John Smithin 1 York University

Transcription

1 Introduction (Draft of December 2012) Keynes s Theories of Money and Banking in the Treatise and the General Theory John Smithin 1 York University Some years ago Rogers and Rymes (2000) described in detail the disappearance of Keynes s nascent theory of banking (emphasis added) between the publication of the Treatise on Money in 1930 and the General Theory in This paper will also discuss the implications of the changes in approach between these two important books by Keynes, but covers a wider range of topics in monetary theory (in addition to banking). From the outset it should be kept in mind that the Treatise and General Theory were the last two books of a trilogy on monetary theory, covering the entire period of crisis in the global monetary system between the two world wars of the twentieth century, and comprising A Tract on Monetary Reform (1923), A Treatise on Money (1971/1930), and The General Theory of Employment Interest and Money (1964/1936). Therefore, although this paper focuses on the two best-known of Keynes s books, it must be stated that if we are eventually to reconstruct a complete monetary theory along Keynesian lines, at least some of the ideas from the Tract would have to be included (Smithin 1996, 2003). In particular, attention should be given the discussion of the distribution of income between the Earner, the Investing Class and the Business Class in chapter 1 of the Tract, as well as to the discussion about interest parity conditions and the forward exchanges in chapter 3. The latter would provide the essential clue about how Keynesian policy could be implemented in a single jurisdiction, if necessary. 2 In what follows, however, the discussion will deal mainly with the differences between the Treatise and General Theory. What are the Key Issues in Monetary Theory? In both editions of Controversies in Monetary Economics (Smithin 1994, 2003), and later in Essays in the Fundamental Theory of Monetary Economics (Smithin 2013a), I pointed to a number of standard themes/debating points, that have been in contention throughout the history of discussion about money. There seem to have been five main recurring themes in all, which together have some claim to be considered the key issues in monetary theory. Each of these is briefly discussed below: 1

2 Ontological Confusion about Money - Its Nature and Functions. The first question to be asked, presumably, is the basic one of how the social constructs of money and credit came into existence in the first place, how they are reproduced and maintained. A prospective monetary theorist must consciously or unconsciously take up a definite position on the ontology of money. This is bound up with the entire logical/historical question of how the capitalist institutions, particularly the basic concept of production for sale in the market, specifically for monetary reward, came to exert such a dominating influence in our social life. The answer to this question tends to be taken for granted in the orthodox economic literature, which invariably espouses one version or another of the view of Menger (1892) that the market itself is the primary concept, and that money logically emerges an initial state of barter. There is, however, an alternative point of view, which has a much stronger claim to be a realist social ontology (Mendoza 2012). This is that the idea or concept of money itself is the starting point. Rather than money emerging from the market, the market emerges from money. Money is a social relation (Ingham 2000, 2004, 2005/1996) involving a money of account, means of payment, credit facilities and so on, and is actually a pre-condition for the existence of such things as rational accounting and pricing. It is therefore also a pre-requisite for market exchange and monetary production. The seemingly purely analytical questions in dispute, discussed below, are closely tied to the positions taken in this logically prior debate and the fundamental ontological issues deserve far more attention than is usually given them in textbook treatments. How Does Money Get into the Economy? How is it Created and Destroyed? Bougrine and Seccareccia (2010) have raised one the most important of the subsequent analytical questions when they asked directly how money gets into the economic process. As they put it, "how is it [money] created and destroyed?. This is a reference to the ongoing debate over whether money should be treated as an exogenous or endogenous variable, terms originally taken from mathematics or engineering. An exogenous variable is something determined outside the system being studied, whereas an endogenous variable is determined within the system. In the case of the financial system, we can ask, specifically, whether the quantity of money should be treated as fixed amount, capable of precise determination by some such agency such as a central bank. Or, rather should it be treated as a wholly endogenous variable which expands residually whenever the volume of credit granted by financial institutions increases (and contracts when the volume of credit contracts). In a modern credit economy, after all, the statistically-defined money supply consists mainly of the deposit liabilities of financial institutions such as banks, to 2

3 the point where the expression total bank deposits becomes more-or-less synonymous with the quantity of money. The debate about exogenous or endogenous money has therefore been going on as long as banking institutions have been in existence. The monetarists of the mid-twentieth century took the former point of view, arguing that (even in a developed credit economy) the central bank can control the money supply quantitatively via the money multiplier which supposedly links the monetary base to the statistically-defined monetary aggregates. More recently there has been a considerable revival of the endogenous money approach, particularly among heterodox economists such as the Post-Keynesian horizontalists (Kaldor 1986, Moore 1988, Lavoie 1992) members of the circuit school (Rochon 1989, Parguez and Seccareccia 2000, Graziani 2003), and neo-chartalists or the adherents of MMT (Mosler 2011, Wray 2012). 3 On this view loans make deposits not deposits make loans (Smithin 2013b). Therefore monetary control, to the extent that it can be exercised at all, can only be indirect, for example, by interest rate changes. What Determines the Real Rate of Interest on Loans of Money? A third equally important issue, and another ubiquitous analytical theme, is how the real rate of interest on loans of money is determined. By far the most popular assumption among professional economists at all times, including the present, has been that the rate of interest is determined fundamentally by real forces outside the monetary system, such as the demand for and supply of physical capital. The underlying doctrine is of the existence of a natural rate of interest. Something like this has appeared, and re-appeared, in one guise or another, throughout the whole history of discussion of monetary topics. Once such an assumption is made, however, it immediately forecloses any debate over whether monetary changes can have any lasting significance for the real economy. By definition they will not. If money does affect inflation, it might be accepted by this majority school that monetary changes can affect the nominal interest rate. However, by assumption monetary changes are not simply not allowed to affect the real (inflation-adjusted) interest rate, and there is no question of any long-lasting impact of monetary changes on any of the most important real economic variables. The alternative point of view, in this case, is that the real rate of interest itself is essentially a monetary phenomenon, and that the economy must adjust to the interest rate established in the monetary/financial system, not the reverse. The outcome of the debate over monetary policy is no longer a foregone conclusion. Money and monetary policy will be fundamentally non-neutral and a genuine monetary analysis (Schumpeter 1994/1954) will be possible. 3

4 There have been basically two versions of the argument. The first of these was Keynes s notion of liquidity preference, properly regarded precisely as an alternative theory of interest rate determination, rather than simply an alternative theory of money demand. The second monetary theory of the real rate of interest Burstein (1995) is the view that the rate of interest is determined administratively/politically, by the policy decisions of the central bank. An important issue that has arisen is how to reconcile these two positions. The answer must be along the lines that the central bank directly sets the interest rate under its own control, the policy rate, and that interest rates elsewhere in the system are affected by both the policy rate and also by liquidity preference-type considerations. What is the Main Monetary Policy Instrument that can be Employed by the Central Bank to Determine the Pace of Credit and Money Creation? The above discussion of real interest rate determination is relevant to, and necessarily touched upon, another of the main continuing themes, the long running debate about the appropriate monetary control variable or monetary policy instrument (Taylor 1993; Walsh 1998, 2009; Lavoie and Seccareccia 2004; Smithin 2007; Lavoie 2010). This is not quite the same issue as what determines real interest rates and needs separate discussion. This topic is therefore another of our continuing themes. The question, now, is whether control of the money supply must necessarily be exercised indirectly (as certainly implied in the argument of the previous section, for example via the manipulation of nominal interest rates). Or, can the nominal quantity of money, on the contrary, be directly controlled by the central bank. As already mentioned, the monetarist school favoured direct control of the rate of growth of some statistically defined measure of either the money supply itself, or of the monetary base. In economic textbooks, moreover, the assumption is still frequently made that there is some unique quantity of dollars, M, the money supply, which is the exogenous monetary policy variable. As a practical matter, however, before and after the period of monetarist ascendancy, the actual monetary control instrument typically employed by central banks was usually a short-term nominal interest rate of some kind (B. Friedman 2000; Goodhart 2002; Smithin 2009). This, in turn, raises the awkward questions of trying to reconcile this fact both with the orthodox literature and various preconceptions of how interest rates would theoretically be determined in a barter economy. Some of these questions were prefigured by Keynes (1971/1930) in his discussion of the three strands of thought, on bank rate policy in the Treatise. As pointed out above, if we have already decided that the true interest rate is determined otherwise than by central bank policy or in the monetary/financial sector, it then becomes absolutely necessary to differentiate between the 4

5 interest rate that is set by the central bank and the hypothesized natural rate. Ultimately, it becomes necessary to make the outcome of economic policy decisions hinge on the discrepancy. This is the route taken in both Wicksellian and modern neo-wicksellian models. Otherwise, Keynes s idea of interest rate policy becoming an independent influence becomes a real possibility. How Do Monetary Changes Affect Other Economic Variables? The last of the recurring monetary themes must evidently be the clash of views about the impact (if any, according to some) of changes in the monetary variables on other important economic variables, particularly on the so-called real variables. How do such changes affect output and employment, and their rates of growth, inflation, real wages, profits, the exchange rate, the balance of payments, and so on? The views taken on these issues will again partly depend on each of the positions adopted so far, exogenous or endogenous money, whether there is a natural rate of interest, and base control versus interest rate control. The main dividing line is on whether or not monetary factors can permanently affect real variables, as opposed to some short-run or temporary impact. The latter is often (but not always) conceded. Clearly if the initial presumption is that, by definition, monetary factors cannot permanently affect any of the natural rate of interest, the natural rate of growth, or the natural rate of unemployment, then the most basic question has been decided without any need for further analysis. Remarkably, as indicated something of the kind has almost always been the orthodox view. But then, a major difficulty arises because of the need to explain why au contraire monetary issues and problems have so often been at the forefront of real-world political debate. If eventual monetary neutrality is taken for granted, the only remaining avenue for a coherent explanation of actual monetary problems would be the highly artificial Marshallian, and therefore dubious, distinction between the short-run and the long-run. The calendar length of this short-run must become flexible enough to encompass whatever economic events need to be explained in any particular historical episode. A Preferred Theory of Money? The above discussion of the different attitudes that might be taken to each of the five main debating points leads on naturally to the question of what a preferred theory of monetary economics might look like. I now suggest what seem to me to be the desirable features of such a theory (Smithin 2003, 2009, 2013a). Firstly, the preferred theory must set out an explicit ontology of money. It must explain that money is a 5

6 fundamentally a social institution, or social relation (Ingham 2005/1996), not a commodity. Then, why, nonetheless, this social relation can have substantial causal effects in the material world. It must also explain how credit creation and its residue in money creation are absolutely needed in order to generate any sort of profit surplus (Smithin 2011, 2012) and why there is a hierarchy of money (Bell 2005/2001). It should be made clear, above all, that the establishment of this set of institutions is a necessary condition for the system of capitalism (Weber s method of enterprise and Keynes s entrepreneur economy ) to exist in the first place. This entails clarifying the role of the constitutional state (as opposed to either a confiscatory or a totalitarian state) in sustaining the types of monetary and financial institutions necessary for the system of enterprise to function (Smithin 2013b). Second, the preferred theory should include each of the concepts of endogenous money, credit creation, and the monetary circuit, treated as integral parts of the economic process, and should fully explain their roles. The supply of money is not fixed. An important insight deriving from this basic point is that the theory of inflation, on the one hand, and theories of economic growth and employment, on the other, are in principle quite separate things. They are linked only because of the dual role of banking and credit. Once this point is grasped it should then be possible to explain all of the different combinations of inflation and growth that have occurred in reality and all the possible sources of inflation (Smithin 2012b). The resulting theory should therefore be less vulnerable to the empirical problems (Smithin 1996) that have frequently arisen in economic orthodoxy whenever something occurs in reality at odds with what the textbooks say. There have been serious problems with the textbook narrative in each of the last four decades, when confronted with "stagflation" in 1970s, high unemployment in the 1980s, the "tech boom" of the 1990s, or the financial crisis in the first decade of the 21st century. The preferred theory would not simply be "inflationist" for the sake of it. It may be readily admitted that it would be a good thing, from the point of view of incentives under capitalism, if the value of money was more stable. However, what does this mean? Does it really mean that there should literally be zero inflation, no price changes at all on average? This seems quite incompatible with the basic spirit of capitalism. On the other hand, if the nominal value of money holdings at least keeps pace with inflation (which means, in the modern world, that the real interest rate received on bank deposits is non-negative), this would still mean that the "value of money" would be stable, and as stable as it is ever likely to be (Smithin 2013a). On the question of interest rate determination there must be, as mentioned a monetary theory of the real rate of interest (Burstein 1995, emphasis added). If fundamentally, money is a social relation, then real rate of 6

7 interest on money must also be a social relation (involving money.) This in itself is sufficient to explain why, at the end of the day, the central bank usually has a good deal of power over interest rate determination but also why liquidity preference considerations similarly play a key role. The real rate of interest on loans of money loans is not simply a technical economic variable, but very much a part of the social fabric itself and this inevitably brings up the essentially political question of what the real rate of interest, on loans of money, should be. Note also that the counterpart in international economics to a monetary theory of the real rate of interest must be a "monetary theory of the real exchange rate". The real exchange rate should also be thought of as an endogenous variable, rather than being assumed to be fixed by non-monetary factors such as the hypothetical barter terms of trade. It should clearly be recognized that, in normal circumstances, the monetary policy instrument is indeed usually a nominal interest rate, the policy rate of the central bank. This brings us back to the question of how the real interest rate changes discussed above can actually be achieved. In fact, simply adjusting the nominal policy rate for changes in reported inflation will stabilize what Smithin (2013a) has called the inflation-adjusted real policy rate, and if adherence to such a rule can be achieved in practice, this will also tend to stabilize real interest rates in general. It should be noted that if the nominal interest instrument is to be employed successfully to change real rates there may actually have to be at least some inflation in the system. This seriously calls into question the idea that a policy aiming at no price changes whatsoever is desirable. When, for example, there is either zero inflation or actual deflation, and also the nominal rate has fallen to zero, it is not possible to reduce real interest rates (if required) just by changing the nominal policy rate. At this point, so-called unorthodox monetary policy (e.g., quantitative easing, or fiscally-based monetary policy) may be required, to attempt to induce inflation from other sources. Emergency situations like this seem to dramatically illustrate the argument that the most desirable monetary policy would be a real interest rate rule, stabilizing the inflation-adjusted policy rate (Smithin 1994, 2003, 2007, 2009, 2013a), not a policy of targeting the inflation rate or the price level per se. A possible alternative interest rate policy, that of pegging the nominal rate interest rate at whatever level (including zero), would unfortunately only lead to instability, which could be of either an inflationary or deflationary character (Smithin 2013a). Finally, it should be recognized that both monetary and fiscal policy changes are inevitably non-neutral, in both the short-run and the long-run. They affect not only the ups and down of the business cycle, but also the average growth rate of the economy itself. According to the preferred theory, it would therefore be a good thing if Lerner s notion of functional finance, could be revived (Nell and Forstater 2003). Changes in government 7

8 spending and taxation would be judged by the impact that they actually have on the economy, and not on arbitrary financial rules such as the size of the government budget deficit, or national debt, relative to GDP; nor of any perceived need to balance the budget. In spite of appearances to the contrary, note that this sort of statement does not really impinge on the ideological debate about the scope or size of government, in relation to the overall economy. It does not necessarily support a tax and spend agenda, for example. The real issue, which was well understood by Keynes, is rather the extent to which, when there is government spending, this is financed as loan expenditure (Keynes 1964/1936) instead of increasing taxes to pay for it. Keynes Our main subject, John Maynard Keynes who was writing before the era of monetarism, between the two World Wars of the twentieth century, may well have been the most famous monetary economist of all time. He claimed to have decisively refuted the classical economic theory, only for the monetarists, and others, later to claim to have decisively refuted Keynes himself. What position did Keynes take on each of our five main themes or debating points in monetary economics? The answer to this question turns out to be something of a mixed bag. On the question of the ontology of money there are, in fact, three chapters at the beginning of the first volume of the Treatise on Money (Keynes 1971/1930) in which Keynes does present a detailed view of the nature and functions of money. Moreover, the overall approach in these chapters seems to run very much along creditist and chartalist lines. There is also the additional evidence of the missing draft chapter (Keynes 1973) of The General Theory (Keynes 1964/1936), in which Keynes does make clear the difference between a monetary or "entrepreneur" economy and a barter exchange economy, and that the forthcoming General Theory is intended to apply only to the former (Dillard 1988; Smithin 2009, 2013a). Unfortunately, however, this crucial chapter was not included in the final published version of the book. On the contrary, for the General Theory, a decision was made to let "technical monetary detail [fall] into the background" (Keynes 1964/1936). This was an error in retrospect. Had the draft chapter been kept in, or even some reference made back to the extended discussion of "representative money" (Keynes 1971/1930) in the Treatise, it might well have been easier for the later audience to understand the argument that was being made in the second book. Another difference in presentation between the Treatise and the General Theory is that in the former, as already suggested Rogers and Rymes (2000) locution of a nascent theory of banking, there is already a fairly 8

9 extensive discussion of banking, central banking, credit creation, money creation, and so on, in the Treatise. Hence, implicitly there is a discussion of endogenous money. This is also reflected in the various explicit references to Wicksell in that work (Hicks 2005/1967). By the time of The General Theory, however, the money supply M seems to have been treated by Keynes mainly as an exogenous variable. It becomes a "given" at any point in time, and the quantity of money therefore appears to be something that can only be changed by a deliberate act of policy. Within the Post Keynesian school there has been some dispute about whether Keynes actually meant to argue this (Rochon 2012). 4 In practice, however, whatever Keynes intention may have been, the entire discussion of the money supply thereafter becomes almost impossible to distinguish from that later put forward in monetarism. The same point explains why the Keynesian theory of the demand for money (in this case, quite famously) was also essentially similar to the succeeding monetarist approach (Leeson 2003a, 2003b, Smithin 2004). Throughout the crucial first phase of the "monetarist versus Keynesian" debates of the mid-twentieth century most of the discussion was therefore entirely misdirected to mere empirical nuances such as the interest elasticity of the demand for money. It was not at all focused on more fundamental matters of monetary theory. The competing frameworks that were advertised as either monetarist or Keynesian were simply not as different from each other as might have been supposed from the (apparent) ferocity of the mainstream academic debate. From the later Post Keynesian point of view the apparent assumption of an exogenous supply of money in the General Theory also caused difficulties for attempts to put forward alternative explanations of inflation. Logically, when the money supply is fixed, is it impossible to bring in wage-push or cost-push theories of inflation, for example. On the other hand, in spite of ending up with fairly standard notions of money supply and demand in The General Theory, Keynes clearly did intend to break with orthodoxy quite radically on the question of interest rate determination. It is true that a weakness of the discussion in both the Treatise and the GT is that Keynes failed to make clear the distinction between real and nominal interest rates, in the way familiar to modern economists. This was, in some ways, a strange omission because there was certainly a clear understanding of the issue in the first book of the trilogy, The Tract on Monetary Reform. The reason for the failure to emphasize the difference was probably the change in economic circumstances between the early 1920s and early 1930s. In particular, it should be remembered that by the early 1930s nominal interest rates had already fallen to very low levels (a somewhat similar set of circumstances to the present day). Nonetheless, with three-quarters of a century of hindsight I would say that there is little doubt that Keynes was indeed trying to provide (what was labeled above) a monetary theory of the 9

10 real rate of interest via the liquidity preference theory of interest rates. The concept of liquidity preference was not just a question of providing a new name for the theory of the demand for money, as the famous article by Tobin seemed to imply (Tobin 1958), nor of simply inserting an interest rate argument into the demand for money function. According to Keynes's own statements it seems to have been a definite attempt to provide an alternative theory of interest determination in general, based ultimately on speculation in the financial markets. Confusion reigns, however, because in the form that Keynes presented the argument in the General Theory the liquidity preference theory unfortunately does not succeed in providing the desired alternative theory of interest rates. Again the failure seems to have come about as result of the implicit assumption of a fixed nominal supply of money. This makes the theory vulnerable to the argument that with a sufficiently large deflation (a sufficiently large fall in prices) the real supply of money M/P can always increase enough to satisfy any conceivable degree of liquidity preference. This describes the so-called "Pigou effect" or "real balance effect" originally discussed by such economists as Pigou (1943) and Patinkin (1948). As the dates of these publications indicate the idea was very quickly picked up by orthodox economists, thereby enabling them to simply dismiss the Keynesian interest rate theory on supposedly sound theoretical grounds. To point out the vulnerability of the argument of Keynes s General Theory on this score, however, is not the same thing as to argue that the overall concept of liquidity preference itself is unimportant. This has been certainly been conclusion drawn by very many well-known economists such as Leijonhufvud (1981a), for example. However, in the light of the endogenous versus exogenous money debate, this hardly seems to be the correct response to difficulties caused by a technical error in presentation. On the contrary, liquidity preference properly reinterpreted, does indeed turn out to have a vital role to play in explaining many features of the business cycle, and of financial crises (Davidson 2009, Smithin 2013a). It must by all means be kept in the model. However, in the specific historical situation of the debates of the 1930s and 1940s it was not the vital clue, that Keynes thought it would be, to an alternative non-materialistic fundamental explanation of interest. This last statement immediately raises the question of where such an explanation is actually to be found, and the answer must be at the level of social ontology as already discussed. Money is a social relation, social institution, or social fact and, if this insight can be used to explain why (the concept of) money matters, it must also explain why it is that the real interest rate on money matters, primarily as socio-economic or socio-political concept (Smithin 2009, 2013a). As for Keynes s views on the appropriate monetary policy instrument, in the Treatise the policy rate of the 10

11 central bank was correctly identified as the lynch-pin of the system. This was called "bank rate" in the terminology of the early 20th century (Keynes 1971/1930). However, by the time of The General Theory, published a few years later than the Treatise, in keeping with the absence of technical monetary detail in the later book, there is much less discussion about the modus operandi of bank rate, as Keynes had called it in In the later work, as explained, it now seems that changes in the money supply itself (changes in M) represent changes in monetary policy. This is another example of how this particular shift in focus between Keynes's two major books, tends to weaken rather than strengthen the overall argument (Rogers and Rymes 2000). On the fundamental question of monetary neutrality versus non-neutrality there also seems little doubt that in The General Theory, and as in the case of the theory of interest rates, Keynes did at least have the intention to demonstrate that both monetary and fiscal policy (changes in government spending and taxation) were non-neutral. Moreover to show that they were non-neutral in both the short run and the long run. It would be difficult, however, to convincingly argue that Keynes ever really succeeded in persuading the rest of the economics profession that he had achieved this goal (Salant 1985; Meltzer 1988; Smithin 2004). The most popular interpretation of Keynes's massive intellectual effort has been that he had only provided the same sort of arguments as (some of) the classical economists had already done centuries before, and that the later monetarists would go on to do again (Laidler 1999). Although Keynes claimed to have shown that an economy could be permanently depressed, the majority of people in the economics profession, at the time and ever since, have thought that he simply re-cycled the age-old argument that temporary wage rigidities, or the like, can cause temporary unemployment. One of the reasons for this failure of communication, I think, is that, seemingly for the first time in the General Theory, and even though the topic was macroeconomics and money, Keynes nonetheless wanted to couch his arguments very much in terms of standard Marshallian microeconomics. This therefore meant using marginalist mathematics, at least to the limit of his own personal technical range. Even more importantly, in fact, crucially, it also required the assumption that conditions of perfect competition prevail in all markets. These dubious features were presumably included to prove his bona fides to the rest of the economics profession in this important work. Why, though, was it felt necessary to do this? I suppose that everyone who has spent any time at all studying economics in graduate school, or has taught in a university, will be familiar with the sort of peer pressure that might lead someone go down this route (Harcourt 2012, King 2012). Nonetheless, it is profoundly discouraging to have to think that even so great a figure as Keynes apparently felt he had to do this in order to get his message across. In any 11

12 event it was bound to lead to mistakes. As already hinted probably the more important of the two points in the present context has to do with the assumption of perfect competition, rather than the issue of marginalism. This is because, when one thinks about it, among the conditions defining a state of perfect competition in the microeconomic textbooks is the idea that each firm in each industry is so small, in relation to the market, that it simply believes that it can sell all the output that it wants to at the "going price." However, if firms do believe this, and act on it, then it should already be quite clear that there is no theory that could ever be devised that will make changes in the aggregate demand for goods and services permanently alter a firm s behavior. The only escape from such a conclusion would again by such devices as nominal wage or price rigidities that serve to "throw a spanner in the works" (Leijonhufvud 1968, 1981a) and temporarily prevent some actors from carrying out their plans (but only temporarily, and this is the whole point). Making the assumption of perfect competition clearly says nothing at all about what markets conditions are actually like in the real world. The idea of perfect competition in all markets is self-evidently a theoretical abstraction, although this seems not always to have been apparent to many of the more enthusiastic theorists themselves. It is, however precisely this abstraction that makes it impossible to give any coherent theoretical account of how aggregate demand changes could ever have anything more than a transitory effect. Compared to this basic theoretical point the use of marginalism, as such, might actually be regarded as something of a side issue, were it not such a sticking point for the "experts" (Kaldor 1983, 1985). In reality, marginalism is neither a necessary nor a sufficient condition to arrive at a macroeconomic theory that ignores the demand side (Smithin 2004a, 2009a). It is important mainly to budding mathematical economists in order, again, to prove their bona fides, to demonstrate to their colleagues that they can "do the math" (i.e., solve optimization problems). It is true that for historical reasons so-called marginalist methods did come to be closely associated with the very first subjectivist theories of value that emerged in the late 19th century. Nonetheless, it would be logically quite possible to subscribe to a subjectivist theory of value with or without marginalism. It can also be shown with or without marginalism that under imperfect competition (rather than perfect competition) real monetary demand does matter (Smithin 2007b 2009, 2013a). If both these statements are true, however, then apart from the motivation to fit in and to have a conventionally successful academic career, the analyst may as well simply abandon marginalism for most practical, rather than rhetorical, purposes. It is not, after all, anything like a realistic description of the way in which decisions are actually made. Moreover, as just shown, it does not really bring 12

13 anything else useful to the table except whatever intellectual pleasure is to be gained from working out problems in differential calculus. There is one caveat to this position, I suppose, in the case of some types of firm behavior particularly financial and banking firms. This is because, given the economic sociology of the system, it may be the case that such types of firms are constrained to behave in a certain way (that is, to maximize profits) if they are to succeed on their own terms of reference in the existing environment. 5 No comparable case can be made, however, that it makes any sense to describe the behavior of households, individual consumers, or workers, in this way. Could a Complete Monetary Theory Eventually be Re-Constructed from the Materials Originally Provided by Keynes? The answer to this question is almost certainly yes, but it must be stressed that Keynes personally did not do this, and that the theory regarded as his legacy has therefore been vulnerable in practice, to criticisms from many different sources. An interesting project in the history of economic thought would be to try to identify the various materials in Keynes s whole body of work that could be combined in an attempt to reconstruct a more watertight theory. This cannot be attempted here in the space available, but the following notes may be relevant to some future effort along these lines: The Nature and Functions of Money It has already been mentioned that there are materials in the first few chapters of the Treatise, and early drafts of The General Theory from 1933, that do address these issues in some detail. An obvious starting point, therefore, would be not to leave this discussion in the background, but rather to develop it further along the lines suggested (e.g.) in the section on the preferred theory of money above. Reviving the Nascent Theory of Banking If there was indeed a nascent theory of banking in the Treatise on Money then an obvious route to take that might lead back to a correct understanding of endogenous money, is to revive and develop this. Rochon (2012) has recently discussed the extent to which Keynes did have a fully-fledged theory of endogenous money, and comes to an ambiguous conclusion. In that case, it would remain for the current generation to fill in the gaps, drawing on the wide variety of heterodox theories of endogenous money that have been developed in recent decades by Post Keynesians, circuitists, MMT theorists, and others. 13

14 A Monetary Theory of the Real Rate of Interest It does seem clear that (in modern terminology) Keynes did intend to argue that the real rate of interest on money is determined in the financial markets. As explained above, however, it cannot really be claimed that this idea was fully articulated in the GT. 6 One of the problems in following Keynes s discussion of interest rates is the notion that, at the aggregate level, it is reasonable to speak about just one interest rate (be it real or nominal) the rate of interest, which can stand as a macroeconomic proxy for the whole complex of interest rates in general. This cannot really work in a system with central banking, commercial banking, and an endogenous supply of money. A minimum requirement is to distinguish between the policy rate and the commercial bank lending rate and to investigate the relationship between them (Kam and Smithin 2012, Smithin 2013a) If we do eventually come to conclusion that the rate of interest is essentially a socio-political concept, we would then faced squarely with the ethical question of what this real interest rate should be. In this connection, it is very important to clearly distinguish interest from profit (Smithin 2009, 2012,) which Keynes, in fact, had already done in his earlier Tract on Monetary Reform. This is a clear instance in which materials from Keynes s broader corpus of work that can be used in the construction of a more comprehensive monetary theory. Thinking about the question of the functional distribution of income in this way, leads eventually to the conclusion that basic trade-off in income distribution is that between profit, real wages, real interest, and the tax burden (Smithin 1982, 2009, 2013a). Disputes over the interest burden are therefore fundamentally similar to disputes about taxation. The Instrument of Monetary Policy In the mid-1930s, Keynes was faced with much the same sort of problem in talking about monetary policy that bedevils the field today, the notion of the zero bound on interest rate policy. This may be why Keynes was less clear on the importance of bank rate in his writings of the 1930s as compared to the 1920s. The fact that nominal policy rates are currently low, however, really has no bearing on their general theoretical and practical importance within the system, or on what might happen in the future. Nor does it affect the proposition that what matters is always the real rather than nominal rate of interest. A nominal policy rate of approximately zero could be associated with very high real interest rates if there is deflation at the same time, and vice versa if there is inflation. The lesson to be drawn here, presumably, is that the construction of theory per se, should not to be too much influenced by 14

15 current conditions and the necessarily ephemeral current policy debate. Keynes, in fact, did continue to argue that the general level of interest rates can be determined by policy decisions long after The General Theory had been published and the immediate problems of the depression years were over. Defending the Bretton Woods agreement in the British House of Lords in 1944, for example, and as quoted in (Smithin 2009), Keynes asserts that, we are determined that the external value of sterling shall conform to its internal value as set by our domestic policies, not the other way round we intend to retain control of our domestic rate of interest, so that we can keep it as low as suits our own purposes whilst we intend to prevent inflation at home we will not accept deflation at the dictate of influences from outside (emphasis added). Whether this would actually have been possible under the specific conditions of Bretton Woods is, of course, a debatable point. Monetary Non-Neutrality Harcourt (2012) has addressed the relationship of Keynes s General Theory to the imperfect competition revolution that had occurred in the same university (Cambridge) only 2 or 3 years before. According to Harcourt, Keynes apparently did not see that there was any connection. This is therefore a case in which the materials for a rehabilitation of the theory cannot be directly found in Keynes. 7 Smithin (2007, 2009, 2012) however, has shown that the situation can be repaired by following up the leads suggested by Kaldor (1983, 1985), which do in fact, propose to re-integrate the two Cambridge revolutions. It is interesting that in an interview with David Colander (Colander and Landreth 1996), Paul Samuelson once remarked that [w]e always assumed that the Keynesian underemployment equilibrium floated on a substructure of imperfect competition and administered prices. Moreover, the reason that these ideas were never really formalized was that there was no need to. Presumably, this was because the point was thought to be so obvious. However, it apparently had not been so obvious to Keynes and to very many other people in the economics profession. Conclusion It seems to me that one sort of tactical mistake that modern Keynesian and Post Keynesian economists do frequently make is to tend to argue, or imply, that there already exists a ready-made Keynesian theory somewhere in the General Theory, in the drafts of 1933, in the various explanatory articles published in the Economic Journal and Quarterly Journal of Economics in 1937 or in Keynes s later official discussions about exchange rate regimes and 15

16 the management of the public debt (Tilley 2010) that now simply needs to be revived, re-applied to current affairs, and all will be well. It has been argued in this paper that unfortunately this is not so. There is certainly a wealth of material in these sources that could employed, and has been employed by many over the years, as a starting point and inspiration, as well as in the construction of many of the actual building blocks for a more complete theory. Nonetheless, there remains a good deal of work for the present generation to do. Notes 1. John Smithin is Professor of Economics in the Department of Economics and the Schulich School of Business, York University, 4700 Keele Street, Toronto, Ontario, Canada M3J 1P3; tel: +1 (416) , ext.33623; e- mail: jsmithin@yorku.ca. 2. Capitalism in one country as Eric Kam and I have called it (Kam and Smithin 2008, 2011). 3. MMT stands for modern money theory. 4. Rochon refers, in particular, to the work of Sheila Dow on this topic. 5. For further discussion see the exposition in Kam and Smithin (2012). 6. Although a key statement from that book was Keynes s (1964/1936) explicit repudiation of the natural rate theory. According to Keynes: I am now no longer of the opinion that the concept of a natural rate of interest which previously seemed to me a most promising idea has anything very useful or significant to contribute to our analysis. 7. There is just one place in the GT, in chapter 18, where Keynes somewhat obliquely refers to the issue, stating that, for the purposes of the summary of the theory in that chapter, and among other things, the degree of competition is taken as given. This would certainly not be adequate if the whole theory was supposed to be floating on this point. References Bell, Stephanie The role of the state and the hierarchy of money. In Concepts of money: interdisciplinary perspectives from economics, sociology and political science, ed. Geoffrey Ingham, , Cheltenham: Edward Elgar. (First published in Cambridge Journal of Economics, 2001). Bougrine, Hassan and Mario Seccarecccia What is money? How is it created and destroyed?editors introduction. In Introducing Macroeconomic Analysis: Ideas, Questions, and Competing Views, eds. H. Bougrine and M. Seccareccia, 33-4, Toronto: Esmond Montgomery Publications. Burstein, Meyer Classical Monetary Economics for the Next Century, manuscript, York University, Toronto. Colander, David and Harry Landreth. eds The Coming of Keynesianism to America. Cheltenham: Edward Elgar. Davidson, Paul The Keynes Solution: The Path to Global Economic Prosperity, New York: Palgrave Macmillan. 16

17 Dillard, Dudley The barter illusion in classical and neoclassical economics. Eastern Economic Journal 14: Friedman, Benjamin The role of interest rates in Federal Reserve policymaking. NBER Working Paper 8047, December. Goodhart, Charles, The endogeneity of money. In Money, Macroeconomics and Keynes: Essays in Honour of Victoria Chick, eds. P. Arestis, M. Desai and S.Dow, 14-24, London: Routledge. Graziani, Augusto The Monetary Theory of Production. Cambridge: Cambridge University Press. Harcourt, G.C On Skidelsky s Keynes and Other Essays. London: Palgrave Macmillan. Hicks, John Critical Essays in Monetary Theory. Oxford: Clarendon Press. (First published 1967). Ingham, Geoffrey Babylonian madness: on the historical and sociological foundations of money. In What is Money?, ed. J. Smithin, 16-41, London: Routledge The Nature of Money. Cambridge: Polity Press Money is a social relation, In Concepts of Money: Interdisciplinary Perspectives from Economics, Sociology and Political Science, ed. G. Ingham, Cheltenham: Edward Elgar. (originally published in Review of Social Economy, 1996) Kaldor, Nicholas Keynesian economics after fifty years. In Keynes and the Modern World. eds. David N. Worswick and James R. Trevithick, 1-27, Cambridge: Cambridge University Press Economics without Equilibrium, Armonk, NY: M.E. Sharpe The Scourge of Monetarism. Oxford: Oxford: University Press. (first published 1982) Kam, Eric and John Smithin Unequal partners: the role of international financial flows and the exchange rate regime, Journal of Economic Asymmetries 5: Capitalismo en un sólo país? Una re-valoración de los sistemas mercantilistas desde el punto de vista financier. In Las Instituciones Financieras y el Crecimiento Económico en el Contexto de la Dominación del Capital Financiero, eds. Noemi Levy Orlik and Terasa López González, 37-58, Mexico City: Juan Pablo A simple theory of banking and the relationship between commercial banks and the central bank. Journal of Post Keynesian Economics 34: Keynes John Maynard A Tract on Monetary Reform. London: Macmillan The General Theory of Employment Interest and Money. London, Harcourt Brace and Company. (originally published 1936) A Treatise on Money (2 vols). Collected Writings, Vols. V & V1, ed. Donald Moggridge, London: Macmillan. (originally published 1930) The Collected Writings of John Maynard Keynes, XIII, The General Theory and After, Part 1. ed. Donald Moggridge, Cambridge: Cambridge University Press. Laidler, David Fabricating the Keynesian Revolution: Studies in the Inter-War Literature on Money, the Cycle and Unemployment. Cambridge: Cambridge University Press. 17

18 Lavoie, Marc Foundations of Post-Keynesian Economic Analysis. Aldershot: Edward Elgar Changes in central bank procedures during the sub-prime crisis and their repercussions on monetary policy. International Journal of Political Economy 39: Lavoie, Marc and Mario Seccareccia. eds Central Banking in the Modern World: Alternative Perspectives, Cheltenham: Edward Elgar. Leeson, Robert. ed. 2003a. Keynes, Chicago and Friedman, v. 1, London: Pickering & Chatto ed. 2003b. Keynes, Chicago, and Friedman, v. 2, London: Pickering & Chatto. Leijonhufvud, Axel On Keynesian Economics and the Economics of Keynes, New York: Oxford University Press a. Keynes and the Keynesians: a suggested interpretation. In Information and Coordination: Essays in Macroeconomic Theory, 3-15, New York: Oxford University Press. (First published in American Economic Review, 1967) b. The Wicksell connection: variations on a theme. In Information and Coordination: Essays in Macroeconomic Theory, , New York: Oxford University Press. King, John The Microfoundations Delusion: Metaphor and Dogma in the History of Macroeconomics. Cheltenham: Edward Elgar. Meltzer, Allan Keynes s Monetary Theory: A Different Interpretation: Cambridge: Cambridge University Press Mendoza d Espana, Alberto d Ansi Three Essays on Money, Credit and Philosophy: A Realist Approach Per Totam Viam to Monetary Science. Ph.D thesis in Economics, York University, Toronto. Menger, Carl, On the origin of money. Economic Journal 2: Moore, Basil Horizontalists and Verticalists: The Macroeconomics of Credit Money, Cambridge: Cambridge University Press. Nell, Edward and Matt Forstater. Eds Reinventing Functional Finance: Transformational Growth and Full Employment. Cheltenham: Edward Elgar. Patinkin, Don Price flexibilty and full employment. American Economic Review 38: Pigou, A.C The classical stationary state. Economic Journal 53: Parguez, Alain and Mario Seccareccia The credit theory of money: the monetary circuit approach. In What is Money? ed. John Smithin, , London: Routledge. Mosler, Warren The Seven Deadly Innocent Frauds of Economic Policy. Guildford CT: Valence Co. Inc. Rochon, Louis-Phillipe Credit, Money and Production, Cheltenham: Edward Elgar Mr. Keynes and the horizontalists: a suggested interpretation. Paper presented at the annual meetings of the History of Economics Society, Brock University, St, Catherines ON, June. Rymes, T.K. and Colin Rogers The disappearance of Keynes's nascent theory of banking between the Treatise and the General Theory. In What is Money?, ed. J. Smithin, , London: Routledge. 18