displace the mechanical and industrial processes. Not more mechanical and artisan labor is applied because of marketing activities, but less.

In the aggregate, therefore, all gainful ways of employing labor that are not also socially productive are waste and nothing more; the social dividend suffers with no diminution in the number of claimants against it. But in the competitive process the distribution of the diminished product is also modified. The increased emphasis for gain-getting purposes on the activities displacing mechanical and industrial labor, redistributes the total share to be allotted to the labor factor to the marked disadvantage of artisan wages. There is, that is to say, a distributive change in the relative shares allotted to the different grades and varieties of labor, a redistribution within the labor group. As the substitute and competing laborers get more, the artisans and mechanics get less. Not merely do these last share in the loss attending a diminished social income of concrete things, but also they are subjected to a disproportionate share of the loss. They are the chief sufferers through the entire process the distributary legatees of the cumulative deficit.

All this is not so much a novelty in distributive theory as, in the main, merely a neglected application of principles long established. Where certain of the factors in a joint process are becoming more expensive the others suffer. Those becoming relatively plenty become also relatively cheap. But it has not been quite clearly recognized that in the processes of competitive rivalry, and by the standards of these processes, there are other than mechanical and industrial factors. There are business factors, equally productive by the test of gain. Advertising, for example, is such a factor. Salesmanship in the large is a group of such factors.

Thus it must be recognized that all parasitisms, predations and wastes bear adversely on the distributive process in the large, because they bear adversely on the total of goods to be distributed. Incomes have, then, to suffer in the average, wages with the rest. But some of these adverse influences bear in especial degree on wage earners in the spending of their incomes. Others impose in addition on a particular class of laborers, the artisan and mechanic class, certain peculiar distributive disadvantages. The theory of wages needs in general, then, supplementation to take account of the difference between competitive acquisition and social production. It needs to take special account of the peculiar bearing of certain gainseeking activities on the wages of artisan and industrial laborers, not merely (1) in what their dollars will buy after they are collected, but (2) on the number of dollars that are received.

H. J. DAVENPORT.

CORNELL UNIVERSITY.

THE TAXATION OF LUXURIES AND

THE RATE OF INTEREST

SUMMARY

I. Luxury defined as consumption which does not increase capacity for labor, 298. - Superfluities include consumption necessary for business connection, 300.- Short and long periods, 301. Capacity for labor varies with consumption of necessaries, 301. Classification of commodities, 302. II. Demand for necessaries inelastic because (a) their consumption benefits both present and future, 304. (b) Capacity for labor is limited, 306. - A tax on luxuries would divert labor to production of capital, also lessening demand, 307. — Labor would become more productive and real wages would rise, 308.- Wealth would be increased, 309; and transferred to people whose effective desire of accumulation is strong, 310. - III. Possible decrease in quantity of labor expended, 314. Self-development not labor, 315.Leisure, defined as time not devoted to labor, is either necessary or superfluous, 316. Annual product varies inversely with superfluous leisure enjoyed, and might be either decreased or increased, 318. Rise of wages might check fall of interest; convergence of altruism and foresight, 319.

I

THE proposition that " a tax for purposes of revenue should have the least possible prohibitive effect "1 seems fundamental to the system of laissez-faire, and has been accepted, expressly or tacitly, even by writers who in general regard that doctrine with disfavor. But such a proposition cannot be established unless not only some but all of the repressive effects which taxes may have are examined, and found on the whole to be undesirable. It cannot therefore be a bar to the consideration of a tax on luxuries. In the present essay an

1 E. A. Ross, "A New Canon of Taxation," Political Science Quarterly, December, 1892, p. 579.

attempt will be made to show that such a tax would have a tendency to reduce the normal rate of interest. To this task the discussion will as far as possible confine itself. The question of general expediency, which would seem to involve a careful comparison of incommensurable evils, must be understood to transcend the limits of this inquiry.

"Con

It is convenient to begin with a definition. sumable commodities," writes Adam Smith, are either necessaries or luxuries." If we adopt Professor Fisher's conception of income, and use the word consumption as the equivalent of what he calls "enjoyable objective services" that is, including labor which satisfies wants directly - we may say that all consumption must be either of necessaries or luxuries. Let us now define necessary consumption as that which increases the consumer's capacity for labor. Luxury then becomes definable as consumption which does not increase this capacity.

At the outset we notice that, owing to the division of labor, an individual cannot attain his maximum productiveness without coöperating with others. To be able to do this with advantage, he must in a measure enjoy their confidence and good opinion. That of possible employers or customers is of most importance, but a person cannot regard anyone's estimation of him as a matter of perfect indifference, as is shown by the pecuniary value occasionally attached to reputation, which, be it noticed, is not entirely dependent on the direct satisfaction which a good reputation affords, but in all cases bears some analogy to business connection. Now A's opinion of B partly depends on whether B's consumption, or expenditure, is sufficient to bring his standard of living into conformity with what A con

1 Wealth of Nations, Bk. V, chap. ii.

siders suitable to one in B's position. And A's ideas on that head are chiefly determined by what is consumed by others in circumstances similar to those of B. Thus, to attain his maximum productiveness, B must make his consumption conform to a purely arbitrary standard, and all expenditure within those limits is, from his individual point of view, in the strictest sense necessary.

The point of view of the private individual, however, must be distinguished from that of society as a whole. From the latter standpoint, it appears that part of every man's consumption is necessary in order that he may develop his productive powers as they are affected by his physical, mental, and moral qualities; and part is necessary merely because other people are accustomed to indulge in such expenditure. The latter part of consumption might be called relatively, or, if I might venture to adopt Dr. Marshall's term, conventionally necessary.1 From the social standpoint, however, such consumption is not necessary, but is purely a luxury, inasmuch as it does not, in itself, increase capacity for labor. The term superfluity might be used to denote consumption covered by the above definition of luxury from the social point of view, including both what an individual would regard as luxury, and that which is relatively necessary. This terminology would appear conformable both to economic authority 2 and to general

usage.

The term superfluity, then, is more comprehensive than the term luxury. The distinction between them, however, is not only vague and hard to draw, but is constantly changing. It is often said that the luxuries of one generation become the necessaries- that is, con

1 Principles of Economics, Bk. II, chap. iii, § 4. The term is there used in a sense which I believe will not be found to differ substantially from the above. See also Dr. Marshall's whole treatment of the subject of Necessaries.

See Sidgwick, Principles of Political Economy, Bk. I, chap. iii, § 3.