Keynes' General Theory of Interest

Fiona MacLachlan

• Publisher: Routledge
• Year: 2005
• ISBN:
  9781134896806
• Edition: 1
• Topic: Business
• Subtopic: Business General
• Index: Business

In Keynes' General Theory of Interest Fiona Maclachlan rehabilitates the largely discredited liquidity preference theory of interest, providing an original and rigorously reasoned restatement of the theory. Her provocative book draws on the methodological tenets of the Austrian school and is grounded firmly both in the history of economic thought and in real world economic institutions.

Chapter 1

---

Introduction

---

The theory of interest has for a long time been a weak spot in the science of economics, and the explanation and determination of the interest rate still gives rise to more disagreement among economists than any other branch of general economic theory.

(Gottfried Haberler [1939] 1958:195)

Following the marginalist revolution of the late nineteenth century, economists were able to reach an unprecedented degree of consensus about the basic principles of the determination of price in a competitive market economy. Yet when these general principles were applied in the explanation of the rate of interest, the result was a set of several distinct theories. Controversy continued to rage over the rudimentary question of the ultimate cause of interest: was it due solely to time preference (Fetter [1914] 1977), or to the productivity of capital (Brown 1913,1914), or to both (Fisher 1907, 1930)? Eventually, the debate subsided and Fisher's theory came to dominate in textbook presentations; but many of the issues were never adequately resolved. As a result, controversy brewed again when the Cambridge, England critics of neoclassical capital and interest theory pointed to logical defects in the standard textbook treatment.

Within the controversy, the liquidity preference theory of interest has played a very limited role. In Daniel Hausman's (1981) thorough study of modern interest rate theory, for instance, it is not mentioned at all. ‘It seems that the liquidity preference theory of the rate of interest is now widely regarded as either unimportant, incorrect or both’ (McGregor 1985: 89). The theory survives to a certain extent in macroeconomic theory, but it has been neglected by economists studying the fundamental problems of value and distribution.

In the following work, we hope to show that the liquidity preference theory that was imperfectly set out by Keynes in his General Theory and subsequent articles, deserves a place in any serious debate about the theory of interest. Our primary aim is not to determine what Keynes really said, but rather to attempt a rational reconstruction of his theory. Keynes' work provides the inspiration but the end result bears a debatable resemblance to his seminal contribution. The issue of whether our restatement of the liquidity preference theory is consistent with Keynes' original intention is one that we consider subsidiary to our central thesis and will be dealt with separately in a later chapter.

AN OVERVIEW OF THE LITERATURE ON LIQUIDITY PREFERENCE

For many economists, the strongest association with the liquidity preference theory is the IS–LM model. The model has proven much more resilient than Keynes' original writings and the theory of interest implicit in the model has eclipsed Keynes' suggestive prose presentation of his own theory. The exact origins of the model are not clear (Young 1987) but Hicks is justly credited with popularizing it in his 1937 article, ‘Mr Keynes and the Classics’, in which Hicks complains that in his original exposition of the theory, Keynes did not adequately capture the interdependence that exists between his key variables. Keynes organizes the theory into a series of sequential steps and summarizes his theory as saying:

1 the money supply in relation to the state of liquidity preference determines the interest rate;

2 the rate of interest in relation to the marginal efficiency of capital determines the level of investment; and

3 the level of investment together with the marginal propensity to consume determines the level of aggregate income.

Hicks notes, however, that since Keynes is using the term liquidity preference to refer to the total demand for money and since the total demand for money depends on the level of income, the theory is circular. Hicks sets about to remedy the circularity in the reasoning by casting the argument in the form of simultaneous equations. His IS–LM model can be represented as a system of two equations in two unknowns:

M^3s = L(r,Y)

I(r,Y) = S(r,Y)

where M^3s is the supply of money, L() is the demand for money, r the rate of interest, Y the level of income, I the level of investment, and S the level of saving. The second equation is supposed to represent equilibrium in the goods market. Unlike conventional general equilibrium models in which individual prices adjust to clear the goods markets, Hicks' IS–LM model has the peculiar feature that all prices except the interest rate are taken as given. Thus, it is only through adjustments in income and the interest rate that equilibrium in the goods markets is restored. The model suggests that if the market for pizza ovens, say, is out of equilibrium then equilibrium will be restored through an adjustment in the interest rate and incomes which in turn react on the amount of saving. It is not clear, however, that a mechanism equilibrating saving and investment is part of Keynes' original system. In Keynes' theory ‘(ex ante) saving and investment are not brought into equality by anything whatever, save the purest accident. For they do not confront each other simultaneously in one and the same market’ (Shackle 1967: 239). A disaggregated view has equilibrium in the pizza oven market achieved essentially through an adjustment in the prices or quantities, or both, of pizza ovens. Any effect on saving is likely to be indirect and weak: it will not be central in the equilibrating process.

Yet despite this seemingly obvious criticism of the model, Keynes did not object to Hicks' 1937 paper. He wrote to Hicks about a draft of the article: ‘At long last I have caught up with my reading and have been through the enclosed. I found it very interesting and really have next to nothing to say by way of criticism’ (Keynes [1937] 1973,14: 79). The reaction is ambiguous. The term ‘interesting’ can be used as a compliment or it can be used as a way of evading saying something more definite, either negative or positive. It is certainly not a ringing endorsement. Keynes' not wanting to present criticisms might indicate that he accepted it but it might also mean that he did not want to be bothered to devote the careful thought necessary to discover any defect. At the time Keynes was under the gun of a great many critics. He may have chosen to let Hicks' paper pass without serious scrutiny because Hicks could be counted as a supporter: with so many dissenters to deal with, it might not have seemed worthwhile to find fault with a follower.

In any event, there appear to be several aspects of the IS–LM model that are inconsistent with Keynes' original position. Besides the use of the condition that savings equal investment to represent equilibrium in the goods market that we have already mentioned, there is Hicks' use of the liquidity trap to explain Keynes' insistence that the interest rate determines the marginal efficiency of capital and not vice versa. Hicks claims that Keynes' result is properly regarded as a special case of his theory that obtains only when the LM curve is horizontal, that is, only in the case in which the demand for money is infinitely elastic. Keynes (1936: 207), however, had discussed the liquidity trap and concluded that although it was a possibility, it had never occurred. It seems implausible, then, that it was the liquidity trap that he had in mind when he denied that the marginal efficiency of capital determines the interest rate.

Another aspect of the IS–LM model that appears inconsistent with Keynes is its neglect of the causal process through which the interest rate is determined. In a simultaneous equation model the question of the ‘determination’ of a variable is simply the mathematical question of which value of the variable simultaneously satisfies all the equations. It leaves to the side all questions of what actually happens in which markets to cause the variable to take on a particular value. Pasinetti (1974:46) says that Hicks ‘has in fact broken up Keynes’ basic chain of arguments. The relations have been turned into a system of simultaneous equations, i.e., precisely what Keynes did not want them to be'. In order to properly represent a causal process, one needs to abandon the simultaneous equation approach and frame one's theory in the context of historical time.

Hicks was later to express reservations about his 1937 article. Interestingly, the problem he points to is the way in which the model abstracts from time.

I must say that [IS–LM] is now much less popular with me than I think it still is with many other people. It reduces the General Theory to equilibrium economics; it is not really in time. That, of course, is why it has done so well.

(Hicks [1976] 1983: 289–90)

Hicks explains that there are two sides to Keynes' theory. One side dealing with liquidity preference and the marginal efficiency of capital which is ‘unquestionably in time’ (op. cit.: 289), and the multiplier theory which, like equilibrium economics, is out of time. Hicks argues that it was the latter side that had been developed by Keynes' successors while the former side was neglected.

The ‘Keynesian revolution’ went off at half-cock; so the line, which I believe to be a vital line, was smudged over. The equilibrists … thought that what Keynes had said could be absorbed into their equilibrium systems; all that was needed was that the scope of their equilibrium systems should be extended. As we know, there has been a lot of extension, a vast amount of extension; what I am saying is that it has never quite got to the point.

(op. cit.: 289)

The vast amount of extension to the IS–LM model to which Hicks refers begins with Modigliani's seminal 1944 article. In the article, Modigliani (1944: 48) states his main task to be ‘to clarify and develop [Hicks’] arguments taking into account later theoretical developments.' The major improvement made by Modigliani was his relaxation of Hicks' assumption of rigid real wages. In doing so, he reached the startling conclusion that if real wages are flexible, then liquidity preference does not determine the interest rate. Rather the interest rate is determined purely by real, or non-monetary, factors.

Modigliani's version of the IS–LM model consists of the following equations:

1 X = F(N)

2 W/P = F′(N)

3 N = N(W/P)

4 S = S(r,W/P,X)

5 I = I(r,W/P,X)

6 S = I

7 M^s = L(r,WP/W,X)

Where X is aggregate output, N is aggregate employment, F() is the production function with a given stock of capital, N() is the labour supply function, W is the nominal wage rate, P is the price level, S is saving, I is investment, M^s is the money supply and L() is the demand for money. The model is structured so that it can be divided into three separate sub-systems (Hahn 1955: 55) with 1–3 determining X, N, and W/P; 4–6 determining S, I, and r; and 7 determining P.

It is on the basis of this feature of the model that Modigliani reaches the conclusion that saving and investment determine the interest rate and liquidity preference determines the price level. There is some ambiguity, however, with the term ‘determines’, as we have mentioned already. The theory that Keynes was attacking in the General Theory was one that suggested that saving and investment play a causal role in the determination of the interest rate. Within Modigliani's model, on the other hand, determination has a mathematical meaning that does not necessarily imply any causality. It is only in Modigliani's verbal discussion that he reveals his belief that the interest rate is determined in the causal, as well as the mathematical, sense by saving and investment, independently of monetary factors. He buttresses the result of his model with three verbal arguments.

The first arises in his discussion of long-run equilibrium. He defines it to be a state in which the quantity of money that people desire to hold as a store of wealth is constant (Modigliani 1944: 60). The implication is that in long-run equilibrium, wealthholders cease to be operative figures in the loanable funds market and hence have no effect on the interest rate. Modigliani successfully rules out any long-run monetary effect on the interest rate through his choice of a definition of long-run equilibrium. But the question of the appropriateness of his definition arises. It might be plausible to define a long-run equilibrium as a state in which markets clear because there are tendencies pushing the economy in that direction. But there are no similar tendencies causing wealth-holders to hold onto a fixed amount of money as a store of wealth. As long as there is uncertainty about future rates of interest, there will be the possibility of variable levels of speculative balances, and there are no apparent tendencies towards certain expectations of the course of future rates. The environment in which wealth-holders make their decisions is inherently uncertain and in a constant state of flux. In eliminating that flux, Modigliani's definition of long-run equilibrium may be abstracting from an important feature of the loanable funds market.

The second verbal argument presented by Modigliani takes the form of a conceptual experiment of the type we shall employ in Chapter 4. He supposes that the marginal efficiency of capital falls from r₀ to r₁ and asks about the process through which equilibrium is restored. He then asserts — without explanation — that in ‘order for the system to reach a new position of equilibrium, it is necessary that the interest rate fall to [r₁]‘ (1944: 72). In saying so he is admitting that he believes the marginal efficiency of capital directly determines the interest rate. He ignores the possibility that the causality might run the other way and that the equilibrating process would involve either a reduction in the level of investment or a fall in the price of investment goods, both of which would cause the marginal efficiency of capital to rise back up towards r₀.

The third argument used by Modigliani to support his conclusion that the interest rate is determined by saving and investment involves the presentation of some empirical evidence. He notes that in developed countries with high rates of saving and capital accumulation, interest rates are low; while in the less developed countries in which savings rates and the level of capital accumulation are lower, interest rates are high. This empirical evidence is far from conclusive, however, since other factors can be adduced to explain the discrepancy. One can cite at least three:

1 in less developed countries there is a greater degree of distrust in the central monetary authority's commitment to keeping inflation under control and thus a higher inflationary premium is built into the interest rate;

2 in less developed countries there are fewer established and credit-worthy firms and thus the risk-premia are bound to be higher on average; and

3 lenders in less developed countries do not have the same access to reliable sources of information about creditworthiness and will add a margin onto the interest rate to compensate for the uncertainty.

Modigliani's 1944 version of the IS–LM model stimulated criticism which led to improvements. The criticism, however, is not along the lines of our observations above. Rather than being concerned with his substantive conclusion about the role of monetary factors in the determination of the interest rate, the criticism is about the logical structure of the model. Patinkin (1948,1949) questions Modigliani's assertion that the price level is determined by the money equation alone, while all the other variables are determined independently of the money equation. Patinkin argues that if money is allowed in the utility function so that it is a substitute for the other goods entering the utility function, then the demand for at least one good must depend on the absolute price level. If a change in the price level does not affect the demand for any other good, as Modigliani supposes, then it must mean that the amount of money people hold is always the equilibrium amount regardless of the price level. If that is the case, however, then the money equation cannot be used to determine the price level.

Hahn (1955) presents a similar argument. He notes that Walras's Law implies that the value of total demand for goods and money must be equal to the value of their total supply and concludes that it ‘is therefore of considerable importance to [Modigliani's] ...

Table of contents

• Cover
• Half Title
• Full Title
• Copyright
• Contents
• Preface
• 1 Introduction
• 2 Methodology and definitions
• 3 The theory of interest from an ‘essentialist’ perspective
• 4 A causal process analysis of equilibrium in the cash–debt market
• 5 Keynes and the liquidity preference theory of interest
• 6 The liquidity preference versus loanable funds debate
• 7 Intertemporal coordination and the liquidity preference theory of interest
• 8 Concluding summary
• Notes
• References
• Index