[23] POSNER, Circuit Judge, concurring and dissenting.

Central Illinois Light Company, a retail distributor of natural gas, bought natural gas from the defendant, an interstate gas pipeline company, at prices allegedly inflated because of violations of the antitrust laws by the defendant, and resold the gas to its customers. The question — and it is a difficult one — is whether any of those customers can sue the pipeline company on the theory that CILCO passed on the entire cost of the overcharge to them in the form of higher rates.

Hanover Shoe, Inc. v. United Shoe Machinery Corp., 392 U.S. 481, 88 S.Ct. 2224, 20 L.Ed.2d 1231 (1968), held that it is not a defense to a damages action that a buyer forced to pay a higher price because of the seller's antitrust violation passed on the cost of the violation to his own customers by raising his prices to them — unless the buyer had a cost-plus contract with them. Illinois Brick Co. v. Illinois, 431 U.S. 720, 97 S.Ct. 2061, 52 L.Ed.2d 707 (1977), announced a corollary to Hanover Shoe: the "indirect purchaser" (that is, the customer of the buyer, who is the "direct purchaser") cannot sue to recover the part of the overcharge that the buyer passed on to him. The Court again recognized an exception for the cost-plus contract, noting that it insulates the direct purchaser from "any decrease in its sales as a result of attempting to pass on the overcharge, because its customer is committed to buying a fixed quantity regardless of price." Id. at 736, 97 S.Ct. at 2069. Fastening on the words "fixed quantity," my brethren hold that the cost-plus exception is never available when the indirect purchasers are free to vary the quantity they buy from the direct purchaser.

This is not the first court to confine the exception so. See Mid-West Paper Products Co. v. Continental Group, Inc., 596 F.2d 573, 577 n. 9, 580 (3d Cir. 1979); In re Midwest Milk Monopolization Litigation, 730 F.2d 528, 533 (8th Cir. 1984); Lefrak v. Arabian American Oil Co., 487 F. Supp. 808, 819 (E.D.N.Y. 1980); cf. Arizona v. Shamrock Foods Co., 729 F.2d 1208, 1212 n. 2 (9th Cir. 1984). But the previous cases involved privately negotiated cost-plus contracts rather than cost-plus contractual provisions required by public utility regulation, which may make a difference as we shall see. Nor is it clear from the Supreme Court's opinion in Illinois Brick whether the reference to fixed quantity was intended to state an independent requirement of the cost-plus exception or merely to describe a normal contract situation, where the buyer's obligation is to buy the quantity specified in the contract. Not all contracts have this feature, but there is no indication that the Court meant to distinguish between fixed-quantity and variable-quantity contracts. Not only is the reference to fixed quantity dictum, because the case did not involve a fixed-quantity contract; the Court's entire discussion of cost-plus contracts is dictum, because the case did not involve a cost-plus contract but merely an argument (which the Court rejected, see 431 U.S. at 744, 97 S.Ct. at 2073) that a buyer's practice of rule-of-thumb cost-plus pricing should be enough to allow his customers to sue.

Although several district courts have rejected an exception for cost-plus regulation, see Go-Tane Service Stations, Inc. v. Ashland Oil, Inc., 508 F. Supp. 200, 204 (N.D. Ill. 1981); City of Cleveland v. Cleveland Electric, Illuminating Co., 538 F. Supp. 1320 (N.D.Ohio 1980); U.S. Oil Co. v. Koch Refining Co., 518 F. Supp. 957 (E.D.Wis. 1981), the cases are factually distinguishable; and in the case whose facts are most like those of the present case the district court held that indirect purchasers could sue, because public utility regulation had created "a straight cost passthrough." In re New Mexico Natural Gas Antitrust Litigation, 1982-1 Trade Cases ¶ 64,685, at p. 73,722 (D.N.Mex. 1982). Cf. County of Oakland v. City of Detroit, 628 F. Supp. 610, 613 (E.D.Mich. 1986); Illinois v. Borg, Inc., 548 F. Supp. 972, 975-76 (N.D.Ill. 1982). (An additional wrinkle in that case, however, was that the direct purchasers were in cahoots with the defendants; this was an independent ground for allowing the indirect purchasers to sue.) It is possible to allow indirect purchasers to sue in a case such as the present one without embracing the ill-defined "functional equivalent" approach of In re Beef Industry Antitrust Litigation, 600 F.2d 1148 (5th Cir. 1979); see also, e.g., Gulf Oil Corp. v. Dyke, 734 F.2d 797, 809 (T.E.C.A. 1984), effectively criticized in In re Midwest Milk Monopolization Litigation, 529 F. Supp. 1326, 1337-38 (W.D.Mo. 1982), aff'd, 730 F.2d 528 (8th Cir. 1984); Comment, A Legal and Economic Analysis of the Cost-Plus Contract Exception in Hanover Shoe and Illinois Brick, 47 U.Chi.L.Rev. 743, 756-70 (1980); cf. Abbott Dairies Division v. Butz, 584 F.2d 12, 16-17 (3d Cir. 1978), and found to be inapplicable to conditions in the beef industry itself in In re Beef Industry Antitrust Litigation, 710 F.2d 216, 219-20 (5th Cir. 1983). However, for reasons to be explained, I would not allow all the indirect purchasers in this case to sue — just the residential consumers.

To determine whether (or to what extent) this case is within the rule of Illinois Brick, we must consider the reasons for confining the right to sue to the direct purchaser; for it is the reasons behind a rule that determine its scope. First, the direct purchaser is closer to the violation, hence more likely to discover it. We therefore want to make sure that he has a powerful incentive to bring the violator to book, and we do this by holding out to him the prospect of recovering the entire damages for the violation if he wins the suit. Second, it is difficult to apportion damages between direct and indirect purchasers by the methods of litigation. A direct purchaser who finds himself paying a higher price for inputs would love to pass on all of the additional cost to his customers in the form of a higher price, but he can't, because a price that much higher will so reduce the demand for his product that his profits will fall unacceptably. The optimal adjustment to the increased cost of the input will be a price increase smaller than the increase in input cost. But this means that the increased cost will be divided between the two tiers, the direct and indirect purchasers — but in what proportions will often be hard to determine. An additional complication is that the higher input price may induce the direct purchaser to use more of an alternative input, and this substitution will affect the proportion of the initial overcharge that the direct purchaser can recoup.

Where the direct purchaser has a cost-plus contract with his customers that requires them to buy a fixed quantity, the reasons for confining the right to seek damages to the direct purchaser cease to be fully persuasive. There is no longer a problem of apportionment, because the whole of any price increase will have been passed on to the customers. Yet the other reason for confining the right to seek damages to the direct purchaser survives: he has better information about the violation. And, despite the cost-plus nature of the contract, he has everything to gain from suing. He will not have to share any of the damages that he recovers with his customers unless the contract contains a clause (or a court is persuaded to adopt an imaginative conception of unjust enrichment) that entitles them to any rebate which he might receive, probably years later, on an input used in performing his side of the bargain.

In the present case, where cost-plus pricing is the product of public utility regulation rather than a purely private contract, the reasons balance out slightly differently, but the case for applying the cost-plus exception of Illinois Brick is no weaker once the balance is restruck. The public utility has less to gain from suit than the direct purchaser in the purely private contract case has because the public utility commission may force the utility to pass on to the consumers any and all damages that the utility recovers. If so, the utility will have no incentive to sue; and it is no surprise that CILCO did not sue (and apparently the statute of limitations has now run). And although the amount of gas purchased by the utility's customers is not fixed in their contract with the utility, there is no problem of apportionment with respect to the only class of customers that I would allow to sue, the residential customers.

To see this, notice first that the retail distribution of natural gas through a grid of pipes to the customers' homes or places of business is a classic natural monopoly, Omega Satellite Products Co. v. City of Indianapolis, 694 F.2d 119, 123, 126 (7th Cir. 1982); so, in the absence of regulation, a gas utility would charge a monopoly price. Although the efficacy of public utility regulation has been questioned (see, e.g., Stigler Friedland, What Can Regulators Regulate? The Case of Electricity, 5 J. Law Econ. 1 (1962); Moore, The Effectiveness of Regulation of Electric Utility Rates, 36 So.Econ.J. 365 (1970)), the facts of this case suggest, and the parties seem to agree, that residential natural-gas rates in Illinois are lower because of regulation than they would be in an unregulated market, implying that CILCO has unused monopoly power in that market. The situation in the industrial market is different. Some industrial consumers of natural gas have good alternatives, and as to them CILCO apparently had no unused monopoly power that would enable it to shift the whole of the cost increase to them. Instead CILCO sought and obtained regulatory permission to reduce its profit margin on sales to these customers, thereby offsetting in part the higher rates enabled by the automatic pass-through provision.

The residential consumers have no good alternatives to natural gas, and as to them CILCO's profit-maximizing course of action was, it appears, to allow its rates to rise by the exact amount of the increase in its gas costs. Rate regulation evidently had succeeded in keeping CILCO's rates to a level where the demand for its gas was inelastic. In this region of a firm's demand curve, an increase in price will by definition increase the firm's revenues, because the price increase will not be offset by an equal or greater proportional decline in quantity demanded — that is what it means to say that demand is inelastic. And since the increase in price will reduce the firm's costs (by causing demand for its product to fall, since consumers will buy less at the higher price) at the same time that it causes the firm's revenues to rise, the firm's profits must increase. CILCO had therefore (with a qualification to be considered shortly) every incentive to raise its price by the full amount allowed by the regulatory commission — that is, by the full amount of the gas overcharge. This implies that the entire overcharge was passed on to the utility's residential consumers on all sales made to these consumers.

Put more simply, if because of regulation a utility's rates are below what it would like to charge, it will raise those rates by the full amount allowed by the regulatory commission unless such an increase would carry the utility above its optimal level. It is unclear from the record filed in this court how large the allowance was; conceivably the commission allowed CILCO to double its rates — yet even so, if regulation had forced CILCO to charge half or less of its preferred price, CILCO would pass through the entire cost increase. No one suggests that CILCO in fact absorbed any of the increase, so far as sales to its residential customers were concerned.

The mechanics of the pass-through provision are important. The "Uniform Purchase of Gas Adjustment Clause" that the Illinois commission required CILCO to include in its contracts not only entitled but directed CILCO, if it paid Panhandle Eastern Pipe Line Company an extra penny per million cubic feet of gas, to add exactly one penny to each customer's bill for every Mcf of gas sold to that customer. So if all of its customers had continued buying the same amount of gas, CILCO would have suffered no loss on account of the overcharge. To the extent that the utility lost residential sales because it was charging a higher price — and no doubt it lost some sales, because demand is never perfectly inelastic — the loss was a loss to the utility, was not passed on to its customers (at least in any sufficiently direct way to escape the rule of Illinois Brick), and hence is not an allowable factor in computing the customers' damages.

To illustrate the distinction between the two types of loss, suppose that the price to CILCO's residential customers before the overcharge to CILCO by Panhandle Eastern Pipe Line Company was $1 per Mcf, the overcharge was 10 ¢ per Mcf, and therefore by operation of the Uniform Purchase of Gas Adjustment Clause the retail price rose to $1.10 because CILCO made no offsetting reduction in another component of the price, as it did with its industrial customers. Suppose further that at the new, higher price CILCO sold only 950,000 Mcf, whereas at the old price it had sold one million Mcf; and suppose that nothing plausibly accounts for the decrease in sales except the price increase, which induced consumers to use less gas. Cf. Illinois Power Co. v. Commissioner, 792 F.2d 683, 687-88 (7th Cir. 1986). The loss to each consumer would be the number of Mcf he bought at the new price times 10 ¢ the loss to CILCO would be its lost profits on the sales it did not make.

Clearly, then, there is no problem of apportionment in the suit by the residential customers. Those customers are not seeking damages for gas they didn't buy, and the damages for the gas they did buy can simply be read off from their gas bills. (Well, almost: at the beginning of each year the utility estimates the amount of rate increase necessary to cover any increase in its gas bill, and there is an adjustment at the end of the year based on actual experience.) The only problem would come if CILCO had tried to sue for its lost sales, for then there would be more than one set of plaintiffs. But each set would be suing in respect of different sales — not, as in a Hanover Shoe or Illinois Brick case, the same sales. There would be no haggling over how much of the overcharge on each sale had been borne ultimately by the direct purchaser and how much by the indirect purchaser. And the proof of CILCO's lost sales would be straightforward — at least as straightforward as is possible in an antitrust case. It would involve comparing the utility's sales before and after the increase in gas prices, correcting for other factors, besides the increase, that might have affected those sales. Such correction is not always easy but is a conventional aspect of calculating damages in antitrust cases; it has to be done in every case where the plaintiff claims to have lost sales because of the defendant's unlawful conduct and the defendant argues that the loss was due partly or entirely to other factors. More important, it is a computational problem that has nothing to do with the problem that concerned the Supreme Court in Hanover and Illinois Brick. To repeat, the Court was concerned with the situation where two purchasers of the same thing — the initial purchaser and the purchaser from the initial purchaser — are or could be complaining that both had been hurt, and the problem is to apportion the loss between them. Here only the residential consumers could complain about a loss from the overcharge on the gas they bought, while only CILCO could complain about a loss caused by the overcharge on gas that the residential consumers did not buy.

Finally, unless the indirect purchasers are allowed to sue, the antitrust violation is likely to go unremedied, because the direct purchaser has even less incentive to sue than in a contractual cost-plus setting. CILCO might have an incentive to sue in respect of its lost sales (a distinct item of damages, as we have seen), but this would depend on how many sales it lost and on whether it would have to pass through any damages to its customers. In fact the incentive was not enough to induce it to sue.

I am not suggesting that the rule of Illinois Brick has no application to cases where the direct purchaser is subject to rate regulation. Although cost-plus is the spirit of rate regulation, the flesh is weak and often therefore the utility has considerable flexibility in pricing, much like an unregulated firm. Indeed, to the extent that the utility operates free from effective rate regulation — either because it faces competition that depresses its rates below the regulated level (apparently CILCO's situation with its industrial customers), or because the regulators are unable to prevent it from charging monopoly prices to its captive customers — its situation is identical to that of an unregulated seller. But in the case of CILCO's residential customers, regulation apparently succeeded in forcing the utility to operate deep in the inelastic region of its demand curve, with the result that 100 percent passing on of any cost increase was the optimal strategy for the utility to follow.

With the rapid and unanticipated increases in fuel prices during the 1970s, utilities pressed for and obtained the right to include automatic fuel pass-through provisions in their contracts with customers, provisions that would allow the utility to pass on every dollar in higher prices that it paid for gas or other fuels to its customers without going through the time-consuming process of obtaining regulatory authorization to raise rates. Such provisions are also found in unregulated contracts and should be treated the same there when the contract requires the buyer to take either a fixed quantity or his requirements, since a buyer cannot reduce his purchases under a requirements contract merely because he is dissatisfied with the terms of the contract as they have worked themselves out. See Wilsonville Concrete Products v. Todd Building Co., 281 Or. 345, 352, 574 P.2d 1112, 1115 (1978); Royal Paper Box Co. v. E.R. Apt Shoe Co., 290 Mass. 207, 195 N.E. 96 (1935); Fort Wayne Corrugated Paper Co. v. Anchor Hocking Glass Corp., 130 F.2d 471, 473-74 (3d Cir. 1942); White Summers, Handbook of the Law Under the Uniform Commercial Code 126 (2d ed. 1980). (This example shows, by the way, why a rigid requirement of fixed quantity would be a senseless limitation on the cost-plus exception of Hanover Shoe and Illinois Brick: a buyer under a requirements contract does not have discretion as to the amount to take under the contract.) If on the other hand the buyer has complete flexibility as to how much to buy, a cost-plus provision is ineffectual; the buyer can always condition an agreement to buy a specific amount on the seller's agreeing to modify the contract by reducing the price. This is another reason for supposing that an agreement to take a fixed quantity, or, what is equivalent for these purposes, an agreement to take one's requirements, is implicit in the cost-plus exception rather than being an independent requirement for invoking it. The unexhausted monopoly power of a regulated utility takes the place of a fixed-quantity or requirements provision. The utility can force the whole of the cost increase through to its residential customers without sacrificing any profits, and did so.

I would not allow the industrial customers to sue, however. By cutting its profit margin to them, CILCO raised its price by less than the increase in gas cost; and while the apportionment of that increase between the utility and its industrial customers is easy to make — precisely because the utility was required to get approval for reducing its profit margin — I would be reluctant to complicate the administration of the Illinois Brick rule by trying to distinguish between difficult and easy apportionment cases. And the Court seemed unwilling to listen to such arguments. However, for every cubic foot of gas bought by a residential customer, we know that the whole overcharge was passed on to the customers, in accordance with the fuel pass-through provision.

It might seem an unimportant detail whether a buyer reacts to an overcharge by raising its price by less than the overcharge (as CILCO did with its industrial customers), thus losing fewer customers, or by raising its price by the full overcharge and thereby losing more customers than it would if it swallowed part of the overcharge. But in the second case the problem of apportioning losses on the same sales does not arise. It might also seem impermissible under Illinois Brick to inquire into the amount of passing on but the Court made an explicit exception for cases where there is a cost-plus contract. There is such a contract here — the automatic fuel pass-through provision — and although it is not a contract for a fixed quantity or (what I contend is equivalent for purposes of the exception) the buyer's requirements, the existence of public utility regulation is an adequate substitute in the circumstances.

We can never be absolutely certain that regulation has resulted in a 100 percent pass through; for all we know, CILCO would have sought a rate increase but for the gas overcharge, and by forbearing to do so in effect absorbed part of the overcharge. But by the same token, the seller under a fixed-quantity cost-plus contract might forbear to insist on a 100 percent pass through in order to curry favor with the buyer for the sake of future deals. No counterfactuals are certain, but the doubts here are too small to warrant us in insisting that this potentially serious antitrust violation, which may have imposed on consumers of natural gas aggregate damages of almost $50 million, go unremedied, as apparently it will.

The suit by the residential customers is within the scope of the cost-plus exception to the rule of Illinois Brick, and I would therefore affirm the denial by the district court of the defendant's motion to dismiss the complaint insofar as the complaint seeks damages on behalf of CILCO's residential customers.

ORDER

On consideration of the petition for rehearing and suggestion for rehearing en banc filed by counsel for plaintiff-appellee in the above-entitled cause, a vote of the active members of the Court was requested. A majority of the judges in regular active service above named[**] voted to GRANT the petition and suggestion for rehearing en banc. Accordingly,

IT IS ORDERED that the aforesaid petition for rehearing and suggestion for rehearing en banc be, and the same are hereby, GRANTED.

IT IS FURTHER ORDERED that the judgment and opinion entered in this case on January 22, 1988 be, and are hereby, VACATED. This case will be reheard en banc on Thursday, May 26, 1988, at 9:30 AM.