Geoff Hodgson on Methodological Individualism

16 July 2007 at 9:51 am 10 comments

| Nicolai Foss |

Geoff Hodgson is no doubt a very thoughtful economist. I admire much of his work. But I have always been disturbed by a sustained theme in his writings: His relentless criticism of methodological individualism. To me, MI is “trivially correct,” to paraphrase Jon Elster, and I have viewed Hodgson’s critiques as bizarre and idiosyncratic, particularly because I have not thought that he provided any good reasons to reject MI.

However, the Journal of Economic Methodology has just published a very interesting piece by Hodgson, “Meanings of Methodological Individualism,” in which he offers some serious reasons why MI is problematic (and perhaps more than that). Hodgson argues that MI are surrounded by a number of ambiguities: 1) It is unclear whether it is intended to be something that is specific to “pure economics” or to the social sciences in general; 2) it is unclear whether MI is about social ontology or about social explanation, and 3) it is unclear whether it refers to “explanation in terms of individuals, or indivuals alone.

Now, 1) doesn’t really seem to me to be an ambiguity. While indeed Schumpeter, the inventor of the term, thought of MI as something that applied to pure economics alone, it is quite clear that modern proponents of MI think of it as applying generally to the social sciences. 2) is a red herring, for while MI is about explanation it is rooted in the ontological argument that only individuals act. 

However, Hodgson is right in arguing that 3) represent an ambiguity in the history of MI. He himself takes the position that explanation purely in terms of individuals is impossible and points (drawing on Arrow) to general equilibrium as a theory that, although it has been claimed to explain solely in terms of individuals, actually relies fundamentally on assumptions on social structure. The position that a social science explanation involves individuals as well as the structures within which they interact has been called “institutional individualism.” It may be found in the famous Coleman diagram in the argument that “downwards causation” from structures to the conditions of individual actions is possible (see the first chapter of this) (for this reason, Mario Bunge has argued that Coleman is really a “closet systemist”, i.e., semi-holist).

Hodgson argues that “institutional individualism” isn’t individualism proper. He is right that the position looks like a confused half-way house. But is it really? The structures or institutions in question are themselves and in turn explainable in terms of individuals. Of course, it may be argued that these individuals in turn interacted within structurs, so we get bogged down in the wellknown regress problem of individuals and structures. This argument, I submit, is ultimately another red herring, as only individuals act, not structures or institutions. The idea of a regress gets this wrong.

Entry filed under: - Foss -, Methods/Methodology/Theory of Science.

Education Quote of the Day John Hagel’s Blog

10 Comments Add your own

  • 1. Steve Phelan  |  16 July 2007 at 4:15 pm

    Doesn’t Giddens’ theory of structuration address this point:

    As Giddens puts it, ‘Society only has form, and that form only has effects on people, in so far as structure is produced and reproduced in what people do’ (Giddens & Pierson, 1998: 77).”

    So people’s everyday actions reinforce and reproduce a set of expectations – and it is this set of other people’s expectations which make up the ‘social forces’ and ‘social structures’ that sociologists talk about. (

    Giddens’ point is that individuals are not free to do anything they want. They are constrained by social norms governing how they interact (right down to the level of speech and thought). These norms do not emerge de novo but are always the product of human action (albeit past human actions in many cases) and, as a result, can always evolve and change.

    I think there is plenty of room for Austrian style subjectivism in structuration. All agents are not socialised in identical ways so everyone will have different norms and thus different expectations.

  • 2. jonfernquest  |  17 July 2007 at 6:52 am

    “…the famous Coleman diagram in the argument that “downwards causation” from structures to the conditions of individual actions is possible”

    Thanks for this lead. Historiography since Braudel has tended to over stress the influence of structure (as opposed to human agency). There was a move away from narrative history where human agency overshadows the influence of structure, but has since moved back to thicker narratives that incorporate a mix of structure and agency, but alas not in the geographical area I specialise in which is still stuck in a Braudellian never never land without human agency, a sort of cultural determinism.

    With this Coleman diagram it’s interesting to see a more abstract sociological analogue to issues in historiography, that could perhaps be applied productively to historiography.

  • 3. jc  |  17 July 2007 at 10:42 am

    Thanks for bringing Hodgson’s recent paper to our notice; though it seems to sustain the bizarre-ness noted earlier. While Hodgson’s bibliographic research is certainly useful to us all, especially in the way it presents his determination to get to the bottom of the various ‘meanings of MI’, the paper threatens to become a collection of tautologies sitting atop his bibliographic notes. Now none of us ever does any better, since all theorizing is tautologous – as Hodgson himself allows at the bottom of page 7 – but the paper shows us something of the difference between getting bogged down in tautologies and getting some intellectual leverage over them.

    To illustrate, to say – as in Section 6 – that ‘social phenomena have not ever been explained in terms of individuals alone’ is tautologous in the sense that implied is a discourse based on individuals alone, in which case what can we possibly mean by ‘social phenomena’? Actually there is nothing much wrong with saying, in a world that denies every relationship aspect of ‘social phenomena’, that what we might mean by ‘social’ is simply the aggregation of individuals between whom there is no interaction.

    Unfortunately while this is a perfectly viable ‘pure’ abstraction, it can have little to do with our experience of the world – be that social or physical. Even inorganic entities, such as planets, grains of sand or atoms of gas, interact when massed, and it is their order that requires explanation. Thus an MI explanation based on ‘individuals alone’ ignores and so denies the concept of order among these individuals, and so makes explanation irrelevant. For once we have assumed the existence of the individuals (i.e. relieved ourselves of explaining their existence) only their order calls for explanation. The kind of extremist MI Hodgson implies here is totally irrelevant, whether as the basis of a description of our world or for theorizing – the two kinds of explanation Hodgson cites Schumpeter as allowing. Consequently it is not clear to me that the ‘meaning’ Hodgson invokes here has ever been proposed by any serious commentator and in that sense he dismisses a ‘straw man’ of his own building.

    Which gets us to the real agenda behind MI, well captured in the Popper quote in Section 2 about ‘never being satisfied with an explanation in terms of so-called ‘collectives’. But, conversely, there can be no explanation of social or economic activity without some ordered social dimension. The real questions raised by MI are to do with theorizing the interactions (or modeling the relations) between the individual and social entities one assumes – in the manner of the ‘elements which are taken as given’ – and the interactive phenomena one observes, which thereby drags in the ‘ontological’ red-herring issue Nicolai notes.

    What exactly is meant by ‘the individual’ or by ‘social relations’? One cannot make much progress towards understanding MI if you leave these prefacing issues out of account. Otherwise you end up comparing utterly different notions and supposing you get to some intelligence. Thus I feel it absurd to compare, say, Coleman’s notions with Lachmann’s since their axiomatic bases are utterly incommensurate.

    Can one say anything useful about MI beyond simply expressing one’s assumptions? I think the answer is yes, obviously. The reason is that MI’s real agenda is the economist’s way of thinking about the phenomena of social order and relations. Both Hayek and Schumpeter have something profound to say to naive collectivists of the Durkheimian tradition who have washed the Enlightenment notions of individualism and agency out of their discourse. Likewise ‘pure’ economists are inclined to wash collectives out of their discourse on the grounds that all social phenomena are, or should be, manifestations of ‘invisible hands’. Thus MI is a profoundly intelligent and empirically grounded search for a middle ground that admits both agentic individuals and the social structures that delimit or in some other way shape those individuals’ agentic opportunities and actions.

    But the search is complicated by the chicken-and-egg ontology issue to which Hodgson alludes on page 7. Either you start off assuming that both societies and individuals exist and that you do not need explanation for either – all that you are trying to explain is their interaction – or you feel some compulsion to reduce one assumption to the other, to explain society in terms of its constituting individuals or vice-versa.

    Personally I see this reductionist ‘grand theorizing’ as part the compulsive aspect of our contemporary ‘scientific’ pathology, our need to frame social and economic enquiry ‘rigorously’ in terms of a ‘pure natural science’ that does not, as we know, even operate among those who do real natural science. Thus I see the essence of the MI discussion which Hodgson reports as a social realist’s (i.e. one who accepts a certain ontology of social order and relations) attempt to focus on understanding the interaction between individuals – as assumed – and social collectives – as assumed. This conversation throws up the supporting insights to be gained from economical thinking, as distinct from those of political theory and sociology and might yield immediate policy possibilities not evident in those discourses. Thus Agassi’s ‘institutional individualism’ or Hodgson’s ‘individualistic institutionalism’ both point us to a middle ground which re-invokes the origins of economics itself, i.e. ‘political economy’, the study of how thinking based on rational actors who act in their self-interest bears on a society grounded on political power.

    As I see it, the way to move the MI agenda forward beyond tautological circling is to surface the assumptions to be adopted i.e. the ‘model of man’ and the complementary ‘model of social order’. We know already that there are many models of man, especially ones other than the ‘rational economic man’ which underpins much neoclassical economic discourse. Likewise there are many models of society, especially ones other than the naïve Durkheimian or the social contract models alluded to in Hodgson’s paper. This variety will give rise to several branches of MI. Personally I think only a small subset of the various models already present in the literature are capable of providing the middle ground I believe drove Schumpeter and Hayek to frame the MI agenda as an expression of their fundamentally political concerns. The point of raising MI again is surely to help us identify these economics-based possibilities rather than to back off into ‘purism’ and abandon pondering the reality our circumstances in the hope of resolving the axiomatic contradictions necessary to create the middle ground they sought.

  • 4. Paolo MARITI  |  17 July 2007 at 11:57 am

    I think that the following Foss’s passage is paradoxicallymuch clarifying:

    …” Hodgson is right in arguing that 3) represent an ambiguity in the history of MI. He himself takes the position that explanation purely in terms of individuals is impossible and points (drawing on Arrow) to general equilibrium as a theory that, although it has been claimed to explain solely in terms of individuals, actually relies fundamentally on assumptions on social structure. The position that a social science explanation involves individuals as well as the structures within which they interact has been called “institutional individualism.”
    Hdgson is not right As many do, he fails to distinguihs the variuos stages of analysis:
    1)individuals have their goals and constrains on their behaviour. Within such constrains, what is the best way to achieve goals?
    2)assuming that individuals know how to solve the problem and behave accordingly, the solution is found. Constrains are then changed to see if solutions vary.
    3) an Individual’s solution affects other individuasl constrains
    e.g. a walraisina eq

  • 5. P M(continued)  |  17 July 2007 at 12:11 pm

    e.g. a Walrasian economic system (or institution)is one in which individuals interact, taking market prices as given In other circumstances individuals interact, taking actions of others (and rules) as constrains

  • 6. JC  |  17 July 2007 at 2:07 pm

    The ambiguity of (3) is not an ambiguity in the sense of being something indefinite or under-determined.

    It is the contradiction of assumptions that creates the conceptual space for the MI discussion. Thus we have, for instance, independent rational self-maximizing individuals on the one hand and power-manifesting social structures on the other. The axioms underpinning these two assumptions are incompatible – i.e. incommensurate in the philosophical sense. This makes the interaction nonsensical in the ‘pure theory’ sense. But it makes it relevant to the human condition which is, after all full of contradiction.

    Thus the various possible MI discourses offer analyses of a different type – different, that is, to a political theory-based analysis or most sociological analyses.

    If the axioms underpinning the model of man and the model of social structure is made consistent, the MI-conceptual space collapses.

  • 7. jonfernquest  |  18 July 2007 at 8:19 am

    “…the way to move the MI agenda forward beyond tautological circling is to surface the assumptions to be adopted i.e. the ‘model of man’ and the complementary ‘model of social order’.”

    I agree, but I would see this as zooming up a lot closer to an actual “man” embedded in a much smaller “social order”

    To get concrete, there is always a tension in a large corporation between standardising the different parts and making them controllable (a social order if you will), and replaceable (perhaps through IT) and, on the other hand, instilling the humans who work in the organisation with a sense of initiative and purpose, making them believe that they themselves can transform the organisation and add value to it, not just lay their daily union allotment of bricks and go home or play computer solitaire for the rest of the day. This seems to be particularly important in media companies that derive their income from creativity.

    I wonder what MI would make of the Battle of Waterloo, probably the one historical event analysts tried to pick apart microscopically almost from the day it finished (Siborne, 1844), an attempt at reductionism to the actions of individuals that still fails to show how the whole is a sum of the parts, one of the most important parts, the arrival of the Prussian army, was not even known until recently, and then seems to have a very chance character to it.

    Anyway, my point is, even if you get every social actor in the events you are trying to explain, write down what they experienced, which is what happened almost immediately after Waterloo, hundreds of years later, if you’re still trying to figure it out, you still might not get it right. Even if as Elster says “all social phenomena (their structure and their change) are in principle explicable only in terms of individuals – their properties, goals, and beliefs’ we are still very limited in our capacity to do so.

  • 8. jc  |  18 July 2007 at 12:35 pm

    It is interesting that you think of the Battle of Waterloo, and thereby bring in history and ‘historical explanations’. The Methodenstreit was a battle between those like Menger, who sought universalistic theories based on such secure axioms as rationality, self-interest, and utility maximization, versus those like Schmoller (incidentally the father of the BSchool case study) who focused on induction, statistical analysis, and generalizations from amassed experiences – a contrast, we might say, between rigor and relevance, or rather between ‘pure’ and ‘historical’ explanations.

    Hodgson tells us that theorists like Schumpeter saw these as two quite different intellectual enterprises. To mix them up is to fail to understand the epistemological nature of the ‘strategic choices’ involved in setting up any workable science. In particular, these two enterprises entail different expectations of what the resulting research can produce. We might say ‘pure’ research produces universal theorizing of limited relevance, while the historical approach produces insights and illumination relevant to our experience, but of limited generality. To think this way is itself to adopt a particular methodological position. Inter alia it means assuming there is no happy-medium, triangulated middle-ground Archimedean vantage-point from which to view the universe of the human condition. Rather it is to understand the costs of methodological choice, for to choose one method is to prefer one set of possible outcomes while making others unavailable. “How happy I could be with either, were t’other dear charmer away!”

    It seems the discussion about MI can be either ‘pure’ or ‘historical’. I would presume most of our community thinks ‘pure’ and want to find some way of following Menger’s dictum that ‘the collective ideas of economics have to rest on individualistic components’, thus society has to be seen as grounded on whatever properties we associate with individuals. This implies a specific (and universal) model of man. If we presume something like rational economic man and think ‘pure’, we invoke a model of society that is something like social contract theory or some other kind of rational functionalism.

    This puts enormous theoretical weight on the selected definition of the individual. So one critique of this kind of MI is focused on how well one’s definition of the individual relates to one’s experience of others and, even more to the point, of oneself. Eventually this critique shifts us towards other models of man in which we allow that social structures shape the individual e.g. people from bad homes have greater criminal propensity, or people of faith are less likely to commit suicide. This, in turn, leads us into grounding the entire analysis on a presumed model of society. Nowadays we might call this ‘social constructionism’ or ‘social constructivism’. Then, instead of putting all the weight on our definition of the individual, we end up putting all the theory’s weight on our chosen definition of society. Again we get to wonder how well that definition captures our experience – since we cannot NOT have experience, which one of the ways in which the social sciences differ from the natural sciences.

    All of this is about theorizing a coherent model that embraces both the individual and her/his social context. I am not sure whether this is what MI is really about. In fact I suspect it is absolutely NOT what it is about.

    The alternative to a ‘pure’ approach is one driven by ‘experience’, from the mind we move to the body – to recall Coase’s comment. Here we shift the theoretical base from the conceptualized nature or ‘ontology’ of the individual and/or the social and onto the nature of human experience. Some people think this is what pragmatism is about, though I would disagree and argue it is closer to what ‘radical constructivism’ is about. Whether or not that is the case, and it is not immediately important to sort out the philosophical issues, we can see that an experience-based view of the world is fundamentally NOT coherent precisely because we lack – indeed, ignore the need for – the totalizing theory the ‘pure’ approach seeks.

    The point of my first posting to this thread was to suggest the MI discussion is about a particularly interesting group of economists’ views of a socio-economic reality that can be illuminated by insights gained from experience but is too complex to be understood in its entirety. This is another application of what came to be called ‘bounded rationality’ and is about adopting an ‘empirical methodology’ to oppose the ‘pure’ one.

    If this conjecture is correct, the methodological issue is to choose an empirical underpinning for MI that prioritizes experience over framing and part of setting this up is to make it impossible to collapse the discourse into either of the two ‘pure’ positions mentioned earlier. This can be achieved by the method of dialectic – by, for instance, choosing an underpinning model of man that is axiomatically incommensurable with the chosen model of social structure, order or process. Thus we might have a logic of conformity to existing power structures colliding with a logic of rational self-maximization, or a logic of market-making (Veblen, Triffin or Galbraith, perhaps) colliding with a logic of market-meeting (Walras or Friedman), and so forth.

    This approach’s Achilles Heel is that there can be no theory-free data, so we immediately have multiple theories swimming about in the discourse, indeed more than we can ever identify – which is one of the many reasons for using the term ‘tacit’. But our gain is in relevance; historical discourse’s relevance grows out of the multiple but familiar theories floating around, just as a painting can profit from the use of multiple colors and familiar modes of representation (tropes). This is not mystification. On the contrary, it is realism for we are used to holding multiple incomplete and tentatively held theories in mind as we negotiate the world. It is what we do all the time.

    In short, I suspect that MI is best understood as an attempt to rehabilitate political economy itself. It is a method of using the insights of Enlightenment individualism to help us break out of the 20th century political and sociological discourses which became too dependent on taken-for-granted collectives such as political parties, power holding groups, family and religious entities, and so forth. Indeed, I would go so far as to think MI was an effort by a particularly politically aware and conscious groups of contrarian thinkers to rehabilitate the Enlightenment vision itself – and in this effort I see much of the ‘libertarian’ discourse as dogmatic, having missed the deeper methodological issues around eschewing ‘pure’ theorizing.

  • 9. Paolo MARITI  |  19 July 2007 at 12:08 pm

    your comment is very fair and illuminating. I would only like to remind that such terms as individuals, society, poltical parties ….are not arytmomorphic, but intrisecally dialectic.And that is why we talk much about them and their relationships. We need at least two interacting individuals for a society. Yet two individuals “non dant collegium.”

  • 10. Rafe Champion  |  29 April 2008 at 2:01 am

    A late contribution. As suggested by the term “methodological individualism”, we are talking about methods, not ontology, and specifically the methods required to understand human actions (and their consequences). Thus, as stated in the second last sentence in the post, it is individuals who act.

    Ian Jarvie in his book “Concepts and Society) circa 1973 summarised an exasperating debate over MI involving anthropologists who referred to kinship structures as “non individual” causes of things that happen in society. The debate was exasperating because it was not resolved after a score of published articles by a dozen contributors over several years, and the protagonists appeared to be irretrievably at cross purposes.

    Maybe the answer is to keep the emphasis on the M part of MI and accept without protest that there are all kinds of things that are just THERE as far as actors are concerned, whatever historical or evolutionary explanations might be advanced to account for them. That includes the landscape, the built urban environment, kinship structures and the laws of the land. The purpose of MI is to provide illuminating and robust explanations of the judgements, decisions and actions of people in situations that include all kinds of constraints. The challenge to critics of MI is to provide better explanations. Protagonists of MI have to avoid being forced to defend atomistic individualism, or the ultimate wisdom and rationality of actors and such like.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

Trackback this post  |  Subscribe to the comments via RSS Feed


Nicolai J. Foss | home | posts
Peter G. Klein | home | posts
Richard Langlois | home | posts
Lasse B. Lien | home | posts


Former Guests | posts


Recent Posts



Our Recent Books

Nicolai J. Foss and Peter G. Klein, Organizing Entrepreneurial Judgment: A New Approach to the Firm (Cambridge University Press, 2012).
Peter G. Klein and Micheal E. Sykuta, eds., The Elgar Companion to Transaction Cost Economics (Edward Elgar, 2010).
Peter G. Klein, The Capitalist and the Entrepreneur: Essays on Organizations and Markets (Mises Institute, 2010).
Richard N. Langlois, The Dynamics of Industrial Capitalism: Schumpeter, Chandler, and the New Economy (Routledge, 2007).
Nicolai J. Foss, Strategy, Economic Organization, and the Knowledge Economy: The Coordination of Firms and Resources (Oxford University Press, 2005).
Raghu Garud, Arun Kumaraswamy, and Richard N. Langlois, eds., Managing in the Modular Age: Architectures, Networks and Organizations (Blackwell, 2003).
Nicolai J. Foss and Peter G. Klein, eds., Entrepreneurship and the Firm: Austrian Perspectives on Economic Organization (Elgar, 2002).
Nicolai J. Foss and Volker Mahnke, eds., Competence, Governance, and Entrepreneurship: Advances in Economic Strategy Research (Oxford, 2000).
Nicolai J. Foss and Paul L. Robertson, eds., Resources, Technology, and Strategy: Explorations in the Resource-based Perspective (Routledge, 2000).

%d bloggers like this: