Theoretical accuracy in cost accounting, important for the manufacturer, is unnecessary for tariff purposes. Although a two per cent. error in costs may be objectionable from the standpoint of the manufacturer, it is of slight importance to the legislator who is hesitating between a duty of 25 per cent. and 35 per cent. Many of the technical difficulties of exact cost accountinginterest, the rate of depreciation, the allocation of overhead and general expenses, and the capital vs. revenue problem, need not be worked out to the limit of scientific accuracy. In the pulp and paper report of

the Tariff Board under the administration of President Taft, for example, an arbitrary figure for depreciation was adopted after careful consultation with two leading engineering concerns in that industry.10 The cost problem is, in fact, not so difficult as it seems to the expert, because, in the first place, a fairly accurate approximation is possible, and in the second place, an approximation is entirely adequate for tariff purposes.

The gathering of comparative cost data is really the function of an international commission such as will be discussed subsequently.11 But it is not beyond the ability of a domestic commission to get foreign costs sufficiently accurate for tariff purposes. The United States Tariff Commission has power to investigate the "conditions, causes, and effects relating to competition of foreign industries with those of the United States, including dumping, and cost of production."2 War conditions have prevented the Tariff Commission from mak

10Report on the Pulp and Newsprint Paper Industry'' (1911), pp. 24 and 73.

11 Chapter XVI.

12 See the Act creating the Tariff Commission, Sec. 704, printed in Appendix IV.

ing any extended investigations abroad, but some interesting comparisons are found in a study of the sugar industry.13

Sugar is essentially an agricultural product, and like other agricultural products is subject to the law of varying costs. In the domestic beet-sugar industry, as in the cane-sugar industry in Cuba, Hawaii, Louisiana, and Porto Rico, there are competing factories producing at widely varying costs per ton of sugar produced, and the forces of demand and supply, far from tending to reduce all of these varying costs to a common level, tend rather to increase the disparity between the highest- and lowestcost producers who can continue to compete in the same market. Leaving out of consideration a small percentage of abnormally high-cost concerns which the forces of competition are in fact tending to eliminate, there will, therefore, be found in each region a "marginal" firm whose costs are such as to leave it a bare living profit and many lower-cost firms that receive a progressively greater and greater profit, technically known as "economic rent." It is obvious, therefore, that it is only the marginal firms whose costs must be met by the prevailing price in order that they may survive. All of the others can suffer a greater or less diminution in price and still do a profitable business.

The table on page 124 shows the output and cost per ton of sugar produced for beet-sugar factories in the continental United States and for cane-sugar factories in Cuba for 1916-17.

These statistics are cited here merely to illustrate how costs may be of service in tariff making. The sugar

tariff should not be determined from this table.

The

13 United States Tariff Commission, "Costs of Production in the Sugar Industry."

OUTPUT AND COST OF PRODUCTION PER TON OF SUGAR PRODUCED FOR BEET SUGAR FACTORIES IN THE CONTINENTAL UNITED STATES AND FOR CANE-SUGAR FACTORIES IN CUBA, CLASSIFIED BY COSTS IN GROUPS OF FIVE FOR THE YEAR 1916-17

Before

*Cuban sugars are raw and beet sugar is granulated. being entered in the column the costs of the granulated beet sugar as computed from original data were reduced by the estimated difference in cost of producing raw and granulated sugar.

year 1916-17 was an abnormal one and the figures of a pre-war year should be considered also. They would show a wider difference between domestic and Cuban costs. We import from Cuba about one-half of our annual sugar consumption, and it is the cost of the Cuban marginal producer that is reflected in the New York price. Importations from other foreign countries are negligible, and although they would increase if Cuba's preferential

rate were removed, they would not be a more serious factor in competition than the Cuban sugar.

How the costs in the table are to be used in tariff making depends on policy. Representative factories in the United States and Cuba might be determined upon and the tariff fixed to cover their difference in costs with a fair margin. Or it may be decided what marginal American producer should in the circumstances be allowed to continue in business and the tariff fixed at such a point as will protect him. In the table the Cuban marginal cost ($78.65) lies between the costs of the eleventh and twelfth group in the American cost column. Assuming that the figures in the foregoing table alone were used for determining the tariff rates, if the present tariff were removed and the price fell to the cost of the Cuban marginal producer, the American factories in the first 11 groups would, other things being equal, be able to survive and the others would be eliminated. But other things would not be equal. If some American producers were forced out, new Cuban producers would take their place, the Cuban marginal cost would rise, and more American factories would be able to stay in business. Also if the tariff were removed and the price of sugar fell in the domestic market, its consumption. would increase and the fall in price would not correspond in amount to the duty removed. Other factors, however, should be considered likewise. Statistics, as has been said, are only of service in tariff making when used with judgment and commonsense. The tariff is not a problem in mathematics.

To obtain cost data from abroad is difficult but not impossible. Only in exceptional cases, it is true, is it possible to examine the actual cost sheets of foreign concerns, but an intelligent study by men trained in eco

nomics and accounting methods of the competitive facts in foreign countries will enable them to draw conclusions sufficiently accurate for tariff purposes. In almost every case information is available upon the sources and prices of raw materials, wage scales, and the general efficiency of labor, the amount of machinery used, the hours which employees work, and the general facts about overhead expenses.

One of the most successful studies of comparative costs of production was made by the Taft Tariff Board in studying the cost of producing cotton yarns in the United States and Great Britain.14 As an illustration of the results obtained, the comparison of yarn conversion costs15 in England and the United States is given in the table on the opposite page.16

17

Elaborate comparative data on cost of production were also obtained by the Taft Tariff Board on wool growing and on the manufacture of tops, woolen and worsted yarn and cloth, and ready-made clothing. The application of these data to wool legislation proposed in the second session of the Sixty-second Congress is made in an article reprinted as an appendix to this volume.18 To take an example here: Sample No. 28 on which the Tariff Board obtained comparative costs was a man's fancy woolen suiting weighing 13 ounces per yard. The costs are on the basis of one pound of cloth, and the differences between this country and Great Britain are given. The difference in the conversion cost for mak

14 Cotton Manufactures," H. R. Doc. No. 643, 63d Cong. 1st Sess. (1913), vol. ii, pp. 398-426.

15 Conversion cost is the cost of converting the raw material into the finished product and does not include cost of raw materials. 16 Op. cit., p. 414.

17 Wool and Manufactures of Wool" in four volumes, H. R. Doc. No. 342, 62d Cong., 2d Sess. (1912).

18 Appendix V.