The total expenses reported by the study carriers after adjustment to reflect January 1974 wage levels and February 1974 fuel prices were separated into four categories of expense, that is terminal loading and unloading, terminal-other expense, line haul, and general overhead. To the extent possible expenses were assigned to these four categories. Expenses which could not be directly assigned were allocated on the basis of driver-hours, equipment utilization, or estimates provided by the carriers. The expenses of each carrier at each terminal were separately tabulated. Expenses assigned or apportioned to general overhead were allocated to the other three categories on the basis of total expenses charges to these categories. Revenues covered by accounts 3120—Freight Revenue-Local Cartage, account 3130—Intercity Transportation For Other Carriers, and account 3900—Other Operating Revenue, were deducted from operating expenses at each terminal except in one instance where account 3130 mistakenly included traffic handled by the reporting carrier.

Each of the carriers reported the investment in carrier-operating property and the reserve for depreciation at the beginning and end of 1973, and from this information the average depreciated investment in 1973 was computed. An allowance for working capital was based on 5 percent of operating expenses for the year. Interest on depreciated investment and working capital was based on a 6-percent interest rate. The total operating expenses of all carriers were combined and the resulting figure was divided by appropriate service units to arrive at a single set of unit costs applicable to operations under the considered tariffs. Expenses assigned or allocated to loading and unloading of $2,958,840 were divided by 409,503 automobiles, the number of vehicles loaded or unloaded by drivers operating from the study terminals in 1973, to arrive at a unit cost of $7.22 per vehicle. Total operating expenses for terminal operations, other than loading and unloading, came to $2,147,781. This figure divided by 449,392 vehicles results in a unit cost of $4.78 per automobile. This divisor is somewhat larger because it relates to total automobiles moving through the terminals, which for some carriers, is different from the loading and unloading count because drivers from other terminals are sometimes used in transporting cars through the terminals. Expenses for such drivers are included in the total expenses at the other terminals. Thus the total terminal cost comes to $12 per vehicle made up of $7.22 for loading and unloading plus $4.78 for other terminal handling.

The total line-haul expenses of the 7 study carriers were $11,526,969. Division of the latter figure by 10,100,720 loaded trailer-miles results in a unit cost of $1.14120 per loaded trailer-mile. This procedure spreads the line-haul expense over loaded trailermiles associated with the round trip movement of equipment from the various study terminals. Costs per automobile mile were computed to be 19.020 cents for a six carload and 16.303 cents for a seven carload.

The foregoing study, as briefly outlined, permitted the development of fully allocated costs from the principal origins to selected destinations and opened the way for comparison of costs with the assailed rates from and to the same points. The costs include 100 percent of the operating expenses, rents, taxes, and interest expense on depreciated investment including an allowance for working capital. No allowance was made for profit.

Costs are shown from Baltimore, Port Newark-Elizabeth, Providence, Boston, Toledo, and Chicago, to destinations which are representative of the points served by the study carriers in 1973 from the standpoint of traffic volume and geographic location. The costs for each of the considered movements are compared with the rates under investigation and rate to cost ratios were developed. The results of the study are shown in table A of respondent's cost statement dated April 1, 1974. The assailed

column B rates are compared with the costs incurred for truckloads comprising seven cars, while the column A, C, and D rates are compared with the costs for truckloads comprising loads of six cars.

As shown in table A, the column A rates failed in most instances to cover costs, and on the average accounted for 94 percent of costs thus failing by 6 percent to cover full costs. The column B rates are shown to be remunerative in some instances but not in others. The ratio of rates to fully allocated costs ranges from 89 to 118 percent with an average of 100 percent, or exactly equal to full costs with no margin for profit. The column C rates with but a single exception exceeded the full cost level with an overall average ratio of rates to costs of 112 percent or an excess of 12 percent over costs. The column D rates exceeded costs in all instances and on the average the rate to cost ratio was 131 percent or 31 percent over full costs.

In operations conducted in 1973 by the seven study carriers from their 13 terminals, the average load per trailer ranged from 6.26 to 7.04 vehicles or an average of 6.73 units. Thus the average cost per automobile mile for all loads transported would fall somewhere between the seven carload cost of 16.303 cents per mile and the six carload cost of 19.020 cents a mile. The average is made up of predominantly six and seven carloads but there were a few less than six and eight carloads.

Protestant is said to customarily limit itself to loading seven cars on a trailer. Though seven small cars such as the Volkswagen can be loaded on a trailer, it is very often the case that one or more larger cars are mixed with the shipment with the result that one less car is loaded. Most shipments under the considered tariffs comprise a mixture of vehicles of two weight brackets or occasionally three weight brackets though vehicles in the under 2,500 pound category predominate. Except for shipments of cars in the under 2,500-pound group, there are very few full loads. of the heavier weight cars. Almost all of the heavier cars move in mixed loads with the smaller cars. The column A and B rates, that is those on vehicles under 2,500 pounds, apply on the predominant portion of the protestant's traffic. Only a small percentage fall in the column Crates, or vehicles in the range of 2,500 to 3,000 pounds.

The aforestated costs were developed to reflect full loads of six and seven cars from origin through to destination. No adjustment was made to give effect to reductions in the average load resulting from split delivery because practically all Volkswagen movements are in full truckload quantities to a single destination. Though some small car movements, other than Volkswagens, entail split deliveries, in general this distribution characteristic is more typical of the heavier more expensive cars. Thus the ratio of the column D rates to costs overstates the earnings achieved on the heavier cars. In the opinion of respondent's cost consultant, the use of an average load factor for movements of the heavier cars, offset to some extent by credits for split delivery charges, results in higher costs and lower ratios of rates to costs for movements under the column D rates. As noted the rates in column D are substantially higher than those in the other columns. The vehicles which move under these rates are luxury or high performance machines with a value of up to five times that of the vehicles which move under the column A or B rates.

Respondent emphasizes that as the vast majority of Volkswagen automobiles move under the column A and B rates, and as these rates account for 94 percent and 100 percent, respectively, of costs, the proposed increase is clearly justifiable. In many studies made by the Commission's Cost Finding section a standard allowance for profit, or revenue need, was included in the costs on the basis of an operating ratio of 93 percent. An allowance for profit based on a 93-percent operating ratio would provide a profit margin of about 7.5 percent above costs, or stated differently the ratio

of rates to costs would have to amount to 107.5 percent to achieve an operating ratio of 93 percent. Thus for the column A rates to achieve the level of earnings represented by a ratio of 107.5 percent, they would have to be increased by an average of 14.4 percent, whereas the column B rates would have to be increased by an average of 7.5 percent.

Cost presentation

PROTESTANT AND RESPONDENT'S REPLY

Protestant's cost expert after evaluating respondent's study and examining the underlying workpapers makes a two-pronged attack on the study. Respondent replied to the criticisms in a statement dated August 23, 1974. For convenience and a more ready understanding of the dispute protestant's objections and respondent's reply will be treated together.

Protestant's first major thrust is that no validity can be accorded the cost study because of certain lapses or failings. Neither the cost study nor the underlying workpapers provide a way for measuring or correcting for the failures and protestant submits the study must be rejected as invalid resulting in the failure of respondent to sustain its burden of justifying the lawfulness of the assailed increases.

Protestant objects first to the manner in which the study carriers were selected, pointing out that a total of 17 carriers participate in the tariffs, but data was only gathered from 7. This is not a statistically valid sample and there is no showing that the seven carriers studied are representative of the average carrier in the universe. Respondent in reply submits that there was no selectivity in fixing the identity of the study carriers because every carrier handling significant volume of imported vehicles under the considered tariffs, with one exception, was included in the cost study. Of the nine initially polled carriers, one, Kenosha, transported 14,164 vehicles during the first 11 months of 1973 from four terminals of which 9,566 units moved out of Chicago. In contrast, Clark handled 54,816 vehicles in 1973 from Chicago alone. To confine the study to workable limits, costs for the Chicago area were based on information developed by Clark and no further consideration was given to Kenosha. Respondent states that the lack of data from Arco has no material effect on the results of the cost study because the two origins which this carrier serves, Chicago and Port Newark, are represented by Clark at the former and M & G, ATI, Boutell, and Nu-Car at the latter.

In response to protestant's crticism respondent's cost expert directed the respondent tariff bureau to determine the extent of the participation in the movement of imported vehicles of the eight carriers not originally polled. Four of the eight indicated they handled no traffic, three others stated that small or insignificant volumes were moved, while one, P & M Auto transport, handled about 66,000 vehicles from Port Newark of which 39,000 were delivered locally with the balance moving in joint-line service. Data for operations at Newark was supplied by M & G, ATI, Boutell, and Nu-Car. Respondent does not believe the inclusion of data for P & M would reduce costs, but rather would have the opposite effect because of the steep expenses associated with delivery in the New York metropolitan area. Finally it is pointed out that protestant knows which carriers are handling the predominant portion of its traffic, but it has not produced any facts tending to cast doubt on the representativeness of the data submitted by the study carriers.

Protestant questions certain of the underlying data which the study carriers supplied to their cost expert. It notes that Nu-Car reported handling 27,015 vehicles from

Boston, Providence, and New Haven, whereas protestant's records indicate that a total of 30,617 Volkswagens were handled out of the aforementioned terminals. The same carrier reportedly handled 53,018 automobiles from Toledo, whereas Volkswagen itself shipped 52,566 from this same point, suggesting that all but 452 of the cars handled by Nu-Car were Volkswagens. The above data clearly suggests that Nu-Car failed to report all vehicles handled.

Respondent in reply contends that protestant used the wrong count for cars handled by Nu-Car at Boston, Providence, and New Haven. The correct figure as reflected in the cost study 39,839 units. Thus it is clear that Nu-Car handled substantially more than the 30,617 Volkswagens which moved through the three port area. In addition a carrier identified as C & M Trucking Company handled a large volume of imported vehicles through Boston destined to points in Massachusetts in 1973, but these vehicles did not move under the considered tariffs. However, the addition of these automobiles to the total moving through Boston would further widen the spread over the number of Volkswagens handled via the three ports.

Respondent emphasized that it does not claim that the number of imported vehicles reported by the study carriers represents all imported traffic moving through the terminals. The underlying workpapers, a copy of which was furnished protestant, disclosed that Arco handled during the first 11 months of 1973, 16,863 units through Port Newark and 12,356 through Chicago. Moreover Kenosha, which was excused from the study, transported 14,164 vehicles during the same period through its Chicago, St. Paul, and Toledo terminals. These cars were not included in the study. Square Deal and Gate City as well as Kenosha reported that the Volkswagen was the principal car which moved through the Toledo terminal. Thus it is not suprising that Volkswagen should comprise such a large percentage of the total movement through this point.

Respondent emphasizes that loading and unloading expenses were computed on a total of 409,503 vehicles, but that terminal expenses, other than loading and unloading, were based on 449,302 units. The difference is represented by vehicles moving through terminals at Port, Newark, Boston, Providence, and New Haven and is the number of vehicles where drivers from other terminals are used in transporting cars moved through the study terminal. The total number of cars moved through the study terminals is, therefore, 449,392 rather than 409,503.

Protestant is critical of the method used by respondent in allocating expenses to the various service categories. It points out, for example, that some method is needed to apportion driver wages to "line-haul service,” “terminal loading and unloading service," and "other terminal expenses.” The Commission's cost formulas are said to require special studies in order that the aforementioned allocations may be made, but respondent conducted no such studies, relying instead on such subjective tests as "carrier judgement" of "consensus of carriers." Thus the portion of driver wages assigned to loading and unloading was 30 minutes per car. Hours of line-haul service was determined by dividing miles operated by 40 miles per hour. Equipment maintenance expense was evenly divided between the loading and unloading and the line-haul categories of expense. All of the aforementioned assignments were made on the basis of estimates or guesses supplied by the study carriers and protestant believes they have seriously distorted cost allocations. Reference is made to information developed by protestant's tariff consultant, who submitted a verified statement on other matters, which indicates that seven vehicles can be loaded within 45 to 60 minutes. This contrasts with respondent's allocation of 105 minutes to loading.

There is considerable doubt concerning the integrity of the study carriers terminal expense reports. There is no indication who prepared the reports and they are not verified. The terminal expense report of respondent Gate City is simply to the effect that 50.7 percent of system expenses are allocable to the terminal expense category, and is based simply on the fact that system costs bear the same relation to the terminal portion of system revenues.

In reply respondent states that there is no requirement that allocation factors be based on special studies. Protestant has introduced no evidence or showing that the allocation factors used which concededly were based on "carrier judgement" or "consensus of carriers" are incorrect. Concerning the element of loading time, the verified statement of the tariff consultant makes no reference to loading time. Undoubtedly the loading time information was relayed to the witness by unkown sources on his May 1974 visit to Dundalk Marine Terminal, but there is no solid evidence to this effect.

Respondent concedes that 15 minutes per vehicle was allocated for loading and the same for unloading. This factor was based on a “consensus of the carriers" which took into account, with respect to loading, such related work items as check in, paperwork, and contact time, and with respect to unloading such additional work items as contacting consignees, skid drops, and paperwork. A labor-management agreement applicable at Edison, N.J., the production site of the Ford Pinto, provides that the loading service consumes 15 minutes per vehicle. Thus for a trailer loaded with seven vehicles drivers receive loading time pay of one and three-quarters hours, or 15 minutes a vehicle which is exactly the time allocated by the study.

Concerning the criticism of the study carriers' terminal expense reports, respondent states that protestant, had it asked, would have been furnished the names of the officials which provided the reports. Gate City allocated 50.7 percent of system expenses to the Toledo terminal based on the revenue relationship of the company's Toledo and Detroit operations. There is no reason to believe that the expense distribution between the two terminals is not closely related to the revenue distribution.

Protestant questions the validity of the 34 different ratios used by respondent in restating the full year 1973 expense figures to the January-February 1974 level. None of the ratios was based on a reliable study. All are simply "carrier estimates." Protestant holds that the assailed 5-percent increase in rates does not justify respondent in showing similar increases in the categories of purchased transportation and equipment rents without drivers. A study of owner-operators actual costs is said to be required.

Respondent concedes that no study of owner-operator costs was made but its submits that carrier payments for purchased transportation in developing terminal and line-haul costs is proper. Furthermore, the decision in Motor Carrier Rates-OwnerOperators, 431 I.C.C. 28, permits the use of regional average costs as a test of the compensativeness of a given rate, and relying thereon respondent compares the costs developed by its study with the regional average cost per vehicle mile shown in I.C.C. statement No. 2C1-71. Because most of the considered traffic moves between New England and the Middle Atlantic Region and within the Middle Atlantic and Central regions, the variable cost per vehicle mile (loaded or empty) in these regions as shown in the aforementioned statement is multiplied by the study carriers' loaded and empty mile figure of 19,216,879 to give the total line-haul expense. Dividing the latter by the loaded mile figure of 10,100,720 miles gives the cost per loaded vehicle mile. Recognizing that motor carrier costs increased by 10 percent from July 1971 to July