I’m racing against the clock. TPM café is closing at midnight and I’ve got a lot to say and I want to say it. There may be more people reading this stuff than if I had the lead article in the American Economic Review – something I know is never going to happen. That makes this too good an opportunity to miss. Somehow or other the mainstream of the economics profession is temporarily outflanked by this weird “internet” stuff.

At the same time I’m dealing with Brad DeLong, who is a heavyweight in the style of Muhammad Ali - “float like a butterfly, sting like a bee”. He jabs straight and hard with references to the early Keynes, feints with what looks like a Jim Tobin upper-cut, and effortlessly pulls a rope-a-dope with references to Marx. I know I’ve got to be careful, but I’m feeling pretty good. I studied Keynes at Oxford and Yale. DeLong studied at Harvard.

In an earlier post I referred to the absence of papers on macroeconomics with a Keynesian structure published in the American Economic Review since the early 1980s. And in a later post I described “new Keynesianism” as a form of intellectual cuckoo that took over the Keynesian nest and pushed out real Keynesian ideas. Brad DeLong has challenged these claims. Minor details aside, the real issue is the legacy of Keynes, and that is what I want to write about.

My starting point is that Keynes’ General Theory, published in 1936, represents a dramatic break in his intellectual life. In effect, there is a pre-General Theory Keynes and a post-General Theory Keynes, though of course the later Keynes grew out of his thinking about issues tackled earlier in The Economic Consequences of Mr. Churchill (1925) and the Treatise on Money (1930). What’s the evidence for this claim? The last paragraph of the beautifully crafted Preface that comes at the very beginning of The General Theory – not near the beginning, or in the middle, or at the very end. Here, Keynes bares his soul:

“The composition of this book has been for the author a long struggle of escape, and so must the reading of it be for most readers if the author’s assault upon them is to be successful, -- a struggle of escape from habitual modes of thought and expression. The ideas which are here expressed so laboriously are extremely simple and should be obvious. The difficulty lies, not in the new ideas, but in escaping from the old ones, which ramify, for those brought up as most of us have been, into every corner of our minds.”

That says it pretty clearly. Keynes felt he had made a break with his intellectual past.

That in turn raises the question of what’s new in The General Theory. Here are my top three things. First, the concept of demand determined equilibrium whereby producers’ expectations of aggregate demand determine the level of output. That means if producer demand expectations are low an economy can get trapped below full employment.

Second, the construction of the relationship between saving and investment that allows increased investment demand to draw forth the output and saving to fund it. This means saving is not a constraint on investment if an economy has spare productive capacity.

Third, Keynes’ monetary theory of interest rate determination, which emphasizes the role of liquidity preference in determining interest rates.

Any great theorist must do three things. First, take down the theory of opponents: second, put up a new theory: and third, make sure the new theory is properly defended. The General Theory does one and two very well, but within a decade of publication it began to run into trouble on three. That trouble concerns the question of whether price and nominal wage adjustment can restore full employment, and that question has become the defining issue of macroeconomics and the legacy of Keynes.

It is also patently clear from Chapter 19 of The General Theory that Keynes felt price and nominal wage flexibility could not solve the problem, and would in all likelihood worsen it. However, Keynes did not give this issue enough attention and the arguments in Chapter 19 are inadequate. This lack of attention may have been because Keynes felt the answer was self-evident given the deflation of the 1930s and the persistence of unemployment. Additionally, The General Theory had lots of other fish to fry, getting across the new theoretical accounts of output and interest rate determination.

Immediately after publication the focus of debate was on interest rate determination, as evidenced by the exchanges with Denis Robertson. One set of issues concerned how interest rates are determined – the so-called “loanable funds vs. liquidity preference” controversy. The second set of issues concerned whether lower interest rates could restore full employment, and here the focus was on interest rate floors (the famous liquidity trap) and the sensitivity of spending to lower interest rates.

Keynes won both of these arguments handsomely, in part with the help of the famous ISLM model introduced by John Hicks (1937). That model quickly became accepted as the appropriate interpretation of The General Theory, but it is also where the rot set in and Keynes’ legacy was nabbed by orthodoxy. This is because Hicks’ ISLM representation of The General Theory, allowed lower prices and nominal wages to restore full employment. That was shown by Franco Modigliani in his famous 1944 paper, and Modigliani was a seminal figure in “MIT Keynesianism” that DeLong refers to and was schooled in.

The ISLM model has had a truly major historical intellectual impact. On the good side it helped Keynes win the early interest rate debates with Denis Robertson. Its elegant mathematical and diagrammatic formulation also meant it was quickly adopted in classrooms, thereby spreading Keynes’ insights about the determination of output and the relation between saving and investment.

However, on the downside it fundamentally undermined Keynes’ argument by reducing his theory of unemployment to a special case in which prices and wages were downwardly rigid. This gave rise to so-called “fix-price Keynesianism”, which is MIT Keynesianism and not the economics of The General Theory.

A second problem with the ISLM model is its representation of the monetary sector and interest rate determination, which does not do justice to the mechanism of liquidity preference and the role of the banking system in interest rate and output determination. This latter problem has lbeen emphasized by heterodox Post Keynesian economists, but it has not been recognized by orthodox economists or MIT Keynesians.

For a long while the fix-price misinterpretation of Keynes did not matter much. This was because in the twenty-five years after World War II everyone considered themselves Keynesian from a policy perspective. The assumption was prices and nominal wages are downwardly rigid in practice and policy has to live with that. Therefore, since full employment cannot be restored by lower prices and nominal wages because they are downwardly rigid, policy should do it by Keynesian methods of stimulating demand.

All of this worked well until the late 1960s when Milton Friedman and the Chicago boys came to power. They claimed that prices and nominal wages were not downwardly rigid and the economy could quickly return to full employment by itself. Government should get out of the way and let prices fluctuate.

Additionally, the Chicago School implicitly noted that if fixed prices and nominal wages are the problem, then policy should make prices and nominal wages flexible so that full employment can rule. That opened the way to what has become known as the “labor market flexibility” agenda – bust unions, eviscerate the minimum wage, erode unemployment insurance, and remove employee protections. That agenda aims to make wages flexible, thereby ensuring full employment.

What looks like a microeconomic policy program in fact has its rationale in macroeconomics. Moreover, MIT Keynesians (what East Anglian Cambridge U.K. Keynesians called “bastard Keynesians”) also have to go along with this flexibility agenda because that is what is implied by their fix-price Keynesianism.

This leads to “new Keynesianism” about which DeLong raves in his “Chopped Liver” post. The so-called new Keynesians attempt to save fix-price Keynesianism, but in doing so they have moved even further away from Keynes. Whereas the “old” MIT/bastard Keynesians simply assumed prices and wages were downwardly sticky, new Keynesians seek to explain price stickiness in terms of rational optimizing behavior. To do this they reach into the microeconomic tool kit of rational choice and profit maximizing firms.

This has had significant effects on economics. First, it has changed the methodology of macroeconomics, causing the abandonment of the old Keynesian macroeconomic modeling paradigm based on market demand curves – hence my claim that no papers of this type have been published by the American Economic Review since the early 1980s. In place of earlier macroeconomic models, new Keynesians now use applied general equilibrium models that turn macroeconomics in the direction of orthodox neo-classical microeconomics.

Most importantly, unemployment now arises because of frictions on the supply side of the economy that prevent prices and nominal wages from adjusting when demand fluctuates. That makes unemployment significantly a supply-side phenomenon, which has nothing to do with The General Theory.

Additionally, new Keynesians appear to significantly abandon Keynes’s core theoretical innovations about interest rate determination, diminishing the role of liquidity preference. Instead, there is slippage back to loanable funds theory whereby interest rates are determined by balancing saving and investment to ensure full employment. That certainly seems to be the hallmark of new Keynesian policy discussion, with its emphasis on need for increased saving and budget deficit phobia. That said, new Keynesianism is a cottage industry so there may be models with a tweak or twos difference – which is why it is so slippery and hard to pin down.

From a policy standpoint there are two implications. One implication is to smooth demand and avoid demand fluctuations. The other implication is to reduce supply-side frictions. This is because the logic of new Keynesian models tends to suggest that increased price and nominal wage flexibility likely lowers unemployment – though there can be exceptions to this. The models also unambiguously imply that perfect price and wage flexibility guarantees full employment, which is definitely not what Keynes wrote in Chapter 19.

This brings me to DeLong’s invocation of the late James Tobin as supporting new Keynesianism. For most of his professional life Tobin was an MIT Keynesian. However, he never embraced the microeconomic parable methodology of new Keynesianism, and in the latter part of his life he seems to have moved away from MIT Keynesianism. In a famous paper from 1975 Tobin showed that deflation could be destabilizing, but perfectly flexible prices would restore full employment. However, by 1980 in his book Asset Accumulation and Economic Activity, Tobin had moved away from that position. Effectively, Tobin discovered the impact of debt (the so-called Fisher debt effect) - perhaps as a result of conversations with the heterodox economist Hyman Minsky. Owing to debt burdens, perfectly flexible prices may not restore full employment and macroeconomics has recovered the legacy of Keynes.

The second great piece of the Keynes debate concerns the theory of income distribution. Here, for heterodox economists it is not a case of claiming the legacy of Keynes, but rather a case of correcting the error of Keynes.

Today’s orthodoxy supports the marginal product theory of income distribution – which essentially states that competitive markets ensure workers are paid their contribution to output. Reading The General Theory there can be little doubt that Keynes believed in marginal product theory. That is the logic of Chapter 3 describing the determination of wages and employment, and it is also the logic behind Chapter 11 describing the demand for capital..

For heterodox economists Keynes got the theory of income distribution wrong, and heterodox economists should acknowledge that. However, Keynes got macroeconomics right, and price and nominal wage flexibility may not be able to ensure full employment. That means capitalist economies can get trapped with unemployment, as happened in the 1930s Great Depression. This creates a role for government stabilization policy.

From a heterodox perspective the bottom line is that New Keynesians reject the part that Keynes got right (price flexibility is unlikely to restore full employment) and hold on to the part he got wrong (marginal productivity theory of income distribution). The so-called “sterile” Cambridge U.K. Keynesian program reverses that, recognizing the part Keynes got right and working to correct the part he got wrong.

Has the MIT Keynesian program been more successful and fruitful? Yes, absolutely when measured in terms of its professional domination – but it may also be teaching things that just ain’t so, which means the fruit is rotten.

Lastly, getting back to the original focus of Chris Hay’s article, DeLong openly admits that MIT Keynesianism is part of the economics profession mafia in its belief that it is legitimate to prevent of graduate students getting exposure to the so-called “fuzzy thinking” of Cambridge U.K.