January 21, 2022
Marc Lavoie, University of Ottawa and University of Sorbonne Paris Nord
When Stephen Marglin came to Paris in 2016, he presented parts of his book to students and teaching staff of the University of Paris 13, where I had just arrived. My souvenir is that it was difficult to understand what he was up to, and I remember that it was even more challenging to follow the comments of the professor who had been assigned the task of starting the discussion period! Marglin sent me the pdf files of the draft of his various chapters, but I must say that I was daunted by the task of reading such a huge work and (forever) set it aside. This symposium has now forced me to remedy to this, focusing, as we have been asked, on Marglin’s chapter 13, ‘Taking money seriously’. I take it that the symposium is not about whether an economy with perfect competition, flexible wages, flexible prices, and flexible interest rates can on its own come back to a position of full employment, as discussed in detail by Marglin in his earlier chapters.
In a nutshell, my understanding of what Marglin is saying is that the analysis of liquidity preference and of money in Keynes’s General Theory is a mess: ‘There is no theory of money in the General Theory. Or rather there is no coherent theory of money’.[1] I never thought of putting it this way, but for a start, I myself also always thought that Keynes’s chapter 17 on own rates of interest was ‘impenetrable’, as Marglin puts it.[2] Like Marglin, I also got frustrated at Keynes assuming that the money supply was one of his three independent variables. Now, there are scholars out there, for instance Sheila Dow, who tell us that Keynes took the money supply as given, and not as exogenous, and that this makes a huge difference.[3] This distinction is overly subtle for my second-language English, so I could never make any sense of it. All I could resolve was that Keynes was well aware that the central bank could just as well set its sights on the short-term interest rate instead of the money supply, as is obvious from a reading of his previous Treatise on Money, as pointed out in particular by Basil Moore in his book on endogenous money,[4] but that for some reason he chose to do otherwise in the General Theory.
More surprising perhaps is Marglin’s assessment that Keynes’s liquidity preference theory was rather unsuccessful and that ‘Keynes doesn’t deliver on his promise to replace the conventional theory of interest, in which productivity and thrift rule the roost, by his own theory of liquidity preference’.[5] But we should only be half surprised. Keynes’s liquidity preference theory in his General Theory was nearly an afterthought. Confronted with the classical theory in which variations in the rate of interest brought investment and saving into equilibrium, Keynes asserted instead that changes in income ensured this equality. As a consequence, the determination of the rate of interest was left in the air, and it took some time for Keynes to come up with a theory that would be an alternative to the conventional theory of the interest rate.[6] But because Keynes limited his analysis of the motives for money to instantaneous positions rather than to a process analysis, this led Keynes to be embroiled in a controversy with his critics, who accused him of having neglected the need for the creation of liquidity, thus inducing Keynes to invent what he called the finance motive for money, the role of which is well explained by Augusto Graziani when making the distinction between the initial finance of production or financement on the one hand, and final finance or saving on the other hand.[7]
There are two issues tackled by Marglin with which I agree wholeheartedly. The first is the statement that ‘the quantity equation holds, but not the quantity theory’.[8] The quantity equation is simply the identity MV = PQ , whereas the quantity theory adds the claim that causality runs from M to P, meaning that increases in the stock of money M will lead to increases in the price level P, a claim found in all mainstream introductory textbooks. This claim is rejected by Marglin and myself. The alternative view, endorsed since 1830 by authors such as Tooke and Fullarton and nowadays by post-Keynesian economists, is that causality runs the other way, from the level of nominal income PQ towards the level of the stock of money. In other words, the level of economic activity in nominal terms determines the demand for money, which in turn determines the stock of money that will be supplied by the banking system and the monetary authorities.
The second issue is also related to causality. Within the Wicksellian framework, the real rate of interest must converge towards the natural rate of interest, set by the confrontation of productivity and thrift. In Marglin’s view, the real rate of interest is determined by the nominal interest rate decided by the monetary authorities less the rate of price inflation, determined outside of the monetary realm. Linked to this is Marglin’s claim that: ‘The rate of interest turns out to be a conventional phenomenon rather than a reflection of market equilibrium….There is no “natural rate of interest” that emerges from the free play of market forces and regulates the accumulation of capital’.[9] I have argued so much by writing that ‘the rate of interest is in fact a distributive variable based on monetary conventions’.[10] Alfred Eichner similarly said that the basic rate of interest, meaning the target fed funds rate, ‘is a politically determined distributional variable rather than a market determined price’,[11] with my fellow Canadian, John Smithin, adding that ‘the real economy must adjust to the policy-determined interest rate, rather than vice-versa. This is therefore the precise opposite to the natural rate doctrine’.[12]
Further to Keynes’s theory of liquidity preference, Marglin asserts that Keynes’s theory of interest is ‘full of ambiguities’ and ‘at best incomplete: it is not a theory of the rate of interest but a theory of interest spreads’.[13] This statement is repeated elsewhere a number of times, for instance when Marglin says that ‘liquidity preference becomes a theory of interest rate differentials or spreads’[14] or when he adds that ‘liquidity preference could never be a theory of interest …liquidity preference can never be more than a theory of interest-rate differentials’.[15]
While I fully agree with Marglin, I think that he overstates the originality of his claim. Marglin gives the impression that his reformulation is a revelation, a new and important discovery: at long last somebody has cleared up the concept of liquidity preference by adding short-term assets that carry interest in addition to long-term assets and cash – an insight that he attributes to a comment made by Dennis Robertson in 1935. Thus, liquidity preference is about interest rate spreads and not about the absolute value of the rate of interest. But we have known this for a long time. For instance, in a book published 30 years ago,[16] I argued that:
The long-term rate of interest, or more precisely the spread between the long-term and the short-term rates of interest, reflects the liquidity preference of households and non-banking financial institutions.[17] According to this view, the spread between long-term and short-term rates is determined by the preferences of the rentiers towards liquid or illiquid positions… We can sum up the above by saying that liquidity preference sets all other rates of interest around the level determined by the central bank. The central bank acts as a sort of price leader, which the banks follow more less loosely. This means that liquidity preference sets the term structure of interest rates.[18] But it cannot set the base rate, upon which all the other rates depend. The base rate itself, that is the discount rate or the money rate set by the central bank, is the truly exogenous factor. It is a convention.
Obviously, as we can see from this quote, other authors have understood and argued a long time ago that liquidity preference does not determine the interest rate, it could only determine spreads, unless we maintain that the target rate set by the central bank results from its own liquidity preference, presumably influenced by various lobbying groups, such as the CEOs of commercial banks. The fact that liquidity preference determines spreads only reinforces the rejection of the concept of the natural rate of interest, since it means that the long-term rate of interest then depends essentially on the base rate set by central bankers, provided they act with enough resolution. This lesson seems to have been learned by contemporary central bankers: interest rates are still low in all industrial countries despite the huge public deficits and rising inflation rates due to the Covid-19 pandemic, thus avoiding for now the explosion of public debt that occurred in the 1980s as a consequence of Volcker and other central bankers cranking up real interest rates to absurdly high levels.
Marglin points out, rightly, that before 2008 the central bank fixed the short rate while the long rate was left to the financial markets to determine. Liquidity preference, as reinterpreted, was still key, as it helped determine the long rate of interest, which is the hurdle rate for investment, as Marglin and bankers call it.[19] The entire demand for money is for transaction purposes, M1 (as banks provide loans); the speculative side, which ought to be M2, is entirely decided between short and long securities, not money. Marglin’s Figure 13.1 is quite useful and revealing. With it, we are back to what he calls the first pass at Keynes’s model, where the long-term rate of interest is fixed independently of economic activity or the size of monetary transactions, depending only on the target short-term rate of the central bank and the spreads arising from liquidity preference. This representation is nearly identical to the causal structure described by Luigi Pasinetti when assessing what he believes to be the economics of effective demand of Keynes’s General Theory, while objecting to Hicks’s IS/LM model.[20] Pasinetti describes instead a system of the decomposable type, with causal ordering, à la Herbert Simon, where the interest rate, set by liquidity preference, determines investment which then determines output, as does Marglin. But Marglin’s version is even better, because it excludes the impact of the supply of money, since the size of the flow of transactions (nominal GDP) determines the transaction demand for money which itself generates the needed supply of money. As I will describe below, however, this representation remains overly simplified: the focus is on the stock of money; there is no role for the flow of bank credit in determining aggregate demand, whereas studies at the Levy Economics Institute for instance show that this variable plays a significant role, independently of any interest rate.
What happened after 2008? Central banks tried to have some more direct control over long-term rates of interest, through swaps between short-term and long-term assets and through purchases of long-term bonds and asset-based securities – quantitative easing in the latter case. As Marglin points out, after the Lehman Brothers catastrophe, the Fed accelerated the institutional change that was supposed to be enacted much later by granting interest payments on bank reserves, thus setting up what can be called a floor system, where the target fed funds rate is the rate of interest on reserves (within a certain band, because of the peculiar institutional features of the US payment system, as noted by Marglin). In such a floor system, in contrast to the more standard corridor system or the deficient system that was in place in the USA before 2008, there is no need to adjust the supply of reserves to its forecasted demand. There is no need at all to track the quantity of reserves for the fed funds rate to be on target. However, by writing that ‘the new regime of paying interest has produced an unprecedented accumulation of excess reserves’ because ‘for banks there is little reason to prefer comparable assets like T-bills or high-grade commercial paper’, Marglin gives the illusory impression that if there is a huge amount of reserves in banks, it is because interest is paid on reserves, so that banks are induced to forego granting new loans or the acquisition of securities. Rather, as long as banks were not previously indebted to the central bank, there is nothing that the banks can do in the aggregate to reduce the amount of excess reserves arising from quantitative easing policies.[21] When pursuing open-market purchases à outrance, as Keynes would say, the Fed (and other central banks pursuing quantitative easing) had to impose an interest rate on reserves, for otherwise the fed funds rate could not have been any different from zero – a point also often made by Modern monetary theory authors.
The illusory impression mentioned just above is likely to have arisen from Marglin’s most puzzling feature of his analysis of money – his description of the banking system. The choice of the monetary system always seems to be limited to the choice between commodity money and fractional reserve banking, where banks are dependent on and limited by the supply of reserves granted by the central bank. The expression fractional-reserve banking appears no less than 56 times in the book. As repeated a number of times, ‘the crucial assumption here is that banks always attempt to deploy their assets so as to have no excess reserves, that they always try to stay fully loaned-up’.[22] In other words, the money supply is exogenous with commodity money, still exogenous with fractional-reserve banking and fully loaned-up banks, partially endogenous when banks are not fully loaned-up, and much more endogenous when banks can get around reserve requirements by getting loans off their books by creating CDOs. This money endogeneity is of the Minskyan-structuralist type, arising from financial innovations. It is unrelated to the intrinsic elasticity of monetary systems, which arises from reversing the causality between deposits and reserves and between credits and deposits, where central banks play an essentially defensive role, as argued by Nicholas Kaldor, Joan Robinson and modern post-Keynesians. With one exception,[23] they seem absent from his stories, although he must certainly be fully aware of their views on money endogeneity. What is the meaning of this fractional-reserve banking system when reserve requirements have been abolished in Canada since the mid-1990s and when reserve requirements at the Fed are nil since March 2020?
Marglin very clearly rejects the first mainstream view of banking, that which is based on loanable funds or on banks as intermediaries. But he keeps making references to the second mainstream view, the money multiplier theory of banking, otherwise called the fractional-reserve theory of banking. Unless I have misunderstood him, he neglects the post-Keynesian view, otherwise known as the credit-creation view of money or the financing-through-money-creation view of banking, which have been clearly endorsed recently by central bankers in opposition to the money multiplier view.[24] Why go with the second mainstream view when, with fully loaned-up banks, ‘a world of fractional-reserve banking is then no different from a world of commodity money because banks simply multiply an exogenously given quantity of outside money’?[25] Perhaps it has to do with Marglin being ‘committed to engaging mainstream economics on its own turf’.[26]
I finish by going back to the earlier chapters where Marglin argues that an analysis based on comparative statics (or ‘logical time’, as Joan Robinson would have it) does not necessarily achieve the same results as an analysis based on disequilibrium dynamics in real time (in ‘historical time’, as Robinson would say), which is what Keynes had in mind. Presumably, to understand the real world, the analysis should be pursued in real time. It seems to me that the stock-flow consistent (SFC) approach proposed by Wynne Godley attempts to do just that. When dealing with real time, Marglin does consider the role of bank debt and hence is briefly attuned to the credit-creation view of banking endorsed by post-Keynesians. Marglin makes the quite correct claim in his epilogue that in real time, in contrast to comparative statics, ‘a fall in prices leads to a reduction in transactions demand, but this does not lead to greater amounts of money pouring into wealth portfolios’ because ‘reductions in the demand for money to transact business are reflected in reductions in bank lending, not in an influx of money into wealth portfolios’.[27] This is an essential observation, and it is exactly what occurs in the SFC model of Lavoie and Godley.[28] Such SFC models show that the relationship between economic activity, liquidity preference and the various interest rates is not an easy one and can change between the short term and the long term.[29] We can use those models to pick up where Marglin left us. But this requires to abandon mainstream assumptions such as fractional-reserve banking, pure competition, rising marginal costs, profit maximization, and to build models where there are no black holes and where debt stocks and credit flow aggregates are explicitly determined and taken into account.
[1] Stephen A. Marglin, Raising Keynes: A Twenty-First Century General Theory (Cambridge: Harvard University Press, 2020), 522. BACK TO POST
[2] Ibid., 499. BACK TO POST
[3] Sheila Dow, ‘Endogenous money’, in A Second Edition of the General Theory, eds. .C. Harcourt and P.A. Riach (London: Routledge, 1997), 63. BACK TO POST
[4] Basil Moore, Horizontalists and Verticalists (Cambridge: Cambridge University Press, 1988). BACK TO POST
[5] Marglin, Raising Keynes, 499. BACK TO POST
[6] John Maynard Keynes, ‘Alternative theories of the rate of interest’, Economic Journal (June 1937): section III. BACK TO POST
[7] Augusto Graziani, ‘The debate on Keynes’s finance motive’, 1 Economic Notes (1984), 5-33. BACK TO POST
[8] Marglin, Raising Keynes, 519. BACK TO POST
[9] Ibid., 520. BACK TO POST
[10] Marc Lavoie, Foundations of Post-Keynesian Economic Analysis (Aldershot: Edward Elgar, 1992), 193. BACK TO POST
[11] Alfred Eichner, The Macrodynamics of Advanced Market Economies (Armonk, New York: M.E. Sharpe, 1987), 860. BACK TO POST
[12] John Smithin, ‘Real interest rates, inflation, and unemployment’, in The Unemployment Crisis: All for Nought?, eds. B.K. MacLean and L. Osberg (Montreal and Kingston: McGill-Queen’s University Press, 1996), 47. BACK TO POST
[13] Marglin, Raising Keynes, 377. BACK TO POST
[14] Ibid., 411. BACK TO POST
[15] Ibid., 514. BACK TO POST
[16] Lavoie, Foundations of Post-Keynesian Economic Analysis, 195-96. BACK TO POST
[17] Paul Wells, ‘A Post Keynesian view of liquidity preference and the demand for money’, 5(4) Journal of Post Keynesian Economics (1983), 533. BACK TO POST
[18] Tracy Mott, ‘Towards a post-Keynesian formulation of liquidity preference’, 8(2) Journal of Post Keynesian Economics (1985-86), 224. BACK TO POST
[19] Marglin, Raising Keynes, 514. BACK TO POST
[20] Luigi L. Pasinetti, ‘The economics of effective demand’, in Growth and Income Distribution: Essays in Economic Theory, (Cambridge: Cambridge University Press, 1974), 44. BACK TO POST
[21] Marglin, Raising Keynes, 515. BACK TO POST
[22] Ibid., 506. BACK TO POST
[23] Ibid., 507. BACK TO POST
[24] Marc Lavoie, ‘Advances in the Post-Keynesian analysis of money and finance’, in Frontiers of Heterodox Macroeconomics, eds. P. Arestis and M. Sawyer (London; Palgrave Macmillan, 2019), 89-130. BACK TO POST
[25] Marglin, Raising Keynes, 253. BACK TO POST
[26] Ibid., 661. BACK TO POST
[27] Ibid., 783. BACK TO POST
[28] Wynne Godley and Marc Lavoie, Monetary Economics: An Integrated Approach to Credit, Money, Income, Production and Wealth (Basingstoke: Palgrave Macmillan, 2007), ch. 9. BACK TO POST
[29] Marc Lavoie and Severin Reissl, ‘Further insights on endogenous money and the liquidity preference theory of interest’, 42(4) Journal of Post Keynesian Economics (2019), 503-526; Marc Lavoie and Gennaro Zezza, ‘A simple stock-flow consistent model with short-term and long-term debt: A comment on Claudio Sardoni’, 32(3) Review of Political Economy (2020), 459-473. BACK TO POST