Methodological Arrogance

A few weeks ago, I posted a somewhat critical review of Kartik Athreya’s new book Big Ideas in Macroeconomics. In quoting a passage from chapter 4 in which Kartik defended the rational-expectations axiom on the grounds that it protects the public from economists who, if left unconstrained by the discipline of rational expectations, could use expectational assumptions to generate whatever results they wanted, I suggested that this sort of reasoning in defense of the rational-expectations axiom betrayed what I called the “methodological arrogance” of modern macroeconomics which has, to a large extent, succeeded in imposing that axiom on all macroeconomic models. In his comment responding to my criticisms, Kartik made good-natured reference in passing to my charge of “methodological arrogance,” without substantively engaging with the charge. And in a post about the early reviews of Kartik’s book, Steve Williamson, while crediting me for at least reading the book before commenting on it, registered puzzlement at what I meant by “methodological arrogance.”

Actually, I realized when writing that post that I was not being entirely clear about what “methodological arrogance” meant, but I thought that my somewhat tongue-in-cheek reference to the duty of modern macroeconomists “to ban such models from polite discourse — certainly from the leading economics journals — lest the public be tainted by economists who might otherwise dare to abuse their models by making illicit assumptions about expectations formation and equilibrium concepts” was sufficiently suggestive not to require elaboration, especially after having devoted several earlier posts to criticisms of the methodology of modern macroeconomics (e.g., here, here, and here). That was a misjudgment.

So let me try to explain what I mean by methodological arrogance, which is not the quite the same as, but is closely related to, methodological authoritarianism. I will do so by referring to the long introductory essay (“A Realist View of Logic, Physics, and History”) that Karl Popper contributed to a book The Self and Its Brain co-authored with neuroscientist John Eccles. The chief aim of the essay was to argue that the universe is not fully determined, but evolves, producing new, emergent, phenomena not originally extant in the universe, such as the higher elements, life, consciousness, language, science and all other products of human creativity, which in turn interact with the universe, in fundamentally unpredictable ways. Popper regards consciousness as a real phenomenon that cannot be reduced to or explained by purely physical causes. Though he makes only brief passing reference to the social sciences, Popper’s criticisms of reductionism are directly applicable to the microfoundations program of modern macroeconomics, and so I think it will be useful to quote what he wrote at some length.

Against the acceptance of the view of emergent evolution there is a strong intuitive prejudice. It is the intuition that, if the universe consists of atoms or elementary particles, so that all things are structures of such particles, then every event in the universe ought to be explicable, and in principle predictable, in terms of particle structure and of particle interaction.

Notice how easy it would be rephrase this statement as a statement about microfoundations:

Against the acceptance of the view that there are macroeconomic phenomena, there is a strong intuitive prejudice. It is the intuition that, if the macroeconomy consists of independent agents, so that all macroeconomic phenomena are the result of decisions made by independent agents, then every macreconomic event ought to be explicable, and in principle predictable, in terms of the decisions of individual agents and their interactions.

Popper continues:

Thus we are led to what has been called the programme of reductionism [microfoundations]. In order to discuss it I shall make use of the following Table

(12) Level of ecosystems

(11) Level of populations of metazoan and plants

(10) Level of metezoa and multicellular plants

(9) Level of tissues and organs (and of sponges?)

(8) Level of populations of unicellular organisms

(7) Level of cells and of unicellular organisms

(6) Level of organelles (and perhaps of viruses)

(5) Liquids and solids (crystals)

(4) Molecules

(3) Atoms

(2) Elementary particles

(1) Sub-elementary particles

(0) Unknown sub-sub-elementary particles?

The reductionist idea behind this table is that the events or things on each level should be explained in terms of the lower levels. . . .

This reductionist idea is interesting and important; and whenever we can explain entities and events on a higher level by those of a lower level, we can speak of a great scientific success, and can say that we have added much to our understanding of the higher level. As a research programme, reductionism is not only important, but it is part of the programme of science whose aim is to explain and to understand.

So far so good. Reductionism certainly has its place. So do microfoundations. Whenever we can take an observation and explain it in terms of its constituent elements, we have accomplished something important. We have made scientific progress.

But Popper goes on to voice a cautionary note. There may be, and probably are, strict, perhaps insuperable, limits to how far higher-level phenomena can be reduced to (explained by) lower-level phenomena.

[E]ven the often referred to reduction of chemistry to physics, important as it is, is far from complete, and very possibly incompletable. . . . [W]e are far removed indeed from being able to claim that all, or most, properties of chemical compounds can be reduced to atomic theory. . . . In fact, the five lower levels of [our] Table . . . can be used to show that we have reason to regard this kind of intuitive reduction programme as clashing with some results of modern physics.

For what [our] Table suggests may be characterized as the principle of “upward causation.” This is the principle that causation can be traced in our Table . . . . from a lower level to a higher level, but not vice versa; that what happens on a higher level can be explained in terms of the next lower level, and ultimately in terms of elementary particles and the relevant physical laws. It appears at first that the higher levels cannot act on the lower ones.

But the idea of particle-to-particle or atom-to-atom interaction has been superseded by physics itself. A diffraction grating or a crystal (belonging to level (5) of our Table . . .) is a spatially very extended complex (and periodic) structure of billions of molecules; but it interacts as a whole extended periodic structure with the photons or the particles of a beam of photons or particles. Thus we have here an important example of “downward causation“. . . . That is to say, the whole, the macro structure, may, qua whole, act upon a photon or an elementary particle or an atom. . . .

Other physical examples of downward causation – of macroscopic structures on level (5) acting upon elementary particles or photons on level (1) – are lasers, masers, and holograms. And there are also many other macro structures which are examples of downward causation: every simple arrangement of negative feedback, such as a steam engine governor, is a macroscopic structure that regulates lower level events, such as the flow of the molecules that constitute the steam. Downward causation is of course important in all tools and machines which are designed for sompe purpose. When we use a wedge, for example, we do not arrange for the action of its elementary particles, but we use a structure, relying on it ot guide the actions of its constituent elementary particles to act, in concert, so as to achieve the desired result.

Stars are undersigned, but one may look at them as undersigned “machines” for putting the atoms and elementary particles in their central region under terrific gravitational pressure, with the (undersigned) result that some atomic nuclei fuse and form the nuclei of heavier elements; an excellent example of downward causation,of the action of the whole structure upon its constituent particles.

(Stars, incidentally, are good examples of the general rule that things are processes. Also, they illustrate the mistake of distinguishing between “wholes” – which are “more than the sums of their parts” – and “mere heaps”: a star is, in a sense, a “mere” accumulation, a “mere heap” of its constituent atoms. Yet it is a process – a dynamic structure. Its stability depends upon the dynamic equilibrium between its gravitational pressure, due to its sheer bulk, and the repulsive forces between its closely packed elementary particles. If the latter are excessive, the star explodes, If they are smaller than the gravitational pressure, it collapses into a “black hole.”

The most interesting examples of downward causation are to be found in organisms and in their ecological systems, and in societies of organisms [my emphasis]. A society may continue to function even though many of its members die; but a strike in an essential industry, such as the supply of electricity, may cause great suffering to many individual people. .. . I believe that these examples make the existence of downward causation obvious; and they make the complete success of any reductionist programme at least problematic.

I was very glad when I recently found this discussion of reductionism by Popper in a book that I had not opened for maybe 40 years, because it supports an argument that I have been making on this blog against the microfoundations program in macroeconomics: that as much as macroeconomics requires microfoundations, microeconomics also requires macrofoundations. Here is how I put a little over a year ago:

In fact, the standard comparative-statics propositions of microeconomics are also based on the assumption of the existence of a unique stable general equilibrium. Those comparative-statics propositions about the signs of the derivatives of various endogenous variables (price, quantity demanded, quantity supplied, etc.) with respect to various parameters of a microeconomic model involve comparisons between equilibrium values of the relevant variables before and after the posited parametric changes. All such comparative-statics results involve a ceteris-paribus assumption, conditional on the existence of a unique stable general equilibrium which serves as the starting and ending point (after adjustment to the parameter change) of the exercise, thereby isolating the purely hypothetical effect of a parameter change. Thus, as much as macroeconomics may require microfoundations, microeconomics is no less in need of macrofoundations, i.e., the existence of a unique stable general equilibrium, absent which a comparative-statics exercise would be meaningless, because the ceteris-paribus assumption could not otherwise be maintained. To assert that macroeconomics is impossible without microfoundations is therefore to reason in a circle, the empirically relevant propositions of microeconomics being predicated on the existence of a unique stable general equilibrium. But it is precisely the putative failure of a unique stable intertemporal general equilibrium to be attained, or to serve as a powerful attractor to economic variables, that provides the rationale for the existence of a field called macroeconomics.

And more recently, I put it this way:

The microeconomic theory of price adjustment is a theory of price adjustment in a single market. It is a theory in which, implicitly, all prices and quantities, but a single price-quantity pair are in equilibrium. Equilibrium in that single market is rapidly restored by price and quantity adjustment in that single market. That is why I have said that microeconomics rests on a macroeconomic foundation, and that is why it is illusory to imagine that macroeconomics can be logically derived from microfoundations. Microfoundations, insofar as they explain how prices adjust, are themselves founded on the existence of a macroeconomic equilibrium. Founding macroeconomics on microfoundations is just a form of bootstrapping.

So I think that my criticism of the microfoundations project exactly captures the gist of Popper’s criticism of reductionism. Popper extended his criticism of a certain form of reductionism, which he called “radical materialism or radical physicalism” in later passage in the same essay that is also worth quoting:

Radical materialism or radical physicalism is certainly a selfconsistent position. Fir it is a view of the universe which, as far as we know, was adequate once; that is, before the emergence of life and consciousness. . . .

What speaks in favour of radical materialism or radical physicalism is, of course, that it offers us a simple vision of a simple universe, and this looks attractive just because, in science, we search for simple theories. However, I think that it is important that we note that there are two different ways by which we can search for simplicity. They may be called, briefly, philosophical reduction and scientific reduction. The former is characterized by an attempt to provide bold and testable theories of high explanatory power. I believe that the latter is an extremely valuable and worthwhile method; while the former is of value only if we have good reasons to assume that it corresponds to the facts about the universe.

Indeed, the demand for simplicity in the sense of philosophical rather than scientific reduction may actually be damaging. For even in order to attempt scientific reduction, it is necessary for us to get a full grasp of the problem to be solved, and it is therefore vitally important that interesting problems are not “explained away” by philosophical analysis. If, say, more than one factor is responsible for some effect, it is important that we do not pre-empt the scientific judgment: there is always the danger that we might refuse to admit any ideas other than the ones we appear to have at hand: explaining away, or belittling the problem. The danger is increased if we try to settle the matter in advance by philosophical reduction. Philosophical reduction also makes us blind to the significance of scientific reduction.

Popper adds the following footnote about the difference between philosophic and scientific reduction.

Consider, for example, what a dogmatic philosophical reductionist of a mechanistic disposition (or even a quantum-mechanistic disposition) might have done in the face of the problem of the chemical bond. The actual reduction, so far as it goes, of the theory of the hydrogen bond to quantum mechanics is far more interesting than the philosophical assertion that such a reduction will one be achieved.

What modern macroeconomics now offers is largely an array of models simplified sufficiently so that they are solvable using the techniques of dynamic optimization. Dynamic optimization by individual agents — the microfoundations of modern macro — makes sense only in the context of an intertemporal equilibrium. But it is just the possibility that intertemporal equilibrium may not obtain that, to some of us at least, makes macroeconomics interesting and relevant. As the great Cambridge economist, Frederick Lavington, anticipating Popper in grasping the possibility of downward causation, put it so well, “the inactivity of all is the cause of the inactivity of each.”

So what do I mean by methodological arrogance? I mean an attitude that invokes microfoundations as a methodological principle — philosophical reductionism in Popper’s terminology — while dismissing non-microfounded macromodels as unscientific. To be sure, the progress of science may enable us to reformulate (and perhaps improve) explanations of certain higher-level phenomena by expressing those relationships in terms of lower-level concepts. That is what Popper calls scientific reduction. But scientific reduction is very different from rejecting, on methodological principle, any explanation not expressed in terms of more basic concepts.

And whenever macrotheory seems inconsistent with microtheory, the inconsistency poses a problem to be solved. Solving the problem will advance our understanding. But simply to reject the macrotheory on methodological principle without evidence that the microfounded theory gives a better explanation of the observed phenomena than the non-microfounded macrotheory (and especially when the evidence strongly indicates the opposite) is arrogant. Microfoundations for macroeconomics should result from progress in economic theory, not from a dubious methodological precept.

Let me quote Popper again (this time from his book Objective Knowledge) about the difference between scientific and philosophical reduction, addressing the denial by physicalists that that there is such a thing as consciousness, a denial based on their belief that all supposedly mental phenomena can and will ultimately be reduced to purely physical phenomena

[P]hilosophical speculations of a materialistic or physicalistic character are very interesting, and may even be able to point the way to a successful scientific reduction. But they should be frankly tentative theories. . . . Some physicalists do not, however, consider their theories as tentative, but as proposals to express everything in physicalist language; and they think these proposals have much in their favour because they are undoubtedly convenient: inconvenient problems such as the body-mind problem do indeed, most conveniently, disappear. So these physicalists think that there can be no doubt that these problems should be eliminated as pseudo-problems. (p. 293)

One could easily substitute “methodological speculations about macroeconomics” for “philosophical speculations of a materialistic or physicalistic character” in the first sentence. And in the third sentence one could substitute “advocates of microfounding all macroeconomic theories” for “physicalists,” “microeconomic” for “physicalist,” and “Phillips Curve” or “involuntary unemployment” for “body-mind problem.”

So, yes, I think it is arrogant to think that you can settle an argument by forcing the other side to use only those terms that you approve of.

About these ads

19 Responses to “Methodological Arrogance”


  1. 1 Diego Espinosa February 26, 2014 at 9:53 pm

    David,
    I appreciate the Popper references, but I’m not sure I agree with your view of reductionism. Yes, reductionism takes the form of explaining the whole by reference to the behavior of its parts. The question is, is macroeconomics reductionist ex-microfoundations? I think yes: in that it imagines the parts of the macroeconomy as independent and ergodic, with initial conditions lacking influence. Macroeconomics is reductionist in its clockwork view of equilibrium; in its assumptions about the representative agent and his preferences; in its lack of recognition of the sources of complexity and its principle phenomenon, emergence. Perhaps one could argue that it is reductionist without a very good understanding of its parts, and that is the problem microfoundations is trying to fix. Maybe it is the wrong problem. The real one is its incorrect assumptions about how the economy really works, and for the degree of uncertainty that applies to understanding the behavior of the whole. Given your post, I think Popper may have agreed with that.

  2. 2 Blue Aurora February 27, 2014 at 1:22 am

    Out of curiosity, David Glasner…are you aware of any research in experimental economics or decision theory or behavioural economics that would support your point(s)?

  3. 3 dmerciar February 27, 2014 at 3:14 am

    Reblogged this on MERCIAR BUSINESS CONSULTING and commented:
    Most pertinent piece on the interdependence of micro and macro, extending the critique of micro foundations in macro…

  4. 4 Dan Thorn February 27, 2014 at 5:39 am

    How do microfoundations deal with the paradox of thrift? or more generally with the set of problems where what is good for the individual doesn’t sum up to something that’s good for everyone and in turn ends up not being good for the individual?
    Also, how do microfoundations deal with ‘traffic light’ problems. Where coordination is required. obviously each individual at an intersection is best off going first – but if everyone did, no one makes it through. Coordination problems lead naturally to a theory of government and a theory of the firm that I never see addressed in microfoundation.
    I agree that your point that the sum is different than the whole also requires some interaction term which so far seems left out.
    And the ever popular point variously referred to as reflexivity, the observer effect, path dependency and no doubt other ways is similarly in need of being addressed by microfoundations.

    How the heck do you have time to write these posts? your breadth and depth seems to me to be unrivaled in the econ blogosphere.

  5. 5 Pablo Bortz February 27, 2014 at 6:31 am

    Excellent post. I think that the term “foundations” should be scrapped and replace by the word “connections”. “Foundations” carries a weight, and it imparts an a priori, unfounded (sorry for the redundancy) judgement. It is fair to ask for the “microconnections” of a model, or the “macroconnections”. But to demand an analysis to be “founded” on micro behaviour, is very arrogant, like you perfectly described.

  6. 6 Erik February 27, 2014 at 7:11 am

    Valuable thougts, but somewhat wasted on Mr. Athreya I think. Sure enough unaxiomatic thinking differs from axiomatic thinking. This proves nothing though for the validity or substantive rationality of the axiom, in Atheya’s case the rational expectations axiom. Axioms cannot be shown to be right or wrong. They are set arbitrarly – as a starting point for deduction – by definition. Their plausibility must be determinated with regards to the reality they are supposed to represent.

  7. 7 Non nobis solum February 27, 2014 at 7:51 am

    David,
    I am truly grateful for this most insightful posting.
    The world of economics is a complex place, and you certainly help clarify things. This posting together with Paul De Grauwe’s article

    http://www.voxeu.org/article/top-down-versus-bottom-macroeconomics,

    are the two best econ articles I have read to date.

  8. 8 Digital Cosmology (@DCosmology) February 27, 2014 at 10:46 am

    It is possible that there is no definable link between macro and micro the same way what there is no apparent link between the indeterminism of the microcosm of quantum mechanism and the determinism of the macrocosm of general relativity. One must chose sides in total ignorance of what is connecting these two worlds. One answer is that one of the two worlds, the micro or the macro, is an epiphenomenon, but which is it? Answer first this question. It appears that a reasonable call to action is this: http://www.digitalcosmology.com/Blog/2014/02/12/a-call-to-action-abolish-economics-departments-now/

    Then, start thinking that the reason that the rules seem to break all the time is that this world is a video game: http://www.digitalcosmology.com/Blog/2012/09/14/is-our-world-a-simulation-or-even-an-interactive-computer-game/

  9. 9 David Glasner February 27, 2014 at 2:05 pm

    Diego, I agree with all your criticisms of macroeconomics. Where do you think we are in disagreement?

    Blue Aurora, My knowledge of the three fields you mention is pretty meagre, at best, so I can’t really say? What do you think?

    dmerciar, Thanks.

    Dan, I don’t know how microfoundations would deal with the paradox of thrift. The paradox of thrift cannot characterize an economy operating on a unique equilibrium time path. There has to be some disequilibrium or a transition from optimistic to pessimistic equilibrium paths to generate something like the paradox of thrift in a standard microfounded framework. Coordination problems are largely dismissed by the microfoundations school (see my review of Athreya’s book). Roger Farmer is very concerned about coordination problems and he models them using multiple potential equilibria whose realization depends on what is expected.

    Actually, I had a hard time writing this post, it took me the better part of two weeks to get it done.

    Pablo, The terms microfoundations has been around for over 40 years, so it would be pretty hard to get people to switch to a new term, but I understand your point. I wrote a post a while back about the multiple meanings lurking in the background of microfoundations.

    Erik, I wanted to discuss the axiomatic approach, which Popper also criticized as a relatively unproductive exercise in science, but that would have stretched a long post way too far. Maybe I’ll talk about Popper’s views on the axiomatic method in another post another time.

    Non nobis solum, You’re welcome, and thank you for the link to De Grauwe’s piece, which looks very interesting.

    Digital Cosmology, Interesting point, but I’m afraid it’s over my head.

  10. 10 Jason February 27, 2014 at 2:59 pm

    Excellent post!

    I have two additions:

    1. Since one can formulate a theory at level (12) “ecosystems” in the table that makes no reference whatsoever to e.g. level (4) “molecules”, the information about the detailed theory of (4) must have been destroyed (and it is possible that many theories of “molecules” could be consistent with the known theory of “ecosystems”). The analogy between this an economics was the entire purpose of my comments on your earlier post. The details of any microeconomic theory are likely lost at the macroeconomic level.

    2. [Random note] T-duality in string theory makes the table somewhat of an ouroboros. The theory of the very large is equivalent to a theory of the very small (in a particular sense). The Universe as a whole at the large scale where it is almost flat is described by general relativity which is equivalent (has the same physical content) to some strongly interacting string theory on a small scale.

  11. 11 Dan Thorn February 27, 2014 at 3:34 pm

    Well, micro-foundation will no doubt add value one way or the other. I suppose the basic point is that it isn’t THE truth. Which should be too obvious to need to state.

    How dull to be so certain, and how embarrassing to be so narrow.

  12. 12 Jonathan Goldberg February 27, 2014 at 7:03 pm

    The comments on economics are interesting. I’m sorry, but the references to Popper don’t do any work. What you say after you substitute you concerns for Popper’s seems to me to be the meat of this post. The Popper itself doesn’t add anything.

    Popper wrote a long time ago and Philosophy of Science has covered a lot of ground in the meantime.

    To me, the point about micro foundations being dependent on macro equilibrium assumptions is absolutely crucial and cannot be overstated.

    I’ve just received Athreya’s book and will be reading it with your caveats in mind.

    Warning: my study of Philosophy of Economics is at a an early enough stage that I should be cautious about making strong claims. As you can see, this is easier advice to give than to take.

  13. 13 David Glasner March 3, 2014 at 7:19 am

    Jason, Thanks. Your point about the loss of information between molecules and ecosystems might well be true, but one could at least imagine the possibility that it might not. But that aside, I don’t think that argument would work for macroeconomics and microeconomics, which do seem to overlap to a substantial degree. Your random note is lost on me, but I hope other readers will find it of interest.

    Dan, Popper is right that we should aspire to achieve reduction as an endpoint, not insist on it as our starting point.

    Jonathan, Popper may have been superfluous from your point of view as the reader, but I could not have formulated my position without having his discussion in front of me.

  14. 14 Blue Aurora March 11, 2014 at 11:41 pm

    Although my response to you is belated, Dr. Glasner…I meant that experimental economics and decision theory and behavioural economics can illuminate the legitimacy of the microeconomic foundations project. I personally don’t have a problem with “bottom-up aggregation”, but I do think that the definition of rationality used in orthodox economics is excessively narrow and relies on criteria for information that are too strong. I personally would suggest that you read Daniel Ellsberg’s article in the November 1961 edition of the Quarterly Journal of Economics for more reference, along with the contributions of Daniel Kahneman and Amos Tversky.


  1. 1 On macroeconomic arrogance | LARS P. SYLL Trackback on February 27, 2014 at 9:03 am
  2. 2 Microfounded Economics or Macrotheory? | Digital Cosmology Blog Trackback on February 27, 2014 at 10:57 am
  3. 3 Somewhere else, part 117 | Freakonometrics Trackback on March 5, 2014 at 8:37 am
  4. 4 Methodological arrogance | LARS P. SYLL Trackback on July 20, 2014 at 5:40 am
  5. 5 Explaining the Hegemony of New Classical Economics | Uneasy Money Trackback on September 30, 2014 at 9:24 pm

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




About Me

David Glasner
Washington, DC

I am an economist at the Federal Trade Commission. Nothing that you read on this blog necessarily reflects the views of the FTC or the individual commissioners. Although I work at the FTC as an antitrust economist, most of my research and writing has been on monetary economics and policy and the history of monetary theory. In my book Free Banking and Monetary Reform, I argued for a non-Monetarist non-Keynesian approach to monetary policy, based on a theory of a competitive supply of money. Over the years, I have become increasingly impressed by the similarities between my approach and that of R. G. Hawtrey and hope to bring Hawtrey's unduly neglected contributions to the attention of a wider audience.

Enter your email address to follow this blog and receive notifications of new posts by email.

Join 276 other followers


Follow

Get every new post delivered to your Inbox.

Join 276 other followers

%d bloggers like this: