Famous economists’ famous errors

Bob T., a loyal MR reader, asks the following:

10 (or more) most famous mistakes in economics.
Viner on costs and Feldstein on Social Security come to mind. Malthus? Not
talking about old vs. new economics, but simple analytical errors and bad

That’s a good start.  What else might be listed?  Just to circumvent various hobby horses in the comments section, let’s avoid Marx and Marxists, Keynes, and the last twenty years. 

1. Kenneth Arrow confusing risk subdivision and risk multiplication, in arguing that government should use a riskless rate of discount.

2. The Cambridge, Mass. economists having to admit, finally, that capital reswitching could be quite a general phenomenon (though is it, really?)

3. Ricardo’s prediction that most of national output would end up going to the landlords.

4. Paul Samuelson praising the economic performance of Soviet central planning in his Principles text.

5. 93 percent of all proclamations made about the demand for money in macroeconomics.

6. The more exaggerated claims about the Laffer Curve.

7. Various claims that the Fed should have let the money supply fall during the Great Depression.

8. Jevons’s claim that England (or was it the world?) would soon run out of coal.

9. Welfare analysis done in overlapping generations models (the standard welfare theorems do not generally hold in such models).

And dare I offer up a controversial pick:?

10. Those who think that the difference between "capital" and "ideas" in a Solow growth model is actually well-defined.

What else can you think of?


The belief, common to most macroeconomists 30 years ago, that the national debt is not a burden on future taxpayers "because we owe it to ourselves", and at the same time believing that bond-financed tax cuts increased net wealth and so would increase consumption demand.

Mises saying that middle-of-the-road policy would lead to socialism. Some will say that 2008 makes him look better, but not much better, I think.

Keynes wrote something about a future of agricultural shortages, I think in Economic Consequences of the Peace.

The second comment adds Paul Ehrlich's bet on the future price of commodities, but he was not an economist.

In fact, it was the economist Julian Simon that pointed out the folly and subsequently won the bet.

Dow 36000 is the one who stand out in my mind.

But one would argue Modigliani-Miller theorem deserves some mention in the list given the current evidence.

How about Starve the Beast? I'd also argue against Friedman's 1953 'The Methodology of Positive Economics' but I suspect I'm in a very small minority on that one.

Anything that Abba P. Lerner wrote, including the letter
"w," as in "we owe it..."

The long list of efficient markets theorists, which includes Samuelson.

Anyone who ever used calculus to try to theorize (and that's the horrible word for it) about economics.

Whoever founded econometrics.

That a consumption tax would lead to more productivity than an income tax. Taxing people on the front end introduces more risk and uncertainty.

Often, it's just too soon to tell.

perhaps not in the top 20, but on the Solow growth model:

XX) Everybody who uses Solow's model for *anything* without reading and understanding Robert Ayres' papers on it. Really, economists should try and understand useful work and energy, even though most will dogmatically resist even the idea.

as for Laffer, why even include him? The Laffer curve error, if attributed should probably be attributed to Keynes, no? There was nothing new about the Laffer curve, afaik, only the way it was politically misused. The blame on that should fall on the politicians - not that Mr Laffer himself escapes without major brain bloopers:


In general I think we should post this as numero uno:

1) Building a theory with total disregard to reality and when proven wrong by fact from the real economy proceeding to claim that economy is 'wrong' and actors are 'irrational'

That is the cardinal sin.

"Building a theory with total disregard to reality..."

Hey, maybe I'm not the only one who sees Friedman (1953) as a mistake!

Mistakes in economics is a dull subject.
The truly electrifying inquiry would be to name 10†¦ well†¦ 5†¦ OK just 3 incidents when economics made indisputably correct (in a rigorous scientific sense – since economics dare to call itself a science) prediction/forecast on a more or less significant scale.

Noah Yetter beat me to the punch with the Phillips Curve, but I'll nominate the Gordon growth model for its assumption of a constant (and permanent) rate of dividend per share growth.

On a closely associated note, any assumption linking GDP growth to the growth of stock prices.

Also, for first place, any economic theory predicated on stochastic assumptions where interdependencies among the underlying data are evident.

Don't you mean to say that they had to admit the formal reality of reswitching (as a formal property of their formal models of "capital"), in the first instance. The empirical reality of the phenomena is a secondary question. On this secondary question, I might recommend a paper by Roger Garrison, "Reflections on Reswitching and Roundaboutness":


>>The Cambridge, Mass. economists having to admit, finally, that capital reswitching could be quite a general phenomenon (though is it, really?)<<

This would also be an interesting list:

What are the most famous mathematical blunders of famous economists?

Who would be on the list?

Samuelson got caught in a famous mathematical error, if I'm not mistaken. Who else?

Let's add Irving Fisher:

"The stock market crash of 1929 and the subsequent Great Depression cost Fisher much of his personal wealth and academic reputation. He famously predicted, a few days before the Stock Market Crash of 1929, "Stock prices have reached what looks like a permanently high plateau." Irving Fisher stated on October 21st that the market was "only shaking out of the lunatic fringe" and went on to explain why he felt the prices still had not caught up with their real value and should go much higher. On Wednesday, October 23rd, he announced in a banker’s meeting “security values in most instances were not inflated.† For months after the Crash, he continued to assure investors that a recovery was just around the corner."

-- from Wikipedia

Hayek's biggest formal economic mistake?

Assuming that Bohm-Bawerk's non-micro version of capital theory -- his "average period of production" -- could serve as a non-problematic short-hand stand in for the micro capital theory in Hayek's work on the trade cycle.

Hayek's biggest professional mistake?

Assuming that British and American economists were capable of understanding what he was doing. By 1936 Hayek seems aware that there was a profound disconnect there. Hayek's most influential papers "Economics and Knowledge", the scientism papers, "The Use of Knowledge in Society" were written as attempts to bridge the gap. In the 1960s, 70s, and 80s Hayek seems to admit to himself that -- at most fundamental level -- he'd effectively failed in his effort to be understood by his peers.

Don't forget John von Neumann's 1955 prediction that, thanks to the advent of nuclear power, "in a few decades hence energy may be free...."

Quoted from "Can We Survive Technology? [June 1955] Fortune, reprinted in John von Neumann, "John von Neumann on Technological Prospects and Global Limits [Mar. 1986], Population & Development Review 12:117, 120.

Doomsayers like Ehrlich are not the only fortune tellers who get the future spectacularly wrong.

A special prize in 1974-75. The newly installed President Ford was derided as an idiot, but look who's talking?

Ford was facing a problem of high inflation and a developing severe recession. He called a summit of the country's leading economists and personally presided over the meeting, asking ideas about what should be done. Maybe Obama will do the same!

In the first meeting, they all asked for tighter monetary and fiscal policy to fight inflation. The recession widened and deepened, and Ford again called a summit, and the economists called for easier monetary and fiscal policy. Ford sat there clenching his pipe in his teeth. Poor deluded man thought economists had something useful to say.

The economists understood aggregate demand, I think, because of the heavy emphasis from the Keynesian years. But they did not get aggregate supply, whether it was the role of oil prices, the impact on supply of inflation, the distortions of price/wage controls, the Phillip's curve, the importance of taxation on supply, or labor search theories.

I like to think we understand aggregage supply better. Do we?

I like #7, because creating money to buy worthless assets totally makes those assets worth while.

Who knows, after the Fed is done slamming the presses mabye the old quants will go on sale again?

Speaking of Baumol: I would think that the concept of `Contestability' could also be counted as an `error', or at least as something wildly oversold at the time.

Friedman: ending the mandatory licensing of doctors

The entire set of economists who thought Black Scholes Merton Option pricing models with its idiotic Gaussian return assumptions to be useful, should qualify as people who made a big mistake no? Surely if LTCM related mess caused the current crisis as Taylor speculates, the Black Scholes Merton mistake must count as pretty grave.

Ozornik, I'll take you up on that challenge. Here are a few predictions of economics that I would regard as (1) experimentally well-confirmed and (2) non-trivial enough that non-economists have, either explicitly or implicitly, denied them:

That price controls (e.g. rent control) lead to under-supply, non-cash settlements, and Pareto-inefficient allocations.

That, becuse of Laffer effects, "static scoring" will consistently under/over-estimate the revenue effects of tax hikes/reductions.

That specialization and trade are, in aggregate, welfare-enhancing.

That marginal rather than average effects determine economic behaviors.

That PPP predicts currency exchange rates in the long run.

Oh, and I agree with other commenters that the Phillips curve definitely belongs at the top of any list of embarassments.

The randomised trials of development projects of Ms.Duflo and co and the idea that to "do development economics" is meaningful than theorising development economics from a policy point of view.

"10. Those who think that the difference between "capital" and "ideas" in a Solow growth model is actually well-defined." --- Tyler Cowen

1. As Notsneaky noted, a relevant concern here is what Tyler means by "well-defined." Within the Solow model, capital-inputs clearly refer to physical capital as one input to growth in per capita income over time . . . conceptualized in standard neoclassical form as a homogeneous aggregate, whose contributions could be measured by its marginal product. And of course the growth in labor is the other input . . . it too clearly defined in quantity terms, at any rate in the original non-augmented Solow model. (Don’t worry. I’ll get to the role of “ideas† or knowledge-capital in a few seconds.)

2) In noting all this, I pass over the assumptions that quickly became contestable: competitive markets, constant returns to scale, and the controversies over the nature of "homogeneous" capital and whether it could be measured in a neo-classical sense.


Because here, it seems, Milton Friedman's instrumentalist view of economic theories and modeling applies to Solow's "exogenous growth model". Namely: we have no easy way, maybe any way, to assume the correspondence-truth between any theory in the sciences --- including social science --- and the slice of the "real" world we are studying. What counts are two other things:

(i) Does a model lead to good predictions that other models don't about economic behavior?

(ii.) And does it inspire good follow-up empirical work that leads to fruitful research and improvements or not in the original model and, no less important, new spin-off research and augmented additions to the theory or improved theories.


3. Note here, to drive home these points out an instrumental approach as opposed to a "realist" epistemology that Friedman and all other monetarists have always had trouble pinning down the "true" definition of money: M1, M2, M3, and "near money" that has proliferated in the last 20 years.

Similarly, no one really believes that your average consumer, investor, or firm-manager has as much knowledge about the economy's workings as a Nobel prize-winning economist's model . . . a postulate at the core of rational-expectations theory, with in fact "average" itself an abstract concept (non-measurable in this case) and with additionally the "model" that is assumed not specified. The epistemological issue here is whether this Lucas-inspired theory and its offshoots --- with its emphasis on informational-problems (or breakdown) to explain recessions or inflationary bouts --- leads to continued useful research and good predictions or not . . . always compared with other, competing theories.


4) Back to the Solow model.

What that model does entail, as Dennison and others quickly showed, is that the "residual" in growth-accounting models was a large part of what explained the differences across various countries' per capita income growth over several years . . . longer anyway than a short-erm business cycle.

That finding, a fruitful outcome, underscored a key point that Solow didn't emphasize as much: the "measure of ignorance" attaching to the residual would be interpreted as first total factor productivity and later multi-factor productivity. Either way, it drew attention to the role of technology as the long-term driving force in economic growth . . . a major advance in economic modeling, including if you want the challenge to exogenous growth theory launched by "endogenous" theory of the Romer sort. And our knowledge of what accounts for the differences in per capita income across the world’s 200 countries increased considerably.


5) More concretely, the use of “multifactor productivity† led to additional or refined variables in the inputs to growth: (i) labor quality; (ii.) technological progress understood as new knowledge (ideas if you want), whether embodied in machines or in better ways to organize and manage firms; and (iii), more recently, “social capital† . . . institutional influences, including cultural factors as to why there are more risk-taking entrepreneurs in some countries than others.
In short, the Solow growth model has been a progressive paradigm that has inspired a great deal of improved understanding of why, since the late 18th century,

(i.) Malthusian economics was transcended by first European and European-offshoot countries and more recently in Asia and to an extent in parts of Latin America

(ii.) And why convergence catch-up growth, entailed by the original Solow model, has in various augmented offshoots only occurred in certain formerly backward countries . . . where human capital is sufficiently advanced to work diligently with the latest technologies and where sufficient social capital and certain governmental policies have exploited the opportunities created by convergence processes. Such as, despite the huge corruption and CP power-monopoly in China, governmental policies and Chinese human capital that led to a steady infusion of technological transfers from abroad, not least in multinational implants, and a fair amount of diffusion within and across certain manufacturing sectors.

(iii.) Similarly, by extension, the augmented social capital (or institutional) growth model suggest why, in the next decade or two --- perhaps more quickly still --- China’s economic model of growth will probably reach a crisis-level major changes in institutional obstacles are drastically altered . . . above all, the huge obstacle of a CP power monopoly and of the pervasive corruption and lopsided growth domestically that it has entailed across regions, between city-and-countryside, and among social classes, not to forget the astonishing harm to the environment that clearly can’t continue without a major heatlh-crisis.


6) Against this background, the failed predictions in the original Solow model --- to stay with an instrumentalist view of economic theories) --- aren’t significant. Above all, the original theory and model entailed a prediction that convergence catch-up growth would lead to a future in which the levels of productivity and per capita income would be similar across the countries of the world.

It was this failure --- along with a growing interest in the “residual† and augmentations or modifications to the original model --- that led to the kinds of understanding that we have today about the matters discussed in 5) above.


Michael Gordon, AKA, the buggy professor


You are telling us what, that people do not go before boards of companies to propose scenarios of capital
investments under differing assumptions about the rate of interest, or whatever you or somebody else wants
to label it? Such things go on all the time. And, it is easy to construct scenarios where reswitching will
occur under such scenarios. The key is having periods of both positive and negative net returns at different
times in the future. Do companies never face such scenarios?

In 1966 Samuelson admitted in the QJEthat he had made a mistake earlier in his assertions about the aggregate
capital in regard to the reswitching debate. He concluded in his confession that "economic theory is built upon
foundations of sand." The first time I met him (in the early 1970s) I asked him about the controversy, and he
said that one should consider heterogeneous capital and the returns on each form of capital.

In his Pure Theory of Capital, Hayek came to abandon the traditional version of the roundaboutness theory on
the basis of recognizing the basic phenomenon that underlies the reswitching problem, the reality of complicated
streams of net returns arising from various capital investments.

Greg Ransom has provided a link to Robert Vienneau on empirical examples of reswitching. One of those was due
to me and Raymond Prince back in 1985 in Growth and Change between cattle grazing and strip mining of coal in
the southwestern US, based on then current numbers. The switch points were 2 and 7 percent, within real world
relevant ranges.

"9. Welfare analysis done in overlapping generations models (the standard welfare theorems do not generally hold in such models)."

I don't get this. The welfare theorems don't hold, but that doesn't mean one cannot do welfare analysis. The original Samuelson article did exactly that and did it right.

The idea that tax cuts do not increase tax revenue.

Any tax cut that results in economic stimulus will eventually increase tax revenue, it's just a question of when.

Barkley, if you look closely, I think you'll see that in the first instance in both Hayek and Bohm-Bawerk "roundaboutness" is simply about choice theory and a single good -- and the ability of a delay in consumption of a good to produce a superior ability to consume at a later date. It's an insight conceived in real terms, but not in an "economy". It's a very profound but limited microeconomic insight, it's not full blown capital and interest theory within an equilibrium framework.

You drink a jug of grape juice, or you put it aside and drink wine down the road.

No complicated streams. No variety of capital investments.

One might argue that there is a background to this choice situation (and Hayek did), but the "roundaboutness" insight in the first instance is conceived in isolation from this background.

I'll get you a page citation from Hayek on that. (And maybe Bohm-Bawerk)

Barkley wrote:

"In his Pure Theory of Capital, Hayek came to abandon the traditional version of the roundaboutness theory on
the basis of recognizing the basic phenomenon that underlies the reswitching problem, the reality of complicated
streams of net returns arising from various capital investments."

Here are the two most egregious errors in the last sixty years: -- one of commission, the other of omission:
1. The (emotional?) fascination of general equilibrium theory which has thrown macro, growth, and welfare economics into dead ends;
2. Not developing the fundamental implications of the Impossibility Theorem for a better understanding of welfare theory and politics which has led to all kinds of overreliance on bad economic advice over the years.
Of course, both of these issues have Arrow stamped all over them.

... or was that Krugman? Hard to keep them straight.

A great deal of what John Kenneth Galbraith wrote 1-2 generations ago reads as ludicrous today.

All those who edit our very top journals (famous by definition?) who too cavalierly (and frequently) publish "theory" that clearly isn't; for example, on the "paucity of theory in the journal of economic theory" see:


Marx's argument that the profit is exploitation of surplus labor. Wrong. Popular, too.

wonderful article.

Comments for this post are closed