Thursday, April 24, 2014

Methodological Individualism – A Reformulation

It has not been unpopular from some scholars to attack methodological individualism as an unnecessary constraint on economic analysis. Individual action does not occur in an institutional vacuum. To reduce economics, or any social science, to only individual choice ignores the significance of emergent phenomena that shape the world we live in. Geoff Hodgson makes a passionate call for a reasoned compromise:

Despite the impression given by many adherents of methodological individualism, it is quite legitimate to deny that only purposes or actions of individuals are explanatory… These beliefs [about individual autonomy] are not undermined by a conviction that whilst people are, on the one hand, purposeful and have real choices, they are, on the other, moulded by their cultural and institutional environment. (Hodgson 1986, 222)

Hodgson claims that Austrian economists overlook this feature in their research. Just as macroeconomics can benefit from microfoundations, macro – or emergent – foundations are always implicit in microeconomic analysis.

Though his claim has legitimacy, Hodgson’s critique of his opponents is more sweeping than merited. In the same work, Hodgson lumps together the views of Hayek and Mises concerning methodological individualism and social wholes. Hayek’s formulation of social exchange and development clearly includes collectives as actual entities, even if he is not always consistent. These entities exist in the form of rules that guide action, which when adopted, create norms not dependent on any single person. “Society can thus exist only if by a process of selection rules have evolved which lead individuals to behave in a manner which makes social life possible (Hayek, 1973, 44 quoted in Andreozzi 2005).” This insight from Hayek presents those who seek to reformulate the nature of the individual in society a preliminary goal post from which to begin analysis. Society can be analyzed, in part, by the rules that it embodies. But first I shall review the conception of methodological individualism often promoted by Austrian economists.

In the textbook version of Austrian economics, only individual action matters. This leads to a tendency to disregard the formative nature of institutions. Institutions serve individuals, not the other way around. As Ludwig von Mises described in his canonical work Human Action,

…Methodological individualism, far from contesting the significance of such collective wholes, considers it as one of its main tasks to describe and to analyze their becoming and their disappearing, their changing structures, and their operation. And it chooses the only method fitted to solve this problem satisfactorily.

First we must realize that all actions are performed by individuals. A collective operates always through the intermediary of one or several individuals whose actions are related to the collective as the secondary source… The hangman, not the state, executes a criminal. (1996, 43)

But the hangman receives authority from the “state”. The state is an entity that is bigger than any one person. While any action requires an individual, the state practically subsumes them when they act as its agent. Individual action is constrained by the state apparatus. Individuals might move in and out of the state, but the state remains. This may have been part of the decisions of particular individuals, but those individuals are logically replaceable by others. Methodological individualism in the strict sense that Mises presents it breaks down. But this need not be so if we accept an interaction between institutions and individuals, which in some instances Mises comes close to acknowledging.

Mises’s argument is likely skewed due to his context which appears to have promoted confusion between political individualism and methodological individualism in his writing. During his career, he battled collectivist perceptions that had been common in Germany like the idea of a fatherland, the volk, and so on. These colloquial references to the collective lack much substance beyond culture. But many collectives are well formed. This does not prevent Mises from denouncing analysis of collectives more generally,

It is illusory to believe that it is possible to visualize collective wholes. They are never visible; their cognition is always the outcome of the understanding of the meaning which acting men attribute to their acts. We can see a crowd, i.e., a multitude of people. Whether this crowd is a mere gathering or a mass (in the sense in which this term is used in contemporary psychology) or an organized body or any other kind of social entity is a question which can only be answered by understanding the meaning which they themselves attach to their presence. And this meaning is always the meaning of individuals. Not our senses, but understanding, a mental process, makes us recognize social entities. (43)

In terms of imaginary collectives that lack substantial structure, Mises is more correct than not. Collectives like government, corporations, and religious groups, however, do not fall under the same category as a crowd. We might better understand these institutions as rules and structure that are epiphenomena of particular networks. As rules become formalized through building of indigenously introduced exogenous institutions, they are embodied in a real, existent entity that govern interactions within particular networks (Boettke, Coyne, and Leeson 2008). Analysis then includes not only the individual, but also the system. Interactions occur within network and between networks, with different networks embodying different norms and rule sets. These rules more tightly constrain the actions of agents within the system than do informal norms, thus we can observe the interaction between the individual agent and emergent phenomena concretized by the creation of formal institutions.

This process has been recognized within development literature, but it has not been explicictly stated in regard to methodological individualism. In their presentation of the process of social development, North, Wallace, and Weingast present three doorstep conditions for moving into what they call an open-access order where political violence is constrained and political power is systematized and contestable (2009). Their second doorstep condition for entering an open-access order is the creation of perpetually lived institutions in both the public and private spheres of life:

The capacity to form and support perpetually lived organizations has direct consequences for a society’s ability to structure social relationships over time. The creation of legal personalities for organizations constitutes an essential element of perpetual life; it is inherently impersonal because it is defined without reference to any specific individuals. (2009, 158)

The growth of well-developed institutions both constrain and expand the option sets of individuals. They perform a coordinating function amongst individuals by providing boundaries for individual action. When the state takes action, it is not so important who serves as its agent, only that someone does. Something bigger than the individual molds the actions of every individual within their network.

These coordinating functions also interact with one another. When institutions collide, some amount of conflict results. To be clear, action always occurs at the individual level. When robust institutions are included and, in this case, when they conflict, the nature of the action changes. One might say that two nation’s militaries battled one another. In reality, many individuals with weapons attacked one another, but in doing so, they acted as agents of the state. Their constraints change the moment they put on uniforms and are depending on the role they adopt or are assigned. This sort of analysis makes little sense without recognizing the existence and importance of emergent institutions like the state and the constraints placed upon its agents.

The world is dynamic. Social economy must be modeled as such. Methodological individualism need not be discarded. It must be integrated into a deeper form of analysis. Its general spirit is correct. Action occurs at the individual level. The individual, however, is constantly engaged in a conversation with society about what his or her actions should and can be. Society, in this sense, is defined not simply as everyone else, but also as the formal and informal institutions that constantly influence one’s actions. Mises and Hayek may not have been amenable to studying why people act as they do, as this is not a question of prime interest to economists. Economists should ask how and why one’s options are shaped and constrained by emergent institutions. The search for an answer to this will enable us to form a more robust method of analysis.

Andreozzi, Luciano. “Hayek Reads the Literature on the Emergence of Norms.” Econ Stor Working Paper no. 0503. 2005.

Boettke, Peter, Coyne, and Leeson. “Institutional Stickiness and the New Development Economics.” The American Journal of Economics and Sociology 64, no. 2 (April 2008): 331-358.

Hodgson, Geoff. “Behind Methodological Individualism.” Cambridge Journal of Economic 10 (1986): 211-224.

Mises, Ludwig von. Human Action: A Treatise on Economics Irvington-on-the-Hudson, NY: Foundation for Economic Education, 1996 [1949].

North, Douglass, Wallace, and Weingast. Violence and Social Orders New York: Cambridge University Press, 2009.


  1. "But this need not be so if we accept an interaction between institutions and individuals"

    In other words, if we abandon methodological individualism! Or, looked at in a different way, if we incorporate all you want into our analysis (which is certainly a good idea!), why in the world would we still call this "methodological individualism"? It seems like China stubbornly insisting on calling a society with stock markets and billionaires "communist."

  2. Not quite. I propose that we connect institutions with individuals. That is their starting place. Over time, these institutions may not be dependent on any single individual. This still doesn't change that in its early stages, an institution represents a network of particular individuals.

    1. OK, so you've not quite given up the ghost? OK, then contemplate this: while viewed in one way, the starting place of institutions is individuals, in another sense, the starting place of every concrete individual is in institutions: without a family none of us would have made it out of infancy, without the institution of language, which pre-existed both of us, we could not have this conversation.

      Hayek abandoned methodological individualism: you can too.

    2. Your line about Hayek would make a great article title.

      Agents exist within a network. This is inevitable for anyone not named Robinson Crusoe. The existent of emergent phenomena like the family is not incompatible with methodological individualism. We may have to formulate a new title for a broader, inclusive methodology, as it is not the same blunt that Mises refers to in the above passages.

      This is not the same as incorporating "all I want" into my analysis. It's adding a place for emergent phenomena. (You won't catch me saying that the state doesn't exist...)

    3. "Agents exist within a network."

      This implies an pre-existing agent that could be unplugged from one network and dropped into another unchanged. But that is not reality: in reality, the agents would not exist at all without the network. The network and the agents are mutually dependent. Either without the other is a mere abstraction.

      "The existent of emergent phenomena like the family is not incompatible with methodological individualism."

      Yes, but this is still backwards: in the actual world, families don't emerge from individuals: individuals emerge from families!

  3. Yes agents are born into networks, and their they are dependent upon those networks.Some networks allows for individuals to move in and out of them, others do not. This is not a rule, just a description. I don't believe we are at disagreement there. Adherence to a naive MI is not a viable strategy.

    But how did that family come to exist? It came to exist by the engagement of two individuals. Although interesting, the chicken/egg argument does not get us very far. Individuals effect institutions and institutions constrain and shape individuals.

    I'll have something more substantive to contribute to this argument after I spend the next year learning ABM and learning network theory.