2006
HOW, AND FOR HOW LONG, DID KEYNES MAINTAIN THE TREATISE THEORY?
BY
TOSHIAKI HIRAI
I. INTRODUCTION
Many historians of economic theory have by now studied how Keynes developed his theory from the Tract on Monetary Reform through the Treatise on Money to the General Theory. After the pioneering studies by Moggridge (1973) and Patinkin (1976; 1982), there followed Dimand (1988), Amdeo (1989), Clarke (1988;1998), Meltzer (1988), Moggridge (1992), Skidelsky (1992), Laidler (1999) and others. This is no wonder, for the Keynesian Revolution remains the most singular phenomenon that economic theory and policy have ever seen.
Although the objective of our entire project has been to shed new light on this important and interesting phenomenon, examining and analyzing the processes of theory-building and re-building which constitute Keynes’s intellectual journey (see Hirai(1977-79)), the present paper focuses solely on one chapter in the long story of the transition through his three major works. The very fact of addressing the questions, “How did Keynes maintain the theory developed in the Treatise after its publication (October 1930), and for how long?” narrows the period under study to roughly two years which span roughly from October 1930 through October 1932. Our scrutiny will range over the original texts and primary material such as manuscripts, lecture notes, and correspondence produced over this period, and our findings will rest on the meticulous analysis of material of crucial importance for a clear understanding of Keynes’s theoretical situation. We will also offer our comments on the earlier efforts insofar as they relate to the period in question.
When we have answered our questions we will be able to say clearly when Keynes had finished with the Treatise theory, and when the ground was laid for the arrival of the General Theory.1
The period concerned has so far been examined in a very fragmentary way. To take a few examples, Patinkin (1976) deals with the MTP manuscript in less than one page (see p.86), while Amadeo (1989) confines his treatment of it – and only from the development of the multiplier – to one page alone and Dimand (1988) offers no examination of it.
Before entering upon the main topic, we must outline our basic view of Keynes’s theoretical development. The point we wish to stress is that Keynes was able to reach his theory of employment (and unemployment) as a result of abandoning “Keynes’s own theory” (to be explained below) and accepting the criticisms of several people with whom he corresponded. Although Keynes still clung to the idea of the TM supply function until late in 1932, his manuscript “The Parameters of a Monetary Economy”, which was written at the end of 1932, marks a clear turning point from the Treatise toward the General Theory. We have already clarified that the TM supply function was abandoned in this manuscript2 and Keynes’s abandonment of it shook the foundations of his theory as it had stood up to mid-1932: the main components of the General Theory were to be built as a consequence. We must also point out that the core of the Keynesian Revolution lies in the theory of underemployment equilibrium, which emerged in 1933, and for which we have three manuscripts.3
After a spell of “seven years off and on” since the publication of A Tract on Monetary Reform Keynes published the Treatise on Money in October 1930).4 Although the Treatise sill showed some traces of Dennis Robertson’s great influence on his thought, Keynes was setting off on his own path. And the Treatise received an extraordinary degree of attention.5
As we shall see later, the Treatise has two theories: “Keynes’s own theory” and a “Wicksellian theory”. For some time after the publication of the book, Keynes went on using and extending “Keynes’s own theory”, stressing the TM supply function in the face of criticism from various sides, including Robertson, Hawtrey, the “Cambridge Circus” and Hayek6 (interestingly enough, Keynes seldom refers to a “Wicksellian theory”.)
After explaining the theory developed in the Treatise in preparation, our examination proceeds thus, looking into: (1) Hawtrey’s criticism; (2) the sources from June 1931 to early 1932, which include two interesting manuscripts not contained in JMK; (3) a manuscript consisting of four chapters (JMK.13, pp. 381-396; hereafter “The Monetary Theory of Production” manuscript or the MTP manuscript7); (4) Criticism of the “Cambridge Circus”; and (5) two 1932 tables of contents.
II. THE TREATISE THEORY
In our view, the most significant feature of the Treatise theory is the coexistence of a Wicksellian theory and “Keynes’s own theory”.
1. Wicksellian Theory8
The monetary economics stemming from Wicksell and subsequently developed by various economists can be called the “Wicksell Tradition”. Wicksell’s basic insight involved a separation between the theory of relative prices and the theory of money prices in neo-classical economics.9 Wicksell put forward the “cumulative process” theory to explain money prices, while Myrdal, Hayek and Mises constructed their own versions of monetary economics based on Wicksell’s theory of cumulative process, criticizing the neoclassical system per se.
The Treatise also belongs to the Wicksell Tradition. Putting the market rate of interest together with the natural rate at its centre, distinguishing between investment and saving, and accepting Wicksell’s three conditions of monetary equilibrium, it explains fluctuations of money prices and recommends stability of the price level as an objective. Keynes’s Wicksellian theory, in which the second fundamental equation is used, plays a crucial role in the second volume of the Treatise.
2. KEYNES’S OWN THEORY10
Keynes develops his own theory in two parts, one of which addresses the determination of variables relating to consumption goods and investment goods in “each period”.
(Mechanism 1) The cost of production and the volume of output are determined at the beginning of the current period. Once the expenditure for consumption goods is determined on the basis of earnings, it is automatically realized as the sale of consumption goods proceeds, and the price level and amount of profit are simultaneously determined.
It should be noted that Mechanism 1 is substantially the same as the first fundamental equation.
(Mechanism 2) The cost of production and the volume of output are determined at the beginning of the current period. The price level of investment goods is determined either in the stock market (bearishness function) or as the demand price of capital goods. As a result, profit is determined.
The other part of Keynes’s theory concerns the determination of variables from one period to the next.
(Mechanism 3) The TM supply function
The behaviour of entrepreneurs is such that, if they make a profit (take a loss) in the current period, they expand (contract) output in the next.
Now “Keynes’s own theory” can be expressed as a dynamic process composed of Mechanisms 1 and 2 working through Mechanism 3. As a result, the economy may or may not reach long-period equilibrium. This interpretation sees the Treatise theory as articulating a dynamic process inclusive of price levels and volumes of output.
III. HAWTREY’S CRITICISM
1. HAWTREY’S ECONOMICS
Hawtrey’s trade cycle theory originates in Hawtrey (1913). It is a cumulative process theory of the banker-dealer connection based on two lags: the lag in the rate of interest behind change in rising or falling prices, and the lag in the demand for currency behind change in credit.
Hawtrey (1919) makes use of his key concepts of “consumers’ income” and “consumers’ outlay”, although the theory presented here is a revised version of the theory set out in Hawtrey (1913).11
Then we have Hawtrey (1928) and Chapter III of Hawtrey (1932), which further elaborate his theory and demonstrate that he held on to his earlier idea for many years. Let us explain his theory starting with the key concepts:
The total of the incomes which people … have to spend I call the consumers’ income; the total which they do spend I call the consumers’ outlay (Hawtrey, 1928, p. 83).
He finds the essence of the trade cycle in the variations of “effective demand” ─ variations, that is, in the consumers’ outlay ─ and the cause of these variations in the movement of bank credit .12
The trade cycle takes place in the transitional period during which the consumers’ income and outlay show divergence. If the banks judge that the level of their reserves relatively to the amount of credit money is lower (higher) than is appropriate, they then proceed to raise (lower) the rate of interest. In turn, the dealers, who hold stocks of commodities and stand between manufacturers and consumers, contract (expand) their bank credit and decrease (or increase) their orders to the manufacturers, who produce less (more) and reduce (increase) employment.
A contraction (expansion) of bank credit induces a return of currency to the banks (release of currency from the banks) and a decrease (increase) in consumers’ income, which in turn brings about a decrease (increase) in consumers’ outlay. A decrease (increase) in consumers’ outlay brings about an accumulation of unsold stocks, so that dealers give fewer orders to producers, who in turn reduce their production and employment. Production then decreases and unemployment increases, which induces a decrease (increase) in consumers’ income. During this process prices and profits fall (rise), followed – with a lag – by wages.
The process has a tendency to accelerate. That is, the divergence between consumers’ income and consumers’ outlay becomes wider and wider with the help of the above-mentioned lags.
Eventually the banks notice that the ratio of the reserves over the amount of credit money is higher (lower) than is appropriate, and lower (raise) the rate of interest.
Thus Hawtrey concludes that trade depressions must be due to deficiency in consumers’ outlay owing to credit squeezes.
2. HAWTREY’S CRITICISM
Hawtrey testified before the Macmillan Committee on 10 and 11 April 1930, and subsequently submitted three papers to the Committee.
On 23 April (Tm/1/2/8413) Keynes sent Hawtrey a batch of the proofs for Volume I of the Treatise, followed by a batch of proofs for Volume II on 24 June (Tm/1/2/89). Hawtrey responded to Keynes with long critical notes on 7 (Tm/1/2/94-151) and 9 (Tm/1/2/152-166) July.14
Throughout these exchanges Hawtrey stuck firmly to his own theory outlined above.
Here let us take a look at a few passages from his critical notes.
Mr Keynes’ formula only takes account of the reduction of prices in relation to costs, and does not recognise the possibility of a reduction of output being caused directly by a contraction of demand without an intervening fall of price (Tm/1/2/106).
The sequence here assumed is first a fall of prices, and then a contraction of output. With that assumption the unemployment inevitably appears as consequential upon the excess of saving over investment, [i.e.] the fall of prices (relative to costs) under another name (Tm/1/2/107).
It is … misleading to treat the discrepancy between investment and saving as an operative cause of monetary phenomena. … The cause of the divergence [between prices and costs] … is to be found in a change in demand, i.e. in the consumers’ outlay (Tm/1/2/110-111).
… while a windfall loss … produces a tendency to a reduction of output, this has not been … a contributory cause of actual … unemployment (Tm/1/2/116).
Hawtrey, questioning the TM supply function, emphasizes “present demand” based on the consumers’ “income” and “outlay”. He also stresses the role which the adjustment of goods in stock plays, attributing unemployment to a contraction of demand.
3. KEYNES’S RESPONSE
Keynes took his time before responding to Hawtrey’s criticism, on 28 November 193015, which shows just how much the TM supply function had been occupying his thoughts.
Firstly, Keynes was of the opinion that realized profits influence anticipated profits.
… I have laid too much stress on realised profits in respect of the production period just ended as influencing anticipated profits in respect of the production just beginning … (JMK.13, p.145).16
… there is not likely to be more than a transitory departure from the optimum level of output unless there is an actual or anticipated profit disequilibrium … (JMK.13, p.145).
Secondly, he explained that because ‘how much reduction of output’ is not a monetary problem, it is not dealt with in detail.
The question how much reduction of output is caused, … by a … fall of price17 …, is important, but not strictly a monetary problem. I have not attempted to deal with it in my book [the Treatise] , though I have done a good deal of work at it (JMK.13, p.145).
Thirdly, he feels the need to link monetary theory to the theory of short-period supply.
… I am not dealing with the complete set of causes … determin[ing] volume of output. For this would have led me an endlessly long journey into the theory of short-period supply and a long way from monetary theory…. If I were to write the book again, I should probably attempt to probe further into [making the two theories run together]; but I have already probed far enough to know what a complicated affair it it is.
As it is I have gone no further than that anticipated windfall loss or profit affects the output of entrepreneurs and their offers to the factors of production; but I have left on one side the question how much output is affected and also whether output can be affected in any other way (JMK.13, pp.145-146. The parentheses are mine).
It would be interesting to know how he thought the Treatise should be rewritten if time permitted in November 1930. We can say that the above idea was to take shape in the General Theory.
It should be noted that Hawtrey’s theory anticipates the General Theory although it is not a theory of employment or income, but one of income-outlay repercussion induced by bank credit. Hawtrey’s criticism of the Treatise’s fundamental equations and his stress on the importance of effective demand seem to have made a great contribution to weaning Keynes from the Treatise framework, to set off – after some hesitation – in the direction of the General Theory.18
IV. JUNE 1931-EARLY 1932
1. JUNE 1931
In June 1931 Keynes delivered the lectures19, “An Economic Analysis of Unemployment” in Chicago (JMK.13, pp. 343-367).
Bridel (1987, p. 152) correctly regards these lectures as belonging to the Treatise framework. He tends to take the Treatise theory in terms of interest rate as equalising investment and savings, although he takes the quantity adjustments into account. In our view, the former is related to a Wicksellian theory, the latter to Keynes’s own theory. Vicarelli (1984) considers that with these lectures “Keynes had made another fundamental break with the theory of the Treatise” (p. 106).
In these lectures Keynes discussed a trade cycle mechanism, emphasizing the TM supply function:
... when ... the value of current investment [I] is less than the savings [S] …, the receipts of the entrepreneurs will be less than their costs, so that they make a loss [profit = I - S]. That is … the clue to the scientific explanation of booms and slumps .... [W]hen [I] increases [decreases] (with no change in S) business profits increase [decrease] (JMK.13, p. 353).
A given deficiency of investment causes a given decline of profit [which] … causes a given decline of output (JMK.13, pp. 355-356).
In these lectures Keynes argued that the economy is inclined to proceed cumulatively, and might reach “a kind of spurious equilibrium”20 (JMK.13, p. 356).
Patinkin (1976, p. 68) rightly points out that this differs from the unemployment equilibrium of the General Theory. It would lead to the long-period equilibrium of the MTP manuscript.
In the face of the Great Depression Keynes emphasized the importance of reviving investment:
The cure of unemployment involves improving business profits [, which] can come about only by an improvement in new investment relative to saving (JMK.13, p. 362).
To increase investment, he argues, the solution is to lower the long-term rate of interest.21 Although he believes in the direct effect of the short-term rate on the long-term rate, Keynes suggests three other ways to this end:
(i) to increase the quantity of liquid assets;
(ii) to diminish the attractions of liquid assets by lowering the deposit rate of interest;
(iii) to increase the attractions of non-liquid assets.
He also mentions three lines of approach for the Great Depression, attaching great value to (3)22:
(1) restoration of confidence to lender and borrower;
(2) new construction programmes under the auspices of the public authorities;
(3) a reduction in the long-term rate of interest.
He then goes on to question the practicality of (2), while fully accepting its theoretical validity.
Theoretically… [(2)] can play an extremely valuable part in breaking the vicious circle. … I applaud the idea and only hesitate to depend too much in practice on this method alone … (JMK.13, p. 364).
Together with the lectures, we also have Keynes’s contributions to the Round Table on “Unemployment as a World Problem” (JMK.13, pp. 367-373). In one noteworthy instance he proposed:
… let us consider the totality of industries. You have over a short period something of the nature of a supply curve which tells you that for a given level of prime profit there will be a given level of output… Every increase in aggregate prime profit will enable somebody to expand, … so if you have a supply curve…, you could only increase … output by increasing prime profit (JMK.13, p. 368).
[Y]ou then aggregate all the curves and you can get a supply curve for industry as a whole in which the quantity of output is unequivocally related to the aggregate excess receipts over prime costs [profit] … (JMK.13, p. 372).
These observations attest to Keynes’s view that the quantity of output, both for an individual industry and for the whole industry, is related to profit, which keeps him tied firmly to his conception of the supply function in the Treatise.23
2. 20 SEPTEMBER 1931 AND TWO MANUSCRIPTS
In his letter to Kahn (20 September 1931. JMK.13, pp.373-375)24, Keynes argues out the possibility of underemployment equilibrium, which now makes its first appearance:
… if, starting with equilibrium, an increase of I makes Q positive, O increases and S increases but Q/O gradually diminishes. If Q/O reaches zero before O reaches maximum, we have ‘long-period unemployment’, i.e. an equilibrium position short of full employment. Similarly if a decrease of I decreases Q, O decreases and S decreases with the result that Q gradually increases until it is zero, which will be likely to occur before O is zero (JMK.13, pp. 374-375).
Behind this observation lie an idea that ‘profit is the difference between the value of investment and saving’ and a TM supply function — the theoretical framework of the Treatise. And, based on the TM supply function, a state of equilibrium, ‘long-period unemployment’, is explained in which employment stands at a level between zero and full. This idea might be the starting point for the idea found in GT, p.254 — “we oscillate … round an intermediate position …”.
For the same period we also have the manuscripts, “Why Are the Equations for Consumption-Goods and Investment-Goods Asymmetrical?”25 and “The Determination of Price”.26
In the manuscript, “Why Are the Equations for Consumption-Goods and Investment-Goods Asymmetrical?”, which may be closely related to Chapter 1, “The Differential of Consumption-goods and Capital-goods”, of Book III, “The Determination of Price”, of the First Table of Contents (probably written prior to April 1932. to be dealt with in Section VII below), Keynes sets out to explain, in response to Robertson’s criticism, why the equation determining the price of consumption goods is asymmetrical with that determining the price of investment goods.
Keynes’s main argument is that, given E, Q, and M, P1 depends on the difference between saving and investment, or on the propensity to save defined as S = φ (E, Q, P1, P2) while P2 depends on the propensity to hoard defined as M = ψ (H, P2) (E denotes the earnings, Q profit, M money supply, P1 and P2 the price levels of consumption goods and investment goods respectively, S saving and H hoarding).
Dealing with P1, he comes to the first fundamental equation of the Treatise.
Equalling the flow of money directed towards stomach-goods [consumption-goods] with the product of the price and quantity of such goods we have
E – S = … = R.P1
It is this equation which I re-analyse in my Treatise into the form
P = E/O + (I’ – S)/R where I’ = (E/O).(O-R) (GTE/1/27-28)
With regard to P2, he develops the following argument:
… we can equal the flow of new money … against the product of the quantity and the price of new hump-goods [investment goods] coming into the market. … Thus S + Q + dM = CP2 – (A3 +B3)P2
… We can, if we like, deduce from this P2 = - dM/ (A3 +B3)
But this is only another way of saying
P2 =ψ’(M) (GTE/1/33-34)
(C denotes the quantity of investment goods, A3 the release of consumption goods by the banks, B3 the release of investment goods by the banks.)
Here Keynes explains how the price of investment goods is determined in relation to money supply. He also uses the term, “bull-bear position”, which shows that he has not dropped the Treatise theory of the bearishness function.
The manuscript entitled “The Determination of Price” is very similar in content to the above manuscript. It explains how the price level of consumption-goods is determined, and then goes on to see how the price level of capital-goods is determined, emphasizing the difference between the two.
A significant difference from the above manuscript is, however, to be seen in the presence of the supply schedules, which make their first appearance here.
Let P1 = f1(A1)
and P2 = f2(A2) be the supply schedules of the two types of goods in the sense that A1 and A2 will be the quantity of each marketed for money … in response to prices P1 and P2 (GTE/5/469)
(where A1= O1+ Δ B1 and A2= O2+Δ B2. Oi and ΔBi respectively denote output, and the change in the stock of goods i [i= 1, 2] held by the public, where goods 1 are consumption goods and goods 2 capital goods.)
Keynes explains the price level of the consumption-goods based on the first fundamental equation of the Treatise, arguing that it depends on the “saving propensity”. With respect to the “saving propensity” S, Keynes argues as follows:
S will depend, apart from fundamental changes in popular psychology, on O, the community’s real income, on Q [profit] as determining the distribution of the real income and on P1 itself relatively both to P1 and to P2 (GTE/5/474)
This anticipates the propensity to save in the General Theory.
Keynes accounts for the price level of capital-goods taking into consideration the “curve of bearishness” or “hoarding propensity”. His argument runs thus:
P2 = ΔM/ΔB2 =Φ(M)
Where M is the total stock of money and Φ(M) what I have called the curve of bearishness, i.e. Φ(M) is the price for capital-goods at which … the public desire to hold an amount of money M rather than either to increase by purchase or to decrease by sale their holdings B2 of capital-goods (GTE/5/473).
Both manuscripts probably precede the manuscript “The Monetary Theory of Production”, for they adopt the same line as the Treatise on the bearishness function, and on the determination of the price levels of consumption and investment goods. In the manuscript “The Monetary Theory of Production”, as will be seen below, we find the theory of liquidity preference in embryo, while in the two manuscripts money is treated solely in relation to the price level of capital-goods, without referring to the rate of interest.
However, there is a possibility that the two manuscripts may have been written immediately after “The Monetary Theory of Production” manuscript, for the “supply schedule”, defined as the function of the price, appeared in the manuscript entitled “The Determination of Price”, while the “propensity to save” appeared in the other two manuscripts.
V. “THE MONETARY THEORY OF PRODUCTION” MANUSCRIPT (THE MTP MANUSCRIPT)
This manuscript is probably preceded in time by the preface to the Japanese edition of the Treatise (5 April 1932. TM.1, pp. xx-xxvii) which states the determination of the price level of capital goods in terms of “bearishness” (“the propensity to hoard”) but contains neither the determination of the rate of interest nor the term “liquidity preference”, whereas the MTP manuscript reveals the seeds of the liquidity preference theory.
Let us review the main points of the manuscript, and then assess it in terms of Keynes’s theoretical development.
1. THE SHORT-PERIOD ANALYSIS
In the MTP manuscript, the fluctuations of the economy and the possibility it faces of cumulative deterioration are analyzed using the TM supply function.
(A) The TM Supply Function
Keynes’s analysis below might be considered to be a defensive response to Hawtrey’s criticism (see Section III-2 above), showing deeper insight into the TM supply function. In “Notes on the Definition of Saving” which Keynes sent to Robertson on 22 March 1932, he places stress on the TM supply function: “it is not a matter of indifference in what proportions the income E′ of the community is divided between E [earnings] and Q [profit]. For if Q is positive, entrepreneurs will be under a stimulus to increase output, whilst if Q is negative they will tend to decrease output. This is a good reason for wishing to split up E′ into its constituents E and Q, and for “bothering" about the effect on Q of changes in F [expenditure on consumption] and I ′ [cost of investment]” (JMK.13, p. 279).
The following passage clearly demonstrates the importance that Keynes attaches to the TM supply function:
The essence of the monetary theory of production ... can be expressed quite briefly, starting from the equation
Δ Q =Δ I - Δ S,
or, …
Δ Q = Δ I + Δ F -Δ E,
or Δ Q = Δ D - Δ E
where Q stands for profit, I for investment, S for saving, F for spending, and D for disbursement. …
…we have started with the assumption that …entrepreneurs tend to increase [decrease] their output according as their profit is increasing [decreasing]. Thus we are led to…the vital generalisation that increases [decreases] in the volume of output … depend upon the changes in disbursement relatively to earnings ... or in investment relatively to savings.... Throughout this Book we shall be engaged in developing … this central generalisation (JMK. 13, p. 381. E denotes earnings).
Q is formulated as either I - S or D - E. To obtain the first formulation from the second, the equation F = E - S is required.
The MTP manuscript assumes an economy in which the total volume of output, fixed at the beginning of each period, is sold in that period. This is related to the TM supply function, in the context of which, as in the Treatise, both investment I and spending F are taken to be the sale proceeds realized as a result of the total volume of output being sold. The intrinsic logic of this function is developed as follows.27
The quantitative effect on output of a given decrease - Δ Q … will depend on: (i) the margin between each entrepreneur’s receipts and his variable costs...; (ii) the distribution of the total reduction - Δ Q between different entrepreneurs; and (iii) the duration of the period of diminished profit relatively to the durability of his fixed capital....[If (i) is sufficiently large initially, (ii) equal and (iii) short], … there is no reason to expect any significant decline of output, ....
Nevertheless, even so, an initial movement - Δ Q is likely to aggravate itself.... As a net result of … [a cumulative deterioration of - Δ Q, and the existence of firms in which the initial (i) is insufficient, the inequality of (ii) and the prolongation of (iii)], it is … reasonable to expect that a point will be reached at which there is some elasticity of supply in response to diminished profit, the initial reduction in entrepreneurs’ profit aggravating itself until, having reached an amount - Δ Q, it causes a reduction - Δ O … and a reduction - Δ E … (JMK.13, pp. 382-383).
From this we gather that Keynes regards the TM supply function as a useful analytical tool operating with some time lag.
(B) THE POSSIBILITY OF CUMULATIVE DETERIORATION
Using profit and the TM supply functions as pivotal concepts, Keynes argues that the economy might proceed along a course of cumulative deterioration:
... an initial movement - Δ Q is likely to aggravate itself. For the reduction in entrepreneurs’ profit will have a tendency to retard new capital development [investment] in respect both of value and volume…; and at the same time it may stimulate economy by diminishing both the [consumption] expenditure of entrepreneurs whose incomes are reduced and also the [consumption] expenditure of other consumers who can maintain their previous standard of life at a lower money cost. In short, the initial decline in disbursement is likely to generate ... a further decline in disbursement ... (JMK.13, pp. 382-383).
Here Keynes argues that a reduction in profit generates a further decline in disbursement, which in turn generates a further decline in profit, so that the deterioration becomes self-reinforcing. Behind this argument lies the idea that the volume of production of both investment goods and consumption goods is determined by the TM supply function.
He also discusses the case in which the economy deteriorates cumulatively even if the extent of the decline in profit remains the same as before:
Let us...assume that … the amount of disbursement declines by about the same amount as the decline in earnings.... On this assumption, the deficiency in the profit of entrepreneurs as a whole will remain exactly the same in absolute amount as it was before. But it will be spread over a smaller number of units of production, as a result of a certain number having fallen out of production. Consequently the average loss will be greater than it was, with the result that the next most vulnerable section of entrepreneurs now falls out of production.
Thus if we can imagine the entrepreneurs ranged in a continuous series according to what percentage reduction in their receipts impels them to close down production, it might be that a very small diminution -ΔQ in the total receipts of entrepreneurs might by a sort of inverse tontine process gradually close down one after another of them, until production was at a total standstill (JMK.13, pp. 384-385).
Thus Keynes argues that the economy in which a decline in disbursement initially generates a decline in profit can enter upon a process of cumulative deterioration, due either to a falling off in profit itself or to an increase in the average loss as a result of a decrease in the number of units of production.
2. THE LONG-PERIOD EQUILIBRIUM
How long does the economy continue to deteriorate cumulatively? Is there a possibility that it may reach a long-period state of equilibrium? Keynes answers the latter question in the affirmative.
In the MTP manuscript he emphasizes the relation between spending (consumption expenditure) and earnings (income). This relation is similar to that in the Treatise, and not directly connected with the GT consumption function. The following passage is to be read in this context:
... it is natural to expect that, as the earnings of the public decline, a point will eventually be reached at which the decline in total expenditure … will cease to be so great as the decline in [earnings] (JMK.13, p. 386).
Here “total expenditure” does not mean disbursement but consumption expenditure: the passage implies that the marginal propensity to consume (in terms of the General Theory) increases as earnings increase.
In the Treatise, the relation between spending and earnings is a factor causing a collapse in the credit cycle while, in the MTP manuscript it is regarded as warranting a kind of long-period equilibrium after a series of short-period equilibria. What the two share is the idea that spending has no influence on the production of consumption goods, but determines their price level, and that the profit from consumption goods determines their volume of production in the next period.
Keynes argues that the long-period equilibrium thus attained is characterized as a state of underemployment:
… provided that spending always increases [decreases] less than earnings increase [decreases] … [with investment stable] ...any level of output is a position of stable equilibrium. For any increase [decrease] of output will bring in a retarding [stimulating] factor, since ΔS will be positive [negative] and consequently I being assumed constant, Δ Q will be negative [positive] … (JMK.13, p. 387).
Here it is regarded as an important condition for stable equilibrium that the rate of the change in spending be smaller than that in earnings. Equilibrium generally brings about a state of underemployment equilibrium.28
... there is no presumption…that the equilibrium output will be anywhere near the optimum output. The essence of the above process is that the real income of the community has to be forced down to a level at which the rate of saving is not so excessive relatively to investment at the current rate of interest as to produce a crescendo of business losses… (JMK.13, p. 387).
3. THE RELATION OF INVESTMENT TO THE LEVEL OF OUTPUT
The MTP manuscript stresses the role of investment in determining the level of output, and of the government in promoting investment. However, it is no easy task:
... if we regard the response of individual spending to any given conditions of earnings and profits as something …determined by nature and habit…, then the level of output, which will be a stable level, entirely depends on the policy of the authorities as affecting the amount of investment.... [W]hen the output of the community increases a point comes eventually … when its spending … ceases to increase as rapidly as its earnings; … though we may … for a time maintain disbursement as a whole by increasing investment as rapidly as [savings] rises, a critical point comes when…we cease to be able to increase investment at an adequate pace, with the result that forces come into operation which prevent a further increase of output (JMK.13, p. 388).
We need to consider the relation between the short-period analysis applying the TM supply function and the proposition that the level of output depends on investment. Keynes seems to be saying that if investment increases in the initial period, then output will continue to increase over a period of time, resulting in an increase in output at a new long-period equilibrium. When he says “the level of output, which will be a stable level, entirely depends on the policy of the authorities as affecting the amount of investment”, he is stressing the influence that both investment and the investment policy of the authorities have on the level of output at a long-period equilibrium.
Keynes also says that “output will finally settle down to a position of equilibrium which is stable, so long as no extraneous influence interposes to change the value of I” (JMK.13, p. 388). He may seem here to be expressing reliance on an automatic mechanism driving the economy towards long-period equilibrium, but actually this is not the case. Having serious doubts about any such mechanism, he stresses the importance of a low interest rate policy for attaining optimum output: “there is no safeguard against savings increasing faster than … investment, except a monetary policy deliberately aimed at making a rate of interest sufficiently stimulating to investment” (JMK.13, p. 396).
As for the relation between monetary policy and long-period equilibrium, Keynes explains it clearly in the lecture draft for 14 November 1932.
4. THE ORIGIN OF THE THEORY OF LIQUIDITY PREFERENCE
The Treatise brings the focus on the “degree of bearishness” which is concerned with a portfolio selection between money as bearing interest, and equities. Contrastingly, the General Theory, which does not assume identity of equities and capital goods, brings the focus on the liquidity preference which is concerned with a portfolio selection between money as bearing no interest, and debts.
The term appears for the first time in the title of Chapter 20 in the table of contents, “The Monetary Theory of Production”29. Bridel (1987) argues that “ the ‘liquidity preference’ doctrine was already part and parcel of the Treatise” (p. 149), expressing it as ‘MD = M１(Y) + M(r)’ (p. 133) which he even describes as an “anticipation of the liquidity-preference equation of the General Theory'” (p. 207). In our view, this is misleading, for the rate of interest as a policy variable in the Treatise is directly concerned with the prices of equities (and the price level of investment goods). The similarity between the theory of the bearishness function and the theory of liquidity preference should be noted, without, however, ignoring the difference. Otherwise it would be impossible to understand Keynes’s subsequent development in this field.
In the MTP manuscript we see the liquidity preference making its first appearance.
... as output and prices decline, the proportion of the stock of money to income will...tend to increase. This growing relative abundance of money will, unless the general desire for liquidity relatively to income is capable of increasing without limit, lead…to a decline in the rate of interest (JMK.13, p. 395).
This may be read in terms of the General Theory as follows:
(i) a decrease in the demand for money due to the transactions motive makes money relatively abundant;
(ii) the abundance of money causes a decline in the rate of interest, unless the liquidity preference due to the speculative motive is infinitely elastic.
Here the MTP manuscript abandons two of the ideas in the Treatise: (i) that the price level of investment goods is determined in the financial market dealing with savings deposits and equities30; (ii) that the rate of interest is a policy variable.
In short, the role of the financial market switches from determining the price level of investment goods to determining the rate of interest. The rate of interest thus determined is expected to influence investment, which in turn influences the level of output through the TM supply function.
This change represents an enormous transformation in monetary theory. From the viewpoint of Keynes’s theoretical development, however, it is nothing like the change in the theory of the commodity market, for it does not undermine the grounds for the MTP manuscript to be argued within the Treatise framework.
5. THE PLACE OF THE MTP MANUSCRIPT
Patinkin (1976, pp. 71-73) finds several original ideas leading to the General Theory, although he correctly regards the MTP manuscript as being within the Treatise framework. Amadeo (1989, pp. 74-75) sees the embryo of the multiplier mechanism in the MTP manuscript, although he points out that “the adjustment process is based on the same causation scheme used in the Treatise” (p. 74).
We are now in a position to evaluate the place of the MTP manuscript31 in the development from the Treatise to the General Theory. Five points are worth noting in particular.
(1) Its fundamental framework belongs to the Treatise theory, for it maintains (i) profit and the TM supply function as determining the level of output, (ii) the relation, similar to Mechanism 1, between consumption and earnings, in that it determines not the level of output but the price level of consumption goods, and (iii) the value of investment, determined in the same way as in Mechanism 2.
(2) The basic theoretical structure is put forward as the dynamic process of Mechanisms 1 and 2 working through the TM supply function, while Wicksellian theory is no longer used.
(3) The TM supply function and profit are discussed in detail. This is particularly significant, for they are insufficiently discussed in the Treatise. ( (2) and (3) are not mentioned in Patinkin (1976; 1977)).
(4) The theory of liquidity preference appears for the first time – an important advance in monetary theory.
(5) We can pick out three salient features of the MTP manuscript – (i) the possibility of cumulative deterioration, (ii) underemployment equilibrium, and (iii) the role of investment and its fragility – back to the Treatise32 in (i) and the first half of (iii), and to the Chicago lectures “An Economic Analysis of Unemployment”33 in (ii) and the second half of (iii).
In the MTP manuscript Keynes describes the fluctuations of the economy using the TM supply function. He also describes long-period equilibrium as the economy reaching through a series of short-period equilibria, and thereby attempts to analyze short-period disequilibrium and long-period equilibrium in tandem. This was to come in for criticism from the Cambridge Circus – criticism which he resisted for some time, but which was eventually to prompt a great transformation in Keynes’s thinking.
VI. THE CRITICISMS OF THE CAMBRIDGE CIRCUS: MAY 1932
Immediately after the publication of the Treatise, several economists, including Kahn, Sraffa, Meade, and Austin and Joan Robinson, formed a group – the “Cambridge Circus” (January - May 1931) – to study it. The members attended Keynes’s lecture of 2 May 1932 and responded with a “Manifesto”, to which Keynes replied.
Patinkin (1976, Chapter 7) identifies three stages in the supply of criticism by economists that contributed to Keynes’s theoretical development: (i) Cambridge Circus, Hawtrey and Hayek’s criticism of the Treatise; (ii) discussions with Kahn and J. Robinson in 1932 and 1933; (iii) Hawtrey and Robertson’s criticisms of the galley proofs of the General Theory. Patinkin endorses (i), but not (ii) and (iii); he feels some difficulty in judging (ii) because of lack in documentation.
1. THE 2 MAY 1932 LECTURE
We can reconstruct Keynes’s lecture from some surviving fragments, probably his preparatory notes (JMK.29, pp. 39-42). The picture that emerges is substantially similar to the MTP manuscript, for the proposition, ‘the volume of output and employment depends predominantly on the amount of investment’, is stressed in terms of the TM supply function.
The argument begins by making two assumptions: (i) Δ O and Δ E΄ have the same sign; and (ii) Δ E΄ - Δ F and Δ E΄ have the same sign, where O denotes the volume of output, E΄ earnings plus profit, F spending, and Δ X an increment in X.
On these assumptions Keynes attempts to establish a positive relation between the volume of output and that of investment.
He then goes on to examine the case in which this relation does not hold. By looking closely into his procedure we can understand the relation between the positive output/investment relation and the TM supply function:
If..., whenever there was an increase in investment, there should also be such an increase in rates of earnings that the increase in aggregate earnings on the basis of the old output was greater than the increase in investment, and if earners were to save these increased earnings whilst entrepreneurs maintained their expenditure at their previous level, then every increase in investment would be associated with a decrease in profit and therefore in output. / Thus we are left with the remarkable generalisation that…the volume of employment depends on [that] of investment, and that anything which increases [decreases] the latter will increase [decrease] the former (JMK.29, p. 40).
Keynes argues for this relation in the context of “generalisations of far-reaching practical importance” (JMK.29, p. 40), stressing Δ Q = Δ I + Δ F - Δ E, which appeared in the MTP manuscript. He discusses the relation between the positive output/investment relation and the TM supply function in terms of not only the long period but also the short:
The general upshot of this…seems to be that the fluctuations of output and employment for a given community over the short period…depend almost entirely on the amount of current investment (JMK.29, p. 41).
The key to a proper understanding of the lecture lies in Keynes’s attempt to link the positive output/investment relation with the TM supply function. His analysis centres on the short-period process; long-period equilibrium is simply the point towards which the economy is moving. This is essential in understanding the differences between Keynes and the Circus.
2. THE CAMBRIDGE CIRCUS
The lecture drew criticism from the Circus in the form of a “Manifesto” by Kahn and the Robinsons (JMK.29, pp. 42-45), which, arguing that Keynes’s procedure for the proposition ‘an increase in investment (I) brings about an increase in the volume of output (O)’ lacks generality, puts forward an alternative way of proving it:
The problem seems to us to be susceptible to treatment by the method of Supply and Demand. For the truth of the proposition … , the two following conditions appear to us to be sufficient…:
(a) That an increase in I will lead per se to a rise in the demand for consumption goods, ….
(b) That the conditions of supply of consumption goods are not affected by a change in I.
When these conditions are fulfilled, an increase in I will lead to a rise in the demand curve for consumption goods without raising the supply curve, and so must lead to an increase of output of consumption goods, and a fortiori to an increase in total output (JMK.29, pp. 43-44).
Keynes put forward the proposition in terms of the TM supply function, while the Circus did so in the framework constructed by Kahn (1931)34, disregarding the function. The Circus’s position is very clear in Joan Robinson’s letter to Keynes:
You begin by increasing I. Now … tell me the elasticity of supply of capital goods i.e. how much increase in output does this ΔI entail. Then I will tell you for any set of conditions of supply of consumption goods what increase in E would be necessary to prevent O from increasing ....
In this sense I consider our method more general than yours. You announce in advance that yours only works when Δ Q and Δ O have the same sign. Ours is designed to overcome that limitation (JMK.29, p. 47).
The Circus’s argument was based on two considerations: (i) the elasticities of supply in capital and consumption goods industries; and (ii) the influence of an increase in investment on the demand for consumption goods.
On 8 May Keynes discussed the whole matter with Kahn and Joan Robinson. The next day he wrote a letter to Robinson, revealing some doubt about his own argument, while having no less doubt about Robinson’s:
[W]hich is the best of two alternative exegetical methods [?] ... my present belief is that ... your way would be much more difficult and cumbersome. … I lack at present sufficient evidence to the contrary to induce me to scrap all my present half-forged weapons (JMK.13, p. 378).
The “two alternative exegetical methods”35 are: (a) the credit cycle theory by way of the TM supply function; (b) the multiplier theory, including the method of supply and demand. The point of contention is unmistakably the supply function.
Kahn’s stance had already become clear in Chapter 4, ‘Supply Schedule of an Industry under Perfect Competition’ of Kahn (1929;1989). He was a virtual co-author of J. Robinson (1933). Moreover, Keynes ‘exercised a stimulating influence on [Kahn’s] progress’ (Kahn, 1989, p. xi). Taking these into consideration, it is evident that Keynes intentionally adopted method (a).
Robinson replied to Keynes on 10 May.36
Then the supply price for each output is (on your view) the average prime cost + the profit per unit just sufficient to retain the marginal entrepreneurs .... I believe that like the rest of us you have had your faith in supply curves shaken by [Sraffa]. But what he attacks are just the one-by-one supply curves that you regard as legitimate. His objections do not apply to the supply curve of output …37 (JMK.13, p. 378).
In mid-1932 Keynes was still using the average principle of the Treatise. The first sentence can be taken to mean that profit per unit is defined as the difference between the (supply) price as determined in the period, and the (average) prime cost as determined at the beginning of the period.
Let us now see how Robinson’s answer continues. First, she suggests that Keynes thinks that a supply curve, in the ordinary sense, is useful for the analysis of an individual firm or an individual industry, while Sraffa rejects this. Second, the supply curve of output [of the economy as a whole] lies outside Sraffa’s theory.
Here Robinson’s reference is to an article by Sraffa (1926) of epoch-making importance for the imperfect competition revolution, in which he maintains that increasing returns are applicable to the industry in the narrower range:
the more nearly [an industry] includes … only those undertakings which produce a given type of consumable commodity ― the greater will be the probability that the forces which make for increasing returns will predominate … [if the more nearly [an industry] includes all the undertakings, … decreasing returns will predominate] (Sraffa, 1926, p. 538).
Thus the “supply curve of output” works under diminishing returns, while the “one-by-one supply curves” work under increasing returns.38 Robinson seems to have this in mind.
If we understand Robinson’s comments aright, it follows that she fails to recognize that: (i) Keynes thinks that the TM supply function is useful for analysis not only of the economy as a whole, but also of an individual industry, considering that the ordinary supply curve is not useful in either case; (ii) Keynes had so far never used the “supply curve in the ordinary sense”. It was in mid-1933 that he came to accept the so-called “first postulate of the classical economics”.39
Indeed, Keynes stressed the TM supply function in a letter to Joan Robinson of 12 May:
About all one can say is that ... an increment in aggregate profit can reasonably be expected to produce an increment of aggregate output ― which is in substance what I have said.
… even when one is dealing with separate industries, or separate groups of industries, my supply curve is one which relates output and profit, not one which relates output and price (JMK.13, p. 380).
By the end of 1932, however, he had come to accept the Circus criticisms. He abandoned the TM supply function and adopted “the method of Supply and Demand”, following the lines laid down by Kahn and Robinson.
These lines were also adopted by Harrod (1936). Harrod criticised the analysis in terms of the TM supply function in the Treatise on the grounds that (i) it does not take into consideration a temporary equilibrium state; (ii) it does not take into consideration a marginal entrepreneur (p. 66).
This is not, however, to say that Harrod’s criticism of the Treatise was comparable with that of the Cambridge Circus. For ‘Keynes’s Treatise Influence on Harrod, 1930-39'’ see Young (1989, pp. 48-50). Also see Harrod (1969, Chapter 7, esp. pp. 163-165) which emphasises the importance of the Treatise in terms of (i) the distinction between cost inflation and demand inflation, (ii) the treatment of investment and saving.
VII. TWO TABLES OF CONTENTS (1932)
We now have extant two draft Tables of Contents that Keynes wrote in 1932 while he was in the early stages of developing the General Theory. This material is indispensable for a proper understanding of Keynes’s theoretical development.
1. KEYNES AND KITOH
In the preface to the Japanese edition of the Treatise (5 April 1932), Keynes announced his intention “to publish a short book of a purely theoretical character, extending and correcting the theoretical basis … in Books III and IV …” (Tm/1/3/255). Kitoh, the translator, asked Keynes about the progress of the book (30 June. Tm/1/3/270). Keynes replied that “…[it] is…still many months off” (20 July. Tm/1/3/271). On 18 September he remarked to his mother that “I have written nearly a third of my new book on monetary theory” (JMK. 13, p. 380).
When Kitoh repeated his enquiry to Keynes (20 October 1933. Tm/1/3/273), he replied: “…[My book] may possibly be published some time next year” (9 November. Tm/1/3/272), followed by a letter saying that “I am hard at work on my further book on the Pure Theory of Money …. But it will be some months more before … printing” (22 June 1934. Tm/1/3/274).
2. TWO TABLES OF CONTENTS
We have two tables of contents drawn up in 1932 (JMK.29, pp.49-50: hereafter the First Table of Contents, and the Second Table of Contents). It is a pity that the two tables contain nothing of the relevant text, but they nevertheless offer help in filling out the picture we are piecing together.
As we shall see below, the work to which Keynes referred in his letters of 1932 both to Kitoh, dated 20 July, and to his mother, dated 18 September probably had to do with a third (missing) table of contents.
(A) The First Table of Contents
The table of contents entitled “The Monetary Theory of Production” has several points worth noting:
(i) The titles of Chapters 1 - 4 (Book I) suggest that they may argue disbursement, profit, earnings, and output in the same way as the MTP manuscript in terms of the equation ΔQ = ΔI +ΔF - ΔE =ΔD - ΔE;
(ii) The title of Chapter 2, “The Relation of Profit to Output”, may point to the TM supply function;
(iii) Chapter 6, “Generalisations” (Book I) may have to do with the “vital generalisation” referred to in Section 4 (1(A));
(iv) Chapter 7, “Historical Retrospect” (Book I) may be related to a note, “Historical Retrospect” (JMK.13, pp. 406-407);
(v) Chapter 1, “The Differential of Consumption-Goods and Capital-Goods” (Book II) may be related to the manuscript “Why Are the Equations for Consumption-Goods and Investment-Goods Asymmetrical?”
(vi) The title of Chapter 2, “The Meaning and Consequences of ‘Bearishness’ ” (of Book III) may suggest the Treatise theory of money.
(B) The Second Table of Contents
The untitled table of contents, whose main chapters are almost the same as in the First Table of Contents, shows the following features:
(i) The title of Chapter 20, “The Factors Determining Liquidity Preference”, may propose the liquidity preference theory;
(ii) Chapters 23 - 27 (Book IV) bear titles regarding economic policy. No such chapters re-appear thereafter.
(C) Chronological Order
Let us consider the chronological order of this material. Obviously the First Table of Contents, referring to “bearishness”, precedes the Second Table of Contents, referring to “liquidity preference”.
The MTP manuscript was probably written on the basis of a third table of contents subsequent to these two table of contents, as indeed is suggested:
(i) by the chapter titles of the MTP manuscript, which have no corresponding titles in the two tables of contents;
(ii) by the fact that the MTP manuscript contains the seeds of the theory of liquidity preference.
However, this is not to deny that the two tables of contents are closely related to the MTP manuscript.
VIII. Conclusion
With the following three steps, it is possible to answer clearly and conclusively the question, “For how long did Keynes maintain the theory he had developed in the Treatise on Money?”
(i) Clarification of how Keynes set about defending and extending his own theory developed in the Treatise, expressed as a dynamic process composed of Mechanisms 1 and 2 through the TM supply functions.
Through the Chicago lectures, “An Economic Analysis of Unemployment” and the Round Table on “Unemployment as a World Problem” (June 1931) we saw how much stress Keynes placed on the TM supply function in his analysis. I also argued that the MTP manuscript (possibly prior to April 1932) analyzes the fluctuations of the economy and its possibility of cumulative deterioration by means of the TM supply function, and describes long-period equilibrium as a point reached after a series of short-period equilibria. Moreover, we saw that the MTP manuscript stresses the role of investment in determining the level of output, and of the government in promoting investment, and contains the seeds of liquidity preference. It should be noted that the MTP manuscript discusses the TM supply function and profit in more detail than the Treatise does.
(ii) The significance of two criticisms of the Treatise.
The first criticism came from Hawtrey in 1930, who, questioning the TM supply function, emphasized effective demand. The correspondence between Hawtrey and Keynes is very instructive in the sense that Hawtrey criticized Keynes’s theory from the standpoint of his own theory, which to some extent anticipates the General Theory, while Keynes reveals his state of mind concerning the TM supply function and the need to link monetary theory to the theory of short-period supply.
The second came from the “Cambridge Circus”, which, questioning the TM supply function, put forward the proposition in terms of Kahn (1931). Around May 1932 there was a controversy over the credit cycle theory between Keynes (the TM supply function) and the Cambridge Circus (the multiplier theory, including the method of supply and demand).
By the end of 1932, however, he had come to accept the Circus criticisms. 40 He abandoned the TM supply function and adopted “the method of Supply and Demand”, following the lines laid down by Kahn and Robinson.
(iii) A careful examination of two manuscripts probably written prior to the MTP manuscript: “Why Are the Equations for Consumption-Goods and Investment-Goods Asymmetrical?”, and “The Determination of Price”.
Both can be seen to argue that the price level of consumption goods is determined by Mechanism 1, the price level of investment goods by Mechanism 2, and in this respect they are in line with the Treatise theory. However, they also contain the beginnings of the propensity to save, and the latter manuscript has the “supply schedule” as a relation between price and output appearing for the first time. These indicate some change toward the General Theory.
From these crucial developments we can clearly trace out the time span between October 1930 and October 1932 during which Keynes held to his Treatise theory. The latter date calls for some explanation. It has to do with the Michalemas lecture (October – December 1932) and the manuscript “The Parameters of a Monetary Economy” written at the end of 1932. Although Keynes still clung to the idea of the TM supply function, “The Parameters of a Monetary Economy” manuscript drawn up at the end of 1932 together with the sixth (14 November) and seventh lecture (21 November) nevertheless mark a turning point from the Treatise toward the General Theory.41
Finally, although the present paper focuses solely on the period during which the Treatise theory was maintained, it must be said that the core of the Keynesian Revolution lies in the theory of underemployment equilibrium, which emerge clearly from the three 1933 manuscripts.
Notes
1) One of the major obstacles to studies of Keynes’s developmental process is the established understanding formed prior to serious studies based on the primary material. According to this “common knowledge”, the Treatise deals with the determination of price levels only, assuming given quantities of income, while the General Theory treats the determination of income assuming given price levels. This understanding permeates not only the Income-Expenditure Approach but also the Disequilibrium Approach. The Keynesian Revolution, therefore, is seen in the light of this widely shared perception to lie in the fact that Keynes established the theory of income in the General Theory, abandoning the theory of prices in the Treatise. If we view Keynes’s developmental process with this conviction in mind, there is a real risk of serious misunderstanding. Indeed, the material between the two major works seems to have received arbitrary treatment. Our understanding is that both the Treatise and the General Theory treat both prices and quantities, albeit in a different way.
2) On which, see Hirai (2004).
3) On which, see Hirai (1997-9, Chapter 10).
4) On which, see Hirai (2004).
5) On which, see Keynes Papers TM/1/3.
6) Interestingly enough, Pigou (1931) rates the Treatise highly. For example, he remarks: “[Keynes] claims for his new equation … that it enables the causal sequence, in many sorts of industrial disturbances, to be followed with a surer eye. This is, I think, a valid claim” (544). The only point of criticism is concerned with the TM supply function: “… as it seems to me, the relation of falling prices to industrial activity can be studied more effectively on the lines made familiar by Professor Irving Fisher than on those that … are followed by Mr. Keynes” (544). The correspondence between Pigou and Keynes survive very fragmentarily (see JMK.13, 214-218). Keynes stresses the advantages of his fundamental equations over the Cambridge equation in his letter dated 11 May. We have an impression that Pigou’s understanding of the Treatise as revealed in the letter dated May 1931 is on the right track. See Bridel and Ingrao (Macuzzo=Rosselli, 2005, pp.157-158).
7) This is not Keynes’s paper in the Festschrift for A. Spiethoff (JMK.13, pp.408-11).
8) On which, see Hirai (2005).
9) See, for example, Wicksell (1936, p.23).
10) On which, see Hirai (1997-9, Chapter 7).
11) See Hawtrey (1913, pp. 42-44, 107-110 and 124-126).
12) See Hawtrey (1928, p.84 and p.94).
13) “Tm/a/b/c” (where a, b, c stand for numbers) is a paging system actually adopted in Keynes Papers.
14) It should be noted that “ADDENDUM” (JMK.13, pp.150-164) is an excerpt produced from these by the editor. We shall, therefore, make use of the original ones.
15) The typescript is, however, dated 19 November 1930. See Tm/1/4/54.
16) We quote this passage as attesting to Keynes’s acceptance of Hawtrey’s criticism.
17) This seems to be Keynes’s response to Hawtrey’s above-mentioned comment, “The sequence here… prices”. JMK.13, 152’s reference to “8” might not be appropriate.
18) See Davis (1980), Cain (1982) and Deutscher (1990, p. 105).
19) See two sets of lecture notes prepared for the New School of Social Research in New York (11 June 1931. Keynes Papers, AV/1/40-53. “AV/1/bc” (where a, b, c stand for numbers) is a paging system actually adopted there). He advocates raising prices by increasing investment with lowering of the long-term rate of interest. The first fundamental equation and a kind of multiplier theory underlie his argument.
20) This is cited by Samuelson (1946) as adumbrating the General Theory.
21) See JMK.13, p. 366.
22) Keynes did not advocate deficit financing here, nor, let it be noted, did he in any other context throughout his entire career. See Clarke (1997) and Bateman (2005) in criticism of Buchanan and Wagner (1977).
23) The quotation above shows that he does not accept marginal analysis, for it is a response to H. Schultz’s argument based on the short-period marginal cost curve. See JMK.13, p.372.
24) Although Patinkin (1976, p. 84) erroneously regards the letter as showing an equilibrating mechanism, it is still within the Treatise framework.
25) Keynes Papers, Reel 33, GTE/1/23-41 (GTE/a/b. “GTE/a/b” (where a and b stand for numbers) is a paging system actually adopted there). Pen-written, and not contained in JMK.
26) Keynes Papers, Reel 35, GTE/5/474-499. Pencil-written, and not contained in JMK.
27) We can ascertain from the Keynes Papers that Chapter 8 of the MTP manuscript was initially entitled ‘Notes on the Effect of Changes in Output’. Another manuscript with the same title survives (Keynes Papers, Reel 35, GTE/5/424-437). Crossed out in the former is the passage: “So far we have assumed that the quantity of output is constant. And, indeed if we did not distinguish between S and S ′, we should have no clue to the effect on output of changes in the Investment and Economy Factors. But the problem can be attacked starting from the equation ΔQ = ΔI + ΔF - ΔE. Let us take the case where ΔQ becomes negative as a result of I decreasing more than F is increasing”. The former manuscript seems to have been written earlier than the latter.
28) Keynes expresses this with the term “involuntary unemployment” in a note, “Historical Retrospect” (1932. JMK.13, pp. 406-408), which is still argued in terms of the I-S difference. Here Keynes refers to Mercantilist and protectionist policies, anti-usury laws, etc. It may quite possibly precede the final lecture of the Michaelmas term of 1932: the title of the note is the same as that of Chapter 7 in the table of contents of ‘The Monetary Theory of Production’ (JMK.29, p. 49). Judging from the titles of Chapters 1 - 4, the table of contents is indubitably related to the MTP manuscript, although this note is thought to have been written after the table of contents.
29) See JMK.29, p.50.
30) The MTP manuscript maintains an idea that the price level of investment goods is determined by discounting the prospective yields at the rate of interest.
31) There exist two sets of material related to the MTP manuscript: (i) “Notes on the Definition of Saving”, where the TM supply function (JMK.13, p. 279) and the possibility of a cumulative deterioration of the economy (pp. 288-289) are argued: Keynes wrote this note immediately before the MTP manuscript, and, by using the fundamental equation ΔQ = ΔI + ΔF (where Q denotes profit, I the value of investment, F spending, Δ an increment in X), developed a critique of the theory of forced saving which was advocated by Robertson and others; (ii) the 1932 spring lectures (JMK.29, pp. 35-48): the draft for the lecture of 2 May is in line with the MTP manuscript.
32) See TM. 1, pp. 259-261.
33) See JMK.13, pp. 355-358 and 364 respectively.
34) On this paper in relation to the aggregate supply function, see Marcuzzo (2002, Section 3). Kahn’s fellowship dissertation (Kahn, 1989) was submitted to King’s College in 1929. The “short period” was evidently based on Marshall’s idea, which Robinson (1962) admired. The dissertation was to influence greatly Keynes’s path to the General Theory. See Kahn (1989, p. xi). It is also to be borne in mind that Kahn had delivered his first lectures on the Short Period in the Lent Term of 1931: on the position adopted in his lectures, see, e.g., Kahn’s Michaelmas lectures for 1932 (Kahn Papers).
35) Aslanbeigui, N. and Oakes, G. (2002, p.20) argue that “[Keynes] had no theoretical analysis to support his proof”. In fact, however, he had theory (a) from the Treatise on. This is why he hesitated to adopt theory (b).
36) There were heated discussions on the supply curve between J. Robinson, Kahn, and Sraffa around this period, for which see Marcuzzo (2003, pp.3-4). See also Robinson’s remark made in the summer of 1931; “Actually the supply of goods in the short period is likely to be fairly inelastic …, but not completely so” (1933, p.82).
37) For a theoretical rift between Kahn=Robinson and Sraffa, see Marcuzzo (2003).
38) Sraffa (1926, p. 543) says that “a very large number of undertakings … work under … diminishing costs”.
39) Keynes first discussed it critically in ‘The First Manuscript’ (1933. JMK.29, p. 66). He accepted it in ‘the Second Manuscript’ (1933. JMK.29, p. 72) as well as in the 1933 Michaelmas lectures (JMK.13, p. 420). See Hirai (1997-1999, Chapter 10).
40) By contrast, however, Keynes did not accept Roberson’s and Hayek’s criticisms.
41) See Hirai (2004). “The Parameters of a Monetary Economy” manuscript has been almost entirely neglected — a significant point distinguishing our interpretation of what marked the end of the Treatise period from the others.
REFERENCES
Amadeo, E., 1989. Keynes’s Principles of Effective Demand, Edward Elgar.
Aslanbeigui, N. and Oakes, G., 2002. “The Theory Arsenal: The Cambridge Circus and the Origins of the Keynesian Revolution.” Journal of the History of Economic Thought, no.1.
Backhouse, R. and Bateman, W. eds. 2006. The Cambridge Companion to Keynes, Cambridge University Press.
Bateman, B. 1996. Keynes’s Uncertain Revolution, University of Michigan Press.
Bateman, B. 2005. “Scholarship in Deficit: Buchanan and Wagner on John Maynard Keynes.” History of Political Economy, Summer.
Bridel, P. 1987. Cambridge Monetary Thought, London: Macmillan,.
Buchanan, J. and Wagner, R. 1977. Democracy in Deficit, Academic Press.
Cain, N. 1982. “Hawtrey and the Multiplier.” Australian Economic History Review, 20, pp. 1-27.
Clarke, P. 1988. The Keynesian Revolution in the Making 1924-1936, Clarendon Press.
Clarke, P. 1997. “Keynes, Buchannan, and the Balanced Budget Doctrine” in Maloney, J. ed.
Clarke, P. 1998. The Keynesian Revolution and its Economic Consequences, Edward Elgar.
Davis, E.G. 1980. “The Correspondence between R.G. Hawtrey and J.M. Keynes on the Treatise.” Canadian Journal of Economics and Political Science, XIII, pp.716-724.
Deutscher, P. 1990. R.G. Hawtrey and the Development of Macroeconomics, London: Macmillan.
Dimand, R. 1988. The Origins of the Keynesian Revolution, Edward Elgar.
Harrod, R. 2003. The Collected Interwar Papers and Correspondence of Roy Harrod, Electronic
Version, ed. Besomi, D., Cheltenham, Edward Elgar
(http://economia.unipv.it/harrod/edition/editionstuff/contentsfr.htm?rfh.htm~target).
Hawtrey, R. 1913. Good and Bad Trade, London: Longmans, Green and Co.
Hawtrey, R. 1919. Currency and Credit, London: Longmans, Green and Co.
Hawtrey, R. 1928. Trade and Credit, London: Longmans, Green and Co.
Hawtrey, R. 1932. The Art of Central Banking, London: Longmans, Green and Co.
Hirai, T. December 1997 – March 1999. “A Study of Keynes’s Economics (I) - (IV).” Sophia Economic Review, 43(1[pp. 67-136], 2[pp. 13-121]), 44 (1[pp. 35-127], 2[pp. 29-96]).
Hirai, T. 2004. “Keynes’s Theoretical Development from A Tract on Monetary Reform to A Treatise on Money” read at the History of Economics Society Annual Meeting, Toronto, June.
Hirai, T. 2004. “The Turning Point in Keynes’s Theoretical Development,” History of Economic Ideas, XII-2, pp.29-50.
Hirai, T. 2005. “How Did Wicksell’s Theory of Cumulative Process Influence Keynes and His Contemporaries?” read at the History of Economics Society Annual Meeting, Tacoma, June.
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).
Kahn, R. 1929, 1989. The Economics of the Short Period, London: Macmillan.
Kahn’s Papers, RFK 4/15/1-13 (Kahn’s Lecture Notes in the Michaelmas Term of 1932 taken by Tarshis), King’s College, Cambridge.
Keynes, J.M., Keynes Papers, Reel 27, Treatise on Money, Tm/1/2.
Keynes, J.M., Keynes Papers, Reel 27, Treatise on Money, Tm/1/3.
Keynes, J.M., Keynes Papers, Reel 33, GTE.
Keynes, J.M., Keynes Papers, Reel 35, GTE.
Keynes, J.M., Keynes Papers, Reel 41, American Visits 1931 and 1934, AV/1.
Keynes, J.M. 1930, 1971. A Treatise on Money I & II, London: Macmillan..
Keynes, J.M. 1936, 1973. The General Theory of Employment, Interest and Money, London: Macmillan.
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).
Laidler, D. 1999. Fabricating the Keynesian Revolution, Cambridge University Press.
Laidler, D. and Sandilands, R., “An Early Harvard Memorandum on Anti-Depression Policies”, History of Political Economy, 2002, 34/3, pp.515-552.
Leijonhufvud, A. 2006. “Keynes as a Marshallian” in Backhouse and Bateman eds.
Maloney, J. ed. 1997. Debt and Deficit, Edward Elgar.
Marcuzzo, M.C. 2002. “The Demise of the Quantity Theory of Money”, History of Economic Ideas, X/1, pp.49-62.
Marcuzzo M. C. 2002. “The Collaboration between J.M. Keynes and R.F. Kahn from the Treatise to the General Theory”, History of Political Economy, 34(2), pp. 421-447.
Marcuzzo M.C. 2003. “Joan Robinson and the Two Strands of Cambridge Economists”, Prof. K. Yagi’s Workshop, Kyoto Univ., Dec.
Marcuzzo M. C. 2004. “From Market ‘Imperfections’ to Market Failures: Some Cambridge Challengers to Laissez-Faire”, Annals of the Society for the History of Economic Thought, 45, pp. 1-10.
Marcuzzo, M. C. and Rosselli, A., eds., Economists in Cambridge: A Study through Their Correspondence 1907-1946, Routledge, 2005.
Meltzer, A. 1988. Keynes’s Monetary Theory, Cambridge University Press.
Moggridge, D. 1973. “From the Treatise to the General Theory”, History of Political Economy, Spring.
Moggridge, D. 1992. Maynard Keynes, Routledge.
Patinkin, D. 1976. Keynes’ Monetary Thought, Durham: Duke University Press.
Patinkin, D. 1982. Anticipations of the General Theory? And Other Essays on Keynes, University of Chicago Press.
Patinkin, D. and Leith, J. eds. 1977. Keynes, Cambridge and the General Theory, London: Macmillan.
Pigou, A.C., 1931. “Mr. Keynes on Money”, The Nation and Athenaeum, 24 January, pp.544-545.
Robinson, J., 1933. “The Theory of Money and the Analysis of Output.” Review of Economic Studies, 1(1), pp. 22-26.
Robinson, J., 1933. The Economics of Imperfect Competition, London, Macmillan.
Robinson, J., 1962. “The General Theory after Twenty-Five Years” (in The Collected Economic Papers, Vol. III, Basil Blackwell, 1965, pp. 100 - 102).
Skidelsky, R. 1992. John Maynard Keynes. Vol.2. The Economists as Saviour 1920-1937, Macmillan.
Sraffa, P. 1926. “The Laws of Returns under Competitive Conditions.” Economic Journal, 36, pp. 535-550.
Tarshis, L. 1979. “The Aggregate Supply Function in Keynes’s General Theory.” In Boskin, M.J. ed., Essays in Honor of Tibor Scitovsky, New York: Academic Press.
Vicarelli F. (translated by Walker, J.) 1984, Keynes, Philadelphia: University of Pennsylvania Press.
Wicksell, K. (translated by Kahn, R.) 1936. Interest and Prices, Macmillan (original in 1898).
