We are once again confronted with "another episode in the long and continued struggle between the railroads and competing barge lines." A. L. Mechling Barge Lines, Inc. v. United States, 376 U.S. 375, 376, 84 S. Ct. 874, 11 L. Ed. 2d 788 (1964). In this return engagement to the reviewing courts, Waterways Freight Bureau (WFB or petitioner), representing common carrier barge lines handling almost all of the regulated traffic along the Mississippi River system1 petitions this court to review and set aside2 a report and final order of the Interstate Commerce Commission, acting through its Division 2,3 which granted an application filed by member railroads of the Southwestern Freight Bureau for special relief from the Interstate Commerce Act's general prohibition against long- and short-haul rate discrimination.4 Interstate Commerce Act § 4(1), 49 U.S.C. § 4(1) (1970). We vacate the Commission's decision and remand for further proceedings.

In an effort to attract a greater share of the traffic then moving by the barge lines, the applicant-railroads filed an application with the Commission in September 1972 in which they proposed to reduce their rates 16.7 percent, from $16.80 to $14.01 per ton, on iron and steel shipments from St. Louis and its environs to Houston. Since the existing rates would continue in effect for intermediate destinations along the St. Louis-to-Houston route and would be higher than the proposed rate in certain cases, the railroads were required at the same time to seek relief from the general proscription of section 4(1). See generally Intermountain Rate Cases, 234 U.S. 476, 34 S. Ct. 986, 58 L. Ed. 1408 (1914). This "fourth-section" application was supported by one shipper, Granite City Steel Company, which during 1971 had shipped 92.6 percent of more than 37,000 tons of various steel products by barge from St. Louis to Houston. J.A. 332. The proposed rate of $14.01 per ton was calculated at 110 percent5 of the shipper's aggregate costs for comparable transportation of its product by barge. Both the railroads and the shipper claim that this rate was determined upon lengthy negotiations to be "the highest rate which would permit movement of the traffic by rail."6 Id.; Intervenor's Brief at 3. The barge lines, on the other hand, argue that the proposed rate is too low.

The parties on appeal are in substantial agreement on the governing legal principles.7 Section 4(1) of the Act, which is as old as the Commission itself, Act of Feb. 4, 1887, § 4, 24 Stat. 380, essentially prohibits a railroad, pipeline or water common carrier from collecting a greater total charge for shipments over shorter distances than over longer, when both involve the same route and the same direction.8 This general prohibition, however, is subject to an exception which provides

(t)hat upon application to the Commission and after investigation, such carrier, in special cases, may be authorized by the Commission to charge less for longer than for shorter distances for the transportation of passengers or property, and the Commission may from time to time prescribe the extent to which such designated carriers may be relieved from the operation of the foregoing provisions of this section, but in exercising the authority conferred upon it in this proviso, the Commission shall not permit the establishment of any charge to or from the more distant point that is not reasonably compensatory for the service performed . . . .

49 U.S.C. § 4(1) (1970) (emphasis added).

The burden on an applicant common carrier seeking special relief from this section is a heavy one. See, e. g., Mechling Barge Lines, 376 U.S. at 385, 84 S. Ct. 874, 11 L. Ed. 2d 788; Louisville & N. R. Co. v. United States, 225 F. 571, 581 (W.D.Ky.1915), aff'd, 245 U.S. 463, 38 S. Ct. 141, 62 L. Ed. 400 (1918); J.A. 338; Petitioner's Brief at 36. Thus, to be " reasonably compensatory" so as to qualify for dispensation from section 4(1), a proposed rate must

(1) cover and more than cover the extra or additional expenses incurred in handling the traffic to which it applies; (2) be no lower than necessary to meet existing competition; (3) not be so low as to threaten the extinction of legitimate competition by water carriers; and (4) not impose an undue burden on other traffic or jeopardize the appropriate return on the value of carrier property generally . . . .

Transcontinental Cases of 1922, 74 I.C.C. 48, 71 (1922) (emphasis added); e. g., Rules to Govern the Filing of Fourth Section Applications, 310 I.C.C. 275 (1960); Igert v. United States, 211 F. Supp. 42, 44-45 (N.D.Ala.1962). Of these four criteria, we need concern ourselves on this appeal only with the second, for WFB's essential position collapses basically to the argument that the railroads failed to meet their burden of showing that the proposed $14.01 rate was "no lower than necessary to meet existing competition."9

This case consequently boils down to little more than a dispute over the discrete cost items to be included in determining the shipper's total cost of moving the goods to Houston via barge. The railroads based their proposed rate on the following calculation:

Origin loading cost    --  $ 2.00 per net ton
Barge rate (St. Louis
  to Houston)          --    7.98
Houston transfer       --    1.60
Interstate switching   --    1.16
                           ------
        Total              $12.74 per net ton
        10% Add on     --    1.27
                           ------
        TOTAL              $14.01 per net ton

See Petitioner's Brief at 12; J.A. 332.

Petitioner discredits this calculation, claiming that the total cost of utilizing barge transportation actually is significantly higher because most ex-barge traffic in Houston travels by truck, not railroad, and that at least half of all such truck deliveries to Houston-area customers suffer delays that result in additional dockside storage and reloading charges. While it does not dispute the first three above-described cost items, petitioner vigorously argues that to reflect accurately the actual total cost of getting the shipper's products to their ultimate destination, the Houston-area customer, the ex-barge interstate rail switching charge of $1.16 per net ton should be replaced by a "drayage" charge of $2.00 per net ton for land movement in Houston and a Houston pier transfer and reloading charge of $1.00 per net ton. Petitioner therefore contends that the actual cost of the subject barge and ex-barge shipments, and perforce the lowest railroad rate "necessary to meet existing competition" by its barge line members, is not $14.01 per net ton as proposed but rather $16.04.10 Quite obviously, the higher the railroads' rate the lower the intermodal competition and the less the barge lines need fear the railroads' encroachment into the market for the shipper's transportation requirements, such as it now is.

In passing upon WFB's claims we are mindful of the limited nature of our review. If the Commission's action in this fourth-section proceeding is not strictly one of ratemaking in the usual sense, as counsel for the barge lines asserted during oral argument, it is at least for our current purposes functionally indistinguishable from ratemaking. See Igert v. United States, 211 F. Supp. at 44; Seatrain Lines, Inc. v. United States, 168 F. Supp. 819, 826 (S.D.N.Y.1958). This being so, as we believe it clearly is, the rate-determining action in the instant case is a product of the agency's informal rulemaking powers, see 5 U.S.C. §§ 551(4), 553 (Supp. V 1975); see also United States v. Florida East Coast Railway Co., 410 U.S. 224, 93 S. Ct. 810, 35 L. Ed. 2d 223 (1973), for which the proper standard of review is that specified by 5 U.S.C. § 706(2)(A)-(D) (1970), and principally by paragraph (A). Basically, then, we may disturb the Commission decision only if we find that it is "arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law . . . ." Id. § 706(2)(A). We may require no more than that a rational basis be discernible "in the Commission's opinion . . . for its treatment of the evidence." Bowman Transportation, Inc. v. Arkansas-Best Freight System, Inc., 419 U.S. 281, 290, 95 S. Ct. 438, 444, 42 L. Ed. 2d 447 (1974); see, e. g., United States v. Allegheny-Ludlum Steel Corp., 406 U.S. 742, 748-49, 92 S. Ct. 1941, 32 L. Ed. 2d 453 (1972); National Association of Food Chains, Inc. v. ICC, 535 F.2d 1308, 1313-14 (D.C.Cir.1976).

This less stringent standard of review is especially appropriate in ratemaking cases, since

(t)he process of rate making is essentially empiric. The stuff of the process is fluid and changing the resultant of factors that must be valued as well as weighed. Congress has therefore delegated the enforcement of transportation policy to a permanent expert body and has charged it with the duty of being responsive to the dynamic character of transportation problems.

Board of Trade of Kansas City v. United States, 314 U.S. 534, 546, 62 S. Ct. 366, 372, 86 L. Ed. 432 (1942); see, e. g., ICC v. Union Pacific Railroad Co., 222 U.S. 541, 550, 32 S. Ct. 108, 112, 56 L. Ed. 308 (1912) ("there is no possibility of solving the question as though it were a mathematical problem to which there could only be one correct answer").

Notwithstanding these strict limits upon our review, we conclude that the decision challenged in this appeal cannot survive the minimal requirements of rationality on the present state of the administrative record.

The Administrative Law Judge (ALJ) found that the total transportation cost from the shipper to the Houston-area receivers more closely approximated WFB's calculation than the applicant railroads', J.A. 198, 335, though he ultimately held that the second criterion of the four-prong test enunciated in Transcontinental Cases of 1922, had been rendered nugatory by the enactment of 49 U.S.C. § 15a(3). The railroads did except to this "finding" in favor of the WFB's proffered figure, J.A. 289-90, but, on appeal, Review Board No. 4's ratio decidendi again entailed a section 15a(3) analysis and not an analysis of whether the proposed rate was "no lower than necessary to meet existing competition." J.A. 301. The Commission, however, expressly disagreed with the ALJ's cost analysis and his conclusion that the WFB's figure was the more accurate one. J.A. 340. It held that "the actual costs incurred to the shipper of moving these commodities by barge more closely approximates (the applicants' figure) . . . ." J.A. 342 (emphasis added). The Commission's report gives the following reasons in support of this conclusion:

We find no basis in evidence to substantiate the $2.00 per net ton drayage charge. As noted in the initial decision, much of the beyond-Houston tonnage (approximately 90 percent according to applicants' evidence) is carried by truck in the customers' private, unregulated vehicles, and this is not disputed by protestants. Drayage charges for these shipments are generally paid by the customer to the Houston stevedores, and not by the shipper. Generally, all additional and potentially required movements after destination at Houston, such as drayage, are paid by the receiver and not the shipper, whether the line-haul shipment is by barge or rail and whether moved afterwards by common carrier or private carriage. The proposed rail rate of $14.00 ($14.01) per net ton has been computed on a basis of 110 percent of the shipper's total through costs of moving these commodities by barge. Since drayage is generally paid by the receiver at Houston, this charge would not be included in the computation of the shipper's total barge costs, and therefore would not (sic) used by applicants in the construction of their rate. Since a portion of the ex-barge traffic does move by rail, we find the $1.16 per net ton switching charge the best evidence of record in calculating the shipper's total costs, and find its use applicable in this situation.

While protestants' assertions of fact in regard to additional unloading and reloading at Houston can be properly accepted, we find no basis in evidence to substantiate the extra $1.00 per net ton charge beyond the $1.60 per net ton direct transfer charge. As in the situation with drayage charges noted above, 90 percent of the beyond-Houston tonnage moves by private carriage. These charges are generally paid by the customer to the Houston stevedores, and could not possibly be computed on a per net ton basis to all of Granite City's private carriage receivers which require the extra handling. Further, additional common carrier ex-barge movements above the direct transfer charge, such as extra rehandling, are generally paid by the receiver and not by the shipper as part of his total costs. As such, this extra charge would not be included in the computation of the shipper's total barge costs, and therefore would not be used by applicants in the construction of their rate. Further, assuming the extra handling charges for even a small portion of this traffic would be paid for by the shipper, we could not possibly compute on a per net ton basis this extra cost to the shipper, because there is no method with which we could determine the percentage of traffic in which the shipper would pay these charges so as to verify a $1.00 figure or any per net ton figure. Based on the above, we find the additional $1.00 per net ton extra handling charge inapplicable in the computation of the shipper's total costs of moving the issue commodities by barge.

J.A. 340-42.

Petitioner challenges the assumption underlying the above-quoted reasoning, namely that the primary issue involved is one of cost to shippers only and that, consequently, the charges generally paid by receivers are not relevant inputs into the fourth-section calculus. It argues rather that the relevant inquiry is the total cost from origin to ultimate destination, no matter who pays the various shipping charges, because "all charges incurred in the use of a particular route eventually will be reflected in the price paid for goods and ultimately must be considered in any rational selection of the lowest cost method of transportation." Petitioner's Reply Brief at 4.

In support of this argument, petitioner refers us to two Commission decisions which it claims demonstrate that heretofore the agency in fourth-section cases has always included all charges incurred in using barge service irrespective of whether the consignor or consignee paid any or all of them. Steel Pipe From The East To The Southwest, 301 I.C.C. 203, 223 (1957); Pig Iron From New York Points To Detroit, 296 I.C.C. 747, 750 (1955); see Petitioner's Brief at 33-34; Petitioner's Reply Brief at 4. Neither the Commission nor the intervenor railroad has troubled to distinguish or even cite these two cases, apparently choosing instead to cast a veil of silence over the difficulties they raise. We cannot ignore these cases, and neither should the Commission.11

In the Steel Pipe case, supra, the Commission held that certain proposed all-rail rates were lower than necessary to meet the existing competition and, consequently, denied the fourth-section applications. Recognizing that establishing differential rate treatment of competing services was not an accurate science, the agency noted that the difference in value of the competing services over their respective routes must nonetheless be considered in passing upon the applications.12 See 301 I.C.C. at 223. It then went on to analyze and discuss various intermodal "value differences" that were evident from the record before it. One of these resulted from the fact that some producers did not have direct access to water service and, hence, had to pay additional costs to get their goods from their plants to the barges. This, of course, presents a different factual situation from our immediate case because Granite City Steel maintains a barge dock only a short distance from its steel mill, J.A. 194, and origin loading costs of $2.00 per net ton were included in the railroads' calculation. J.A. 332. The Commission in Steel Pipe, however, did emphasize, in apparent contrast to its present position, that it is not necessarily who pays any immediate shipping charge but rather who ultimately bears the transportation costs and influences or makes the traffic-routing decisions that is analytically germane.

The (railroads) consider all costs which are incidental to the barge loadings as outside the competitive scope of the respective rates in that the consignee, who pays the freight, is interested only in the charges beyond the loading points. But this record is not convincing that the producers who would bear these incidental costs would have no influence in the routing so as to avoid such expenses.13

301 I.C.C. at 223 (emphasis added). We believe that the record in the present case is similarly barren.

Even more to the point, the Commission in Steel Pipe also stressed that the ex-barge costs at the destination port were a material difference between barge and rail services warranting consideration.

The (railroads) apparently constructed their all-rail rates on the belief that transfer charges at the ports were generally absorbed by the motor carriers, and that some shipments move directly through the ports and do not incur the charge. But other than the testimony of 1 shipper . . . there is little evidence of record to establish that such charge is not generally incurred. On the contrary, several shipper witnesses supporting the (railroads) readily admitted that such charges were being incurred, and, as seen, absorption of such charges by the motor carriers was found to be unlawful in Oilfield Equipment to and between the Southwest, supra. In their exceptions, the (railroads) argue that barge-truck service without storage is available and alone necessitates this competitive level of rail rates. In a particular situation this might be true. But where, as here, an extensive adjustment of rates is under consideration, all factors must be viewed in their proper perspective to the whole, and we cannot ignore the fact that barge traffic generally is stored at the ports and is subject to additional charges therefor. These factors should have been given some consideration in the establishment of the all-rail rates.

Id. at 223-24 (emphasis added).

In the other case emphasized by WFB, Pig Iron From New York Points To Detroit, the railroads filed a fourth-section application proposing a reduced rate to enable them better to compete for the pig iron traffic from Buffalo to Detroit, of which they then transported only about two percent. Again, the parties disputed the total costs of the barge service, including whether barge loading and certain dock storage and incidental handling costs at the destination port should be included in the relevant fourth-section cost equation. In concluding that both sets of costs should be so included, 296 I.C.C. at 751-52, the Commission explained:

We have consistently held that, in order to justify rail rates lower than normal rail rates to meet water competition, the proposed rates should not be lower than necessary to meet the competition, which is indicated by the cost to the shipper for the transportation over the water route, plus such incidental expenses as are incurred in using the competitive service. Due to somewhat less desirable service, it is often necessary that the water carriers maintain rates lower than those in effect over competing all-rail routes, but the proper relation of the rates depends upon the circumstances in each case.

Id. at 749-50 (emphasis added).

These two cases referred to by the petitioner may not ultimately be dispositive of the cost-evidence issue in the instant proceeding, but they do indicate that in other cases, not dissimilar to the one sub judice, the Commission has required that the cost calculation in a fourth-section application include all incidental costs incurred in shipping via the competing service and has required more than conclusory factual assertions in determining which cost-bearers may influence the routing decision.14

It may well be, as our intervenor argues, see Intervenor's Brief at 7-9, that in cases where the receiver is not on a railroad siding and must thus use truck service from the destination port the ex-rail costs will approximate the ex-barge costs (i. e., shipments by either barge or rail would incur similar reloading and drayage charges) so that neither should be included in computing an acceptable fourth-section rail rate. This argument must await the Commission's attention, however, for the evidence of record as it currently exists cannot adequately support a factual finding necessary to resolve the point either one way or the other.

Although the Commission is owed considerable deference when bringing its expertise to bear in ratemaking action, it cannot be permitted simply to abandon the rules of the game ad libitum. Some explanation is due both the parties and the reviewing court when a decision is made that conflicts with relevant precedent. The apparent departure from a method of analysis previously applied in resolving other intermodal rate competition cases calls for a reexamination and some reasoned explanation by the Commission which is now missing from the record we review. See generally Parkhill Truck Co. v. United States, 559 F.2d 834 (D.C.Cir.1977).

On remand, the Commission should reconcile its disposition of these proceedings with the principles applied in the Steel Pipe and Pig Iron cases, supra, and take a hard look15 at the respective ex-rail and ex-barge service disabilities in considering whether the proposed rate at issue is no lower than necessary to meet the competition of the barge lines. Specifically, the Commission should address itself once again to the question whether the within-Houston ex-barge costs are actually within or without the competitive scope of the respective barge and rail rates. We accordingly vacate the Commission's order in this case and remand the record for further proceedings not inconsistent with this opinion.

So ordered.