''The General Theory of Employment, Interest and Money''

Stickiness of wages in money terms

Say’s Law depends on the operation of a market economy. If there is unemployment (and if there are no distortions preventing the employment market from adjusting to it) then there will be workers willing to offer their labour at less than the current wage levels, leading to downward pressure on wages and increased offers of jobs.

The classics held that full employment was the equilibrium condition of an undistorted labour market, but they and Keynes agreed in the existence of distortions impeding transition to equilibrium.^[3] The classical position had generally been to view the distortions as the culprit and to argue that their removal was the main tool for eliminating unemployment. Keynes on the other hand viewed the market distortions as part of the economic fabric and advocated different policy measures which (as a separate consideration) had social consequences which he personally found congenial and which he expected his readers to see in the same light.

The distortions which have prevented wage levels from adapting downwards have lain in employment contracts being expressed in monetary terms; in various forms of legislation such as the minimum wage and in state-supplied benefits; in the unwillingness of workers to accept reductions in their income; and in their ability through unionisation to resist the market forces exerting downward pressure on them.

Keynes accepted the classical relation between wages and the marginal productivity of labour, referring to it on p5^[4] as the ‘first postulate of classical economics’ and summarising it as saying that ‘The wage is equal to the marginal product of labour’.

The first postulate can be expressed in the equation y' (N ) = W / p, where y (N ) is the real output when employment is N, and W  and p  are the wage rate and price rate in money terms. A system can be analysed on the assumption that W  is fixed (i.e. that wages are fixed in money terms) or that W / p is fixed (i.e. that they are fixed in real terms) or that N  is fixed (e.g. if wages adapt to ensure full employment). All three assumptions had at times been made by classical economists, but under the assumption of wages fixed in money terms the ‘first postulate’ becomes an equation in two variables (N  and p ), and the consequences of this had not been taken into account by the classical school.

Keynes proposed a ‘second postulate of classical economics’ asserting that the wage is equal to the marginal disutility of labour. This is an instance of wages being fixed in real terms. He attributes the second postulate to the classics subject to the qualification that unemployment may result from wages being fixed by legislation, collective bargaining, or ‘mere human obstinacy’  (p6), none of which can be identified with the marginal disutility of labour and all of which are likely to fix wages in money terms.

Outline of Keynes’s theory

Keynes’s economic theory is based on the interaction between demands for saving, investment, and liquidity (i.e. money). Saving and investment are necessarily equal, but different factors influence decisions concerning them. The desire to save, in Keynes’s analysis, is mostly a function of income: the richer people are, the more wealth they will seek to put aside. The profitability of investment, on the other hand, is determined by the relation between the return available to capital and the interest rate. The economy needs to find its way to an equilibrium in which no more money is being saved than will be invested, and this can be accomplished by contraction of income and a consequent reduction in the level of employment.

In the classical scheme it is the interest rate rather than income which adjusts to maintain equilibrium between saving and investment; but Keynes asserts that the rate of interest already performs another function in the economy, that of equating demand and supply of money, and that it cannot adjust to maintain two separate equilibria. In his view it is the monetary role which wins out. This is why Keynes’s theory is a theory of money as much as of employment: the monetary economy of interest and liquidity interacts with the real economy of production, investment and consumption.

The choice of units

Keynes sought to allow for the lack of downwards flexibility of wages by constructing an economic model in which the money supply and wage rates were externally determined (the latter in money terms), and in which the main variables were fixed by the equilibrium conditions of various markets in the presence of these facts.

Many of the quantities of interest, such as income and consumption, are monetary. Keynes often expresses such quantities in wage units (Chapter 4): to be precise, a value in wage units is equal to its price in money terms divided by W, the wage (in money units) per man-hour of labour. Keynes generally writes a subscript w  on quantities expressed in wage units, but in this account we omit the w. When, occasionally, we use real terms for a value which Keynes expresses in wage units we write it in lower case (e.g. y  rather than Y ).

As a result of Keynes’s choice of units, the assumption of sticky wages, though important to the argument, is largely invisible in the reasoning. If we want to know how a change in the wage rate would influence the economy, Keynes tells us on p266 that the effect is the same as that of an opposite change in the money supply.

The identity of saving and investment

The relationship between saving and investment, and the factors influencing their demands, play an important role in Keynes’s model. Saving and investment are considered to be necessarily equal for reasons set out in Chapter 6 which looks at economic aggregates from the viewpoint of manufacturers. The discussion is intricate, considering matters such as the depreciation of machinery, but is summarised on p63:

Provided it is agreed that income is equal to the value of current output, that current investment is equal to the value of that part of current output which is not consumed, and that saving is equal to the excess of income over consumption... the equality of saving and investment necessarily follows.

This statement incorporates Keynes’s definition of saving, which is the normal one.

The marginal propensity to consume and the multiplier

The marginal propensity to consume, C '(Y ), is the gradient of the purple curve, and the marginal propensity to save S '(Y ) is equal to 1 – C '(Y ). Keynes states as a ‘fundamental psychological law’ (p96) that the marginal propensity to consume will be positive and less than unity.

Chapter 10 introduces the famous ‘multiplier’ through an example: if the marginal propensity to consume is 90%, then ‘the multiplier k is 10; and the total employment caused by (e.g.) increased public works will be ten times the employment caused by the public works themselves’ (pp116f). Formally Keynes writes the multiplier as k = 1/S '(Y ). It follows from his ‘fundamental psychological law’ that k  will be greater than 1.

Keynes’s account is not intelligible until his economic system has been fully set out (see below). In Chapter 10 he describes his multiplier as being related to the one introduced by R. F. Kahn in 1931,^[5] but the two have little in common. The mechanism of Kahn’s mutliplier lies in an infinite series of transactions, each conceived of as creating employment: if you spend a certain amount of money, then the recipient will spend a proportion of what he or she receives, the second recipient will spend a further proportion again, and so forth. Enough meaning can be extracted from Keynes’s account of his own mechanism (in the second para of p117) to see that it makes no reference to infinite series. It also differs from Kahn’s multiplier in being attached to investment rather than to spending in general, and in having a value determined by the marginal propensity to consume rather than by the marginal propensity to spend.

The rate of investment

Keynes’s schedule of the marginal efficiency of capital

Book IV discusses the inducement to invest, with the key ideas being presented in Chapter 11. The ‘marginal efficiency of capital’ is defined as the annual revenue which will be yielded by an extra increment of capital as a proportion of its cost. The ‘schedule of the marginal efficiency of capital’ is the function which, for any rate of interest r, gives us the level of investment which will take place if all opportunities are accepted whose return is at least r. By construction this depends on r  alone and is a decreasing function of its argument; it is illustrated in the diagram, and we shall write it as I_s(r ).

This schedule is a characteristic of the current industrial process which Irving Fisher described as representing the ‘investment opportunity side of interest theory’;^[6] and in fact the condition that it should equal S (Y,r ) is the equation which determines the interest rate from income in classical theory. Keynes is seeking to reverse the direction of causality (and omitting r  as an argument to S () ).

He interprets the schedule as expressing the demand for investment at any given value of r , giving it an alternative name: “We shall call this the investment demand-schedule...” (p136). He also refers to it as the ‘demand curve for capital’ (p178). However it has many of the properties of a supply curve (for instance not being constrained by income). It is analogous to the curve giving the amount of gold which can be extracted from the soil at a price less than p. Nonetheless the schedule of the marginal efficiency of capital was a demand function in Keynes’s eyes.

For fixed industrial conditions, we conclude that ‘the amount of investment... depends on the rate of interest’ as John Hicks put it in “Mr Keynes and the Classics”.

Interest and liquidity preference

Keynes proposes two theories of liquidity preference (i.e. the demand for money): the first as a theory of interest in Chapter 13 and the second as a correction in Chapter 15. His arguments offer ample scope for criticism, but his final conclusion is that liquidity preference is a function mainly of income and the interest rate. The influence of income (which really represents a composite of income and wealth) is common ground with the classical tradition and is embodied in the Quantity Theory; the influence of interest had also been noted earlier, in particular by Frederick Lavington (see Hicks’s “Mr. Keynes and the Classics”). Thus Keynes’s final conclusion may be acceptable to people who question the arguments along the way. However he shows a persistent tendency to think in terms of the Chapter 13 theory while nominally accepting the Chapter 15 correction.

Chapter 13 presents the first theory in rather metaphysical terms. Keynes argues that:

It should be obvious that the rate of interest cannot be a return to saving or waiting as such. For if a man hoards his savings in cash, he earns no interest, though he saves just as much as before. On the contrary, the mere definition of the rate of interest tells us in so many words that the rate of interest is the reward for parting with liquidity for a specified period.^[7]

To which Jacob Viner retorted that:

By analogous reasoning he could deny that wages are the reward for labor, or that profit is the reward for risk-taking, because labor is sometimes done without anticipation or realization of a return, and men who assume financial risks have been known to incur losses as a result instead of profits.^[8]

Keynes goes on to claim that the demand for money is a function of the interest rate alone on the grounds that:

The rate of interest is... the “price” which equilibrates the desire to hold wealth in the form of cash with the available quantity of cash...^[9]

which, as Frank Knight commented,^[10] seems to assume that demand is simply an inverse function of price. The upshot from these reasonings is that:

Liquidity-preference is a potentiality or functional tendency, which fixes the quantity of money which the public will hold when the rate of interest is given; so that if r  is the rate of interest, M  the quantity of money and L  the function of liquidity-preference, we have M  = L(r ). This is where, and how, the quantity of money enters into the economic scheme.^[11]

And specifically it determines the rate of interest, which therefore cannot be determined by the traditional factors of ‘productivity and thrift’.

Chapter 15 looks in more detail at the three motives Keynes ascribes for the holding of money: the ‘transactions motive’, the ‘precautionary motive’, and the ‘speculative motive’. He considers that demand arising from the first two motives ‘mainly depends on the level of income’ (p199), while the interest rate is ‘likely to be a minor factor’ (p196).

Keynes treats the speculative demand for money as a function of r  alone without justifying its independence of income. He says that...

what matters is not the absolute  level of r  but the degree of its divergence from what is considered a fairly safe  level...^[12]

but gives reasons to suppose that demand will nonetheless tend to decrease as r  increases. He thus writes liquidity preference in the form L₁(Y ) + L₂(r ) where L₁ is the sum of transaction and precautionary demands and L₂ measures speculative demand. The structure of Keynes’s expression plays no part in his subsequent theory, so it does no harm to follow Hicks by writing liquidity preference simply as L (Y,r ).

The money supply is treated as fixed and exogenous. This was the normal assumption made by economists at the time and is almost explicit on p200. At no point does Keynes mention ‘repercussions’ on the supply of money or credit arising from changes elsewhere in the economy.

Keynes does not put a subscript ‘w ’ on L or M, implying that we should think of them in money terms. This suggestion is reinforced by his wording on p172 where he says “Unless we measure liquidity-preference in terms of wage-units (which is convenient in some contexts)... ”. But seventy pages later there is a fairly clear statement that liquidity preference and the quantity of money are indeed “measured in terms of wage-units” (p246).

Keynes’s economic model

In Chapter 14 Keynes contrasts the classical theory of interest with his own, and in making the comparison he shows how his system can be applied to explain all the principal economic unknowns from the facts he takes as given. The two topics can be treated together because they are different ways of analysing the same equation.

Keynes’s presentation is informal. To make it more precise we will identify a set of 4 variables – saving, investment, the rate of interest, and the national income – and a parallel set of 4 equations which jointly determine them. The graph illustrates the reasoning. The red S  lines are shown as increasing functions of r  in obedience to classical theory; for Keynes they should be horizontal.

Graphical representation of Keynes’s economic model, based on his own diagram at p180 of the General Theory.

The first equation asserts that the reigning rate of interest r̂  is determined from the amount of money in circulation M̂  through the liquidity preference function and the assumption that L (r̂ ) = M̂.

The second equation fixes the level of investment Î  given the rate of interest through the schedule of the marginal efficiency of capital as I_s(r̂ ).

The third equation tells us that saving is equal to investment: S (Y ) = Î. The final equation tells us that the income Ŷ  is the value of Y  corresponding to the implied level of saving.

All this makes a satisfying theoretical system.

Three comments can be made concerning the argument. Firstly, no use is made of the ‘first postulate of classical economics’, which can be called on later to set the price level. Secondly, Hicks (in “Mr Keynes and the classics”) presents his version of Keynes’s system with a single variable representing both saving and investment; so his exposition has three equations in three unknowns.

And finally, since Keynes’s discussion takes place in Chapter 14, it precedes the modification which makes liquidity preference depend on income as well as on the rate of interest. Once this modification has been made the unknowns can no longer be recovered sequentially.

Keynesian economic intervention

The state of the economy, according to Keynes, is determined by four parameters: the money supply, the demand functions for consumption (or equivalently for saving) and for liquidity, and the schedule of the marginal efficiency of capital determined by ‘the existing quantity of equipment‘ and ‘the state of long-term expectation’ (p246).

Adjusting the money supply is the domain of monetary policy. The effect of a change in the quantity of money is considered at p298. The change is effected in the first place in money units. According to Keynes’s account on p295, wages will not change if there is any unemployment, with the result that the money supply will change to the same extent in wage units.

We can then analyse its effect from the diagram, in which we see that an increase in M̂  shifts r̂  to the left, pushing Î  upwards and leading to an increase in total income (and employment) whose size depends on the gradients of all 3 demand functions. If we look at the change in income as a function of the upwards shift of the schedule of the marginal efficiency of capital (blue curve), we see that as the level of investment is increased by one unit, the income must adjust so that the level of saving (red curve) is one unit greater, and hence the increase in income must be 1 / S' (Y ) units, i.e. k units. This is the explanation of Keynes’s multiplier.

It does not necessarily follow that individual decisions to invest will have a similar effect, since decisions to invest above the level suggested by the schedule of the marginal efficiency of capital are not the same thing as an increase in the schedule.

The equations of Keynesian and classical economics

Keynes’s initial statement of his economic model (in Chapter 14) is based on his Chapter 13 theory of liquidity preference. His restatement in Chapter 18 doesn’t take full account of his Chapter 15 revision, treating it as a source of ‘repercussions’ rather than as an integral component. It was left to John Hicks to give a satisfactory presentation.^[13] Equilibrium between supply and demand of money depends on two variables – interest rate and income – and these are the same two variables as are related by the equation between the propensity to save and the schedule of the marginal efficiency of capital. It follows that neither equation can be solved in isolation and that they need to be considered simultaneously.

The ‘first postulate’ of classical economics was also accepted as valid by Keynes, though not used in the first four books of the General Theory. The Keynesian system can thus be represented by three equations in three variables as shown below, roughly following Hicks. Three analogous equations can be given for classical economics. As presented below they are in forms given by Keynes himself (the practice of writing r  as an argument to V  derives from his Treatise on money ^[14]).

Here y  is written as a function of N, the number of workers employed; p  is the price (in money terms) of a unit of real output; V (r ) is the velocity of money; and W is the wage rate in money terms. N, p  and r  are the 3 variables we need to recover. In the Keynesian system income is measured in wage units and is therefore not a function of the level of employment alone since it will also vary with prices. The first postulate is written in a form which is only admissible if one allows prices to be represented by a single variable. Strictly it should be modified to take account of the distinction between marginal wage cost and marginal prime cost.^[15]

The classics took the second equation as determining the rate of interest, the third as determining the price level, and the first as determining employment. Keynes believed (perhaps incorrectly: see wage unit) that the last two equations could be solved together for Y  and r , which is not possible in the classical system. He accordingly concentrated on these two equations, treating income as ‘almost the same thing’ as employment on p247 but looking more closely at the relation between them in Chapter 20. Here we see the benefit he has gained by simplifying the form of the consumption function. If he had written it (slightly more accurately) as C (Y, p /W ), then he would have needed to bring in the first equation to get a solution.

The classical theory of employment for wages fixed in money terms. (The three curves have different vertical scales.)

If we wish to examine the classical system our task is made easier if we assume that the effect of the interest rate on the velocity of circulation is small enough to be ignored. This allows us to treat V  as constant and solve the first and third equations (the ‘first postulate’ and the quantity theory) together, leaving the second equation to determine the interest rate from the result. We then find that the level of employment is given by the formula

.

The graph shows the numerator and denominator of the left-hand side as blue and green curves; their ratio – the pink curve – will be a decreasing function of N  even if we don’t assume diminishing marginal returns. The level of employment N̂  is given by the horizontal position at which the pink curve has a value of , and this is evidently a decreasing function of W. This diagram cannot have been in Keynes’s mind when he wrote (Chapter 2, p12) that “one would have expected the classical school to argue” that if money-wages change then the effect on prices would leave “the level of unemployment practically the same as before”.

Chapter 3: The principle of effective demand

The theoretical system we have described is developed over chapters 4–18, and is anticipated by a chapter which – amidst a proliferation of symbols – interprets Keynesian unemployment in terms of ‘aggregate demand’.

The aggregate supply Z  is the total value of output when N  workers are employed, written functionally as φ(N ). The aggregate demand D  is manufacturers’ expected proceeds, written as f (N ). In equilibrium Z  = D. D  can be decomposed as D₁ + D₂ where D₁ is the propensity to consume, which may be written C (Y ) or χ(N ). D₂ is explained as ‘the volume of investment’, and the equilibrium condition determining the level of employment is that D₁ + D₂ should equal Z  as functions of N. D₂ can be identified with I_s(r ).

The meaning of this is that in equilibrium the total demand for goods must equal total income. Total demand for goods is the sum of demand for consumption goods and demand for investment goods. Hence Y = C (Y ) + S (Y ) = C (Y ) + I_s(r ); and this equation determines a unique value of Y  given r.

Samuelson’s Keynesian cross is a graphical representation of the Chapter 3 argument.^[16]

Chapter 5: Expectation as determining output and employment

Chapter 5 makes some common-sense observations on the role of expectation in economics. Short-term expectations govern the level of production chosen by an entrepreneur while long-term expectations govern decisions to adjust the level of capitalisation. Keynes describes the process by which the level of employment adapts to a change in long-term expectations and remarks that:

the level of employment at any time depends... not merely on the existing state of expectation but on the states of expectation which have existed over a certain past period. Nevertheless past expectations, which have not yet worked themselves out, are embodied in to-day’s capital equipment... and only influence [the entrepreneur’s] decisions in so far as they are so embodied...^[17]

Chapter 11: Expectation as influencing the schedule of the marginal efficiency of capital

The main role of expectation in Keynes’s theory lies in the schedule of the marginal efficiency of capital which, as we have seen, is defined in Chapter 11 in terms of expected  returns. Keynes differs here from Fisher^[18] whom he largely follows, but who defined the ‘rate of return over cost’ in terms of an actual revenue stream rather than its expectation. Keynes was certainly correct in this, and the point has a particular significance in his theory.

Chapters 14, 18: The schedule of the marginal efficiency of capital as influencing employment

Keynes differed from his classical predecessors in assigning a role to the schedule of the marginal efficiency of capital in determining the level of employment. The classics regarded this, as they regarded other supply curves, as simply influencing the allocation of resources. But here too, even from a classical point of view, Keynes can be seen to be correct, at least assuming fixed wage rates. The classics (who had written little about unemployment) had not taken account of Lavington’s observation^[19] that interest is the opportunity cost of holding money. They should have reasoned that an increase in the marginal efficiency of capital leads to an increase in the interest rate which increases the cost of holding money and therefore leads to faster circulation; that this has the same effect on prices as an increase in the money supply; and that higher prices at a given wage level lead to higher employment. They might have thought this to be a weak effect, but they could not with consistency have denied its existence.

On the other hand Keynes may have considered the effect to be larger than it really is. He defines his multiplier k  as 1 / S '(Y ), and this gives the response of income to a change in the schedule of the marginal efficiency of capital for a given interest rate. However the interest rate is not given, and Keynes should have looked at the income component of the joint response of income and interest rate to a change in the schedule using his Chapter 15 model of liquidity preference.

In his ‘restatement’ of Chapter 18 (pp248f) he again treats the interest rate as given and discusses the effect of ‘an increase in the rate of investment’, concluding that ‘the ratio... between an increment of investment and the corresponding increment of aggregate income... is given by the multiplier’. He then mentions the ‘repercussions’ arising from changes in liquidity preference due to an increase in employment, thereby treating the role of interest in determining liquidity preference as primary and the role of income as a secondary correction separate from the main analysis. Hansen comments:

But this is wrong. His mistake follows from the fact that he often, perhaps generally, made the rate of interest depend exclusively on liquidity preference and the quantity of money.^[20]

Keynes compounds this error when he writes in his reply to Viner that...

...there is always a formula, more or less of this kind, relating the output of consumption-goods which it pays to produce to the output of investment-goods; and I have given attention to it in my book under the name of the multiplier.^[21]

This relates the multiplier to the ‘output of consumption-goods’ rather than the output of all  goods and ignores the ‘repercussions’. In fact a different multiplier κ can be defined to relate changes in income to changes in the schedule of the marginal efficiency of capital under the Chapter 15 model of liquidity preference. If the schedule of the marginal efficiency of capital increases by a small unit in the region of the equilibrium position, then income increases by κ units where:

Here 0 ≤ κ ≤ k. The lower bound is attained under a pre-Lavington classical model in which interest  has no effect on liquidity preference, and the upper bound under the Chapter 13 theory in which income  has no effect; in general the position of κ between the extremes will depend on the relative magnitudes of the two influences.

Chapter 12: Animal spirits

Chapter 12 discusses the psychology of speculation and enterprise.

Most, probably, of our decisions to do something positive, the full consequences of which will be drawn out over days to come, can only be taken as a result of animal spirits – of a spontaneous urge to action rather than inaction, and not as the outcome of a weighted average of quantified benefits... Thus if the animal spirits are dimmed and spontaneous optimism falters, leaving us to depend on nothing but a mathematical expectation, enterprise will fade and die.^[22]

Keynes’s picture of the psychology of speculators is less indulgent.

In point of fact, all sorts of considerations enter into the market valuation which are in no way relevant to the prospective yield... The recurrence of a bank-holiday may raise the market valuation of the British railway system by several million pounds.^[23]

(Henry Hazlitt examined some railway share prices and found that they did not bear out Keynes’s assertion.^[24])

Keynes considers speculators to be concerned...

...not with what an investment is really worth to a man who buys it ‘for keeps’, but with what the market will value it at, under the influence of mass psychology, three months or a year hence...

This battle of wits to anticipate the basis of conventional valuation a few months hence, rather than the prospective yield of an investment over a long term of years, does not even require gulls amongst the public to feed the maws of the professional;– it can be played by professionals amongst themselves. Nor is it necessary that anyone should keep his simple faith in the conventional basis of valuation having any genuine long-term validity. For it is, so to speak, a game of Snap, of Old Maid, of Musical Chairs – a pastime in which he is victor who says Snap neither too soon nor too late, who passed the Old Maid to his neighbour before the game is over, who secures a chair for himself when the music stops. These games can be played with zest and enjoyment, though all the players know that it is the Old Maid which is circulating, or that when the music stops some of the players will find themselves unseated.

Or, to change the metaphor slightly, professional investment may be likened to those newspaper competitions in which the competitors have to pick out the six prettiest faces from a hundred photographs, the prize being awarded to the competitor whose choice most nearly corresponds to the average preferences of the competitors as a whole; so that each competitor has to pick, not those faces which he himself finds prettiest, but those which he thinks likeliest to catch the fancy of the other competitors, all of whom are looking at the problem from the same point of view. It is not a case of choosing those which, to the best of one’s judgment, are really the prettiest, nor even those which average opinion genuinely thinks the prettiest. We have reached the third degree where we devote our intelligences to anticipating what average opinion expects the average opinion to be. And there are some, I believe, who practise the fourth, fifth and higher degrees.

Chapter 21: Wage behaviour

Keynes’s theory of the trade cycle is a theory of the slow oscillation of money income which requires it to be possible for income to move upwards or downwards. If he had assumed that wages were constant, upward motion of income would have been impossible at full employment, and he would have needed some mechanism to frustrate upward pressure if it arose in such circumstances.

His task is made easier by a less restrictive (but nonetheless crude) assumption concerning wage behaviour:

let us simplify our assumptions still further, and assume... that the factors of production... are content with the same money-wage so long as there is a surplus of them unemployed... ; whilst as soon as full employment is reached, it will thenceforward be the wage-unit and prices which will increase in exact proportion to the increase in effective demand.^[25]

Chapter 22: The trade cycle

Keynes's theory of the trade cycle is based on ‘a cyclical change in the marginal efficiency of capital’ induced by ‘the uncontrollable and disobedient psychology of the business world’ (pp313, 317).

The marginal efficiency of capital depends... on current expectations... But, as we have seen, the basis for such expectations is very precarious. Being based on shifting and unreliable evidence, they are subject to sudden and violent changes.^[26]

Optimism leads to a rise in the marginal efficiency of capital and increased investment, reflected – through the multiplier – in an even greater increase in income until ‘disillusion falls upon an over-optimistic and over-bought market‘ which consequently falls with ‘sudden and even catastrophic force’ (p316).

There are reasons, given firstly by the length of life of durable assets... and secondly by the carrying-costs of surplus stocks, why the duration of the downward movement should have an order of magnitude... between, let us say, three and five years.^[27]

And a half cycle of 5 years tallies with Jevons’s sunspot cycle length of 11 years.

Income fluctuates cyclically in Keynes’s theory, with the effect being borne by prices if income increases during a period of full employment, and by employment in other circumstances.

Wage behaviour and the Phillips curve

Keynes’s assumption about wage behaviour has been the subject of much criticism. Hazlitt pointed out that wage rates adapted partially to depression conditions^[28] and Keynes can hardly have been unaware of this. The consequence may be that effects on employment are weaker than his model implies, but not that they disappear.

Lerner pointed out in the 40s that it was optimistic to hope that the workforce would be content with fixed wages in the presence of rising prices, and proposed a modification to Keynes’s model. After this a succession of more elaborate models were constructed, many associated with the Phillips curve.

Keynes’s optimistic prediction that an increase in money supply would be taken up by an increase in employment led to Jacob Viner’s pessimistic prediction that “in a world organized in accordance with Keynes’ specifications there would be a constant race between the printing press and the business agents of the trade unions”.^[29]

Models of wage pressure on the economy needed frequent correction and the standing of Keynesian theory suffered. Geoff Tily wrote ruefully:

Finally, the most destructive step of all was Samuelson’s and [Robert] Solow’s incorporation of the Phillips curve into ‘Keynesian’ theory in a manner which traduced not only Phillips but also Keynes’s careful work in the General Theory, Chapter 21, substituting for its subtlety an immutable relationship between inflation and employment. The 1970s combination of inflation and stagnating economic activity was at odds with this relationship, and therefore ‘Keynesianism’, and by association Keynes were rejected. Monetarism was merely waiting in the wings for this to happen.^[30]

Keynes’s assumption of wage behaviour was not an integral part of his theory – very little in his book depends on it – and was avowedly a simplification: in fact it was the simplest assumption he could make without imposing an unnatural cap on money income.

Explanation based on hoarding

A natural interpretation of Keynes’s view is that he draws a demarcation line after the first four items in the list. It is certainly true that saving can find an outlet in hoarding whose demand is not a demand for investment. This was the line pursued by Jacob Viner, who wrote of Keynes that:

He finds fault with the “classical” economists for their alleged neglect of the gulf between the desire to save and the desire to invest, i.e., for their neglect of “liquidity preferences”... It was a shortcoming of the Ricardian wing of the classical school that... they steadfastly adhered to their position that hoarding was so abnormal a phenomenon as not to constitute a significant contributing factor to unemployment...^[59]

Hoarding constitutes a fairly small proportion of saving, and Viner accordingly doubted that it could play the role he thought Keynes assigned to it. Keynes, in his reply, did not seem to understand why Viner had imagined hoarding to be an important part of his theory.^[60] In fact his model of liquidity allows no scope for either the supply or the demand for hoarding: the quantity of money is fixed and each individual person desires to hold a static amount.

Supporters as well as critics of Keynes have adopted hoarding explanations. Paul Davidson, after rejecting Samuelson’s reading of the General Theory, reported that a 1977 paper by ‘the perennial renegade’ F. H. Hahn had generalised monetary hoarding to ‘non-reproducible assets’:

Some forty years after Keynes, Hahn rediscovered Keynes’s point that a stable involuntary unemployment equilibrium could exist even in a Walrasian system with flexible wages and prices  whenever there are “resting places for savings in other than reproducible assets”... [since] any non-reproducible asset allows for a choice between employment inducing and non-employment inducing demand.^[61]

Joan Robinson, on the other hand, rejected hoarding explanations with characteristic vigour as...

...simply an error... [which] arises, no doubt, from the desire to find where the vanished savings have got to... It is of no use to search for the non-existent savings either in “hoards” or anywhere else.^[62]

Of course the mystery is not that savings themselves have gone missing, but that there is a demand to save which is neither the demand for capital goods nor the demand to accumulate money in lieu of spending it.

Explanation based on lending

Lending and borrowing are two sides of the same transaction which necessarily cancel out, and which the classics therefore ignored for most purposes. It may be that Keynes saw the demand for lending as a demand for saving which was not the same thing as a demand for investment, thus drawing his demarcation under the first item in the list. He says that:

So long as it is open to the individual to employ his wealth in hoarding or lending money, the alternative of purchasing actual capital assets cannot be rendered sufficiently attractive...^[63]

and that:

...there is always an alternative to the ownership of real capital-assets, namely the ownership of money and debts...^[64]

A more decisive, though somewhat involved, statement occurs in his discussion of the multiplier on pp128ff. He says in a footnote that:

It is often convenient to use the term ‘loan expenditure’ to include the public investment financed by borrowing from individuals and also any other current public expenditure which is so financed. Strictly speaking, the latter should be reckoned as negative saving...^[65]

and it is difficult to interpret this as not implying that public demand for borrowing corresponds to a demand for dis-saving – not just for the exchequer (which is immaterial to the level of employment) but for society as a whole.

Explanation based on the endogeneity of money supply

Hansen suggested that adjustments to the money supply might absorb the gap between demands for saving and investment:

If expectations are favorable for investment, though funds are currently lacking, means of purchase can readily be made available in a society with an elastic money and credit system.^[66]

This draws a line after the second item in the list.

Explanation based on the separation of roles

Alternatively it might be said that when money is lent, the decision to lend is made by the saver and the decision to invest by the borrower, and that these may not conform with each other. This draws a line somewhere around the middle of the list and seems to be the assumption made by Frank Knight, who commented that:

It almost seems as if the money which is saved is completely distinct from the money which is lent and borrowed, and that the former, if it ever reaches a bank, or any lending agency, is still kept entirely separate.^[67]

The significance of the separation of roles is rejected every bit as forcefully by Keynes himself in §V of Chapter 7.

This explanation too had acceptance among some of Keynes’s ‘Book I’ supporters. Joan Robinson wrote:

Decisions to save and decisions to invest are taken quite independently of each other... Under a completely socialist system the government would decide how much investment was desirable... But in the system under which we live the decision to save and the decision to invest are not bound together, and the motives governing them are quite different... The individual saver has no direct influence upon the rate of investment. If entrepreneurs see a profit to be made by investment, investment will take place, and if they do not it will not. The initiative lies with the entrepreneurs, not with the savers. The savers, as a group, are helpless in the hands of the entrepreneurs...^[68]

She doesn’t tell us whether decisions to save are independent of decisions to lend, or whether savers’ decisions to lend are independent of entrepreneurs’ decisions to borrow, or whether entrepreneurs’ decisions to borrow are independent of their decisions to invest.

Hansen mostly follows this explanation. He writes that:

Under modern conditions savers and real investors are to a high degree different groups.^[69]

Denial of a causal link from saving to investment

The previous explanations have sought to identify particular forms of saving which do not contribute to the demand for investment. A more extreme viewpoint is that all  forms of saving have this property. Gordon Fletcher deemed it a ‘myth’ that ‘saving finances investment’:

Keynes attacked the widespread belief that investment is somehow financed by saving.^[70]

The position has several advantages. It avoids drawing a line which might look arbitrary between different types of saving. Moreover the whole of Keynes’s theory is based on saving qua  saving, not on the special properties of particular forms. The propensity to save embraces all forms of saving, and it is total saving (over all forms) which equals total investment. Sometimes this seems to be Keynes’s own view:

If... an act of saving does nothing to improve prospective yield, it does nothing to stimulate investment...^[71]

an observation which can be applied to summerhouses as easily as to industrial plant.^[72] On the other hand capital expenditure constitutes saving under Keynes’s definition, and it seems perverse to deny that investment is financed (or stimulated) by capital expenditure.

Explanation based on the gap between plans and reality

Schumpeter provided the following summary of Keynes’s theory:

Current saving and current investment, being identically equal, cannot determine anything. Planned (ex ante) saving and planned (ex ante) investment determine income (total net output)...^[34]

G. L. S. Shackle later wrote that:

Myrdalian ex ante language would have saved the General Theory from describing the flow of investment and the flow of saving as identically, tautologically equal, and within the same discourse, treating their equality as a condition which may, or not, be fulfilled.^[73]

All forms of saving may depart slightly from their planned levels.

Explanation based on Chapters 11 and 14

The IS  equation emerges in Keynes’s theory in Chapter 14 as common ground with the classics: he says that he would not dispute the classical proposition that “the current rate of interest must lie at the point where the demand curve for capital... cuts the curve of the amounts saved” (p178). It can also be argued from the identity of saving and investment. Harrod’s and Hicks’s readings of Keynes did not go far beyond this: see Mr Keynes and the Classics.

But if one starts from the IS  equation, and interprets the schedule of the marginal efficiency of capital as a demand curve (as Keynes may not have been the first to do, and as Hicks was willing to follow him in doing), then it is a natural step to add it to the demand curve for consumption to get the Keynesian cross argument of Chapter 3. One thus arrives at a concept of effective demand which does not derive from considerations of what earners seek to do with their money, but which needs to make sense in terms of them, leading to the other explanations listed above.

The respending multiplier as explained by Samuelson

The multiplier was introduced to economic theory by Richard Kahn in a 1931 paper^[74] cited by Keynes in his own discussion. Kahn’s mechanism is based on looking at an initial expenditure and noting that it will end up in the pockets of various employees and suppliers who in turn will spend a proportion of what they receive, thereby creating more employment; and so forth. Robinson, Samuelson and Hansen include it in their accounts. Mantoux wrote that “The entire demonstration, it would seem... rests on this function”.

Samuelson explained the multiplier in these terms:

Let’s suppose that I hire unemployed resources to build a $1000 woodshed. My carpenters and lumber producers will get an extra $1000 of income... If they all have a marginal propensity to consume of 2/3, they will now spend $666.67 on new consumption goods. The producers of these goods will now have extra incomes... they in turn will spend $444.44... Thus an endless chain of secondary consumption respending  is set in motion by my primary  investment of $1000.^[75]

Samuelson is being arbitrary in describing secondary employment as arising from carpenters’ consumption spending but not from their investment spending, while seeing the chain as being set in motion by a primary investment rather than by primary consumption. If the carpenters spend their remaining $333.33 on housing and infrastructure, the builders of these goods too will have extra incomes; and if Samuelson had instead spent $1000 on Napa Valley wine, it is reasonable to assume that the growers would likewise have spent their revenue on consumption and investment goods giving rise to similar ‘secondary respending’.

Keynes’s multipliers

There are two slightly incompatible accounts of the multplier in the General Theory : an opaque definition followed by a satirical illustration in Chapter 10 and a perfectly lucid account of its theoretical role in Chapter 18 – the Book IV ‘restatement’. Since it is the theoretical functioning of the multiplier which matters, and since Chapter 18 makes its meaning clear, one is free to dispense with the description in Chapter 10. As we might expect, the earlier account is the more radical, the later the more classical; but in neither is there any sign of an argument based on the repeated spending of the same money as we saw with Kahn and Samuelson.

The Chapter 18 ‘investment multiplier’ simply measures the response of income to a change in a system parameter, namely the schedule of the marginal efficiency of capital. There will be an associated response for employment, and given Keynes’s use of wage units this too can be expressed as a dimensionless constant. It might reasonably be termed the ’employment mulitplier’ and one might well ‘assume (as a first approximation) that the employment multiplier is equal to the investment multiplier’ (p248).

Keynes assigns a value 1 / (1–c ) to his investment multiplier, where c  is the marginal propensity to consume. This is correct under his Chapter 13 theory of liquidity preference but he doesn’t notice that it needs to be modified in the light of his Chapter 15 theory. A respending multiplier would be expected to take take the value of 1 / (1–x ) where x  is the marginal propensity to spend.

Keynes describes the Chapter 10 multiplier as being set in motion by an ‘increment of investment’ (p117) and gives the example of ‘increased public works’ (p116). He explicitly differentiates it from Kahn’s multiplier, which he terms the ‘employment multiplier’ but which is still not the same as the employment multiplier which might be seen as implicit in Chapter 18.

Although the Chapter 10 multiplier appears to be associated with acts of investment, Keynes describes it in the same chapter as a ‘logical’ mutliplier which ‘holds good continuously, without time-lag’ ^[76] suggesting that it is in fact a response to the schedule of the marginal efficiency of capital.

The four multipliers are summarised in the table.

Keynes makes an exception to the limitation to investment as a trigger for his multiplier in Chapter 10 by generalising this to ‘loan expenditure’ which also includes public dis-saving. The passage explaining this exception occurs in a footnote justifying the inclusion of a satirical passage which may have been written much earlier. It has already been quoted:

It is often convenient to use the term ‘loan expenditures’ to include the public investment financed by borrowing from individuals and also any other current public expenditure which is so financed. Strictly speaking, the latter should be reckoned as negative saving...^[77]

Hansen’s attempted reconciliation

According to Hansen,^[78] the main task of Kahn’s paper was to show that the multiplier was finite. Kahn used the term cul-de-sac  for any use of money which removed it from the respending cycle (Hansen preferred ‘leakage’). Kahn paid a lot of attention to the flow of money abroad to fund imports, but since Keynes worked with a closed economy this form of leakage can be ignored, and it appears that hoarding is the only one left. This is why a respending multiplier may be expected to take the value 1 / (1–x ), where x, the marginal propensity to spend, is not the same as Keynes’s c.

Hansen sought to show that Keynes’s multiplier had the same properties as a respending multiplier. The first difference to address lies in how the multiplier is put in motion. The initial expenditure for a respending multiplier is conventionally described as ‘investment’ but this term was dismissed by Hazlitt as a Pickwickianism^[79] since, evidently, consumption outlay can also be respent; any expenditure will do. Hansen does not deny the Pickwickianism, nor does he doubt that Keynes associated his own multiplier with investment in a non-Pickwickian sense. He bridges the gap by defining ‘investment expenditure’ as including ‘private-consumption outlays’ on the grounds that:

Whatever the initial increase in expenditure, whether private or public investment or simply an increase in private-consumption outlays resulting from tax reduction... the effect... is the same.^[80]

Hansen then discusses the numerical value of a respending multiplier in a laborious case-by-case analysis (pp90–96). The third and sixth cases are based on repeated rather than one-off expenditure; the remaining four are summarised in the table on the right. Only the second makes a distinction between consumption and expenditure. Hansen draws the conclusion that a respending multiplier should take the value 1 / (1–c ) rather than 1 / (1–x ), thus putting it in harmony with Keynes’s formula (p96).

The Book IV multiplier

The Chapter 18 multiplier is simply an intermediate term in a calculation. Any presentation of Keynesian theory which ignores it, or simply embodies it in equations, is consistent with Keynes’s Book IV since there is no need to focus attention on a single intermediate term. Hicks’s account is a case in point: he says that ‘the third equation becomes the multiplier equation, which performs such queer tricks’ and leaves the matter there.

The role which the multiplier receives in Keynes’s presentation slightly understates its significance. The response of income to changes in the schedule of the marginal efficiency of capital is discussed qualitatively  in Chapters 14 and 22: it is essential to the reasoning of the latter chapter (on the trade cycle) that the magnitude should be appreciable, but this assumption is not stated overtly. The relevance of the multiplier to Keynes’s theory of the trade cycle is more clearly seen in his reply to Viner.^[81]

Shackle regarded Keynes’s departure from Kahn’s multiplier as...

...a retrograde step... For when we look upon the Multiplier as an instantaneous functional relation... we are merely using the word Multiplier to stand for an alternative way of looking at the marginal propensity to consume.^[82]

§I: Say’s Law

Keynes reiterates his denial that an act of saving constitutes an act of investment without shedding light on his grounds. A formulation of classical macroeconomics in three equations was given above as follows:

• y' (N ) = W / p        i_s(r ) = s (y (N ),r )        M̂ = p ·y (N ) / V (r )

The role of Say’s Law in Keynes’s interpretation of them can be seen if we split the second equation into two components:

• i_s(r ) = i_d(y (N ),r )        i_d(y (N ),r ) = s (y (N ),r )

the first of which asserts the equilibrium between the supply and demand for investment, and the second of which identifies the demand for investment with the desired level of saving. In rejecting the second component Keynes denies that the total demand for goods in an economy is identical with its total income – i.e. that supply creates its own demand – and is therefore able to make their equality an equilibrium condition.

To justify his position Keynes needs to argue that some earnings do not enter into the demand for any commodity (consumption or capital) and he needs to say what the earner intends to do with his or her money which does not constitute the purchase of goods. He says:

An act of individual saving means – so to speak – a decision not to have dinner to-day. But it does not  necessitate a decision to have dinner or to buy a pair of boots a week hence or a year hence or to consume any specified thing at any specified date.^[86]

... an individual decision to save does not, in actual fact, involve the placing of any specific forward order for consumption, but merely the cancellation of a present order.^[87]

Then he adds that:

The absurd, though almost universal, idea that an act of individual saving is just as good for effective demand as an act of individual consumption, has been fostered by the fallacy... that an increased desire to hold wealth, being much the same thing as an increased desire to hold investments, must, by increasing the demand for investments, provide a stimulus to their production; so that current investment is promoted by individual saving to the same extent as present consumption is diminished.^[88]

But he makes no attempt to explain how an absurd view had become ‘almost universal’ or what error its adherents had made. In fact there do exist obvious reasons to consider it absurd, but they are of no help to Keynes. A person may be happy to lend his or her money with no concern for what use the borrower makes of it; but the classics saw lending as cancelled by borrowing (both in supply and demand) and Keynes does not say why this should not be so. Or the person may prefer to accumulate cash, and this is a genuine and serious issue which the classics had been prone to neglect, but which Keynes also excluded from his system.

At any rate, the classics assumed that every penny earned would be lent, spent, or accumulated; that net lending was zero; that spending was split between consumption and capital goods; and that total income was therefore equal to total demand for goods plus total demand for hoarding: and like Keynes they ignored the last factor.^[89]

Keynes saw savers as having the same options but came to a different conclusion. It is in this chapter that he mentioned “the ownership of money and debts” as “an alternative to the ownership of real capital-assets” (p212). On the same page he draws attention to the error of...

... believing that the owner of wealth desires a capital-asset as such, whereas what he really desires is its prospective yield.

§ II–IV: The declining yield of capital

Keynes argues that the value of capital derives from its scarcity and sympathises with ‘the pre-classical doctrine that everything is produced  by labour ’ (p213). The preference for direct over roundabout processes will depend on the rate of interest.

He wonders what would happen to ‘a society which finds itself so well equipped with capital that its marginal efficiency is zero’ while money provides a safe outlet for savings. He does not consider this hypothesis far-fetched: on the contrary...

... a properly run community... ought to be able to bring down the marginal efficiency of capital in equilibrium approximately to zero within a single generation...^[90]

He asserts that...

... the position of equilibrium, under conditions of laissez-faire, will be one in which employment is low enough and the standard of life sufficiently miserable to bring savings to zero’.^[91]

The misery does not depend on any assumption of static wages. If the return to capital falls to zero then according to Keynes’s theory there will be no investment, and income must collapse to the point at which the propensity to save disappears. However his conclusions are not pessimistic because he postulates that steps may be taken to adjust the interest rate to ensure full employment (p220), that ‘enormous social changes would result’ and that ‘this may be the most sensible way of getting rid of many of the objectionable features of capitalism’ (p221).