just touched, that savings are supplied largely by people who are relatively rich, and who become richer when the rate of interest rises.  For at this point it is necessary to be careful.  It is easy to slide from the above conclusion into an argument of the following kind.  A higher rate of interest leads to more saving; it is thus necessary to evoke more saving; it is thus required as an incentive to induce people to incur the sacrifice of waiting; this sacrifice represents the “real cost” for which interest is paid.

This terminology of incentive, inducement and sacrifice is of very dubious validity.  A rich man, who is made richer by a rise in the rate of interest, will probably save more, but it will be rather because he has become richer than because he is tempted by the higher rate:  and the less we talk about his sacrifice the better.  Nor is it clear that the attraction of a high rate of interest is an operative factor on the mind of a man to whom saving means a real sacrifice of immediate comfort or enjoyment.  Certainly it is only one among many factors, and seldom an important one.  A really poor man will think not so much of the annual income which will accrue from his savings, as of the capital sum upon which he or his family can fall back if a rainy day should come.  And for this purpose he might save as much as he saves now, even if there were no interest to be obtained thereby.  He might even be prepared to lend what he had saved, at least to banks (a deposit with a bank is in effect a loan), for the mere advantage of safe custody.  The people who save rather for the sake of the capital sum that can be realized than for that of the annual interest are very numerous, and probably include many men in receipt of quite considerable earned incomes.  Moreover, those who consider mainly the future annual income which their savings will yield them, are usually more concerned with its absolute amount than with the ratio it bears to the amount they must save in order to acquire it.  For this reason, as has been often recognized, they may save less when the rate of interest rises, since a smaller quantity of savings will insure to them the future annual income they desire to obtain.  There is no need to be dogmatic upon any of these points.  The psychology of saving is both complex and obscure.  Our conclusion must be the negative one that we have insufficient evidence to warrant the assertion that the particular rate of interest which happens to prevail is a measure of the sacrifice involved in saving, even in the case of what we might regard as the “marginal saving.”  And, if we cannot assert this, we must be careful not to assume it as the basis of other arguments, or as part of a general analysis of price or exchange value.