biological evolution, as it seems to be, then we have a clear example of a “social structure” that cannot be “fully reduced” to human actions. Besides the actions of our hominid (= great apes) ancestors, there is the action of natural selection, to which the concept of methodological individualism applies.
[…] קלאהן פרסם ב ThinkMarkets פוסט מעניין אשר עוסק במתודולוגיה כלכלית, ובהרהורים על המתודולוגיה […]
[…] אנושית. אני מאוד ממליץ (למי שיש עוד כוח) לקרוא גם את התגובות הרבות שנכתבו לפוסט הזה. קלאהן מגיב למספר הערות ונוצר דיון ממש […]
“The MI requirement is that the explanation must go through the individual, not that it must stop there.”
Alright, I declare a new doctrine: methodological culturalism. It requires the explanation must go through the cultural. Is this requirement any less valid than MI’s?
“Societies, nations, customs, and languages cannot be apprehended directly…”
“If [MI] means that in the social world only individuals are observable, it is evidently false. Some social phenomena simply can be observed (as both trees and forests can): and indeed, many features of social phenomena are observable (e.g. the procedure of a court) while many features of individuals are not (e.g. intentions).” — Lukes
‘You cannot define a “custom” without reference to individuals.’
And you cannot define individuals with reference to culture! Yes, it is clear: you believe individuals are somehow prior to society, while I think it is plain that they are not. Again, see Lawson’s quote above for what I think is the correct understanding of the relationship here.
An alien comes to observe earth. The first place he goes is the Isis on the day of a rowing race. What he “directly observes” are large, oblong creatures with multiple fins and lots of hairy bumps racing along the surface of the water.
The second place he goes is to a football stadium, where he sees a crowd doing ‘the wave.’ What he directly observes is some sort of giant beast with thousands of tentacles, perhaps trawling the air for food.
He would have to learn that these creatures can ‘come aprt.” As Lukes notes, the contention that we can ‘never observe’ social phenomena is evidently false.
Gene Callahan suggests, following Tony Lawson, that Mises and Hayek’s stress on a single form of explanation in the social sciences was in line with the unity of science movement. It seems to me unlikely that this movement influenced Mises and Hayek, as they both often stress the differences between the methods of the physical and human sciences.
Also, Gene defends his claim that their explanations were in practice better than their methodological individualist prescriptions from the objection that their explanations were in accord with their precepts, not in violation of them, in an odd fashion. He cites Steven Lukes as saying that if one allows these “better in practice” explanations to count as MI accounts, then it is hard to see why they insist on calling these explanations by that name. What strikes me as odd about this response is this: Suppose Lukes is right and it is hard to see why the Mises-Hayek explanations count as MI. It would not follow from this that Mises and Hayek held in theory a stricter conception of MI from which they deviated in practice. All that follows, if Lukes is right, is that they should have not have called what they taught MI.
I wasn’t going to comment on Gene’s post for 2 reasons. First, I didn’t really understand what his problem with MI was. Second, I consider methodological debates to be sterile.
After 72 responses, I go back to the first one from Pietro. What do people think MI is? Or anti-MI
I go back to Menger’s original fomrulation, which is how does one explain how social phenomena come into existence w/o the intent of any one of the actors. They are the uninintended outcome of human action.
MI explains exactly what Gene suggested it does not.
Nor is MI the invention of Mises or Hayek, as some seemed to suggest. Nor even of Menger. Nor of Smith. It goes back to the Spanish Scholastics at least. It is the originary problem of social science without which there is no science.
With respect, Gene’s account is unhistorical. Holism was a response to MI, not the other way around.
I will end where I began. I don’t understand Gene’s problem. And methodological discussions are sterile.
“It would not follow from this that Mises and Hayek held in theory a stricter conception of MI from which they deviated in practice. All that follows, if Lukes is right, is that they should have not have called what they taught MI.”
Correct, David. I have tried to make the point, throughout the discussion, that EITHER you preach MI1, and I believe you are mistaken, OR you preach MI2, and I don’t understand your terminology. As Paul Lewis notes, I am hardly the first person to remark upon this difficulty with MI2.
Here is Hodgson’s abstract (cite in Lewis comment above):
“Advocacy of ‘methodological individualism’ is widespread, especially among economists. However, the term is rarely defined with adequate precision and some crucial ambiguities are explored in this article. Among these is the commonplace ambivalence over whether explanations should be in terms of individuals alone, or in terms of individuals plus relations between them. It is shown that a great deal hinges on this subtle and often overlooked distinction in explanantia. In particular, explanations in terms of individuals alone have never, as yet, been achieved. Furthermore, the more feasible version of explanations in terms of individuals plus relations between them amounts to the introduction of social structure alongside individuals in the explanantia. Serious questions remain whether this version warrants the one-sided emphasis on individuals in the term ‘methodological individualism.'”
Just my point.
Gene,
It seems to me that now you’ve reduced your argument to, “I just don’t like the term, though the actual method as actual practiced by the actual people who actually came up with it and actually used it is great!” Well, I don’t like the term “capitalism,” but I’m not going to write a diatribe saying how capitalism had its time and place, but now we need economic pluralism when all I really don’t like is the term being used, because I don’t think it captures the kind of economic system I think works best (more and more I like Hayek’s term “catallaxy”). And worse, people misuse and misunderstand it. On the other hand, big deal. I suspect that “methodological individualism” as a term will be easier to change than “capitalism,” but in the end, if people don’t like the method or system itself, no name change is going to help you. It’s just petty quibbling. The Austrians always used individualism to mean a socially embedded individual, not as a description of a sociopath. Thus, methodological individualism was always already a social method as well — it just wasn’t a collectivist method.
“t seems to me that now you’ve reduced your argument to, “I just don’t like the term…”
Yes, it may seem that way to you, but that’s nothing like what I said.
“And methodological discussions are sterile.”
Nonsense, Jerry. Why in this very thread, I discovered how few English words rhyme with ‘culture.’
While I appreciate the spirit of Jerry’s comments, it may be worth noting that the doctrine of MI was articulated only in 1909 or so by Joseph Schumpeter. I have read that Schumpeter was pointing to Max Weber’s methodology as given in the first bit of his Economy & Society. Presumably, Weber was reacting, in part at least, to the younger historical school. Thus, I think there is a sense in which MI goes back only about 100 years. And it is true that Mises and Hayek pushed it and are normally associated with it in greater degree than most other economists. But I agree, of course, that examples of appropriate *practice* go way back.
Methodological discussions do have a tendency to get mired in silliness, but so do applied and theoretical economics. It’s just that the silliness comes in different forms. Still, I think it is helpful for each of us to be clear about methods. I think my story about Glen vs. Sober & Wilson illustrates that claim.
Argh I can’t help myself, I need to make another post here because I know that several people are still reading. If I can help just one person…
Gene, it took me a half-day to digest it, but I think I have the answer to your clever examples about an alien who thinks the crowd in the football stadium is a giant creature etc., and that therefore (you claim) this is an example of an observer directly perceiving a “social phenomenon.”
No, it’s not. To the alien observer, it is an individual creature. Note that in your own description, you employed methodological individualism and in fact employed praxeology too, for you had the alien perceive it as a single creature and attributed a *motive* to the creature’s behavior.
As the noted scholar Gene Callahan explained in a recent ThinkMarkets post, what is a “fact” can differ from theorist to theorist. The alien observer operated under the theory that the various movements in the football stadium corresponded to the functional (purposive?) action of a single creature, and so for the alien theorist he observed a “fact” of individual behavior. There was no question of “group behavior” or a “general will” going on.
Now it’s true, some other theorists would analyze the situation differently; they would see a “fact” of thousands of drunk people. And they too would employ an individualist, praxeological analysis to explain each individual’s purposive action.
Now if you want to say that the determination of who the “individual actor” is, can be a thorny issue, I agree. I recently wrote an econ textbook for junior high kids, and in the opening methodological part I was getting bogged down with all sorts of weird cases like split personality (is it one actor or many?), chimpanzees (do they have an ordinal preference scale?), and so forth. There is no doubt that it’s hard to apply the Misesian approach in some of these cases.
But I think the situation is analogous to, say, Mises’ theory of money. In certain cases, it’s not obvious if something is just a medium of exchange, or if we go ahead and call it money. But there’s no doubt that the intellectual framework concerning money qua money is solid, and we just aren’t sure when to apply it.
The same thing with praxeology. There is a huge gulf between a theorist interpreting physical motions as the result of mindless natural processes, versus the result of an intentional being. (I’m just repeating the central lesson of your FEE talk about a year ago.) And if you go down the path of interpreting something as action in the Misesian sense, then certain results pop out. In particular, you must have in mind ONE INDIVIDUAL ACTOR who is doing the action; it doesn’t make sense to say “this is an action in the Misesian sense, but I can’t locate an actor even in theory” or “this is an action in the Misesian sense, but I cannot break it down to fewer than 3 actors.”
So that’s why methodological individualism is a crucial component of Misesian praxeology. It’s like the issue of ordinal preferences. Neuroscientists and others tut-tut Misesians for painting themselves into a corner with their insistence on ordinal utility, but these critics don’t realize that Mises isn’t making an empirical claim about brain chemistry or the nervous system. Rather, he’s making a logical claim about how praxeology works. The implied “value scale” that corresponds to the interpretation of things as purposive actions, necessarily requires an ordinal value scale. Maybe people really do have “cardinal utility functions,” but then you would be explaining human behaviors with a different framework in that case.
So maybe Misesian praxeology will be discarded at some point because social scientists come up with something better. But methodological individualism is a necessary component of it. The essence of MI is not whether individual’s act because of “inborn” motivations or because they can be influenced by other people. No, the essence of MI is that your explanation is built upon the actions of individuals. So a guy painting his face is no counterexample, and an alien interpreting a crowd of fans as a single creature that’s trying to find food is no counterexample.
If you need another rhyme, don’t forget mulcher or even gulcher. Given the culture in which many of us have chosen to participate, they could acquire unexpected (but certainly delightful) extra meanings.
Not that my opinion counts for much but, as I see it, from a standpoint of expediency a strict version of neither approach is complete. While a perfect economics could simultaneously comprehend all gestalt phenomena and the full depth of meaning of each individual’s value scales, both in each instant a choice was made and overall, we do not possess the capacity to construct such a science. The requisite knowledge is beyond human limitations. Within the limits of what efforts are possible both approaches are practically useful. A truly holistic/collectivist approach which makes no consideration of acting man as such has no basic foundation; it is an approach to a field that does not know its own subject and is therefore ultimately worthless.
An individualism which cannot even apprehend the possibility of a phenomenon arising out of something more than any one person’s actions is equally worthless – it is an approach to a field which, in this case, is guaranteed to go nowhere. Phenomena like the structure of production (a point which I think the Austrians are right to pride themselves for articulating a more sophisticated view than some other schools) are completely beyond its understanding.
I don’t think anyone here is guilty of either. Part of the appeal of such terms as culture or custom is their utility in subsuming a great deal of information in a short span of words. While it may be true that it only exists as an extension of each individual human action, that it is our minds forming the pattern which ‘sees’ the culture in action, that does not mitigate its usefulness one bit. If we accept as valid the arguments for strictly ordinal value scales, alongside the impossibility of direct cardinal comparison of values or the measurability of satisfaction, then a perfectly individualist approach – especially in historical analysis – is impossible. It would have to know what it admits it cannot know. It may also be true that a perfect explanation and understanding of social phenomena would ultimately describe everything relevant in terms of individual choices, but I think this is impossible.
Without taking a side in this debate, I have to say, Gene generates great threads. Maybe he has some Penelope gene (pun unintended but not unwelcome). In general, Jerry is surely right that that methodological discussions tend to be sterile. But not here!
I think Bob’s posts have been far and away the best of this exchange.
Certainly it’s not what you initially argued, but at one point you essentially argue that you just don’t like the term. Of course, I think I and most everyone else here have pointed out to anyone open to listening to evidence and reason that you misunderstand MI, particularly the fact that it was always already social in nature, regardless of the use of the term “individual.” Well, that’s not entirely fair. On the one hand, to the extent that you think that they violate the principles of the method they created and used in a consistent fashion “incorrectly” — meaning, using it for a social mammal rather than a sociopath — you do seem to understand the method. But to the extent that you think they could possibly be using their own method incorrectly precisely because you apparently correlate individualism with sociopathy, then you don’t have a clue what you’re talking about. MI is a theory of human beings as a social mammal. It always was. It is one of the things that differentiates the Austrians from the economists who prefer using the sociopath known as Homo economus.
Let’s suppose we take up Gene’s “Methodological Culturalism”. So, each preference has a social cause. How can we do that without also using “Methodological Individualism” too? After all each preference is personal, even if it is cultural too.
I see the important point of “Methodological Individualism” as being to point to informal social theorising using generalities. In Gene’s face painting example there is a whole theory behind the word “because it is his culture”. The point of the methodological individualism is to recognize this and to recognize that the social theorising involved (which may be contained in our language) is open to criticism.
I think I have to disagree here. If a choice system in has multiple corner solutions we can see situations that can be described as “Germany invaded Poland” the choices are all made individually, but the outcome seems collective. This is in part because my utility is influenced by the choices around me.
Languages are a good example, people within a geographic location tend to speak the same language because what language you end up learning is dependent on who you are trying to communicate with. For small children it is primarily their parents, but the adults also want their kids to be able to speak to other people in the area. This individual choice (I mean they could have taught them Klingon instead) seems to create a mass groups of people who are making a single choice, but its really individual choices that happen to be reinforced by nearby individual’s choices.
For cases that are corner solutions, it makes little difference if we say “the collective did x, or y number of individuals al did z together so it looked like x.” But in cases where the solutions aren’t corner solutions, and the system doesn’t discrete-ify itself, we can speak of groups as if they are single entities but its not obvious what we mean. We can say, “the american public likes hotdogs,” what does this mean? Don’t a lot of people hate hot dogs? What percent of people calling themselves ‘Americans’ actually like hot dogs? What percent is enough for it to be true? 51% 90%? Maybe just 30% if no one has really strong negative feelings about hot dogs.
This is what I call “self reinforcing solutions.” The more people believe the solution, the more likely others are to believe it. So a large group collectively has the same solution. We can use this, methodological individualism and praxology provide the answer. Solutions that are self reinforcing we can speak of collectives and solutions that are not self reinforcing we must only speak of individuals.
I had to opt ought to get some other work done, but I just want to wrap up with a few remarks:
Roger: “While I appreciate the spirit of Jerry’s comments, it may be worth noting that the doctrine of MI was articulated only in 1909 or so by Joseph Schumpeter.”
Right. And while the Late Scholastics used individual explanations, that’s not what MI is — it’s the doctrine that only MI explanations are valid, and the Late Scholastics couldn’t believe that, since, as Catholics, they had to believe in a huge non-individual actor in history, namely, the Church.
David G: “It seems to me unlikely that this movement influenced Mises and Hayek, as they both often stress the differences between the methods of the physical and human sciences.”
David, you seem to making the startling claim that, if thinker A disagreed on some point with doctrine D, then it is unlikely that D influenced A. But surely thinkers are influenced all of the time by doctrines with which they disagree in some way or another.
“Of course, I think I and most everyone else here have pointed out to anyone open to listening to evidence and reason that you misunderstand MI…”
Right, Troy. Except somehow I understood it perfectly well when I wrote my book. I must have been hit in the head or something. And so must Paul Lewis, and Tony Lawson, and Geoff Hodgson, and Bruce Caldwell. Have pity on us brain-damaged ones!
“No, it’s not. To the alien observer, it is an individual creature. ”
Right, Bob. Sometimes social phenomena display a unity that makes them appear as an individual, active (not I didn’t say acting entity.
You have just conceded my case. Thank you, thank you, QED and all.
Gene, Mises and Hayek explicitly rejected the central tenets of reductionism; they did not just disagree with this view “in some way or another”. If you think that they nevertheless embraced the view that all social phenomena should be explained following a single explanatory framework under the influence of the unity of science movement, I’d say you bear a substantial evidential burden.
Gene,
You have committed several logicla fallacies here.
One, having published a book proves nothing except that you can find a publisher to publish your book. I have a book. I have a couple upcoming publications on spontaneous order — but I don’t replace logical discourse with those facts.
Two, people can and do change their minds from what is true to what is untrue. I see it all the time. In the same way that bad money drives out the good, bad information or ideas can and sometimes do drive out the good.
Three, you have committed the expert fallacy. Pointing to experts is no replacement for reasoned argumentation. Unless you’re a postmodernist who rejects logic and reason, that is.
In other words, you haven’t proven to me (or, apparently, to many other people here) that you do in fact understand that MI is a theory/approach that thinks of humans as social beings — as its used by Mises and Hayek, for example, prove. Austrian MI is different both from Homo oeconimus of classical economics and the collectivist approach of Marxism and other socialist approaches. Unlike each of these approaches (you seem to wrongly conflate the former with MI), MI is a theory that sees humans as social and is specifically nonreductionist. That, after all, was one of the things the Austrians were about: a nonreductionist approach to economics. Thus, their opposition to scientism.
YOu might also consider the fact that you seem to be talking about Methodological Atomism rather than MI, as discussed here:
http://hope.dukejournals.org/cgi/pdf_extract/39/1/47
For a contrary take, check out Eliezer Yudkowsky on reductionism.
“but I don’t replace logical discourse with those facts…”
No, that’s true — you replace logical discourse with an endless repetition of “You don’t get it. You don’t get it.”
YOU might consider that the “thing” you keep telling me I don’t get, in fact, I stated RIGHT IN THE INITIAL POST THAT STARTED THIS THREAD: “Now, it may be objected, to the above, that Mises, Hayek, and other Austrians were well aware that individuals are socially embedded, and that their choices are ineluctably situated within a social context. Just so, I respond!”
It seems to me that an impasse has been reached. I and others argue that MI is inherently social and was intended to be understood as social — and that the way that the people who developed MI used is proves precisely that — and you argue that it is not. Apparently, no evidence anyone has provided will convince you, and you haven’t provided any evidence that convinces us. I argue that you can’t tell the difference between MI and Methodological Atomism. You don’t respond — to that, or to any of the actual evidence provided. Thus, it seems, we are at an impasse.
[…] the neuron as their unit of analysis. In his ongoing quest to alienate his old Austrian friends, Gene Callahan has pronounced methodological individualism to be alright for its time but wanting in comparison to […]
Individuals reach their goals (e.g. make a ‘good’ painting) by catering to the consumers. Calling that set of consumers ‘society’ does not refute individualism.
Can any social science, or indeed any “hard” science, actually have legitimate grounding outside The Word (or its meta-equivalent “God Particle”)?
Check out this chapter of this book: http://www.sagepub.com/upm-data/27354_02_Brown_&_Stenner__CH_02.pdf
Also, consider this bit from the article above: “…social explanations should be framed in terms of whatever level of reality makes the phenomenon in question most comprehensible”
That seems a lot like dumbing things down in a Philosophy of As If kind of way.
One more thing: doesn’t methodological individualism have more to do with psychological motivations and praxeology than with post-hoc statistical analysis (which of course also necessitates some human action to “crunch the numbers”)? Didn’t each of those German soldiers need to take action as an individual before anyone could ever analyze what “the German Army” did?
Bottom line: the analysis of previous praxeology does not transform the past into having been some kind of aggregated group action, although that is the way in which people best seem to be able to remember what happened (i.e. memory must by biological necessity “dumb things down” inside the brain). Otherwise it would be ridiculous to ever speak of any individual ever doing anything at all (history would need to be expressed as one big temporal mega-word similar to the fictional language of Tlön: http://www.coldbacon.com/writing/borges-tlon.html
Ok, after digesting some of the commentary discussions, this all appears to be a case of desired reconciliation for the “splitting” of man.
Adam into Adam-and-Eve, Nights 482-531 of 1001 Nights, etc. For that matter, take a look at any number of modern film scripts having to do with identity and self and loss and redemption (e.g. The Zero Theorem). For every Jekyll there’s an alchemical something-something monster that demands special attention (on a more fundamental level is the analogy of historical rivalry between grammar and dialectics).
Does Homo Economicus exist? What trigger goes off when someone sees the Queen of Diamonds? Who is the real Slim Shady?