POSNER, Circuit Judge.
Field Container Corporation, a shipper, complained to the Interstate Commerce Commission that the C & NW railroad had unreasonably assessed some $19,000 in demurrage against it. The Commission dismissed the complaint and ordered Field to pay the assessment, and Field has petitioned us to set aside the Commission’s decision. “Demurrage” in railroad parlance is the charge for a shipper’s holding the railroad’s cars for loading or unloading beyond a specified time, thereby depriving the railroad of their use; on the history of railroad demurrage see Chrysler Corp. v. New York Central R.R, 234 I.C.C. 755, 759-61 (1939). Field’s petition for review raises interesting questions with respect to the form of demurrage known as the “average agreement,’’ but before we can reach them we must satisfy ourselves that we have jurisdiction.
Section 2321(a) of the Judicial Code gives the courts of appeals jurisdiction, made exclusive by section 2342(5), to review final orders of the ICC, except where Congress has otherwise provided, as it has done in section 1336(a) by giving the district courts jurisdiction to enforce or enjoin “any order of the Interstate Commerce Commission for the payment of money or the collection of fines, penalties, and forfeitures.” So if this is a proceeding to enjoin an order to pay money, we have no jurisdiction.
The divided review of ICC orders is a vestige of legal evolution — about as useful, and as mischievous, as the human appendix. It goes back to 1906, when judicial hostility to the Commission, among other factors, led Congress in the Hepburn Act, 34 Stat. 584, to overhaul the procedure for judicial enforcement and review of the Commission’s orders. See 1 Sharfman, Interstate Commerce Commission 24r-25, 46-48 (1931). Section 5 of the Hepburn Act distinguished between judicial review of ICC orders for the payment of money and judicial review of other *ICC orders, and as to the latter created a right of direct appeal to the Supreme Court from the initial review courts, which were the circuit courts (not to be confused with the courts of appeals, then called circuit courts of appeals). 34 Stat. 591-92. Continued dissatisfaction with judicial attitudes toward the Commission led Congress a few years later to shift judicial review of ICC orders other than for the payment of money to a new court, the Commerce Court, a short-lived experiment in specialized federal appellate courts. See Frankfurter & Landis, The Business of the Supreme Court 153-74 (1928). Review of orders for the payment of money remained in the circuit courts. In 1913, amidst a chorus of complaints about the Commerce Court, see 50 Cong.Rec. 4531, 4537, 4540-41 (1913), Congress abolished it and transferred its jurisdiction to three-judge district courts, 38 Stat. 219; see United States v. ICC, 337 U.S. 426, 442, 69 S.Ct. 1410, 1419, 93 L.Ed. 1451 (1949), but without altering the jurisdiction of single-judge district courts — which had succeeded to the jurisdiction of the circuit courts upon their abolition two years before, 36 Stat. 1087 — to review orders for the payment of money.
Given the judicial hostility in this period to regulatory agencies such as the ICC, and a bearable caseload in the Supreme Court, the dual system of judicial review made a certain amount of sense. Orders that might have national impact received initial scrutiny by three judges with a right of direct appeal to the Supreme Court. But since invalidating an order limited to the payment of money had little potential to disrupt the regulatory system, such orders were reviewed by a single district judge with a right of appeal only to the court of appeals, though the Supreme Court could of course review on certiorari. However, as the Supreme Court’s caseload waxed and judicial hostility to regulatory agencies waned, the extraordinary review procedure for nonpayment orders became an ana[254]*254chronism and it was finally abolished in 1975. 88 Stat. 1917. But instead of making all ICC orders reviewable in the standard modern fashion — that is, in the courts of appeals — Congress, without explanation, gave those courts only the review jurisdiction that the three-judge district courts had had, and left jurisdiction to review orders for the payment of money in the district courts. As a result, the only difference in judicial review of the two types of order is that the less important, the order for the payment of money, gets an extra tier of review — review by a federal district court, in addition to review by the court of appeals, and by the Supreme Court on certiorari.
When an administrative order is based on a record, as ICC orders for the payment of money are, the reviewing court has no factfinding function, so there is no point in enlisting the federal district courts — trial rather than review courts — to provide an initial tier of judicial review. Denberg v. United States Railroad Retirement Bd., 696 F.2d 1193,1196 (7th Cir.1983). And it is worse than pointless to do so with respect to only some of an agency’s orders, so that proper classification becomes a litigable issue, and a party’s mistake in classification can result in his losing his right to judicial review. But that is the system we have for ICC orders, and until Congress changes it we must make what sense we can out of it.
Although Field’s complaint challenged the reasonableness of a demurrage assessment that it had not yet paid, the C & NW intervened and got the Commission to order Field to pay, and if that order, a payment order, were the only order Field was asking us to set aside, we would have to dismiss its petition for review. In Pullman-Standard v. ICC, 705 F.2d 875, 879-80 (7th Cir.1983), we joined the District of Columbia Circuit in holding that the form of the order determines whether the district court or the court of appeals has jurisdiction to review it. See Consolidated Rail Corp. v. ICC, 685 F.2d 687, 694 (D.C.Cir. 1982). Other courts have held that not form but basis and potential impact determine jurisdiction. The Third Circuit held recently that an order denying a shipper’s complaint for remission of demurrage was reviewable directly in the court of appeals because the petition involved “a challenge to the legal bases of the Commission’s action .... ” Empire-Detroit Steel Division of Cyclops Corp. v. ICC, 659 F.2d 396, 397 (3d Cir.1981). But this is one of the decisions that Pullman-Standard rejected, see 705 F.2d at 878, in adopting the formal approach. Although it would be better if the courts of appeals had exclusive review jurisdiction over all ICC orders, Congress has divided jurisdiction between the courts of appeals and the district courts, and the division should be clearly marked. This requires that all orders to pay, the consequential and the inconsequential alike, be reviewable only in the district courts.
But as it is well settled that review of both a money and a nonmoney order must be sought in the court of appeals (Pullman-Standard was such a case), we must consider whether there is anything else that Field is seeking review of besides the order to pay demurrage. Its petition for review just asks us to set aside the Commission’s decision, which dismissed Field’s complaint.
Free access — add to your briefcase to read the full text and ask questions with AI
POSNER, Circuit Judge.
Field Container Corporation, a shipper, complained to the Interstate Commerce Commission that the C & NW railroad had unreasonably assessed some $19,000 in demurrage against it. The Commission dismissed the complaint and ordered Field to pay the assessment, and Field has petitioned us to set aside the Commission’s decision. “Demurrage” in railroad parlance is the charge for a shipper’s holding the railroad’s cars for loading or unloading beyond a specified time, thereby depriving the railroad of their use; on the history of railroad demurrage see Chrysler Corp. v. New York Central R.R, 234 I.C.C. 755, 759-61 (1939). Field’s petition for review raises interesting questions with respect to the form of demurrage known as the “average agreement,’’ but before we can reach them we must satisfy ourselves that we have jurisdiction.
Section 2321(a) of the Judicial Code gives the courts of appeals jurisdiction, made exclusive by section 2342(5), to review final orders of the ICC, except where Congress has otherwise provided, as it has done in section 1336(a) by giving the district courts jurisdiction to enforce or enjoin “any order of the Interstate Commerce Commission for the payment of money or the collection of fines, penalties, and forfeitures.” So if this is a proceeding to enjoin an order to pay money, we have no jurisdiction.
The divided review of ICC orders is a vestige of legal evolution — about as useful, and as mischievous, as the human appendix. It goes back to 1906, when judicial hostility to the Commission, among other factors, led Congress in the Hepburn Act, 34 Stat. 584, to overhaul the procedure for judicial enforcement and review of the Commission’s orders. See 1 Sharfman, Interstate Commerce Commission 24r-25, 46-48 (1931). Section 5 of the Hepburn Act distinguished between judicial review of ICC orders for the payment of money and judicial review of other *ICC orders, and as to the latter created a right of direct appeal to the Supreme Court from the initial review courts, which were the circuit courts (not to be confused with the courts of appeals, then called circuit courts of appeals). 34 Stat. 591-92. Continued dissatisfaction with judicial attitudes toward the Commission led Congress a few years later to shift judicial review of ICC orders other than for the payment of money to a new court, the Commerce Court, a short-lived experiment in specialized federal appellate courts. See Frankfurter & Landis, The Business of the Supreme Court 153-74 (1928). Review of orders for the payment of money remained in the circuit courts. In 1913, amidst a chorus of complaints about the Commerce Court, see 50 Cong.Rec. 4531, 4537, 4540-41 (1913), Congress abolished it and transferred its jurisdiction to three-judge district courts, 38 Stat. 219; see United States v. ICC, 337 U.S. 426, 442, 69 S.Ct. 1410, 1419, 93 L.Ed. 1451 (1949), but without altering the jurisdiction of single-judge district courts — which had succeeded to the jurisdiction of the circuit courts upon their abolition two years before, 36 Stat. 1087 — to review orders for the payment of money.
Given the judicial hostility in this period to regulatory agencies such as the ICC, and a bearable caseload in the Supreme Court, the dual system of judicial review made a certain amount of sense. Orders that might have national impact received initial scrutiny by three judges with a right of direct appeal to the Supreme Court. But since invalidating an order limited to the payment of money had little potential to disrupt the regulatory system, such orders were reviewed by a single district judge with a right of appeal only to the court of appeals, though the Supreme Court could of course review on certiorari. However, as the Supreme Court’s caseload waxed and judicial hostility to regulatory agencies waned, the extraordinary review procedure for nonpayment orders became an ana[254]*254chronism and it was finally abolished in 1975. 88 Stat. 1917. But instead of making all ICC orders reviewable in the standard modern fashion — that is, in the courts of appeals — Congress, without explanation, gave those courts only the review jurisdiction that the three-judge district courts had had, and left jurisdiction to review orders for the payment of money in the district courts. As a result, the only difference in judicial review of the two types of order is that the less important, the order for the payment of money, gets an extra tier of review — review by a federal district court, in addition to review by the court of appeals, and by the Supreme Court on certiorari.
When an administrative order is based on a record, as ICC orders for the payment of money are, the reviewing court has no factfinding function, so there is no point in enlisting the federal district courts — trial rather than review courts — to provide an initial tier of judicial review. Denberg v. United States Railroad Retirement Bd., 696 F.2d 1193,1196 (7th Cir.1983). And it is worse than pointless to do so with respect to only some of an agency’s orders, so that proper classification becomes a litigable issue, and a party’s mistake in classification can result in his losing his right to judicial review. But that is the system we have for ICC orders, and until Congress changes it we must make what sense we can out of it.
Although Field’s complaint challenged the reasonableness of a demurrage assessment that it had not yet paid, the C & NW intervened and got the Commission to order Field to pay, and if that order, a payment order, were the only order Field was asking us to set aside, we would have to dismiss its petition for review. In Pullman-Standard v. ICC, 705 F.2d 875, 879-80 (7th Cir.1983), we joined the District of Columbia Circuit in holding that the form of the order determines whether the district court or the court of appeals has jurisdiction to review it. See Consolidated Rail Corp. v. ICC, 685 F.2d 687, 694 (D.C.Cir. 1982). Other courts have held that not form but basis and potential impact determine jurisdiction. The Third Circuit held recently that an order denying a shipper’s complaint for remission of demurrage was reviewable directly in the court of appeals because the petition involved “a challenge to the legal bases of the Commission’s action .... ” Empire-Detroit Steel Division of Cyclops Corp. v. ICC, 659 F.2d 396, 397 (3d Cir.1981). But this is one of the decisions that Pullman-Standard rejected, see 705 F.2d at 878, in adopting the formal approach. Although it would be better if the courts of appeals had exclusive review jurisdiction over all ICC orders, Congress has divided jurisdiction between the courts of appeals and the district courts, and the division should be clearly marked. This requires that all orders to pay, the consequential and the inconsequential alike, be reviewable only in the district courts.
But as it is well settled that review of both a money and a nonmoney order must be sought in the court of appeals (Pullman-Standard was such a case), we must consider whether there is anything else that Field is seeking review of besides the order to pay demurrage. Its petition for review just asks us to set aside the Commission’s decision, which dismissed Field’s complaint. If Field had paid the demurrage and all its complaint asked was that the railroad be ordered to repay it, the Commission’s decision would be reviewable in the district court because an order denying reparations, although it is not literally an order for the payment of money, is treated as one for purposes of determining which court has jurisdiction. United States v. ICC, supra, 337 U.S. at 441, 69 S.Ct. at 1418. But Field had not paid the demur-rage, and therefore was not asking for its return. Its complaint asked for an order “commanding [C & NW] to cease and desist from the aforesaid violations of said act, ... [and] directing] the waiver [in whole or part] of the sum of demurrage charges demanded of complainant by [C & NW], ... and that such other and further order or orders be made as the Commission may consider proper in the premises.”
[255]*255This language suggests two grounds for treating the Commission’s order dismissing the complaint as more than a payment order. First, in asking that the railroad be forbidden to collect the demur-rage that it had assessed, Field was asking for an injunction against the railroad, and an injunction is not a payment order. This may seem an unduly technical argument. The nature of the dispute is the same whether the railroad asks for reparations first or the shipper asks the Commission first to decide whether reparations are due. By suing first Field precipitated the railroad’s suit to collect demurrage and was therefore in effect bringing a declaratory judgment action, though not called that under the Commission’s rules. A declaratory judgment, like an injunction, is equitable; but where the only dispute between two parties is over whether a particular sum of money is due from one to the other, it might seem that the spirit of section 1336(a), if it has a spirit, would require that review jurisdiction be in the district court. Otherwise there would be the same kind of arbitrary distinction between court of appeals and district court jurisdiction that led the Supreme Court in United States v. ICC, supra, to equate denials of reparations with reparation orders. But that case was decided at a time when the existence of a cumbersome procedure for the review of non-money orders provided a reason for interpreting the predecessor of section 1336(a) broadly. See 337 U.S. at 443, 69 S.Ct. at 1419. The abolition of that procedure argues for interpreting section 1336(a) narrowly, in accordance with the usual presumption that direct review of administrative action belongs in the court of appeals unless there is a factfinding function to be performed by the reviewing court.
An alternative and perhaps solider basis for our jurisdiction is that although in asking that the C & NW be ordered to cease and desist from its alleged violations Field may have meant no more than that it wanted the railroad ordered to stop dunning it for the $19,000, its complaint did allege “that the exaction of demurrage due to a condition not ‘attributable to consignor or consignee’ is unreasonable ...,” which suggests that in asking for a cease and desist order Field may have wanted the average agreement on demur-rage reformed for the future. If so, it was asking for the partial cancellation of a tariff, as in Pullman-Standard, supra, 705 F.2d at 880, and we would have jurisdiction. Although it is not clear that this is what Field was asking for, the policy of routing review proceedings to the courts of appeals directly where the statutes will allow this entitles us to resolve our doubts in favor of jurisdiction — and come at last to the merits.
Demurrage is assessed according to one of two nationwide tariffs. The first provides for “straight demurrage,” and if the shipper and railroad do not provide otherwise they are automatically governed by it. Under straight demurrage, for every day the shipper holds back a car beyond the period (usually two days) that he is allowed to keep it for loading he must pay the railroad $10, but after four days this rises to $20 and after six days to $30, the amount over $10 per day being referred to as a penalty. The shipper gets no credit for returning cars early but on the other hand is not assessed demurrage if severe weather or other circumstances beyond his control prevent him from returning the cars on time. The other type of demurrage is the “average agreement,” under which the shipper gets a $10 credit against demurrage for every car that he returns within 24 hours of receipt but loses the excuses allowed under straight demurrage except for delay caused by floods, earthquakes, tornadoes, or hurricanes. A railroad cannot force a shipper into the average agreement — a point that gives the petitioner’s argument that the average agreement is a “contract of adhesion” a hollow ring. Not only must the shipper agree to join the average agreement but he can on short notice unilaterally return to straight demurrage.
The average agreement has been held to be reasonable, see, e.g., Monongahela Power Co. v. ICC, 640 F.2d 504 (4th [256]*256Cir.1981), despite its narrowing of the shipper’s excuses; and the petitioner’s suggestion that it is somehow “anti-shipper” borders on the frivolous. Since the shipper can always stay with or return to straight demurrage, the average agreement simply enlarges the shipper’s options — and you cannot be harmed by being given additional choices. Every contract allocates risks; the average agreement simply allocates them differently from straight demurrage. In straight demurrage, unexpectedly good weather is a boon to the railroad, because it gets its cars back faster without having to pay anything for this benefit, but unexpectedly bad weather hurts it. Under the average agreement.unexpectedly good weather is a boon to the shipper because he gets demurrage credits for prompt return of cars but unexpectedly bad weather hurts him because he has to pay demurrage even when the weather prevents him from returning the cars within the specified period. The shipper who is pessimistic about the weather or doubtful of his ability to cope with bad weather will choose straight demurrage, and the shipper who is optimistic about the weather or about his ability to cope with bad weather will choose the average agreement. But no shipper is coerced to adopt the average agreement; it is his choice, and having chosen he must take the bad with the good. It was Field’s misfortune that while operating under the average agreement — as it had been doing happily for 10 years — it was struck by the worst Chicago winter in memory, the winter of 1978-1979, and it was then that the demur-rage in question accrued. Had Field stayed with straight demurrage the risk of unusually bad weather would have remained on the C & NW.
Besides challenging the principle of the average agreement, Field makes two narrower arguments. The first is that the C & NW never delivered the cars on which demurrage was assessed, and could not have done so, because they were frozen into C & NW’s yard. But the fact that the cars never left the C & NW’s yard is not important by itself. The average agreement provides for what is called “constructive placement,” which means that for purposes of assessing demurrage the cars are deemed delivered when they are ready to be switched onto the shipper’s loading dock.' The reason for this rule is plain enough. The shipper cannot be allowed to defeat the railroad’s right to demurrage by refusing to accept delivery of cars after he has ordered them, and it would be wasteful to require the railroad, as a condition of preserving its demurrage rights, to trundle the cars up to the shipper’s loading dock only to be told to take them back to the yard.
Field argues, however, that the principle of constructive placement is inapplicable if though the cars have arrived at the yard the railroad cannot deliver them to the shipper’s loading dock. Field submitted an affidavit which stated that on the days on which it was assessed demurrage the cars were frozen in at the C & NW’s yard and could not have been switched onto Field’s siding even though, according to the affidavit, Field had cleared the siding of ice and snow and wanted and could receive cars. These allegations were not specifically controverted and the Commission made no finding that they were untrue. The Commission also accepted Field’s contention that if weather prevents the carrier from delivering rather than the shipper from receiving cars the average agreement does not allow demurrage to be assessed. But the Commission noted that under the practice long followed by the parties the C & NW’s duty to deliver cars, once it informed Field that the cars destined for Field had arrived at the local C & NW yard, was triggered by Field’s calling the yard and requesting delivery of the cars, and that on the days in question no calls were made. In the Commission’s view the failure to call barred Field from claiming that the cars had not been constructively placed because of the railroad’s alleged inability to deliver the cars on those days.
At first glance this ground for awarding demurrage seems the very pinnacle of technicality. The railroad could not have been hurt by Field’s failure to call for cars that the railroad could not have delivered. The [257]*257only function of the call would have been to prevent the railroad from claiming demur-rage. If Field was able and willing to accept delivery on those days, its failure to make the phone calls was an inadvertent omission, for which $19,000 is a big penalty. But the “if” is also big. The practice of calling for the cars after being notified of their arrival at the yard was well established. If Field really could have received the cars at its siding on the days in question it had every reason to make calls in order to avoid having to pay demurrage. Its failure to do so casts doubt on the accuracy of its affidavit, thus providing practical support for the Commission’s decision to treat calling as a condition precedent to Field’s avoiding its obligations under the tariff.
Although the result may seem harsh, we cannot say that the condition itself is an unreasonable one, let alone an unconscionable one; and as Professor Farnsworth has pointed out recently, “In exercising their freedom of contract the parties [to a contract] are not fettered by any test of materiality or reasonableness. If they agree, they can make even an apparently insignificant event a condition.” Contracts 538 (1982). This is what the Commission found the parties had done here. The Commission has been interpreting the average agreement for many years; and although contract interpretation is technically a question of law rather than fact, we think it simple prudence to defer to the Commission’s interpretation when it is reasonable, whether or not it seems correct to us as an original proposition. Cf. Illinois Terminal R.R. v. ICC, 671 F.2d 1214,1216-17 (8th Cir.1982). Interpreting constructive placement under the average agreement as having occurred when Field failed to order the cars before the end of the day on which it was notified of their arrival at the switching yard simplifies the administration of the average agreement by eliminating the type of factual dispute that Field wanted the Commission to resolve. We cannot call this an unreasonable interpretation.
Field’s second argument is that even if it was unable to receive cars because of the bad weather, a force majeure clause should be read into the agreement. Although acts of God and other unexpected events that prevent a party from performing his contractual obligations frequently do excuse performance under the rubrics of impossibility, impracticability, frustration, vis major, or force majeure, they do so only when it is reasonable to suppose that if the parties had negotiated expressly with regard to the unexpected contingency that has prevented performance they would have wanted the performing party to be excused. It is not always reasonable. As Holmes pointed out a century ago, you can bind yourself to perform acts over which you have no control. “In the case of a binding promise that it shall rain to-morrow, the immediate legal effect of what the promisor does is, that he takes the risk of the event, within certain defined limits, as between himself and the promisee.” The Common Law 300 (1881). For modern authorities see references in Farnsworth, supra, at 684-85. So we must ask whether the average agreement implicitly assigned the risk of a terrible winter that would prevent Field from receiving cars to Field or to the C & NW, and the answer is to Field. The whole purpose of the average agreement is to shift the risk of unexpectedly bad weather to the shipper. The express exception for floods, earthquakes, tornadoes, and hurricanes leaves little doubt that the exclusion of severe winter storms was deliberate. Indeed, without such an exclusion it would be difficult to find any quid pro quo for the railroad’s giving demurrage credits for early return of cars.
Last, Field argues that the Commission should have forgiven at least the penalty portion of the demurrage. But the so-called penalty is an intrinsic part of the average agreement which Field chose with its eyes open. The Commission was not obliged to rewrite the agreement in its favor. Monongahela Power Co. v. ICC, supra, 640 F.2d at 507-08; Cleveland Elec. Illuminating Co. v. ICC, 685 F.2d 170, 174 (6th Cir.1982) (per curiam); Illinois Central Gulf R.R. v. ICC, 702 F.2d 111, 114 (7th Cir.1983). [258]*258Field does not argue that the common law doctrine that penalty provisions in contracts are unenforceable is applicable to demur-rage.
The decision appealed from is Affirmed.