intended to show that the weighted-average ratios of revenues to variable and fully allocated costs for representative shipments on shipper-owned or supplied equipment are in excess of 200 percent and that earnings per car-mile for so-called similar "high-wide" commodities are substantially less than earnings per car-mile for heavy electrical machinery. This data is based on a sample of shipments of heavy electrical machinery costed on the basis of Public Statement ICI-69, supra. However, we find this data unreliable as will be shown.

Cost statements of the Commission are public documents prepared from annual reports of carriers and do not purport to represent the exact costs of any particular carrier or carriers but instead provide general average-cost information which can be useful in establishing rates. The Commission accepts cost studies based on these statements and accords them whatever weight the circumstances warrant. See Steel Bars, Lemont, Ill. to Iowa, Kans., Mo., Minn., & Nebr., 319 I.C.C. 292, 302 (1963). Complainants have made errors in their cost estimating which preclude our according it evidentiary weight.

At the outset, complainants' cost estimates are based on data obtained from Public Statement ICI-69, supra, and applied to Rail Form A. Station clerical expenses, interchange switching costs, and road train-to-industry switching costs were applied on a pershipment basis. Cost data obtained from Public Statement ICI-69, however, are computed on a carload basis. Similarly, Rail Form A derives unit costs for interchange switching and other terminal costs on a carload basis, specifically by dividing the total expenses of a given service by the number of carloads given that service. Thus, carloads are the basic service unit divisor in Rail Form A, and costs derived therefrom are on a carload and not a shipment basis. Therefore, in instances where multiple-car and train-load shipments, in fact, are involved, the carload cost per car shown in Public Statement ICI-69, supra, to be relevent needs adjustment downward to reflect the economies associated with multiple-car handlings. Adjustments to the carload costs the carload costs to reflect such economies are discussed in Investigation of Railroad Frt. Rate Structure Coal, 345 I.C.C. 493, 540-542 (1976). Complainants, however, calculated costs on a per-shipment basis equating carload costs and shipment costs, and thereby understate costs in instances where the number of cars involved in a service is greater than one. The costs associated with handling more than one car at one time (switching and clerical costs, for example) while increasing with the

number of cars in the shipment, do not do so proportionally. Complainants' analysis fails to recognize such factors.

Moreover, where complainants attempt to quantify such additional expenses as delays for high and wide movements, additional yard switching, and additional origin inspection, they provide no basis for the estimates used. Yet, it is clear that the underlying data on which these judgments are made should be part of the record. See Boots or Shoes From Massachusetts to Texas, 315 I.C.C. 264, 265 (1961). Similarly, while complainants have included many of the special costs of moving heavy or wide shipments, they omit costs for special equipment for clearance handling, the assignment of additional personnel for handling and supervising the routing of shipments requiring special clearance, and the special services rendered such shipments. These omissions again result in an understatement of the costs. Also, the costs for train delay due to high, wide, or heavy shipments, while a reasonable estimate would have been acceptable, included only those costs incurred specific to that train, namely locomotives, freight car, and crew expenses. The added costs for a clearance bureau appear to have been included only as part of general overhead. The net effect of this oversight is again to understate costs.

Complainants also erred in their cost computations in movements where idler cars are used. Complainants have considered movements which require idler cars to be multiple-car movements moving as a single unit. (This is shown by their application of terminal and interchange costs on a shipment rather than a carload basis.) But, to the extent idler cars and the cars carrying heavy electrical equipment are handled separately, either at origin or destination, classification or intermediate yards, complainants' consideration of the movement of these cars as a multiple-car unit understates costs.

In addition, complainants erred in computing the weightedaverage ratios of revenues to costs. Complainants used the weight of the shipment as the weighting factor for each shipment's revenuecost ratio. The proper method, however, would have been to divide total revenue of all shipments under study by the total variable or fully allocated costs of those shipments. The computation of both revenues and the costs from which the ratios are computed reflect the influence of both shipment weight and distance. This method incorrectly introduces shipping weight a second time.

Moreover, complainant's computations of car-miles per day for purposes of estimating car ownership costs are not reliable.

Complainants applied the results of a 1969 study to 1968 traffic data. A 1968 sample would have been more appropriate. Additionally, it appears the traffic data is based on the months of June and July 1969, but these months would not necessarily be representative of an entire year since transit averages are likely to increase and car-miles per day decrease in direct proportion to adverse weather conditions. Transit time would tend to be higher and car-miles per day lower if a winter month had been selected for the study. The period during which a cost study is made should be of such length that it will include all of a regularly recurring cycle and at the same time will not distort the results by unduly weighing certain parts of the period. See Des Moines Union Ry. Switching, 231 I.C.C. 631, 640 (1939); Reciprocal Switching at Mobile, Ala., 208 I.C.C. 151, 154 (1935). Additionally complainants include terminal switching time in determining car-miles per day. This method, however, combines time spent in line-haul service and terminal service and ignores a part of terminal switching such as switching service prior to loading and after loading. The correct costing method would have been to break down the total time between linehaul and terminal service.

Complainants also submitted a comparison of car-mile revenues shown in appendix B hereto. This data would indicate that in almost every instance the assailed rates yield car-mile revenues which are substantially higher than those yielded by the compared rates. But we do not consider the comparison valid since revenues from the assailed rates are computed on a 150,000-pound minimum weight, whereas the compared revenues are computed on shipments ranging from 40,000 to 80,000 pounds. Moreover, and as shown in the prior report, the commodities compared are not, in fact, comparable. See generally Public Serv. Com. of N. Dak. v. Great Northern Ry. Co., 340 I.C.C. 739 (1972); American Colloid Co. v. Akron, C. & Y. R. Co., 321 I.C.C. 91, 113 (1963); General Motors Corp. v. New York Central R. Co., 311 I.C.C. 622, 624 (1960), affirmed General Motors Corporation v. United States, 207 F. Supp. 641 (1962).

Complainants also assail the present basis of rates for movements on shipper-owned or supplied equipment as relatively unreasonable. In numerous proceedings we have stated the principle that an important test of reasonableness is the comparison of an assailed rate with others which present a proper standard of comparison. Whenever the reasonableness of a rate is so measured, an examination of the relationship between rates is entailed. Thus, in an appropriate case where the compared rates or some other

standard satisfy the test of a proper basis of comparison, we may find a rate unreasonable by comparison. This rate relationship is cognizable under section 1 but distinguishable from much broader inquiries under section 3. See Chicago Board of Trade v. Illinois Central R. Co., 329 I.C.C. 529, 532-535, remanded on other grounds sub nom. City of Chicago v. United States, 291 F. Supp. 858 (N.D. III. 1968), affirmed per curiam 394 U.S. 717 (1969).

In this proceeding, however, we are not persuaded that the rate comparisons or cost estimates provide that suitable standard. Our criticisms above are equally applicable here, and the comparison of the assailed rates with the rates on various allegedly comparable commodities shows the rates are almost identical with two major exceptions. One exception is the compared rates on structural steel beams. Here the assailed rates are sometimes twice the compared rates on steel beams but the defendants defend the rate difference on the basis of differing transportation characteristics, especially value. They show that the 1972 per-pound value of structural carbon steel was 8 cents in contrast to $1.20 for electric transformers and $2.25 for other heavy electrical machinery. While complainants point out that the value of 8 cents for structural carbon steel must be increased to allow for such processing as sizing and milling, they have not indicated what the increased value is. The other exception is the compared rates on automobile parts. The defendants defend the difference in rates not only because heavy electrical machinery has a value per pound substantially greater than automobile parts, but also because the heavy electrical machinery moves at substantially lesser volume each year than automobile parts. For example, in 1971 the Penn Central alone transported a total of 7,662,406 tons of automobile parts. As noted above, in 1968 an approximate total of only 121,175 tons of heavy electrical equipment moved over all the defendants' lines. Although these figures for the year 1968 are not as timely as those of Penn Central for 1971, there is no indication that the volume of this traffic has varied substantially from year to year. In the prior report (349 I.C.C. 502, at 507-508) we also indicated many other problems with the comparisons-including, for example, the abnormally high risk in transporting complainants' products. These reasons presented by defendants constitute constitute a valid justification for maintaining substantially higher rates on heavy electrical machinery in comparison to the rates simultaneously maintained on structural 'This Commission has frequently approved lower rates on commodities moving in great volume. See Alpha Picoline from Detroit, Mich., to Indianapolis, Ind., 310 I.C.C. 31. 34.

5

steel beams and automobile parts. Thus there is no reason on this record for our concluding that the assailed rates are relatively unreasonable in comparison to the compared rates on other commodities.

As noted, complainants also contend that the assailed shipper allowance of 4 cents per mile does not cover the cost of providing the expensive equipment necessary to transport heavy electrical machinery. However, complainants' cost data are severely limited in this respect. These cost data are composed primarily of general estimates of depreciation expenses relative to purchase prices and monies received in annual revenues under current allowances, but the cars themselves have widely varying original purchase prices, widely varying use and widely varying miles run per year. As such, complainants' evidence cannot be used as a basis for establishing or determining the unreasonableness of the present allowance. Moreover, the adequacy of mileage allowances for shipper-owned or supplied cars was recently considered in Ex Parte No. 334, Car Service Compensation-Basic Per Diem Charges-Formula Revision, served August 10, 1977, in accordance with The Railroad Revitalization and Regulatory Reform Act of 1976. Therein, because of limited data and information pertaining to the adequacy of the present mileage allowance system of compensation, the Commission decided that a separate rulemaking proceeding will be instituted for the determination of the method of compensation for all privately owned cars, other than tank cars.

Finally, the court asks us to explain why we did not reopen this case for further hearing to enable complainants to present additional evidence as the basis for a rate prescription. This Commission has frequently noted the fundamental difference between finding rates unlawful for the past and prescribing rates for the future, and has pointed out how much more detailed evidence is necessary to prescribe rates which would be just and reasonable for the future. Moreover, the prescription of rates here is a special request of the complainants and thus they have the burden of proof. They have failed to meet this burden, and thus the complaint will be dismissed. See Hanna Mining Co. v. Missouri Pac. R. Co., 332 I.C.C. 166, 171, and Investigation of Railroad Freight Rate Structure, 345 I.C.C. 1364, 1394.

Upon reconsideration, we find the assailed rates and shipper allowance are not shown to have been or to be relatively or per se unjust and unreasonable. The complaint will be dismissed.