Krugman on Keynes

In 2007 the Royal Economic Society (in conjunction with Palgrave Macmillan) published a new edition of J M Keynes's, The General Theory of Employment, Interest and Money,1 with a new introduction by this year’s winner of the Nobel Prize for Economics,2 Paul Krugman. In that introduction, Krugman argued that while some aspects of the The General Theory are better understood against the peculiar circumstances of the 1930’s, the fundamental ideas are widely accepted — ‘we are all Keynesians now’. Since that introduction was written, there have been dramatic changes in the world economy which make The General Theory even more relevant. To mark Krugman’s honour we are printing a shortened version of his essay.

The message of Keynes
It is probably safe to assume that the ‘conservative scholars and policy leaders’ who pronounced The General Theory one of the most dangerous books of the past two centuries have not read it. But they are sure it is a leftist tract, a call for big government and high taxes. That is what people on the right, and some on the left, too, have said about The General Theory from the beginning. ....

But Keynes was no socialist — he came to save capitalism, not to bury it. And there’s a sense in which The General Theory was, given the time it was written, a conservative book.... Keynes wrote during a time of mass unemployment, of waste and suffering on an incredible scale. A reasonable man might well have concluded that capitalism had failed, and that only huge institutional changes — perhaps the nationalization of the means of production — could restore economic sanity. Many reasonable people did, in fact, reach that conclusion: large numbers of British and American intellectuals who had no particular antipathy toward markets and private property became socialists during the depression years simply because they saw no other way to remedy capitalism’s colossal failures.

Yet Keynes argued that these failures had surprisingly narrow, technical causes. ‘We have magneto [alternator] trouble’ he wrote in 1930, as the world was plunging into depression. And because he saw the causes of mass unemployment as narrow and technical, he argued that the problem’s solution could also be narrow and technical: the system needed a new alternator; there was no need to replace the whole car....

Stripped down, the conclusions of The General Theory might be expressed as four bullet points:

• Economies can and often do suffer from an overall lack of demand, which leads to involuntary unemployment

• The economy’s automatic tendency to correct shortfalls in demand, if it exists at all, operates slowly and painfully

• Government policies to increase demand, by contrast, can reduce unemployment quickly

• Sometimes increasing the money supply won’t be enough to persuade the private sector to spend more, and government spending must step into the breach.

To a modern practitioner of economic policy, none of this — except, possibly, the last point — sounds startling or even especially controversial. But these ideas were not just radical when Keynes proposed them; they were very nearly unthinkable. And the great achievement of The General Theory was precisely to make them thinkable.

How Keynes did it
....Keynes’s struggle with classical economics was much more difficult than we can easily imagine today. Modern macroeconomics textbooks usually contain a discussion of something they call the ‘classical model’ of the price level. But that model offers far too flattering a picture of the classical economics Keynes had to escape from. What we call the classical model today is really a post-Keynesian attempt to rationalize pre-Keynesian views. Change one assumption in our so-called classical model, that of perfect wage flexibility, and it turns back into The General Theory. If that had been all Keynes had to contend with, The General Theory would have been an easy book to write.

The real classical model, as Keynes described it, was something much harder to fix. It was, essentially, a model of a barter economy, in which money and nominal prices do not matter, with a monetary theory of the price level appended in a non-essential way, like a veneer on a tabletop. It was a model in which Say’s Law applied: supply automatically creates its own demand, because income must be spent. And it was a model in which the interest rate was purely a matter of the supply and demand for funds, with no possible role for money or monetary policy. It was a model in which ideas we now take for granted were literally unthinkable.

If the classical economics Keynes confronted had been what we call the classical model nowadays, he would not have had to write Book V of The General Theory, ‘Money-wages and prices’. In that book Keynes confronts naïve beliefs about how a fall in wages can increase employment, beliefs that were prevalent among economists when he wrote, but play no role in the model we now call ‘classical’.

So the crucial innovation in The General Theory is not, as a modern macroeconomist tends to think, the idea that nominal wages are sticky. It is the demolition of Say’s Law and the classical theory of the interest rate in Book IV, ‘The inducement to invest’. One measure of how hard it was for Keynes to divest himself of Say’s Law is that to this day some people deny what Keynes realized — that the ‘law’ is, at best, a useless tautology when individuals have the option of accumulating money rather than purchasing real goods and services. Another measure of Keynes’s achievement may be hard to appreciate unless you’ve taught introductory macroeconomics: how do you explain to students how the central bank can reduce the interest rate by increasing the money supply, even though the interest rate is the price at which the supply of loans is equal to the demand? It is not easy to explain even when you know the answer; think how much harder it was for Keynes to arrive at the right answer in the first place.

Mr Keynes and the moderns
There is a widespread impression among modern macroeconomists that we have left Keynes behind, for better or for worse. But that impression, is based either on a misreading or a non-reading of The General Theory. Let us start with the non-readers, a group that included me during the several decades that passed between my first and second readings of The General Theory.

If you do not read Keynes himself, but only read his work as refracted through various interpreters, it is easy to imagine that The General Theory is much cruder than it is. Even professional economists, who know that Keynes was not a raving socialist, tend to think that The General Theory is largely a manifesto proclaiming the need for deficit spending, and that it belittles monetary policy. If that were really true, The General Theory would be a very dated book. These days economic stabilization is mainly left to technocrats in central banks, who move interest rates up and down through their control of the money supply; the use of public works spending to prop up employment is generally considered unnecessary. To put it crudely, if you imagine that Keynes was dismissive of monetary policy, it is easy to imagine that Milton Friedman in some sense refuted or superseded Keynes by showing that money matters.

The impression that The General Theory failed to give monetary policy its due may have been reinforced by John Hicks, whose 1937 review essay ‘Mr Keynes and the classics’ is probably more read by economists these days than The General Theory itself. In that essay Hicks interpreted The General Theory in terms of two curves, the IS curve, which can be shifted by changes in taxes and spending, and the LM curve, which can be shifted by changes in the money supply. And Hicks seemed to imply that Keynesian economics applies only when the LM curve is flat, so that changes in the money supply do not affect interest rates, while classical macroeconomics applies when the LM curve is upward-sloping.

But in this implication Hicks was both excessively kind to the classics and unfair to Keynes. I have already pointed out that the macroeconomic doctrine from which Keynes had to escape was much cruder and more confused than the doctrine we now call the ‘classical model’. Let me add that The General Theory does not dismiss or ignore monetary policy. Keynes discusses at some length how changes in the quantity of money can affect the rate of interest, and through the rate of interest affect aggregate demand. In fact, the modern theory of how monetary policy works is essentially that laid out in The General Theory.3

Yet it is fair to say that The General Theory is pervaded by skepticism about whether merely adding to the money supply is enough to restore full employment. This was not because Keynes was ignorant of the potential role of monetary policy. Rather, it was an empirical judgment on his part: The General Theory was written in an economy with interest rates already so low that there was little an increase in the money supply could do to push them lower.

Many of today’s most prominent macroeconomists came of intellectual age during the 1970s and 1980s, when interest rates were consistently above 5 per cent and sometimes in double digits. Under those conditions there was no reason to doubt the effectiveness of monetary policy, no reason to worry that the central bank could fail in efforts to drive down interest rates and thereby increase demand. But The General Theory was written in a very different monetary environment, one in which interest rates stayed close to zero for an extended period.

Modern macroeconomists do not have to theorize about what happens to monetary policy in such an environment, or even plumb the depths of economic history, because we have a striking recent example to contemplate. There are hopes as I write this that the Japanese economy may finally be staging a sustained recovery, but from the early 1990s at least through 2004 Japan was in much the same monetary state that the US and UK economies were in during the 1930s. Short-term interest rates were close to zero, long-term rates were at historical lows, yet private investment spending remained insufficient to bring the economy out of deflation. In that environment, monetary policy was just as ineffective as Keynes described. Attempts by the Bank of Japan to increase the money supply simply added to already ample bank reserves and public holdings of cash while doing nothing to stimulate the economy. (A Japanese joke from the late 90s said that safes were the only product consumers were buying.) And when the Bank of Japan found itself impotent, the government of Japan turned to large public works projects to prop up demand.

Keynes made it clear that his skepticism about the effectiveness of monetary policy was a contingent proposition, not a statement of a general principle. In the past, he believed, things had been otherwise. ‘There is evidence that for a period of almost one hundred and fifty years the long-run typical rate of interest in the leading financial centres was about 5 percent, and the gilt-edged rate between 3 and 3 ½ percent; and that these rates were modest enough to encourage a rate of investment consistent with an average of employment which was not intolerably low.’ [307-308] In that environment, he believed, ‘a tolerable level of unemployment could be attained on the average of one or two or three decades merely by assuring an adequate supply of money in terms of wage-units.’ [309] In other words, monetary policy had worked in the past — but not now.

Now it is true that Keynes believed, wrongly, that the conditions of the 1930s would persist indefinitely — indeed, that the marginal efficiency of capital was falling to the point that the euthanasia of rentiers was in view. I will mention why he was wrong in a moment.

Before I get there, however, let me consider an alternative view. This view agrees with those who say that modern macroeconomics owes little to Keynes. But rather than arguing that we have superseded Keynes, this view says that we have misunderstood him. That is, some economists insist that we’ve lost the true Keynesian path — that modern macroeconomic theory, which reduces Keynes to a static equilibrium model, and tries to base as much of that model as possible on rational choice, is a betrayal of Keynesian thinking.

Is this right? On the issue of rational choice, it’s true that compared with any modern exposition of macroeconomics, The General Theory contains very little discussion of maximization and a lot of behavioural hypothesizing. Keynes’s emphasis on the non-rational roots of economic behaviour is most quotable when he writes of financial market speculation, ‘where we devote our intelligences to anticipating what average opinion expects average opinion to be.’[156] But it is most notable, from a modern perspective, in his discussion of the consumption function. Attempts to model consumption behaviour in terms of rational choice were one of the main themes of macroeconomics after Keynes. But Keynes’s consumption function, as laid out in Book III, is grounded in psychological observation rather than intertemporal optimization.

This raises two questions. First, was Keynes right to eschew maximizing theory? Second, did his successors betray his legacy by bringing maximization back in?

The answer to the first question is: it depends. Keynes was surely right that there is a strong non-rational element in economic behaviour. The rise of behavioural economics and behavioural finance is a belated recognition by the profession of this fact. On the other hand, some of Keynes’s attempted generalizations about behaviour now seem excessively facile and misleading in important ways. In particular, he argued on psychological grounds that the average savings rate would rise with per capita income (see p. 97.) That has turned out to be not the case at all.

But the answer to the second question, I would argue, is clearly ‘no’. Yes, Keynes was a shrewd observer of economic irrationality, a behavioural economist before his time, who had a lot to say about economic dynamics. Yes, The General Theory is full of witty passages about investing as a game of musical chairs, about animal spirits, and so on. But The General Theory is not primarily a book about the unpredictability and irrationality of economic actors. Keynes emphasizes the relative stability of the relationship between income and consumer spending; trying to ground that stability in rational choice may be wrong-headed, but it does not undermine his intent. And while Keynes didn’t think much of the rationality of business behaviour, one of the key strategic decisions he made, as I have already suggested, was to push the whole question of why investment rises and falls into the background.

What about equilibrium? Let me offer some fighting words: to interpret Keynes in terms of static equilibrium models is no betrayal, because what Keynes mainly produced was indeed a static equilibrium model. The essential story laid out in The General Theory is that liquidity preference determines the rate of interest; given the rate of interest, the marginal efficiency of capital determines the rate of investment; and employment is determined by the point at which the value of output is equal to the sum of investment and consumer spending. ‘[G]iven the propensity to consume and the rate of new investment, there will be only one level of employment consistent with equilibrium.’ [28]

Let me address one issue in particular: did Paul Samuelson, whose 1948 textbook introduced the famous 45-degree diagram to explain the multiplier, misrepresent what Keynes was all about? There are commentators who insist passionately that Samuelson defiled the master’s thought. Yet it is hard to see any significant difference between Samuelson’s formulation and Keynes’s own equation for equilibrium employment, right there in Chapter 3:

Represented graphically, Keynes’s version looks a lot like Samuelson’s diagram; quantities are measured in wage units rather than constant dollars, and the nifty 45-degree feature is absent, but the logic is exactly the same.

The bottom line, then, is that we really are all Keynesians now. A very large part of what modern macroeconomists do derives directly from The General Theory; the framework Keynes introduced holds up very well to this day.


Editor’s Notes:

1. The book is available to members at a discount.

2. Strictly speaking, the Sveriges Riksbank Prize in Economic Sciences in Memory of Alfred Nobel

3. See, for example, the diagram in Bank of England Quarterly Bulletin, May 1999 p. 163. The ECB publishes a diagram which is virtually identical but interestingly, in the light of recent events, it begins with the need for the central bank to be able to influence market rates. See The Monetary Policy of the ECB, 2004 (Frankfurt: ECB) p.43.


Page Options