Lucas on Keynes Via De Vroey

Print/Save PDF

I am on my annual get-away-from-Chicago-winter sojourn in San Miguel, Mexico. As suggested by the previous post, part of my poolside recreation has been rereading parts of Michel De Vroey’s A History of Macroeconomics from Keynes to Lucas and Beyond that most interested me the first time around. This post concerns one of the most arresting bits: De Vroey on Lucas’s assessment of Keynes’ contribution to macro modeling. Coming clean about my own priors, I believe that Keynes did more than any other theorist to establish macroeconomics as a distinct branch of economic theory. He and his work were of the utmost importance in making the  field a useful guide to the behavior of highly specialized economies.

It is then probably unsurprising that  I believe Robert Lucas’ well-known public harshness toward Keynes and Keynesianism was childishly nasty, way out of line with what the great British theorist and his followers actually accomplished. This famous example is from a Lucas speech published in 1980: “Keynesian economics is dead…. You cannot find a good, under 40 economist who identifies himself and his work as ‘Keynesian’. Indeed, people even take offense if referred to in this way. At research seminars, people don’t take Keynesian theorizing seriously any more – the audience starts to whisper and giggle to one another.” Insults, beyond being gratuitous and false, are particularly hard to take when coming from a theorist whose own accomplishments pale in comparison to his favorite piñatas, Keynes and the formidable Early Keynesians.

That statement, of course, needs elaboration. Let’s be bold and illustrate my assessment with Lucas’ most lionized work: the use of rational expectations to “destroy” both the Early Keynesian Phillips curve and EK macroeconomics itself. Instead of the promised devastation, it is Lucas’ analysis turns out to be wrong. If you are going to be uncivil in tearing down others while praising yourself, your modeling has better hold water. (It doesn’t; see below.)

Objections to Keynes. De Vroey identifies three major Lucas criticisms of Keynes and The General Theory. “First, Lucas criticized Keynes for abandoning his initial aim of constructing a theory of the business cycle by redirecting his attention to the apparently simpler issue of explaining the existence of involuntary unemployment at one point in time.” If macroeconomists (and policymakers) believe that mass involuntary unemployment is the most salient characteristic of business cycles, it is not unreasonable to organize analysis of cyclicality around that phenomenon. Would Lucas have economists concerned with the welfare costs of massive forced layoffs simply ignore such unemployment? (Oops, that is indeed what he recommends.)

“Lucas’s second indictment of Keynes bore on his abandonment on the market-clearing assumption and its underpinning by what he calls the ‘equilibrium discipline,’ the decision to depict agents as behaving in an optimizing way.” Especially in evidence-consistent modeling, there is no justification for assuming that “market clearing” and the “equilibrium discipline” are the same thing. Lucas conveniently assumes that demonstrably false equivalency.

To Lucas, Keynes’ modeling of consumption, investment, liquidity preference, labor pricing and use were little more than rules of thumb, that added up to “bad social science: an attempt to explain important aspects of human behavior without reference either to what people like or what they are capable of doing.” (Lucas (1981)) This is at best a self-involved criticism that ultimately argues that Lucas’ preferred class of model-building (mathematical derivation) is inherently superior to Keynesian descriptions based on close observation of the economic activity in question. The Project’s analysis of worker preferences when engaged in routinized employment is illustrative of how to model human behavior: construct a mathematical spine around which to rationally organize close observation of the activity in question. Mathematics alone is ill-suited to go it alone in insightfully modeling human behavior, an outcome vividly  demonstrated by Lucas’ own work.

“A third conceptual criticism addressed by Lucas to Keynes is that his analysis was based on ill-defined and basically useless concepts, in particular those of involuntary unemployment and full employment. In his eyes, it would have been better had these concepts never been introduced in the theoretical lexicon.” Sympathetic students of Lucas cannot avoid confronting his conclusion that macroeconomics would be better if the subject of involuntary unemployment were banned. It may seem reasonable at this point to search for a civil version of stupidity. As a believer in civility, I have sought a kinder response. In another context, I hypothesized perhaps the only reasonable excuse for Lucas at his most bizarre. Quoting myself: “Lucas is too careful to deny the obvious existence of involuntary job loss (IJL). His relevant quote (1981, p.243) is: ‘Involuntary unemployment is not a fact or a phenomenon which it is the task of theorists to explain.’ I believe he is [correctly] arguing that meaningful  IJL cannot exist in friction-augmented general market equilibrium (FGME) modeling. If theorists choose to work within that framework, which he believes Keynes did not, the fact of IJL must be ignored.” Note that if my defense of Lucas is indeed the best available, his macro analysis should always be accompanied by a red-letter warning: DANGEROUS FOR USE IN REAL-WORLD DECISION-MAKING. For any practical use, it is really bad social science.

The lesson. The Lucas affair earns attention from the GEM Blog mostly because it illustrates a debilitating mistake made by some of the most energetic modern market-centric macro theorists. In making the case for adherence to the classical tenets of optimization and equilibrium in proper model-building, they boldly push aside whether or not their theories align with the evidence. In making microfoundations the only model characteristic that matters, they willingly accept models that badly mislead private and public decision-makers

The Fed which I spent many years advising never hid their scorn for Lucas’ modeling, especially during macro crises. In my later career in the profit-seeking financial sector, I would have been fired if I tried to convince the other members of the management committee or our clients that millions of involuntary unemployed is not something about which we should be concerned.            

As promised above. If readers are curious about the falsity of Lucas’ famous destruction of EK macroeconomics by injecting rational expectations into the EK Phillips Curve, the case will be made in the next series of posts on the GEM Blog. It is not a difficult argument.

Blog Type: Lucas’ Macro Theory

Write a Comment

Your email address will not be published.