Full opinion text
TABLE OF CONTENTS Page I. Background....................................................... 1034 II. Maximum Rate Exkmition.................................... 1036 A. The Commission Decision............................. 1036 B. Abuse of Market Power............................... 1038 1. The standard of review.......................... 1038 2. “General” market constraints.................. 1039 3. The Conrail study.................................. 1040 4. Particular commodities............................ 1042 5. Conclusion as to rail carriers’ market power 1042 C. The Scope of the Maximum Rate Exemption. 1043 III. Joint Rates and Through Rates.......................... 1044 A. The Commission Decision............................. 1045 B. Carrying Out the Rail Transportation Policy.. 1046 1. Incentives for large carriers to close efficient routes........................................... 1047 2. Division of joint rates............................ 1048 IV. Car Hire Decision............................................. 1051 A. Background................................................. 1051 B. Analysis...................................................... 1053 V. The Alaska Railroad......................................... 1060 A. Predatory Pricing and Practices.................... 1061 B. Discriminatory Ratemaking........................... 1064 VI. The Canadian Railroads..................................... 1066 VII. The Board of Port Commissioners for the City of Oakland............................................................ 1068 VIII. Conclusion........................................................ 1070 Before ROBINSON, Chief Judge, and WALD and MIKVA, Circuit Judges. Opinion PER CURIAM. PER CURIAM : We today confront the relationship between the Interstate Commerce Commission (ICC or Commission), railroad carriers and shippers, and the deregulatory objectives embodied in the Staggers Rail Act of 1980, Pub.L. No. 96-448, 94 Stat. 1895 (1980). Petitioners in this case seek review of four related Commission decisions that exempt boxcar traffic from rate regulation and that substantially modify the basis upon which railroads may use boxcars owned by others. These decisions, affecting approximately twenty-five percent of all rail traffic moving throughout the United States, apply to all geographic regions, to all railroads, and to all types of boxcars —equipped, unequipped, refrigerated, and livestock. Thus, it is hardly surprising that numerous parties petitioned this court to review the Commission’s decisions and that our opinion today addresses the merits of over thirty consolidated cases. Petitioners assert a host of challenges to the Commission’s decisions. The petitioners raise questions concerning the Commission’s statutory authority and questions regarding the substance of many Commission conclusions. In response, the Commission rests on that section in the Staggers Act, 49 U.S.C. § 10505(a) (Supp. V 1981), which allows the ICC to deregulate the railroads, and argues that its decisions are supported in the record and are consistent with congressional goals. For the reasons set forth below, we find merit in some of the petitioners’ arguments and conclude that in parts of its decisions the Commission failed to consider certain factors and, additionally, exceeded the scope of its statutory authority. We thus affirm in part, vacate in part, and remand the case to the Commission. I. Background On May 22, 1981, the Consolidated Rail Corporation (Conrail) petitioned the Interstate Commerce Commission (ICC or Commission) to deregulate all aspects of the transportation Conrail provided in boxcars. Conrail argued that the rates it received for boxcar traffic did not cover its variable costs. “The rates are wrong, the divisions are too low, and the car hire rates are unfair.” Conrail pointed to the nationwide surplus of boxcars as evidence of problems with the current regulations. In response, the ICC instituted a rulemaking proceeding, proposing the application of Conrail’s deregulation petition to all boxcar traffic nationwide. 47 Fed.Reg. 4100 (1982). Following the Commission’s receipt of many negative comments, Conrail offered a less extensive proposal. In a series of four decisions, issued from April 1983 through December 1983, the Commission adopted Conrail’s modified proposal on a national basis. In Exemption from Regulation—Boxcar Traffic, 367 I.C.C. 424 (1983) (Boxcars I), the Commission reached its initial decision. In Exemption from Regulation—Boxcar Traffic, 367 I.C.C. 747 (1983) (Boxcars II), the Commission denied petitions for reconsideration and further refined its analysis. In Exemption from Regulation—Boxcar Traffic, served November 30, 1983 (not printed) (Boxcars III), the Commission denied petitioners’ motions for a stay pending appeal. And, in Exemption from Regulation— Boxcar Traffic, served December 19, 1983 (not printed) (Boxcars IV), the Commission addressed certain issues raised by the Association of American Railroads, specifically focusing on questions of antitrust liability. The Boxcars decisions became effective January 1, 1984, with the exception of the car hire modifications as applied to Class III carriers (small railroads). As to Class III carriers, the decisions were to become effective July 1, 1984. The Commission initially based its decision on its power to deregulate, 49 U.S.C. § 10505(a), and on its power to regulate the compensation paid for the use of freight cars, 49 U.S.C. § 11122. See Boxcars I, 367 I.C.C. at 456 (“We recognize, however, that [Conrail’s modified proposal] could be construed in some respects as being new regulation. To allay any doubt about the sufficiency of section 10505(a) as authority for our approval of Conrail’s modified proposal, we shall take this action also under section 11122.”). In Boxcars II, however, the Commission withdrew any reliance on its regulatory powers under section 11122, Boxcars II, 367 I.C.C. at 759. Thus, the Commission’s decision rests entirely on its deregulatory powers. The Commission’s deregulatory powers trace to section 10505(a) of the Staggers Act. That section provides: (a) In a matter related to a rail carrier providing transportation subject to the jurisdiction of the Interstate Commerce Commission under this subchapter, the Commission shall exempt a person, class of persons, or a transaction or service when the Commission finds that the application of a provision of this subtitle— (1) is not necessary to carry out the transportation policy of section 10101a of this title; and (2) either (A) the transaction or service is of limited scope, or (B) the application of a provision of this subtitle is not needed to protect shippers from the abuse of market power. 49 U.S.C. § 10505(a) (Supp.V 1981). In turn, section 10101a, referenced in subparagraph 1, lists fifteen different national rail transportation policies. 49 U.S.C. §§ 10101a(l)-(15) (Supp.V 1981). The first focus of the Boxcars decisions is the Commission’s regulation of the rates that a railroad carrier may charge for the transportation of freight in boxcars (freight rate). Prior to the decisions under review, the Commission had jurisdiction to find rates unreasonably high when such rates exceeded specified ratios of revenue to variable cost. 49 U.S.C. 10701a(c)(4)(B). Moreover, the Commission also exercised certain supervisory powers over the rates for shipments transported jointly by two or more connecting carriers — that is, joint rates. See, e.g., 49 U.S.C. §§ 10705, 10705a. The Boxcars decisions removed all regulations on freight rates, including joint rates. The Commission found that freight rates could be deregulated under section 10505(a) of the Staggers Act because the regulations were unnecessary to implement the national transportation policy and because the regulations were not needed to protect shippers from an abuse of market power. The linchpin in the Commission’s “abuse of market power” analysis was the pervasive pattern of competition between trucks and rails, as evidenced at least in part by an admittedly controversial study submitted by Conrail. The Commission, however, indicated that it would reimpose freight rate regulations to any commodity in which shippers subsequently introduced evidence revealing an abuse of market power and a need for protection. See Boxcars I, 367 I.C.C. at 40-41. The Commission also concluded that its deregulatory posture would not harm Class III carriers (small railroads). These carriers feared that the deregulation of joint rates would allow large carriers to establish rates that would undermine the competitiveness of the joint route, or would otherwise lead to the demise of many small carriers. The Commission found this argument unpersuasive and concluded that, in light of the pervasive truck competition, large carriers would not cancel joint rates with Class III carriers where the Class III carrier provided an efficient route. Boxcars II, 367 I.C.C. at 753-54. Moreover, the Commission concluded that the economic disaster forecast by the Class III carriers would not occur because small carriers have sufficient bargaining power to protect themselves. Id. at 766-67. The Commission conceded that some inefficient Class III carriers might be driven out of business, but concluded that the cancellation of such inefficient routes was in the public interest. Id. at 754. The Commission retained jurisdiction over mandatory interchange, reciprocal switching, and the joint use of terminal facilities. The next focus of the Boxcars decisions is car hire — the financial relationship between boxcar owners (originating carriers) and the railroad over which the cars travel (destination carriers). Prior to the Boxcars decisions, a carrier that owned or leased boxcars received compensation from the railroad over which its cars were traveling. The amount of such compensation, known as the per diem rate, was based on a Commission-established formula that distributed the costs of car ownership evenly over the days in which the car was in service. This per diem rate accrued for all periods that the boxcar was off line, regardless of whether it was empty or full. Thus, the destination carrier paid per diem even after the car was sitting idly on its tracks. This system of compensation, the Commission found, created incentives for the originating carrier to load its own cars and to return other cars empty. The Commission argued that as a result of this incentive, American railroads were inefficiently devoting substantial resources to the movement of empty boxcars. Moreover, the Commission suggested that the per diem structure contributed to the current surplus of boxcars by guaranteeing a return on boxcar costs, regardless of the ratio of supply to demand. The Boxcars decisions modify the relationship between the originating carrier and the destination carrier. The Commission believed that these modifications, which it considered a “partial exemption from regulation subject to conditions”, would ameliorate market efficiencies because incentives to ship empty boxcars would be reduced and because market factors would gain importance in the pricing of car hire. Although leaving untouched the basic per diem structure, the Commission modified the car hire relationship in three significant ways. First, the destination may now impose storage charges on the originating carrier for boxcars that have been empty for seventy-two hours. Because these storage charges are designed to offset the per diem rate that the destination carrier owes the originating carrier, the storage charges may not exceed the corresponding per diem. Second, the destination carrier may charge a fee for the return of an empty boxcar when such return is requested by the originating carrier. The charge, however, cannot exceed thirty-five cents per mile. Third, carriers that enter agreements governing car hire rates, empty movements, and storage are freed from the Commission’s car hire rules. Our opinion addresses each major issue seriatim. In Part II, we find that the Commission’s “abuse of market power” analysis withstands review. In Part III, we hold that the Commission’s analysis of the deregulatory impact on small carriers and joint rates is arbitrary and capricious. In Part IV, we hold that the Commission exceeded the scope of the Staggers Act exemption provision in adopting the new car hire rules. In Part V, we hold that the Commission improperly included the Alaska Railroad within the scope of its rate exemption. In the remainder of the opinion, we address, and reject, the complaints of the Canadian carriers and the Port of Oakland. II. Maximum Rate Exemption A. The Commission Decision The Commission exempted freight rates that carriers charge shippers for boxcar transportation from all regulation because it concluded that rate regulation was not needed either to further the rail transportation policy of 49 U.S.C. § 10101a, or to protect shippers against abuses of market power. Its conclusions were based on a finding that boxcar freight rates would not increase unreasonably in the absence of regulation because rail carriers did not have sufficient market power in the transportation market for goods that travel by boxcar to inspire unreasonably high prices. This finding was relevant to both inquiries under the Staggers Act exemption provision, 49 U.S.C. § 10505(a). First, one aspect of the rail transportation policy is the maintenance of reasonable rates. 49 U.S.C. § 10101a(6). . Second, the existence of unreasonable rates is a symptom of market power abuse. Focusing primarily on intermodal and intramodal competition, the Commission concluded that “the market itself places an effective ceiling on rail rates for boxcar transportation, and regulation is unnecessary to assure that boxcar rates do not rise to unreasonably high levels.” Boxcars I, 367 I.C.C. at 433. The linchpin in the Commission’s analysis is the presence of pervasive truck competition — its conclusion that goods transportable by boxcars can also, in the vast majority of cases, be carried by trucks. But the ICC relied as well on intramodal competition, the shipper’s ability to select alternative railroad routes, to use other non-boxcar railroad equipment to carry goods, or finally to transport goods by trailer on flat car (TOPC) or container on flat car (COFC). The Commission further reasoned that rates would be restrained by the carriers’ reluctance to set a rate so high that the shipper could not compete in the ultimate product market. Thus, it concluded, the carrier would not fix its rates at a level that would drive the shipper out of business because in that event the railroad itself would lose the shipper’s business. It also focused on the fact that many large shippers operate from several locations. “Thus, even if a carrier should find itself in a position to charge an unreasonably high rate to the company at one location, it normally would refrain from doing so to avoid a retaliatory loss of the shipper’s business at that or at other locations where competition exists.” Boxcars I, 367 I.C.C. at 434. In addition to these generic constraints on boxcar rates the Commission looked at data on past boxcar transportation in assessing railroads’ market power and the likelihood of abuse of such power. In particular it looked at the percentage of total traffic handled by boxcars and by trucks, and at revenue to variable cost ratios (r/vc) for boxcar transportation of commodities. According to the Commission, this data supported a conclusion that railroads lacked the necessary degree of market power over the transportation of commodities that travel by boxcar to pose a potential for abuse. The primary factual basis for the Commission’s conclusion was a 1980 Conrail study exploring 18 groups of commodities moving to, from, or within the Northeast. This report indicated that for each commodity group in that region, trucks had a substantial share of the market and r/vc did not exceed the relevant percent threshold for ICC authority to review rates. See 49 U.S.C. §§ 10701a, 10709. Within these groups, however, the percentage of truck carriage varied substantially, from very high to very low. The Commission also cited a DOT 1977 nationwide study indicating that for 25 groups of commodities, all had a truck market share of at least 20%. Finally, the Commission reviewed specifically “those few commodities about which the greatest cause for concern has been presented.” Boxcars I, 367 I.C.C. at 436. The Commission focused on auto parts, paper and forest products, grain, and metals, and concluded that in none did the railroads have market power for section 10505(a) purposes. In sum, the Commission relied on the pervasive and increasing role of trucks in transporting commodities that typically travel by boxcar, buttressed by the Conrail study, the DOT study, and general findings about competition in the boxcar freight transportation market, to conclude the regulations generally were unnecessary to protect the shippers from an abuse of market power. It then looked at several specific commodities — those which petitioners’ comments identified as presenting the closest questions — and decided that its conclusions that rail carriers held no market power over shippers held true for those specific commodities as well. The Commission recognized that its determination that railroads overall were not in a position to abuse market power could mask individual situations vis-a-vis particular shippers, commodities or locations that might hold a potential for carrier abuse. But in light of the strong deregulatory thrust of the Staggers Act, especially of the exemption provision, it decided that an exemption for boxcar rates was nonetheless justified, stating: Conceivably, there exists a specific commodity that for some reason can be transported only by boxcar and on which the rate levels, being uncontrolled by intramodal competition or other market forces, would rise to very high levels in the absence of regulation. We do not totally rule out the possible existence of such a situation. Congress desires that we move aggressively forward in extending the exempt sector of rail transportation without laboriously erasing every shadow of a doubt before we act____ Congress declared that we should instead make exemptions effective and then deal with any special problem “after the fact,” not on a priori possibilities. Boxcars I, 367 I.C.C. at 440-41. Petitioners, in the main, challenge the Commission’s decision to exempt boxcar rates in the face of these acknowledged uncertainties about future abuses of market power in particular situations. B. Abuse of Market Power 1. The standard of review This court must review the Commission’s exemption of boxcar freight rates from regulation, implementing the exemption provision of the Staggers Act, 49 U.S.C. § 10505(a), to see if it was arbitrary and capricious, an abuse of discretion or otherwise contrary to law. See 5 U.S.C. § 706; National Small Shipments v. Civil Aeronautics Board, 618 F.2d 819, 826 (D.C.Cir.1980) (applying section 706 to exemption of airlines from regulation); American Trucking Associations v. ICC, 656 F.2d 1115, 1125 (5th Cir.1981) (applying “arbitrary and capricious” standard to exemption of trailer or flat car transport from regulation under 49 U.S.C. § 10505(a)). Petitioners here challenge the ICC’s finding that continued regulation is not necessary to protect shippers from abuses of market power on the ground that this finding was arbitrary and capricious. “The scope of review under the ‘arbitrary and capricious’ standard is narrow and a court is not to substitute its judgment for that of the agency.” Motor Vehicles Manufacturers Association v. State Farm Mutual Automobile Insurance Co., 463 U.S. 29, 103 S.Ct. 2856, 2866-67, 77 L.Ed.2d 443 (1983) [hereinafter cited as Airbags ]. Nevertheless, the agency must consider all critical aspects of the problems before it, and must articulate a reasoned explanation for its action, including “a rational connection between the facts found and the choice made.” Burlington Truck Lines v. United States, 371 U.S. 156, 83 S.Ct. 239, 9 L.Ed.2d 207 (1962). Petitioners here suggest that judicial review must begin with a presumption that significant changes in current policy from traditional norms of regulation must be explained, and “where, as here, the challenged action represents ‘a departure from prior policies and precedents,’ the Court’s ‘level of scrutiny’ is heightened.” Brief of Petitioners and Intervenors on Common Issues at 35 (quoting Office of Communication of United Church of Christ v. Federal Communications Commission, 707 F.2d 1413, 1425 (D.C.Cir.1983)) [hereinafter cited as Joint Brief for Petitioners]. But, the Supreme Court has recently emphasized that abandonments of existing rules and policies are not to be reviewed under a heightened standard of scrutiny, see Airbags, 103 S.Ct. at 2865; rather the agency must explain why the original reasons for adopting the rule or policy are no longer dispositive. See id. at 2866 (“an agency changing its course by rescinding a rule is obligated to supply a reasoned analysis for the change”); International Ladies’ Garment Workers’ Union (ILGWU) v. Donovan, 722 F.2d 795, 813 (D.C.Cir.1983). We would note as well that this case differs materially from Airbags and ILGWU in one respect. Congress itself has found that the structure of the transportation industry has changed so that “many of the Government regulations affecting railroads have become unnecessary and inefficient,” H.R.Rep. No. 1430, 96th Cong., 2d Sess. 79 reprinted in U.S.Code Cong. & Ad.News 1980 pp. 3978, 4110, 4111 (conference report) [hereinafter cited as Conference Report], and has furthermore commanded the Commission to remove by exemption “as many as possible of the Commission’s restrictions on changes in prices and services by rail carriers.” Id. at 105, U.S.Code Cong. & Admin.News 1980, p. 4137. Given that explicit congressional mandate, we do not believe the Commission need as exhaustively review and explain away its original justifications for abandoned regulations as if it were operating under the same statute it always had. The Commission, on the other hand, argues that we must be especially deferential in reviewing its exemption decisions since they inevitably involve judgments and predictions of economic consequences and behavior which are inherently uncertain in nature but which Congress expressly delegated to the Commission. Brief for Respondents at 33-34; see also American Trucking Associations, 656 F.2d at 1127. We do, of course, recognize that in some circumstances “complete factual support in the record for the Commission’s judgment or prediction is not possible or required.” Federal Communications Commission v. National Citizens Committee for Broadcasting, 436 U.S. 775, 814, 98 S.Ct. 2096, 2122, 56 L.Ed.2d 697 (1978); see also National Small Shipments v. Civil Aeronautics Board, 618 F.2d 819, 829 (D.C.Cir.1980) (applying principle to predictions about deregulation’s effects on airline price competition and price discrimination). But we must point out as well that the predictive nature of the ICC’s findings does not, by itself, alter our basic standard of review, which focuses on the reasoned nature of the Commission’s decision, taking into account the nature of that decision and of the components that can be reasonably expected to go into it. Thus, when the facts relied upon by the Commission are insufficient, by themselves, to support its ultimate conclusion with certainty, it must identify the uncertainties, see ILG-WU, 722 F.2d at 814 n. 33, explain why it acted prior to “engaging in a search for further evidence,” Airbags, 103 S.Ct. at 2871, and state what considerations led it to resolve the uncertainties as it did. See Small Refiner Lead Phase-Down Task Force v. Environmental Protection Agency, 705 F.2d 506, 520 (D.C.Cir.1983). 2. “General” market constraints Petitioners contend that the Commission had no basis in the record for its conclusions regarding general market constraints such as alternative forms of transportation, leverage of large shippers, and product and geographic competition. For each constraint they cite record evidence showing substantial numbers of shippers who do not enjoy the benefits of the constraints. See Joint Brief for Petitioners at 67-72. They further attack the Commission’s economic analysis, disagreeing with its conclusion that the existence of product and geographic competition, without more, constrains market abuses against shippers. But it seems to us petitioners misapprehend the significance the Commission allotted to each of these factors. The Commission recognized that the factors enumerated above were only general constraints in the transportation market and that the circumstances of some individual shippers were such that they might not be constrained by any or all such factors. See Boxcars I, 367 I.C.C. at 433-34. While the Commission did not always cite specific record evidence for its conclusions about market constraints, our perusal of the record convinces us that it did in fact rely on uncontroversial facts that were within its expert knowledge and which petitioners do not contest. See, e.g., Boxcars I, 367 I.C.C. at 433 (most items that can be loaded in a boxcar can be loaded in a truck; motor carriage tends to be faster, more accessible, and more convenient; alternate rail routes and TOFC/COFC service give many shippers the benefit of intramodal rail competition). Furthermore, while we find merit in the petitioners’ argument that absent evidence about cross-elasticities of particular commodities with substitute products or commodities from other geographical locales, existence of product and geographical competition says little about the constraints such competition places on carriers vis-a-vis shippers, we do not find the Commission relied on product or geographical competition for more than its limited worth. See Boxcars I, 367 I.C.C. at 434. In sum, we believe that the Commission recognized full well the generalized nature of some of its predicted constraints on market abuse. While reliance on any one of these factors alone might not justify a finding that railroads could not abuse their market power, it was not unreasonable for the Commission to conclude that together the factors assured this. The Commission further bolstered these contraints with evidence that, in fact, market abuse was rare, if not non-existent. The ICC noted that “the present record identifies no commodity moving chiefly by boxcar on which rate levels are now being controlled by Commission maximum rate prescriptions rather than by market forces.” Boxcars I, 367 I.C.C. at 440; see also Brief for Respondent at 54-55 & n. 29 (claiming that there is only one case since the passage of the 4-R Act in 1976 in which the Commission found either market dominance or charging of an unreasonable rate by a railroad for boxcar transportation of freight). It also retained jurisdiction to require reciprocal switching and joint use of terminals allowing one carrier to use another carrier’s routes and other facilities, see Boxcars I, 367 I.C.C. at 433 n. 20, and promised to modify or revoke the exemption of maximum rate regulation if “the exemption does not work as planned.” Boxcars III, slip op. at 4. Finally, the Commission considered the Conrail and DOT studies, which it found further supported a conclusion that railroads did not have market dominance. 3. The Conrail study This brings us to petitioners’ major attack on the Commission decision, which focuses on the study of railroad/truck market shares and r/vc submitted by Conrail. Petitioners argue that this study, upon which the ICC heavily relied, does not rationally support the Commission’s “finding that railroads could not abuse their market power with respect to ‘any’ commodity transported by boxcar” because the data it reports are too aggregated to justify such a conclusion about particular commodities. Joint Brief for Petitioners at 61-62. Petitioners point to differences in the types of boxcar services required for different commodities, and contend that each of the commodity groups in the study represented several, rather than a single, market for transportation. For example, commodities like certain automobile parts require boxcars with special racks that make these cars unsuitable for transportation of other goods. According to petitioners, these special parts comprise a distinct transportation market that the ICC should have looked at separately. See Boxcars I, 367 I.C.C. at 436. Petitioners claim that railroads enjoy high market shares and r/vc in many of these submarkets. Because the data from these submarkets are averaged with data from submarkets where railroads have very little market share and low r/vc, they argue, the aggregated data of the Conrail study masks significant potential for railroads to abuse market power. Petitioners cite Chesapeake and Ohio Railway v. United States, 704 F.2d 373 (7th Cir.1983). In that case the Seventh Circuit rejected the ICC’s reliance on aggregated data about through routes as arbitrary and capricious. In Chesapeake the ICC had granted Conrail’s petition to cancel several thousand joint rates under 49 U.S.C. § 10705(e) based on a study Conrail submitted showing that cancellation of all routes, in the aggregate, would result in an average reduction of transit time for the routes remaining open. Chesapeake held that the aggregated data was not sufficiently probative of the efficiency of individual routes for the court to conclude that substantial evidence supported the cancellations as being in the “public interest.” Id. at 379. The statutory scheme of section 10505(a) nonetheless leads us to conclude that the ICC’s reliance on the Conrail study was justified. In the Staggers Act, Congress amended this provision explicitly to authorize Commission exemptions from regulations that are not limited in scope. See Conference Report, supra at 104-05, 1980 U.S.Code Cong. & Ad.News at 4137. It explained this change in the Staggers Act conference report: The conferees expect that, consistent with the policies of this Act [the Staggers Act], the Commission will pursue partial and complete exemptions from remaining regulation. The conferees anticipate that through the exemption process the Commission will eventually reduce its exercise of authority to instances where regulation is necessary to protect against abuses of market power where other federal remedies are inadequate for this purpose. Particularly the conferees expect that as many as possible of the Commission’s restrictions on changes in prices and services by rail carriers will be removed and that the Commission will adopt a policy of reviewing carrier actions after the fact to correct abuses of market power. Conference Report, supra at 105, 1980 U.S.Code Cong. & Ad.News at 4137. Thus, Congress encouraged the Commission to apply its exemption authority under section 10505(a) in a manner of “general applicability,” and therefore “it is not a condition of [the] validity [of such an exemption] that there be adduced evidence of its appropriateness in respect to every railroad to which it will be applicable.” United States v. Allegheny-Ludlum Steel, 406 U.S. 742, 749, 92 S.Ct. 1941, 1947, 32 L.Ed.2d 453 (1972) (quoting Assigned Car Cases, 274 U.S. 564, 583, 47 S.Ct. 727, 734, 71 L.Ed. 1204 (1927)). Of course the evidence that the Commission does consider must be probative of the findings it is required to make as part of its exemption decision and the Commission must explain why more probative evidence, if available, was not collected. Here the ICC claimed that a commodity-by-commodity approach would involve thousands of commodities and was not feasible. See Boxcars I, 367 I.C.C. at 436. At argument counsel for petitioners stated that it did not expect the Commission to do a commodity-by-commodity study, but that the study should have grouped commodities according to their transportation characteristics. We do not believe the Commission’s commodity groups were inappropriate even when viewed, as petitioners suggest, in terms of transportation characteristics. Although aggregated data is imperfect, it is reasonable to assume that there is some correlation between types of commodities and transportation-related attributes of the goods. For example, we expect that ears that carry different types of auto parts are more likely to be interchangeable with each other than either is with cars that carry grain products, and the routes that auto parts travel are more likely to overlap with each other than either is with the routes that grain products travel. Thus, the Commission did not act arbitrarily and capriciously in concluding that the Conrail study “divided boxcar commodities into reasonably related groups under a standard classification system.” Boxcars I, 367 I.C.C. at 436. Nor do we believe that Chesapeake, 704 F.2d at 379, mandates a reversal of the ICC decision here. The Chesapeake court reviewed a decision to cancel particular (albeit numerous) through routes under 49 U.S.C. § 10705(e). That section envisions a more circumscribed ICC decision based on an evidentiary hearing focusing on the particular through routes of particular parties. Information therefore that may be of little probative value with respect to a particular route, may be highly probative for the purposes of a policy decision of general applicability, such as the one Congress envisioned, and the ICC here effectuated under section 10505(a). Additionally Chesapeake noted that even the aggregate data submitted in that case was probably meaningful enough for Conrail to meet its initial burden of production under section 10705(e). The court did not so find only because the aggregated data “was built up from individual through route comparisons [and] Conrail easily could have presented the data underlying those comparisons along with the summary statistics^] ... the burden of production would not have been greater____” Chesapeake, 704 F.2d at 379. In Chesapeake, the Commission could not give any reason why the full study was not included in the administrative record. In this case the Commission concluded that presenting meaningful less-aggregated data would impose an impossible burden time and energy-wise, see Boxcars I, 367 I.C.C. at 436, and petitioners point to no record evidence undermining this conclusion. We thus conclude that ICC reliance on the aggregate data of the Conrail study was not per se arbitrary and capricious. Petitioners assail the Conrail study because it is limited to rail transportation to, from, or within the Northeast. They argue that the attributes of freight transportation markets in other regions of the country are sufficiently different from those in the Northeast to render the ICC’s conclusions about national rail transportation based on this data arbitrary and capricious. The ICC, however, did not consider the Conrail data in isolation; it relied on many other factors, most importantly a Department of Transportation (DOT) nationwide study done in 1977. In addition one of the petitioners itself submitted comments indicating that Conrail’s data was representative of the whole country. See J.A. 1418-21. The Commission thus had before it nationwide data as well as that from the Northeast and indications that the Northeast data was representative of the whole country. It therefore had a basis for informed judgments about boxcar freight transportation markets throughout the country. 4. Particular commodities The conclusion that the ICC did not err in relying on the Conrail data does not, however, completely put to rest the challenge to its finding that rate regulation is not needed to prevent abuse of market power, since the Commission must consider all relevant factors in the comments it receives. Petitioners submitted evidence demonstrating that four commodity groups — auto parts, paper and forest products, grain, and metals — included submarkets for transportation within which railroads had high market shares and in one case were earning sufficiently high r/vc to subject them to regulation. For these commodity groups the Conrail study would not be enough to support the Commission’s conclusion on market abuse since its aggregated data cannot refute this more specific evidence submitted by petitioners. In every one of these potentially problematic groups, however, the ICC considered record evidence independent of the Conrail study and concluded that railroads had no market dominance or otherwise were not in a position to abuse market power within any commodity group. The petitioners argue that the Commission ignored evidence they submitted tending to show market dominance in certain submarkets. See Joint Brief for Petitioners at 75-82. While we agree that there is some evidence in the record tending to support their contentions about market power as to certain commodities, we do not find it, on review, compelling enough to render the Commission decision arbitrary and capricious. It is not our function to second-guess the ICC on factual matters. See Allegheny-Ludlum Steel, 406 U.S. at 749, 92 S.Ct. at 1946 (reviewing court is not to weigh the evidence before the Commission). The ICC, in its decision, drew on evidence of such factors as transportation alternatives, trends in the transportation of these commodities, shipper market power, etc., from which it was reasonable to conclude that railroads either do not dominate these submarkets or could not abuse any market power they did have. See Boxcars I, 367 I.C.C. at 436-40. Thus, despite the lack of the Conrail study’s probativeness for the four potentially problematic groups, the ICC had ample record evidence supporting its conclusions as to rail carriers’ market power over transportation of commodities within these groups. 5. Conclusion as to rail carriers’ market power In sum, we conclude that the ICC did not act arbitrarily and capriciously in finding that rail carriers did not have monopoly power in the market for transport of commodities shipped by boxcars. We are relying as well on the Commission’s position that its authority to revoke the exemption is an appropriate mechanism to correct any post-exemption market abuse. See Boxcars I, 367 I.C.C. at 440-41. This court has previously endorsed such an approach in approving promulgation of general rules, see, e.g., The Process Gas Consumers Group v. United States Depart- merit of Agriculture, 694 F.2d 728, 745-46 (D.C.Cir.1981); adopted en banc, 694 F.2d 778, 783 n. 3 (D.C.Cir.1982), cert. denied, — U.S. -, 103 S.Ct. 1874, 76 L.Ed.2d 807 (1983), and we believe it especially appropriate here for three reasons. First, the ICC considered substantial evidence about particular commodities as well as commodity groups without finding any monopoly power by railroads. Second, Congress itself envisioned after the fact review to correct isolated market abuses that may follow the lifting of protective regulations under section 10505(a). See Conference Report, supra at 105, 1980 U.S.Code Cong. & Ad.News at 4137 (“conferees expect ... that the Commission will adopt a policy of reviewing carrier actions after the fact to correct abuses of market power”). Finally, counsel for the Commission assured us at argument that the ICC would seriously consider revocation of the rate exemption and investigation of the boxcar freight rates charged any shipper upon a demonstration that the shipper had no meaningful transportation alternatives and r/vc for the rail carrier was sufficiently high to subject the rate to the Commission’s jurisdiction. In the absence of a showing by petitioners that the railroads enjoy such market power over the transport of any specific commodity, and assured by the Commission that it will likely revoke the exemption with respect to any such commodities later identified, we refrain at this juncture from disturbing the exemption of boxcar traffic from maximum rate regulation. C. The Scope of the Maximum Rate Exemption Petitioners argue that because this exemption was so much broader than those previously adopted by the ICC, the Com- . mission must therefore explain why it chose to proceed in the sweeping manner it did. See Joint Brief for Petitioners at 82-83. As previously stated, we believe that where Congress itself finds that generally continued regulation is unnecessary and anticipates deregulation of the entire railroad industry to the maximum extent possible in conformity with the national rail transportation policy, the Commission need not make an exquisitely detailed showing of why the broad scope of its exemption was proper. In this case the Cotnmission noted that “the premise of both the Railroad Revitalization and Regulatory Reform Act of 1976 (4R Act) [Pub.L. 94-210, 90 Stat. 31 (1976) ] and the Staggers Rail Act of 1980, [Pub.L. 96-448, 94 Stat. 1895 (1980) ] was that trucking competition had rendered continued railroad regulation, to a large degree, both undesirable and unnecessary.” Boxcars I, 367 I.C.C. at 427. It relied on this premise and the fact that “the typical boxcar commodity is ... the same type of commodity that is typically transported in motor carrier van trailers.” Id. at 433. It was quite reasonable then, for the Commission to conclude as it did that a commodity-by-commodity approach would be unnecessarily burdensome. One petitioner, the Port of Oakland, however, attacks the ICC’s authority to exempt all boxcar traffic, claiming that this exemption covers more than “a person, class of persons or a transaction or service,” the terms Congress used in section 10505(a). See Supplemental Brief of Petitioner Board of Port Commissioners at 8-9. The Port attempts to set up a distinction between a “service,” which the ICC can exempt from regulation, and a “transportation,” which the Port argues includes non-service activity that the ICC cannot exempt. The Port points to the Interstate Commerce Act’s (ICA) definition of “transportation” to support this distinction. The ICA defines “transportation” [to] include— (A) ... equipment of any kind related to the movement of passengers or property ... (B) services related to that movement, including receipt, delivery, elevation, transfer in transit, refrigeration, icing, ventilation, storage, handling, and interchange of passengers and property. 49 U.S.C. § 10102(25). The Port argues that the blanket exemption of boxcar transit is not an exemption of a “service” because, under that definition, a service involves activity ancillary to the movement of freight, but does not include the actual movement itself. Because the boxcar exemption involves rates for movement of freight, «the Port says it is an exemption of “transportation,” which it sees as a much broader classification than the “service” alluded to in section 10505(a). We find the Port of Oakland’s position also to be without merit. At the outset we note that the ICA does not separately define “service.” While the ICA definition of “transportation” is broken down into “equipment” related to the movement of freight and “services” related to that movement, it is actually silent as to whether the movement of freight is itself a “service.” On the other hand, the legislative history makes it quite clear that Congress envisioned exemptions of movements, like that here, under section 10505(a). The House report on the Staggers Act explicitly approved the ICC’s efforts under the narrower pre-Staggers Act exemption provision to exempt “the carriage of fresh fruits and vegetables ... and ... trailer-on-flat-car traffic” from regulation. H.R.Rep. No. 1035, 96th Cong., 2d Sess. 60, reprinted in 1980 U.S.Code Cong. & Ad.News 3978, 4005. Neither fruit and vegetable transport nor TOFC traffic fit into the Port’s narrow interpretation of the term “service.” We are supported in our reading of section 10505(a) as authorizing exemption of boxcar carriage by American Trucking Associations, 656 F.2d at 1120-21, which affirmed the ICC’s exemption of TOFC carriage from all regulation under section 10505(a). The Port argues that the TOFC exemption was within the ICC’s authority only because Congress specifically authorized exemptions of intermodal transportation in 49 U.S.C. § 10505(f). But American Trucking Associations viewed the special treatment of intermodal transportation in section 10505(f) as a clarification that such exemption of such transportation is authorized by the general provision of section 10505(a), and it specifically relied on the latter section as “the source of the Commission’s authority to grant [the TOFC] exemption____” Id. at 1120. Given the ambiguity of the language of section 10505(a) in the face of its clear legislative history, and given further the persuasive view of the Fifth Circuit in American Trucking Associations, we hold that the Commission acted within its lawful authority in exempting freight boxcar rates from regulation. III. Joint Rates and Through Routes In exempting boxcar transit from rate regulation, the Commission also exempted railroads from the statutory provisions dealing with “joint rates” charged for boxcar service over “through routes.” See 49 U.S.C. §§ 10705, 10705a, 10707. A through route is one in which two or more rail carriers participate with each carrier transporting a shipment over part of the route. A joint rate for that through route is a single rate charged by the carriers that together carry a shipment over the route. “In contrast to ‘combination rate’ service, in which each carrier collects its charges separately, in ‘joint rate’ service, the delivering carrier bills and collects for all participating carriers and payments are divided among the participants according to a ‘division’ formula [that the participants agree upon prior to setting the joint rate].” Ford Motor Co. v. ICC, 714 F.2d 1157, 1159 (D.C.Cir.1983). Prior to the ICC exemption, two statutory provisions governed modification and cancellation of joint rates for boxcar transit. See 49 U.S.C. §§ 10705, 10705a. If a carrier elects to modify or cancel a joint rate under section 10705(e), it has to file its new tariff or cancellation 10 or 20 days prior to the effective date of the change, see 49 U.S.C. § 10762(c)(3), and the Commission reviews the filed tariff or cancellation to ensure it complies with the ICA (as amended). See 49 U.S.C. § 10707. As this court previously noted, a carrier proceeding under section 10705(e) “takes the chance that the ICC might suspend [the tariff] under section 10707.” Southern Railway Co. v. ICC, 681 F.2d 29, 32 (D.C.Cir.1982). If the ICC does suspend the tariff, the filing carrier has to show, in a full evidentiary proceeding, that the change or cancellation of the joint rate is in the public interest. If the carrier elects to proceed under section 10705a, upon 45 days notice, see 49 U.S.C. § 10705(a), (f)(2), it can add a surcharge to or cancel a joint rate if its revenue from the joint rate is less than 110% of the variable cost of providing service over its portion of the route. See 49 U.S.C. § 10705a(a)(l)(A). The carrier, however, has no right under section 10705a to impose a surcharge in such a manner that it receives more than the 110% threshold figure, or even to cancel a rate if its co-participant counters with a revenue division that ensures the cancelling carrier a 110% return. If it tries to do so, the Commission either sua sponte or upon protest by other participating carriers may impose a new joint rate that grants the surcharging or cancelling carrier his 110% return. See 49 U.S.C. §§ 10705a(a)(2), 10705a(c)(2), (3), (6). In addition, section 10705a provides special protections for class III railroads that include the right to challenge a surcharge or cancellation as adversely affecting competition. See 49 U.S.C. § 10705a(i)(l). If the Commission determines that the public interest requires a return below the 110% threshold to avoid anticompetitive action and to maintain service on the route, it may order the carrier to provide such unprofitable service. See 49 U.S.C. § 10705a(i)(2). Since the exemption went into effect, carriers have been free to cancel or modify joint rates for boxcar transport as they see fit. They thus have been relieved of any obligation to provide prior notice of changes and to file changes with the ICC. In addition, to the extent sections 10705(e) and 10705a give other participating carriers a right to challenge changes in joint rates, the exemption extinguishes such rights. In short, while sections 10705(e) and 10705a give carriers “no unfettered right of immediate joint rate cancellation [or modification],” Southern Railway, 681 F.2d at 34 n. 11, the exemption gives them precisely such a right, A. The Commission Decision The ICC briefly discussed the exemption of joint rates from regulation in its initial decision. It responded to comments that joint rate exemption would allow elimination of joint rates and closing of through routes in derogation of the national rail transportation policy “to ensure the development and continuation of a sound rail transportation system with effective competition among rail carriers ...,” 49 U.S.C. § 10101a(4), and “to foster sound economic conditions in transportation ...,” 49 U.S.C. § I0101a(5). The Commission first noted, “[t]here is no reason to expect that an exemption would lead to cancellation of all joint boxcar rates.” Boxcars I, 367 I.C.C. at 444. It reasoned that a carrier would have no incentive to “insist on a division of profits so disproportionate as to force the other [carriers] to withdraw from the movement ... [as that would] forfeit the traffic for [all participating] carriers.” Id. In the next paragraph, however, it essentially admitted that such incentives exist since “route closings might occur if a carrier wishes to close routes that involve long hauls over the lines of connecting railroads in order to concentrate traffic on routes that involve long hauls over its own lines.” Id. In essence, the Commission recognized that where a carrier’s own route competed with the through route it might benefit from closing the through route since that would reduce intramodal competition. The Commission dismissed this possibility by a single cryptic remark that “[nonetheless, it is unlikely that a carrier acting in an economically rational manner would close efficient routings.” Id. In its decision denying reconsideration of the boxcar exemption, the ICC again addressed the exemption of joint rates from regulation. See Boxcars II, 367 I.C.C. at 753-54. There the Commission focused on the concerns of class III carriers that the exemption was inconsistent with the protections for small railroads contained in 49 U.S.C. § 10705a, which allows carriers under limited circumstances to add surcharges to or cancel joint rates. The Commission first stated that its joint rate exemption did not open small railroads to abuses against which 49 U.S.C. § 10705a protected. It noted that section 10705a was merely a shortcut to allow railroads that can support certain claims of nonprofitability to quickly impose a surcharge on or cancel particularly unprofitable joint rates. The ICA (as amended) provides an alternative ground for a carrier to cancel joint rate that the carrier can show is not in the public interest. See 49 U.S.C. § 10705(e). The ICC pointed out that under this provision, a carrier could cancel almost any inefficient or unprofitable rate, and reasoned that the exemption of joint rate regulation merely allowed such cancellations to occur without administrative delay and expense. Boxcars II, 367 I.C.C. at 753 & n. 23. The Commission also noted that the specific provisions of section 10705a do not prevent the exemption of joint rates from regulation. Id. at 753. The whole point of the exemption provision was to eliminate the requirements of such provisions once the ICC shows they neither further rail transportation policy nor protect shippers from market abuse. The Commission went on to explain why it believed the joint rate exemption would not harm class III carriers. It stated: Boxcar traffic is especially subject to diversion to other modes and consequently will not bear excessive rate increases. What matters to the shipper is the origin-to-destination rate, not the size of the connecting railroad of origin or destination. In other words, after excessive rate increases on boxcar traffic, diversion would take place regardless of whether the traffic moved jointly with a class III railroad or moved solely on a class I [i.e., large] railroad. This gives long haul class I railroads an incentive to hold down rates that they would charge for transportation to junctions with class III carriers in the absence of joint (single factor) rates for the origin to destination movement. Boxcars II, 367 I.C.C. at 753-54. In essence, the Commission noted that intermodal competition puts some cap on the rate a large carrier will set for its portion of transit in which a small carrier also participates. The Commission further reasoned that the exemption of boxcar traffic generally would attract boxcar traffic from shippers, and the benefits of this increased traffic would inure to small and large carriers alike. Id. at 754. In Boxcars II, the ICC also repeated its assertion that large carriers would cancel only inefficient routes, this time explaining that “[c]arriers cannot afford the luxury of preserving inefficient routes when competitive pressures force them to reduce costs as much as possible.” Id. at 754. The ICC was satisfied that the exemption was consistent with the rail transportation policy because it would result in no cancellations of efficient through routes. The ICC revisited the joint rate exemption one final time in Boxcars III. There it briefly repeated, without elaboration, its determinations that carriers will not cancel efficient joint routes, and that therefore small carriers will not suffer from the joint rate exemption. Boxcars III, slip op. at 6. B. Carrying Out the Rail Transportation Policy Petitioners contend that before the ICC concludes that a particular provision of the ICA is “not necessary” to carry out the rail transportation policy at 49 U.S.C. § 10101a, it must consider why Congress adopted the provision in the first place. See Joint Brief of Petitioners at 37. They claim that the Commission did not adequately consider the balance Congress struck in section 10705a, “a key objective of which was to protect small railroads against anticompetitive actions by large carriers.” Id. at 38. At least in this case we believe that the Commission must consider the relationship between section 10705a and the national rail transportation policy as well as the exemption provision invoked here. Such consideration is mandated by the Staggers Act itself which requires the ICC to consider whether regulation is needed to further the transportation policy set out in the Act. Additionally, the legislative history shows that section 10705a, adopted at approximately the same time and as part of the same Act which contains both the transportation policy in section 10101a and the exemption authorization in section 10505a, was specifically designed to address one of the components of the rail transportation policy set out in section 10101a, i.e., “the development ... of a sound rail transportation system____” The Commission counters that it has indeed considered the interests of small railroads and concluded that large carriers will not divert traffic from efficient through routes in which small carriers participate. It further asserts that this is all the rail transportation policy requires. We find, however, that the legislative history of these three provisions clearly demonstrates that Congress anticipated that the ICC would engage in a far broader and more thorough inquiry into the need for continued joint rate regulation before granting a total and unconditioned exemption of joint rates from any oversight or regulation. 1. Incentives for large carriers to close efficient routes The first problem with the ICC joint rate exemption is the Commission’s failure to adequately explain its crucial assertion that large long-haul carriers will not close off efficient routes of small short-haul carriers. After admitting that large carriers have an incentive to close off efficient competing routes to gain monopoly power over a haul from a given origin to a given destination, it simply dismissed this possibility as unlikely. 367 I.C.C. at 444. Such bald assertions do not qualify as reasoned decisionmaking. The government’s brief, attempting to fill this reasoning gap, explained that in such a monopolization situation where “a carrier seeks to foreclose a smaller carrier from markets it [the small carrier] could serve as part of an efficient route, antitrust remedies are available.” Brief for Respondent at 63. While the ICC itself mentioned antitrust laws as a possible remedy for predatory pricing by large railroads against small railroads generally, see Boxcars I, 367 I.C.C. at 442, it never mentioned antitrust remedies either as a deterrent to potential monopolizing stemming from the relationship created by co-participation in joint rates or through routes or as an after-the-fact remedy that would correct such abuses by large carriers, see Boxcars I, 367 I.C.C. at 443-44. We therefore must reject the argument in the government’s brief as a post hoc rationalization of counsel and not the Commission’s reason. See Airbags, 103 S.Ct. at 2780. Even if we attended counsel’s argument, however, our perusal of the legislative history of the Staggers Act creates doubt that merely noting the existence of the antitrust laws sufficiently satisfies the Commission’s statutory burden of making a finding that large carriers will not try to drive class III railroads from providing service over efficient routes. Under pre-Staggers Act law, the Commission had authority to investigate all joint rate changes and cancellations in order to ensure they were in the public interest. See Southern Railway, 681 F.2d at 33-34 & n. 11. Presumably, cancellations that significantly reduced competition over a particular haul would not be approved. When Congress enacted section 10705a, which took away the Commission’s authority to review certain joint rate cancellations, it was careful to preserve a mechanism by which class III railroads could challenge the cancellation before the Commission as anticompetitive. It did so “to avoid the need and burden of complex litigation that might arise out of the need for class III carriers to resort to the antitrust laws, although an available alternative, if they feel a carrier is engaging in anticompetitive activities.” H.R. Rep. No. 1035, 96th Cong., 2d Sess. 64, reprinted in 1980 U.S.Code Cong. & Ad. News 3978, 4009. It is thus clear that Congress itself rejected the mere existence of antitrust remedies as a justification for removing the Commission’s authority to review joint rate cancellations. We do not mean to suggest that the antitrust remedies are therefore irrelevant to the Commission’s inquiry into the need for joint rate regulation. If the Commission had, for example, shown that the types of monopolistic practices which the joint rate exemption might permit are in fact antitrust violations not likely to result in suits involving difficult questions of proof and long drawn out trials, antitrust remedies might negate the need for continued regulation. But neither the ICC nor its counsel made any such showing which would convince us that it considered factors beyond those