Opinion for the Court filed by Circuit Judge D.H. GINSBURG.
D.H. GINSBURG, Circuit Judge:
Under the Staggers Rail Act of 1980, Pub.L. No. 96-448, 94 Stat. 1895 (codified at various sections of 11, 45 & 49 U.S.C.), a railroad rate set at or below an inflation-adjusted base level is presumed to be reasonable. In 1989 the Interstate Commerce Commission decided, over the objection of Consolidated Rail Corporation (Conrail) and the Association of American Railroads, in calculating the inflation-adjusted base level, to include a deflator for railroad productivity gains. Over the objection of a group of railroad customers (the Shippers), the ICC also decided to implement the productivity adjustment prospectively only. Both the Railroads and the Shippers now seek review of the Commission’s decision. We conclude that the Commission acted within its statutory authority and reasonably in all respects.
I. Background
The Staggers Act allows a railroad to set or change a rate without advance approval from the Interstate Commerce Commission. See 49 U.S.C. § 10701a(a). In general ei[223]*223ther a shipper or the ICC may challenge the reasonableness of the new rate (subject to limitations not relevant here). See id. §§ 10701a(b), 10707(a), 10709(d), 11701. The Act places certain limits, however, upon the ability both of a shipper and of the ICC to challenge a rate increase, insofar as it merely keeps pace with inflation. To that end, § 229 of the Act provides that a rate in effect on October 1,1980 that was not challenged by March 30, 1981 became a “base rate” that could not later be assailed as unreasonable. Id. § 10701a note. Section 203 of the Act further mandates that such an unchallengeable base rate be adjusted quarterly for inflation. A rate set at or below the “adjusted base rate” is likewise immune from challenge. Id. § 10707a. When a railroad files a rate above the adjusted base rate and it is challenged, the carrier bears the burden of proving that the rate is reasonable. In practice, however, filed rates have generally been below adjusted base rates. See Railroad Cost Recovery Procedures—Productivity Adjustment, 5 I.C.C.2d 434, 436 (1989) (hereinafter Productivity Adjustment ).
The manner in which the ICC calculates the multiplier by which it adjusts base rates for inflation is the primary subject of this case. The Staggers Act requires that the Commission use an index of railroad costs, making “appropriate adjustments to reflect the changing composition of [such] costs, including the quality and mix of material and labor.” 49 U.S.C. § 10707a(a)(2)(B). In 1981 the Commission began publishing quarterly a Rail Cost Adjustment Factor which, when multiplied by a base rate, creates the challenge-free adjusted base rate that just keeps up with inflation.
Before the Commission began publishing the RCAF, the Shippers had vigorously urged that it should take into account any increase in the railroads’ productivity. A gain in productivity occurs when a railroad is able to produce a rail service (output) from a smaller quantity of materials and labor (inputs) than previously. (Productivity may be increased by a managerial innovation, the use of new physical capital, workforce training, the relaxation of restrictive work rules, or otherwise.) Citing (1) “grave” methodological problems, (2) the disincentive to increase productivity that might result, and (3) inadequate revenues throughout the rail industry, the Commission declined to incorporate a productivity adjustment into its index. As a consequence, the original RCAF measured only changes in the cost of railroad inputs, as approximated by a fixed market basket of inputs commonly used by railroads.
In Western Coal Traffic League v. United States, 677 F.2d 915 (D.C.Cir.1982), a group of shippers challenged the Commission’s decision, claiming that the Staggers Act requires the ICC to use a RCAF that reflects changes in railroad productivity, or alternatively, that the ICC acted arbitrarily and capriciously in excluding productivity changes from its calculation of the RCAF. After reviewing the text and the legislative history of the Staggers Act, we concluded that the Act does not require the Commission to include productivity gains in the RCAF. Id. at 925-26. We further stated that
the difficulties of accurately measuring productivity growth on a current basis, coupled with the Commission’s assessment of the importance of assuring railroads unencumbered rate increases to cover increased costs, support the Commission’s conclusion that a productivity adjustment is inappropriate under the present circumstances.
Id. at 931. We nonetheless urged the ICC to continue refining its RCAF methodology, “including consideration of measurable productivity gains.” Id.
Shortly after we decided Western Coal, the Commission issued an Advance Notice of Proposed Rulemaking seeking comments on the possibility and desirability of including productivity gains in the RCAF. See 47 Fed.Reg. 32,176 (1982). In response, several shipper groups proposed that the ICC use the so-called “Caves-Christensen” methodology, which seemed to address satisfactorily many of the Commission’s previous concerns. The Commission then published a Notice of Proposed Rulemaking [224]*224requesting further comments on the Caves-Christensen approach. See 49 Fed. Reg. 38,319 (1984).
In 1985, as the Commission considered whether to adopt a productivity adjustment, the Congress tasked the Railroad Accounting Principles Board to “establish ... principles governing the determination of economically accurate railroad costs.” See 49 U.S.C. § 11162(a). The RAPB concluded that including a productivity adjustment in the RCAF was both feasible and desireable. The ICC then hired a consultant to evaluate the various proposals and the consultant determined that a refined Caves-Christensen methodology could serve as a workable measure of railroad productivity gains.
In November 1988 the Commission proposed to adjust the RCAF for productivity. See 53 Fed.Reg. 17,558. Following a period for comment and oral argument, the Commission adopted a modified version of the Caves-Christensen productivity measure in March 1989. See Productivity Adjustment, 5 I.C.C.2d at 436. In the second quarter of 1989, the Commission began adjusting the RCAF for changes in productivity, lagged by two years.
The Commission gave three reasons for adopting a lagged productivity adjustment. First, the modified Caves-Christensen methodology largely satisfied its longstanding concern about accuracy. Second, the “rail industry had made substantial improvement from what had been a tenuous financial position and ..., importantly, this improvement had not come from the widespread use of rates at or near the maximum levels permitted by the RCAF.” Id. Finally, the Commission concluded that lagging the adjustment for productivity would avoid unduly dampening the railroads’ incentive to achieve additional productivity gains. Id. at 444-45.
II. The 1989 Productivity Adjustment
Petitioner Conrail urges that the ICC does not have statutory authority to , include a productivity adjustment in the RCAF. According to Conrail, in § 203 of the Staggers Act the “Congress struck a deliberate regulatory bargain, whereby shippers were to be protected from excessive rates in exchange for the railroads’ right to retain all productivity gains.” Alternatively, the Railroads (meaning the AAR and Conrail jointly) argue that the Commission’s decision is not supported by the administrative record. For their part, the Shippers claim that the ICC should have adjusted the 1989 RCAF downward to take into account productivity gains achieved from 1982 to 1989.
A. Does the Staggers Act Prohibit a Productivity Adjustment?
Applying the teaching of the Supreme Court in Chevron USA, Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837, 104 S.Ct. 2778, 81 L.Ed.2d 694 (1984), we first consider whether, as Conrail claims, the Congress specifically prohibited the ICC from including a productivity adjustment in the RCAF. In Western Coal we considered the cognate question whether the Congress specifically required the ICC to include a productivity adjustment in the RCAF. We did find indications in the legislative history that the Congress had contemplated a productivity adjustment. Ultimately, however, we reported that “while ... the purpose of section 203 was to allow railroads to keep up with cost increases, we [could not] conclude that this purpose mandates adoption of [a productivity adjustment] regardless of the accuracy of such an index or the feasibility of [it].” Id. at 926 (emphasis deleted). We urged the Commission, however, to “revise the index in an appropriate fashion, including consideration of measurable productivity gains, as the circumstances warrant.” Id. at 931.
Conrail argues that Western Coal does not and should not control this case. (“Any language in [that decision] respecting the ICG’s discretion to adopt a productivity adjustment was dictum and should in any event be overruled as manifestly incorrect.”) To counter the analysis in Western Coal, Conrail points out that when a productivity adjustment is included in the RCAF, a prolonged period of gains in pro[225]*225ductivity could bring adjusted base rates down below the base rates deemed lawful by § 229 of the statute.
We are not persuaded by Conrail’s argument. One might equally well note that during a prolonged period of deflation the RCAF could erode the base rates even if no productivity adjustment were included in its calculation. Neither of these theoretical but unlikely possibilities significantly advances our understanding of the Congress’ actual intent as expressed in the Staggers Act.
We think it more to the point that in Western Coal, after delving deeply into the Staggers Act and its legislative history in order to determine whether the Congress intended to require the ICC to adjust the RCAF for productivity gains, we concluded that the “plain meaning” of § 203 neither mandates nor prohibits a productivity adjustment. 677 F.2d at 922-23. Likewise, we reported that neither the legislative history nor the structure of the Act compels the use of any particular methodology for calculating the RCAF. Id. at 923-26.
Conrail claims that these conclusions are dicta because we undertook the inquiry in Western Coal not to determine whether the Act prohibits, but only whether it requires, a productivity adjustment; we did not need to resolve the question of prohibition vel non in order to decide that case. In the course of unearthing the intent of the Congress relative to the requirement question, however, we necessarily brought to light the answer to the prohibition question at the same time. In any event, putting aside the matter of whether we are hound by Western Coal as the law of this Circuit, we are persuaded for the reasons given in that opinion that the ICC is authorized by statute to include a productivity adjustment in the RCAF. See id.
B. Did the ICC Act Arbitrarily in Adopting the Productivity Adjustment?
In the alternative the Railroads argue that the Commission acted arbitrarily by adopting the productivity adjustment without adequate record support for its (assertedly) implicit premise that present rates are unreasonably high. “In the absence of evidence that a productivity adjustment is needed to protect shippers,” say the carriers, “even slight injury to the railroads tilts the balance against an adjustment;” and not surprisingly, the carriers argue that a productivity adjustment does harm some railroads.
Stripped of embellishment, this argument reduces to the assertion that the Railroads need higher rates more than the Shippers need rate relief. Whether that is a true, a false, or an incoherent statement, the argument proceeds from an erroneous view of the nature and purpose of the RCAF. In establishing the system of challenge-free, inflation-adjusted base rates, the Staggers Act neither prescribes nor caps railroad rates. The scheme simply enables a railroad to increase its rates without having to defend its action any further than to note that it is only keeping pace with inflation. As we stated in Western Coal, the purpose of the § 203 inflationary cost index “is to provide a mechanism for recovering inflation-based costs, not to enhance [real] revenues.” 677 F.2d at 925.
Therefore we reject the Railroads’ suggestion that it was arbitrary for the ICC to adopt an across-the-board productivity adjustment “without any showing of its need.” We think it axiomatic that the ICC has discretion to improve the accuracy of the RCAF when and how it can. Although the productivity adjustment will undoubtedly have some effect upon the Railroads’ profitability, “increasing [or decreasing] net revenues is not a proper reason for refusing to make otherwise appropriate adjustments.” Id.
We turn, therefore, to the Railroads’ more relevant concerns, viz. with the operation of the productivity adjustment. The Railroads complain first that the productivity adjustment penalizes them because competitive pressures already force most of them to pass their productivity gains through to shippers in the form of lower rates. From that premise, the carriers adduce a complicated argument, which we [226]*226quote at length in order to avoid any paraphrastic imprecision:
The railroads’ [double passthrough argument] rested on evidence that (1) railroads do not achieve productivity gains evenly for all services; (2) railroads do not share [with shippers] productivity gains evenly for all services, and may share all, none, or more or less than 100 percent of productivity gains, depending on market conditions; and (3) in the aggregate, at least 100 percent of productivity gains have been transferred to shippers in recent years.
An ICC-imposed productivity adjustment equal to 100 percent of productivity gains requires additional rate reductions (or holddowns of increases) for those services where the rate of productivity growth, or the passthrough of productivity growth, is below the industry average. Such an adjustment thus results in a passthrough equal to the greater of the industry-average rate of productivity or the competitively-induced passthrough in each market. Unless a railroad is able to offset the productivity adjustment with individual rate increases, the overall rate of passthrough resulting from competition and regulation in tandem substantially exceeds ... 100 percent.
(Emphases original.)
The effect of which the Railroads complain is inherent in the use of an industry-wide average. It cannot be avoided unless the ICC abandons altogether the use of averages in calculating productivity adjustments (the methodology it uses also when calculating the inflation adjustment). It is not arbitrary, however, for an industry-wide rate regulatory scheme to use industry-wide average cost data. See Permian Basin Area Rate Cases, 390 U.S. 747, 805-06, 88 S.Ct. 1344, 1379-80, 20 L.Ed.2d 312 (1968) (whether to set natural gas rates based upon an industry-wide average of net production investment is “a question peculiarly within the Commission’s discretion”); see also 1 Alfred E. Kahn, The Economics of Regulation 45-46 & n. 62 (1970) (use of industry-wide average in rate setting provides some incentive for a firm to achieve above-average efficiency). This is especially so in a setting where any railroad disadvantaged by below-average productivity gains for a particular service may file for that service a rate higher than the adjusted base rate. At bottom, the Railroads’ complaint seems to be that the Congress has undertaken to regulate prices in a competitive industry. The Railroads will have to take that case directly to the legislature.
Next the Railroads complain that the ICC acted arbitrarily in concluding that a productivity adjustment will not adversely affect railroad productivity. The Commission did not pretend that the productivity adjustment will have no impact upon productivity. Indeed, the agency devoted a good deal of attention to the problem and specifically, in order to reduce the disincentive effect, inserted a two-year lag time between the Railroads’ realization of a productivity gain and inclusion of that gain in the RCAF. This allows the Railroads to retain for two years (except to the extent that competitive pressures independently dictate otherwise) all of the gains from their improved productivity.
In the end, the ICC was faced with the difficult problem of creating an incentive for a price-regulated industry to make prudent business decisions without at the same time allowing it to charge unreasonably high rates free of challenge. See generally 2 Kahn at 59-63 (describing various regulatory schemes to reward efficiency and innovation). Although the ICC’s solution is perhaps not optimal from the Railroads’ perspective or that of economic efficiency, we cannot conclude, in the statutory context that frames our evaluation, that it is arbitrary.
Finally, the Railroads claim that the methodology the ICC adopted for calculating the productivity adjustment is so inaccurate as to be arbitrary. Specifically, the Railroads contend that because the ICC does not cover a full business cycle when measuring changes in productivity, the adjustment greatly overstates the significance, in the longer run, of the productivity gains made in good times. The ICC did not [227]*227blink at this problem either. On the contrary, in order to address any inequity that may result, the Commission pledged to “continue to lengthen the time period used to calculate the trend as data becomes [sic] available,” and to “include the time frame used to calculate the productivity trend as one of the items to be considered in Ex Parte No. 290 (Sub-No. 7),” the subdocket for comments relating to implementation of the productivity adjustment. See Productivity Adjustment, 5 I.C.C.2d. at 464-65. Indeed, the ICC has recently lengthened to eight years the period used to calculate the productivity adjustment. See Railroad Cost Recovery Procedures—Productivity Adjustment, 8 I.C.C.2d 177, 179 (1991).
Not content with the ICC’s on-going approach to tackling this issue, the Railroads apparently still maintain that the Commission should not have instituted a productivity adjustment until data from a full business cycle are available. We do not think that the Commission acted arbitrarily, however, by using the data that are available now and undertaking to incorporate more data into its analysis as they become available in the future. See Potomac Elec. Power Co. v. ICC, 744 F.2d 185, 191-92 (D.C.Cir.1984). “[T]he best should not be the enemy of the good. It is reasonable to assume that Congress did not intend the infeasible perfect to oust the feasible good.” Commonwealth of Pennsylvania v. ICC, 535 F.2d 91, 96 (D.C.Cir.1976).
C. Did the ICC Act Arbitrarily in Making the Productivity Adjustment Prospective Only?
Not content with the ICC’s decision to take productivity gains into account henceforth, the Shippers argue that the Commission acted without statutory authority, or at least arbitrarily, in refusing “to correct the RCAF to reflect accumulated past productivity gains.” The Shippers would have the ICC reduce the second quarter 1989 RCAF by approximately 8%, which they claim represents the cumulative productivity gain that the railroads have been allowed improperly to retain owing to the want of a deflator in the RCAF for eight years.
In Western Coal we said that the Staggers Act requires the Commission to “revise the index in an appropriate fashion, including consideration of measurable productivity gains, as the circumstances warrant.” 677 F.2d at 931. Building upon that statement, the Shippers argue that once a reliable method for measuring productivity gains was devised, the ICC was obliged to apply it retrospectively so as to “restate” the 1989 RCAF, lest it perpetuate into the future base levels inflated by the non-use of a productivity deflator in the past. Additionally, the Shippers contend that both the ICC’s own decision in Railroad Cost Recovery Procedures, 3 I.C.C.2d 60, 65 (1986), forcing railroads to adjust RCAF-based rates downward in periods of deflation, and the RAPB’s Final Report support this conclusion. Both interpret the Staggers Act to require that the RCAF be used only to allow cost recovery and not to enhance real (i.e. inflation-adjusted) revenues.
Although the Shippers are correct in pointing out that the RCAF was not intended to enhance real railroad revenues, it does not follow that all adjusted base rates antedating the productivity adjustment must be reduced. Certainly, our decision in Western Coal, finding reasonable a RCAF that did not measure productivity gains, suggests no such rigid requirement. See 677 F.2d at 928-29.
We review the Commission’s decision not to adjust the RCAF retroactively for productivity, as we reviewed the underlying decision to adjust at all, pursuant to Chevron: We ask only whether the Commission reasonably interpreted the statute, which requires it to make “appropriate adjustments” to rates in order to reflect changing costs but does not specifically address the question of retroactivity. In deciding not to restate the RCAF, the ICC cited four concerns: (1) a railroad might be made “financially liable for rate actions that were protected when they were taken”; (2) data limitations make it difficult to calculate accurately productivity gains made pri- or to 1986; (3) there is a lack of evidence that present rates are unreasonably high; [228]*228and (4) restatement would have an “unknowable but potentially substantial” impact upon railroads’ earnings. Productivity Adjustment, 5 I.C.C.2d at 470-71.
The ICC’s desire to preserve settled expectations, expressed in the first and perhaps implied in the second and third of these concerns, is not unreasonable. The Shippers strain to argue that restating the second quarter RCAF for 1989 would not, as a matter of law, require a railroad to adjust previously filed rates. We express no opinion on that vexing issue; it is clear enough that restatement would at least make reliance upon past rates an uncertain proposition. For restatement would generate the very sort of litigation that the Congress sought to avoid; if the Commission were to reduce the RCAF by 8%, any railroad that set a rate at 92% or more of the supposedly unchallengeable adjusted base rate for the first quarter of 1989 would surely find that rate now subject to challenge after all.
The Commission also appears to be justifiably concerned that data limitations preclude accurate retroactive recalculation of the RCAF. The productivity measurement it adopted requires five-year trend data that simply are not available for the early 1980s. Accordingly, although the ICC is “certain that restatement would cause a substantial drop in the existing index, the correct measure of the restated index ... cannot be readily developed.” Productivity Adjustment, 5 I.C.C.2d at 470-71. We do not think it unreasonable for the Commission to avoid a venture as speculative as restatement would necessarily be in this circumstance. As we said in Western Coal, using an inaccurate productivity adjustment “could over-deflate the cost index, thereby denying the railroads the timely and unencumbered rate increases they deserve.” 677 F.2d at 928.
Either of these two reasons is enough to support the Commission’s decision not to apply the productivity adjustment retroactively. Therefore, without in any way reflecting upon their merit, we need not concern ourselves with the agency’s other grounds.
III. The 1986 Rate Rollback Decision
When the cost of railroad inputs goes down, the adjusted base rate likewise declines. Nothing in the Staggers Act or the ICC’s 1984 RCAF regulations, however, requires railroads correspondingly to reduce their RCAF-based rates. Thus, when the RCAF initially declined, the ICC permitted the railroads to leave their rates in effect even though those rates exceeded the downward-adjusted base rates. See Ex Parte No. 290 (Sub-No. 2) Railroad Cost Recovery Procedures, decided Nov. 21, 1984 (refusal to order rate reduction in response to decline in RCAF); Ex Parte No. 290 (Sub-No. 2) Railroad Cost Recovery Procedures, (not printed) decided Apr. 19, 1982 (same).
The ICC first ordered the railroads to rollback their RCAF-based rates in 1986, when the price of railroad inputs dropped significantly. See Railroad Cost Recovery Procedures, 3 I.C.C.2d 60 (1986). The Shippers and the AAR petitioned this court for review of that so-called Yoyo decision, but the AAR later withdrew its petition because (it says) it doubted that price declines would recur with much frequency. Conrail, which had not participated directly in the Yoyo proceedings before the ICC nor filed its own petition for review, then moved to substitute itself individually in place of the AAR, pursuant to Fed. R.App.P. 43(b), or to intervene in the controversy, pursuant to Fed.R.App.P. 15(d). We rejected Conrail’s attempt to enter the lists under its own banner, stating that to permit Conrail “to substitute itself for the now-departed trade association or to intervene at this late juncture not only sanctions an undisputed failure to comply with applicable statutes and rules, but ... would squarely conflict with this court’s interpretive approach to the Hobbs Act.” Alabama Power Co. v. ICC, 852 F.2d 1361, 1367 (D.C.Cir.1988) (citing Simmons v. ICC, 716 F.2d 40, 42 (D.C.Cir.1983)). The Hobbs Act, 28 U.S.C. § 2344, provides that a petition for review of a final ICC order must be brought within sixty days of its issuance.
[229]*229Meanwhile, in August 1987 the ICC had given notice that it was splitting off from Ex Parte No. 290 (Sub-No. 2) a new docket, Ex Parte No. 290 (Sub-No. 5), in which it would deal with ministerial and methodological issues concerning the quarterly calculation of the RCAF. See 52 Fed.Reg. 32,613 (1987) (“Comments [in Sub-No. 5] may not properly include substantive issues regarding the propriety of the rules themselves____ Comments suggesting changes to our rules should be filed in Ex Parte No. 290 (Sub-No. 2)”). Despite this notice, in 1989 and 1990 Conrail and the AAR made brief arguments challenging the ICC’s authority to order a rate rollback, not in Ex Parte No. 290 (Sub-No. 2), but in the course of their submissions in Ex Parte No. 290 (Sub-No..4) relating to the productivity adjustment issue.
The Railroads now ask the court in effect to review for a second time the ICC’s 1986 Yoyo decision, claiming that it exceeds the agency’s statutory authority. The Railroads insist that their petition is properly before the court, notwithstanding the 60-day limitations period for filing a petition for review, because “the 1989-90 [productivity adjustment] decisions, by substantially expanding the scope of the rollback requirement, ... entitle the railroads ... to a renewed opportunity to challenge the requirement.” The Railroads rely upon MCI Telecommunications Corp. v. FCC, 765 F.2d 1186, 1190 (D.C.Cir.1985), for that proposition. Alternatively, the Railroads rely upon NLRB Union v. FLRA, 834 F.2d 191, 195 (D.C.Cir.1987), for the proposition that whenever a rule is applied to it, a party may challenge that rule on the ground that it exceeds the agency’s statutory authority. Conrail, in particular, insists that it “has a special claim of entitlement to challenge the rollback order now, having been denied an adequate opportunity to test” the Yoyo rule when it was first promulgated.
First, we do not agree that the Commission’s productivity adjustment decision so altered the Yoyo decision as to entitle the Railroads to another opportunity to challenge it. This case is not analogous to MCI, where the petitioner became aggrieved only when the agency changed a permissive regulation into a mandatory one. 765 F.2d at 1189-90. Because the agency had reopened the issue and “fundamentally altered” the previous rule, we held that a party aggrieved by the change could challenge the agency’s authority to make it. See id. at 1190. In the present case, by contrast, the Commission’s decision to take account of productivity gains in the RCAF did not reopen its 1986 decision requiring that RCAF-based rates be lowered when the RCAF declines; it simply altered the manner in which the RCAF would be calculated, a detail that the Railroads may (and do above) challenge at the margin without implicating the earlier decision.
We also reject the Railroads’ claim that under NLRB Union they may now challenge the ICC’s statutory authority for the Yoyo rule because it was in some sense applied to them in the later productivity adjustment decision. NLRB Union, which suggests that a party may challenge an agency’s statutory authority whenever the agency applies a rule against it, is simply not applicable here. In adding a productivity adjustment to the Yoyo rule, the ICC did not apply the Yoyo rule to any particular carrier nor extend it to any new circumstance.
As we recently noted:
[A] claim that agency action was violative of statute may be raised outside a statutory limitations period, by filing a petition for amendment or rescission of the agency’s regulations, and challenging the denial of that petition.
Public Citizen v. Nuclear Regulatory Comm’n, 901 F.2d 147, 152 (D.C.Cir.1990); see also Geller v. FCC, 610 F.2d 973, 978 (D.C.Cir.1979); Functional Music, Inc. v. FCC, 274 F.2d 543, 546 (D.C.Cir.1958). Absent such a petition to the agency, we have reviewed a rule only when
the agency’s action [1] did not reasonably put aggrieved parties on notice of the rule’s content, or [2] clearly remained unripe for judicial review throughout the statutory review period, ... [or] in extreme cases involving [3] gross violations [230]*230of statutory or constitutional mandates, or [4] denial of an adequate opportunity to test the regulation in court.
Raton Gas Transmission Co. v. FERC, 852 F.2d 612, 615 (D.C.Cir.1988) (citations omitted).. When the aggrieved party neither petitions the agency to change the rule nor presents a “valid excuse” for failing to mount a timely challenge to the initial regulation, considerations both of finality and of exhaustion dictate that the litigant should be bound by that regulation. See Investment Co. Inst. v. Bd. of Governors, 551 F.2d 1270, 1281-82 (D.C.Cir.1977).
In a supplemental brief for which the court called after the oral argument in this case, the Railroads also seem to argue that we should treat as a petition for rescission the challenge to the productivity adjustment that they filed in docket Sub-No. 4. We cannot do that. The Commission plainly designated Sub-No. 2 as the proper docket in which to file substantive comments regarding the RCAF, and, barring extreme arbitrariness, we defer to the agency’s decision regarding the management of its own docket. See Nader v. FCC, 520 F.2d 182, 195-96 (D.C.Cir.1975).
Nor have the Railroads presented a persuasive argument that they should be excused from the petition requirement. In their supplemental brief the Railroads claim that because the ICC has “defended its authority to [adopt the Yoyo decision] in this Court, [and] has expanded the scope of the requirement to include productivity-related RCAF changes ... any hope that a different docket caption on the railroads’ 1989 comments might have induced the ICC to consider the matter afresh is unfounded.” See Randolph-Sheppard Vendors of America v. Weinberger, 795 F.2d 90, 105 (D.C.Cir.1986) (exhaustion not required when “following the administrative remedy would be futile because of certainty of an adverse decision”).
In fact the ICC has not in this case defended on the merits its authority to adopt the Yoyo decision; rather it has stood pat upon the ground that that question is not now before the court. More important, we cannot conclude that merely because the agency in 1989 adhered to its 1986 position it would certainly do so now if the Railroads, by filing a proper petition, were to require the agency to give a current account, subject to judicial review, of its reasons for rejecting their argument. This is especially so in view of the Agency’s unqualified assertion in its post-argument brief that “[e]xhaustion would not be pointless here.”
Finally, we reject Conrail’s suggestion that it has “a special claim of entitlement” to contest the rollback order now because it missed its chance to do so in Alabama Power, 852 F.2d at 1368. Conrail lost the opportunity to challenge the Yoyo decision then because it had been “asleep at the switch.” Id. Its challenge is not even now before the court only because of its own failure to follow the procedures clearly prescribed by the agency to which its petition must be addressed in the first instance.
IV. Conclusion
The Commission did not exceed its statutory authority by including railroad productivity gains in the RCAF. Neither did it act arbitrarily; after ensuring that a productivity adjustment was desireable as policy and feasible methodologically, the ICC adopted the measure while pledging to continue to improve its accuracy. Nor was the Commission unjustified in its decision not to implement the productivity adjustment retroactively.
Finally, we decline to hear the Railroads’ argument that the Agency exceeded its statutory authority by requiring a rate rollback when the RCAF declines. Because the Railroads failed to follow agency procedures, their argument is not properly before us now. Accordingly, both the Railroads’ and the Shippers’ petitions for review are
Denied.