The Theoretical Mechanism Of Saving

The Theoretical Mechanism Of Saving

For a long time saving was regarded as such a simple economic action that it was hardly thought necessary to make a theory of it. Adam Smith and Turgot expatiated mainly on the economic benefits that it yields, and paid little attention to the way it works. Through­out the nineteenth century authors of the most opposite views rivalled Adam Smith in singing its praises. Karl Marx considered it the essential function of the capitalist, and Keynes, who was later to criticize it so acutely, explained eloquently in his famous book on Reparations how the savings of the older countries before the First World War served to fertilize the industry of the whole world—one of the most striking pictures that his pen has drawn. Some discordant voices were heard, however—the most celebrated being that of Lord Lauderdale at the end of the eighteenth century. In a book that was widely read—it was even translated into French—and in which he often finds fault with Adam Smith's views, he protests against the too rapid liquidation of the English debt. Dupont de Nemours, in a very elaborate note on Turgot's Reflexions, vigorously condemns all saving that does not consist in 'productive expenditure.' Malthus, for his part, spoke in praise of luxury, and at the end of the nineteenth century certain less-known writers, such as Hobson, regarded excessive saving as one of the sources of the distress of the working-classes.

But many points in the theory still remain obscure, and we must call attention to them here or we shall not understand the con­troversies that they have given rise to. For example, economists have not managed to reach agreement as to the influence that may be exerted on the growth of saving by the rate of interest. Normally, it would seem, every rise in the rate of interest should tend to increase saving and every fall to reduce it. But experience does not seem to bear this out. Pareto thinks that saving would continue even if the rate of interest fell to zero. Another Italian author, Ricci, thinks that the amount of saving is a function of the rate of interest that follows a curve that first rises and then falls after reaching a maximum. Gassel's view is that below a certain rate the reduction in saving will automatically produce a recovery in the rate and prevent its falling to zero. Divisia believes that the problem can only be solved induc­tively, and that we should stick to observed facts. But this is extremely difficult, and the answer to the question, he thinks, remains in abeyance. The uncertainty is greater still when it comes to analysing the mechanism of saving. The most satisfactory description is that given by Bohm-Bawerk. Saving, according to him, is merely a new direction given to the use of income. It consists in applying income not to the consumption of present goods, but to the making of instruments or the construction of factories or dwellings which will facilitate the future increase of immediately consumable goods and services. This description aims at showing how, even if all the forces of labour in an economic system are engaged, saving none the less provides the means for increasing the real future income of this economic system. If, on this hypothesis, an individual decides to save—that is, to withdraw from the consumption of his income money that was formerly applied to con­sumption—the process started by this saving will evidently be as follows. The former demand for consumption goods will be diminished by the amount of the savings, while on the other hand the demand for production goods (tools, machinery, etc.) will be increased by the same amount. The price-level of the former will tend to fall and that of the latter to rise. There follows a call for labour (which has become useless for the now reduced production of consumption goods) on the part of businesses making production goods. Saving, by diminishing consumption, will thus have served to set free labour forces that will henceforth serve to increase future production at the expense of the production of immediately consumable goods.

Nothing could be clearer than this. And it has, besides, the advan­tage of showing plainly that we are concerned here with community saving alone, and not with individual saving. Indeed, if we take the individual point of view only, we see at once that the savings of some may be consumed by others. The thrifty man who buys a farm, a factory, or some stocks or shares does not know whether the seller is not going to consume at once the money he receives. So these transactions do not necessarily increase the capital of the community. The best-known example is that of a State which consumes the savings of the public in war expenditure. There is no saving for the community unless, when account is taken of the consumption by some people of the savings of others, there remains an actual balance which is invested in new undertakings or in improvements of old ones. It is this actual balance, thus invested, to which the term 'savings' is applied in political economy, and to which Bohm-Bawerk's considerations apply— considerations which are very pertinent but far from exhausting the complexities of the subject.

In the mechanism here described everything turns, in fact, on the hypothesis of non-consumption, or, if preferred, reduction of consump­tion, by the saver. Now, this reduction is in very many cases useless. We have only to imagine a saver receiving an addition to his income and deciding not to consume it but to procure with it some machines that are indispensable to increase his future production. He has then no need to reduce his normal consumption. As a matter of fact, in the economic world of to-day, savings are mainly drawn from additions to income. It is the increased profits of an industry or an individual that are set aside and invested, without any diminution in the personal consumption of the saver or of society as a whole. From this follows the important consequence that the fall in the prices of consumption commodities, which was just now considered as bound up with the phenomenon of savings is in no way necessary, since the demand for these commodities has not changed. It may even have increased, if only a part of the additional income is saved, the rest being spent in the purchase of consumable products. And that is what most frequently happens.

In periods of a general rise in incomes the increase in saving is accom­panied by a general increase in consumption. Even in periods when incomes in general are falling, many communities or individuals who have remained prosperous continue to save without thereby reducing their consumption. The reductions in consumption that we see, therefore, are due to the general fall in consumable incomes of the rest of the population, and not to the fact that some people save. We cannot, then, accept without serious limitations the general proposi­tion enunciated by Keynes in his General Theory of Employment, Interest and Money (and asserted by many other writers as an obvious truth): "It is common ground that increased individual saving will cause a fall in the price of consumers' goods" (p. 192). On the contrary, an increase in saving, or its persistence even in a period of diminishing incomes, contributes to the fall in price of consumable commodities only on the hypothesis adopted by Bohm-Bawerk, which is the least often realized in the ordinary course of events.

But still further modifications are needed in a theory constructed exclusively for cases where all economic forces are employed and nominal incomes are stationary. Even if savings are deductions from the normal consumption of the savers, even if for them they mean retrenchment, it does not follow that for the community this deduction will involve a reduction in consumption. We have only to imagine— what is, in fact, the commonest case—that the community includes some unemployed workers. In this case the new savings, by providing an income for workers who hitherto had received none, are spent on consumption. The choice of commodities consumed is altered, but the aggregate demand for consumption products, measured in pur­chasing power, remains unchanged. At the beginning of a boom period, when unemployment, born of the preceding crisis, is generally fairly widespread, the effect of increased saving is simply to set to work a larger number of workers and thus to provide them with incomes, and to compensate, by their demand for consumable goods, for the reduced demand of the savers. That is undoubtedly the case that Adam Smith had in mind when he said, in what seems at first sight a rather paradoxical statement, "What is annually saved, is as regularly consumed as what is annually spent."

There is no need even to adopt the hypothesis of unemployment. The mere growth of population throws on the market every year a new generation of workers, whose numbers depend on the birth-rate. These new teams of workers, hitherto without any income of their own absorb the new savings, and while creating new capital prevent the demand for consumable goods from being restricted. Let us imagine a still more familiar hypothesis, large-scale immigration, such as used to take place into the United States, or a large import of foreign labour, as in France. It is obvious that in a great many circumstances new saving does not result in any diminution of the aggregate demand for consumption products, but merely changes the nature of the things demanded.

Such are the immediate effects of an increase in saving at the moment when it takes place—that is to say at the time of that change in demand which is its first manifestation, and which must not be confused with a reduction in demand. But this is only the first phase, as has been so well stressed by Bresciani-Turroni. When saving has attained its object—i.e., when savings have been transformed into more powerful or more numerous machines which serve to multiply consumption commodities—then these commodities are in more plentiful supply on the market than before. In this new phase a fall in prices is inevi­table unless other circumstances counteract it, such as a greater abundance of gold or paper money. The deep and lasting effect of an increase in annual saving, or simply of its maintenance at the normal level it has reached at a given moment, is a fall in the prices of consumable commodities and services "as a natural consequence of their greater abundance." That is the most certain result of an increase in general productivity due to saving. Whereas in the first phase, the creation of new savings, it is the alteration in demand that interests the economist, in the second, when saving has attained its end (the increase of production), it is the supply. This has been well brought out by Bresciani-Turroni in a noteworthy article in Economica, in which he sums up with admirable conciseness the fundamental notions that make up the modern theory of saving. Now, recent dis­cussions on the influence of saving on crises are concerned essentially with the phenomena of the first phase—those that result from the new direction given to demand—and neglect those of the second phase which result from the increase in supply.

If we take both phases into account, we see how hard it is then to follow in actual events the effects of a relative increase or decrease in saving. For all the effects of the first phase are constantly being joined by those of the second, which either counterbalance or reinforce them. In our industrial communities there is a continual creation of savings. Normally part of the net income is saved and only the rest is consumed: it is precisely this that characterizes an economically progressive community, according to the plan set forth long ago by Walras. Can we believe that the fluctuations of savings around: his normal point are sufficient to start a crisis? Entrepreneurs who extend their produc­tion (e.g., the makers of machines) reckon that normally the weavers or spinners who use the machines will also extend their production of cotton or cloth by saving part of their profits with which to buy more machines. They may be wrong, just as the weavers or the dress­makers may be wrong in counting on the consumers of cloth and dresses to increase their purchases. Have the mistakes of the machine-makers more to do with starting crises than those of the makers of cloth or dresses? There seems no reason why it should be so. All we can say is that the more far removed the manufactured products or the extracted raw materials are, in the chain of their successive transformations, from the finished product, the more is it possible for mistakes in forecasting, if there are any, to continue before the entre­preneurs are enlightened by the eventual sale or non-sale of that finished product.

So a shifting of expenditure from investment to consumption, or vice versa, will by itself produce hardly any lasting effects. If, for a given aggregate income, the demand for consumable products is increased1 their prices will rise, and there will be increased profits for the industrialists who make them, and they will find it advantageous to invest these profits in new manufactures. In other words, savings and investment will increase from the very fact of the increased demand for consumable products. And conversely, if it is the savings that have increased, and investments along with savings, then the new invest­ments will result in an additional supply of consumable goods, their prices will fall, and this will stimulate purchases by consumers and again reduce that part of income that is invested. Thus an automatic correction comes into play, in whichever direction the mistake is made.

We can but agree with the conclusion reached by Bresciani-Turroni, as follows:
The deeper we attempt to look into the saving process, the more this latter reveals to us unexpected intricacies and manifold forms. This process is by no means so simple and smooth as the conceptions still prevailing not many years ago among economists would have us believe.

And Divisia, for his part, asserts that "it is clear that if the growth of saving has not itself, strictly speaking, an actually contradictory character, nevertheless the analysis of its manner of working is full of pitfalls."2 In particular, adds Bresciani-Turroni, if you wish to use a theory of saving to explain crises, a careful distinction must be made according to whether or not the increase in saving takes place at a time of'full employment.' If there is unemployment, an extension of production may continue for a considerable time without leading to any reduction in the making of consumable commodities, for the savings provide incomes for workers hitherto without wages, and these incomes they hasten to consume. If, on the other hand, the increase in savings comes at a time when there is no unemployment—which is the case when a boom is at its height—the result will be to withdraw workers from the manufacture of consumable commodities and attract them into that of production goods. This gives rise to a reduction in the quantity of consumption commodities, followed by a rise in their prices and a nominal rise in wages, with a consequent reduction in profits in all industries with a tendency for production to slacken.

There we have the germ of an interesting explanation of crises. The creation of savings at the end of a boom would account in particu­lar for what was observed in Germany in 1924 by Bresciani-Turroni, a great authority on German economy,3 that at the moment when the 1924 crisis broke out it was consumable goods that were in short supply and whose price rose. This is in direct contradiction of the opinion of those economists who think that crises originate in a fall in the prices of consumption products, but it contradicts also the idea that they originate in a scarcity of capital. Their origin would lie rather in a shortage and dearness of labour.

Bresciani-Turroni, however, does not believe that these transfers of savings (whose effects correct themselves) are sufficient to start a crisis. His view is that crises would not occur without excessive credit. Industrialists very often start new business in production goods with the help of credit, which they count on being able to pay back with future savings. But if they are mistaken about the influx of future savings (and nothing is easier), the temporary expansion of production born of credit will be faced at a certain moment by an insufficiency of savings. If this mistake is widespread, a crisis may be the result. It thus has its origin in excessive credit, not excessive saving. It is obvious that the mechanism of saving becomes difficult to grasp when we try to follow it in detail.

The most certain conclusion of the general theory of saving is that by the increase of productive forces it tends to increase the quantity of products and services and consequently—in the absence of any exterior causes operating in the opposite direction—to lower their prices. The real income of the community is increased, and the mass of products and services to be shared between its members grows unceasingly, but the price-level falls. On the whole its effects are beneficial and bring increased well-being to every one.