Lucas Defending Macroeconomics: Not Enough

by Mario Rizzo

Reportedly, Queen Elisabeth asked why didn’t any of the top-tier macroeconomists predict the financial meltdown and Great Recession. So now The Economist and other loyal subjects are trying to answer that question. The Chicago economist Robert Lucas, an American, has also added his opinion. It is deeply flawed but perhaps a comfort to the “Neoclassical Top Brass” (as my late colleague Ludwig Lachmann used to say).

This is not the time to go into detail regarding Lucas’s views of macroeconomics and on forecasting. However, I am intrigued that he uses the Efficient Markets Hypothesis (EMH) in a fairly strong form, I should think,  to show why if the high-tech macroeconomists did predict the the meltdown it would have come earlier (immediately). There is an obvious paradox here.

In other words, he maintains EMH to show the impossibility of prediction. Does he not appreciate the degree to which strong EMH is under attack and in doubt?

I am in the strange position of agreeing with Lucas on the impossibility of prediction. However, I’d rather use the analysis of Friedrich Hayek in his “Pretence of Knowledge” and “Theory of Complex Phenomena”  as my basis.

But what Hayek makes clear is that it would not be too much to ask of a theory that it at least point in the general direction of a problem. Lucas, and others, continue to think of recent events as purely a bubble phenomenon — and so there is a degree of randomness involved in the phenonemon, in their view. However, Hayek’s cycle theory pointed to an interest rate effects on higher-order investments that generally or systematically produce over- and mal-investment. But it says little about timing. Lucas seems oblivious to both.

15 thoughts on “Lucas Defending Macroeconomics: Not Enough

  1. Keynesians seem oblivious to both as well, but Keynesians are attempting to use the recession as an attempt to make their failed paradigm the dominant one once again.

  2. Asset markets are hard to predict. EMH says they are hard to predict because we’re all so smart that prices reflect all available information. (Different versions of EMH come from different ideas of what information is “available.”) Hayekian economics says they are hard to predict because we’re all too stupid to properly sift the available information. EMH allows “rational” bubbles; Hayekian economics allows non-rational bubbles. I think the we’re-too-stupid explanation beats the we’re-so-smart explanation. I think it makes a substantive difference for policy proposals and institutional design whether we do “dumb-onomics” or “smart-onomics.” Clearly, dumb-onomics is the only one consistent with the Hume-Smith tradition.

    Algorithmic information theory bolsters dumb-onomics and the we’re-too-stupid version of asset market unpredictability. This theory points out that we are often unable to compress data. A theory compresses data and lets you get the answer ahead of time, instead of just waiting for events to unfold. But it turn out that for computability reasons, we can’t always do much data compression. I can compress 3.14159 . . . by telling you to compute the ratio of a circle’s circumference to its diameter. But most numbers cannot be compressed in this way and are, in that sense, “random.” If the human mind is finite, then we can’t do much data compression and most events will look random to us, even if they are “determined” in some metaphysical sense.

    For a user friendly introductions to these issues, see

    Chaitin, Gregory. 2006. “The Limits of Reason,” Scientific American, March 2006.

    Koppl, R. (2008). Thinking impossible things: A review essay, Journal of Economic Behavior and Organizaiton, 66, 837-847.


    Click to access sciamer3.pdf

  3. I always enjoy Roger’s posts, but I winced at his choice of words. Is Lucas man smarter than Hayek man? If I don’t know something, I am not stupid, but ignorant. As our late friend and colleague, Ludwig Lachman, was fond of putting it: the future is unknowable, but not unimaginable. Men imagine the future in asset markets. So the difference is in the conception of how particpants interact, i.e., how markets work. And the conflation of knowledge and imagination.

  4. These physics worshipers need to learn some physics, i.e. complex non-linear dynamics.

    Physicists know that a sandpile with every increasing sand piling on its top will eventually collapse and spill down its sides — but there physics tells them they can not know when it will do that, the timing is not predictable. There is a lot of physics like this. They can know what will happen, and that what will happen will have systematic characteristics, but they can’t know when it will happen or what its precise details will be.

    That’s science — what Lucas is dishing up is something else.

  5. In science, theory can’t always do this:

    “A theory compresses data and lets you get the answer ahead of time, instead of just waiting for events to unfold.”

    The three body problem is a problem within Newtonian mechanics, a paradigm of a scientific theory. And how things come out is not compressible. We have here both theory and non compressibility.

  6. Jerry,

    I think ignorance and “stupidity” are both issues. Of course “stupidity” is a bit of a jocular term. But I think Hayek was more sensitive to our cognitive limits than Lucas and, generally, ratexers tend to be. I realize that ratex does not necessarily say anything about individual cognition and instead refers to the distribution of expectations. But there is no room in ratex to seek out the consequences of cognitive infirmity and I think ratex models sometimes assume more cognitive prowess than we really have. This was an issue in my Big Players work with Butos, Yeager, and others. People had trouble understanding why we didn’t just “model” the Big Player, in which case we would see that the small players could simply use that same model. You have to make a diagonal argument such as Hayek used at the end of The Sensory Order to show why that approach is mistaken. So you have computability issues in Hayek, but not Lucas. See also Markose’s paper in the 2005 EJ symposium on computability issues.

  7. Roger: The primary definition of stupidity is “lacking ordinary activity and kenness of mind.” It is an inherently relative concept, and suggests that “ordinary” people know more. Not everyone can be stupid, but everyone can be ignorant. I’m suggesting the real contrast is between Lucas and Lachmann about what is knowable. Hayek falls somewhere between the two. We agree that ratex does not specify how much individuals know. Both Lucas and Hayek agree that markets gather specific information, which is reflected in prices. I’m not saying that Hayek himself believed in ratex (though, in an early paper, Lucas credited Hayek with anticipating the theory). But Hayek is one source of thought for ratex. Lachmann, clearly not.

  8. I think we’re just keen on different, but complementary points, Jerry. Plus I made the mistake of using the word “stupid,” thinking it would be all punchy and clever. But I was just talking about bounded rationality, and we can all have bounded rationality. Admittedly, I was talking about a particular kind of bounded rationality whereby the imagined bounds are equal to what a universal Turing machine can compute. But I think that’s still bounded rationality.

  9. Greg Ransom,

    You make the point that models of self-organized criticality cannot predict the future. Quite true. But don’t such models have instrumental value in the sense that they show us how a handful of parameters can affect the statistics of the system as a whole? For example, the size of grain and their stickiness affect the avalanche statistics.

    If I had to pick one “worst mistake” of rational expectations theory it would be its lack of instrumentalism — i.e., its tendency to close down debate about why market prices are changing in time. In this sense, models inspired by nature (including SOC models) cannot but be an improvement since they ask new questions.

  10. I really like Roger’s point, with Jerry’s modification of the language. But I should point out that in “Individualism: True and False,” when Hayek is talking about Smith’s program in political economy he does invoke the word “stupid” — or more accurately “more often stupid” to describe man and why robust institutions are so important in inducing us.

    However, I do want to raise a different issue. During the period of 1990 to say 2005 the actually experienced economic life was “pretty good” (put aside the distortions that were being created). At the HES meetings I actually saw a paper presented by Roger Backhouse and Brad Bateman on Keynesianism. In that presentation it was pointed out that Keynesian ideas in public policy were reintroduced into the policy world in the late Clinton and early Bush years. In short, prior to the reintroduction of Keynesianism, the policy rules were governed by the ideas of modern macroeconomics. Think of that as the Taylor Rule of inflation targetting in monetary policy.

    To a modern macroeconomist such as Lucas then couldn’t he argue that “we” in fact had done well from the mid-1980s thru the 1990s until politics intervened and discretionary policy deviated from the policy rules of ratex?

    So rather than our current mess being a mess caused by modern macroeconomics, it is a mess by the resurrection of older ideas in macroeconomics that also gained political acceptance again. We deviated from the teachings of modern macroeconomics and we ended in this mess and the culprit is Keynesian justified discretionary policy.

    Isn’t this what we see when we look at John Taylor’s Getting Off Track? Thought experiement is what would have happended in our world IF Fed policy did not deviate from the Taylor Rule?

  11. Economic history teaches us two things.

    First, financial/economic crises happen, perhaps once every 10 years or so.

    Second, no one – not economists, not speculators, not regulators, not politicians – can predict the advent of a crisis in any way that is timely or useful for preventing that crisis.

    There is a logical reason for this – and I think this is Lucas’ main point. Any predictive value an economic theory has will be quickly incorporated into the expectations of economic agents – and thus washed away.

    These facts give discussion of the “causes of this mess” a certain other-worldly quality. Have you noticed how the “causes” of each successive crisis change from crisis to crisis? What is the significance of this phenomenon? I think it is more evidence for Lucas’ assertion that successful macroeconomic cycle prediction is a forlorn hope.

    I know, I know. All crises are caused by investment that, ex post, is unproductive. But even if true in some sense, this dogma is of little use for prediction.

  12. Pete raises an interesting issue. My first reaction is to question the Great Moderation thesis. In my paper on the crisis, published now in the Cato Journal, I point out the many financial crises here and abroad. Just in the period 1990-2005, we had the Asian and Russian financial crises; the failure of LTCM; the dotcom bubble and bust; and the housing bubble. Note the crises began occurring more often than the 10-year cycle suggested by Carl Futia. We know there is one common cause: easy credit. (So we cannot forecast them, but we can predict them). Additionally, we now have an increasingly unstable financial system due to the government’s underwriting the major players. I’m not sure this is Keynesianism.

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 )

Google photo

You are commenting using your Google 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