9
Keynes’s Economics in the Making
Toshiaki Hirai
Introduction
Since the publication of Keynes’s General Theory (1936) and the advent of the “Keynesian Revolution,” many scholars have attempted to interpret Keynes’s book and its impact. These Keynesian scholars have put forward their own Keynesian theories or their own versions of Keynes’s theories: among the various approaches, the income-expenditure approach, normally taken as the orthodox position; the Post-Keynesian approach, of which there are many versions; and the disequilibrium approach have proven influential. We are, of course, well acquainted with the fierce attacks launched by outsiders, such as the exponents of monetarism and the New Classical macroeconomics, but controversies proving no less fierce also developed within the “Keynes Camp.”
Almost all the above interpretations focus solely on the General Theory, disregarding Keynes’s other economic writings. Since the mid-1970s, however, a new line of research on Keynes’s economics has been developed, with studies of Keynes’s economics based on the Keynes Papers.1 The present study belongs to this genre and has two purposes. The first is to clarify a feature of the Cambridge school in the first half of the twentieth century through examination of business cycle theories, identifying in particular two streams—the “Marshallian tradition” and the “Wicksell-type theories.”.
The second purpose is to clarify Keynes’s theoretical development from the Treatise on Money to the General Theory. Here I will address two tasks, namely, analysis of the theoretical structures of the Treatise and the General Theory and examination and interpretation of each stage of development based on various manuscripts and correspondence contained in the Keynes Papers.
Two strands within the Cambridge school—with regard to business cycle theories
Within the Cambridge school, two trends can be distinguished in business cycle theories. One trend, represented by Pigou and Lavington, drew its inspiration from Marshall’s theory, the “Marshallian tradition,” while the other, represented by Robertson, Hawtrey, and Keynes (in the Treatise), was—consciously or unconsciously—to take up a Wicksell-type theory and came to contribute greatly to the development of monetary economics, along with Lindahl and Myrdal (the Stockholm school) and Mises and Hayek (the Austrian school).
Marshall and the Marshallian tradition
Marshall’s economic system is comprised of three theories: the theory of exchange through supply-demand equilibrium, the (cash balance) quantity theory of money, and the theory of business cycles.2
Marshall put forward the theory of exchange, in Principles of Economics (1890), as a theory of stable equilibrium of normal supply and demand, which, assuming the constancy of both the marginal utility of money and the general purchasing power of money and confining the object of analysis to one commodity, provides an effective way of working out intertemporal problems.
Marshall presented his version of the quantity theory of money around 1871. He was doubtful of the transaction approach as formulated by Fisher on the grounds that it does not specify the factors that govern the velocity of the circulation of currency. Instead, Marshall stresses the “influence which the credit of a currency exerts on the willingness of the population to hold much of their resources, either directly in the form of cash in hand and at a bank; or indirectly in the form of debentures and other stock exchange securities” (Marshall 1923, p.º47). This idea is the ultimate germ of Keynes’s theory of liquidity preference. It should be noted that Marshall also considers that the so-called Marshallian k moves in inverse proportion to an increase in the supply of money, so that an increase in the supply of money would decrease the value of each unit of money more than proportionately. Marshall develops the above argument subject to general credit being normal.
Marshall’s theory of business cycles was first expounded in the Economics of Industry (Marshall and Marshall 1879). Here he analyzes the case where the general credit structure is shaken. This theory of business cycles is characterized by three points given particular stress. First, there is the role played by the public psychology, and in particular by moods of confidence or diffidence, in causing fluctuations in the level of economic activity. Confidence breeds confidence in the upward phase, while diffidence breeds diffidence in the downward phase.
Second, we have the “multiplier process,” invoked in explaining the working of the economy. In both the upward and downward phases, change occurs first in the investment goods sector. This is followed by change in the demand for consumption goods, induced by the increased incomes of those newly employed in the investment goods sector. Third, there is the role played by speculators in causing excessive economic fluctuations.
Essentially, Marshall sees the chief cause of depression as a lack in confidence, which undermines credit. In consequence, the economy fails to adjust means to ends. The solution he comes up with is to ensure that credit is managed within narrower limits. Compared with the theory of exchange developed in Principles of Economics, Marshall’s theory of business cycles was rather patchy. The work of developing it was left to his disciples, Pigou and Lavington.
Pigou developed his argument in Industrial Fluctuations (1927), which is an extended version of part VI, “The Variability of the National Dividend,” of Economics of Welfare (1920). Pigou takes fluctuations in the level of employment, which is given by the intersection of the supply schedule of labor with the demand schedule for labor, as the main statistical indicator of industrial fluctuations. Starting with this fundamental proposition, Pigou argues that the proximate causes of industrial fluctuations lie in the deviations in the demand schedule from its general trend. He goes on to argue that changes in the demand for labor, in turn, come about either through changes in the expectations of entrepreneurs with regard to the real yield obtainable from investment (“industrial spending”) or through changes in the size of real income, and he concludes that the dominant causal factor is the former rather than the latter. Thus, in Pigou’s view, the varying expectations of business people are the most important factor contributing to industrial fluctuations (see Pigou 1927, p.º34).
Pigou’s theory is based mainly on two analyses: (a) of the factors governing the amplitude of industrial fluctuations and (b) of the factors governing their rhythmic nature.
With regard to analysis a, Pigou argues that the amplitude of industrial fluctuations depends both on the initial causes and on the environmental conditions in which the initial causes have effect. The initial causes of variations in expectations of the profit to be had from industrial spending include real causes, such as harvest variations, new inventions, industrial disputes, wars, and changes in fashion; psychological causes, particularly errors of optimism or pessimism; and autonomous monetary causes.
The environmental conditions mentioned as influencing the amplitude of industrial fluctuations are monetary and banking arrangements (the principal problem here is forced saving); the policy adopted by industrialists with regard to spoiling the market; and the policy pursued by laborers aiming at making wage rates rigid.
Reacting to the environmental conditions, the initial causes that set industrial fluctuations in motion determine their amplitude. In this argument, Pigou places great emphasis on the multiplier process and the influences that excessive oscillations in the psychology of entrepreneurs have on the economy.
With regard to analysis b, Pigou identifies two major sets of causes. The first is causes that, though sporadic in nature, start off wave movements once they have come into play: the tendency of human constructions to wear out after a certain interval; the alteration of optimistic and pessimistic errors among entrepreneurs; and processes relating to money. The other causes comprise recurrent factors linked to the periodicity of actual industrial movements.
Pigou’s theory of industrial fluctuations is deeply rooted in psychological forces. He stresses that alternation of optimistic and pessimistic errors is the main factor governing both the amplitude and the rhythmic nature of industrial fluctuations. It should be noted, however, that in developing his argument Pigou pays attention to both real and monetary causes, as well as the multiplier process.
Lavington was a Cambridge economist, best known for his book The English Capital Market (1921). Lavington first argues that the service that the money market is able to render consists of two activities, namely, the manufacture of money and the transportation of capital. To provide the latter service is to facilitate “the movement of this stream of money, of Command over Capital, whereby the control over a part of the productive resources of society which is available for capital uses is transferred into the hands of those by whom it can most effectively be employed” (Lavington 1921, p.º12). This service is thus related to forced saving. Lavington then proceeds to analyze the money market, both in the normal state (i.e., where the level of credit is normal) and in the abnormal state (i.e., where the confidence to grant credit has been shaken). Lavington’s theory of the trade cycle consists in an analysis of the money market in the abnormal state.
Because the treatment in the English Capital Market is somewhat dispersed, it is better to follow the more straightforward account in the Trade Cycle (Lavington 1922). The key ideas there are drawn from Marshall, Pigou, and Robertson (in A Study of Industrial Fluctuation, 1915).
Lavington’s basic proposition is two-pronged. On the one hand, he presents the conditions conducive to rhythmic variations in the level of business activity. Modern industrial systems are equipped with features inducing rhythmic patterns and/or cumulative changes—the organization of production by independent entrepreneurs, the decision-making process applied by entrepreneurs based on forecasts rather than facts, and the mutual interdependence of entrepreneurs (Lavington 1922, pp.º52–53).
On the other hand, he exposes the causes that, operating under these conditions, actually induce the rhythmic patterns. The main cause is to be found, he argues, in changes in the level of business confidence. Lavington develops this basic proposition as follows:
First, he assumes that a rise (or fall) in the level of business confidence always induces an increase (or decrease) in business activity.
Second, he emphasizes both the multiplier process and changes in the level of business confidence, taking the effective purchasing power of society into consideration.
Third, he does not consider monetary factors to be important in causing industrial fluctuations. He argues that the principal factors inducing industrial fluctuations have direct relation neither to the price level nor to the monetary or nonmonetary nature of the economy. That said, he accepts that we do actually live in a monetary economy, in which a rise in the level of business confidence causes an increase in purchasing power and thus induces a rise in the price level, which in turn leads to a rise in the level of confidence. This process develops in a cumulative way.
Fourth, and most important, Lavington argues that it is essentially the changes in the level of business confidence that determine the amplitude and periodicity of the trace cycle (Lavington 1922, p.º92).
Pigou’s and Lavington’s theories of industrial fluctuations represent the Marshallian tradition, in which business confidence, as the psychological factor, and the multiplier process, through which the impact of this confidence is diffused, occupy key positions.
The Wicksell-type theories
In the 1920s and 1930s, immanent criticism of neoclassical economics, as composed of the theory of relative prices and the quantity theory of money, gave rise to an array of monetary economics deeply influenced by Wicksell (1898). Wicksell himself endorsed the theory of relative prices in the form of Walrasian general equilibrium theory-cum-Böhm-Bawerk’s capital theory, while he put forward the theory of the cumulative process of capital formation as an alternative to the quantity theory.
Accepting the essential points made by Wicksell, the interwar economists, including Lindahl and Myrdal (the Stockholm school) as well as Mises and Hayek (the Austria school), put forward their own brand of monetary economics, going beyond Wicksell and criticizing the neoclassical dichotomy per se (see Hirai 2008, chap.º2). Let us call them “Wicksell’s influences.” Robertson, Keynes, and Hawtrey were the “Wicksell-type theorists” in interwar Cambridge.
Lines of thinking departing from the Marshallian tradition were initiated in Cambridge by Robertson (1926) under the strong influence of Aftalion and Cassel. In his book, Robertson incorporated the relation between savings, credit creation, and capital accumulation into the nonmonetary argument developed in Robertson (1915). Although Robertson was not directly influenced by Wicksell, his theory shares some of the same essential components, and indeed Myrdal and Hayek evaluated Robertson highly on the grounds that he was dealing with similar problems using similar methods. It would not, therefore, be inappropriate to view Robertson’s theory as a sort of Wicksell-type theory.
Keynes was greatly influenced by Robertson, and it was very much thanks to this influence that he was able to evolve from his position in A Tract on Monetary Reform to that of the Treatise. In his case, however, he was aware of his theory’s place among Wicksell’s influences . (Keynes is dealt with more below.)
It is also appropriate to consider Hawtrey in this context. Hawtrey developed his theory of the trade cycle in Good and Bad Trade (1913), which may be regarded as independent of both the Marshallian tradition and the Wicksell’s influences but closer to the latter than the former.
Robertson.Æ The title “Banking Policy and the Price Level” encapsulates Robertson’s main theme in that work: that a policy of credit creation on the part of the banking system enables entrepreneurs to purchase the real capital necessary to boost production, which in turn induces an increase in money supply and thus in the price level of consumables. Production increases in the following period. Although he adopts a variant of the quantity theory in this argument, Robertson denies, in substance, the classical dichotomy.
Unlike monetary theories such as those found in Hawtrey (1913) and Keynes (1923), and also unlike psychological theories such as that set forth in Pigou (1927), Robertson’s theory takes the interaction between real and monetary factors seriously. He divides the fluctuations in output into “appropriate” and “inappropriate” categorizations. Appropriate fluctuations, which are closely related to the technical and legal structure of the modern economy, follow a rhythmical pattern to be justified. Inappropriate fluctuations, which are constituted by the excess of actual fluctuations in the volume of output over what would be appropriate ones, occur due to the use of large and expensive equipment.
Monetary factors are related to the activities of the banking system. In order to increase the volume of output, real circulating capital is correspondingly required in advance. In order to purchase it, the need arises to procure the corresponding short lacking. When firms cannot arrive at the whole amount of short lacking required, it is the banking system that provides the balance. The consumption goods produced in the economy are either consumed by the public or used by the firms as circulating capital. When the firms’ demand for circulating capital is strong, the public must of necessity be deprived of consumption goods. This is facilitated by the banking system providing credit to the firms.
Robertson develops his theory in two cases: short lacking and short and long lacking (where investment goods play an important role), but for the sake of brevity, let us here focus on the latter case. (Lacking is Robertson’s somewhat awkward term for savings. )
In this case, the idea of the “demand and supply of short lacking” plays a central role. In accordance with the phases of the trade cycle, changes in the level of demand for and supply of short lacking follow different courses. How the economy deals with this difference depends on the behavior of the banking system. Here let us examine the upward phase.
In order to increase the volume of output, circulating capital needs to be purchased. For this, procurement of short lacking is indispensable. The supply of short lacking, however, is made by “hoarding.” Hoarding is a kind of lacking that occurs where money obtained by selling current output is saved without any guarantee of being applied to the creation of capital. The supply of short lacking does not increase flexibly enough to cope with a large and discontinuous increase in the demand for it.
Large excess demand for short lacking is met only with provision of credit by the banking system (see Robertson 1949, pp.º50 and 72). The consequence for the economy as a whole means forced saving.
When short lacking is thus met, the required circulating capital is procured, so that the volume of production increases. The amount of money then increases, and consequently the price level rises. Once this rise occurs, excess demand for short lacking persistently arises due to an increase in consumption or investment by firms, a lengthening of the period of production, a dis-hoarding by consumers, and a direct short lacking by firms. If the banking system goes on supplying money in order to cope with the excess demand, the price level goes on rising cumulatively.
If the banking system provides no credit, firms cannot procure the short lacking required; thus, circulating capital cannot be sufficiently obtained and the volume of production shows no sufficient increase. In this case, the amount of money does not increase, and the price level remains stable.
Thus, stability in the volume of production is placed in a trade-off relation with that in the price level. Such is the feature of the modern capitalistic economy revealed by Robertson’s theory.
Hawtrey.ÆHawtrey’s theory of the trade cycle is uniquely monetary in the sense that it sees the main cause of the business cycle in the behavior of banks. The banks attempt to adjust the amount of credit money to a level appropriate to their cash holdings. When they judge the ratio to be inappropriate, they raise (or lower) the rate of interest to decrease (or increase) the amount of credit money. This behavior is argued to induce industrial fluctuations, for it causes the fluctuations in cash (mainly payment to wages) to lag behind those in credit money (which makes up the purchasing power).
Now, let us consider the case in which the banks judge the ratio to be in excess and raise the (short-term) rate of interest to reduce the excess. Hawtrey goes on to explain two major aspects of the fluctuations thus induced. One is that the dealers reduce their orders to the manufacturers due to the increased cost of holding goods. The manufacturers, in turn, cut back their production. Both the dealers and the manufacturers then reduce their borrowings from the banks, which decreases the amount of credit money. The other aspect is that due to the decrease in the amount of credit money, the purchasing power directed toward goods decreases.
Due to the interaction between these two aspects, the economy spirals downward. During this course, wholesale prices, retail prices, and wages go on falling. Between interest rate, wages, and prices, it is the prices that fall most drastically, so that the rate of profits comes to diverge increasingly from the rate of interest. Although wages go on falling as unemployment develops, the fluctuations in wages are not as large as those in prices. Thus, the banks’ cash holdings come to be in excess relative to the amount of credit money, and the banks reduce the (short-term) rate of interest to increase the amount of credit money. The process of expansion then begins to develop, reversing the sequence just explained.
The development of Keynes’s economics
In the above section, we saw two strands in Cambridge economics concerning business cycle theories. Where does Keynes stand in all this? Finding the answer requires some careful work uncovering the evolution of Keynes’s thought between the Treatise and the General Theory.
Keynes published the Treatise in 1930 after seven years’ intellectual struggle following A Tract on Monetary Reform (1923).3 The Treatise has a Wicksellian influence. Keynes made great efforts to defend the Treatise from both internal and the external criticisms up until the end of 1932. He then began to change the line of his argument in a direction that was to lead up to the General Theory. This can be characterized as a zigzag process. Finally, I will describe the main features of the General Theory and draw comparison between the Treatise and the General Theory. This will then allow us to see the ultimate outcome of the two influences upon Keynes’s own theory of the business cycle.
The Treatise theory
The most significant feature of the Treatise theory might arguably be considered the coexistence of a Wicksellian theory and “Keynes’s own theory” (see Hirai 2008, chap.º5).
On the one hand, the Treatise has Wicksellian influences, for Keynes tries to construct his own theory of monetary economics, criticizing both Marshall’s quantity theory of money (see Keynes 1930, vol. 1, p.º205) and business cycle theory, which emphasizes the behavior of speculators while embracing Wicksell’s view.
The Wicksellian strand of thought in the Treatise can be discerned in the following points: the explanation of the fluctuations in price level in terms of a relative relation between the natural rate and money rate of interest, and explanation of the working of the economy based on this relation; the stress on a bank-rate policy for stabilizing the price level; and the acceptance of an equivalence between Wicksell’s three conditions for monetary equilibrium.4
The principal grounds on which Keynes himself regards his theory as belonging to the Wicksellian stream, however, lie rather in his adoption of the idea that the bank rate influences investment and saving. This idea is used to provide a mechanism to the effect that economic stability (stability of the price level and the volume of output) can be attained by means of interest rate policy.
On the other hand, it should be noted that the Treatise has “Keynes’s own theory” as well. This theory consists of two parts. The first 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 as well as the amount of profit are simultaneously determined.
(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 or as the demand price of capital goods. Profit is determined as a result.
The second part deals with the determination of variables between one period and the next:
(Mechanism 3) The behavior of entrepreneurs is such that, if they make a profit (loss) in the current period, they expand (contract) output in the next.
I will call this last relation the “TM supply function.” Stimulated by the profits (losses) realized in the two sectors, the firms behave in such a way as to expand (or contract) output in the next period. Given output thus determined, Mechanisms 1 and 2 function accordingly. “Keynes’s own theory” can thus be expressed as the dynamic process consisting of Mechanisms 1 and 2 working through Mechanism 3.
As a result of this process (which can be called short-period oscillations), the economy may or may not reach long-period equilibrium. Keynes argues that the duty of the monetary authority is to achieve long-period equilibrium by means of bank-rate policy.
In long-period equilibrium, profits are supposed to become zero, investment is supposed to become equal to saving, and the price level is supposed to become stable. Keynes remarks that “every change towards a new equilibrium price level is initiated by a departure of profits from zero” (Keynes 1930, vol. 1, p.º142).
This interpretation sees the Treatise as articulating a dynamic process that includes the determination of both the price levels and the volumes of output.
It should be noted that the Treatise theory thus interpreted involves three kinds of “dualities,” offering alternative and not always compatible analyses. These are the duality of the theory concerning the price level of consumption goods, a theory dependent on either (a) earnings or (b) the rate of interest; the duality of the theory concerning the price level of investment goods, either (c) the bearishness function theory or (d) the idea that prospective yields are discounted by the rate of interest; and finally, the duality of Wicksellian theory and Keynes’s own theory.
The third duality rests on the other two. The Wicksellian theory, which is mainly used in the argument of economic policy by means of the bank rate in the second volume of the Treatise, adopts b and d, while Keynes’s own theory adopts a and c.
The process of development after the Treatise
What characterizes the period up to mid-1932 was that Keynes maintained and improved upon “Keynes’s own theory,” disregarding the Wicksellian theory.5 It is important to appreciate how Keynes dealt with the relation between profits and the volume of output. In the Treatise, the “TM supply function” was stressed as expressing the dynamic mechanism. Keynes adhered to the function (see his letter to Joan Robinson, CWK 13, p.º380) in the face of considerable criticism. This stance emerges clearly from the manuscript entitled “The Monetary Theory of Production” drafted in mid-1932 (CWK 13, pp.º381–396).
Toward the end of 1932 (see Hirai 2004; 2008, chap.º7), Keynes abandoned the TM supply function, albeit with some hesitation, and put forward a new formula of a system of commodity markets in the manuscript entitled “The Parameters of a Monetary Economy” (CWK 13, pp.º397–405). Here the TM supply function virtually disappeared from the analysis of commodity markets, as a result of which there emerged the model consisting of a system of simultaneous equations based on the equality of investment and saving in which profits do not relate to the determination of prices and output, thereby departing sharply from “Keynes’s own theory.” This was a turning point toward the General Theory.
In the three manuscripts of 1933 (see Hirai 2008, chap.º8), two functions are emphasized—the “pseudo-TM supply function” (see Hirai 2008, pp.º104–105 and 107) and the “pseudo-TM supply function mk2.” I use the prefix “pseudo” because the functions, in substance, differ from the (original) TM supply function, although Keynes tends to regard these two functions as continuous with the TM supply function.
The three manuscripts constitute the origins of the General Theory’s chapter 3. They conceivably discuss both an equilibrium condition for the level of employment and its stability condition, although no concept corresponding to the General Theory’s aggregate supply function appears. In the “First Manuscript” (CWK 13, pp.º62–66), Keynes first put forward a system for determining the level of employment. This was, indeed, a breakthrough.
Keynes’s way of formulating the system, however, suffers from certain ambiguities. The central problem in examining the three manuscripts is how consistently one can explain the “Second Manuscript” (CWK 29, pp.º63, 66–73, 87–92, 95–102), which accepts the first postulate of the classical economics, using the concept of the accounting period, and yet succeeds the pseudo-TM supply function mk2. I interpret the pseudo-TM supply function mk2 as describing the stability condition for the equilibrium level of employment. But the argument in terms of the function remains unclear. Moreover, the argument in the “Third Manuscript” (CWK 29, p.º76–101; CWK 13, pp.º421–422), which emphasizes “effective demand,”6 is somewhat confusing, for it is still developed in terms of the sale proceeds and variable cost. The arguments recognizable in the Second and Third Manuscripts were to disappear thereafter. This has very much to do with the fact that the role played by profit was to change drastically. These ambiguities show how Keynes struggled to work out a new employment theory.
In the “First Undated Manuscript” (CWK 29, pp.º102–111; see Hirai 2008, chap.º9), Keynes appears to have first established the fundamental psychological law and the multiplier theory, while in the “Second Undated Manuscript” (CWK 29, pp.º112–120), he first put forward an investment theory close to that of the General Theory (both were written either toward the end of 1933 or in the first half of 1934).
Thus, the end of 1933 saw Keynes making a great advance toward the General Theory. Piecing together the arguments made in the fragments concerned, we can confirm that Keynes had by then established the following points: the system of determining the level of employment, the consumption function, the fundamental psychological law, the liquidity preference theory,7 the marginal efficiency of capital8 (albeit not yet that of the General Theory), and the multiplier theory.
In the manuscript “The General Theory” (CWK 13, pp.º423–456; see Hirai 2008, chap.º10, “The Eve of the General Theory”) of the spring of 1934, and also in “The Summer Manuscript” (CWK 13, pp.º471–484) of the same year, which comprises the revised versions of chapters 8 and 9 of “The General Theory,” Keynes puts forward almost the same theoretical framework as that of the General Theory in the area of consumption and investment theories. However, it should be noted that some theoretical ambiguity still remains in the concept of effective demand and the theory of determination of the level of employment. Thus, we might call this period the “Eve of the General Theory.”
The proofing process of the General Theory (see Hirai 2008, chaps. 11–12) continued from the summer of 1934 up to publication. Here I set myself two objectives, namely, to determine (a) how Keynes went through the proofing process and (b) what the main features are.
With regard to objective a, the following points emerge: Galley1(I) (chaps. 1–19, between early December 1934 and mid-January 1935) represents the most considerably revised work on the topics covered in the table of contents shared by three galleys (see CWK 13, pp.º525–526), while Galley2 (chaps. 1–15, between January and April 1935) and Galley3 (chaps. 2–6, June 1935) represent stylistic revisions of Galley1(I); all the chapters, except for chaps. 4, 5, 12, and 13, were completed in Galley1(I) both in contents and at the stylistic level. (There is also Galley1(II), chaps. 20–25, from March 1935 on.) The largest change after Galley1(III) (chaps. 26–28, June–July 1935) occurred in “The Great Revision” (chaps. 3 and 6–9, in August–October; see Hirai 2008, pp.º159–171). The definitions of some fundamental concepts changed due to alterations both in the definition of “user cost” and in its treatment; in the Michaelmas lectures (November 18), the precautionary motive was argued to be dependent on the rate of interest.
In the case of objective b, I focused on what kind of changes or difficulties can be detected in the latter half of Keynes’s process of development. Here we find two major changes worth noting, one regarding the “employment function,” the other concerning fundamental concepts.
In “The General Theory,” the Summer Manuscript, and the Pre-first Proof Typescript (summer 1934), the employment function was used as both supply and equilibrium concept. This duality disappears in Galley1(I) up to and including Galley3. In the General Theory, however, the duality reappears and overshadows its theoretical structure (see Hirai 2008, pp.º186–191).
Some fundamental concepts in Galley1(I) are related to those in the General Theory as follows:
The difference in effective demand, investment, and prime cost in definition depends on whether they include user cost (Galley1(I)) or not (the General Theory). From Galley1(I) to Galley III, for example, Keynes argued that income as realized value consistently differed from effective demand as expected value by user cost, emphasizing that effective demand inclusive of user cost matters in determining the level of employment. In the General Theory, however, this idea has disappeared.
The definitions of income, profit, and saving are the same.
The equation U2 = U1 - B is vital in grasping the relation (where U2 is the user cost in the General Theory; U1 is user cost in Galley 1(I); and B is cost of the maintenance and improvement of the initial capital equipment).
The main thread runs as follows: In the Treatise, the “TM supply function” is stressed as expressing the dynamic mechanism. Keynes adhered to this function after the Treatise, in spite of much criticism. Toward the end of 1932, he abandoned it with some hesitation, and from the First Manuscript of 1933 on put forward a new theory of employment, which led to the General Theory.
The General Theory
Through the above zigzag process we have seen, in February 1936 Keynes finally published the General Theory, destined to spark off the Keynesian Revolution. At this point, we must take a look at the essential points of the General Theory (see Hirai 2008, chap.º13).
There are three central themes we can identify as running through the General Theory: contrasting potentialities, monetary economics, and underemployment equilibrium.
The General Theory sees the market economy as possessing two contrasting sets of potentialities: stability, certainty, and simplicity, on the one hand; and instability, uncertainty, and complexity, on the other.
Keynes argues that the market economy is equipped with several built-in stabilizers and thus has an inherent tendency to converge to equilibrium. It does not, however, reach an optimum (or full-employment) level but, rather, stays at an underemployment level. Based on this “optimistic” vision, he constructs a theoretical model in which the level of employment is determined where the aggregate demand function intersects the aggregate supply function, making use of such concepts as the consumption function, the marginal efficiency of capital, and the liquidity preference theory. The model is constructed in a simple and straightforward way, providing the foundations upon which Keynes presents his economic policy proposals for attaining full employment.
At the same time, however, Keynes repeatedly argues that the stability to which the market economy tends cannot set in unless some conditions are met; failing these, the market economy is doomed to instability. In this respect, we are faced with a structure built on fragile foundations, reflecting the uncertainty and complexity to be observed in the market economy.
Keynes argues that the working of the market economy depends on various psychological factors, such as short-term expectations, long-term expectations (the marginal efficiency of capital and the nature of the stock market), liquidity preference, and user cost.
The other element making the market economy unstable is its vulnerability to large changes in some exogenous variables. Here Keynes’s concern is above all about any substantial changes in the quantity of money or in money wages.
Keynes seems confident that the possibility that the market economy will be undermined through the falling away of the above-mentioned conditions is remote and that an economy stuck in underemployment equilibrium could be cured with policies such as public works programs and low interest-rate policies.
Keynes puts forward his theory of underemployment equilibrium as monetary economics, as distinct from the real economics to which “classical economics” belongs. He argues that the monetary economy in which we live can be analyzed only within a framework of monetary economics.
Keynes’s fundamental idea is expounded in chapter 21, where he presents two new ways of dividing up economics. One is a division “between the theory of the individual industry or firm and of the rewards and the distribution between different uses of a given quantity of resources on the one hand, and the theory of output and employment as a whole on the other hand” (1936, p.º293). The other is a division “between the theory of stationary equilibrium and the theory of shifting equilibrium—meaning by the latter the theory of a system in which changing views about the future are capable of influencing the present situation” (ibid.).
In both cases, the criterion of division hinges on money. Indeed, when dealing with the determination of the level of output and employment as a whole in the real world, we must consider the role played by money. This is what Keynes means by monetary economics. Keynes argues that monetary economics in his sense remains “a theory of value and distribution, not a separate ‘theory of money’” (1936, p.º294).
The above considerations show that Keynes’s monetary economics aims at analyzing an economy in which money plays an essential role. It is not surprising, then, that Keynes allocates a lot of room to discussion of the rate of interest (chaps. 13, 14, 15, 17, 23, and 24). Apart from his liquidity preference theory, Keynes, in chapter 17, develops a theory of “own-rates of interest,” accounting for the way in which the behavior of money becomes an obstacle to full employment.
The central message of the General Theory is that left to itself, the market economy will remain in underemployment equilibrium (see 1936, pp.º249–250). Underemployment equilibrium has four features: involuntary unemployment, equilibrium, stability, and fluctuation.
We may, indeed, go as far as saying that the theoretical structure of the General Theory can be characterized as the monetary economics of underemployment equilibrium.
Comparison
The General Theory would not have appeared but for the Treatise, and yet it is an achievement independent of it.
First, in the Treatise, the TM supply function plays an essential role in the dynamic movement of the system; the General Theory addresses the question of how the level of employment is determined.9
Second, the Treatise provides no theoretical account of investment and consumption; in the General Theory, theories of both investment and consumption are put forward and play important roles in determining the volume of employment.
Third, although we can find some continuity between the concepts of bearishness and liquidity preference (respectively , the view that the banking system and the public, with their psychological inclinations, behave interactively; and classification of the motives for holding money), the role assigned to money differs considerably.
Fourth, in the Treatise, the rate of interest is a policy variable through which the banking system is supposed to adjust the supply of money to the public’s bearishness; in the General Theory, the rate of interest is supposed to be adjusted in such a way that the supply of money, which is a policy variable, meets the liquidity preference.
Fifth, the Treatise and the General Theory nevertheless have the following points in common: both belong to the field of monetary economics and were pitched against neoclassical orthodoxy, prices and output are treated as endogenous variables, and the importance of both monetary and fiscal policies is stressed.
Conclusion
This chapter began by identifying two strands in Cambridge with regard to business cycle theories in order to position Keynes correctly. It should be remembered that Keynes was ranked as highly as Pigou, Robertson, and Hawtrey before the General Theory, although he was the most famous of the four outside the academic world.
I then put forward my interpretation, based on the Keynes Papers, of how Keynes developed his theory from the Treatise to the General Theory.
From the late nineteenth century on, economists increasingly concentrated their attention on the phenomenon of exchange in the market. They were concerned with the problem of how resources are exchanged through the price mechanism, assuming full employment and accepting Say’s law.
A new approach to economics was initiated by Wicksell at the turn of the century. He put forward the theory of cumulative process as an alternative to the quantity theory. A number of economists of diverse intellectual backgrounds emerged in the interwar period to follow up Wicksell’s lead. They were united in their desire to construct a new monetary economics, criticizing the quantity theory, the classical dichotomy, and Say’s law. Keynes was one of these economists.
Keynes, however, dispensed with the Wicksellian influence soon after the Treatise and ultimately arrived at the General Theory. In doing so, he returned to his own roots in Cambridge cycle theory. Here we have a monetary economics that demonstrates underemployment equilibrium. It was an independent achievement, and it generated the Keynesian Revolution.
Having said that, I think that Keynes’s methodology, as well as his economics, is of extreme importance at present. It was never as ends in themselves that Keynes constructed his theoretical models but always for the purpose of analyzing and diagnosing the economic problems facing the real economy and proposing polices to deal with them. While observing the real economy, Keynes sought out useful analytical tools or concepts or, when necessary, contrived new tools or concepts applying his own intuition and introspection. These he would then set about elaborating with his powers of speculation.
Reference here to “theoretical models” does not necessarily imply only mathematical models. Keynes perceived the complexity of the economy, which could not be dealt with using only mathematical models. Thus, as argued above, Keynes’s theoretical models are built in such a way that the market economy is equipped with two contrasting potentialities. To this end, Keynes took great pains in constructing his models, showing how the level of employment is determined on the one hand, while at the same time giving due weight to the aspect of uncertainty on the other.
The last thirty years have seen the emergence of the “New Classical” school (see Lucas 1975; Kydland and Prescott 1982) as the dominant orthodoxy in macroeconomics. The exponents of this school explicitly and utterly rejected Keynes’s economics, his way of thinking, and his social philosophy, arguing the superiority of their models as being constructed rigorously and mathematically from microeconomic foundations. But Keynes would surely have been highly critical of them as being based on the “representative agent,” assumed to maximize his/her utility and to be able to form “rational expectations.” By assuming this type of homo oeconomicus, the New Classical economists construct their models with “pseudo-mathematical methods” (Keynes 1936, p.º297), which do not go a long way toward analyzing the real economy. I believe that we should welcome the return of Keynes in this sense, rather than blindly following the New Classical school.
Notes
1. Keynes Papers, King’s College Archive Centre, Cambridge.
2. This section is based on Hirai (2003, sec. 2 of chap.º2).
3. Marshall’s economic system is examined in Hirai (2007a; 2008, chap.º4).
4. Wicksell’s three conditions for monetary equilibrium are equality of the natural rate and the money rate of interest, equality of investment and saving, and price level stability.
5. See Hirai (2008, chaps. 6–12). For studies focusing on the same theme, see Asano 1987; Dimand 1987; Amadeo 1989. Rymes (1989) is also an important source of information.
6. It should be noted that the concept of “effective demand” was to undergo several changes before reaching the General Theory.
7. Keynes’s liquidity preference theory is first developed in “The Monetary Theory of Production.”
8. Keynes’s theory of the marginal efficiency of capital is first developed in “The Second Undated Manuscript.”
