Although the Commission approved the use of a 3 percent rate of depreciation (10 percent salvage value) for determining the annual unit costs of tank cars in the 1967 Tank Car Allowances case, MCA employed the rate of 2.65 percent (with a 20 percent salvage value) for the private covered hopper car fleet. Respondents argue that this important factor was not adequately explained or justified by MCA in this proceeding. MCA did not develop any engineering or cost study of its own to determine the rate of depreciation reflected in its cost formula. Rather, it applied a rate drawn from AAR Mechanical Division Interchange Rule 112, which specifies that a depreciation rate of 2.65 percent, on a straight line basis and with a salvage value of 20 percent, will be used among the carriers themselves in arriving at settlement values for covered hopper cars destroyed or badly damaged in interchange ser vice. A primary justification for Commission approval of the 3 percent depreciation - 10 percent salvage value factor in the 1967 Tank Car Allowances case, 329 I.C.C. 466, at 516, was the fact that that factor had been in use by the railroads under the provisions of rule 112. Since the same is true here as to the depreciation rate for covered hopper cars, the 2.65 percent ratio has prima facie validity for a covered hopper car allowance formula. Respondents' characterization of that rate as "a totally unjustified and unexplained component" of the MCA's proposed allowances does not weaken that validity. The examiners find its use in the manner proposed by MCA to be just and reasonable. Respondents also argue that the proposed allowances may not be approved by this Commission in any event because, even assuming the private hopper car cost data employed herein is accurate, there is nothing in the record to show the cost of railroad controlled covered hopper cars which make up 75 percent to 80 percent of the total covered hopper car fleet. Respondents contend that the value of shippers' furnishing hopper cars to the carriers cannot exceed the cost to the carriers of their own hopper cars, apparently on the theory that the carriers' cost experience on these cars should mark a reasonable comparative level for those similar cars furnished by shippers. In taking this position, respondents appear to be relying upon the dual test the Commission has applied in the past to determine the maximum reasonable level of allowances: In administering the provisions of section might be paid should not exceed the reasonable The test upon which respondents rely is neither as narrow nor as absolute as they maintain. In the present context, the test of justness and reasonableness has relevance to the interests of the shippers being paid allowances as well as to the interests of the carriers themselves. Shippers should not be deprived of allowances which appear to be just and reasonable judged by shippers' costs and other related data solely because the carriers' comparable costs are not disclosed. The carriers themselves are in control of their freight car cost information, and they may, at their own election, organize it and relate it to protestants' proposals. The examiners conclude that the lack of specific data as to the respondents' costs for their own covered hopper cars does not bar our inquiry into the justness and reasonableness of the allowances placed in issue in this case. If the railroads enjoyed full-time use of the private covered hopper car fleet, mileage allowances could be determined merely by dividing the above-described annual costs by the mileage recorded by the fleet for the same year. The AAR's 1966 study of the covered hopper fleet showed an average of 10.007 loaded miles per car per year. MCA concedes that private covered hopper cars, for example, are used to some extent by shippers for purposes such as storage, not directly related to transportation, and that, therefore, a portion of these cars' costs should be excluded from allowances. To measure that "nontransportation use" of private hopper cars, MCA analyzed a wide range of statistical data and, ultimately, developed a "transportation use ratio" of 90 percent. MCA then used this ratio in adjusting the annual car costs to exclude costs not related to "transportation use". To develop a transportation use ratio, MCA first undertook to compute the number of days that private hopper cars spent in active transportation service. Here and in subsequent areas of the MCA transportation use formula, respondents criticized the MCA's interdependent use of statistical data from various independent sources, such as officially published statistics of the AAR and averages selected from a carload waybill sample taken by this Commission. MCA made various computations from this data and by applying Rail Form A factors to make adjustments to data from certai special studies of the private covered hopper fleet -- translated it into active car days; time consumed by private covered hopper cars in repairs; loading and unloading time; and in "waiting time" between loads. MCA developed its - transportation use ratio by comparing the results of these computations with similar computations related to railroad controlled covered hopper cars. Lacking comprehensive data on the actual operations of the covered hopper fleet, MCA relied heavily upon certain "judgment factors" built into the Rail Form A costing procedures. For example, MCA used an average of one half day on each side of an interchange and a half day for each intertrain and intratrain switching, based on Form A. The number of intertrain and intratrain switches was computed from average round trip mileage on the basis of an average of 202.91 car miles per switch, again derived from use of calculations indicated in Rail Form A. The number of car days consumed in switching to and from the points of loading and unloading, and in the process of loading or unloading itself, were average factors taken from Rail Form A. Respondents criticized the use of these factors as arbitrary, unreliable, and unrelated to actual covered hopper car operations. The examiners conclude, however, that used in the manner accomplished by MCA, the Rail Form A factors are useful indicators. In the context of these proceedings, the application of these Rail Form A costing factors to known averages and other sound statistical data would be adequate to establish a prima facie validity for a structure of interim allowances in this procedure proposed by MCA. In the kind of case presented here, the shippers furnishing cars to the carriers are not, by themselves in any position to conduct the kind of comprehensive engineering survey of the carriers' operations that would be required to develop actual operating factors, as opposed to those derived from statistical data. On the other hand the carriers are fully familiar with the Rail Form A costing factors and with the engineering and statistical analyses upon which they are based. Thus, the carriers are in a Dosition either to furnish specific factual operational data to refute protestants' utilization of the Form A factors, or to present specific arguments why such factors may not validly be used as MCA has used them. One of the key elements in the MCA determination of the time spent by covered hoppers in transportation use is the computation of the average number of shipments per year per car. Respondents contend that this computation is the result of an unsupported and arbitrary selection by MCA among conflicting statistical alternatives. MCA made a special questionnaire survey of the 1966 operations of 23 users of private covered hopper cars, including 18 members of MCA and 5 companies which ship grain. The questionnaire returns provided the number of cars, number and total miles operated. A total of 3,139 cars was embraced in the survey, shown as averaging 840 miles per trip, 10.87 trips per year, and 9,130 average miles per car per year. In contrast, the AAR's general survey of the pri vate covered hopper fleet for 1966 had recorded an average of 10,997 average miles per year per car. The MCA concluded that the only statistically sound average developed in its own special study was the 840 average miles per trip, since the questionnaire responses included data for cars controlled by the study shippers for portions of the year only. The responses of most of the reporting MCA companies did not provide a basis for determining whether the cars were controlled and used for 12 months in 1966, or for any specific proportion of that year. MCA, therefore, used the annual average of 10,997 miles drawn from the AAR general survey, in its transportation utilization formula, together with the 840-mile average per trip to determine the average number of shipments per car per year, 13.1 shipments. The examiners conclude that the MCA's statistical method for computing average annual shipments for private cars is a sufficiently reliable and reasonable procedure for its intended purpose in the absence of other, complete data. The 10,997 miles figure is drawn from a generalized survey conducted by the AAR covering some 15,000 cars' operations, and it is entitled to substantial weight in these computations. Similarly, the 840-mile average per trip is based on a broad sample covering more than 34,000 test shipments in 1966. A relation between these statistics would be likely to produce a reliable and credible average. - more It is on a succeeding procedure in the MCA formula, however, that the computation of the proposed utilization ratio becomes dependent upon an invalid statistical device. Using data from the 1966 I.C. C. carload waybill sample a computation of 501 miles as the average haul for all covered hoppers together with a number of computations based on Rail Form A factors, MCA determined that covered hopper cars in 1966 averaged 100.3 car miles per active car day (exclusive of loading and unloading time). Concluding that privately-owned cars experience a substantially longer haul per shipment than rail-owned equipment, and that the use of the 100.3 car miles per day would not be proper, MCA proceeded to adjust the computed average for the differing uses of covered hopper cars under railroad control and private control. To make the adjustment, MCA examined the records of 1967 covered hopper car operations by one of its members, which unlike most MCA members, had regularly been keeping records of its transit times and miles traversed. From these records, MCA computed the actual experience of the study company in terms of car miles per day, and mathematically converted these averages, by the method of least squares, into a statistical trend line. The trend line purports to show the increase in active car miles per day which would accompany a given increase in the loaded and empty mileage per shipment. The computed trend line for the actual experience of the study company would, for example, indicate that a movement of 500 miles would tend to be accomplished at an average of about 36 car miles per active car day, while a 1,000 mile movement would reflect about 64 car miles per day, and a 1,500 mile movement about 91 car miles per day. MCA observed, however, that the study company's experience did not equal the 100.3 car miles for active car day average computed by use of Rail Form A factors as described above. Therefore, MCA utilized the computed 100.3 car-mile average as the basis for an adjusted trend line paralleling the trend line drawn from the study company's actual experience. The adjusted trend line showed about 72 car miles per active car day, instead of 36 car miles, for a 500 mile movement; approximately 100 car miles per day instead of 64 car miles, for a 1,000 mile movement; and, approximately 127 car miles per day instead of 91 car miles, for a 1,500 mile movement. The adjusted trend line was used by MCA to determine car miles per active car day for railroad controlled covered hopper cars as compared with privately-controlled hoppers. The car miles per day determination is critical to the ultimate computation by MCA of its projected transportation use ratio, since such factors as active car days in railroad possession, total car days in railroad possession, and transportation related "waiting time" between shipments are dependent upon that basic determination. The examiners conclude that the car miles per active car day trend line projected by MCA has not been shown to be a valid measuring device for use in this proceeding. The adjusted trend line based first on actual experience and then modified by a further computation based on Rail Form A factors purports to reflect useful statistical experience for use of all covered hopper cars, under either railroad control or private control, but it is founded initially on the actual experience of only one company. MCA represents the study company, on the one hand, to be "fairly representative" of other users of covered hopper cars, but, on the other hand, MCA rejects the real experience of that company as the basic factor under study - the average number of car miles per active car day. By accepting, and adjusting for formula utilization, only the trend line discernible in the study company's experience, and rejecting that company's actual experience in car miles per car day, MCA's arguments are self-contradictory. Even if we could accept the experience of a single company study for the use MCA proposes, MCA's own rejection of that experience impeaches it validity. As shown above, the differences are very substantial between that experience and the Rail Form A based computations of car miles per active car day upon which MCA chooses to rely. |