The Economics of Welfare
§ 1. THE economic welfare of the country is intimately associated with the size of the national dividend, and changes in economic welfare with changes in the size of the dividend. We are concerned to understand, so far as may be, the nature of these associations. To this end an essential preliminary is to form clear ideas as to what precisely changes in the size of the dividend mean. It will be convenient, in the first instance, to postulate that the size of that group whose dividend we are studying remains unchanged.
§ 2. The dividend is an objective thing, consisting in any period of such and such a collection of goods and services that flow into being during the period. Since it is an objective thing, we should naturally wish, if we were able, to define changes in the size of it by reference to some objective physical unit, and without any regard to people's attitude of mind towards the several items contained in it. I do not mean that changes in public tastes would be thought of as incapable of affecting the size of the national dividend. They are obviously capable of affecting it by causing changes in the objective constituents of the dividend. I mean that, given those objective constituents, the size of the dividend should depend on them alone, and not at all on the state of people's tastes. This is the point of view which everybody intuitively wishes to take.
§ 3. If the national dividend consisted of one single sort of commodity only, there would be no difficulty about this. Everybody would agree that an increase in the size of the dividend should mean an increase, and a decrease a decrease, in the number of units of this commodity. In like manner, if the dividend consisted of a number of different commodities, but the quantities of all of them always varied in equal proportions, there would be no difficulty. The dividend would at any time consist of a certain number of complex units, each of them made up of so much of each commodity, and increases and decreases in the dividend would mean increases and decreases in the number of these complex units.
§ 4. If the national dividend consisted of a number of different sorts of things, the proportion between which was not fixed, but some pre-established harmony made it impossible for the quantity of any one of them to diminish when the quantity of any other was increasing, we should no longer be able to say that the dividend at any moment consisted of such a number of units and at another moment of such another number of units. But we should still always be able to determine by a physical reference whether the dividend of one moment was greater or less than the dividend of another moment: and this, for many purposes, would be all that anybody would need.
§ 5. In actual life, however, the national dividend consists of a number of different sorts of things, the quantities of some of which are liable to increase at the same time that the quantities of others are decreasing. In these circumstances there is no direct means of determining by a physical reference whether the dividend of one period is greater or less than that of another; and it becomes necessary to seek for a definition along other lines. Plainly the definition chosen must be such that, supposing the dividend consisted of one sort of thing only, we should always be able to say that an increase in the quantity of this thing constituted an increase in the size of the dividend. A definition that did not admit of this would be paradoxical. From this starting-point we are led forward as follows. Considering a single individual whose tastes are taken as fixed, we say that his dividend in period II. is greater than in period I. if the items that are added to it in period II. are items that he wants more than the items that are taken away from it in period II. Passing to a group of persons (of given numbers), whose tastes are taken as fixed and among whom the distribution of purchasing power is also taken as fixed, we say that the dividend in period II. is greater than in period I. if the items that are added to it in period II. are items to conserve which they would be willing to give more money than they would be willing to give to conserve the items that are taken away from it in period II. This definition is free from ambiguity. However the technique of production has altered,—though it has become more costly to make one thing and less costly to make another, though it has become possible to make some entirely new things and at the same time impossible to make some things that used to be made before,—it can yield one conclusion, and one only, as to the effect on the size of the national dividend of any change in its content that may have taken place. If, then, tastes and the distribution of purchasing power were really fixed, there would be nothing to set against the advantages of this method of definition. It would be the natural and obvious one to adopt.
§ 6. As a matter of fact, however, tastes and the distribution of purchasing power both vary. The consequence of this is that our definition leads in certain circumstances to results which, in appearance at least, are highly paradoxical. Thus in period I. tastes are such and such, and in period II. they are something different; in period I. the dividend is a collection C1 and in period II. a collection C2. It may happen both that the group with period I. tastes would give less money for the items added in period II. than for the items subtracted in that period, and also that the group (of equal numbers) with period II. tastes would give more money for the items added in period II. than for the items subtracted in that period. In this case our definition makes C2 both less than C1 and also greater than C1; which is a violent paradox. The only escape from this is to admit that, in these circumstances, there is no meaning in speaking of an increase or decrease in the national dividend in an absolute sense. The dividend decreases from the point of view of period I. tastes, and increases from the point of view of period II. tastes; and there is nothing more to say.*56 It is easy to see that the same paradox may arise, and the same solution be forced upon us, when the distribution of purchasing power alters between period I. and period II. Here again we can only speak of an increase (or decrease) in the size of the dividend from the point of view of period I. distribution or from the point of view of period II. distribution: we cannot speak of an increase or decrease in any absolute sense.*57
§ 7. We are thus confronted with the awkward fact that there are likely to be certain changes in the constitution of the national dividend, of which it is not possible to say that they are either increases or decreases in an absolute sense. Plainly there is serious objection to a definition which leads to this result. On the other hand, though it will rarely happen that a modification of the dividend, which constitutes an upward (or downward) change of so much per cent from the point of view of period I., will constitute an equal percentage change from the point of view of period II., if between these two periods tastes or distribution have altered, yet it will, we may reasonably expect, usually constitute a change in the same direction from the point of view of period II. Most causes, in short, will increase the dividend from both points of view or diminish it from both points of view. Usually, therefore, we can say, without circumlocution or complicated reference to two points of view, that a given cause either has or has not increased the size of the national dividend. The defect in our definition is thus not a fatal defect. Moreover, continued reflection fails to reveal any other definition that is not even more defective. In spite, therefore, of all that has been said, I propose, for the purposes of this volume, to define an increase in the size of the dividend for a group of given numbers as follows. From the point of view of period I. an increase in the size of the dividend is a change in its content such that, if tastes in period II. were the same as those prevailing in period I. and if the distribution of purchasing power were also the same as prevailed in period I., the group would be willing to give more money to conserve the items added in period II. than they would be willing to give to conserve the items that are taken away in period II. Waiving the distinction, discussed in Chapter II., between desire and the satisfaction that results when a desired thing is obtained, we may state the above definition alternatively thus. From the point of view of period I. an increase in the size of the dividend for a group of given numbers is a change in its content such that, if tastes in period II. were the same as those prevailing in period I., and if the distribution of purchasing power were also the same as prevailed in period I., the economic satisfaction (as measured in money) due to the items added in period II. From the point of view of period II. an increase in the dividend is defined in exactly analogous ways. From an absolute point of view an increase in the size of the dividend is a change which constitutes an increase from both the above two points of view. When, of two dividends, one is larger from the point of view of one period and the other from that of the other, the two are, from an absolute point of view, incommensurable.
§ 8. Hitherto we have been concerned with groups containing equal numbers. As between groups of different sizes a direct comparison of dividends would be of little service. We may, however, in imagination reduce the numbers—all classes of persons being treated equally—in the larger group in the proportion required to make it equal to the smaller group, and reduce its money income in an equal proportion. The dividend of the group so obtained may then be compared, on the lines of the preceding analysis, with that of the smaller group. The result is roughly a comparison of the per capita dividends of the two original groups.
Notes for this chapter
An exactly analogous difficulty emerges when we attempt to compare the size of the national dividend, as defined above, in two countries. Thus, if the German population with German tastes were given the national dividend of England, they might get less economic satisfaction than before; while, if the English population with English tastes were given the German national dividend, they also might get less economic satisfaction than before. The proposed definition would, in these circumstances, compel us to say both that the English dividend is larger (from the English point of view) than the German dividend, and also that the German dividend is larger (from the German point of view) than the English dividend. It may be added, though the point is not strictly relevant, that differences in comparative tastes between the people of two countries can sometimes, though not always, be detected by statistical methods. For example, Germans before the war would not eat mutton though it was a penny cheaper than pork, while Englishmen ate it readily (Cd. 4032, pp. xlviii and xlix). Again Germans eat rye bread, whereas English people eat white bread. We know that this is not due merely to the fact that rye bread is relatively cheap in Germany and that Germans are poorer than Englishmen, because, if it were cheapness alone that was responsible for the consumption of rye in Germany, there would presumably be a higher consumption of white bread among better-to-do Germans. This, however, is not found. Hence, we may legitimately infer that Germans have a taste for rye bread, as against wheaten bread, different from the English taste.
Cf. Dr. Bowley's observation: "The values included in incomes are values in exchange, which are dependent, not only on the goods or services in question, but also on the whole complex of the income and purchases of the whole of a society... The numerical measurement of total national income is thus dependent on the distribution of income and would alter with it" (The Measurement of Social Phenomena, pp. 207-8). Cf. also stamp, British Incomes and Property, pp. 419-20.
Part I, Chapter VI
End of Notes
Return to top