This case requires us to consider again the proper scope of judicial review of decisions by the Federal Energy Regulatory Commission ("FERC" or "Commission") concerning the reasons given to justify oil pipeline rate suspensions. On March 18, 1983, FERC issued an order suspending for seven months a proposed rate filed by Exxon Pipeline Company for the Trans Alaska Pipeline System ("TAPS"). A seven-month suspension is the maximum period permitted by the Interstate Commerce Act.1 On March 31, 1983, FERC issued a similar order suspending the effective date of a proposed rate filed by BP Pipelines Inc. ("BP"), also for seven months. In this consolidated action, Exxon and BP challenge the adequacy of the reasons provided by the Commission for the length of the suspensions.2 We conclude that, under the very narrow standard of review set out below, FERC has articulated adequate reasons to withstand further judicial intervention. Accordingly, we affirm.

The Trans Alaska Pipeline System carries crude oil from producing fields in Alaska to a marine terminal on the Pacific coast of the state. Exxon and BP are two of eight oil companies that hold an undivided interest in TAPS facilities. Each owner files with FERC its own proposed rate for transportation of crude oil through its share of the total capacity.

On February 17, 1983, Exxon filed FERC Tariff No. 268 to increase its tariff by 97 cents, from $5.30 per barrel to $6.27 per barrel, with a proposed effective date of March 19, 1983. On March 7, 1983, the State of Alaska, an intervenor in this action, filed with FERC a protest and petition for a seven-month suspension and an investigation, as authorized by section 15(7) of the Interstate Commerce Act. Alaska argued, inter alia, that the rate was unjust and unreasonable, in violation of section 1(5) of the Act.3 Exxon responded that the tariff was just and reasonable and that, in any event, a one-day rather than a seven-month suspension would be consistent with established FERC policy. It cited the Commission's decision in Buckeye Pipe Line Co., 13 FERC (CCH) p 61,267 (Dec. 24, 1980), which announced a FERC policy of suspending oil pipeline rates for one day only.

On March 18, 1983, one day before the proposed effective date of the tariff, FERC issued its order suspending the rate for seven months and instituting a formal investigation.4 The FERC order explained that the rate had "not been shown to be just and reasonable, and may be unjust, unreasonable, unduly discriminatory or otherwise unlawful," 22 FERC (CCH) at 61,569, and that it would therefore be suspended. Addressing the length of suspension, the Commission acknowledged its established policy that oil pipeline rates generally should be suspended for one day only, absent extraordinary circumstances. But, it continued,

[s]uch circumstances are presented here because the rate increase by Exxon could have a detrimental impact on the citizens of Alaska. This is because a higher TAPS transportation rate results in a lower wellhead price paid to the producer, which in turn means a lower royalty payment to the State of Alaska. We will therefore exercise our discretion and suspend the filing for seven months, or until October 19, 1983, at which time it will be allowed to go into effect subject to refund.

22 FERC (CCH) at 61,569.

BP's tariff increase effort followed a similar course. On March 2, 1983, BP filed FERC Tariff No. 4 to increase its TAPS rates,5 with a proposed effective date of April 1, 1983. The State of Alaska then filed a protest and petition for suspension. On March 31, 1983, one day before the proposed effective date, FERC issued an order suspending the new rates for seven months and instituting an investigation.6 As it had done in the Exxon case, FERC referred to the detrimental impact that the new rates might have on the citizens of Alaska to justify the length of the suspension. 22 FERC (CCH) at 61,633.

In this action, Exxon and BP do not challenge the FERC decisions to suspend the proposed tariffs for some period; nor do they contest the decisions to institute investigations. Rather, their challenges focus on the reasons given for the length of the suspensions, and the alleged failure by FERC to explain adequately why the Buckeye one-day suspension rule was not followed.

We begin our analysis with a brief review of the applicable statutory scheme. Section 15(7) authorizes the Commission to institute a proceeding to determine the lawfulness of a proposed rate. Pending final action in that proceeding, the Commission may suspend the proposed rate for up to seven months beyond the time it would otherwise go into effect by delivering to each affected carrier, and filing with the proposed rate, a statement of reasons for the suspension.

If FERC has not issued a final ruling on the rate at the conclusion of the suspension period, the new rate goes into effect. In that instance, the Commission may require the carrier to account for all amounts received under the new rate. In the event that FERC ultimately finds any part of the increased rate to be in violation of the Interstate Commerce Act, it may require the carrier to refund the appropriate amount, with interest. See 49 U.S.C. Sec. 15(7) (1976). If the proposed rate is determined to be reasonable, the carrier is not reimbursed for money foregone during the suspension period.

Several propositions about the narrow scope of judicial review of Commission actions pursuant to this section are absolutely clear. It is well-established that a court may not review a Commission decision as to whether or not to suspend a rate, at least as long as the agency complies with its statutory obligation to give a reason,7 and in no other way oversteps the bounds of its authority. See Trans Alaska Pipeline Rate Cases, 436 U.S. 631, 638 n. 17, 652-53, 98 S. Ct. 2053, 2058 n. 17, 2065-66, 56 L. Ed. 2d 591 (1978); Arrow Transportation Co. v. Southern R. Co., 372 U.S. 658, 83 S. Ct. 984, 10 L. Ed. 2d 52 (1963); Municipal Light Boards v. FPC, 450 F.2d 1341 (D.C.Cir.1971), cert. denied, 405 U.S. 989, 92 S. Ct. 1251, 31 L. Ed. 2d 455 (1972). It is less well-established, however, whether a court may review the reasons given by FERC to satisfy either a suspension or its length;8 likewise, it is unclear whether a court may or should remand a case to the agency for reconsideration of the reasons where some flaw is perceived in agency decisionmaking. These narrow yet recurring issues are the ones that we confront today.

In addressing these issues, we do not write on a clean slate. As a starting point, we consider a pair of cases in this Circuit that have grappled with judicial review of decisions by FERC concerning the reasons for the length of a suspension, and a pair of Commission decisions that have established policy relevant to the length of suspensions. We recognize that existing precedent in this Circuit is somewhat obscure, primarily because the two leading cases, Connecticut Light & Power Co. v. FERC, 627 F.2d 467 (D.C.Cir.1980), and Delmarva Power & Light Co. v. FERC, 671 F.2d 587 (D.C.Cir.1982), do not sit well together. Our conclusion on the proper scope of review, set out below, is intended to clarify existing case law and enable parties before the Commission, as well as FERC itself, to glean some clear understanding of the potential (and limited) role of the judiciary in these matters.

In Connecticut Light, the electric utility petitioned for review of a FERC order suspending for five months a proposed new rate schedule. Under section 205(e) of the Federal Power Act, as under section 15(7) of the Interstate Commerce Act,9 the Commission may suspend rates as long as it gives a statement of reasons. The court considered whether the order provided the requisite statement of reasons, held that it did not, and remanded to the Commission for an accounting of its reasons.

FERC had justified the suspension in Connecticut Light with "boiler plate" language explaining that rates had not been shown to be just and reasonable and might be unjust, unreasonable or otherwise unlawful. The court ruled that the same boiler plate language might justify suspension in every case but certainly would not justify different lengths of suspension:

Of course, indistinguishable rationales for differing results constitute no rationale at all.... WHY a five-month suspension is justified instead of one-month--or one day--is within the agency's knowledge, because it consciously made a decision to impose a five-month or one day suspension. If the agency had no knowledge, standards, or rationale to guide it in making the choice, then its choice was totally arbitrary and must be set aside. If the FERC has standards or a rationale by which it chooses a period of suspension ranging from one day to five months, then the statute commands that they be articulated, so that their reasonableness and application to the individual case can be known, and so that the regulated entities will have some guidance for the future.

627 F.2d at 471. As is apparent from the above-quoted passage, the court rejected the Commission's assertion that its reasons were unreviewable and held that a court indeed has jurisdiction to determine whether a suspension order complies with the statutory reasons requirement. Id. The court then defined the requirement by stating that the agency must give "reasons for the length of a suspension that fit the fact situation of the relevant case." Id. at 473 (footnote omitted) (emphasis added). As a result, by declaring that it would consider the applicability of FERC's reasons to individual cases, the court's opinion was read by some as opening the door to challenges to Commission suspension decisions.

Connecticut Light also ordered the Commission to establish standards for the exercise of its discretion, either through rulemaking or case-by-case adjudication, "to assure that its discretion is not exercised in an arbitrary way and to give guidance to parties filing and challenging rate increases." Id. FERC responded in Buckeye Pipe Line Co., 13 FERC (CCH) p 61,267 (Dec. 24, 1980), in which it began to articulate a framework for determining oil pipeline suspension lengths. The Commission observed that previously it had formulated a policy to suspend electric utility and natural gas rates for the maximum period permitted by statute. 13 FERC (CCH) at 61,593-94 & n. 9. It concluded, however, that oil pipeline suspensions probably should be governed by different rules, and that such rates should be suspended for one day only, absent extraordinary circumstances.

The Buckeye decision explained that differences in lengths of suspension were justified by the disparate impacts on consumers of rate increases in three different industries. Electric power and natural gas rate regulation is cost-based, and wholesale rate increases "flow through" to retail bills. Oil pipeline rates represent a small regulated portion of an otherwise vast unregulated whole--oil pricing. Market forces, not regulatory concepts, primarily control prices, and "the pipeline charge does not bulk large in the price of the end product." 13 FERC (CCH) at 61,594. Thus, electric power and natural gas rate increases as a rule affect consumers more than pipeline rates. At the same time, for reasons noted below,10 statutory refunds do not sufficiently compensate consumers forced to pay higher rates that later are ruled unreasonable. As a result, pursuant to the decision in Buckeye, in instances where consumers bear the brunt of the increase--as with electric utility and natural gas rates--the enunciated FERC policy was to suspend rates for the maximum period. However, where shippers, not consumers, bear the brunt--as with oil pipelines--the announced FERC policy was to suspend rates for one day only.11

In a second decision, Williams Pipe Line Co., 21 FERC (CCH) p 61,260 (Nov. 30, 1982), appeal pending sub nom. Farmers Union Central Exchange, Inc. v. FERC, Nos. 82-2412 et al. (D.C.Cir.), FERC further refined its general policy on oil pipeline rates and identified TAPS rates as warranting special treatment, in significant part because of the possible impact of pipeline rate increases on Alaska's revenues. 21 FERC (CCH) at 61,600. FERC noted first that there is almost a total identity of interest between TAPS owners and shippers, and that practically all money paid to TAPS owners is a "wash." Id. Second, royalties and severance taxes due the State of Alaska are computed on the oil's value at the wellhead, which is approximately the world market price minus the cost of transportation. The higher the pipeline charge, the lower the state's revenues. The shippers therefore have an interest in paying themselves as much as possible for pipeline transit. "Every time they shift a dollar from one pocket to the other in order to pay themselves for shipping their own oil over their own pipeline they save 27.5 cents in royalties and severance taxes that would otherwise fall into Alaska's coffers." Id. at 61,691-92 n. 229.

As these cases make apparent, FERC heeded the admonition of Connecticut Light and began to formulate FERC suspension length policy on a case-by-case basis. In the meantime, this court revisited the issue of judicial review. In Delmarva Power & Light Co. v. FERC, 671 F.2d 587 (D.C.Cir.1982), we cut back on any broad reading of Connecticut Light's assertions of reviewability. This court held unreviewable a maximum-length FERC suspension, accompanied by reasons, against a challenge of illogic and inconsistency, explaining that "the length of a rate suspension is a decision utterly inappropriate for judicial review and manifestly within the agency's province." 671 F.2d at 594. Judge Wilkey, who also authored Connecticut Light, concluded:

It is one thing to require that a rationale be given [as Connecticut Light had required], it is something else to examine the merits of the rationale so expressed. At oral argument Delmarva claimed that the reasons given by FERC were illogical and inconsistent, and that giving illogical and inconsistent reasons was tantamount to giving no reasons at all. However, we cannot see how reviewing the consistency and logic of an order is anything but reviewing its merits.

Id. at 595. Delmarva is peppered with broad language about the unreviewability of all aspects of a suspension decision, including the reasons given for its length. In direct contrast to Connecticut Light, it implies that a court's review must halt if any reason at all is offered for the suspension and its length. One inference that might be drawn from Delmarva is that FERC may suspend for seven months on sunny days and one month on cloudy days and, offering those weather-related reasons alone, be immune from review. Indeed, that is one reading respondents apparently urge upon us. However, Delmarva nowhere articulates what reasons FERC had given in that case. The actual holding of the case therefore left unanswered precisely what degree of fact-specificity and rationality will suffice to preclude further review.

We today reject any suggestion of absolute unreviewability, as has another panel of this court in reconciling Connecticut Light and Delmarva. See Southern California Edison Co. v. FERC, 686 F.2d 43, 46 (D.C.Cir.1982) (review is permissible to assure reasons satisfy Connecticut Light; if they do, the suspension decision itself is unreviewable). We believe Connecticut Light was something of an aberration, and that it authorized a more probing review of the merits of a suspension decision than is appropriate. Realizing that it perhaps enabled reviewing courts to penetrate too far into the merits of FERC suspension decisions, we attempted in Delmarva to cut back on Connecticut Light. We did not intend in Delmarva, as we do not here, to say that the decision to suspend a rate for a given length of time is wholly unreviewable. Nonetheless, we emphasize that our rejection of that reading of Delmarva in no respect widens the narrow scope of review we carve out today.

The statute imposes on FERC an obligation to state reasons for a suspension and for the length of the suspension. We think it is nonsensical to construe this requirement as permitting FERC to state reasons that have nothing whatsoever to do either with its ultimate inquiry into the reasonableness of the rate or with its interim concern, expressed in Buckeye, for persons or entities irrevocably harmed by the institution of a new rate. We therefore hold that the reasons given must pass a minimal threshold test. They must in some way be related to FERC's interim or ultimate inquiries. We will not, however, take the next step and review the merits of a given case. For us to intrude at the suspension stage would disrupt the Commission's regulatory function, by forcing a consideration of the reasonableness of a proposed rate prior to a final FERC ruling on that very question.

It is clear that the requirement of a statement of reasons is far from a frivolous adornment to the statute. Such a requirement has a rationale of its own, quite separate from facilitating judicial review on the merits. This point was emphasized by the Supreme Court in another context, in Dunlop v. Bachowski, 421 U.S. 560, 95 S. Ct. 1851, 44 L. Ed. 2d 377 (1975) (considering the scope of review of Labor Secretary's decision not to bring a civil action to set aside a union election). In so doing, the Court noted that a "reasons" requirement promotes thought by the decisionmaker and "compels him to cover the relevant points and eschew irrelevancies." Id. at 572, 95 S.Ct. at 1860. Thus, a statement of reasons, "almost essential if there is to be judicial review, is desirable on many other grounds. The necessity for justification is a powerful preventive of wrong decisions.... A statement of reasons may even make a decision somewhat more acceptable to a losing claimant." Friendly, "Some Kind of Hearing," 123 U.PA.L.REV. 1267, 1292 (1975) (footnotes omitted). With these considerations in mind, we conclude, as did the Supreme Court in Dunlop, that "a statement of reasons must be adequate to enable the court to determine whether the [Commission's] decision was reached for an impermissible reason or for no reason at all." 421 U.S. at 573, 95 S.Ct. at 1860.

We note also that, in the particular context that we confront, a reasons requirement plays an additional role. In Connecticut Light, we explained that reasons provided by FERC give guidance to carriers. 627 F.2d at 471. They enable carriers to predict, on the basis of past explanations, how long a suspension they face, thereby permitting them to target an effective date for a new rate, and to file the new rate with the agency in time for it to become effective when needed. Particularly in light of these recognized practical and theoretical rationales for a reasons requirement, we conclude that a court's review should extend as far as--but no farther than--an examination of the reasons to assure they are in some way relevant to FERC's statutory inquiries.

To illustrate the limited scope of this review, we now consider the rate suspensions at issue in this case. In particular, we address petitioners' assertion that FERC's concern about the impact of a new rate on Alaska's citizens is not an acceptable reason to suspend the rate for the maximum period. We disagree. Petitioners apparently claim that the reason is not acceptable because it could not be determinative in FERC's ultimate decision on the reasonableness of the rate. That may be true. But we also know that FERC's consideration of the rate's impact on Alaska's coffers is not wholly irrelevant to FERC's ultimate decision.12 FERC has explained the relationship between that reason and its inquiry, see Order, supra note 4, 22 FERC (CCH) at 61,569, and we have no reason to doubt that explanation. The reason therefore suffices.

As we observed in Connecticut Light, and as we reaffirm today, FERC has broad authority to determine its policy on the length of suspensions either through rulemaking or case-by-case adjudication. As a result, in the vast majority of cases a reviewing court's inquiry will end if reasons given pass the threshold test of relatedness. For now, we can identify only two possible situations in which we believe an appellate court might appropriately remand to FERC for further articulation of reasons even after the reasons offered have passed the threshold test set out above.

First, if we were faced with a situation in which FERC had taken action that was patently arbitrary by imposing two different suspension lengths in cases that were absolutely indistinguishable, and if it had failed to offer even summary reasons to explain the difference, we believe the reasons would be susceptible to remand.13 We do not face such a situation here. For one, these cases are not indistinguishable from a third in which FERC decided differently. Petitioners no doubt would argue that Buckeye fits that bill precisely. However, Buckeye is distinguishable because there FERC did not address TAPS rates. As FERC made clear in its Williams decision, TAPS cases implicate different concerns from most oil pipeline cases and are treated differently. In light of that case, it cannot plausibly be argued that Buckeye is indistinguishable from the Exxon and BP actions.14

Moreover, even if the cases were sufficiently similar to trigger this second-tier analysis, FERC offered an adequate reason for its treatment of Exxon and BP. The Commission's reference to the possible detrimental effect of a rate increase on the citizens of Alaska--a concern not present in the Buckeye case--would have sufficed to end our inquiry.

The second situation warranting further consideration of reasons might occur if a FERC action was plainly and absolutely foreclosed by existing rules or past precedent. In that instance, it might be patently unfair to permit the suspension's length to stand. The facts of the instant case enable us to illustrate the slender opening this exception cuts in the Commission's discretion. Petitioners argue that Buckeye established a norm--one-day suspensions--and identified two possible exceptions to that norm. Thus, they conclude, FERC is bound to the norm unless one of the exceptions applies. Had FERC engaged in rulemaking, that conclusion might be accurate. However, Buckeye did not establish any FERC policy that was intended to be fixed in concrete. First Buckeye, then Williams, and now these cases sequentially formulate pipeline policies for Alaska. In this ongoing case-by-case process, our role surely is not to inhibit the evolution of agency policy as circumstances change and unforeseen issues arise.

Therefore, while it is true that Buckeye did not specifically forecast the result in this case, that fact is irrelevant. What is important is that Buckeye did not foreclose the Commission's action here. In other words, nothing in Buckeye promised that a one-day suspension would always be applied without regard to new or changed circumstances that might confront FERC in future cases.

Accordingly, having reviewed FERC's reasons within the narrow limits set out above, we affirm the two suspension orders before us.

So ordered.

J. SKELLY WRIGHT, Circuit Judge, concurring in the result:

I agree with the majority that our cases concerning the reviewability of FERC rate suspension orders are "somewhat obscure," majority opinion (maj. op.) at 1470, and I sympathize with the majority's desire to clarify the law in this area. I also agree--although perhaps a bit more guardedly--with the majority's result affirming the Commission. However, because I am afraid that today's opinion continues to leave the water somewhat murky, I write separately to suggest an alternative analysis that I believe would provide parties with more certain legal principles and would co-exist more peacefully with general principles of administrative law.

It is useful to begin with the principle that, absent a strong indication that Congress intended to preclude review of a particular type of administrative action, final administrative actions disposing of substantial private rights are generally subject to judicial review. This well-entrenched principle provides half of the foundation for the partnership between courts and agencies that is at the heart of administrative law. The other half of the foundation is the principle that courts will in general only review final agency action. Until an agency has actually done something that affects substantial private rights, a court will not generally interfere in the agency's decisionmaking process. A corollary is that a given agency action should be subject only once to judicial review, and that review is available only at an appropriate time.

I believe these simple principles provide the appropriate framework for analyzing the reviewability of FERC rate suspension orders. They counsel that the key considerations are whether rate suspension orders dispose of substantial private rights and whether rate suspension orders are final. To put the question schematically: If a rate suspension order is not a final agency action that disposes of substantial private rights, it is not reviewable by this court. If a rate suspension order is a final agency action that disposes of substantial private rights, it is reviewable by this court. In the latter case, of course, we will have to elaborate the standard by which the order is to be reviewed. But the first question is surely whether there has been final agency action affecting the substantial rights of private parties. To answer this question, we must examine the place of rate suspension orders in the statutory scheme governing utility rates.

B. The Statutory Scheme Governing Rate Suspensions

The majority accurately summarizes FERC's authority to order rate suspensions; I wish to emphasize only a few aspects of this regulatory scheme. See maj. op. at 1469-1470. Section 15(7) of the Interstate Commerce Act, 49 U.S.C. Sec. 15(7) (1976), governs the rate suspension order at issue in this case.1 When a utility2 files a new rate with FERC, the Commission may either accept or suspend that rate. If the Commission decides to accept the new rate, it goes into effect immediately. Nothing further occurs unless someone--usually a customer of the utility--challenges the rate as unreasonable3; in that case, the Commission will of course adjudicate the challenge and, if the customer successfully shows the new rate to be unreasonable, may order the utility to refund the difference between the new rate charged and the just and reasonable rate.

If the Commission suspends the rate, it may do so for a period of up to seven months. Usually, the Commission will order a suspension only if it has reason to believe that the new rate is not reasonable. If the Commission suspends the rate for any period of time, the old rate remains in effect for the period of the suspension. At the end of that period, the utility may begin charging the new rate, subject to the Commission's final adjudication of the new rate's reasonableness after a full administrative proceeding. A decision to suspend has three additional "procedural" effects. First, the decision to suspend commits the Commission to investigate the reasonableness of the new rate.4 Second, the decision places the burden of proof on the utility--rather than the customer--to demonstrate the reasonableness of the new rate. Third, the decision makes available to the customer a more generous measure of damages if the new rate is ultimately determined to be unreasonable. See note 5 infra.

If the Commission suspends the new rate and then, after full proceedings, determines that the new (presumably higher) rate is not just and reasonable, the Commission may order the utility to refund (with interest, see note 1 supra ) to its customers the difference between the just and reasonable rate and the new rate for the period of time between the end of the suspension and the Commission's final determination. If the Commission suspends the new rate but ultimately determines that the new rate is just and reasonable, the utility may not collect from its customers the difference between the old and the new rates for the period of suspension.

What remains is to apply general principles concerning judicial review of administrative action to the specific statutory regime governing rate suspension orders. My conclusion will be that decisions not to suspend and decisions to suspend for short periods are not final decisions that have a substantial impact on the rights of private parties. Decisions to suspend for lengthy periods, however, are final decisions that can have substantial impact on the rights of private parties. Therefore, decisions not to suspend rates and decisions to suspend rates for very short periods are not reviewable. Decisions to suspend rates for lengthy periods are reviewable. It is our own failure in some recent cases adequately to distinguish between these two kinds of decisions that has caused the unnecessary confusion that the majority and I both deplore. In my examination of the cases I will therefore emphasize the differences between the two kinds of cases; in particular, I hope to show that the cases holding failures to suspend and short suspensions unreviewable provide no support for the proposition that long suspensions are also unreviewable.

1. Decisions Not to Suspend. Although rate suspension orders are at issue in this case, it is useful to begin with the reviewability of Commission decisions not to suspend. It is unnecessary to spend a great deal of time on these decisions because, if there is one point on which the case law is clear, it is that a Commission decision not to suspend a new rate is not reviewable. The reason for the unreviewability is that the decision not to suspend is largely based on the Commission's necessarily rough judgments as to whether the rate is likely to be reasonable and whether the Commission should devote its enforcement efforts to investigating the particular rate in question. To be sure, the decision not to suspend does affect the burden of proof in any subsequent, customer-initiated proceedings on the merits of the new rate and also may affect the measure of damages the customer can obtain if the new rate is ultimately found to be unreasonable.5 But, even after a decision not to suspend, the customer retains the ability to initiate a proceeding to determine the reasonableness of the rates, at the conclusion of which (if the rate is determined to be unreasonable) the customer can get full compensation for any damages suffered. Therefore, realizing that a decision not to suspend has a major impact on the Commission's own docket, ordinarily will cause only minor (if any) injury to the utility's customer, and will cause no injury to the utility (which can put its new rate into effect immediately), courts have generally held that decisions not to suspend rates are within the Commission's discretion and therefore nonreviewable.

The leading case is Southern Railway Co. v. Seaboard Allied Mining Corp., 442 U.S. 444, 99 S. Ct. 2388, 60 L. Ed. 2d 1017 (1979), in which the Supreme Court held unreviewable an ICC decision6 not to suspend--and not to order a hearing to inquire into the lawfulness of--a seasonal rate increase proposed by a group of railroads. The Supreme Court explicitly adverted to the narrowness of its holding, see id. at 454, 99 S.Ct. at 2394 ("Initially, it is important to note the extremely limited scope of the administrative decision that we conclude is not judicially reviewable."), and based that holding on the fact that the Commission's failure to suspend would "not necessarily affect[ ] any citizen's ultimate rights." Id. at 455, 99 S.Ct. at 2394. The Court noted that a shipper aggrieved by a rate increase would still retain the right under Section 13(1) of the Interstate Commerce Act to institute a complaint against a railroad. The subsequent Section 13(1) proceeding would differ from an ICC-initiated proceeding only in that the burden of proof and measure of damages differed in Section 13(1) proceedings from those used in the ICC-initiated proceedings that follow a rate suspension. Since, despite these differences, Section 13(1) still would fully compensate shippers if any illegality in the rate increase were shown, the Court held that the Commission's decision was unreviewable. See Southern Railway, 442 U.S. at 454-455, 99 S. Ct. at 2394.

For present purposes, it is most important to note that the Court's reasoning was based on a pragmatic determination that no substantial rights of private parties would be affected by a Commission decision not to suspend. Given the Commission's discretion to control its own docket, see id. at 456-458, 99 S.Ct. at 2395-2396,7 and given the fact that ultimate review of the merits of the rates was assured, the Court thus used principles similar to those urged above to reach its holding that Commission decisions not to suspend are nonreviewable.

The other cases dealing with decisions not to suspend use similar reasoning to reach the same result. In United States v. SCRAP, 412 U.S. 669, 93 S. Ct. 2405, 37 L. Ed. 2d 254 (1973), for example, petitioners sought to have the District Court order the Commission to suspend new, higher rates. The Supreme Court held that such an injunction would "constitute[ ] a direct interference with the Commission's discretionary decision whether or not to suspend the rates." Id. at 692, 93 S.Ct. at 2418. This result is entirely in keeping with the analysis I propose because, as Southern Railway, supra, was later to hold, it was not a decision that finally disposed of substantial private rights. Once the Commission has decided that it will suspend rates, SCRAP has nothing to say about whether its decision to suspend for a lengthy period--a decision that can injure the substantial rights of private parties--is reviewable.

2. Short Suspensions. If the Commission's action has no substantial (and final) effect on private rights, it ought not be reviewable. In determining whether a short (i.e., one-day)8 suspension is reviewable, it is necessary to inquire whether a short suspension has a substantial effect on private rights. This question, in turn, must be analyzed in terms of the effect of a short suspension on the utility and the effect of a short suspension on other parties, usually the utility's customers.

As for the utility, its primary legitimate interest is in being able to receive the proceeds generated by reasonable rates. If its new rates are ultimately determined to be unreasonable, it has no right to the proceeds of the new rates and cannot suffer legally cognizable injury as the result of a suspension order. If, on the other hand, its new rates are ultimately determined to be reasonable, then it may incur the loss of revenues for the one-day period of the suspension. And this loss is final: as explained above, the utility's customers may not be forced to pay the utility the new rate for the suspension period, even if the new rate is ultimately determined to be perfectly reasonable.

However, even though the utility may suffer some final loss of revenues as the result of a one-day suspension, there are good reasons why we should not deem this loss to be substantial. First, it is a loss of a relatively insubstantial amount--the difference for only one day between the new and old rates--and even then it is a contingent loss that only is realized in cases in which the new rate is ultimately determined to be reasonable. Second, if the utility knows in advance that it is likely to face a one-day suspension, it is relatively easy for it to merely apply for the rates a day earlier than it would have, thus mitigating any loss it might suffer. Finally, the purpose of a one-day suspension is to activate the procedural effects of a suspension--shift of the burden of proof, commitment of agency resources to investigating the rate, etc. Therefore, because Congress must have intended that the Commission have the discretion to order a short suspension to activate these procedural features, and because the utility's contingent loss (if the new rates do turn out to be reasonable) is relatively small and can be mitigated,9 we are justified in concluding that, although a one-day suspension may have some effect on the utility, its rights are not substantially violated.

As for the utility's customer, if the utility's new rates are ultimately determined to be reasonable, he has lost nothing by the one-day suspension. And if the utility's rates are ultimately determined to be unreasonable, the statutory refund mechanism guarantees him that he will be refunded the difference between the new and old rates for the period during which the new rates were in effect. Therefore, no substantial rights of the customer will be finally disposed of by a short suspension.10 Because no substantial rights of either utility or customer are finally disposed of by a one-day suspension, I therefore conclude that short suspensions ought not be subject to judicial review.

In Municipal Light Boards v. FPC, 450 F.2d 1341 (D.C.Cir.1971), cert. denied, 405 U.S. 989, 92 S. Ct. 1251, 31 L. Ed. 2d 455 (1972), the leading case in this court often cited (incorrectly, I think) to show the unreviewability of all suspension orders, see, e.g., Delmarva Power & Light Co. v. FERC, 671 F.2d 587, 592 n. 28 (D.C.Cir.1982), this court held that a one-day suspension order was not reviewable.11 The court explicitly based its reasoning on the fact that the customers challenging the suspension had the right to receive refunds if the utility's rates ultimately were shown to be excessive. See 450 F.2d at 1351. As in the cases in which decisions not to suspend were held unreviewable, the court found that the existence of this "make whole" remedy after the ultimate decision on the reasonableness of the rates militated strongly against reviewability at this earlier stage. The court noted that the only meaningful distinctions between the failure-to-suspend cases and this one-day suspension were what I have called the "procedural" effects of the one-day suspension. The court reasoned that these procedural effects of the suspension had put the customer in a better position than "if the [Commission] had decided, in a plainly nonreviewable exercise of discretion, not to suspend [the utility's] filing at all." Id. The customer had no reason to complain of the one-day suspension, and it should thus be held unreviewable.

Like the failure-to-suspend cases, Municipal Light Boards provides no support for the proposition that lengthy suspensions are unreviewable. Instead, the case squarely stands for one of the propositions urged here: that one-day suspensions are not reviewable. And, given that the court emphasized the existence of an alternate, make-whole remedy for those damaged by a one-day suspension order, the opinion can fairly be read to imply that, in cases where no such make-whole remedy is available (i.e., when the suspension is lengthy), judicial review is appropriate.

Although Arrow Transportation Co. v. Southern R. Co., 372 U.S. 658, 83 S. Ct. 984, 10 L. Ed. 2d 52 (1963), technically did not involve a short suspension, it is useful to consider it in this category.12 The holding of the case was that the courts have no power to enjoin newly-filed rates-- i.e., to lengthen a Commission-ordered suspension or order a suspension where the Commission has not done so. This result provides strong support for my analysis, in that it virtually compels the result that short suspensions are unreviewable; since a utility customer's aim in going before the court is to seek a lengthening of the short suspension, forbidding the court to order such a lengthening is equivalent to forbidding the court to entertain the customer's claim. However, neither the result nor the reasoning in the case are relevant to the question whether a court may enjoin a suspension once the Commission has ordered it. Therefore, although (like Municipal Light Boards) we seem to have occasionally cited the case to show the unreviewability of all rate suspension orders,13 it actually supports only the narrower (and valid) proposition that short suspensions are unreviewable.

One additional case in this court, Papago Tribal Utilities Authority v. FERC, 628 F.2d 235 (D.C.Cir.), cert. denied, 449 U.S. 1061, 101 S. Ct. 784, 66 L. Ed. 2d 604 (1980), has some bearing on the question of the reviewability of short suspensions. Papago involved a challenge by a utility's customers to a FERC decision accepting and suspending for the maximum period14 a rate filing by an electric power wholesaler. However, the petitioners in Papago did not challenge the five-month suspension; they in fact benefitted from the Commission's decision to suspend for the maximum period. Instead, they claimed that the rate filing contained a number of patent errors and they therefore challenged FERC's acceptance of the rate filing, i.e., FERC's decision to let the rate go into effect at all, even at the end of the suspension period.

Analytically, this made the major issue in the case virtually identical to the failure-to-suspend and one-day suspension cases, and our analysis emphasized pragmatic factors in deciding that the Commission's order was unreviewable. These factors included the finality of the order, the irreparability of the injury, and the degree to which the decision is committed to agency discretion. See id. at 244.15 In particular the court emphasized that the Commission's acceptance of the rates was not a final decision on the merits of the rates because "the same issues now posed for summary disposition will also constitute the principal issues in the full administrative hearing." Id. at 240. It also emphasized that the decision did not dispose of substantial private rights, because the plaintiffs would be refunded any overcharges they would incur if the new rates were ultimately determined to be unreasonable.16 The court accordingly held that the Commission's decision to accept the rate filing, like the decisions not to suspend the filings in the above cases, was nonreviewable. Implicit in the court's reasoning is the notion that judicial review of the merits of the new rates would be more effective (and would intrude less on the agency) if it were delayed until the agency had a chance to make a final decision itself in the full administrative hearing. For our purposes, the important point once again is that the two factors emphasized in the Papago opinion--finality of the agency decision and harm to private parties--are just the two factors that militate in favor of judicial review of lengthy suspensions.

3. Long Suspensions. Unlike failures to suspend and short suspensions, long suspensions are fully reviewable. To understand the reasons why long suspensions are to be treated differently from failures to suspend and short suspensions, it is necessary to examine the effects under the current regulatory scheme of a long suspension. Thus, assume that the Commission suspends a new rate for a long period--say, the seven-month maximum permitted under Section 15(7). The question to ask, as above, is whether this suspension is a final decision likely to injure substantial private interests.

The customers of the utility cannot be injured by such a decision. If the new rates are ultimately deemed reasonable, the customer has not lost anything by paying the old (presumably lower) rates during the suspension period. If the new rates are ultimately deemed unreasonable, the customer has also not been injured by the decision to suspend rates (i.e., to charge the old, reasonable rates) for seven months.

The utility, however, is in a different position. If the Commission ultimately determines that the new rate is not just and reasonable, the decision will not have caused any legally cognizable injury to the utility because it has merely deprived the utility of its opportunity to collect its new (by hypothesis, unreasonable) rates for seven months. But if the Commission ultimately decides that the new rates are just and reasonable, the decision does affect substantial private rights. For during the seven-month period the utility has only been able to charge its old rate, and it therefore has lost the difference between the old rate and the new (by hypothesis, reasonable) rate for the seven-month period. Unlike the customer, who can get a refund from the utility if he has been charged a rate that is not just and reasonable, the utility cannot go back and charge the customer for its losses caused by the fact that for seven months it had to charge a rate lower than that determined to be just and reasonable. Moreover, the suspension decision in such a case is a final decision: it finally determines the rate for the seven-month period. Therefore, a long suspension can finally injure the substantial rights of the utility. Because such long suspensions can affect the substantial rights of private parties, they are subject to judicial review.

The respondent cites three cases from this court that deal with the reviewability of lengthy suspension orders: Southern California Edison Co. v. FERC, 686 F.2d 43 (D.C.Cir.1982); Delmarva Power & Light Co. v. FERC, supra, 671 F.2d 587; Connecticut Light & Power Co. v. FERC, 627 F.2d 467 (D.C.Cir.1980). These are all relatively recent cases and, as the majority points out, the apparent conflict between them (and especially between Connecticut Light & Power and Delmarva Power & Light ) has been largely responsible for the unsatisfactory state of our jurisprudence in this area. See maj. op. at 1470. However, although the language of the cases is (to put it charitably) incautious, none of them stands in the way of the reviewability of lengthy suspension orders.

In Connecticut Light & Power the court reviewed the Commission's suspension of a new utility rate for the five-month maximum period permitted by statute. The court held that FERC had acted arbitrarily and capriciously, offering the same boilerplate reasons to justify suspensions of varying length. See 627 F.2d at 471-472. The court therefore vacated the suspension and remanded the case for further consideration by the Commission.

Regardless whether the court's action is characterized as a review of the reasons given for the length of the suspension or as a review of the length of the suspension itself, the fact remains that the Connecticut Light & Power court engaged in precisely the appropriate inquiry. It refused to consider the Commission's determination that the rates may be unjust and unreasonable, because it held that this determination was a mere preliminary to full-scale inquiry into the merits of the rate increase. It was therefore not final, and not subject to judicial review at this early stage. In effect, the court refused to review the Commission's decision, insofar as it was merely a decision (within the Commission's discretion) whether to suspend the rates at all.

On the other hand, the court did review the Commission's decision that the suspension should run for five months.17 Such review is appropriate precisely because, given the statutory scheme, the Commission's order was final with respect to the rates charged during the suspension period and because the regulated entity could suffer substantial (and irreversible) injury if it was not allowed to charge what turned out to be a reasonable rate during this period.

What Connecticut Light & Power hath given, Delmarva hath (seemingly) taken away. The Delmarva court refused to engage in a review of the "consistency or logic" of the Commission's decision to suspend a rate filing for five months. See 671 F.2d at 595. Instead, the court interpreted Connecticut Light & Power to hold that judicial review of suspension orders was for the sole purpose of assuring that the Commission stated a reason (seemingly any reason) for its order.18 Respondent here argues that, under the reasoning of Delmarva, if the court is satisfied that the Commission has stated some reason for its order, the review is at an end and the order must be affirmed.

To the extent that Delmarva represents a repudiation of Connecticut Light & Power and our other earlier cases, there is no persuasive reason why we should follow it. However, I agree with the majority that this extreme interpretation of Delmarva is unnecessary. See maj. op. at ---- - ----. The Delmarva opinion, despite its broad language, in fact hints that it was considerably influenced by its determination that the suspension order was on the merits a consistent and logical application of Commission policy. See 671 F.2d at 595 ("doubt exists that the Commission's actions lack consistency and logic"). Moreover, the scope of review that the Delmarva court envisioned is unclear, given that the opinion does not articulate what reasons had been given for the suspension. Finally, nowhere in Delmarva does the court explain why Commission decisions to order lengthy suspensions should be entirely unreviewable.19

Southern California Edison points the way, albeit tentatively, toward the correct resolution of the reviewability issue. In that case this court once again faced a challenge to a five-month maximum suspension period. The court perceptively noted that "once the reasons behind imposing a rate suspension of a given length have been articulated to the satisfaction of Connecticut Light & Power, the suspension decision itself is 'utterly inappropriate for judicial review.' " 686 F.2d at 46 (quoting Delmarva, 671 F.2d at 595). This is very close to the analysis that I suggest here: the decision whether or not to suspend rates (i.e., "the suspension decision itself") is not reviewable; the decision to suspend rates for a lengthy time (i.e., "the reasons behind imposing a rate suspension of a given length") is reviewable.20 Moreover, the Southern California Edison court's actions speak as loudly as its words, for the court in fact affirmed the five-month suspension only after satisfying itself that the suspension was consistent with Commission policy as articulated on three previous occasions on which the Commission had "scrupulously adhered" to the requirements of Connecticut Light & Power. See Southern California Edison, 686 F.2d at 45. Thus, far from standing in the way of a rationalization of our reviewability holdings, the Southern California Edison case in fact points in precisely the right direction.

To summarize: Short suspensions and decisions not to suspend are not reviewable; long suspensions are reviewable. The reasoning leading to this result is based on the principle that final agency action affecting substantial private rights is reviewable. Under the regulatory scheme governing utility rates, suspensions represent at most final dispositions of the difference between new and old rates for the period of the suspension; they have this effect only when the rates are ultimately determined to be just and reasonable, in which case the statutory scheme operates to deprive the utility of the benefits of its just and reasonable rate for the suspension period. The Commission's action in so depriving the utility should be subject to judicial review.

What has occurred in the recent cases is that we have failed to understand that the same principles that preclude judicial review of failures to suspend and short suspensions mandate judicial review of lengthy suspensions. Although the decisions are sound, they betray a tendency to use language that converts our earlier understanding that decisions whether to suspend are unreviewable into a blanket assertion that decisions relating to suspension are unreviewable. See, e.g., Papago, supra, 628 F.2d at 242 n. 19 (alleging that this court has held, "on the strength of Arrow Transportation, that we have no power to review orders concerning suspension of rates") (emphasis added); Delmarva, supra, 671 F.2d at 594 ("length of a rate suspension is a decision utterly inappropriate for judicial review," citing Papago language quoted above); Connecticut Light & Power, supra, 627 F.2d at 471 ("decision to suspend * * * rates for five months was committed to the Commission's discretion and is not reviewable by us"). It is this tendency that has made our jurisprudence in this area so difficult.

I conclude that this court has never squarely held that long suspensions are unreviewable. This is in part because this court has never explicitly addressed the differences between long and short suspensions under the current statutory scheme. Nonetheless, the reasoning used in most or all of the cases suggests strongly that Commission decisions that are final and can injure the substantial rights of private parties should be subject to review; because Commission decisions to suspend new rates for lengthy periods are final (as to the period of suspension) and can injure the substantial rights of parties (if the new rates are in fact reasonable), the precedents militate strongly in favor of review of such long suspensions.

Under the analysis suggested above, FERC's suspension order in this case is reviewable, because it orders the suspension for a long period of time. It remains to determine the scope of review and the standard to be used in reviewing the Commission's action in this case.

A lengthy disquisition on the standard of review is not necessary, in part because many of the principles that govern our review can be derived from the discussion above and in part because review of this kind of order shares most of its essential features with our review of other kinds of agency actions. In general, we should make use of the "arbitrary and capricious" standard of review that is embedded in the Administrative Procedure Act and in so much of our jurisprudence. See 5 U.S.C. Sec. 706(2)(A) (1982).

Our application of the "arbitrary and capricious" standard in rate suspension cases should concentrate on the consistency and predictability of the Commission's action. To the extent that the Commission's decision to suspend for a lengthy time is a final decision, it is at most a final decision as to the rate to be charged during the suspension. If regulated entities understand the rules that will govern suspensions, they may quite easily see to it that they are not injured by rate suspensions. For if a utility knows that its application for new rates is likely to be suspended for seven months, it can save itself considerable damage by applying seven months in advance for a new rate. On the other hand, if a utility knows that its application is likely to be suspended for a single day, it can wait longer before submitting that application to the Commission. Thus the primary interest to be protected by judicial review of lengthy rate suspension orders is the regulated entity's interest in predictability and certainty; if this interest is protected, the utility can protect itself from the possible injury that a rate suspension could cause.

In Connecticut Light & Power this court balked at the Commission's failure to be consistent or predictable in the length of its rate suspension orders. We noted that "[t]he Commission should have enunciated standards for rate suspension periods through rulemaking or adjudication years ago," and concluded that "[t]he Commission, therefore, must either engage in rulemaking to establish standards for its exercise of discretion, or it must in individual cases spell out its rational [sic ] briefly, and over a period of years build a set of standards by precedent to assure that its discretion is not exercised in an arbitrary way and to give guidance to parties filing and challenging rate increases." 627 F.2d at 473. These observations precisely portray the proper scope of review in rate suspension cases. The Commission has great discretion to adopt a reasonable policy to govern suspension decisions; our task is to see to it that it adopts some policy and applies it with reasonable consistency. It is because this consistency is of great importance to the parties that Congress required a statement of reasons to accompany Commission suspension orders. See 49 U.S.C. Sec. 15(7) (quoted in note 1 supra ). in this sense, although I believe that review is authorized not by the "statement of reasons" requirement of Section 15(7), but by Section 10 of the APA, Connecticut Light & Power's reliance on the "statement of reasons" requirement to justify its review was not misplaced.

It is worth noting that, in accord with both the Commission's practice and sound principles of administrative law, our review will have little to do with the merits of the rate increase itself. Cf. Connecticut Light & Power, 627 F.2d at 472 ("[T]he reasons why a rate is or is not valid ultimately and the data behind such determinations are quite different from the factors involved in the Commission's choice to suspend for one day or five months.") (emphasis in original). Because the very purpose of a rate suspension is to enable the Commission to investigate the reasonableness of the proffered rate, the Commission should have maximum discretion in making its initial, necessarily rough, determination of reasonableness. Judicial review of the reasonableness of rates will have to come after--not before--the Commission has had the opportunity to express its own final judgment on the question pursuant to a full administrative adjudication. Therefore--and this is the kernel of truth in the cases concerning reviewability of rate suspensions--the Commission's preliminary determination that a given new rate is or is not likely to be reasonable is not subject to judicial review. The failure-to-suspend and short suspension cases discussed above effectively stand for this proposition.

Our recognition of the Commission's discretion in making its initial determination of reasonableness does not, however, leave us with no significant role to play. For the Commission does not decide to grant a lengthy suspension in a given case based on the likelihood that the particular new rate in question is reasonable. See maj. op. at 1470 n. 8. Instead, in response to our Connecticut Light & Power decision, the Commission has set out a number of policies governing the length of rate suspensions. See maj. op. at 1471-1472. In Buckeye Pipe Line Co., 13 FERC (CCH) p 61,267 (Dec. 24, 1980), the Commission pointed out that while electric and gas utility rates have a great and virtually immediate impact on consumers, oil pipeline rates generally have minimal impact on the consumer. Therefore, in order to accord consumers maximum protection, the Commission adopted a policy of suspending electric and gas rates for the maximum period permitted by statute. At the same time, the Commission has seen little need to afford extra protection to either oil pipeline companies or the giant oil companies that are their customers. Furthermore, the Commission is fully aware that, under the current regulatory scheme, the existence of the refund mechanism (available only to customers against utilities and not vice versa, see p. 1477 supra) guarantees that, if short suspensions are ordered, the ultimate rate charged will be the reasonable rate approved by the Commission. Therefore, it has adopted a policy of one-day suspensions in oil pipeline cases. See maj. op. at 1478-1480. Our review in a given case should focus on the Commission's consistency and predictability in applying these policies, none of which have anything to do with the ultimate merits of a given rate increase.21

If this case, like Southern California Edison, had merely presented a routine application of these policies, our review could have been rather brief. The policies themselves seem to be reasonable and to lie well within the Commission's discretion. Having determined that the policies were legitimate, we would have had only to inquire whether the policies had been applied consistently; in an oil pipeline case, the question would be merely whether the case was a genuine oil pipeline case and, if so, whether the Commission had ordered a one-day suspension.

Unfortunately, this case is considerably more difficult than a routine application of the above policies. I can not accept the majority's suggestion that "relatedness" (whatever that is) is the determining factor in our review. Nor do I believe that the only cases appropriate for remand would be the kinds of cases described at page 1482 of its opinion.22 Thus, although I reach the same conclusion as the majority, I find the case to be considerably more troublesome. For petitioners argue that the Commission's decision here must be reversed precisely because the Commission in this case did not follow its own established policies: this is an oil pipeline case, yet the Commission suspended the new rates for seven months. Given the purposes of judicial review in rate suspension cases as outlined above, this is the contention that should trigger a fairly close look at the Commission's action.

Nonetheless, I agree with most of the majority's analysis of the particular order in question here. The Commission's policy of suspending oil pipeline rates for only a day was announced in a previous adjudication, and the announcement of that policy does not prevent the Commission from modifying it as necessary in future cases. Although consistency and predictability are at a premium in this area, the Commission certainly may adopt new policies in areas where the presuppositions or goals of its one-day oil pipeline suspension policy are not fully applicable. This seems to be just such a case. While other cases may involve only the interests of consumers, utilities, and their customers, this case implicates the interests of the State of Alaska in its substantial share of the pipeline revenues. See Williams Pipe Line Co., 21 FERC (CCH) p 61,260 (Nov. 30, 1982), at 61,692 n. 229, appeal pending sub nom. Farmers Union v. FERC, D.C.Cir. No. 82-2412 et al. Moreover, FERC has noted that, unlike the situation in the rest of the country, the largest customers of the Alaska pipeline companies are the oil companies that own the pipeline companies. This incestuous relationship may create a special incentive for FERC to be solicitous of interests (like those of the State of Alaska) that may be inadequately protected otherwise. See maj. op. at 1472. The brief justification the Commission has offered in this case, see maj. op. at 1469; 22 FERC (CCH) p 61,327 (Mar. 18, 1983), at 61,569, in the context of the Commission's policy of treating Alaska pipelines in general separately, is sufficient to justify the Commission's action. I therefore conclude that the Commission's order here, although troublesome, ought to be affirmed.

I offer the analysis in this opinion as an attempt to dispel some of the confusion that our cases have created concerning review of rate suspension orders. Because nonreviewability doctrines have as their function to keep courts from premature interference in agency decisionmaking, such doctrines should be clear, simple, and easy to use. The rules for reviewability I have suggested have this feature: short suspensions and decisions not to suspend are not reviewable, long suspensions are reviewable.

I would like to urge one further advantage to the suggested analysis. When courts believe that an agency is illegitimately injuring the substantial rights of a private party, they will find it extremely difficult to apply a doctrine of nonreviewability whose effect would be to shield the agency's action from their scrutiny. As a result, the nonreviewability doctrine becomes riddled with loopholes and the parties and court expend a great deal of time and effort discussing reviewability, all the while causing just the intrusion on agency decisionmaking that the nonreviewability doctrine was intended to preclude.

The solution to the above dilemma is not to become absolutely rigid in applying nonreviewability doctrines, thus ignoring arbitrary and capricious agency action. Instead, the solution is to be certain that agency actions that can cause substantial injury to private parties are subject to review, and that agency actions that are non-final and cause no substantial injury remain nonreviewable. It is in this spirit that I offer this opinion.