value. The composition of this rate can be seen at a glance. It is made up of imports of approximately $7,000 worth of candied apricots having a converted rate of 23.7 per cent; of about $3,800 worth of candied prunes having a converted rate of 24 per cent; of pears to the amount of about $4,300, with a converted rate of 22.9 per cent, and so on for the other items named. Because of the relatively heavy weight of "other fruit" in the table the average for the entire duty bracket is almost the same as the converted rate for "other fruit."

CLASSIFICATION OF IMPORTS

The question of classification of imports by customs officers is involved in some of the tabulations presented in the report. One of the first steps in the valuation study was to determine the principal articles imported under each of the rates of duty to be converted. This information was obtained largely through customs officers. Because of the usual and normal amount of litigation which occurred during the valuation period the classification of some articles covered by this report is still in question, and it may ultimately develop by decisions of the courts that they were improperly classified; that is to say, an item which has been tabulated and the converted rates shown under a particular paragraph may ultimately be classified by court decision in another paragraph. Some of the valuation data, therefore, under a specified duty may not be strictly accurate because of the inclusion of articles not properly dutiable under that bracket. However, because of the uncertainty of the outcome of several pending cases, the commission has been constrained to abide by the classification determined by the customs officers in the first instance.

An illustration of the foregoing point is that of unfinished glass hypodermic syringes, which were classified by customs officers under paragraph 218 (a) at 85 per cent. The Customs Court ruled them to be properly dutiable under paragraph 359 at 70 per cent. The Court of Customs and Patent Appeals reversed the judgment of the Customs Court and sustained the classification of the collector. Our report treats of this article under paragraph 218 (a) at 85 per cent. It has not been found practicable to call attention to all instances of this sort. Generally speaking, pending litigation was not a serious problem in the valuation study.

PLANS VERSUS ACCOMPLISHMENT

The foregoing discussion has outlined the provisions of the law with respect to the valuation investigation, and has given in some detail the methods used by the commission in obtaining and summarizing the field data. The outline of the methods used may give an impression of adequacy of the results obtained beyond that justified by the facts in particular cases. The commission's plan of obtaining tabulations for representative importers for representative samples for representative shipments during the valuation period obviously could be carried out only where the number of importers and the volume of business was sufficiently large to give an adequate basis for the sampling process. Where the number of importers was quite limited and imports sporadic and irregular the facts tabulated necessarily fall short of the adequacy assumed in the commission's plan. The extent to which particular tabulations are adequate or inadequate for valuation purposes has not been appraised by the commission except as far as the published results themselves bear upon the

point. If imports are small or negligible, or if only a few of many companies are represented in the tabulations, or if there is a wide variation in the converted rates for different items dutiable under the same paragraph, the summary tables disclose these facts without comment by the commission.

4. SPECIAL PROBLEMS

In carrying on the work of tabulating and summarizing the valuation data a number of special economic and statistical problems were encountered which are of sufficient importance to be discussed in some detail. The problems encountered were of two kinds: First, those which could be solved by adequate statistical research methods, and second, those which were objective facts about which nothing could be done except by way of reflecting the conditions in the tabulations for the consideration of the Congress.

Several problems of the first type have already been referred to, namely, those relating to the representativeness and adequacy of the data tabulated. One of these, for example, has to do with "weighting" the information obtained for individual commodities or groups of commodities in making up combined averages.

Among the other type of problems, namely, objective conditions in the import trade, about which nothing can be done beyond tabulating the facts as found, the most important one grows out of the variations in the margins or spreads between foreign and domestic value for different products dutiable under the act of 1930 at a single ad valorem rate of duty. Different margins between foreign and domestic value give different converted rates on domestic value for a single duty on foreign value.

VARIATIONS IN CONVERTED RATES BECAUSE OF DIFFERENT KINDS OR

GRADES OF PRODUCTS

The margin problem, to give it a short name, takes several forms. The simplest form occurs when two grades of the same product, dutiable at the same rate of duty, sell in the United States at different percentages of mark-up over foreign value. If, for example, grade A of a product, dutiable at 40 per cent, costs $1 and sells for $2, the converted rate on domestic value would be 20 per cent; whereas if grade B of the same product, dutiable at the same rate, costs 50 cents and sells for 75 cents in the American market, the converted rate would be 26% per cent. Under these conditions an average rate on domestic value would yield the same amount of duty on future imports as the present rate only if the relative quantities of the two grades were the same in the future as during the period covered by the investigation.

VARIATIONS BECAUSE OF DIFFERENT CHANNELS OF TRADE

Another form of the margin problem has been referred to briefly above in connection with the discussion of sales through different channels, and of the difference between sales from stock and on import order. If one importer transacts his business upon the basis of unbroken shipments made after orders are taken, the equivalent converted rate of duty on domestic value may be entirely different from that of the importer who sells in broken lots from stock. This again is an objective fact which can not be eliminated by methods of averaging, etc.

VARIATIONS BECAUSE OF CHANGES IN DUTIES

A special form of the variations in margins between foreign and domestic values, and therefore in converted rates, is found when products were dutiable during the valuation period under the tariff act of 1922 at different rates of duty than under the tariff act of 1930. If a product costing $1 during the valuation period was dutiable at 40 per cent, the domestic selling price was fixed upon that basis. Presumably the selling price would have been different if the product had been dutiable at 60 per cent. Under such conditions the conversion is made upon the basis of the domestic selling price when the 40 per cent duty was in effect. To the extent, therefore, that the increase in the rate of duty from 40 to 60 per cent under the tariff act of 1930 would affect the domestic selling price of the article the converted rate is not representative of prevailing conditions under the tariff act of 1930. In those cases in which the rate in the tariff act of 1930 is higher than under the act of 1922, the converted rates may be higher than if the rates in the tariff act of 1930 had been in effect. When the rates in the tariff act of 1930 are lower than in the previous tariff act the converse would probably be true.

VARIATIONS BECAUSE OF TREASURY AND COURT DECISIONS

Another form of the margin problem occurs when by reason of Treasury or court decisions the classification, and therefore the rate of duty upon products, is changed during the valuation period. In such cases part of the sales data may be upon the basis of a 25 per cent duty, for example, whereas another portion may be upon the basis of the revised classification.

VARIATIONS BECAUSE OF TIME

Another important matter about which nothing can be done by way of statistical presentations for the valuation period is that of the possible change in margins between purchase and sales prices of imported products since the valuation period ended June 30, 1929. If the percentage spread between foreign and domestic prices has changed in recent years, the tabulations for the valuation period are not representative of present conditions.

In connection with all margin problems, however, attention is called to the fact that high or low prices as such are not significant. If in 1929 a product was purchased for $1 and sold for $2, after duty and other charges were paid, the conversion rates would be the same as if the product were purchased at another time at 50 cents and sold for $1. To the extent that importers operate upon an approximate fixed percentage for overhead and profit above landed cost, to that extent will conversion rates be approximately the same regardless of absolute price levels. A change in the duty under the act of 1930 as compared with 1922 would, of course, change the margin if operations were upon the basis of fixed percentages of landed cost.

PROBLEMS ILLUSTRATED

A detailed examination of many of the summary tables published in the report is necessary for sound conclusions as to the significance of certain of the problems discussed above. Because of their complexity, however, it seems desirable to illustrate some of these problems

in further detail in the introductory section of the report, and to that end some of the summary tables will be reproduced and commented upon. It should be emphasized that these tables are reproduced for the purpose of illustrating particular points, and not for the purpose of giving or inferring conclusions about the whole question of valuation.

Methods of weighting.

The first point to be illustrated in the order named on a previous page is that of the method of "weighting" the valuation data. A reference to the sample summary table for candied and glacé fruits on page 12 will show that the average rate of 24.5 per cent for the duty bracket as a whole is composed of the converted rates for apricots at 23.7; prunes at 24.0; pears at 22.9; figs at 24.7; berries at 19.6; and "other fruit" at 24.8. Incidentally, as stated in the previous discussion of the table, the imports of "other fruit" are so much larger than those of the particular kinds named that the "weighting' process gives an average rate for the entire bracket approximately equal to the rate for "other fruit."

It will also be recalled that the footnote to this table states that "converted rates are weighted averages, and therefore can not usually be calculated directly from the unit prices and duties shown." A division of the unit domestic value into the calculated unit duty indicates to what extent the "weighted" averages differ from the converted rates obtained by the simple process of division. By dividing the unit domestic value into the duty the converted rates become for candied apricots 22.6 per cent instead of the "weighted" average of 23.7; prunes become 24.4 instead of 24; pears 23.8 rather than 22.9; figs 24.6 instead of 24.7; berries 19.2 in place of 19.6; and "other fruit" 23.5 instead of 24.8.

Inasmuch as the imports of "other fruit" were sizable items for a number of companies, the weighting problem will be further illustrated by the details back of the figures given in the summary.

In the first place the figure of 24.8 given in the summary table for "other fruit" is composed of a weighted average of the conversion rates for four 6-month periods within the 2-year valuation period. For the six months ended December, 1927, $18,130 were imported by the importers covered and had a converted rate of 24.4 per cent. For the six months ended June, 1928, $3,271 had a converted rate of 24.6. For the six months ended December, 1928, $22,054 had a converted rate of 25.2, and for the six months ended June, 1929, $3,040 had a converted rate of 24.2. The weighted average of these figures is shown in the summary, i. e., 24.8 per cent.

The figures given for each of the four 6-month periods, however, are themselves weighted averages of more detailed data. For example, the weighted average for the six months ended December, 1927, of 24.4 is composed of 17 conversions of different grades and assortments of "other fruit" imported by five companies. The trade names of these fruits and the companies can not be given without disclosing confidential information, but the converted rates by numbered grades can be given for the purpose of illustrating the variability in the rates which make up the average. The value of imports of each grade during this 6-month period is also given for use as a weighting figure. Upon this basis the converted rates and the weighting values are as follows:

The converted rates for individual assortments among the several companies are reasonably consistent with the exceptions of one rate at 7.9 per cent; two others over 12 per cent; and one other at about 15 per cent. Most of these inconsistent rates do not have heavy weights in the calculations. The 7.9 per cent rate has a weight of about $273 out of the total of $18,000. One of the 12 per cent rates has a weight of $77 and another about $136. The 15 per cent rate has a somewhat larger weight, namely, $968 out of approximately $18,000.

These

The above figures by numbered grades of candied fruit falling under the general name of "other fruit" are shown regardless of whether they were imported by one or more companies. Average converted rates by companies are also of some significance. are as follows for the same 6-month period ended December, 1927. Other fruit; converted rates by companies-35 per cent rate converted to domestic basis Company No.:

3.

Per cent

24. 5

20. 8

27.0

9.8

4.

5--

Weighted average_

15. 2

24. 4

The foregoing discussion of the details of weighting has referred only to a single line out of six lines in the summary table, namely, "other fruit" having a converted rate of 24.8. Five additional items of candied fruit are shown in the summary, namely, apricots, prunes, pears, figs, and berries. The next step in the weighting problem is to combine the rates for each of these items into an average for the final rate. As explained in the footnote to the table, this average is weighted by the total value of the imports of importers covered; that is to say, out of a total value of $64,234, the figure of 24.8 per cent for "other fruit" has a weight of $46,495; the 19.6 per cent for berries has a weight of $531, and so on for each of the items named. It so happens that in the particular illustration the dutiable value of imports tabulated was 100 per cent of the total imports of these products of the importers covered. In a number of cases, however, as has been pointed out, it was not practicable to tabulate all the imports of each. importer for the products selected for the entire valuation period. In some cases only 40, 50, or 60 per cent of the imports were tabu