The microfoundationalist delusion

1 August, 2016 at 12:00 | Posted in Economics | 10 Comments

The microfoundationalist’s fantasy has a powerful hold on macroeconomists. They recognize that an agent-by-agent reconstruction of the economy is not feasible, but they argue that it is something that we could do “in principle,” and that the in-principle claim warrants a particular theoretical strategy. The strategy is to start with the analysis of a single agent and to build up through ever more complex analyses to a whole economy …

The implicit argument in favor of representative-agent models as empirically relevant to aggregate economic data runs something like this: a representative-agent model is not itself an acceptable representation of the whole economy … but it is a first step in a program which step by step will inevitably bring the model closer to the agent-by-agent microeconomic model of the whole economy … I call this argument eschatological justification: it is the claim that there is a plausible in-principle game plan for a reductionist program and that the conclusions of early stages of that program are epistemically warranted by the presumed, but undemonstrated, success of the future implementation of the program in the fullness of time …

24958274Analysis using the representative-agent model employs an analogy between the behavior of a single agent and the agents collectively in a whole economy. For example, the representative-agent is typically endowed with a utility function from precisely the same family as those typically assigned to individual agents in microeconomic analysis. Do we have any good reason to accept the analogy? Microeconomists have long known that the answer is, no.

Exact aggregation requires that utility functions be identical and homothetic … Translated into behavioral terms, it requires that every agent subject to aggregation have the same preferences (you must share the same taste for chocolate with Warren Buffett) and those preferences must be the same except for a scale factor (Warren Buffet with an income of $10 billion per year must consume one million times as much chocolate as Warren Buffet with an income of $10,000 per year). This is not the world that we live in. The Sonnenschein-Mantel-Debreu theorem shows theoretically that, in an idealized general-equilibrium model in which each individual agent has a regularly specified preference function, aggregate excess demand functions inherit only a few of the regularity properties of the underlying individual excess demand functions: continuity, homogeneity of degree zero (i.e., the independence of demand from simple rescalings of all prices), Walras’s law (i.e., the sum of the value of all excess demands is zero), and that demand rises as price falls (i.e., that demand curves ceteris paribus income effects are downward sloping) … These regularity conditions are very weak and put so few restrictions on aggregate relationships that the theorem is sometimes called “the anything goes theorem.”

The importance of the theorem for the representative-agent model is that it cuts off any facile analogy between even empirically well-established individual preferences and preferences that might be assigned to a representative agent to rationalize observed aggregate demand. The theorem establishes that, even in the most favorable case, there is a conceptual chasm between the microeconomic analysis and the macroeconomic analysis. The reasoning of the representative-agent modelers would be analogous to a physicist attempting to model the macro- behavior of a gas by treating it as single, room-size molecule. The theorem demonstrates thatthere is no warrant for the notion that the behavior of the aggregate is just the behavior of the individual writ large: the interactions among the individual agents, even in the most idealized model, shapes in an exceedingly complex way the behavior of the aggregate economy. Not only does the representative-agent model fail to provide an analysis of those interactions, but it seems likely that that they will defy an analysis that insists on starting with the individual, and it is certain that no one knows at this point how to begin to provide an empirically relevant analysis on that basis.

Kevin Hoover

Kevin Hoover has been writing on microfoundations for now more than 25 years, and is beyond any doubts the one economist/econometrician/methodologist who has thought most on the issue. It’s always interesting to compare his qualified and methodologically founded assessment on the representative-agent-rational-expectations microfoundationalist program with the more or less apologetic views of freshwater economists like Robert Lucas:

hoovGiven what we know about representative-agent models, there is not the slightest reason for us to think that the conditions under which they should work are fulfilled. The claim that representative-agent models provide microfundations succeeds only when we steadfastly avoid the fact that representative-agent models are just as aggregative as old-fashioned Keynesian macroeconometric models. They do not solve the problem of aggregation; rather they assume that it can be ignored. While they appear to use the mathematics of microeconomis, the subjects to which they apply that microeconomics are aggregates that do not belong to any agent. There is no agent who maximizes a utility function that represents the whole economy subject to a budget constraint that takes GDP as its limiting quantity. This is the simulacrum of microeconomics, not the genuine article …

[W]e should conclude that what happens to the microeconomy is relevant to the macroeconomy but that macroeconomics has its own modes of analysis … [I]t is almost certain that macroeconomics cannot be euthanized or eliminated. It shall remain necessary for the serious economist to switch back and forth between microeconomics and a relatively autonomous macroeconomics depending upon the problem in hand.

Instead of just methodologically sleepwalking into their models, modern followers of the Lucasian microfoundational program ought to do some reflection and at least try to come up with a sound methodological justification for their position.  Just looking the other way won’t do. Writes Hoover:

garciaThe representative-­agent program elevates the claims of microeconomics in some version or other to the utmost importance, while at the same time not acknowledging that the very microeconomic theory it privileges undermines, in the guise of the Sonnenschein­Debreu­Mantel theorem, the likelihood that the utility function of the representative agent will be any direct analogue of a plausible utility function for an individual agent … The new classicals treat [the difficulties posed by aggregation] as a non-issue, showing no apprciation of the theoretical work on aggregation and apparently unaware that earlier uses of the representative-agent model had achieved consistency wiyh theory only at the price of empirical relevance.

Where ‘New Keynesian’ and New Classical economists think that they can rigorously deduce the aggregate effects of (representative) actors with their reductionist microfoundational methodology, they — as argued in chapter 4 of my  On the use and misuse of theories and models in economics — have to put a blind eye on the emergent properties that characterize all open social and economic systems. The interaction between animal spirits, trust, confidence, institutions, etc., cannot be deduced or reduced to a question answerable on the individual level. Macroeconomic structures and phenomena have to be analyzed also on their own terms.

10 Comments »

RSS feed for comments on this post. TrackBack URI

  1. The position of Robert Lucas is flat earther stuff. It is a Newtonian enterprise so backward it hardly needs rebutting. The appropriate response is laughter.

    Five minutes reading on the concept of ’emergence’ is enough to put paid to the ideas that

    a. explanation requires reduction
    b. reduction is explanation

  2. If I remember correctly, several years ago I asked Simon Wren-Lewis if Keynes’s ‘paradox of thrift’ was micro-founded. It was a bit of a mean-spirited question on my part, but he answered politely that although it wasn’t micro-founded, it could be, and that that possibility gave him more confidence in the idea of a paradox of thrift. He didn’t say how it could be micro-founded, and I didn’t ask.

    But now I’m thinking that question was a better question than I realized at the time. How is it logical to hold that the paradox of thrift can be correct, and at the same time, hold that models based on a representative agent are also correct? Wouldn’t these models show that either the paradox always occurred whenever your agent increased savings even just a bit, or that alternatively, it never occurred? And wouldn’t it be true that that is just not how the economy works?

    Now I never learned much about micro-founded models so I may very well be very wrong. Does anyone have an opinion about this?

    • Jerry

      First, the Keynesian cross is a representative agent model, so there is no dichotomy between the paradox of thrift and working with a representative agent.

      Second, in most microfounded new Keynesian models the extent to which the paradox of thrift kicks in depends on the reaction of the central bank — if the CB manages to lower the interest rate enough, the desire to save may be staved off. If, however, the central bank does not react aggressively enough (perhaps because the economy is in a liquidity trap) then the paradox of thrift operates at full blast.

      I hope that helps.

      • Pontus, thank you for the reply. If the Keynesian cross is representative of a representative agent model that means I have no idea what is meant by calling something a representative agent model. It is, for sure, always helpful to know when I am completely wrong. Painful sometimes, but helpful. Thanks again.

      • Jerry,

        A representative agent model is a model in which if you know the behavior of one agent, you know the behavior of all agents. It does not imply that coordination failures cannot arise (for instance, the prisoner’s dilemma is a representative agent model).

        All the best

      • Jerry,
        Please ignore Pontus. His grasp of the history of macro (and macro generally) is slippery. The Keynesian cross model is, of course, NOT microfounded as it was built long before the microfoundations ‘revolution’.
        To answer your original question, no, the paradox of thrift cannot be microfounded as it relies on ’emergent properties’. You can construct a model that react in the same was a Keynesian macro-model. That is, you can build a microfounded model that ‘simulates’ a Keynesian macro-model (Pontus has accurately described that in his comment). But it will function in an entirely different way.
        Simon-Wren Lewis and Pontus suffer from an awful affliction that is very prevalent amongst mainstream economists: they think that because they can simulate something they have thereby imitated it’s inner working. But in reality this is like assuming that dressing up like a police officer gives you the power of arrest.
        Hope that helps. Don’t get pulled in by these guys.

  3. “..the Keynesian cross is a representative agent model..”

    Come again?

    What happened to the heterogeneous agent model?

    • Which is the heterogeneous agent model?
      .
      You can of course always claim that the aggregate demand function in the Keynesian cross is an aggregate arising from several heterogenous agents with identical marginal propensities to consume (since linearity aggregates easily), or that it somewhat magically emerges as the “aggregate behavior” from the sum of millions of intricate demand functions. But such statements involves enormous leaps of faith, and is more of a red herring than anything else (and needless to say, standard “mainstream models” can use similar cop outs, if you think they are more aesthetically pleasing). And more importantly, such explanations are observationally equivalent to the representative agent interpretation.
      .
      I also made a quick google check to see that I’m not alone in this view. Indeed, Nick Rowe makes a similar point here: http://worthwhile.typepad.com/worthwhile_canadian_initi/2015/02/keynesian-cross-as-simultaneous-moves-symmetric-nash-equilibrium.html
      .
      To quote: “But strictly speaking, the Keynesian Cross model is a representative agent model. So you Old Keynesians shouldn’t be throwing stones at representative agent models.”
      .
      Ouch

      • There are no explicit assumptions made about how the model is constructed. It could be constructed in an manner desired.

  4. It’s OK Jerry, Pontus inhabits a neoclassical world of delusion.


Leave a Reply

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

WordPress.com Logo

You are commenting using your WordPress.com 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 )

Google+ photo

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

Connecting to %s

Create a free website or blog at WordPress.com.
Entries and comments feeds.