The accompanying table compares the trend of average ton-mile receipts during the period 1878 to 1900 with an index number of published tariffs. The fact that such tariffs were ignored to a considerable extent in actual rate-making is indicated by the fact that the striking increases in the index in 1882 and 1886 were not reflected in the average ton-mile revenue figures. The effect of regulation in eliminating rebates is shown by the trend in ton-mile revenue as compared with the rate index after 1886-the latter declining 38 points and the former only 6. The stabilizing effect of regulation on rates is also evident in the comparative changes before and after 1887, in either index.

It is evident from a study of the table on pages 81-82 that, following the period of rebate elimination (roughly 1890 to 1900), regulation tended to prevent the roads from exploiting the growth of industry and charging high rates, thus probably reducing their potential profits. More recently (as noted in the analysis of the work of the Interstate Commerce Commission since 1910) regulation has become unduly repressive, though a change in sentiment now indicates that the roads will receive fairer treatment in the future.

Thus far no mention has been made of fluctuations in the average ton-mile revenue caused by changes in the composition of the traffic. It is evident that if a larger proportion of high-grade commodities are shipped in one year than in another, the average ton-mile revenue will be higher even if rates remain the same. Thus, the trend of average ton-mile revenue indicates only approximately the trend of an average of rates. Unfortunately, since only tonnage (and not ton-mileage) statistics are available for groups of commodities for the length of time required for an historical study, it is impracticable to revise the averages by making allowance for traffic changes.

In a very rough way, however, we may measure these changes by the alternate periods of depression and prosper

ity. During depressions the proportion of high-rate commodities might be expected to decline considerably, because the consumption of luxuries is affected to a greater extent than that of necessities. Referring again to the table on pages 81-82, it will be noted that the average ton-mile revenue for all carriers declined steadily from eighty-nine cents to seventy-three cents during the period of depression 1893 to 1900, following which there was a gradual recovery to seventy-eight cents in 1904. About this time rate-making became more fully subject to the regulation of the Interstate Commerce Commission, which may account for the narrow range of the average (between seventy-five and seventyseven cents) during the years 1905 to 1911.

In the previous chapter we studied the relations between the density of traffic and the trend of average ton-mile revenues. These factors combined produce the gross freight revenues per mile of road. The accompanying table gives gross and net revenues per mile for all railroads during the period 1871 to 1920. Several striking facts are evident from a study of the table: a remarkably steady decline in gross and net per mile from 1871 to 1878, representing the effects of the severe depression during that period; a brief recovery, followed by an irregular decline from 1882 to 1895; from that year on, a steady and rapid increase in gross and net up to 1917, explained by heavy traffic growth without much increase in mileage; from 1917 to 1920 a very rapid increase in gross due to war-time activity and rate advances, but a phenomenal decline in net due to mounting costs.

It is evident that the period of railroad growth during the past fifty years may be roughly divided in half: up to about 1895, expansion was in the form of new mileage, and development was of an extensive character; since that time, the growth has been intensive in character, existing mileage being greatly improved so as to handle a volume of traffic two or three times larger than before.

The figures for net revenue per mile in the table do not present a very accurate picture of the trend of railroad profits, especially during the period 1910 to 1920. Operating expenses, while the most important item in cost of transportation, do not make up the total cost, which should also include taxes, rentals, and other items. Moreover, net income from operations does not mean a great deal unless brought into relation to the amount of the property investment.

The table following shows the net earnings from operation in relation to the property investment of the principal railroads in the United States, from 1908 to date. Figures prior to 1908 are not available for the United States, but results for the Eastern railroads (the figures for which were carried back to 1900 in the 1914 Advanced Rate Case exhibits) indicated that earnings increased rather steadily from 1900 to 1907, falling off in 1908 to about the same figure as in 1900which facts agree fairly well with the trend of net revenues per mile for the United States.

The earnings for the years 1918 to 1920 are of little significance because during the period January 1, 1918, to September 1, 1920, the railroads were guaranteed a "standard return" by the Government, the amount in dollars being based (with adjustment for increased investment, etc.) on the average income for the three years ending June 30, 1917. The figures as given in the table represent actual operat

1 The figures from 1908 to 1919 were taken from the Interstate Commerce Commission's thirty-fourth annual report to Congress. The rate of return for 1920 is figured on property investment of the roads as of December 31, 1919, and railway operating income is partly estimated. The 1921 figures are taken from a compilation prepared by the Bureau of Railway Economics.

2 For 1916-21 the figures are for calendar years.

3 Operating income means operating revenues less operating expenses, taxes, and operating rents (as distinct from rents for lease of road and miscellaneous rents).

4 Tentative valuation fixed by Interstate Commerce Commission slightly lower than property investment actually reported by the roads.

ing results under Federal operation set in form comparable to figures for previous years. They do not indicate the amounts actually obtained by the roads.

In September, 1920, the Interstate Commerce Commission, acting in accordance with the policy laid down in the Esch-Cummins Law, raised all freight rates about one-third and all passenger rates about one-fifth.