John Maynard Keynes
The General Theory of Employment, Interest and Money

Book III
The Propensity to Consume

Chapter 8. The Propensity to Consume: I. The Objective Factors

I

WE are now in a position to return to our main theme, from which we broke off at the end of Book I in order to deal with certain general problems of method and definition. The ultimate object of our analysis is to discover what determines the volume of employment. So far we have established the preliminary conclusion that the volume of employment is determined by the point of intersection of the aggregate supply function with the aggregate demand function. The aggregate supply function, however, which depends in the main on the physical conditions of supply, involves few considerations which are not already familiar. The form may be unfamiliar but the underlying factors are not new. We shall return to the aggregate supply function in Chapter 20, where we discuss its inverse under the name of the employment function. But, in the main, it is the part played by the aggregate demand function which has been overlooked; and it is to the aggregate demand function that we shall devote Books III and IV.

The aggregate demand function relates any given level of employment to the “proceeds” which that level of employment is expected to realise. The “proceeds” are made up of the sum of two quantities — the sum which will be spent on consumption when employment is at the given level, and the sum which will be devoted to investment. The factors which govern these two quantities are largely distinct. In this book we shall consider the former, namely what factors determine the sum which will be spent on consumption when employment is at a given level; and in Book IV we shall proceed to the factors which determine the sum which will be devoted to investment.

Since we are here concerned in determining what sum will be spent on consumption when employment is at a given level, we should, strictly speaking, consider the function which relates the former quantity (C) to the latter (N). It is more convenient, however, to work in terms of a slightly different function, namely, the function which relates the consumption in terms of wage-units (C_w) to the income in terms of wage-units (Y_w) corresponding to a level of employment N. This suffers from the objection that Y_w is not a unique function of N, which is the same in all circumstances. For the relationship between Y_w and N may depend (though probably in a very minor degree) on the precise nature of the employment. That is to say, two different distributions of a given aggregate employment N between different employments might (owing to the different shapes of the individual employment functions — a matter to be discussed in Chapter 20 below) lead to different values of Y_w. In conceivable circumstances a special allowance might have to be made for this factor. But in general it is a good approximation to regard Y_w as uniquely determined by N. We will therefore define what we shall call the propensity to consume as the functional relationship χ between Y_w a given level of income in terms of wage-units, and C_w the expenditure on consumption out of that level of income, so that

C_w = χ(Y_w) or C = W.χ(Y_w).

The amount that the community spends on consumption obviously depends (i) partly on the amount of its income, (ii) partly on the other objective attendant circumstances, and (iii) partly on the subjective needs and the psychological propensities and habits of the individuals composing it and the principles on which the income is divided between them (which may suffer modification as output is increased). The motives to spending interact and the attempt to classify them runs the danger of false division. Nevertheless it will clear our minds to consider them separately under two broad heads which we shall call the subjective factors and the objective factors. The subjective factors, which we shall consider in more detail in the next Chapter, include those psychological characteristics of human nature and those social practices and institutions which, though not unalterable, are unlikely to undergo a material change over a short period of time except in abnormal or revolutionary circumstances. In an historical enquiry or in comparing one social system with another of a different type, it is necessary to take account of the manner in which changes in the subjective factors may affect the propensity to consume. But, in general, we shall in what follows take the subjective factors as given; and we shall assume that the propensity to consume depends only on changes in the objective factors.

II

The principal objective factors which influence the propensity to consume appear to be the following:

(1) A change in the wage-unit. — Consumption (C) is obviously much more a function of (in some sense) real income than of money-income. In a given state of technique and tastes and of social conditions determining the distribution of income, a man’s real income will rise and fall with the amount of his command over labour-units, i.e. with the amount of his income measured in wage-units; though when the aggregate volume of output changes, his real income will (owing to the operation of decreasing returns) rise less than in proportion to his income measured in wage-units. As a first approximation, therefore, we can reasonably assume that, if the wage-unit changes, the expenditure on consumption corresponding to a given level of employment will, like prices, change in the same proportion; though in some circumstances we may have to make an allowance for the possible reactions on aggregate consumption of the change in the distribution of a given real income between entrepreneurs and rentiers resulting from a change in the wage-unit. Apart from this, we have already allowed for changes in the wage-unit by defining the propensity to consume in terms of income measured in terms of wage-units.

(2) A change in the difference between income and net income. We have shown above that the amount of consumption depends on net income rather than on income, since it is, by definition, his net income that a man has primarily in mind when he is deciding his scale of consumption. In a given situation there may be a somewhat stable relationship between the two, in the sense that there will be a function uniquely relating different levels of income to the corresponding levels of net income. If, however, this should not be the case, such part of any change in income as is not reflected in net income must be neglected since it will have no effect on consumption; and, similarly, a change in net income, not reflected in income, must be allowed for. Save in exceptional circumstances, however, I doubt the practical importance of this factor. We will return to a fuller discussion of the effect on consumption of the difference between income and net income in the fourth section of this chapter.

(3) Windfall changes in capital-values not allowed for in calculating net income. — These are of much more importance in modifying the propensity to consume, since they will bear no stable or regular relationship to the amount of income. The consumption of the wealth-owning class may be extremely susceptible to unforeseen changes in the money-value of its wealth. This should be classified amongst the major factors capable of causing short-period changes in the propensity to consume.

(4) Changes in the rate of time-discounting, i.e. in the ratio of exchange between present goods and future goods. — This is not quite the same thing as the rate of interest, since it allows for future changes in the purchasing power of money in so far as these are foreseen. Account has also to be taken of all kinds of risks, such as the prospect of not living to enjoy the future goods or of confiscatory taxation. As an approximation, however, we can identify this with the rate of interest.

The influence of this factor on the rate of spending out of a given income is open to a good deal of doubt. For the classical theory of the rate of interest,^[1] which was based on the idea that the rate of interest was the factor which brought the supply and demand for savings into equilibrium, it was convenient to suppose that expenditure on consumption is cet. par. negatively sensitive to changes in the rate of interest, so that any rise in the rate of interest would appreciably diminish consumption. It has long been recognised, however, that the total effect of changes in the rate of interest on the readiness to spend on present consumption is complex and uncertain, being dependent on conflicting tendencies, since some of the subjective motives towards saving will be more easily satisfied if the rate of interest rises, whilst others will be weakened. Over a long period substantial changes in the rate of interest probably tend to modify social habits considerably, thus affecting the subjective propensity to spend — though in which direction it would be hard to say, except in the light of actual experience. The usual type of short-period fluctuation in the rate of interest is not likely, however, to have much direct influence on spending either way. There are not many people who will alter their way of living because the rate of interest has fallen from 5 to 4 per cent, if their aggregate income is the same as before. Indirectly there may be more effects, though not all in the same direction. Perhaps the most important influence, operating through changes in the rate of interest, on the readiness to spend out of a given income, depends on the effect of these changes on the appreciation or depreciation in the price of securities and other assets. For if a man is enjoying a windfall increment in the value of his capital, it is natural that his motives towards current spending should be strengthened, even though in terms of income his capital is worth no more than before; and weakened if he is suffering capital losses. But this indirect influence we have allowed for already under (3) above. Apart from this, the main conclusion suggested by experience is, I think, that the short-period influence of the rate of interest on individual spending out of a given income is secondary and relatively unimportant, except, perhaps, where unusually large changes are in question. When the rate of interest falls very low indeed, the increase in the ratio between an annuity purchasable for a given sum and the annual interest on that sum may, however, provide an important source of negative saving by encouraging the practice of providing for old age by the purchase of an annuity.

The abnormal situation, where the propensity to consume may be sharply affected by the development of extreme uncertainty concerning the future and what it may bring forth, should also, perhaps, be classified under this heading.

(5) Changes in fiscal policy. — In so far as the inducement to the individual to save depends on the future return which he expects, it clearly depends not only on the rate of interest but on the fiscal policy of the Government. Income taxes, especially when they discriminate against “unearned” income, taxes on capital-profits, death-duties and the like are as relevant as the rate of interest; whilst the range of possible changes in fiscal policy may be greater, in expectation at least, than for the rate of interest itself. If fiscal policy is used as a deliberate instrument for the more equal distribution of incomes, its effect in increasing the propensity to consume is, of course, all the greater.^[2]

We must also take account of the effect on the aggregate propensity to consume of Government sinking funds for the discharge of debt paid for out of ordinary taxation. For these represent a species of corporate saving, so that a policy of substantial sinking funds must be regarded in given circumstances as reducing the propensity to consume. It is for this reason that a change-over from a policy of Government borrowing to the opposite policy of providing sinking funds (or vice versa) is capable of causing a severe contraction (or marked expansion) of effective demand.

(6) Changes in expectations of the relation between the present and the future level of income. — We must catalogue this factor for the sake of formal completeness. But, whilst it may affect considerably a particular individual’s propensity to consume, it is likely to average out for the community as a whole. Moreover, it is a matter about which there is, as a rule, too much uncertainty for it to exert much influence.

We are left therefore, with the conclusion that in a given situation the propensity to consume may be considered a fairly stable function, provided that we have eliminated changes in the wage-unit in terms of money. Windfall changes in capital-values will be capable of changing the propensity to consume, and substantial changes in the rate of interest and in fiscal policy may make some difference; but the other objective factors which might affect it, whilst they must not be overlooked, are not likely to be important in ordinary circumstances.

The fact that, given the general economic situation, the expenditure on consumption in terms of the wage-unit depends in the main, on the volume of output and employment is the justification for summing up the other factors in the portmanteau function “propensity to consume”. For whilst the other factors are capable of varying (and this must not be forgotten), the aggregate income measured in terms of the wage-unit is, as a rule, the principal variable upon which the consumption-constituent of the aggregate demand function will depend.

III

Granted, then, that the propensity to consume is a fairly stable function so that, as a rule, the amount of aggregate consumption mainly depends on the amount of aggregate income (both measured in terms of wage-units), changes in the propensity itself being treated as a secondary influence, what is the normal shape of this function?

The fundamental psychological law, upon which we are entitled to depend with great confidence both a priori from our knowledge of human nature and from the detailed facts of experience, is that men are disposed, as a rule and on the average, to increase their consumption as their income increases, but not by as much as the increase in their income. That is to say, if C_w is the amount of consumption and Y_w is income (both measured in wage-units) ΔC_w has the same sign as ΔY_w but is smaller in amount, i.e. dC_w/dY_w is positive and less than unity.

This is especially the case where we have short periods in view, as in the case of the so-called cyclical fluctuations of employment during which habits, as distinct from more permanent psychological propensities, are not given time enough to adapt themselves to changed objective circumstances. For a man’s habitual standard of life usually has the first claim on his income, and he is apt to save the difference which discovers itself between his actual income and the expense of his habitual standard; or, if he does adjust his expenditure to changes in his income, he will over short periods do so imperfectly. Thus a rising income will often be accompanied by increased saving, and a falling income by decreased saving, on a greater scale at first than subsequently.

But, apart from short-period changes in the level of income, it is also obvious that a higher absolute level of income will tend, as a rule, to widen the gap between income and consumption. For the satisfaction of the immediate primary needs of a man and his family is usually a stronger motive than the motives towards accumulation, which only acquire effective sway when a margin of comfort has been attained. These reasons will lead, as a rule, to a greater proportion of income being saved as real income increases. But whether or not a greater proportion is saved, we take it as a fundamental psychological rule of any modern community that, when its real income is increased, it will not increase its consumption by an equal absolute amount, so that a greater absolute amount must be saved, unless a large and unusual change is occurring at the same time in other factors. As we shall show subsequently,^[3] the stability of the economic system essentially depends on this rule prevailing in practice. This means that, if employment and hence aggregate income increase, not all the additional employment will be required to satisfy the needs of additional consumption.

On the other hand, a decline in income due to a decline in the level of employment, if it goes far, may even cause consumption to exceed income not only by some individuals and institutions using up the financial reserves which they have accumulated in better times, but also by the Government, which will be liable, willingly or unwillingly, to run into a budgetary deficit or will provide unemployment relief, for example, out of borrowed money. Thus, when employment falls to a low level, aggregate consumption will decline by a smaller amount than that by which real income has declined, by reason both of the habitual behaviour of individuals and also of the probable policy of governments; which is the explanation why a new position of equilibrium can usually be reached within a modest range of fluctuation. Otherwise a fall in employment and income, once started, might proceed to extreme lengths.

This simple principle leads, it will be seen, to the same conclusion as before, namely, that employment can only increase pari passu with an increase in investment; unless, indeed, there is a change in the propensity to consume. For since consumers will spend less than the increase in aggregate supply price when employment is increased, the increased employment will prove unprofitable unless there is an increase in investment to fill the gap.

IV

We must not underestimate the importance of the fact already mentioned above that, whereas employment is a function of the expected consumption and the expected investment, consumption is, cet. par., a function of net income, i.e. of net investment (net income being equal to consumption plus net investment). In other words, the larger the financial provision which it is thought necessary to make before reckoning net income, the less favourable to consumption, and therefore to employment, will a given level of investment prove to be.

When the whole of this financial provision (or supplementary cost) is in fact currently expended in the upkeep of the already existing capital equipment, this point is not likely to be overlooked. But when the financial provision exceeds the actual expenditure on current upkeep, the practical results of this in its effect on employment are not always appreciated. For the amount of this excess neither directly gives rise to current investment nor is available to pay for consumption. It has, therefore, to be balanced by new investment, the demand for which has arisen quite independently of the current wastage of old equipment against which the financial provision is being made; with the result that the new investment available to provide current income is correspondingly diminished and a more intense demand for new investment is necessary to make possible a given level of employment. Moreover, much the same considerations apply to the allowance for wastage included in user cost, in so far as the wastage is not actually made good.

Take a house which continues to be habitable until it is demolished or abandoned. If a certain sum is written off its value out of the annual rent paid by the tenants, which the landlord neither spends on upkeep nor regards as net income available for consumption, this provision, whether it is a part of U or of V, constitutes a drag on employment all through the life of the house, suddenly made good in a lump when the house has to be rebuilt.

In a stationary economy all this might not be worth mentioning, since in each year the depreciation allowances in respect of old houses would be exactly offset by the new houses built in replacement of those reaching the end of their lives in that year. But such factors may be serious in a non-static economy, especially during a period which immediately succeeds a lively burst of investment in long-lived capital. For in such circumstances a very large proportion of the new items of investment may be absorbed by the larger financial provisions made by entrepreneurs in respect of existing capital equipment, upon the repairs and renewal of which, though it is wearing out with time, the date has not yet arrived for spending anything approaching the full financial provision which is being set aside; with the result that incomes cannot rise above a level which is low enough to correspond with a low aggregate of net investment. Thus sinking funds, etc., are apt to withdraw spending power from the consumer long before the demand for expenditure on replacements (which such provisions are anticipating) comes into play; i.e. they diminish the current effective demand and only increase it in the year in which the replacement is actually made. If the effect of this is aggravated by “financial prudence”, i.e. by its being thought advisable to “write off” the initial cost more rapidly than the equipment actually wears out, the cumulative result may be very serious indeed.

In the United States, for example, by 1929 the rapid capital expansion of the previous five years had led cumulatively to the setting up of sinking funds and depreciation allowances, in respect of plant which did not need replacement, on so huge a scale that an enormous volume of entirely new investment was required merely to absorb these financial provisions; and it became almost hopeless to find still more new investment on a sufficient scale to provide for such new saving as a wealthy community in full employment would be disposed to set aside. This factor alone was probably sufficient to cause a slump. And, furthermore, since “financial prudence” of this kind continued to be exercised through the slump by those great corporations which were still in a position to afford it, it offered a serious obstacle to early recovery.

Or again, in Great Britain at the present time (1935) the substantial amount of house-building and of other new investments since the war has led to an amount of sinking funds being set up much in excess of any present requirements for expenditure on repairs and renewals, a tendency which has been accentuated, where the investment has been made by local authorities and public boards, by the principles of “sound” finance which often require sinking funds sufficient to write off the initial cost some time before replacement will actually fall due; with the result that even if private individuals were ready to spend the whole of their net incomes it would be a severe task to restore full employment in the face of this heavy volume of statutory provision by public and semi-public authorities, entirely associated from any corresponding new investment. The sinking funds of local authorities now stand, I think,^[4] at an annual figure of more than half the amount which these authorities are spending on the whole of their new developments.^[5] Yet it is not certain that the Ministry of Health are aware, when they insist on stiff sinking funds by local authorities, how much they may be aggravating the problem of unemployment. In the case of advances by Building Societies to help an individual to build his own house, the desire to be clear of debt more rapidly than the house actually deteriorates may stimulate the house-owner to save more than he otherwise would; — though this factor should be classified, perhaps, as diminishing the propensity to consume directly rather than through its effect on net income. In actual figures, repayments of mortgages advanced by Building Societies, which amounted to £24,000,000 in 1925, had risen to £68,000,000 by 1933, as compared with new advances of £103,000,000; and to-day the repayments are probably still higher.

That it is investment, rather than net investment, which emerges from the statistics of output, is brought out forcibly and naturally in Mr. Colin Clark’s National Income, 1924-1931. He also shows what a large proportion depreciation, etc., normally bears to the value of investment. For example, he estimates that in Great Britain, over the years 1928-1931,^[6] the investment and the net investment were as follows, though his gross investment is probably somewhat greater than my investment, inasmuch as it may include a part of user cost, and it is not clear how closely his “net investment” corresponds to my definition of this term:

Mr. Kuznets has arrived at much the same conclusion in compiling the statistics of the Gross Capital Formation (as he calls what I call investments in the United States) 1919-1933. The physical fact, to which the statistics of output correspond, is inevitably the gross, and not the net, investment. Mr. Kuznets has also discovered the difficulties in passing from gross investment to net investment. “The difficulty”, he writes, “of passing from gross to net capital formation, that is, the difficulty of correcting for the consumption of existing durable commodities, is not only in the lack of data. The very concept of annual consumption of commodities that last over a number of years suffers from ambiguity”.^[7] He falls back, therefore, “on the assumption that the allowance for depreciation and depletion on the books of business firms describes correctly the volume of consumption of already existing’ finished durable goods used by business firms”. On the other hand, he attempts no deduction at all in respect of houses and other durable commodities in the hands of individuals. His very interesting results for the United States can be summarised as follows:

Several facts emerge with prominence from this table. Net capital formation was very steady over the quinquennium 1925-1929, with only a 10 per cent increase in the latter part of the upward movement. The deduction for entrepreneurs’ repairs, maintenance, depreciation and depletion remained at a high figure even at the bottom of the slump. But Mr. Kuznets’ method must surely lead to too low an estimate of the annual increase in depreciation, etc.; for he puts the latter at less than 1 1/2 per cent per annum of the new net capital formation. Above all, net capital formation suffered an appalling collapse after 1929, falling in 1932 to a figure no less than 95 per cent below the average of the quinquennium 1925-1929.

The above is, to some extent, a digression. But it is important to emphasise the magnitude of the deduction which has to be made from the income of a society, which already possesses a large stock of capital, before we arrive at the net income which is ordinarily available for consumption. For if we overlook this, we may underestimate the heavy drag on the propensity to consume which exists even in conditions where the public is ready to consume a very large proportion of its net income.

Consumption — to repeat the obvious — is the sole end and object of all economic activity. Opportunities for employment are necessarily limited by the extent of aggregate demand. Aggregate demand can be derived only from present consumption or from present provision for future consumption. The consumption for which we can profitably provide in advance cannot be pushed indefinitely into the future. We cannot, as a community, provide for future consumption by financial expedients but only by current physical output. In so far as our social and business organisation separates financial provision for the future from physical provision for the future so that efforts to secure the former do not necessarily carry the latter with them, financial prudence will be liable to diminish aggregate demand and thus impair well-being, as there are many examples to testify. The greater, moreover, the consumption for which we have provided in advance, the more difficult it is to find something further to provide for in advance, and the greater our dependence on present consumption as a source of demand. Yet the larger our incomes, the greater, unfortunately, is the margin between our incomes and our consumption. So, failing some novel expedient, there is, as we shall see, no answer to the riddle, except that there must be sufficient unemployment to keep us so poor that our consumption falls short of our income by no more than the equivalent of the physical provision for future consumption which it pays to produce to-day.

Or look at the matter thus. Consumption is satisfied partly by objects produced currently and partly by objects produced previously, i.e. by disinvestment. To the extent that consumption is satisfied by the latter, there is a contraction of current demand, since to that extent a part of current expenditure fails to find its way back as a part of net income. Contrariwise whenever an object is produced within the period with a view to satisfying consumption subsequently, an expansion of current demand is set up. Now all capital-investment is destined to result, sooner or later, in capital-disinvestment. Thus the problem of providing that new capital-investment shall always outrun capital-disinvestment sufficiently to fill the gap between net income and consumption, presents a problem which is increasingly difficult as capital increases. New capital-investment can only take place in excess of current capital-disinvestment if future expenditure on consumption is expected to increase. Each time we secure to-day’s equilibrium by increased investment we are aggravating the difficulty of securing equilibrium to-morrow. A diminished propensity to consume to-day can only be accommodated to the public advantage if an increased propensity to consume is expected to exist some day. We are reminded of “The Fable of the Bees” — the gay of to-morrow are absolutely indispensable to provide a raison d'être for the grave of to-day.

It is a curious thing, worthy of mention, that the popular mind seems only to be aware of this ultimate perplexity where public investment is concerned, as in the case of road-building and house-building and the like. It is commonly urged as an objection to schemes for raising employment by investment under the auspices of public authority that it is laying up trouble for the future. “What will you do,” it is asked, “when you have built all the houses and roads and town halls and electric grids and water supplies and so forth which the stationary population of the future can be expected to require?” But it is not so easily understood that the same difficulty applies to private investment and to industrial expansion; particularly to the latter, since it is much easier to see an early satiation of the demand for new factories and plant which absorb individually but little money, than of the demand for dwelling-houses.

The obstacle to a clear understanding is, in these examples, much the same as in many academic discussions of capital, namely, an inadequate appreciation of the fact that capital is not a self-subsistent entity existing apart from consumption. On the contrary, every weakening in the propensity to consume regarded as a permanent habit must weaken the demand for capital as well as the demand for consumption.

Author’s Footnotes

1. Cf. Chapter 14 below.

2. It may be mentioned, in passing, that the effect of fiscal policy on the growth of wealth has been the subject of an important misunderstanding which, however, we cannot discuss adequately without the assistance of the theory of the rate of interest to be given in Book IV.

3. Cf. p. 251 below.

4. The actual figures are deemed of so little interest that they are only published two years or more in arrears.

5. In the year ending March 31, 1930, local authorities spent £87,000,000 on capital account, of which £37,000,000 was provided by sinking funds, etc., in respect of previous capital expenditure; in the year ending March 31, 1933, the figures were £81,000,000 and £46,000,000.

6. Op. cit. pp. 117 and 138.

7. These references are taken from a Bulletin (No. 52) of the National Bureau of Economic Research, giving preliminary results of Mr. Kuznets’ forthcoming book.

Contents | Chapter 9 | Keynes Archive