The Staggers Rail Act of 1980 has placed tight time limits on proceedings before the Interstate Commerce Commission to abandon railroad lines; and the most important issue raised by this petition to review an order of the ICC authorizing the Illinois Central Gulf Railroad to abandon a 31-mile line between Clinton and New Holland, Illinois is whether the Commission, in its zeal to expedite the proceeding, denied the opponents of the abandonment (principally shippers) due process of law.

The Act creates three categories of abandonment: the unprotested application, which the Commission must grant within 45 days; the protested application that the Commission decides not to investigate, in which the Commission has 75 days from the filing of the application to decide whether to authorize abandonment; and the protested application that is investigated, in which an initial decision — one subject to review by the full Commission — is due within 165 days. 49 U.S.C. §§ 10904(b), (c)(1), (2), (3). As the Act allows only 30 days from the date of the application for protests to be filed, and only 15 days after that for the Commission to decide whether to investigate, 49 U.S.C. § 10904(c)(1), there are only 120 days between the deadline for the Commission to order an investigation and the deadline for the initial decision.

As we explained in Illinois v. ICC, 709 F.2d 1186, 1191 (7th Cir. 1983), “investigation” is a misnomer in the abandonment context. It just means a somewhat more elaborate evidentiary process. Instead of the Commission’s basing its decision on just the application and the protests, as it does when there is no investigation, when it “investigates” it receives affidavits (“verified statements” in ICC parlance) from the applicant and the protestants and bases its decision on those. In this case the Commission used its “modified” investigative procedure, which means it did not hold an oral hearing. See 49 C.F.R. §§ 1100.43 et seq.; Anderson, ICC: Practice and Procedure 110-11 (1966). Under this procedure the applicant must put in his opening case (i.e., his affidavits) within 15 days after the Commission has ordered an investigation; the protestants must put in their case within 40 days after the Commission’s order; the applicant must put in his reply case within 55 days after the order; and the “parties shall not be permitted to file motions to strike all or part of any evidence submitted.” 49 C.F.R. § 1152.25(d)(6).

Within the 120-day statutory limit a review board (the initial decision-maker within the Commission) authorized abandonment, and the protestants then appealed to the full Commission, which upheld (formally, denied the appeal from) the review board in a brief order. The review board had found that the railroad’s revenues from the line were dropping steadily, with no prospect for a turnaround; by 1981 they were only $25,000, resulting in operating losses that year of almost $250,000. To these losses the board added another $175,-000 representing the income that the railroad would obtain by liquidating the line and reinvesting the assets salvaged from the liquidation. The board also noted that if the line were kept in service the railroad would have to spend hundreds of thousands of dollars rehabilitating it. Although acknowledging that the protesting shippers would be hurt by the abandonment, the board noted that they were all located within a few miles of other rail lines and also could use (and on occasion had used) trucks in lieu of rail to move their shipments. The board concluded that abandonment would be consistent with the public convenience and necessity, the statutory criterion. 49 U.S.C. § 10903(a).

Although the facts found by the board may seem to make a compelling case for abandonment, cf. Chicago & N.W. Transport. Co. Abandonment, 366 I.C.C. 373, 380 (1982), the protestants argue that the factual picture painted by the board is *1346incomplete in important respects. The first relates to operations performed at Lincoln, Illinois. Lincoln is not on the Clinton-New Holland line, but the train that services the line (or rather serviced it, the Commission having refused to stay its order authorizing abandonment pending this review proceeding, although the line remains in place in case we decide to reverse the Commission) —the “Lincoln switcher” as it is called— performs services at Lincoln that generate substantial revenues that the railroad refused to attribute to the Clinton-New Holland line. This was a reasonable decision. The Lincoln switcher will continue to perform those services; it just will not run back and forth between Clinton and New Holland. Revenues unaffected by abandonment are not revenues of the abandoned line.

The second factual issue relates to the alleged labor cost savings from the abandonment. The protestants argue that the railroad’s labor contracts will prevent it from laying off the crews who serve the line and that therefore the wages of those crews should not be counted as a cost that will be saved if the line is abandoned. Regarding the only workers as to whom the protestants are still making this contention, the'firemen, the board found that the railroad would be able to redeploy them elsewhere on its system, bumping less senior employees if necessary. The board seems to have thought that such a finding was all that was necessary to establish that the crews’ wages were a prospective savings from abandonment. It is not. Whether when the dust settles the result of all the bumping will be an actual reduction in the railroad’s payroll, and if so by the full amount of the wages of the firemen on the abandoned line, are separate questions from whether bumping will occur — questions not discussed by the board. But the amount of alleged savings that are in dispute is slight — less than $40,000 a year, including fringe benefits and payroll taxes — relative to the railroad’s annual losses on the Clinton-New Holland line, which are at least $500,000 when the cost of rehabilitating the line, a cost that would have to be incurred if the line were kept in service, is figured in. Thus, even if the review board exaggerated the savings in firemen’s wages, the error was not big enough to affect the board’s decision; it was harmless. Cf. Illinois v. ICC, 698 F.2d 868, 876 (7th Cir. 1983). (More below on harmless error in administrative proceedings.)

The protestants next complain about the refusal of the board (and of the Commission in reviewing the board) to pay any attention to the railroad’s overall financial condition. They submitted as part of their case a magazine article quoting the chairman of the board of directors of IC Industries, the parent of the Illinois Central Gulf Railroad, as having criticized securities analysts’ portrayals of the railroad as “financially troubled,” and an excerpt from IC Industries’ latest annual report, showing that the profits of IC Industries have been rising. The review board, however, held that “The financial condition of IC Industries [the parent] is not in issue in this proceeding. [And] ICG’s [the railroad’s] overall earnings have no relevance to the issues in this proceeding where the uncontroverted evidence shows that ICG is losing revenues on the line proposed for abandonment.”

The protestants argue that the statutory standard of public convenience and necessity requires the Commission to give at least some weight to the overall profitability of a railroad that is asking to be allowed to abandon a line. This is not a frivolous argument. The Supreme Court stated in Colorado v. United States, 271 U.S. 153, 168-69, 46 S.Ct. 452, 456, 70 L.Ed. 878 (1926) (footnote omitted): “In some cases, although the volume of the whole traffic is small, the question is whether abandonment may justly be permitted, in view of the fact that it would subject the communities directly affected to serious injury while continued operation would impose a relatively light burden upon a prosperous carrier.” The Court in Southern Ry. v. North Carolina, 376 U.S. 93, 105, 84 S.Ct. 564, 571, 11 L.Ed.2d 541 (1964), quoted this language and added: “In cases falling within the latter category [i.e., serious injury to communities], such as those involving vital commuter services in large metropolitan areas where the demands of public convenience *1347and necessity are large, it is of course obvious that the Commission would err if it did not give great weight to the ability of the carrier to absorb even large deficits resulting from such services."

These are not isolated statements. They reflect a long and well-documented history of forcing public utilities and common carriers to finance losing operations out of profitable ones-the practice known as "internal subsidization" or "cross-subsidization." See, e.g., Bonbright, Principles of Public Utility Rates 111-12 (1961); Conant, Railroad Mergers and Abandonments 132 (1964); Friedlaender, The Dilemma of Freight Transport Regulation 66-68 (1969); Friedlaender & Spady, Freight Transport Regulation 9-10 (1981); Hilton, The Transportation Act of 1958: A Decade of Experience 136 (1969); Meyer, Peck, Stenason & Zwick, The Economics of Competition in the Transportation Industries 166-67, 194-95~ (1959); MacAvoy, The Regulated Industries and the Economy 40-41 (1979); Morgan,, Cases and Materials on Economic Regulation of Business 423-59 (1976); Nelson, Railroad Transportation and Public Policy 286-301, 331 (1959); Harris, Economies of Traffic Density in the Rail Freight Industry, 8 Bell J.Econ. 556, 563 (1977). A railroad would not abandon a line unless at the rates the Commission allowed it to charge it was losing money, so the only reason the Commission would have for preventing abandonment would be to shift the cost of service from the shippers using the line either to shippers on other lines or to the railroad's owners or creditors. But internal subsidization can, of course, bankrupt a company or an industry. Although the nation's railroads have been in a parlous financial state for most of this century, at the time the statements we have quoted were made (and long after) internal subsidization still seemed, at least to the Interstate Commerce Commission, a viable policy. It is no longer. Not only has the financial condition of the railroad industry become critical, but Congress has decided that excessive regulation is one of the culprits, and in the 4-R Act (see Railroad Revitalization and Regulatory Reform Act of 1976, Pub.L. 94-210, § 802, 90 Stat. 127) and in the Staggers Act it determined to loosen the grip of regulation over abandonments and other practices. See 49 U.S.C. § 10101a; S.Rep. No. 499, 94th Cong., 1st Sess. 2-3 (1975), U.S.Code Cong. & Admin.News 1976, p. 14; S.Rep. No. 470, 96th Cong., 1st Sess. 5-6 (1979); H.R.Rep. No. 1035, 96th Cong., 2d Sess. 38, 53-54 (1980), U.S.Code Cong. & Admin.News 1980, p. 3978; Chicago & N. W. Transport. Co. v. United States, 582 F.2d 1043, 1045-46 (7th Cir. 1978).

Although the focus of legislative reform in the abandonment area has been on reducing delay in processing applications to abandon, it appears that Congress's concerns are not purely procedural; that it believes the railroads cannot continue to support deficit operations out of their all-too-few profitable operations and therefore abandonments should be more freely permitted. See S.Rep. No. 499, supra, at 3, 39-41, 44; S.Conf.Rep. No. 595, 94th Cong., 2d Sess. 219 (1976), U.S.Code Cong. & Admin.News 1976, p. 14; S.Rep. No. 470, supra, at 6. The statutory criterion for permitting abandonment-public convenience and necessity-has not been changed, see Id. at 38-40, but it is elastic enough to allow the Commission to respond to the changing realities of the railroad industry by deciding that a railroad's overall profitability no longer should be given any weight in determining whether to allow the railroad to abandon a line on which it is incurring substantial losses. Cf. S.Rep. No. 499, supra, at 44; S.Conf.Rep. No. 595, supra, at 219.

The protestants also complain, however, about the manner in which the Commission made this decision-one sentence in a review board decision, echoed by a similar sentence in the Commission's order declining even to hear an appeal from the board's decision, with no explanation, or citation of supporting authority, at either level. Although the Commission and the railroad assure us that the Commission's change of policy has been discussed in many recent Commission decisions, none is cited to us and our own research has not brought any to light. In the closest approximation to such a case that we have found, Norfolk *1348& W. Ry. Co. Abandonment, 363 I.C.C. 115, 119 (1980), the Commission stated: “It does not appear logical to require a financially healthy railroad to operate an unprofitable branch line, continue normalized maintenance, and possibly rehabilitate the line when there are no substantial factors weighing against abandonment. This is especially so in light of the economic plight of the railroad industry. We conclude that the overall financial strength of a railroad should not be an overriding factor leading to the denial of an application for permission to abandon an unprofitable line of railroad.” The language we have italicized suggests adherence to the Cobrado standard, as do other recent decisions even more clearly. See Chicago, Rock Island & Pac. R.R. Abandonment, 363 I.C.C. 150, 159 (1980); Illinois Central Gulf R.R. Abandonment, 363 I.C.C. 690, 703-04 (1980); Illinois Central Gulf R.R. Abandonment, 363 I.C.C. 729, 737 (1980). As we remarked a few months ago, “system profitability repeatedly has been disclaimed as a controlling factor in abandonment,” Simmons v. United States, 698 F.2d 888, 897 n. 4 (7th Cir. 1983) (emphasis added) — implying that the Commission thought it one factor; now apparently it thinks it no factor at all.

The Commission may make policy in common law fashion, see NLRB v. Bell Aerospace Co., 416 U.S. 267, 294, 94 S.Ct. 1757, 1771, 40 L.Ed.2d 134 (1974), and therefore may unmake it in the same fashion. But it may not make major policy changes by pure fiat, with no explanation of why the change is consistent with the statute. Atchison, Topeka & Santa Fe Ry. v. Wichita Bd. of Trade, 412 U.S. 800, 807-08, 93 S.Ct. 2367, 2374-2375, 37 L.Ed.2d 350 (1973); Brennan v. Gilies & Cotting, Inc., 504 F.2d 1255, 1264 (4th Cir. 1974); see also Motor Vehicle Mfrs. Ass’n v. State Farm Mutual Automobile Ins. Co., - U.S. -, 103 S.Ct. 2856, 2866, 77 L.Ed.2d 443 (1983). The standard of public convenience and necessity is sufficiently flexible to allow a reasonable ICC at one time to think that it ought to take into account system profitability and at another time to think it ought not to, but if it is to change course so completely it must give a reason why; otherwise its behavior is arbitrary and capricious and therefore a violation of the Administrative Procedure Act. 5 U.S.C. § 706(2)(A); Missouri-Kan.-Tex. R.R. v. United States, 632 F.2d 392, 399-400 (5th Cir. 1980).

But this point does not require reversal in this case. The protestants’ evidence of system profitability was so extraordinarily weak — meaningless, really— that we cannot believe it would have made any difference to the outcome of the case even if the Commission continued to regard system profitability as a weighty factor against granting an application to abandon. The “evidence” consisted of a magazine article containing a general and unsubstantiated reply to some disparagers of the railroad’s profitability and an excerpt from an annual report not of the railroad itself but of its conglomerate parent. The railroad might be losing money even if its parent was making money; profits from other operations might be sufficient to offset railroad losses. Moreover, the review board found no adverse effect of abandonment on “community development,” which seems to be an essential element of the rationale for internal subsidization stated in the Colorado and Southern Ry. opinions.

True, we are not permitted to substitute our own rationale for the Commission’s; we are not permitted to say, as we would be if we were reviewing a district court decision, that though we have doubts about the propriety of the stated ground for decision there is another ground that shows the decision was correct and we affirm on it. SEC v. Chenery Corp., 318 U.S. 80, 87-88, 63 S.Ct. 454, 459, 87 L.Ed. 626 (1943). Congress has not delegated to the courts the making of regulatory policy for the railroad industry. But if we are sure that the agency would if we remanded the case reinstate its decision — if in other words the error in its decision was harmless — a reversal would be futile, and Chenery does not require futile gestures. See Massachusetts Trustees of Eastern Gas & Fuel Assocs. v. United States, 377 U.S. 235, 248, 84 S.Ct. 1236, 1245, 12 L.Ed.2d 268 (1964); Pfizer, Inc. v. Richardson, 434 F.2d 536, 547 n. 21 (2d Cir. 1970); Consolidated Gas Supply *1349Gorp. v. Federal Energy Regulatory Comm’n, 606 F.2d 323, 329 (D.C.Cir. 1979); NLRB v. American Geri-Care, Inc., 697 F.2d 56, 64 (2d Cir. 1982); Friendly, Chenery Revisited: Reflections on Reversal and Remand of Administrative Orders, 1969 Duke L.J. 199. As Judge Friendly has explained, Chenery “was intended only to establish the1 important point that a reviewing court could not affirm an agency on a principle the agency might not embrace — not to require the tedious process of administrative1 adjudication and judicial review to be needlessly dragged out while court and agency engage in a nigh endless game of battledore and shuttlecock with respect to subsidiary findings.” Erie-Lackawanna R.R. v. United States, 279 F.Supp. 316, 354-55 (S.D.N.Y. 1967) (three-judge court), aff’d with modifications, and remanded, sub nom. Penn-Central and N & W Inclusion Cases, 389 U.S. 486, 88 S.Ct. 602, 19 L.Ed.2d 723 (1968). It is inconceivable to us that the Commission1 would have come to a different conclusion in this case even if it had decided to give great weight to the overall financial profitability of any railroad seeking permission to abandon a line. The protestants did not put in enough evidence to create in the collective mind of a rational administrative agency a genuine issue of the railroad’s profitability. A remand would therefore be pointless, and is not required.

Another procedural issue grows out of the Commission’s denial of the protestants’ motion to strike evidence contained in the railroad’s evidentiary submission and to be allowed to cross-examine some of the railroad’s witnesses. The protestants argue that the regulation abolishing motions to strike violates due process of law and that in any event the Commission should have paid attention to the contents of the motion, even if not granting it formally.

The Commission argues that the regulation can be challenged only by a party to the rulemaking proceeding in which the regulation was adopted, and only within the time (which has expired) permitted for judicial review of such proceedings. See 28 U.S.C. §§ 2321(a), 2342(5), 2344. No doubt, these protestants could have become parties to the rulemaking proceeding but it would be unreasonable to require that they have done so, and have brought a review proceeding to challenge the regulation, as. a condition of being allowed to challenge the abolition of motions to strike. When the regulation was adopted in 1981 the particular shippers who are protestants in this case may not have had any proceeding pending or imminent before the ICC, and, even if they did have one, may not have appreciated the possible impact of the abolition of motions to strike on their protests. If a challenge on their part to the rule abolishing such motions would therefore have been premature, clearly their failure to have mounted such a challenge would not bar them from challenging the rule in this proceeding. See Baltimore Gas & Elec. Co. v. ICC, 672 F.2d 146, 149-50 (D.C.Cir. 1982); Geller v. FCC, 610 F.2d 973, 977-78 (D.C.Cir. 1979). But even if a challenge then would not have been premature or unripe, still it would not be sound policy always to force people to challenge regulations, especially procedural regulations, just as soon as they are promulgated. The judicial mind is concentrated by considering issues in context. A rule abolishing motions to strike might seem reasonable in the abstract and yet when judges saw the consequences of applying the rule in particular cases they might realize that actually it was unreasonable.

Although the protestants are entitled to challenge the regulation in this proceeding, their challenge has no merit. What in ICC parlance is called a “motion to strike” is nothing like a motion to strike under Rule 12(f) of the Federal Rules of Civil Procedure; it is an objection to the admissibility of evidence on grounds of hearsay and the like and when it is granted the result is to exclude evidence. See 4 Sharfman, The Interstate Commerce Commission 214 (1937). The rules of evidence were developed to control juries, and although nowadays they also apply in bench trials they do not bind administrative agencies, Opp Cotton Mills v. Administrator, 312 U.S. 126, 155, 61 S.Ct. 524, 537, 85 L.Ed. 624 (1941), which indeed have long been criticized for paying too much attention to them. See, e.g., Gellhorn, Rules of Evi*1350dence and Official Notice in Formal Administrative Hearings, 1971 Duke L.J. 1, 16-17 and n. 66. The Commission therefore acted well within its power when it decided to do away with a type of motion that had served only to present evidentiary challenges that might be important in a jury trial but that are largely a waste of time in an administrative proceeding. Even when motions to strike were allowed, they were rarely granted. The Commission’s standard response was to deny the motion with the statement that it would give appropriate weight to any objectionable testimony. (On the Commission’s traditionally liberal attitude toward hearsay evidence see, e.g., Spiller v. Atchison, Topeka & Santa Fe Ry., 253 U.S. 117, 129-32, 40 S.Ct. 466, 471-72, 64 L.Ed. 810 (1920); Western Trunk Lines Iron & Steel, 47 I.C.C. 109, 122-23 (1917).) Abolition did little more than formalize the agency’s practice. And it was particularly appropriate in reference to proceedings in which Congress had placed the Commission under tight time constraints. It would be a foolish use of the Commission’s limited time for deciding abandonment cases to worry about admissibility as distinct from weight of evidence.

But it does not follow that the Commission was free to ignore completely the contents of the protestants’ motion to strike. In abolishing motions to strike the Commission stated, “Nothing precludes parties from calling to our attention attempts to introduce new or misleading evidence, without filing a formal motion.” Abandonment of Railroad Lines & Discontinuance of Service, 365 I.C.C. 249, 253 (1981). Having extended the invitation the Commission is not free to ignore the response. Yet all the review board said here was, “Joint protestants move to strike various statements made by applicant’s witnesses in the reply statement as hearsay, inaccurate, or new evidence. Joint protestants also request to cross examine the witness. Much of the motion consists of improper rebuttal and will not be considered. Joint protestants’ motion to strike and for cross examination is denied. 49 C.F.R. 1121.36(d)(b).” (The citation is to the identical predecessor of 49 C.F.R. § 1152.25(d)(6), quoted earlier; the “(b)” must be a misprint for “(6).”) Thus the board seems to have denied the motion to strike in its entirety merely because it contained some improper rebuttal; to have ignored completely the request for cross-examination; and also to have ignored completely the claim in the motion to strike that the railroad’s evidentiary submission was inaccurate and contained new evidence.

But having read the motion ourselves, we do not think we should reverse on the basis of the careless treatment that the review board gave it. The motion is only six pages long and most of it is given over to labeling various parts of the railroad’s evidentiary submission as “hearsay” without explaining why the Commission should not give hearsay evidence such weight as is appropriate rather than reject it out of hand. The motion does not indicate the importance of the evidentiary questions that it raises and indeed they all have the appearance of being nitpicks. In their brief to this court the protestants do not indicate how acceptance even of all the points made in the motion might reasonably be expected to have led to a different result. We cannot believe that a reasonable tribunal which thought abandonment warranted before it read the motion to strike would have decided it was not warranted after reading it, even if the tribunal believed every word in the motion. Once again, we think any error was harmless.

The review board had earlier granted (in part) a motion by the railroad to strike some protestant evidence; the protestants argue that this difference in treatment was unfair. But the railroad’s motion was filed before the Commission decided to apply its modified procedure (in which motions to strike are not allowed) to this case, and the protestants’ afterward. Changing procedural rules in the middle of a case may seem the quintessential due process violation but is in fact common. For example, in its recent order promulgating substantial amendments to the Federal Rules of Civil Procedure, the Supreme Court stated that the amendments “shall take effect on August 1, 1983 and shall govern all civil proceedings thereafter commenced and, insofar as just and practicable, *1351in proceedings then pending.” 103 S.Ct. No. 16, at 1 (June 15,1983). We cannot say it was unjust for the Commission to alter procedures midway in this proceeding. The protestants were aware when they filed their motion that motions to strike had been abolished in cases tried under modified procedure and they have not shown how they were harmed by the railroad’s having been allowed to file such a motion earlier.

Although we have decided to affirm the order approving abandonment, we warn the Commission that it must not allow the time limitations under which it labors in abandonment cases to cause it to overlook the procedural rights of opponents of abandonment. Particularly in dismissing the protestants’ evidence of system profitability out of hand and in denying their motion to strike without explicit consideration of any valid points that the motion might have raised, the Commission came perilously close to violating due process and the Administrative Procedure Act. But since we are convinced that the Commission’s procedural pratfalls could not have affected the outcome of the case, the order authorizing abandonment is

Affirmed.