Improvement in observance of tariffs, [431].—Conditions in the eighties, [432].—The depression of 1893-1897, [433].—Resumption of prosperity in 1898, [436].—The rate wars of 1903-1906, [438].—Threatened disturbances in 1909-1911, [439].
The course of freight rates in the United States during the last generation divides naturally into two periods, before and after 1900, respectively.[487] Prior to that date an almost uninterrupted decline took place, which has been followed by a strongly marked upward tendency during the last decade. In respect of freight rates this movement is commonly judged in either of two ways; by comparison of actual rates charged for specified service between given points through a series of years; or, secondly, by means of what is called the revenue per ton mile. Considering first the use of this latter index, the course of events is shown by means of the diagram opposite the next page, covering the period between 1867 and 1910. But before conclusions may safely be drawn from this showing, it is imperative that the true significance of revenue per ton mile statistics should be set forth. For a generation, and particularly in connection with the Roosevelt legislation in 1906, volumes of written and oral evidence upon moot questions were based upon such figures. Specious and misleading reasoning upon a public question was perhaps never more common in the course of our history.[488] It is most important to understand clearly the real significance of this common statistical unit. We shall then, only, be in position to interpret the diagram properly.
The revenue per ton mile for a given road, or for the railway system of the United States, is computed by dividing the total freight revenue for that service, whatever it may be, by the number representing the amount of freight in tons hauled one mile. Thus, for example, if the total freight revenue of a system of roads be $900,000,000, this having been received as compensation for hauling an equivalent of 90,000,000,000 tons of freight one mile, the compensation actually received for each ton hauled one mile is obviously one cent. All that is necessary in order to compute the average revenue per ton mile, then, is to know the total freight revenue and the amount of ton mileage service. Computed in this way the average revenue per ton per mile for the railways of the United States in 1867 was 1.92 cents. From this level a decline took place in 1890 to 0.941 cents—that is to say, the average amount received for each ton of freight hauled one mile had declined about one half. Since about 1897 there has been no considerable change, the corresponding figure for 1911 being 0.757 cents. In other words, at the present time the carriers of the United States receive about three-fourths of a cent for each ton-mile service. From a revenue point of view this unit may seem insignificant in amount; but it should be borne in mind, of course, that it is applied to an immense volume of traffic. Even the slight increase between 1900 and the present time, if applied to the volume of traffic now existing, would make a difference in freight revenue for the entire railway system of the United States of approximately $61,000,000.
MOVEMENT OF FREIGHT RATES 1867-1970
Measurement of the course of freight rates by means of revenue per ton mile possesses one great advantage. It measures the actual return received by the railway without regard to the published tariff, showing accurately, therefore, the degree to which any departures from the published rates take place. For this reason the foregoing diagram probably under-indicates the extent, relatively, of the decline before 1900. For it is indubitable that the published rates have been very much more nearly observed with the passage of time during this decade.[489] On the other hand, revenue per ton mile, as thus used for general purposes, is open to a number of very serious objections. Obviously, like any statistical average it fails to represent the actual payment for any given service. But its disadvantages are more deeply seated than this. Entirely irrespective of any change in the level of rates, revenue per ton mile is affected fundamentally by three distinct sets of conditions. It varies according to the nature of the traffic, whether high grade or low; it is affected by the length of the haul and the proportion of local as distinct from through business; and it is modified profoundly according to the volume of traffic handled.
Before proceeding to the consideration of the above mentioned factors, attention may be directed to the following table, which gives the extreme range of ton-mile revenue for a number of different railways arranged in groups according to the nature of their business.
Each group is graded, moreover, within itself according to the revenue per ton mile. From this showing it appears that for 1910 the range above and below the average for the United States is considerable—being upwards of three times as great for the New Haven system as for the Chesapeake & Ohio, which comes at the foot of the list. It will now be in order to explain the reasons for these wide variations, which are by no means, as is customarily assumed in public discussion, conditioned even primarily by the level or reasonableness of the freight rates charged. Until these attendant circumstances are fully understood, any conclusions as to relative freight rates for a given service based upon revenue per ton mile, are entirely misleading.