HOW DID KEYNES DEVELOP HIS THEORY FROM THE TREATISE TO THE GENERAL THEORY? ─ Main Propositions/ Ideas Examined ─

　　　　　　　　　　　　　

Sophia Symposium (24 September 2005)

<Draft>

HOW DID KEYNES DEVELOP HIS THEORY

FROM THE TREATISE TO THE GENERAL THEORY?

─　Main Propositions/ Ideas Examined ─

Toshiaki Hirai*

I.                   INTRODUCTION

In LKE and SKE (Chs. 7-17) we have looked at the process of development which took Keynes from the Treatise to the General Theory, presenting our analysis. Our main work is to analyze the processes of theory-building and re-building which constitute Keynes’s intellectual journey from the Treatise to the General Theory.

There are two central tasks in such an analysis. The first is to fix our interpretations of the theoretical structures of the Treatise and the General Theory, the starting and finishing points. The second is to interpret the manuscripts, memorandum, lecture notes and letters which Keynes produced in the intervening period, in order to shed light on particular stages in the developmental process.

In the course of our investigation we inevitably encountered the difficult task of identifying the continuities and discontinuities in the development of Keynes’s ideas. We also encountered the equally difficult task of evaluating from the point of view of the evolution of his theories as a whole the judiciousness of Keynes’s own estimation, in each period, of the continuities and discontinuities in his thought. In that sense this sort of investigations are, by nature, endeavors of analyzing and constructing the developmental process theoretically.

This investigations is necessarily accompanied by two tasks: firstly to identify the points of similarity and difference between each manuscript and those which preceded it (backward-looking analysis), and secondly to make the same comparisons with regard to the manuscripts which followed (forward-looking analysis).

To a considerable extent the outcomes of the second task are dependent on the conclusions of the first, of course. Reciprocally, in coming to grips with the second task we substantially enrich our understanding of the major works with which we are concerned. Furthermore, in working through the manuscript material, we must be careful to keep hold of the central threads in Keynes’s thought, so not to lose our way in tracking the complicated developmental trajectories of his ideas.Chs. 7-17 are the product of this interactive approach.

In this paper we set ourselves one task, through which we could considerably deepen our understanding of how Keynes’s major theoretical ideas evolved: to examine in detail the evolutionary process of the propositions and concepts which comprise the central theoretical structure of the General Theory. If our earlier examination of sources forms the woof of our study, then we might call this investigation the warp.

  What we need to do here is to see clearly the whole development of each of the main propositions and ideas. Each of them had its own particular path of development, of course. However, for the purposes of exposition we can conveniently divide the central ideas of the General Theory into three categories, namely:

(i) Propositions Keynes maintained consistently from the time he had arrived at the outlook of the Treatise on (Persistent Propositions);

 

  (ii) Propositions which were transformed in the evolution of Keynes’s thinking from the position he took in the Treatise to that of the General Theory (Changed Propositions);

  (iii) Propositions Keynes came to hold only after he had arrived at the worldview of the General Theory (GT Propositions).

II. PERSISTENT PROPOSITIONS

 

There are four important propositions which Keynes retained from the position of the Treatise in his later thinking.

1. The First Persistent Propositions

 (i)  the central theoretical structure: investment, which is determined by the rate of interest, determines the level of output;

  

(ii) the policy proposal: financial and public investment policies¹ are recommended on the basis of this theoretical structure.

  Keynes came to this position in 1924, the year in which he first began to construct the theoretical outlook of the Treatise. He maintained it consistently up to the time of the General Theory².

  The argument of the Treatise runs as follows: the rate of interest is supposed to determine both saving and the value of investment in the current period. Profit, determined as the difference between the saving and the value of investment in the current period, determines the volume of output in the next period. Regarding the policy suggestions, Keynes laid most stress on financial policy, in particular, manipulation of the rate of interest. Public investment policy was regarded as a second best alternative.

  In the General Theory Keynes argued as follows: the rate of interest, determined in accordance with the theory of liquidity preference together with the dictates of the marginal efficiency of capital, determines the value of investment, which determines the level of output (and therefore the level of employment) through the working of the multiplier. Regarding policy, Keynes now put as much weight on public investment as he did on financial policy (lowering the rate of interest).

  I emphasize that in regarding this proposition as one consistently maintained by Keynes over the period we are considering, I am not saying that the relation between the theoretical structure of the Treatise and that of the General Theory is continuous, in the sense that there are no fundamental differences between the theories of the two books, or that the position of the second is simply a natural development of the position of the first, with no truly revolutionary changes of theory. The question of whether the relation between the theories of two books is continuous in this sense or not can only be answered from a perspective in which they can be compared closely with each other. Here we have looked at the positions represented by the two accounts in a comparatively wide perspective. In a narrower perspective, the relation between the two is indeed discontinuous; that is, abrupt changes and radical developments are undoubtedly apparent. The truth is, however, that when we speak of the ‘Keynesian Revolution’, we are usually taking a comparatively narrow perspective.

2. The Second Persistent Proposition

Left to itself, the capitalist economy has an inherent tendency to fall into a stable state characterized by an employment level lower than full employment but above mass unemployment; that is, into underemployment equilibrium.

  Laissez―faire was the dominant social philosophy in Great Britain in the first half of the nineteenth century. But in the middle of the second half of the century, its influence was surpassed by ‘New Liberalism’ (or Collectivism), the self-correction of English Liberalism under the influence of Hegelianism, as represented by T.H. Green. This outlook allowed for some degree of intervention on the part of the government in some fields of the economy. (The ‘People’s Budget’ of 1909 and the National Insurance Act of 1911, the central figure in each of which was Lloyd George, then Chancellor of the Exchequer in the Asquith Government, reflected the spirit of the times.)

  The leading British economists of the first half of the twentieth century, including Keynes, also belonged to the New Liberalism School. Thus in The End of Laissez―Faire (1926), for example, Keynes pointed out two things as the agenda of government. One was ‘the growth and the recognition of semi-autonomous bodies within the State’³. The other was for the State to take responsibility for the ‘functions which fall outside the sphere of the individual’ and the ‘decisions which are made by no one if the State does not make them’⁴. Among these functions, Keynes mentioned ‘the deliberate control of the currency and of credit by a central institution, and ... the collection and dissemination on a great scale of data relating to the business situation’ for the purpose of ameliorating the inequalities of wealth, unemployment, the disappointment of reasonable business expectations, and the impairment of efficiency and production⁵.

  However, from at least the time of “An Economic Analysis of Unemployment”  (June 1931), in which he used the term ‘a kind of spurious equilibrium’⁶, Keynes   seems to have persistently emphasized the intrinsic proneness of the capitalist economy to fall into underemployment equilibrium. Thus, in a letter to Kahn (20 September 1931), he refers to the phenomenon under the term ‘long-period unemployment’⁷; in “Historical Retrospect” (1932) he refers to ‘involuntary unemployment’⁸; in “The Monetary Theory of Production” he refers to underemployment equilibrium in the long-period⁹; in the Michaelmas lectures (1932, VI) he refers to under-employment equilibrium in the long-period; in the Second Manuscript (1933), he refers to ‘chronic unemployment’¹⁰; in the Michaelmas lectures (1933) he criticizes Say’s Law and refers to the deficiencies in effective demand and the volatility of expected quasi-rents or the marginal efficiency of capital in the modern economy; and in the First Undated Manuscript (the end of 1932-the beginning 1933), he develops the theory of underemployment equilibrium on the basis of the fundamental psychological law¹¹.

  This last argument leads to a view of “The General Theory”:

...when employment falls to a low level, spending will decline by a smaller amount than that by which income has declined, by reason both of the habitual behaviour of individuals and also of the probable policy of governments; and this...is the explanation why a new position of equilibrium can usually be reached within a modest range of fluctuation (JMK.13, p. 446).

This passage is repeated virtually verbatim in GT, p. 98.

In the General Theory Keynes also explains underemployment equilibrium in the following way:

… we oscillate, avoiding the gravest extremes of fluctuation in employment and in prices in both directions, round an intermediate position appreciably below full employment and appreciably above the minimum employment a decline below which would endanger life (GT, p. 254).

 

Keynes’s arguments for underemployment equilibrium do not remain exactly the same throughout, though he was consistent in advocating its reality. Roughly speaking, the First Undated Manuscript marks the major dividing line in his approach.

3. The Third Persistent Proposition

The effect of a cut in money wages on employment is uncertain.

  Early on, Keynes recognized the positive effect of a cut in money wages on employment. For example, in “The Economic Consequences of Mr Churchill”¹², in which he criticized the United Kingdom’s return to the Gold Standard at pre-war parity in April 1925, he also suggested ‘a uniform reduction of wages by agreement, on the understanding that this shall not mean in the long run any fall in average real wages below that what they were in the first quarter of this year’¹³, and attacked the Baldwin Government’s policy of intensifying unemployment by means of credit restriction in order to force down wages.

Moreover, in “Notes on Professor A.C. Pigou’s Memorandum”¹⁴, a paper for the Committee of Economists (25 September 1930), Keynes said:

My view is that, if we were a closed system, the requisite reduction of money wages to restore employment ... would have to be simply enormous; but that, as we are not a closed system, a reduction of money wages would have cured most of the existing unemployment, via international trade before this reduction was large enough to have made more than quite a small effect in improving employment for the reasons which would be operative in a closed system’ (JMK.20, pp. 414-415).

And in ‘Addendum I. Proposals relating to domestic monetary policy to meet the present emergency’ of the Report of the Macmillan Committee, he said:

   the practical courses open to us [for increasing employment] come down to three: (i) A reduction of salaries and wages; (ii) Control of imports and aids to the export industries; (iii) Domestic enterprise assisted by state action, or subsidies to private investment at home.

There is probably no serious dispute as to all these courses, considered apart from their social and long-period consequences, having some effect in the right direction ... [However] we feel...that the practical results of an attempt to reduce salaries and wages are likely to be exceedingly disappointing’ (JMK.20, pp. 286-287).

 

Although Keynes here still recognizes the efficacy of a cut in money wages, there is a noticeable change in his tone. Indeed, he now quite clearly prefers courses (ii) and/or (iii) to course (i):

To meet the immediate problems, arising out of the world slump, a policy intended to direct increased purchasing power into the right channels, both at home and abroad, with a view to restoring equilibrium at the present level of costs, would, therefore, be much wiser, in our judgment, than a policy of trying to cut our costs faster than the rest of the world can cut theirs (JMK.20, p. 291).

 

At the latest, Keynes began to advance the view that the effect of a cut in money wages on employment must be uncertain from the time of ‘Unemployment as a World Problem’ (June 1931), in which he developed the proposition in a rather sophisticated way, making use of the theoretical framework of the Treatise. He expressed his basic position as follows:

my conclusion would be that the effect of a reduction of wages would be determined by whether the adverse effect on saving (and by adverse effect I mean the increase of saving) would be greater or less than the increased amount of investment which might take place by the employers interpreting it, whether rightly or wrongly, as something in their favour.... The net result depends upon quantities which we are not in a position to measure, but when one comes to the practical issue, I am opposed to wage cuts for a much broader ground (JMK.13, p. 370).

 

In “The Monetary Theory of Production” Keynes discussed this problem in Section 9, ‘The Effects of Changes in the Rate of Earnings’. He investigated whether ‘the effect of diminishing W_１ will be to increase Q and O, so that ΔQ and, therefore, ΔD - ΔE and ΔI + ΔF - ΔE will in these circumstances be positive’¹⁵ (W_１: efficiency wages; Q: profit; O: output; D: disbursement; E: earnings; I: investment; F: spending.¹⁶). Keynes concluded that:

on the balance of considerations ... there is no presumption that an all-round reduction of the variable costs of production will prove favourable to the volume of employment (JMK.13, p. 394).

In the Second Manuscript (1933), he argued that ‘firms, taken as a whole, cannot necessarily protect themselves from loss by making revised (i.e. more favourable) money bargains with the factors of production’¹⁷, because the fact that the earnings of the factors of production determine the demand for the output of production must be taken into consideration. This argument is closely related to the one found in the Michaelmas lectures (1933, III)¹⁸.

  In the Michaelmas lecture (1935, VIII) Keynes discussed the problem in roughly the same way as he did in GT, Ch. 19, ‘Changes in Money-Wages’, Section 2, where he applied the method of analysis of the General Theory directly to the problem of the effect of a cut in money wages on the level of employment. That is, he discussed the policy in terms of its repercussions on the propensity to consume, the schedule of the marginal efficiency of capital, and the rate of interest, stating that there is ‘no method of analyzing the effect of a reduction in money-wages, except by following up its possible effects on these three factors’¹⁹. The conclusion of Keynes’s analysis is that any improvement in the level of employment stemming from a cut in money wages will come mainly from ‘an improvement in investment due either to an increased marginal efficiency of capital...or a decreased rate of interest’²⁰.

  With regard to the former, a rigid money wage policy is straightforwardly preferable to a flexible one, for practical reasons.

With regard to the latter, a flexible wage policy can produce the same effects as a change in the quantity of money, but while neither of these two policies is sufficient to attain full employment, changing the quantity of money is superior to a flexible wage policy in a number of respects:

(i)                 with a flexible wage policy there is no means of ensuring uniform wage reductions for each class of labour;

(ii)               changing the quantity of money is superior in terms of social justice and social expediency, since classes are affected more equally;

(iii)             in contrast to a cut in money wages, an increase in the quantity of money will decrease the burden of debt. Keynes concludes that ‘the maintenance of a stable general level of money-wages is, on a balance of considerations, the most advisable policy for a closed system’ and that the same is true for an open system, provided that equilibrium with the rest of the world can be maintained through the mechanism of fluctuating exchange rates²¹.

 The general thrust of Keynes’s remarks regarding the effect of a cut in money wages on employment is that the result cannot be certain because the repercussions on the propensity to consume, the expectations concerning future money wages and other factors are uncertain.

  It is interesting that Keynes maintained three propositions irrespective of the changes in the theoretical framework used to demonstrate them. It is also worth noticing that the first proposition is common to the Treatise and the General Theory (in terms of our wider perspective), while neither the second nor third propositions appeared in the Treatise.

4. The Fourth Persistent Proposition

The heterogeneity of goods should be taken seriously as the foundation of economic analysis.

  In general, it was not Keynes’s analytical procedure to first build a precise micro theory before constructing his macro theory on its basis. In the Treatise the argument is developed for the most part in macro terms, though some discussion in micro terms does appear in a supporting role. However, as he drew nearer to the General Theory, Keynes’s inclination to pay attention to the relation between micro and macro theory did increase.

  The fundamental aspect of Keynes’s viewpoint which placed stress on the heterogeneity of goods was first expressed in “The Parameters of a Monetary Economy” and in the Michaelmas lectures (1932), as is apparent from their use of the term ‘complex’²². The Second Manuscript (1933), Ch.7, III²³ can be said to be the origin of GT, Ch. 4, ‘The Choice of Units’, Section II²⁴. (The Michaelmas lectures (1933) are also worth mentioning in this context.) By way of Chapter 4, ‘The Choice of Units’, of the Pre-First Proof Typescript, the fundamental emphasis on the heterogeneity of goods came to fruition in GT, Ch. 4. Keynes reiterated this point in the Michaelmas lectures (1934).

  For some elucidation of the development of Keynes’s thinking on the logical relation between the micro and macro structures in the period leading up to the General Theory, we can turn to the First Undated Manuscript²⁵.

 We also have two invaluable sources for understanding the logical relation between the micro and macro structures in the General Theory:

(i)                 Kahn (1931), which we can regard as an explicit statement of the mechanism of the consumption goods sector in the General Theory²⁶, and

(ii)               Bryce (1935)²⁷, which is thought to have been written before 3 June 1935.

 This includes sufficient evidence to suggest that the General Theory was developed by means of the aggregate supply functions of the investment and consumption goods sectors.²⁸ The following passage is especially worth noting:

 [Consumption] Expenditure being determined and the [aggregate] supply function of consumable goods, the employment in producing consumable goods is determined. Because investment and the [aggregate] supply function of investment goods are known, employment in investment is easily determined. Therefore total employment is determined (JMK.29, p. 137).

This is the same idea as that developed in A Reconstruction ²⁹.

  Keynes acknowledged Bryce’s paper in a letter to him (10 July 1935), saying that Bryce had ‘got into it [the paper] the main elements in my theory’³⁰.

We should also note the reference to the elasticity of the supply functions of the consumption industries in Keynes’s letter to Beveridge (28 July 1936):

 Take the case of an increase in investment, say, the building of additional houses. The men who are directly employed in building the houses will have a higher income than before. They will spend this income on consumption. If you maintain that there is no such thing as a multiplier, you are maintaining that the state of supply in the consumption industries is always inelastic, even when men and plant are unemployed in those industries.... Yet obviously you cannot really believe this. You would admit that, except when there is full employment, there is an elasticity of supply in the consumption-good industries, and that, if more men are employed on building houses, more men will also be employed in making things for the house-builders to consume. The only reason why the orthodox theory denies the multiplier is because it is in fact assuming that there always is full employment, so that output as a whole has a zero elasticity (JMK.14, p. 58).³¹

III. CHANGED PROPOSITIONS

 

A number of Keynes’s central propositions were significantly transformed in the process of his development from the position of the Treatise to that of the General Theory. Of the major spheres of theory in which Keynes’s propositions underwent such transformation, we shall take up as most important the investment theory, the consumption theory, and the theory of interest.

  1. The Investment Theory

  The investment theory of the Treatise was primarily concerned with the determination of the current period price level (Mechanism 2)³². The volume of investment in the ‘next’ period is determined via the TM supply function (Mechanism 3), and is thus determined by the supply side with a one period time-lag. The demand side plays no role in the determination of the volume of investment. According to this theory, an increase in investment generally brings about an increase in profit through a rise in the price level, so that it leads to an increase in the volume of output in the next period via the TM supply function. In other words, this investment theory includes a mechanism in which an increase in investment brings about an increase in the volume of output.

  We confirmed that the Treatise contains two theories of the determination of the price level of investment goods, which we called Investment Price Theory (1) and Investment Price Theory (2)³³.

Investment Price Theory (1) is the theory of bearishness function itself. In “The Monetary Theory of Production” (1932), Keynes transformed this theory, through the introduction of the concept of liquidity preference, from a theory of determining the price level of investment goods into one of determining the rate of interest in the money market³⁴.

  Investment Price Theory (2) was, in turn, retained in “The Parameters of a Monetary Economy”. There, however, the TM supply function was not adopted.³⁵ The volume of investment is considered to be determined simultaneously with the price level of investment goods (Revised Investment Theory (2)).

 In the Michaelmas lectures (1932), Keynes considered investment to be a function of the cost of production, the rate of interest and the streams of prospective quasi-rents. In the three manuscripts of 1933, we unfortunately find no passages in which the investment theory is discussed, apart from the appearance of the term ‘marginal efficiency of capital’ in the table of contents of the Third Manuscript (1933)³⁶. It is not until the Second Undated Manuscript (1932-33) that an investment theory similar to that of the General Theory is seen³⁷. Here the position of the rate of interest has changed, although the definition of the ‘marginal efficiency of capital’ is still different from that of the General Theory. Then, in “The General Theory” (1934) the investment theory of the General Theory makes its debut³⁸. Keynes reiterated this theory in the Michaelmas lectures (1934)³⁹.

  2. The Consumption Theory

  In the Treatise the theory of saving, in which saving depends on the market rate of interest, plays an important role in the determination of the price level of consumption goods, while the consumption theory, which deals with the functional relationship between earnings and consumption, plays an important role in the determination of the price level of consumption goods and in a collapse in the secondary phase of the credit cycle⁴⁰.

  In “The Parameters of a Monetary Economy” (1932), Keynes discussed the consumption theory in relation to an existence condition for the price complex of consumption goods⁴¹.

A concept close in definition to GT’s consumption function made its first appearance in the First Manuscript (1933)⁴². Keynes then discussed the normal (fundamental) psychological law and the propensity to consume in the Michaelmas lectures (1933, VI)⁴³ and the First Undated Manuscript (1932-33)⁴⁴. This is of considerable importance, because the consumption theory underwent a great transformation due to the introduction of the fundamental psychological law. One could say that, apart from the subjective and objective factors in the propensity to consume, GT’s consumption theory is virtually complete there.

The subjective factors make their first appearance in “The General Theory” ⁴⁵ and appear again in the Michaelmas lectures (1934)⁴⁶. The objective factors also make their first appearance in “The General Theory”, although the list continued to be changed up to the General Theory.

  3. The Theory of Interest

  In the Treatise, the rate of interest played a central role in the determination of the levels of investment and saving. As it was taken to be a policy variable however, no theory of the determination the rate of interest itself was provided.

 The theory of liquidity preference, according to which the rate of interest is determined at the point at which the preference for liquidity equals the supply of money, made its first appearance in “The Monetary Theory of Production” (mid-1932), and appeared again in “The Parameters of a Monetary Economy” and in the Michaelmas lectures (1932).⁴⁷

It is reasonable to conjecture therefore, that in at least an embryonic form, the theory of liquidity preference may well have been established quite early on, as compared with the other elements of the theoretical structure of the General Theory.

As far as the explanation of the motives for liquidity preference goes, however, Keynes continued to change his mind (Unfortunately, apart from the Michaelmas lectures, the documents which might show us the process of his development in this respect are not extant).

In the 1933 lectures, the motives for preferring liquidity are divided into income, business, and precautionary motives, all of which depend upon the business cycle, overdraft facilities and the rate of interest, and the speculative motive, which depends upon the state of bearishness, while in the 1934 lectures they are divided into the transaction motive, which depends upon current operations, and a store-of-wealth motive, which depends upon the rate of interest.

 The story does not end there: in the sixth lecture of 1935, the motives are divided into the transaction motive, which depends upon the volume of business and prices (in the short run), and the precautionary and speculative motives, which depend upon the rate of interest, while in the seventh lecture the classification finally corresponds to that found in the General Theory.

IV. GT PROPOSITIONS

  The most important propositions Keynes adopted after he had come to hold the worldview of the General Theory (that is, after the end of 1932/beginning of 1933), concern the determination of the level of employment, the postulates of classical economics, and the multiplier theory. The first two were not discussed at all until the First Manuscript (1933), while the multiplier theory was not incorporated into Keynes’s main theoretical system until that year, although he had first put it forward even before the publication of the Treatise.

  1.  The Theory of Determining the Level of Employment

 

  A. The Theory of Employment

It was also not until the First Manuscript (1933) that Keynes placed the determination of the level of employment at the centre of his analysis⁴⁸, although the turning point of this change came with “The Parameters of a Monetary Economy” (the end of 1932)⁴⁹. This contrasts with the manuscripts before “The Monetary Theory of Production”, in which the determination of prices (Mechanisms 1 and 2) precedes that of quantities (by the TM supply function).

 However, Keynes’s formulation of the theory of the determination of the level of employment did not remain the same after that, but rather changed several times before he reached the position we find in the General Theory. In discussing this problem, we should keep in mind that these changes accompanied significant changes in the definitions of some fundamental concepts, including the employment function, effective demand, and income.

  Keynes did not fully develop the theory of the determination of the employment level until the First Manuscript (1933), where he articulated the theory under the period concept of an ‘accounting period’, and in terms of on the one hand the aggregate demand function (the sum total of prospective investment and prospective consumption), and on the other the sum total of the cost function and an inverse function of the ‘supply function’ (which we called the pseudo-TM supply function).  

 Then, in the Second and Third Manuscripts (1933), Keynes discussed the stability of the employment level with reference to ‘effective demand’, defined in terms of the excess of aggregate expenditure (sale proceeds) over variable costs (i.e., what we called the pseudo-TM supply function mk2).⁵⁰

  In the Michaelmas lectures (1933), Keynes put forward both the model in which the level of income is determined by Y = C + I, and the model in which the level of employment is determined by N = f₁(N) + f₂(ρ), with N₁ = f₁(N) and N₂ = f₂(ρ), where Y is income, C is consumption, I is investment, N is the total number employed, ρ is the rate of interest, N₁ is the number employed in the consumption goods industry, and N₂is the number employed in the investment goods industry.

  A concept close to the aggregate supply function appeared for the first time under the name ‘employment function’ in “The General Theory” (1934) where the equation determining the level of employment was formulated as follows⁵¹:

  　　　　　　　 F(N) = f₁(N, r, E) + f₂(N, r, E)     (1)

where N is the level of employment, r the rate of interest, E the state of long-term expectation, and f₁(・) (= C_w) and f₂(・) (=I_w) are respectively the expected rate of consumption and the expected rate of investment.

  The employment function is then defined as D_w= F(N), where D_wis effective demand in terms of wage units. Equation (1) is the prototype of the theory explaining the level of employment in the General Theory. The duality to the effect that the employment function is used both as a supply concept and as an equilibrium concept is replicated in the aggregate supply function of the General Theory.

  In the Summer Manuscript (1934) Keynes formulated the theory of the determination of the level of employment in terms of the following equation⁵²:

  　　　　　　N = F₁{Q₁(N, r, e) }+ F₂{Q₂(N, r, e) }        (2)

where F₁(・) and F₂(・) are the employment functions for consumption and investment goods respectively, Q₁(・) and Q₂(・) are the propensity to spend and the propensity to invest respectively, and ‘e’ is the marginal efficiency of capital.

  In the Michaelmas lectures (1934), Keynes put forward the theory of effective demand in which the level of employment is determined at the point where the effective demand function, D = f (N), intersects the employment (or supply) function, D'′= F(N). The formulation is as follows:

                     F (N) = f₁(N₁) + f₂(N₂)

                     N = N₁+ N₂

                     N_１= f(N)

  From the Summer Manuscript (1934) to the Third Galley (June 1935), in defining effective demand Keynes put stress on the user cost, arguing that income differs from effective demand by this amount, and that the effective demand which includes the user cost is vitally important for the determination of the level of employment.⁵³

  From the First Galley I to the Third Galley, Keynes formulated the equation determining the level of employment as follows (the same formulation is used in the Michaelmas lectures (1934)⁵⁴):

  　　　　　　　　　　　D  =  f(N)        (3)

　　　　　　　　　　　　D′ =  F(N)         (4)

　　　　　　　　　　　　D  =  D′         (5)

where equation (3) expresses the ‘state of effective demand’ and equation (4) is the employment function.

  Despite the exclusion of user cost from the definition of effective demand, and although it was mainly as a result of this that Keynes had to change the definitions of several fundamental concepts⁵⁵, he nevertheless followed the above formulation in the General Theory.

 It is noteworthy that in the Michaelmas lectures (1935) Keynes put forward a model in which the level of employment is determined at the point at which the aggregate supply function intersects the aggregate demand function.

  Let us now examine how Keynes changed the definitions ‘effective demand’ and ‘employment function’ in the course of the gestation process.

  B. Effective Demand.

 The term ‘effective demand’ appeared for the first time in the Second and Third Manuscripts (1933). Keynes defined effective demand in terms of the pseudo-TM supply function mk2 and discussed it in relation to the stability of the equilibrium level of employment.⁵⁶

  In the Michaelmas lectures (1933), effective demand is defined as the difference between aggregate expenditure and income, and the fluctuations in the volume of employment were argued to arise from those in effective demand. It is possible that this is the same stance as was taken in the Third Manuscript (1933).

Subsequently, in a letter to Kahn (13 April 1934), Keynes defined effective demand in the following way:

Let W be the marginal prime cost of production when output is O. Let P be the expected selling price of this output. Then OP is effective demand (JMK. 13, p. 422).

  This is followed by the “The General Theory” (1934), in which effective demand is defined as the present value of the expected sale proceeds. Effective demand is equal to the sum total of normal profit and variable costs.⁵⁷

Then, in the Summer Manuscript (1934), Keynes defined effective demand as the sum total of the user cost and aggregate income⁵⁸. He maintained basically the same stance in the Michaelmas lectures (1934). Keynes maintained this definition of effective demand from the First Galley I to the Third Galley, saying that income as a realized value differs from effective demand as an expected value by the user cost⁵⁹.

 Finally, in the Great Revision (August 1935), Keynes defined effective demand as exclusive of user cost. This definition was followed in the Michaelmas lectures (1935) and adopted in the General Theory⁶⁰.

  C. The Employment Function

The term ‘employment function’ appeared for the first time in “The General Theory” (1934)⁶¹. Here the employment function is both a supply concept and as an equilibrium concept, and is in substance the same as the employment function of the General Theory, as it is simply the inverse function of the latter.

The formula of the employment function in the Summer Manuscript is the same as in the General Theory.⁶²

In the Michaelmas lecture (1934) the employment (or supply) function D′= F(N) was defined as the sum which will just make it worthwhile to employ N men (or the supply price of the output of N men).

Then, in the First Galley I, Keynes reverted to the formula of “The General Theory”. In the First Galley II, however, he went back to the formula of the Summer Manuscript. Finally, in the Great Revision Keynes tried to resolve the confusion by distinguishing the employment function from the aggregate supply function⁶³. This move was probably followed in the Michaelmas lectures (1935).

 

  D. Excess Profit

We should take note of the fact that during the course of this process the role played by excess profit gradually dwindled.⁶⁴

 In the Treatise, excess or windfall profit played an essential role in determining entrepreneurs’ decisions on the volume of output in the next period.

 In “The Parameters of a Monetary Economy” (1932) and the Michaelmas lectures (1932), profit was formulated as either investment minus saving or disbursement minus earnings.

 In the First Manuscript (1933) Keynes used excess profit as a factor on the supply side for the determination of the level of employment. In the Second and Third Manuscripts (1933) he used excess profit in connection with the stability condition for the equilibrium level of employment. In the Michaelmas lectures (1933), quasi-rent was defined with reference to the inducement in terms of short-period expectations.

  However, we should pay special attention to the fact that excess profit plays no role in the determination of the level of employment in “The General Theory” (1934), in which excess profit is defined as income minus effective demand. Rather, the stress here seems to be on normal profit⁶⁵.

In the Summer Manuscript (1934), Keynes used, and placed stress on, the user cost⁶⁶, which is akin to depreciation, but we find no argument for excess profit here. In fact, in the Michaelmas lectures (1934), quasi-rent was defined with reference to realized return. Keynes’s inclination to place stress on the user cost continued from the First Galley I up to the Third Galley⁶⁷, but in the Great Revision the user cost was excluded from the theory of the determination of the level of employment, a step probably followed in the Michaelmas lectures (1935), which introduced the concept of supplementary cost.

The above sequence explains why little or no argument for stability is found in GT, Ch. 3.⁶⁸

  2. The Postulates of the Classical Economics

  Keynes first stated the two basic postulates of classical economics in the First Manuscript (1933), rejecting both from the standpoint of the pseudo-TM supply function.

 Then, in the Second Manuscript (1933), he changed his mind, accepting the first postulate, but still rejecting the second ─ the position of the General Theory ─ although the relation between the acceptance of the first postulate and the system determining the level of employment is not clear⁶⁹. In the Michaelmas lectures (1933), Keynes reaffirmed this position.

 Subsequently, he came to attach greater importance than previously to the first postulate and less to excess profit. In “The General Theory” (1934), he repeatedly stressed the importance of the first postulate, and therefore the importance of the maximization of normal profit⁷⁰. In his letter to Kahn (13 April 1934), Keynes commented on this, saying that ‘otherwise the equality of price and marginal cost is infringed. This is the real starting point of everything’⁷¹.

  It should be emphasized that Keynes’s acceptance of the first postulate does not form one of the elements of which the Keynesian Revolution was made, as it was taken over from (neo-) classical economics. However, to simply state this is one thing; to fully articulate Keynes’s real attitude toward the classical postulates is quite another.

  3. The Multiplier Theory⁷²

  An extremely important point regarding Keynes’s development of the multiplier theory is the fact that notwithstanding he grasped the idea before the publication of the Treatise, it was not until 1933 that he actually attempted to incorporate it into his theoretical economics.⁷³

We can surmise from the title of Chapter 15, ‘The Magnitude of Changes in Employment Relatively to Changes in Investment’, of the table of contents of the First Manuscript (1933) that this was the first manuscript in which he described the multiplier theory.

The term ‘multiplier’ itself made its first manuscript appearance in the title of Chapter 16 of the table of contents of the Third Manuscript (1933)⁷⁴.

In the Michaelmas lectures (1933) we find the multiplier theory developed for the first time. It occurs next in the First Undated Manuscript (1932-33)⁷⁵. In the Michaelmas lectures (1935) Keynes also discussed the allowances one should make in calculating the multiplier.

  4. The Motivating Forces

  Let us conclude this section by returning to the question of the motivating forces behind the changes in Keynes’s theoretical economics in the journey from the Treatise to the General Theory.

I have emphasized that Keynes’s position during the course of this transition is always closely related to, and can often be gauged by, his treatment of the TM supply function. The loss entailed by the abandonment of the TM supply function was a heavy blow to the foundations of the theoretical framework Keynes had maintained up to middle of 1932.

The major elements in the theoretical structure of the General Theory, including the consumption theory, the investment theory and the theory of liquidity preference, were all worked out or readjusted in the light of the change of direction this loss forced upon him. The change of direction is clearly visible in “The Parameters of a Monetary Economy” (1932)⁷⁶ as well as in the Michaelmas lectures (1932).

  To put this another way, in the Treatise Keynes’s main concern had been with the analysis of the transitional period, whereas in the General Theory the focus of his concern shifted to the problem of the determination the level of employment in the short period (in Marshall’s sense). In this respect, both the First Manuscript (1933) and the Michaelmas lectures (1933), in which he put forward the equation for determining the level of employment based on the ‘accounting period’ concept, are of vital importance to our understanding of the shift in Keynes’s theorizing.

  Robinson (1933) actually predicted this shift, with quite amazing precision:

   The mechanism of thought involved in the equations of saving and investment compels its exponent to talk only of short-period disequilibrium positions. And it was only with disequilibrium positions that Mr. Keynes was consciously concerned when he wrote the Treatise. He failed to notice that he had incidentally evolved a new theory of the long-period analysis of output’ (1933, p. 25).

Robinson also wrote:

  To regard the profits as a direct cause of the increase in output is apt to be misleading, and since in long-period equilibrium there are no profits in Mr. Keynes’s sense, a theory which regards profits as the mainspring of action is incapable of dealing with long-period analysis (1933, p. 26).

In retrospect, Keynes accepted Robinson’s criticism. Robinson’s insight here was profound. Indeed, we know of no statements which better articulate the essence of the transformation in Keynes’s thought represented by the move from the Treatise to the General Theory. In LKE and STE, through our examination of Keynes’s manuscripts and lecture notes, we have been able to prescience of Robinson’s powerful insight.

V. CONCLUSION

　In this paper we examined in detail how the propositions and concepts which comprise the central theoretical structure of the General Theory evolved from the Treatise to the General Theory.

  On both ends of the time span we have the Treatise and the General Theory from both of which we can construct the theoretical models. In between there lie various material such as manuscripts, memorandum, lecture notes and letters. When we try to analyze how and why Keynes changed his theory, we think that it might be a powerful way for us to identify principal/characteristic propositions and ideas and to chase after them (when they came into being, when and how they changed or disappeared). This is what we have done in this paper.

  We examined this task, classifying main propositions into three categories: Persistent Propositions; Changed Propositions; GT Propositions.

   We should underline the crucial point that what Keynes continued to insist upon throughout, both in his lectures and in his manuscripts, was the need to construct a monetary economics in which the influence of money on production is fully worked out. This is the counterpart of the aspect of classical economics which deals with the real-exchange (or neutral money) economy. To summarize Keynes’s enterprise, by putting forward the theory of effective demand, he endeavored to show that, left to itself, the capitalist economy is characterized by an inherent tendency to fall into underemployment equilibrium. At the same time, he attacked classical economics, with its unquestioning acceptance of Say’s Law and support for laissez―faire, criticizing both the classical theory of interest and the quantity theory of money.

  In this respect, ‘the theory of employment’ (Section IV, 1) is what matters most in understanding Keynes’s theoretical development, followed by ‘the Motivating Forces’ (Section IV, 4).

(The results of the examination described in this paper are summarized as Table 1. Table 2 is a summary of Keynes’s Michaelmas lectures during 1932-35⁷⁷).

  1) With regard to public investment, Keynes remarked that he had ‘written much elsewhere, and need not enlarge on it here’ (TM.2, p. 338). On account of this fact, there is little or no argument for a public investment policy in the Treatise. Keynes ought to have ‘enlarged on it’ though, because the Treatise presents a grand system which includes not only theory but also policy.

2) See LKE, 6, 3(C), and SKE, 6, 2(B(c)).

3) See JMK.9, p. 288.

4) See JMK.9, p. 291.

5) See JMK. 9, p. 292.

6) See LKE and SKE, 8,1.

 7) See LKE, 8, 1(B), and SKE, 8, n.10.

 8) See LKE, 8, n.19, and SKE, 8, n.10.

 9) See LKE and SKE, 8, 2 (B).

 10) See JMK. 29, p. 102.

 11) See LKE and SKE, 11, 2 (A).

 12) JMK.9, pp. 207-230.

 13) JMK.9, p. 228.

 14) JMK.20, pp. 409-416.

 15) JMK.13, p. 390.

 16) See LKE and SKE, 8, 2 (A).

 17) JMK.29, p. 97.

 18) See LKE and SKE, 11, 1.

 19) GT, p. 262.

 20) GT, p. 265.

 21) GT, p. 270.

22) See LKE and SKE, 9, 1 (A).

 23) JMK.29, pp.71-73.

24) See LKE and SKE, 10, 2 (D).

 25) Our ‘third point’ in LKE and SKE, 11, 2 (A) where we looked at Keynes’s description of the market mechanism of the consumption goods sector.

26) See Kahn (1972, pp. 5-7), and JMK.13, p. 340. See LKE and SKE, 9, 3.

27) JMK.29, pp. 132-150.

  28) Bryce (1935) was read at Hayek’s seminar at the University of London. Besides the text, moreover, see the reference to the elasticity of the supply functions of the consumption industries in Keynes’s letter to Beveridge (28 July 1936. JMK.14, pp. 56-59).

  29) See Reconstruction, ch.1, 3, ch.3, 2 etc.

  30) JMK.29, p. 150.

  31) Concerning the development of Keynes’s thought on the logical relation between the micro structure and the macro structure leading to the position he takes in the General Theory, see LKE and SKE, 15, 2 and its appendix, both of which are based upon the detailed examination in my Reconstruction.

32) See LKE and SKE, 7.

33) See LKE and SKE, 7, 3 (B).

34) See LKE and SKE, 8, 2 (D).

35) See LKE and SKE, 9, 1 (A).

36) See LKE and SKE, Table 10-1.

37) See LKE and SKE, 11, 2 (B).

38) See LKE and SKE, 12, 1 (C).

39) See LKE and SKE, 12, 3.

40) See LKE and SKE, 7, 2 (E).

41) See LKE and SKE, 9, 1 (A).

42) See LKE and SKE, 10,1 (A).

43) See LKE and SKE, 11, 1.

44) See LKE and SKE, 11, 2(A).

45) See LKE and SKE, 12, 1.

46) See LKE and SKE, 12, 3.

  47) Clarke (1988) correctly states that the theory of liquidity preference had been established by November 1932. See pp. 229, and 264.

  48) See LKE and SKE, 10, 1 (A).

  49) See ‘Model 1’ in LKE and SKE, 9, 1 (A).

  50) See LKE and SKE, 10.

  51) See LKE and SKE, 12, 1 (A (b)).

   52) See LKE and SKE, 12, 2 (A).

   53) We confirmed this point in LKE and SKE,12, 2 (A) and 14, 1 (E(a)).

   54) See LKE, and SKE,12, 3 (A).

   55) See LKE and SKE,14, 1 (E (a)<1>).

   56) See LKE and SKE, 10, 3. 

   57) See LKE and SKE, 12, 1 (A).

    58)  See LKE and SKE, 12, 2 (A).

    59)  See LKE and SKE, 14, 1 (E).

    60)  See LKE and SKE,14, 1 (E) and 2.

    61)  See LKE and SKE, 12, 1 (A).

    62)  See LKE and SKE, 12, 2 (B).

    63)  See LKE and SKE, 13, 2 (C).

    64)  For a similar opinion, see Tamagaki (1985, p. 77).

    65)  See LKE and SKE, 12, 1 (A (b)).

    66)  See LKE and SKE, 12, 2 (A).

    67)  See LKE and SKE, 14, (β).

68) Professor Yuuichi Shionoya (Hitotsubashi University) considers that ‘The TM supply function is essentially the same as the aggregate supply function in the General Theory’. This differs from the view developed in the present book. I agree with Shionoya’s view that ‘The essential characteristics of the General Theory lie in the introduction of the aggregate demand function’. However, I think that it is vitally important for a proper understanding of the General Theory to pay attention to the fact that the aggregate supply function has a dual aspect, namely as a supply concept and an equilibrium concept. For more on this, see Annuals of the Society for the History of Economic Thought, Vol. 23 (1985), p. 89.

69) See LKE and SKE,10, 1 (A) and 2 (C).

70) See LKE and SKE, 12, 1 (A (c)). After the publication of the General Theory, Keynes (1939) examined, from the point of view of relevance to the real world, the possibility of revising the proposition in the General Theory to the effect that as output increases, money wages increase and real wages decrease. He also refers to the imperfection of markets. However, he clearly thinks that in so far as perfect competition is presupposed, the argument developed in the General Theory is effective. Cf. Hirai (1979a).

   71) JMK.13, p. 423.

72) Robertson opposed the multiplier theory mainly because of the static nature, ‘providing his own dynamic interpretation and consequently remolding the implications of the dynamic multiplier for the forced saving thesis’ (Presley, 1978, p. 169), and because of the improbability of the stability of the propensity to spend. For this see Presley’s penetrating analysis (1978, pp. 169-176), and Robertson (1940, pp. 117-121). As far as the long and detailed correspondence between Hawtrey and Keynes about the galley proofs of the General Theory, which are listed in JMK.13, pp. 565-633, and JMK.14, pp. 2-55, is concerned, what strikes us most is that they did not discuss the multiplier theory at all. Taking Hawtrey as one of the pioneers of the multiplier theory into consideration, what does it mean?

  For recent arguments around the multiplier theory, see LKE, 9, n.41. See also SKE, 9, n.29.

73) See LKE and SKE, 9, 3.

74) See LKE and SKE, 10, 3.

  75) See LKE and SKE, 11, 2 (A).

76) Concerning various evaluations of this manuscript, see LKE, 9, n 18.

77) We examined them in, respectively, LKE and SKE, 9, 11, 12 and 14.

References

“The Monetary Theory of Production” … Before April 1932. JMK.13, pp.381-396.

“Historical Retrospect” (1932) … JMK.13, pp.406-408.

“The Parameters of a Monetary Economy”  (1932) … JMK.13, pp.397-405.

“The First Manuscript” (1933) … JMK.29, pp.62-66.

“The Second Manuscript” (1933) … JMK.29, pp.63, 66-73, 87-92, 95-102.

“The Third Manuscript” (1933) … JMK.29, pp.76-101, and JMK.13, pp.421-422.

“The First Undated Manuscript” … the end of 1932 - the beginning 1933. JMK.29, pp.102-111.

“The Second Undated Manuscript” … the end of 1932 - the beginning 1933. JMK.29, pp.111-120.

“The General Theory” … Spring in 1934. JMK.13, pp.423-456.

The Summer Manuscript (1934) … JMK.13, pp.471-484.

The Pre-First Proof Typescript … Summer in 1934. See JMK.14, p.351.

The First Galley I, the Second Galley, the First Galley II, the Third Galley and the First Galley III … See LKE and SKE, Table 13-1.

The Great Revision … August 1935. See LKE and SKE, 14, 1(A).

  Michaelmas lectures (xxxx, y): the y-th lecture of the year xxxx.

TM supply function … See LKE and SKE, 7, 4.

pseudo-TM supply function …See LKE and SKE, 10, 1.

pseudo-TM supply function mk2 … See LKE and SKE, 10, 2.

LKE: Hirai T. 2003, Looking at Keynes’s Economics from Multiple Points of View,  University of Tokyo Press (in Japanese). LKE, x, y … LKE, Ch.x, Sec. y.

　SKE: Hirai T., December 1997-March 1999, A Study of Keynes’s Economics (I)-(IV), Sophia Economic Review, 43(1[67-136],2[13-121]), 44 (1[35-127],2[29-96]). SKE, x, y … SKE, Ch.x, Sec. y.

The Treatise I or TM.1 :Keynes J. M. 1971 [1930], A Treatise on Money I: The Pure Theory of Money, London, Macmillan (JMK.5).

  The Treatise II or TM.2: Keynes J. M. 1971 [1930], A Treatise on Money II: The Applied Theory of Money, London, Macmillan (JMK.6).

The General Theory or GT :Keynes J. M. 1973 [1936], The General Theory of Employment, Interest and Money, London, Macmillan (JMK.7).

 

Bryce, R., “An Introduction to a Monetary Theory of Employment” (1935.

Reprinted as JMK.29, 132-150).

  Clarke P. 1988, The Keynesian Revolution in the Making 1924-1936, Oxford, Clarendon Press.

Hirai, T. 1981, A Reconstruction of Keynes’s General Theory, Hakuto Shobou.

Hirai, T. 1998, “Recent Japanese Studies in the Development of Keynes’s

Thought: An Evaluation”, Annals of the Society for the History of Economic

Thought, 36, 128-137.

   Hirai, T. How Did Keynes Transform His Theory from the Tract into the Treatise?, presented at the HES, Toronto, June 2004.

Hirai, T. 2004, ‘The Turning Point in Keynes’s Theoretical Development’, History of Economic Ideas, XII-2, pp. 29-50.

  Hirai, T. How, and for How Long, Did Keynes Maintain the Treatise Theory?,

presented at the ESHET, Stirling, 9-12 June 2005.

Hirai, T. How Did Wicksell’s Theory of Cumulative Process Influence Keynes and His Contemporaries?, presented at the HES, Tacoma, June 2005.

Kahn R. 1931, «The Relation of Home Investment to Unemployment», Economic Journal, 41, 173-198 (in R. Kahn 1972, Selected Essays on Employment and Growth, Cambridge, Cambridge University Press, 1-34.

  Keynes J.M. 1939, “Relative Movements of Real Wages and Output”, Economic Journal, March.

Keynes J.M. 1972 [1931], Essays in Persuasion, London, Macmillan (JMK.9).

Keynes J.M. 1973a, The General Theory and After: Part I, Preparation, London, Macmillan (JMK.13).

Keynes J.M. 1973b, The General Theory and After: Part II, Defence and Development, London, Macmillan (JMK.14).

Keynes J.M. 1979, The General Theory and After: A Supplement, London, Macmillan (JMK.29).

Keynes J.M. 1981, Activities 1929-31, London, Macmillan (JMK.20).

Keynes J.M. 1981, Activities 1922-9, Part 1 and Part 2, London, Macmillan (JMK.19).

  Presley J. 1979, Robertsonian Economics, New York, Holmes & Meier.

  Robertson D. 1940, Essays in Monetary Theory, London, P.S. king & Son Ltd.

Robinson J. 1933, “The Theory of Money and the Analysis of Output”, Review of Economic Studies, 1(1), 22-26.

Tamagaki, R., A Structural Analysis of Business Cycles, Iwanami Shoten, 1985 (in Japanese).

---

* Faculty of Economics, Sophia University, Tokyo 102-8554. E-mail: hirai-t@sophia.ac.jp. The paper originates in Hirai (2003, Chapter 16, Section 1).