What Is Essential about Keynes Today?
Annals of the Fondazione Luigi Einaudi, Vol. LI, pp. 5-16 | Thursday, June 01, 2017

In this paper, a transcript of an address given at the 2016 Keynes Conference in Turin, I unpick the various elements of Keynesian revival in economics and explore how different schools of economists have disagreed on which aspects of Keynes’ thought are relevant to modern economic theory. Finally, I conclude with four thoughts on the relevance of Keynes for how one does economics.
There has been a modest revival of interest in Keynes after the collapse of 2008-2009 and the subsequent mediocre ‘recovery’. I daresay graduate students no longer giggle at mention of his name as they were said to in Robert Lucas’s graduate seminar in Chicago (Lucas 1980: 18). However, we must be careful to unpick the different elements of Keynesian revival.
Alan Coddington usefully distinguished between hydraulic and fundamental Keynesians (Coddington 1976). Let me relabel hydraulic Keynesians ‘policy Keynesians’, and divide the fundamentalists into two groups: those interested in Keynes’s theory and those interested in his methodology. There is a large overlap between the two groups of fundamentalists, but we can consider them separately.
1. Policy Keynesians
First, though, the policy Keynesians. I must be careful, as people don’t like being classified or pigeon-holed, but these include Paul Krugman and Joe Stiglitz in the United States; Simon-Wren Lewis, Martin Wolf and David Vines in the UK. Policy Keynesians reject the idea that economies are automatically self-balancing at full employment. Aggregate demand can be deficient, as it is in Europe today. Since 2008 they have been strong critics of fiscal austerity.
But to them, theory is secondary to policy. They have abandoned Keynes’ theory, and instead tinker with the neoclassical model in any way which will justify stimulus policy. They explain unemployment as simply the result of sticky wages and prices, which rapid price adjustment to exogenous shocks. Ultimately, the economy is self-correcting, but in the short-run there is policy space for governments to exploit the short-run Phillips Curve.
They are known as the New Keynesians. Politically they are mainly on the Left. However, I would point out that there is nothing specifically Keynesian about stimulating an economy in a slump.
Keynes himself was a policy Keynesian before he was a theory Keynesian. Many economists, including those at Chicago University, advocated unbalancing the budget during the Great Depression, i.e., before Keynes wrote his General Theory, while maintaining the orthodox view that there would be no unemployment if wages and prices were completely flexible.
Easing monetary policy in a slump was accepted central bank practice long before Keynes. Friedman later claimed that open market operations (today known as quantitative easing) was prevented not by the theory of the time, but by weak leadership at the Fed (Friedman and Schwartz 1963).
In other developed countries, the main obstacle to QE in the Great Depression was commitment to the gold standard because the gold standard was thought to be inflation proof, and as British politician Philip Snowden put it: “the microbe of inflation is always in the air” (Elliot 1993: 42). But the gold standard was strongly challenged by ‘monetary reformers’ before the Great Depression, mainly Fisher (Fisher 1911) and Wicksell (Wicksell
1898). In 1923 Keynes called it a “barbarous relic” (Keynes 1923: 172). All this occurred before the General Theory.
In short, pre-Keynesian (though not pre-Keynes) orthodox theory was sufficiently flexible to provide all kinds of escape clauses when the going got rough, without having to abandon the theory itself.
We now move to the 1950s and 1960s, the era of the neoclassical synthesis. The policy Keynesians were in power. The leading theoretician was Paul Samuelson of textbook fame (Samuelson 1948). Essentially, he bolted macroeconomics onto the existing neoclassical theory. The policy commitment was to maintain full employment mainly by fiscal policy. What did the policy Keynesians take from the General Theory?
First, and foremost, they took a framework of aggregate relationships between consumption, investment, money, etc. yielding a determinate level of income and employment.
The war had given them the necessary statistical concepts and tools to maintain continuous full employment in a closed economy – national income statistics, output gap calculations, multiplier arithmetic. It had also bequeathed a repressed financial system and international institutions like the IMF and World Bank for open economy stabilization.
Second, Hicks claimed that Keynes’s aggregate framework could yield either the classical full employment or the Keynesian ‘special case’ of unemployment equilibrium (Hicks 1937).
Keynes had asserted that, in face of a demand shock, full employment could not be maintained by flexible money wages and interest rates. Even if money wages responded flexibly to an economy wide-shock they could not bring about the required adjustment in the average real wage because a fall in average money wages would cause a proportional fall in aggregate demand, leaving the real wage unchanged. Liquidity preference would prevent interest rates falling enough to restore full employment.
Postwar skirmishes between the economists of the day knocked away these theoretical supports for the unemployment equilibrium possibility. To take just one example: Keynes had asserted that the rate of interest, being the price of money, would rise rather than fall in a slump, owing to an increase in liquidity preference. Against this, it was claimed by Modigliani and others that a reduction in money wages would lead to a fall in demand for transactions balances (Davidson and Smolensky 1964). This would offset the rise in liquidity preference, allowing a fall in the rate of interest. Whether this fall was brought about by a cut in money wages, or, preferably by an increase in money was theoretically unimportant.
In the upshot, Keynes’s theoretical assault on neo-classical theory was deemed to have failed, while leaving the case for stabilization policy intact. This case rested on the existence of sticky wages. Keynes’s theory was a ‘special case’ of the neoclassical theory, of great practical and political significance, but of minimal theoretical importance.
One had only to assert that, with (unexplained) time-lags in the price-adjustment process, any policy which averted a slump was good. This was the essence of the truce between Keynes and the Classics. They carried off the theoretical honours, he won the policy war. The aim of macro policy in the Keynesian era was to maintain full employment. And the policy Keynesians could claim this success: average unemployment was low throughout the Keynesian age.
As this age came to an end, scuppered by the stagflationary episodes of the 70s and 80s, the New Classical and monetarist schools emerged from the wreckage, promising that a laissez-faire approach to policy would yield prosperity and stability.
Working within the boundaries set by the New Classical school – micro-founded optimisation models, rational expectations etc. – the New Keynesians tried to rescue some Keynesian policy space from within a non-Keynesian paradigm, by arguing for small modifications of these models which were meant to acknowledge “market imperfections”. They succeeded in their somewhat limited goal, as central banks moved away from monetarism towards medium term inflation-targeting regimes, creating room for active stabilisation policy. Apart from some theoretical work on the causes of sticky wages and prices, then, the new policy Keynesians are heirs to the neoclassical synthesis of the 1950s.
It has to be said that the New Keynesians were not unduly dissatisfied with the orthodox policies pursued prior to the crash of 2008, since they believed that acceptance of the short-run Phillips curve was the vital element in the central banks’ inflation targeting policy, which they chose to regard as a form of monetary Keynesianism.
So in the battles of the 1980s and 1990s, the New Classicals carried off the theoretical honours, but the New Keynesians won some policy space via monetary policy.
Since the crash, the New Keynesians have used the income determination framework to mount devastating assaults on the policy of fiscal austerity, and have belatedly accepted the weakness of monetary stabilisation. In face of continuous stagnation in much of the developed world, official bodies like the IMF, OECD, and even the ECB have started to call for more expansionary fiscal policy, so New Keynesians feel they are at last starting to claw back their losses in the policy war.
2. Theory Keynesians
The theory Keynesians – often called Post-Keynesians – rest their case for a Keynesian theoretical revolution on certain strands in the General Theory, notably chapters 12 and 17, ‘The State of Long Term Expectations’ and ‘The Essential Properties of Interest and Money’ respectively (Keynes 1936). To these must be added Keynes’ 1937 article The General Theory of Employment (Keynes 1937).
They claim direct descent from one of Keynes’s closest collaborator Joan Robinson, who called the policy Keynesians ‘bastard Keynesians’ (Robinson 1962). There are many strands in this movement, but a leading figure has been Paul Davidson.
The main concern of the theory Keynesians is not with the stability of Keynes’s short-run unemployment equilibrium, but the instability of the capitalist system, i.e. they reject all theorising on the basis of equilibrium. Hyman Minsky’s financial instability hypothesis (Minsky 1992) has added to this theoretical arsenal and has been used to explain the pre-crash fragility of the banking system as against Eugene Fama’s neoclassical efficient market hypothesis model (Fama 1970).
This strand of Keynesian fundamentalism asserts the claims of Keynes as a revolutionary theorist chiefly by reference to his theory of uncertain expectations. Uncertain expectations makes investment radically unstable. As Keynes put it:
By ‘uncertain’ knowledge […] I do not mean merely to distinguish what is known for certain from what is only probable […]. The sense in which I am using the term is that in which the prospect of a European war is uncertain, or the price of copper and the rate of interest twenty years hence, or the obsolescence of a new invention, or the position of private wealth-holders in the social system in 1970. About these matters there is no scientific basis on which to form any calculable probability whatsoever (Keynes 1937: 213-214).
Faced with this situation we “[save] our faces as rational, economic men” by relying on conventional judgments, such as that “the existing state of opinion as expressed in prices and the character of existing output is based on a correct summing up of future prospects”. But the flimsiness of the convention means that it is subject to “sudden and violent changes”. The breakdown of the conventional valuations leads to a flight to liquidity, with money as the “barometer of the degree of distrust in our own calculation and convention concerning the future” (ibid.: 214-216).
It was the uncertainty of the investment function, not sticky wages, which prevented full employment. Since such uncertainty was inherent in the human condition, unemployment, or at least underemployment, was the normal state of affairs. Only a theory which incorporated uncertain expectations could be regarded as general.
In the fundamentalist picture it was the classical theory which assumed full employment which was the limiting case, when the aggregate relationships were in such an alignment as to enable it to occur by accident. Thus Keynes writes in the General Theory:
During the nineteenth century, the growth of population and invention, the opening up of new lands, the state of confidence and the frequency of war…seem to have been sufficient, taken in conjunction with the propensity to consume, to establish a schedule of the marginal efficiency of capital which allowed a reasonably satisfactory level of employment to be compatible with a rate of interest high enough to be psychologically acceptable to wealth-owners (Keynes 1936: 307).
If one takes the General Theory and ‘post-GT’ formulations together with Keynes’s Treatise on Probability (Keynes 1921), one can see that in Keynes’ thought, radical uncertainty itself is a limiting case at the opposite extreme from statistical probability and then absolute certainty, with a rich territory in between.
What New Classicals have done is to identify probability with statistical frequency. All probability is a matter of calculable risk. This move abolishes Keynes’ large intermediate ground of ordinal probability, in which he thought most economic decision-making took place. With ordinal probability, one has enough information to know that something is more or less likely to happen than something else, but not that it is twice or three times as likely.
As a parenthesis, this is the justification which Kaldor for using “stylized facts” as a basis for theorising (Kaldor 1961). This was true to Keynes’s own method.
Rod O’Donnell has argued that whereas statistical frequency is an untenable basis for economic theorising, so it is wrong to attribute to Keynes the general existence of a state of radical uncertainty (O’Donnell 1989). Keynes’ general theory of probability takes ordinal probability to be the general case. The future is not determined but neither is it totally unknown. In most cases, it is indeterminate. If it were completely unknown the market economy would be completely chaotic, which it is not, and the case for the political imposition of certainty on economic actors, in the form of central planning, would become overwhelming.
There is a large policy cost in not taking Keynesian/Knightian (Knight 1921) uncertainty seriously, e.g. by reducing it to statistical risk. The assumption that banks could calculate accurately the risks they were running underpinned the intellectual case for the extensive deregulation of the financial sector and indifference to the explosion of derivative instruments (see e.g. Turner 2009). George Soros rightly noted that the banking collapse of 2007-8 was not due to an exogenous shock but was endogenous to
an unregulated financial system (Soros 2009). And one may similarly comment about the continued indifference of policy makers to the explosion of household debt.
3. Methodology
The insertion of uncertainty into the heart of economics, even in the ‘unradical’ form of ordinal probability, has a profound implication for not just economics, but how one does economics – the territory of methodology.
Uncertainty sets the economic problem to which policy has to respond. Two consequences:
1. It destroys marginalism – and indeed mathematics – as a general, comprehensive method, since marginalism depends on calculable probabilities.
2. It destroys the Robbins definition of economics (Robbins 1932) as the science of scarcity. On the contrary, it strongly suggests that a market economy is dominated by excess supply, i.e. that Say’s Law does not hold as a general rule.
One can see, therefore, that it is wrong to see Keynesian macroeconomics as a bolt on to pre-existing microfoundations. It eliminates the validity of the microfoundational approach, except in special cases.
An analogy here is the relationship between the theory of money and the real economy. Money is not something one adds to the real economy, whose processes can be explained without money. It must enter the explanation “on the very ground floor” (Schumpeter 1954: 278), precisely because of money’s special role in mediating uncertainty.
Against this radical interpretation of Keynes’s method, it will be said that he used marginalist concepts: marginal propensity to consume, marginal efficiency of capital, marginal utility and disutility of wages. Even the decision to hold money was a decision taken at the margin.
But I think it is more accurate to think of Keynes’s marginalism as window dressing, a disguise adopted for political reasons within the profession.
Take the propensity to consume: this is an average, not a marginal concept. The marginal propensity to consume is the tool Keynes introduces for a specific policy purpose: to work out the value of the multiplier. It does no other work in his theory short-run theory, though it has distributional implications for his undeveloped long-run thinking.
Or take the marginal efficiency of capital. This is an expectational, not physical quantity. The inducement to invest in his theory depends on the state of confidence as governed by convention and animal spirits. Large and continuing divergence is possible between expectational equilibrium and ‘objective’ equilibrium because there is no calculable way of getting from the former to the latter.
The preference for holding money does not depend on the kind of calculation of net returns to different classes of assets as in portfolio choice theory, but is the form taken by animal spirits when confidence has collapsed. It is a ‘flight into liquidity’.
Keynes talks about the wage being equal to the marginal utility of labour and the marginal disutility of work. But in fact, there is evidence that he conceived of wages as a normative standard, socially determined by historical and institutional factors (Henry 2015). Stickiness is inherent in the wage system because, as Keynes wrote, the human apple, unlike the stone, has its own ideas about how fast, or whether, it should drop to the ground (Keynes 1938).
Given these facts about the human condition, how should one theorise about economics? To conclude, I have four thoughts:
First: if one tries to place Keynes in the history of economic thought, he appears to be a classical, rather than neoclassical economist, i.e. as someone concerned with aggregate problems – problems of growth, the full use of resources, money and the distribution of wealth. Allocation of scarce resources at the margin becomes important only when the economy is making full use of potential resources; but marginalism is not the mechanism which ensures full use, and therefore economic theorising should not start in this place.
Second, economic theory should start with the organic, not atomic hypothesis. As individuals we are interconnected parts of complex wholes. There is a feedback process; our own decisions affect others’ decisions, whose decisions in turn affect our own. Through these constant interactions the whole shapes the individual parts.
No one can read Keynes – and indeed, neither Anna Carabelli nor Mario Cedrini (e.g. Carabelli and Cedrini 2014) – without becoming aware of the pervasive role in his thinking played by the fallacy of composition. Yet the policy Keynesians have recognised this only in one form: the paradox of thrift. I would submit that behavioural economics, with its focus on the individual, does not properly address the challenge of the interconnectedness of economic decision making. The task for policy is to ensure that the planets are in the right orbits in the macroeconomic universe.
Third, one needs to take seriously Keynes’s claim that economics is a moral science. That is, it deals with introspection and judgments of value; or as I would say, its domain includes the ends of life, and not just the efficient means of realising those ends.
What is essential about Keynes, in other words, includes his essay The Economic Possibilities for our Grandchildren (Keynes 1930a, 1930b), which poses the crucial question: what is economic growth for? To which Keynes’s answer is: “to live wisely, agreeably, and well” (ibid.).
Economics, and economists, must stop trying to separate means from ends and then avoid any discussion of the latter. This is both irresponsible and, ultimately, futile. One should remember that Keynes started as an ethicist and philosopher, and saw economics as a preparation for the good life. This approach has huge implications for policy today in the light of the planetary and environmental challenges we face.
Finally, one should, wherever possible theorise in ordinary language appropriate to the study of indeterminacy. Mathematics is an attempt to give precision to what cannot be made precise, and which if used for that purpose, leads to the wrong advice for policy. It lends economics a false air of objectivity and, as the discipline formalises and ossifies, closes off the possibility for full, open and sound economic debate.
So the challenge, I would submit, is to write a textbook which starts with ‘macrofoundations’, rather than with the microeconomics of individual choice at the margin, and derives behaviour at the micro level from behaviour of the whole. Whether this can be done without surrendering the claim of economics to disciplinary independence I do not know. It certainly spells the end of economics imperialism – the attempt by economics to infiltrate its peculiar methodology into the other social sciences – and suggests rather a sociological and historical economics, cutting across present disciplinary divides.
But that should not surprise or bewilder us. The arrangement of the subjects of study into separate disciplines is already breaking down. At some point, we will see a rearrangement to suit the requirements of the good life of the future.
Carabelli A.M. and M.A. Cedrini 2014, “Chapter 18 of The General Theory ‘further analysed’: economics as a way of thinking”, Cambridge Journal of Economics, 38 (1): 23-47.
Coddington A. 1976, “Keynesian Economics: The Search for First Principles”, Journal of
Economic Literature, 14 (4): 1258-1273.
Davidson P. and E. Smolensky 1964, “Modigliani on the Interaction of Monetary and Real
Phenomena”, The Review of Economics and Statistics, 46 (4): 429-431.
Elliot G. 1993, Labourism and the British Genius: the Strange Death of Labour England?, London: Verso.
Fama E.F. 1970, “Efficient Capital Markets: A Review of Theory and Empirical Work”, The
Journal of Finance, 25 (2): 383-417.
Fisher I. 1911, The Purchasing Power of Money, New York, NY: Macmillan.
Friedman M. and A.J. Schwartz 1963, A Monetary History of the United States, 1867-1960, Princeton: NJ: Princeton University Press for NBER.
Henry J.F. 2015, “Classical Political Economy: The Subsistence Wage, and Job Guarantee
Concerns”, Journal of Post Keynesian Economics, 38: 280-301.
Hicks J.R. 1937, “Mr. Keynes and the ‘Classics’: A Suggested Interpretation”, Econometrica,
5 (2): 147-159.
Kaldor N. 1961, “Capital Accumulation and Economic Growth”, in F.A. Lutz and D.C.
Hague (Eds.), The Theory of Capital, New York, NY: St. Martin’s Press: 177-222.
Keynes J.M. 1938, “Letter to R.F. Harrod, 4 July 1938”, in E. Johnson and D. Moggridge (Eds.), The Collected Writings of John Maynard Keynes, vol. 14: 299-300. Cambridge: Cambridge University Press.
– 1937, “The General Theory of Employment”. The Quarterly Journal of Economics, 51 (2):
– 1936, The General Theory of Employment, Interest and Money, London: Palgrave Macmillan.
– 1930a, “Economic Possibilities for our Grandchildren I”, The Nation and Athenaeum, 48 (2): 36-37.
– 1930b, “Economic Possibilities for our Grandchildren II”, The Nation and Athenaeum, 48 (3): 96-98.
– 1923, A Tract on Monetary Reform, London: Macmillan.
– 1921, A Treatise on Probability, London: Macmillan.
Knight F.H. 1921, Risk, Uncertainty and Profit, Boston, MA and New York, NY: Houghton
Mifflin Company.
Lucas R.E. 1980, “The Death of Keynesian Economics”, Issues and Ideas, Winter 1980. Minsky H. P. 1992, “The Financial Stability Hypothesis”, Working Paper No. 74, Jerome
Levy Economics Institute, Annandale on Hudson, NY.
O’Donnell R.M. 1989, “Keynes on Probability, Expectations and Uncertainty”, in R. M.
O’Donnell (Ed.), Keynes as Philosopher-Economist, London: Macmillan: 3-60.
Robbins L. 1932, An Essay on the Nature and Significance of Economic Science, London: Macmillan.
Robinson J. 1962, “Review of H.G. Johnson’s Money, Trade and Economic Growth”, Economic
Journal, 72: 690-692.
Samuelson P.A. 1948, Economics, New York, NY: McGraw-Hill.
Schumpeter J.A. 1954, History of Economic Analysis, London: Allen & Unwin.
Soros G. 2009, “Do Not Ignore the Need for Financial Reform”, Financial Times, Oct. 25. Turner A. 2009, “The Turner Review: A Regulatory Response to the Global Banking
Crisis”, Financial Services Authority.
Wicksell K. 1898, Geldzins und Güterpreise, Jena: Gustav Fischer.