stated that the maintenance expenditures of the Railroad Administration were being made on a scale which would meet the Government's obligation under the contract proviso based on the test period. It was his belief that at the end of federal control the Government would be indebted to some of the carriers for inadequacy in some elements in maintenance and that on the other hand some of the carriers would be indebted to the Government for overmaintenance in certain elements. On the whole, he believed that these would about offset each other, and in estimating the net financial result of federal control in his last report to the President he made no allowance for claims for undermaintenance.

An examination of the expenditures during federal control would indicate that Mr. Hines had reason to believe that his position was sound. The average yearly expenditures for maintenance (way, structures and equipment) during the test period were $963,970,719. In 1918 they were $1,742,001,074, and in 1919 they were $1,982,793,296. These figures indicate that the increase in expenditures over the test period was 81% in 1918 and 106% in 1919, an average increase of 93% for the two years. The increase in maintenance of equipment was relatively greater than in maintenance of way-105% for equipment and 76% for way and structures.* No information as to the relative cost of labor and prices of material

*Based upon statistics compiled by the Operating Statistics Section of the Railroad Administration. The figures apply to all Class I roads.

in the test and control periods has been made public, but it is unlikely that wage rates and material costs combined were more than 93% greater than in the test period.

None the less, the properties were not fully maintained. In the single item of cross ties the average renewals per year in the test period were 83,885,109. The average yearly renewals in 1918 and 1919 were 71,363,083. This does not give a complete picture because in many cases the renewals in 1918-19 were in ties of lower grade than those used in the test period. The average annual renewals in rails during the test period were 2,041,676 tons. The rail renewals in 191819 were at the rate of 1,821,561 per year. The ballast renewals in the test period were 17,065,599 cubic yards. The comparable figures for 191819 were 16,157,522, but in some cases gravel ballast was used instead of crushed stone.* In addition to the inadequacies in the renewals of ties, rail and ballast, there was inadequacy in the maintenance of structures, mainly in painting. These, however, cannot be stated in exact units.

In the case of locomotives the maintenance was fairly adequate, but both freight and passenger cars were in relatively inferior condition when the roads were returned. The failure to keep up freight car repairs may be attributed in greater part to pooling. With respect to passenger cars it was impracticable to release them from service for the usual periodical shop overhauling, because

*From evidence prepared by Bureau of Railway Economics for presentation before Senate Committee on Interstate Commerce in hearings on Senate Resolution 23, June 15, 1921.

every car was urgently needed in 1918 for the movement of troops from camps to seaboard and in 1919 for the demobilization of the army.

The case of freight cars requires further explanation. Prior to federal control the joint use of freight cars by all roads was governed by the car service and the per diem rules of the American Railway Association and the interchange rules of the Master Car Builders' Association. These rules were entirely suspended or substantially modified during the period of federal control. The car service rules prescribe the manner of joint use; the per diem rules fix the rate of daily rentals; and the Master Car Builders' code defines the responsibility for repairs. We are particularly interested in the question of repairs. In brief, the code holds the car owner responsible for repairs made necessary by ordinary wear and tear in service both when the car is at home or on another road. The user of the car is held responsible for repairs made necessary by "unfair" usage, such as that caused by train accidents. The light running repairs required to make good broken or missing parts are made by the road which has the car, when the need of such repairs becomes apparent. The cost of making good ordinary wear and tear is charged to the owner; the cost of repairs occasioned by rough usage is charged to the road responsible for the damage. In the course of time, usually every two or three years, the cumulative effect of wear and tear and accidental damage necessitates a general over

hauling of the car. In normal times this is commonly done by the owning road when the car is at home. When a car is on a line other than the owning line the repairs are confined ordinarily to those required to put the car in condition to be sent home in service. The using road naturally restricts its work on cars of other companies to that which is necessary to keep the car in condition for day to day service. This means that the typical road takes care of the running repairs of both its own cars and those of other roads while on its rails, and confines the heavy general repair work to its own cars. Prior to federal control the typical road had about 60% of its cars at home and 40% were on other lines.

During federal control the principles of unification were applied in a degree which with respect to freight cars caused a virtual disregard to ownership. Under the instructions no distinction was to be made between "home" and "foreign" cars in making repairs. In theory each road was required to regard every car as its own car and to make the needed repairs, both light and general, to the limit of its capacity. In the distribution of cars little effort was made to return them to the owning road. Consequently the percentage of "home" cars on "home" roads was reduced from 60% to from 10 to 20%.

If the instructions which required each road to assume responsibility for the general repairs of all cars on its lines could have been carried out the results would have been more satisfactory. But as each road had difficulty in obtaining labor

and materials, it was natural that the heavy repairs were devoted principally to the few home cars that happened to be at home and the work on foreign cars was confined to running repairs or patch work which merely kept the car in condition for day to day service. As the heavier work was neglected it was inevitable that the general condition of cars should suffer. It was inevitable also that the larger proportion of patch work or temporary repairs substituted for the thoroughgoing general repairs which would have been given by the owners in normal times meant that the total expenditures would not bear the usual relation to the general condition of maintenance. And in addition to this factor, which explains in greater part why freight cars were returned in inferior condition notwithstanding the expenditure of more than twice the maintenance cost of the test period, there was the factor of relative efficiency of labor. The restrictive rules of the national agreements and the much larger percentage of inexperienced employees had the effect of increasing expenses without increasing output.

A recent statement† by Ralph Budd, president, Great Northern Railway, contains statistics which

*The factor last mentioned was a main point of difference between the Director General and the railroad companies in coming to an agreement on the terms of the final settlement. It is unnecessary to pursue the point further here. It was in controversy also in connection with the guarantee paid by the Government during the first six months under the operation of the Transportation Act of 1920 (March 1 to September 1, 1920). The Interstate Commerce Commission decided that the relative efficiency of labor should not be considered in the settlement.

Railway Review, June 10, 1922, p.875.