74 F.3d 778
64 USLW 2456
GREAT CENTRAL INSURANCE COMPANY, Plaintiff-Appellant,
INSURANCE SERVICES OFFICE, INCORPORATED, Defendant-Appellee.
United States Court of Appeals,
Argued Nov. 29, 1995.
Decided Jan. 17, 1996.
Eugene J. Schiltz, Robert F. Coleman (argued), Coleman & Associates, Chicago, IL, for Plaintiff-Appellant.
James E. Betke (argued), Anne R. Pramaggiore, McDermott, Will & Emery, Chicago, IL, for Defendant-Appellee.
Before POSNER, Chief Judge, and COFFEY and KANNE, Circuit Judges.
POSNER, Chief Judge.
This is a very novel and interesting suit for breach of contract and for tort. A diversity suit governed by Illinois law, it was brought by an insurance company against the nation's largest insurance rating service. The district court granted summary judgment for the defendant, precipitating this appeal.
The plaintiff, Great Central Insurance Company, is, despite the first word in its name, a small insurance company (its total assets in 1987 were only $106 million, though it was and is the wholly owned subsidiary of a much larger company), specializing in writing liability insurance for supermarkets. Half its customers are supermarkets, a concentration that appears to be highly unusual. Hundreds of insurance companies write liability insurance for supermarkets but for the vast majority of these companies, indeed perhaps for all but Great Central, their supermarket business is minute. Indeed, despite its small size, Great Central appears to write more supermarket liability insurance than any other company.
The defendant, Insurance Services Office, is a nonprofit corporation the members of which are insurance companies. Great Central is not a member of ISO but is a customer. ISO sells a variety of services to the property and casualty (liability) insurance industry. Among these are actuarial and filing services. ISO gathers from the members of the industry and consolidates a vast amount of data concerning risks. For example, it collects data concerning slip and fall accidents in grocery stores and supermarkets. These data enable it to calculate insurance rates that will cover the risk of the various insured activities. The calculation of these rates is the actuarial function of ISO and the filing function is the filing of the rates with the state insurance commissions for their approval. Insurance companies are not required to charge the filed rates. Unlike the rates promulgated by the old insurance rate bureaus, ISO's rates are strictly advisory. Insurance companies can file their own rates, also subject of course to commission approval. But thanks to the antitrust exemption created by section 2(b) of the McCarran-Ferguson Act, 15 U.S.C. Sec. 1012(b), ISO is permitted to suggest what rates the members of the insurance industry should charge even though they are competitors, and the insurance companies are permitted to charge the filed rates.
If an insurance company, whatever its size, writes so little of a particular type of insurance that its claims experience is too meager for it to be able to calculate a rate reasonably likely to cover future claims, it is more or less compelled to charge ISO's filed rates; at least we may assume so for purposes of this appeal. The inability of many insurers to compute compensatory rates without an industry-wide pooling of loss experience is the rationale of the antitrust exemption. Great Central may be unique in writing so much supermarket insurance that it can calculate a compensatory rate on the basis of its own loss experience.
All this is by way of background to the events out of which this suit arises. In the early 1980s ISO embarked on an ambitious program of simplifying the rate structure of its Comprehensive General Liability policy, the standard commercial liability insurance policy issued by many hundreds of insurance companies to many thousands of businesses. As part of the simplification ISO combined what had been two separate rate classifications, one for grocery stores and one for supermarkets, into a single classification. The old rate structure had defined a supermarket as a grocery store with annual sales of at least $500,000 and a floor area of at least 3,000 square feet; all other grocery stores were classified in the grocery-store classification. Within each of the two classes, the rates varied by floor area as well as, of course, by coverage (and also by state, but that is not relevant to the appeal). A grocery store having a floor area of 2,000 square feet would pay a higher rate, for the same limits of coverage, than a 1,000 square foot grocery store, while a 5,000 square foot supermarket would pay a higher rate than a 4,000 square foot supermarket. More important, the rate per square foot differed for the two types of store; in some states, the rate for supermarkets was five times greater than the rate for grocery stores.
When it combined the grocery-store and supermarket base rates, ISO did what at first glance seems very strange: rather than averaging the rates, it used the old grocery-store rate per square foot as the basis for calculating rates for the new, combined class. This is not quite so crazy as it sounds, because at the same time that the classes were being combined the method of calculating the rate was being changed. Instead of varying the rate within each class (now within a single class) on the basis of area, the new rate system varied it by volume of sales. Apparently the reason for the change was that the number of slip and fall accidents, and hence the number of claims, are proportional to the amount of "foot traffic" in a store, which in turn, ISO determined, was more closely correlated with sales than with area. In principle at least, the new rate system could have generated higher liability-insurance rates for some supermarkets.
The rates were duly filed with the various state commissions and by the middle of 1986 were in effect. Shortly afterward Great Central realized that the new method of calculating the rates for supermarket liability insurance would yield much lower rates than under the old system. It protested to ISO, which quickly agreed that the rates were indeed "too low." We use quotation marks to flag a significant ambiguity in the concept of "too low" rates. ISO has never conceded, and there is no evidence, that the new rates were inadequate to cover the reasonably predictable risks of claims by supermarkets, together with the other expenses of the insurer. (An insurance rate is composed of the "loss premium," compensating the insurer for the risk of having to indemnify the insured, and the "loading expense," covering the insurer's administrative costs. Since 1989, ISO has computed only loss premiums. The change has no bearing on the issues in this case.) Despite our earlier point that Great Central is probably the only insurer with enough claims experience to realize that a supermarket rate promulgated by ISO might be too low to cover likely future claims, there is no evidence that Great Central believed that the new rates were inadequate in that sense. So far as appears, it squawked only because it saw a threat to its revenues, and thus to its profits, that is, to the difference between revenues and costs, and those revenues, the revenues generated by the previous supermarket rate, may have been well above any reasonable expectation of future claims. Its brief describes the new rates as having had a "catastrophic impact on its profits " (emphasis added). We add that even "inadequate" rates in the sense that we have defined them may not really be inadequate, since an insurance company's revenues include not only the premiums it charges but also the investment income on those premiums before they are expended to cover claims against insureds.
However all this may be, all that ISO has ever conceded was that the new rates would in fact be much lower than the old, and as this had not been intended, and as ISO is a service organization for insurance companies, it was unlikely to resist Great Central's plea and did not. It readily conceded then, and it concedes now, that the new rates were a "mistake," though not that they were "inadequate" in the sense just explained.
According to Great Central, ISO, while acknowledging its mistake, dragged its heels in correcting it by rescinding the new rates. Not until 1994 were they withdrawn from every state in which they had been filed. ISO denies that it dragged its heels. It points out, plausibly enough, that withdrawing a filed rate is easier said than done. State insurance regulators, charged as they are by law with protecting consumer interests, may look askance at a request to withdraw a recently filed rate because the rate is too low and to substitute a much higher rate. ISO wanted to wait until it could sweeten a request to withdraw the low rates with the submission of a low rate for some other type of insurance, and this waiting game, played out over 50 states, was bound to take years. (Great Central calls this waiting game that we've described a "fraud." It is not.) Only one company besides Great Central complained about the new rates. The matter could not have seemed an emergency requiring ISO to circumvent its normal procedures. For purposes of this appeal, however, we assume both that the acknowledged original mistake in combining the grocery store and supermarket rate classifications in the way that ISO did and the slow pace at which ISO corrected the mistake were negligent.
Great Central was not deceived by the new low rates that went into effect in 1986. But, it claims, many of its competitors were. Having too little experience with supermarket liability insurance to know what a proper rate for such insurance was, they, like so many sheep, docilely charged the rates filed by ISO. Great Central claims, surprisingly in view of the McCarran-Ferguson exemption, to have been inhibited by fear of antitrust liability from notifying its competitors of their mistake. See Hartford Fire Ins. Co. v. California, --- U.S. ----, 113 S.Ct. 2891, 125 L.Ed.2d 612 (1993), esp. p. 2901. But McCarran-Ferguson is an exemption only from federal antitrust law; insurers have to worry also about state antitrust law. See, e.g., State v. Insurance Services Office, Inc., 1991-1 CCH Trade Cases p 69,290, 1991 WL 290673 (Texas Dist.Ct.1991).
Although we accept this scenario for purposes of deciding the appeal, it is none too plausible. Remember that there is no evidence that Great Central considered the new rate too low to cover the reasonably predictable claims of supermarkets hit with liability for slips and falls and for other covered accidents that occur in supermarkets. The only evidence is that Great Central realized that the new rates were much lower than the old. Any insurance company that wrote supermarket insurance would realize that. If, however, it wrote very little of this insurance, it might not think the filing of its own, higher rate worth the bother, especially if it anticipated that many of its competitors would not bother to do so and thus would be underpricing it.
Whether Great Central's competitors were misled by ISO's mistake and by its sluggishness in correcting it into charging the new rates, or merely went along with these rates out of inertia or because of the cost of filing its own rates, Great Central was hurt. It could and did, because of its heavy concentration in supermarket liability insurance, file its own higher rates, and so far as appears these rates were approved by the insurance regulators. But the consequence of the higher rates was that Great Central lost business to its competitors who, whether foolishly or not, were offering to sell the same insurance at the lower, ISO rates. The lost profit on this business, and the lost profit resulting from Great Central's not raising its rates all the way to their pre-1986 level for fear of losing too much business to its competitors, are the damages that it seeks in this suit. Its complaint reckons those damages at $11.3 million.
Great Central bases its claim to these damages primarily on its contract with ISO. It interprets the contract as containing a promise by ISO to publish accurate rates and retract any that are found to be mistaken. No such promise can be found in the only writing actually exchanged between the parties--a simple purchase form in which the insurer checks the services that he wishes to purchase from ISO. Nor is there any suggestion of an oral agreement. Rather, Great Central relies on public statements by ISO, primarily to Congress and state regulatory commissions, describing its business as including the sale of "advisory" or "benchmark" rates for use by liability insurance companies. Implicit in these statements, Great Central argues, are a representation that the rates are accurate and that if they are discovered not to be so ISO will take prompt steps to retract them and substitute accurate rates.
In effect though not in so many words Great Central is arguing that ISO's sale of rates by means of a simple purchase form is accompanied by implied warranties that the rates are responsible estimates of proper prices for the lines of insurance in question. The argument is at least plausible, and it will not be necessary for us to consider whether it is actually sound. We have pointed out that an insurance company which writes a limited amount of insurance of a particular type will not know how much it should charge in order to cover future claims; its claims experience will constitute too small a sample of claims to enable a responsible extrapolation of future liability. It will perforce rely on ISO to tell it what its rates, or at least the loss-premium component of its rates, should be. ISO does not merely admit this; it trumpets it: "Only a few insurers enjoy a market share large enough to permit them to develop rates based solely on their own loss experience and actuarial analysis." If ISO makes a mistake, the insurance company that charged ISO's rate could find itself having to pay out claims far in excess of the amount of premium income that it had collected to offset future claims. To repeat an earlier point, this may not be such a case; ISO's new rates for grocery stores and supermarkets may not have been inadequate in the sense of not likely to cover future claims, but merely lower than the previous rates. The case of an inadequate rate is the clearer case for implying a warranty of accuracy, though even then the case is not conclusive. For ISO to act as a guarantor of its customers' rates--as in effect an insurer against the actuarial errors of insurance companies--might entail a potential liability too great to be plausibly imputed to a contract that contains no reference to liability for mistakes or to limitations on that liability. ISO is not a reinsurer, and does not hold reserves against the possibility of having to make good on claims by its "insureds."
But even if, contrary to what we have just suggested, an insurance company that relied on a mistaken rate and as a result charged insurance premiums too low to cover the claims against the policies that it issued would have a claim for breach of contract against ISO, this is not the nature of the claim in this case. Great Central did not rely on the ISO rates. It was not fooled. It knew the rates were mistaken and filed its own rates. It could, we may assume, sue ISO for the return of the money that it paid for the mistaken rates. That is not the relief sought; we cannot find in the record even the prices that Great Central paid ISO (the prices are not in the purchase agreements). The relief sought is damages for the losses caused Great Central by the fact that its competitors followed the ISO rates and as a result underpriced it.
To fit such a claim under the contract between ISO and Great Central would require us to find implicit in that contract a promise not only to sell accurate rates to Great Central but also to indemnify Great Central should a mistake result in its competitors' underpricing it. There is no evidence that such a promise was implicit in the dealings between the parties. It hardly seems likely.
Moreover, Great Central's theory of liability is independent of that or any such promise. The harm it suffered is unrelated to its purchase of rates from ISO. The harm resulted entirely from the purchase of these rates by other insurance companies. Had Great Central never dealt with ISO, its harm would have been the same: the low rates, adopted by its competitors, would have required it either to lower its own rates, or to lose business to its competitors, or to incur both forms or sources of loss.
This is really a tort case, in which a person complains not that another person broke faith with him but that the other person, with whom he had no contractual relation, injured him directly or, as occurred here, indirectly. So it is no surprise that the complaint, though it leads off with the hopeless breach of contract claim, includes tort counts, three in fact. One is intentional interference with contract. In re Estate of Albergo, 275 Ill.App.3d 439, 211 Ill.Dec. 905, 911, 656 N.E.2d 97, 103 (1995); W. Page Keeton et al., Prosser and Keeton on the Law of Torts Sec. 129 (5th ed. 1984); 2 Fowler V. Harper, Fleming James, Jr. & Oscar S. Gray, The Law of Torts Secs. 6.5-6.8 (2d ed. 1986). The argument is that by its mistake and its heel dragging ISO disrupted Great Central's contractual relations with its supermarket customers. This theory of violation is remote from the core of the tort, illustrated by cases in which the defendant lures away the plaintiff's employee, knowing that the employee has a contract with the plaintiff that he is breaking by going to work for the defendant. Lumley v. Gye, 2 El. & Bl. 216, 118 Eng.Rep. 749 (Q.B.1853). Even in such cases, the imposition of tort liability is controversial. It seems inconsistent with the principle that, save in those exceptional cases in which the victim of a breach of contract can obtain a decree of specific enforcement, a contracting party is free to break his contract so long as he is prepared to compensate the promisee. See, e.g., Lillian R. BeVier, "Reconsidering Inducement," 76 Va.L.Rev. 877 (1990). The employee who breaks his contract of employment would by doing so, one might think, merely expose himself to a suit for breach of contract. Why should the victim of the breach have a tort remedy against a person who merely invites the employee to exercise his option of breaking his contract and paying damages? Various answers have been offered. They include history--the tort originated at a time when the status of a slave, who is property, and of an employee, who is not, were not sharply distinguished (the equivocal term "servant" capturing that duality)--and the practical point that an employee who breaks his contract will often be judgment-proof, so that the employer's interest (he may have invested heavily in the employee's skills) is not adequately protected by contract law. It is unnecessary to pursue these interesting byways here. This is not a case in which someone "steals" the plaintiff's promisor. ISO was not "stealing" Great Central's supermarket customers.
This might seem to be the end of the matter so far as Great Central's claim of tortious interference with contract is concerned. But the tort has expanded far beyond its origins. There are many cases, colorfully illustrated by American Surety Co. v. Schottenbauer, 257 F.2d 6 (8th Cir.1958), in which the defendant has been held liable even though he was not trying to make a substitute contract with the plaintiff's promisor. In American Surety the plaintiff was discharged by his employer because the defendant, the workers' compensation insurer of the employer, had falsely advised the employer that the plaintiff had a diseased finger which was creating a serious risk of accident. When invoked in such cases, the tort of intentional interference creates a remedy that overlaps other torts, in American Surety the tort of defamation. The breach of contract is brought about by independently tortious conduct.
The present case is neither one in which a person tries to lure away someone who is under contract to someone else, nor one in which independently tortious conduct deprives the plaintiff of a contractual right. It does not fit the contours of the tort--and for the more fundamental reason that the tort of interfering with contract is an intentional tort. Keeton et al., supra, Sec. 129, pp. 997-1002. ISO had no intention of causing Great Central to lose any of its customers. At worst it acted negligently, and as a remote consequence one of its customers lost some of that customer's own customers.
Great Central's second tort theory is that ISO made negligent misrepresentations concerning the accuracy of the new rates. Negligent misrepresentation is a tort in Illinois, Board of Education v. A, C & S, Inc., 131 Ill.2d 428, 137 Ill.Dec. 635, 546 N.E.2d 580, 591 (1989), but of course requires proof that the plaintiff relied on the misrepresentations. Id.; Cahill v. Eastern Benefit Systems, Inc., 236 Ill.App.3d 517, 177 Ill.Dec. 718, 722, 603 N.E.2d 788, 792 (1992). Great Central did not rely. It saw right through ISO's mistake. The insurance companies that relied were the insurance companies that, having too little supermarket insurance business to know what a proper rate for that business should be, went along with ISO's rates. They might if they were injured have a claim against ISO, though we need not decide this. But Great Central does not have a claim. There is no authority for maintaining a suit for negligent misrepresentation on the ground that other people (unless they were the plaintiff's agents, Bonhiver v. Graff, 311 Minn. 111, 248 N.W.2d 291, 302 (1976); Restatement (Second) of Torts Sec. 552(2), comment h, examples 4-6 (1965)) were deceived by the misrepresentation and as a consequence injured the plaintiff. See Cahill v. Eastern Benefit Systems, Inc., supra, 177 Ill.Dec. at 722, 603 N.E.2d at 792; Restatement, supra, Sec. 552(2), comment i. The imposition of liability would be inconsistent with the requirement that the plaintiff have relied on the misrepresentation of which he is complaining.
It should be apparent from this discussion that what Great Central really wants in this appeal is for us to create in the name of Illinois a new tort. Perhaps a "generic" tort--anyone injured, by whatever route, through the negligence of another should be entitled to obtain damages--or a special tort of actuarial malpractice. It is not clear which. No doubt Great Central would settle for either. Or for an expansion of the tort of interference with contract to include negligent interference, which would have many of the same consequences as the concept of "generic" tort. Cf. Keeton et al., supra, Sec. 129, pp. 100-01. None of these possibilities should be rejected out of hand. If a person injures another, himself faultness, through culpable conduct such as negligence, why should not the innocent victim be able to shift his loss to the culpable injurer? There is no altogether satisfactory answer to this question, but it is obvious that the law recoils from the full implications of it. In Edwards v. Honeywell, Inc., 50 F.3d 484 (7th Cir.1995), we were faced with a case in which a fireman had been killed as a result (or so at least it was contended, with some evidentiary support) of the negligence of the company that had installed and was monitoring the fire alarm in the house in which he was killed fighting a fire. We held that the alarm company was not liable to the fireman, because he was not within the scope of the company's duty of care. This was just a conclusion, but behind it were considerations of policy that despite a certain vagueness have persuaded the courts of all states to place limits on the scope of tort liability through the concept of duty. The same considerations are in play here. ISO is an enterprise rather than an individual, and its ability to prevent its employees from making careless mistakes whether of commission or omission is, as a practical matter, limited. To subject it to unlimited liability for the consequence of those mistakes might be to jeopardize its existence or make it unduly timid about proposing less than astronomical rates, cf. Greycas, Inc. v. Proud, 826 F.2d 1560, 1564 (7th Cir.1987), though we were told at argument that ISO itself carries liability insurance that might have covered its liability in this case if it were adjudged liable. Even though many of its customers are utterly dependent on its rates because they do not have enough claims experience to calculate a compensatory rate on their own, and this dependence could be thought by analogy to the liability of common carriers for injury to their passengers to create a correlative duty of exceptional care, insurance companies can and do protect themselves from the consequences of actuarial mistakes through reinsurance--and even reinsurers protect themselves, by contracts with re-reinsurers, called "retrocessionaires." Great Central's lawsuit seeks to transform ISO into a kind of reinsurer.
Nor, to recur to our earlier distinction between lower rates and inadequate rates, is there any indication that the "mistaken" rates belatedly withdrawn caused any worse harm to Great Central and to its competitors than to reduce the profitability of what may have been excessive rates in relation to the actual risks of supermarket liability. In that event, the effect of the present suit, if Great Central won, would be to recognize an enforceable interest in supracompetitive profits; yet the common law, early recognizing the importance of free markets, never took that step--has never recognized competition (unless predatory or dishonest) as a tort. Keeble v. Hickeringill, 11 East. 574, 103 Eng.Rep. 1127 (K.B. 1706 or 1707). And on top of all this, it is very odd for a firm to be complaining about injuries to rather than by its competitors. The low rates may, for all we know, have caused some of Great Central's competitors to stop writing supermarket liability insurance, which would be to its advantage, while Great Central itself could boast to its stockholders that its specialization in the writing of such insurance had enabled it to avoid incurring the huge losses (given that specialization) that would have ensued from supine adoption of ISO's new rates.
The change in the supermarket rate, moreover, was one of hundreds of changes in rate classifications made by ISO as part of a comprehensive overhaul of its Comprehensive General Liability policy. Some rates were raised, others lowered. Insurers harmed by a lowering of rates on one line of insurance would tend to be benefited by an increase in rates on another line. Only an insurer such as Great Central that had an unusual concentration in one line would be at serious risk of experiencing a substantial net loss of revenues as a result of the overhaul. There is nothing to indicate that ISO knew or should have taken steps to find out that there was such a vulnerable firm. That vulnerability was the deliberate choice of Great Central, which could have reduced it through diversification or reinsurance. ISO is not the insurer's insurer.
Weighing all these factors, any one of which might well convince a court not to recognize the new tort for which Great Central contends, we would be reckless to predict that the Supreme Court of Illinois, were it to hear such a case, would recognize (more realistically, would create) such a tort. We keep warning the bar that a plaintiff who needs a common law departure or innovation to win should bring his suit in state court rather than in federal court. Haynes v. Alfred A. Knopf, Inc., 8 F.3d 1222, 1234 (7th Cir.1993); Shields Enterprises, Inc. v. First Chicago Corp., 975 F.2d 1290, 1297 (7th Cir.1992); Anderson v. Marathon Petroleum Co., 801 F.2d 936, 942 (7th Cir.1986); Murphy v. White Hen Pantry Co., 691 F.2d 350, 355 (7th Cir.1982). Great Central chose to bring this suit in federal court. It is not a case in which the plaintiff seeking the innovation sues in state court, only to be removed by the defendant into federal court. We do not suggest that Great Central's theories of liability have sufficient merit to persuade any court; we rather think the contrary; but in any event more is asked of us than is proper to ask of a federal court trying to predict the evolution of state law.