1 This is an unweighted average of monthly cash fares for street railway or bus service in cities of 25,000 or more population. There is a variation in the number of cities; in 1932, 320 cities were included, while in 1936 there were 268. The data are compiled by the American Transit Association, formerly the American Street Railway Association, and are published in the annual supplement of the Survey of Current Business, U. S. Department of Commerce, Washington.

An average, including all cities of 50,000 or more population, except New York. The average is weighted according to population, the 1930 census having been used back to 1926. Prepared by Mr. Richev, an engineer, this series appears in Moody's Public Utilities. It is here changed from Richey's base, 1913, to a 1926 base.

3 These data are taken from Moody's Public Utilities. The statistics for 1917, 1922, 1927 and 1932 are from the Census of Street Railways; the statistics for the remaining years were obtained from the American Transit Association.

TABLE V.-Index numbers of residential price of typical amounts of manufactured gas in 25 cities 1913-26;1 total and per customer residential consumption of manufactured and mixed gas in identical cities, 1929-36.2

1 Rates were compiled from data supplied by the Wisconsin Public Service Commission, representing a weighted average of 51 cities and three types of service. Telephone usage data were computed from number of telephones owned by and connecting with the Bell system and annual population estimates, both of which are published in U. S. Department of Commerce Statistical Abstract annually.

Price data compiled from Federal Power Commission, Trends in Residential Rates from 1924 to 1936, Washington, 1937. Total consumption data from Edison Electric Institute, Statistical Bulletin No. 4, New York, 1937, for years 1926-36; for earlier years data are those appearing in Moody's Public Utilities, the Electrical World being credited as the source. Average consumption data were compiled from the Edison Electric Institute, cited above.

The price data for the years 1924-34 are as of October 1. The 1935 price is an average of the prices for Jan. 1, 1935 and Jan. 1, 1936; the 1936 price is that for July 1, 1936.

The data are averages weighted according to population, for 150 utilities operating in 132 cities, which include 88 percent of the population in the cities of 50,000 or more persons. The cities included contain 46 percent of the population living in communities of 250 or more persons. Page 21 of the report cited above: "When two or more utilities serve in the same city, the rates of each of the utilities for which this city is the 'major' city served by the utility are shown, and as a result the bills of 150 utilities are reported for the 132 cities."

The price data were presented by Vice President W. G. Vincent of the Pacific Gas and Electric Company in an article entitled "Rate Reductions," Edison Electric Institute Bulletin, June 1936. The 51 cities are those that were employed by the Bureau of Labor Statistics in its computation (discontinued November 1934) of the cost of most popular consumption amounts of electricity, appearing in its retail price bulletins; the method followed duplicates that of the Federal Power Commission in its computation of typical bills.

In the following tables, the relation between tariffs and wholesale prices is set forth in the effort to throw light on the effects of tariffs upon price behavior. In presenting the rates contained in the tariff schedules, note should be made of factors other than the height of the protection afforded by the rates which may affect the importation of goods.

It is convenient to group the more disguised protective measures which act along with and in addition to the custom duties themselves, under the title of "administrative" protection.2 Here we find that the administration of customs, formalities, rules regarding marks of origin, veterinary and sanitary regulations, food and drug regulations, laws concerning the assessment of ad valorem duties, and the process of classification of imports all contribute to the national policy of controlling the flow of goods in international trade. Instances can be cited where the charges for the formalities at the custom border have more than doubled the ad valorem equivalent of the statutory tariff rate.3 Also, many times in many different countries the shibboleth of public health has been used to justify actual prohibition or discrimination in regard to imports of food and similar products. Another, but quite different, form of administrative protection arises out of the technical problem of classifying goods for entry. The gamut of "commercial designation," "legislative intent," "chief use," "in chief value of," "similitude clause," etc., must be run and during such procedure imports may be held up indefinitely. In fact, the uncertainty involved in the awaited decision as well as the cost of bond, etc. may inhibit importation more than if an even higher but certain rate had been applied in the first case.

5

It seems pertinent to point out that in some cases a very low tariff rate on one commodity will furnish more protection than a high rate (ad valorem equivalent) on another commodity. In a situation where there is a slight or no difference between the imported cost and the cost of the domestic article a very low ad valorem rate might give much more protection than a consider

1 Appendix 5 was prepared by Edward C. Welsh,

2 Cf., E. M. Winsiow, Administrative Protectionism, Explorations in Economics, McGraw-Hill Book Co., Inc., N. Y., 1937; also Josef Gruntzel, Economic Protectionism, 1916; also, B. A. Levett, Through the Customs Maze, N. Y., 1923.

For instance, on a port cargo of less than 200 tons destined for Portuguese West Indies, the following consular charges were levied (in English pounds): Rotterdam 4, Antwerp 2, Dunkerque 1, Middlesborough 2, Hull 2, London 2. 1.10, Port Said, 1, Suez, 4. See World Trade, I. C. C., April 1932.

4 Cf., World Trade Barriers in Relation to America Agriculture, Senate Document No. 70, 1933; Sir Frederick Leith-Ross, "Report of the Economic Committee of the League of Nations on Agricultural Protectionism; Board of Trade Journal (Gr. Br.) London, May 1937; etc.

Cf. "American Importer," November 1934; September 1934; World Trade, May 1932; U. S. Customs Encyclopedia, 1934, etc.

ably higher rate levied on another commodity which had a large differential of disadvantage to the foreignmade goods. A usual type of this sort of thing occurs where the cost of transportation of one commodity is much greater per unit of dollar value than the cost of transportation of another article."

Another factor which complicates a study of the relationship between the tariff and prices is one which arises out of the existence of both ad valorem and specific customs rates in the United States tariff schedules. The specific duty, which is a fixed sum of money to be paid on some stated unit of quantity of a commodity, presents more difficulties in this analysis than the ad ́valorem rate, which is a fixed percentage of the value of the imported article. The specific rate acts as a regressive tax in the case where there are several qualities of a commodity imported, i. e., the cheaper qualities tend to get a greater amount (percentage) of protection per unit of value than the more expensive items. In some cases this is offset in part by levying a larger specific duty on the dearer goods and to the extent that such a procedure is followed the specific duty approaches the ad valorem duty in character. It is also of great significance that the weight of the specific duty varies inversely with the price level and particularly that it varies inversely with the price changes of the commodity upon which it is levied. It is true, of course, that the ad valorem type of duty presents difficult problems of appraisement and therefore considerable room for the operation of administrative protection activities. However, its very percentage character prevents it from having such a fluctuating protective role as the specific duty.

7

In spite of these difficulties, a comparison of wholesale prices in the United States with the tariff protection of the commodities to which the prices refer should throw some light on the relation between price behavior and protection. In order that this material might be in such form as to reveal any possible connection between tariff rates and price flexibility, the wholesale prices have been arranged in groups according to their relative frequency of change and their relative magnitude of change. For this purpose, the procedure described in Appendix 2 and the groupings shown in tables II and

For instance, a case of South African apples, weighing 46 pounds, was sent from London to the continent of Europe, a distance of 650 miles. The invoice value was $1.64 and total cost was $6.22. Transport costs were 75.76 percent of total costs. Imports of certain commodities, such as cement and certain fresh vegetables, are restricted to certain areas of a country, usually along the coast or border, largely because of the low unit values of such goods and of the high proportion which transportation costs are of total delivered costs. See Doc. No. 180, 72d Cong.

7 Bureau of Labor Statistics wholesale prices.

III of Appendix 2 were used. In a few instances, items in the Bureau of Labor Statistics wholesale price series which were combined to form a single item in the tables in Appendix 2 carry different tariff rates. In these cases, the separate items are presented in the following tables in place of the composite items which appear in appendix 2 and this fact is noted on the tables.

In table I the items are grouped into 10 groups on the basis of frequency of price change. This grouping corresponds to, and follows the order of Appendix 2, table II.

In table II, the items are grouped into 10 groups on the basis of price sensitivity. This grouping corresponds to, and follows the order of Appendix 2, table III. By reference to Appendix 2, table II, it is possible to see in which sensitivity groups the items in each frequency group are to be found.

The procedure for deriving the tariff rate on items. corresponding to the items in the Bureau of Labor Statistics wholesale price series was as follows:

(1) The commodities in the Bureau of Labor Statistics series were checked in order that further information about the particular nature of each item might be obtained. Description, adequate for accurate tariff classification, was not available in every instance.

(2) With this information to be used as a guide, the Tariff Act of 1930 was used for ascertaining the probable paragraph and rate of the tariff applicable to each commodity.

(3) Next, the United States Department of Commerce publication, Statistical Classification of Imports into the United States was used as a more accurate check on the tariff rates. In some of the more complicated instances, recent Treasury Decisions of the Customs Court were obtained to clarify problems of classification. The coding system used in the above mentioned Commerce publication was followed and recorded so that each item could be located more readily in further examination.

(4) An arithmetic mean of the annual wholesale prices of each of the 734 items was figured for the years 1930 to 1936 inclusive (the years during which the 1930 tariff act has been in force). The purpose of this average was to aid in more accurate classification of those commodities on which the duties vary according to the value of the item, e. g., "valued not over $2 a dozen" and "valued over $2 a dozen" (61548 and 61549). It appeared that an average of these prices over a period of years was a better basis for such classification than the prices for any one year.

(5) Then, by aid of the code numbers obtained in the Department of Commerce publication, further classification checking was done with the more detailed Tariff Commission publication, Comparative Statistics of Imports Into the United States for Consumption (11

volumes 1931-35) and also with the annual publication, Foreign Commerce and Navigation of the United States.

(6) Each item was then checked to ascertain if and to what extent it was affected by the reciprocal trade agreement changes. Each such change was noted.

(7) Using the afore-mentioned Tariff Commission publication for 1931-35 and the latest publication of the same title for the year 1936, the specific duties were translated into equivalent ad volorem rates for each commodity dutiable at specific rates. In cases of combination rates (both specific and ad valorem) the equivalent ad valorem rate shows such combination. The procedure was as follows: The total amount of duty collected on a particular commodity in each of the years studied was compared with the total value of this commodity imported during this particular year. This gave the ad valorem equivalent for a specific duty levied. In cases where there was a combination (specific and ad valorem) duty levied, the number of items, or the volume imported was multiplied by the specific duty; the percentage which this figure was of the total value imported was derived and that percentage was then added to the ad valorem duty quoted to obtain the ad valorem equivalent of the whole duty. In recording these equivalent ad valorem rates for the years in which the Tariff Act of 1930 has been in force, the spread or range of the rates from high to low years was taken rather than an average, since the fluctuation of the weight of the duty is significant, and also since an average would distort the picture.

Tables I and II show the results of the procedure just listed. Column 2 in the tables gives the code number of the different items as they are listed in the Bureau of Labor Statistics wholesale price series. Column 4 gives the code numbers used for corresponding items in the Department of Commerce publication, Statistical Classification of Imports into the United States, and in the Tariff Commission publication, Comparative Statistics of Imports into the United States for Consumption. The code numbers are given without decimal points as they are arranged without reference to decimal points in the Statistical Classification. Columns 5 and 6 give the paragraph in which the item appears and the rate according to the Tariff Act of 1930. In cases where a rate appears in parentheses the rate so enclosed is the changed rate according to the reciprocal trade agreements. Column 7 gives in terms of ad valorem equivalents the range of protection arising from the variations in price and the modification brought about by rate changes made by trade agreements.

The following tariff classification of these commodities should not be used as a final analysis for importers because of inadequate description of the separate items as well as the existence of changes arising out of trade agreements and Treasury decisions.