Marc Lavoie is a professor in the Department of Economics at the University of Ottawa. He is the author of numerous books on post-Keynesian economics. His latest work ‘’, written with the late Wynne Godley, is now available in paperback from Amazon.com.
Interview conducted by Philip Pilkington
Philip Pilkington: Monetary Economics quite consciously departs from the neoclassical paradigm while at the same time setting itself the high task of producing concise and coherent models. So, let’s start from the beginning: why did you and Godley feel the need to depart so forcefully from the neoclassical paradigm? To tear yourselves away from it in such black and white terms as stated in the introduction to the book suggests to me that you and Godley found it lacking at a very fundamental level. Could you explain why this is?
Marc Lavoie: First I should point out that Monetary Economics started out as a project initiated by my co-author Wynne Godley in the early 1990s. Already in a paper published in 1993, and hence written much earlier, Wynne was announcing the preparation of a substantial monograph, in collaboration with a Cambridge colleague. But while Wynne managed to write working papers around his book project, he had only written a rather dry draft of the book itself when I got to meet him for the first time at the end of 1999. So it should be clear that the soul of the book comes from Wynne, while one may say that the body comes from me!
Having dealt with real world economics, through his position at the British Treasury until 1970, Wynne was convinced that mainstream macroeconomics just did not make any sense, whether it was monetarism or the standard Keynesian IS/LM model of the time; in fact he could not understand how any reasonable person, not yet brainwashed by courses in mainstream economics, could grant any credibility to the textbook and to the sophisticated models of mainstream economics. So it is clear that Wynne wished to depart from neoclassical economics, and start from scratch, which is what he did to some extent already when Wynne and his colleague Francis Cripps wrote a highly original book that was published in 1983, Macroeconomics. This book was written because Wynne got convinced that the Keynesians of all strands were losing their battle against Milton Friedman and the monetarists, because Keynesians could only provide very convoluted answers to simple questions such as: “Where does money come from? Where does it go? How do the income flows link up with the money stocks? How is new production financed?”
Our book Monetary Economics also tries to provide appropriate answers to these questions. We agonized for a while between trying to engage in a constructive dialogue with our mainstream colleagues and targeting a non-mainstream audience, or perhaps trying to achieve both goals. In the end, we figured that it would be very difficult to please both audiences, and we chose to focus on a heterodox audience. In any case, I have spent most of my academic career trying to develop alternative views and alternative models of economics – what is now called heterodox economics; this is the literature we know best. So we took our book as a formal contribution to this heterodox literature and more specifically as a contribution to post-Keynesian economics.
PP: I know it’s a rather general question, but what did you and Wynne feel were the most fundamental weaknesses of neoclassical/monetarist theory? In the book you put a lot of emphasis on dynamism and change, was this a major point of departure from the mainstream?
ML: Well, that is a rather grandiose question! Critics of neoclassical economics, and of Friedman’s monetarism, have written dozens of books and thousands of pages about this topic. To start with, I suppose that Wynne objected to basic demand-pull explanations of inflation, where the excess supply of money generates excess aggregate demand and hence inflation. Wynne thought that the money supply is endogenous, brought about by the requirements of the economy, mainly by the time taken by the production process, which involved unfinished goods and goods not yet sold, the production of which had to be financed by credit. Wynne also rejected the mainstream explanation of inflation because his work on pricing theory led him to believe that prices were based on normal unit costs, with no clear relationship between higher output and this normal unit cost, nor between higher output and the markup. So any relationship that could be found between prices and money was likely to be due to a reversed causality. Indeed, Wynne never believed in the concept of the natural rate of unemployment, the NAIRU, or the associated vertical Phillips curve; and the most sophisticated econometric analysis, meta-analysis, has proved him right.
Reversed causation also affected the link between investment and saving: for Keynesians, investment drives saving, whereas the neoclassical view is that saving allows investment. The importance of getting this causality right is clear nowadays: a true neoclassical author would argue that households need to save more so as to provide firms with the funds they need to invest; but if households reduce their consumption expenditures, with firms selling less, why would they want to invest more? Neoclassical macroeconomics is essentially supply-led; this to us is its fundamental weakness: capitalist economies, most of the time, are demand-led. They generally suffer from a lack of effective demand, not from a lack of capacity or a lack of labour resources. Thus we mostly focus on the variables that determine aggregate demand. I used to believe that only advanced economies had spare capacity; but then I read a 1983 book of Lance Taylor, showing the peculiar features of less-developed economies and how they should be modelled: he also assumed spare capacity!
It is also true as you point out that in our book we emphasize the transition from the starting position towards the new equilibrium – that is we go beyond comparative statics, beyond comparing the initial state to the final new state of the model. In part, this is because the short-run results of a change may turn out to be opposite of the long-run impact. For instance, in one of our more sophisticated models, the desire of households to borrow a greater percentage of their personal income has a favourable impact on consumption and GDP in the short-run, but it has a negative impact on these two variables in the long-run. This we do with simulations; but of course, with fancy enough mathematics, you can also show this with the help of algebra, provided the model is not overly complicated.
PP: That’s an impressively succinct answer to a very general question.
I guess since we’re going to have to broach the following question at some stage, now is as good a time as any to ask it since you brought it up. You just said that you and Godley do, in fact, use the notion of equilibrium in your book. But my understanding is that you use it in a rather different way from how it is used in neoclassical theory. My understanding is that your version of equilibrium is not to be located in the Walrasian tradition. Perhaps you could explain how you and Godley use the notion of equilibrium in the book and how this contrasts with the neoclassical/Walrasian usage?
ML: In the neoclassical tradition, the equilibrating mechanism occurs through relative prices, through the law of supply and demand. Prices are said to reflect scarcity. But since the Cambridge capital controversies, we know that this need not be the case, and we know since the mid-1970s and the discovery of the Sonnenschein-Mantel- Debreu conditions that the most sophisticated Walrasian general equilibrium models cannot demonstrate the achievement of equilibrium through price changes, meaning that price changes cannot in general bring back demand in line with supply. Indeed, as everyone has realized with the recent stock market and real estate bubbles, rising prices in financial markets, where one would think that the law of supply and demand would apply, may generate larger demanded quantities – not lower ones.
So this line of thought must be abandoned. In the Cambridge tradition, prices reflect unit costs, or more precisely normal unit costs, that is unit costs estimated under normal operating conditions (at normal rates of capacity utilization or at target inventories to sales ratio). How large is the profit margin – the difference between the price and the normal unit cost – is not so easy to ascertain, but several post-Keynesians argue that it is a function of the growth rate of the industry sales. Thus prices – except perhaps in financial markets and some commodity markets – do not have as a purpose to equate supply to demand. Prices are instead the means to obtain profits sufficient to finance expansion. Variations in output demand are absorbed by fluctuations in quantities or in rates of utilization of capacities, as was argued by Keynes and the Keynesians.
The other adjusting mechanism underlined by Wynne Godley is fluctuations in inventories. Stocks of inventories act as buffers that absorb fluctuations in demand or in production. When demand falls, the slack is taken by a rise in inventories, not by a fall in prices. One of the key notions that we try to emphasize in our book is that all sectors incorporate a buffer. In the case of firms, inventories are the obvious buffer. On the financial side, the counterparty to the (undesired) fluctuations in buffer are the fluctuations in the credit lines drawn upon by firms. It is the existence of these buffers that makes the economy relatively stable.
For individuals like us, the money balances that we hold in our bank deposit accounts are our main buffer. The fluctuations in our income or in our consumption expenditures are absorbed by the variations in our bank account. In the case of banks, their main buffer is the stock of safe and liquid assets that they hold, and which can be acquired or sold. At the level of individual banks, liability management would act as a buffer. In the case of the government, their buffer is the outstanding amount of securities which they have issued. In the case of the central bank, their buffer is the amount of government debt that they hold (in North America) or the amount of advances that they make to the banking system (in continental Europe). Within the Eurosystem, the buffer is the Target2 balances credited or debited at the European Central Bank, which everyone is now talking so much about. Equilibrium, in a much weaker sense than the neoclassicals would have it, is achieved thanks to these buffers. Economic agents then react more or less strongly to variations in these buffers or to how much these buffers are out of line with possible targets. An equilibrium is reached when the target buffer is achieved.
PP: I guess what I’m getting at is a little more fundamental here. My feeling – and I think this was shown in a very famous paper by Imre Lakatos’ student Spiro Latsis – is that neoclassical economics doesn’t allow proper space for agency. Everybody does what they’re supposed to do within the model. You know, you’re supposed to consume in line with your marginal utility and you’re supposed to invest in line with marginal productivity. But your’s and Godley’s models are open, right? I guess this is what I’m interested in. It’s the fact that you’ve constructed coherent models that nevertheless remain open and don’t rely on some sort of, well, ‘metaphysics’, as Robinson would have called it… Perhaps you could talk about this a bit?
ML: There is indeed a fashionable distinction which is being made today by methodologists in economics, between open and closed systems, with the argument that open-system models allow proper space for agency, whereas closed-system models don’t. I am not comfortable with this distinction. If I were a neoclassical economist, I would argue that the results being achieved in a model based on marginal utility depend on the elements that are being included in the utility function; and some people might argue that our stock-flow consistent models are, by construction, closed models. I am more at ease with the argument that our stock-flow consistent models are system-wide models, that can be adjusted to take into account different institutional frameworks, and that are based on more realistic assumptions, with variables that have real counterparts.
In a sense I agree with what you say: investment does not depend on the marginal productivity of capital; it depends on sales, rates of capacity utilization, profitability, the appearance of new products, and other such elements, including the optimism of entrepreneurs.
PP: Actually, that brings me to a quote from Geoff Harcourt that you include in the introduction of the book. You quote Harcourt in saying that older Post-Keynesians, such as Joan Robinson, had tried to look at the economy by segmenting it and examining each of the sections separately. Harcourt goes on to say that it was this approach that Robinson and others had thought a failure. You just said that the models put forward in your book are ‘system-wide’. Is this then the solution to problems raised by the older Post-Keynesian approaches? If so could you perhaps sketch out the difference in the two approaches?
ML: Yes, I would say that this system-wide approach may be the most important contribution of our book to the post-Keynesian literature, and possibly to economics in general. I have argued on a few occasions that the Holy Grail of macroeconomics is to integrate the analysis of the real side of the economy (production, income, employment, investment, and so on) with the monetary side (credit, debt, financial assets, the financing of economic activity, the behaviour of banks and other financial institutions).
When I was a graduate student in Paris, all my teachers were talking about the integration of the monetary sphere into the real economy; however no one knew how to do it. Rediscovering the work of Wynne Godley in 1996, in particular his 1996 Levy Institute working paper where this integration was achieved was a true revelation for me, and I remember telling my long-time colleague Mario Seccareccia at the time that this was what we ought to be doing. In the more sophisticated chapters Wynne and I analyze the flow of production and income, we tie this to the flow of credit and money creation arising from the banking sector, and we link this to the stocks of capital, the stock of financial wealth and the stocks of debt of the households, firms and government.
By contrast, Joan Robinson, in her 1956 book The Accumulation of Capital, which I consider to be one of the most formidable books ever written, focuses on production and technology in the first two thirds of the book, and only discusses money and credit in the last chapters (which hardly anybody ever reads, having been exhausted after reading her intricate analysis of technological choice in the earlier chapters). The two are not integrated. Most of the post-Keynesians did the same, dealing separately with production and income distribution in some papers, and dealing with endogenous money and interest rates in others. Indeed, Wynne often told me how much he regretted that Nicholas Kaldor never took the time or find the time to put together, in a single framework, all of his theories.
Now I am not claiming that one should always pursue this system-wide analysis, or that everyone should be doing stock-flow consistent analysis. But certainly some of us ought to be doing it so as to get the big picture. In any case, I should point out that other authors in the past have in one way or another achieved this system-wide approach to post-Keynesian economics, in particular Peter Skott (1989) and Alfred Eichner (1989) in their books. More recently, Peter Flaschel and his several co-authors have written several books entertaining this system-wide approach, although they sometimes overly rely on neoclassical assumptions to my taste, and Peter Skott is revisiting his earlier work and doing again system-wide analysis.
PP: Actually, you mention the endogenous money approach. Let’s focus on that for a moment. As you probably know there’s been some controversies in the blogs over this theory. I’ve found that an awful lot of the debate between neoclassicals and Post-Keynesians – at least when it comes to monetary theory – is usually based on whether or not an economist subscribes to the endogenous theory of money. I won’t ask you to go into the details of the theory as it has been articulated enough times online already, but why do you think the neoclassicals are so hostile toward it? And why is it so important to your theoretical framework?
ML: Here I can say not better than what Alain Parguez, a replacement professor in Paris when I first arrived there as a graduate student, has been claiming all along: neoclassical economics is based on the concept of scarcity. Without the concept of scarcity, there is no neoclassical economics. By making the stock of money exogenous, the concept of scarcity is recovered, even when discussing money creation. This makes theorizing possible for mainstream economists. Thus there are two kinds of neoclassical economists.
First, there are those who still argue along the textbook view, that there is a given stock of money out there, tightly linked to the amount of reserves determined by a responsible central bank; those are what Basil Moore called the ‘verticalists’, and they are clearly out of touch with reality. Second, there are neoclassical authors who recognize that the money creation process in modern financial systems makes the stock of money endogenous, but who would like the central bank to behave in such a way that the stock of money would be exogenous again. A variant of the second view is to argue that there is a unique natural rate of interest, determined by the supply-side forces of productivity and thrift, and that the central bank ought to set its target interest rate in such a way that the actual rate of interest is equal to the natural rate. There is no alternative for a responsible central bank but to assess the natural rate and act upon it.
By contrast, post-Keynesians argue either that this natural rate of interest does not exist, or that there is a multiplicity of them. In your words, one could say that the post-Keynesian view is that the central bank has room for agency. I would argue once more that aggregate demand matters and will have an impact on potential output, so that the scarcity principle does not apply, even in the long run, and hence the natural rate of interest is not a useful construct, no more useful than the concept of the natural rate of unemployment. Right now, in the midst of the crisis, what is the natural rate?!
PP: Interesting. The idea that the central bank should set its rate of interest in line with the natural rate always struck me as being somewhere between divination and mysticism. Anyway, let’s continue with money for a moment. In the book you and Godley write that you want to “restore to money its natural attributes”. Could you explain a bit what you mean by this and why it is important?
ML: Well, as just pointed out, the textbook version of neoclassical economics and even sophisticated mainstream models assume that money is fixed and exogenous. Our view of money is that money is a buffer.
As said before, variations in the income and expenditures of households will be absorbed by fluctuations in the balances in our bank account; one could argue that these fluctuations of household money balances will be exactly compensated by equivalent inverse fluctuations in the bank accounts of firms; but it is also possible that firms that are indebted towards banks will use the proceeds of additional or unexpected sales to reduce their debit balances at banks, and hence these fluctuations will indeed lead to fluctuations in the stock of money. Thus what we claim is that the central bank cannot control the stock of money, whereas it can act to control a range of interest rates.
Furthermore, because we emphasize accounting consistency, we underline the fact that money as an asset must have a counterpart liability, something which is hard to maintain if you assume that the stock of money is an exogenous variable.