The plaintiffs, former law partners named Goldstick and Smith, brought this diversity suit for breach of contract against John Kusmiersky, U.S. Managers Realty, Inc., and ICM Realty, seeking payment of a legal fee for getting real estate taxes re*458duced on a large apartment complex owned by ICM, leased by Kusmiersky, and managed by U.S. Managers. ICM moved for summary judgment, which was granted, 593 F.Supp. 639 (N.D.Ill.1984). After settling with Kusmiersky and U.S. Managers, Goldstick and Smith appealed the judgment for ICM.
Neither the district court nor any of the parties noticed a serious question of federal jurisdiction. ICM is a real estate investment trust, not a corporation. Its citizenship is not, as all concerned assumed, the state where it is chartered, Maryland. The citizenship of a trust is determined for purposes of diversity jurisdiction by the citizenship of the trustee or, in this case, trustees, all of whom must be citizens of a different state from the plaintiff for the suit to be maintained under the diversity jurisdiction. Navarro Savings Ass’n v. Lee, 446 U.S. 458, 100 S.Ct. 1779, 64 L.Ed.2d 425 (1980). The trust in Navarro was a Massachusetts business trust, though like ICM it had been created to invest in real estate. ICM is a formal real estate investment trust. But this is a distinction without a difference, so far as diversity jurisdiction is concerned. As the Court said in Navarro, “a trustee is a real party to the controversy for purposes of diversity jurisdiction when he possesses certain customary powers to hold, managS, and dispose of assets for the benefit of others.” Id. at 464 100 S.Ct. at 1783 (footnote omitted). ICM’s trustees have all of these powers; although Maryland law places few formal restrictions on the allocation of governance between the trustees and shareholders of a real estate investment trust, see Md. Corps. & Ass’ns Code Ann. § 8-202, ICM’s declaration of trust in fact vests all governance of the trust in the trustees. The special circumstances in Betar v. DeHavilland Aircraft of Canada, Ltd., 603 F.2d 30 (7th Cir.1979), are not present here, making it unnecessary for us to decide whether Betar survived Navarro, as several members of this court have doubted, see Wilsey v. Eddingfield, 780 F.2d 614, 619 (7th Cir.1985) (opinion dissenting from denial of rehearing en banc).
Although the original record did not show what states the trustees of ICM are citizens of, in response to our request for supplemental briefs on jurisdiction the parties submitted an affidavit attesting that on the date the suit was filed all of the trustees were in fact citizens of different states from the plaintiffs. Therefore the requirements of diversity jurisdiction are satisfied. But they easily could not have been; and then we would have had to dismiss the case, and three years of litigation would have gone down the drain. So once again we remind the bench and bar of this circuit of the importance of verifying the existence of federal jurisdiction at the outset of litigation. See, e.g., Kanzelberger v. Kanzelberger, 782 F.2d 774 (7th Cir.1986).
On the merits of the appeal we note first the parties’ agreement that the governing substantive law is that of Illinois and that, under Rule 56 of the Federal Rules of Civil Procedure, ICM in moving for summary judgment had the burden of proving that there was no genuine issue of material fact on any of the liability theories advanced by the plaintiffs — no triable issue of fact. We therefore construe the facts as favorably to the plaintiffs as is reasonable to do, without meaning to suggest that the plaintiffs will succeed in proving these facts at trial.
In 1969 ICM bought an apartment complex in Rolling Meadows, Illinois, from Walter Kassuba, to whom ICM leased back the land and sold the improvements. He went broke, however, and in 1975 ICM leased the land to Kusmiersky, a “work-out specialist,” who also bought the improvements. In a separate agreement ICM promised to advance Kusmiersky $1.5 million to operate the property and pay past-due real estate taxes, hoping that this advance would enable him to resuscitate the property so that ICM could recover its investment in the land. Kusmiersky retained the plaintiffs under a contingent-fee arrangement to seek a reduction in the past-due taxes. ICM approved this arrangement and by 1977 the plaintiffs had succeeded in getting them reduced by some *459$870,000. The plaintiffs billed Kusmiersky $290,000 for their legal services. The bill was never paid, though neither its reasonableness nor its conformity to the terms of the contingent-fee contract has ever been questioned.
The property continued to decline, and Kusmiersky and ICM entered into negotiations to transfer Kusmiersky's interest to Ted Netzky, who agreed to invest fresh capital in the property. Netzky refused to close the deal, however, unless ICM agreed to pay the remaining past-due real estate taxes and the plaintiffs released their claim for legal fees. In an effort to obtain this release Kusmiersky offered the plaintiffs a reduced fee of $250,000, payable over 10 years with interest at 7 percent per annum. Goldstick (for the plaintiffs) refused the offer. ICM drafted a release for the plaintiffs to sign which stated that the $250,000 would be paid over time but only out of the profits of the property; if there were none, the fee would not be paid. Goldstick refused to sign the release until Kusmiersky assured him that ICM was honorable and that something would be worked out, which Goldstick took to mean that he and Smith would receive the $250,000, over time but with interest, regardless of whether the property showed a profit. The plaintiffs signed the release and the deal with Netz-ky closed. Goldstick and Smith continued to negotiate with Kusmiersky and ICM over the terms of payment of their fee, but no agreement was reached, for ICM held steadfastly to the position that the payment of the fee would have to be out of the profits, if any, of the property. There were no profits, and eventually Goldstick and Smith demanded payment of the full $290,000 and when that was refused brought this suit.
The plaintiffs advance several alternative grounds for recovery, all of which the district court rejected. The first is that ICM was a party to their original contract with Kusmiersky to seek a reduction in back taxes in exchange for a contingent fee. If Kusmiersky was merely ICM’s tenant, however, the district judge was certainly correct in concluding that Kusmier-sky was not its agent in the original fee negotiation with the plaintiffs. If a tenant hires a plumber to fix a leaking faucet, the contract of repair is between the lessee and the plumber; the landlord is not a party. See, e.g., County of Kane v. Midway Landfill, Inc., 23 Ill.App.3d 1080, 1083, 321 N.E.2d 91, 94 (1974). If the tenant absconds without paying and the landlord repossesses the property, the plumber might conceivably (though not in a case such as this, as we shall see later) have a claim against the landlord for restitution of benefits conferred: the faucet has been fixed, and the benefit enures to the landlord. But the plumber’s claim would be for unjust enrichment rather than breach of contract and would fail if the landlord had already paid the tenant to fix the faucet, for there would be no enrichment in that case — or in this, at least so far as the original deal with the plaintiffs is concerned. ICM had promised Kusmiersky $1.5 million, which among other things was to be used to pay off the back taxes on the property. If Kusmiersky could get a net reduction in this tax liability by paying a lawyer less than the reduction in taxes that the lawyer obtained, the payment for the lawyer, like the payment for the taxes for which it was a substitute, would come out of the $1.5 million. ICM had paid once, and didn’t want to pay again. Of course ICM would still have to worry about liens being slapped on the property by virtue of its lessee’s operations but it would not have to worry about being sued for breach of the lessee’s contracts — after the property had been written off as a loss — merely by virtue of being the lessor.
What blurs this picture is that the agreement between ICM and Kusmiersky concerning the payment of the real estate taxes, an agreement separate from the lease, provided for only a conditional advance of money to Kusmiersky. He was to try to settle the claims for past-due taxes as cheaply as possible but he had to advise ICM on whether to accept the settlements he negotiated, and ICM would make the *460final decision. The agreement makes no reference to legal fees but Kusmiersky testified in his deposition that it was meant to include fee arrangements too. If so, this meant that ICM had to approve any such arrangement as well as any' tax settlements and as a matter of fact it did approve Kusmiersky’s contract with the plaintiffs, and did so only after the plaintiffs agreed to reduce the contingent fee they had originally negotiated with Kusmiersky.
The district court compared ICM’s role to that of a lender who insists on a right of approval of his borrower’s use of the borrowed funds in order to make sure they aren’t dissipated, which would reduce the lender’s security. That is one possible interpretation but not the only one. ICM owned the land and would be the loser if it were seized for nonpayment of real estate taxes. The agreement with Kusmiersky regarding the payment of those taxes could be interpreted as making Kusmiersky ICM’s agent. If so, ICM was bound as his principal. The proper characterization of the relationship created by the agreement is not so clear on this record that it could be determined on a motion for summary judgment.
We add that it would make no difference whether the agency agreement (if that is what it was) authorized Kusmiersky to make a contingent-fee arrangement — although as a matter of fact there is little doubt that it did, for Kusmiersky could not get the tax liability reduced without hiring a lawyer and the usual method of compensating a lawyer for getting real estate taxes reduced is to give him a percentage of the reduction. ICM’s approval of the contract with the plaintiffs would (assuming that Kusmiersky was ICM’s agent in paying past-due taxes) in any event bind ICM on the contract by the doctrine of ratification, whether or not the agent had authority to make the contract in the first place. See, e.g., Johnston v. Suckow, 55 Ill.App.3d 277, 279-80, 12 Ill.Dec. 846, 849, 370 N.E.2d 650, 653 (1977).
As an original matter one might wonder why ratification should make the principal liable. The effect seems to be to confer a windfall on the other party to the contract, who did not bargain for this additional promisor. The explanations that have been offered are conclusional. See, e.g., Seavey, Handbook of the Law of Agency 57-59 (1964). The best explanation may be that the principal would not have ratified the contract unless he had seen a commercial advantage in doing so, and that advantage would be less if the ratification had no binding effect. Ordinarily a principal ratifies an agent’s unauthorized transaction in order to protect the principal’s relationship with the other party to the transaction, usually a customer or supplier; and for ratification to have this protective effect it has to be more than an idle gesture, signifying nothing because unenforceable. In any event Illinois law is clear in binding the principal to contracts of his agent that he has ratified.
On either theory — actual authorization to make the fee arrangement, or ratification — the existence of an agency relationship is the key question and is a question of fact, Phipps v. Cohn, 139 Ill.App.3d 210, 212, 93 Ill.Dec. 761, 763, 487 N.E.2d 428, 430 (1985), which, if contestable and contested, as it was here, could not be taken away from the jury by granting a motion for summary judgment. E.g., American Nurses’ Ass’n v. Illinois, 783 F.2d 716, 726 (7th Cir.1986). Of course we do not mean to imply that there was an agency relationship — only that there is a jury issue on the question. There may indeed not have been such a relationship. Some evidence that there was not is the provision in the original lease to Kassuba that he would not have any recourse against ICM for any losses he might incur in operating under it. Ordinarily gains and losses are either borne by the principal entirely or shared with the agent, for the agent is acting on behalf of the principal, and the principal hopes to make profits and takes the risk of loss as a normal concomitant. But this evidence is not conclusive, not only because it is unclear from the record whether Kusmiersky assumed the *461lease when he replaced Kassuba but also because as we mentioned earlier the lease is not the only document that defined the relationship between ICM and Kusmiersky. The agreement on taxes was separate and may conceivably have made Kusmiersky ICM’s agent at least for the limited purpose of hiring lawyers to seek a reduction in back taxes that threatened ICM’s ability to recover its investment in the property.
We come to the negotiations with Netzky, upon which the plaintiffs’ other theories are founded. One theory is that the plaintiffs had a contract with ICM to be paid $250,000 in exchange for a release of their claim, but we agree with the district judge that they did not. When Kusmiersky offered to pay $250,000 over 10 years (and with no payment of principal in the early years, just interest at 7 percent), Goldstick turned him down; so there was no contract, express or implied. Later Kusmiersky assured Goldstick that something would be worked out. But just what that “something” was is hopelessly vague. Even if Goldstick could reasonably have understood Kusmiersky to mean that the obligation would be unconditional — and maybe he could, as we shall see — there would still be the question of the period over which the money was to be paid. When Kusmiersky spoke to Goldstick about working something out, it was after Goldstick had refused to accept payment conditioned on the property’s turning a profit. We cannot know whether, faced with ICM’s proposal to make payment conditional in this way, Goldstick would gladly have embraced Kusmiersky’s previous offer, for payment over 10 years; or whether he would have accepted payment over this period only if the fee was restored to its original level of $290,000; or whether he would have insisted on much more if it was to be paid over 10 years — bearing in mind that at a discount rate of 10 percent per annum (for example), $290,000 paid in equal annual installments over 10 years is worth only $178,000 today, and $250,000 is worth only $154,000. It would be impossible to fix within reasonable bounds the present value of the contract that Goldstick and Kusmiersky made, if Kusmiersky’s reference to working something out is treated as an offer and Goldstick’s thereafter signing the release as an acceptance. An offer must be more definite to create an enforceable contract.
Although the Uniform Commercial Code has made inroads into the common law principle that a contract that does not fix a price is unenforceable, see UCC § 2-305, it has done so in order to make contracts for the sale of goods at whatever the market price is on the day when the goods are delivered enforceable, and in such a case the court has an objective basis for determining the contract price. That is not the case here and anyway this contract is not governed by the Code. The common law principle that a contract cannot be enforced if its terms are indefinite (see Farnsworth, Contracts § 3.27 (1982)) retains a core of vitality. If people want the courts to enforce their contracts they have to take the time to fix the terms with reasonable definiteness so that the courts are not put to an undue burden of figuring out what the parties would have agreed to had they completed their negotiations. The parties have the comparative advantage over the court in deciding on what terms a voluntary transaction is value-maximizing; that is a premise of a free-enterprise system.
This is not to say that enforcement should be denied if the parties by inadvertence failed to specify some peripheral term. Such omissions both are the unavoidable consequence of the limitations of human foresight and can be repaired by the courts without undue difficulty. But in this case essential terms were missing: the contract price, the period over which that price would be paid, the interest rate, and the repayment profile (e.g., equal annual installments or a balloon payment at the end?). See Whitelaw v. Brady, 3 Ill.2d 583, 590-91, 121 N.E.2d 785, 789-90 (1954); Hintz v. Lazarus, 58 Ill.App.3d 64, 67, 15 Ill.Dec. 546, 548, 373 N.E.2d 1018, 1020 (1978). We are talking about a possible difference in present value of at least $136,000 (more, if the proper discount rate *462would have been even more than 10 percent, which is entirely possible) — the difference between $290,000, Goldstick’s demand, and $154,000, the present value at a 10 percent discount rate of Kusmiersky’s offer (ignoring interest and the balloon feature of the offer, which would tend to offset each other). The difference is too high a fraction of the value of the “contract” to allow a court to conclude that a contract was made and to enforce the contract. See, e.g., Hintz v. Lazarus, supra; Setterlund v. Firestone, 104 Wash.2d 24, 700 P.2d 745, 746 (1985); Dolanson Co. v. Citizens & Southern Nat’l Bank, 242 Ga. 681, 682, 251 S.E.2d 274, 276 (1978).
The plaintiffs have two other strings to their bow. The first is promissory estoppel. If an unambiguous promise is made in circumstances calculated to induce reliance, and it does so, the promisee if hurt as a result can recover damages. See, e.g., Yardley v. Yardley, 137 Ill.App.3d 747, 754, 92 Ill.Dec. 142, 147, 484 N.E.2d 873, 878 (1985), and cases cited there; Powers Construction Co. v. Salem Carpets, Inc., 283 S.C. 302, 322 S.E.2d 30, 33 (1984). Netzky, when he said he wouldn’t close the deal unless the plaintiffs released their claim for the legal fee, unwittingly armed the plaintiffs with a weapon to use in trying to collect the fee. It was two-edged, because if through their intransigence the plaintiffs killed the deal, Kusmiersky and ICM might abandon the property and with it any incentive to pay the plaintiffs without putting them to the expense and uncertainty of litigation. But it was a weapon, and one that Kusmiersky, acting for ICM (for there is no question that in these negotiations Kusmiersky was ICM’s agent), induced the plaintiffs to surrender by promising to work things out. Made in the context of Goldstick’s refusal to make his and Smith’s fee doubly contingent (it had been contingent on their getting the back taxes reduced, and now ICM wanted to make it contingent on the property’s showing profits), this promise could be understood (or so a reasonable jury conceivably might find) as a waiver of the condition that the property turn a profit. Although much was left in doubt — in particular the payment period — it is a question for a jury whether Kusmiersky made a promise that estops his principal, ICM, to deny that it was assuring Goldstick that the plaintiffs’ fee would be paid by ICM even if the property never showed a profit.
The part of the district judge’s opinion in which he rejected the plaintiffs’ argument for promissory estoppel is very brief and essentially all it says is that the promise was ambiguous because the question whether payment would be contingent on the property’s turning a profit was left unresolved. However, a jury could find that Goldstick reasonably understood Kus-miersky’s statement that something would be worked out to mean that ICM would not insist on conditioning payment on profit. If so, Kusmiersky could be understood to be promising to do better than his previous offer of $250,000 over 10 years with interest at 7 percent but no principal the first few years. How much better is unclear, but that is not critical. If the minimum value of a promise is unambiguous, that should be enough to work an estoppel to the extent of that value, though no more.
Although to our knowledge the precise issue has not arisen before, we find nothing in principle or in the cases to exclude recovery on a theory of promissory estoppel when the promise is unambiguous on the downside but not the upside. In Yardley there was no promise, period. In Levitt Homes Inc. v. Old Farm Homeowner’s Ass’n, 111 Ill.App.3d 300, 314-15, 67 Ill. Dec. 155, 165, 444 N.E.2d 194, 204 (1982), the alleged promise was only a statement of intentions, and there was a written disclaimer of any promises. In Ziese v. Ramada Inns, Inc., 463 F.2d 1058, 1060 (7th Cir.1972), we implied that if the defendant had told the plaintiff “Quit worrying, you’ve got a deal,” there would have been (much as in this case) a promissory estoppel. The evidence held to create a triable issue in Swansea Concrete Products, Inc. v. Distler, 126 Ill.App.3d 927, 933, 81 Ill.Dec. 688, 692-93, 467 N.E.2d 388, 392-93 (1984), was of a promise no more definite *463than what the jury could find had been the minimum promise in this case, while evidence of a less definite promise was held to create a triable issue in Bank Computer Network Corp. v. Continental Illinois Nat’l Bank & Trust Co., 110 Ill.App.3d 492, 497-99, 66 Ill.Dec. 160, 165-66, 442 N.E.2d 586, 591-92 (1982).
The fact that Goldstick rejected Kusmier-sky’s offer does not show that he did not rely on it. He may have regarded it as a floor, once the possibility of its being only contingent was dispelled by the assurance (if that is what it was) that something would be worked out. Thus reassured, he may have been sufficiently encouraged to execute the release that Kusmiersky and ICM needed to complete the deal with Netzky, even though he hoped to get a better settlement than Kusmiersky had offered. At least this is sufficiently plausible to create a triable issue of reliance as well as of the existence of a promise.
Should the plaintiffs prevail on this theory, it will become necessary to consider the proper measure of damages. As a matter of strict logic one might suppose that since the promise was not accepted, the plaintiffs, even if they succeeded in establishing a promissory estoppel, would not necessarily be able to recover the value of the promise — the $250,000 discounted to present value. It would seem they would have to show their actual damages: what they gave up when they signed the release in reliance on Kusmiersky’s promise. Maybe they gave up little. Maybe in the end they would have signed the release without any promises, knowing that if they did not the Netzky deal might fall through and with it all prospects for recovering their fee without a lawsuit. Or maybe as ICM argues the plaintiffs were more optimistic about the prospects of the property’s showing a profit than they have let on in this lawsuit.
This approach treats promissory estoppel as a tort doctrine for purposes of damages, though it is conventionally classified as a contract doctrine, see, e.g., Restatement (Second) of Contracts § 90 (1979); its cousin, equitable estoppel (which requires proof of misrepresentation), is the tort doctrine, A/S Apothekernes Laboratorium for Specialpraeparater v. I.M.C. Chemical Group, Inc., 725 F.2d 1140, 1142 (7th Cir. 1984). Some cases do award just the tort measure of damages in promissory estoppel cases, rather than giving the plaintiff the value of the promise, which he would be entitled to in a breach of contract action. See, e.g., Young v. Johnston, 475 So.2d 1309, 1314 (Fla.App.1985). In Gerson Electric Construction Co. v. Honeywell, Inc., 117 Ill.App.3d 309, 312-13, 72 Ill.Dec. 851, 853, 453 N.E.2d 726, 728 (1983), however, the Illinois Appellate Court held that a plaintiff in a promissory estoppel case could recover damages for the profits he would have made had the defendant kept his promise — provided such an award was necessary to do justice to the plaintiff. Those were the only damages alleged by the plaintiff, so it was either lost profits or nothing. Walters v. Marathon Oil Co., 642 F.2d 1098, 1100-01 (7th Cir. 1981), was a similar case, and was decided the same way, though under Indiana rather than Illinois law. In the present case an alternative damage measure is available which may or may not yield a positive award, depending on whether the plaintiffs can prove that ICM would have paid something for the release of their claims.
There is much to be said for using the value of the promise as the measure of damages, simply on grounds of simplicity. The alternative would be to try to reconstruct the hypothetical bargaining between ICM and the plaintiffs — the former threatening to walk away from the deal, the latter threatening to kill the deal: a classic bilateral monopoly situation, the indetermi-nacies of which we discussed in Chicago & North Western Transportation Co. v. United States, 678 F.2d 665, 670 (7th Cir. 1982), and Milbrew, Inc. v. Commissioner of Internal Revenue, 710 F.2d 1302, 1306-07 (7th Cir.1983). In addition, an expectation measure of damages will frequently cover opportunity costs (which are real, but not out-of-pocket, costs) that the reliance measure would miss. See, e.g., Feinman, *464Promissory Estoppel and Judicial Method, 97 Harv.L.Rev. 678, 685-88 (1984).
Consistently with our leanings, the Restatement of Contracts implies, if somewhat unclearly, that the value of the promise is the presumptive measure of damages for promissory estoppel, to be rejected only if awarding so much would be inequitable. See comment d to section 90. That would not be inequitable here when we recall that no one has ever questioned the reasonableness of the plaintiffs’ original — and higher — legal fee. But we need not decide whether this is the correct approach under the law of Illinois. The issue may wash out on remand; anyway we would want the district judge’s views before deciding a difficult question of state law.
For the same reason we decline to reach ICM’s alternative argument, which the district court did not reach either, that the plaintiffs’ claim for promissory estoppel is barred by the statute of frauds because the promise that the plaintiffs seek to enforce ($250,000 over 10 years) could not be completed within one year. Two difficult questions are presented by the statute of frauds defense. The first is whether the promise was within the statute of frauds at all. Although the Illinois statute requires a writing for “any agreement that is not to be performed within the space of one year from the making thereof,” Ill.Rev.Stat. ch. 59, 111, the Illinois courts have interpreted this to mean, incapable of being performed within one year; if performance though unlikely is possible, that is enough to take the agreement outside the grasp of the statute. Martin v. Federal Life Ins. Co., 109 Ill.App.3d 596, 604, 65 Ill.Dec. 143, 149, 440 N.E.2d 998, 1004 (1982); In re Marriage of Strand, 86 Ill.App.3d 827, 831, 42 Ill.Dec. 37, 40, 408 N.E.2d 415, 418 (1980). Some promises, for example a promise to work for an employer for five years, really cannot be performed within a year. But a promise to pay $250,000 over 10 years can be — by paying the full $250,000 the first year. There is no suggestion that the interest rate (7 percent) was so attractive to the plaintiffs that they would have refused to accept early payment; indeed it is obvious that 7 percent was a below-market rate.
Courts tend to take the concept of “capable of full performance” quite literally. See Farnsworth, supra, § 6.4. They do this because they find the one-year limitation irksome. The limitation is apparently founded on a concern with the tendency of evidence to go stale with the passage of time, but the foundation is weak because the limitation applies even if the promise is broken the day after it is made and suit on it is brought immediately. The Illinois courts evidently share the prevailing dislike for the provision, for in a case very like this the Illinois Appellate Court held that the statute of frauds was not a bar. See Hubbard v. Logsdon, 56 Ill.App.3d 366, 373-74, 14 Ill.Dec. 296, 301, 372 N.E.2d 101, 106 (1978). As here, the agreement called for payment over 10 years, and there was evidence that full payment at once would have been acceptable to the prom-isee. Direct evidence of acceptability has not been offered here but the inference from the low interest rate is very strong. We very much doubt that an Illinois court faced with the statute of frauds defense in this case would hold that the defendant’s alleged promise is not within it, although we hesitate to so hold when the district judge has not ruled on the question.
We may seem to have overlooked an even more direct route to the conclusion that the statute of frauds is not a bar: our decision in R.S. Bennett & Co. v. Economy Mechanical Industries, Inc., 606 F.2d 182, 187-89 (7th Cir.1979), which held that Illinois’ statute of frauds does not apply in promissory estoppel cases. We hesitate to rely on Bennett because there is a question whether it is a correct statement of Illinois law. Two decisions of the Illinois Supreme Court had rejected the position that we later adopted in Bennett. One of them, it is true, Ozier v. Haines, 411 Ill. 160, 103 N.E.2d 485 (1952), preceded the modern development of promissory estoppel. And the other, Sinclair v. Sullivan Chevrolet Co., 31 Ill.2d 507, 202 N.E.2d 516 (1964), did not discuss promissory estoppel, though *465a comparison with the Illinois Appellate Court’s opinion in the case shows that it implicitly rejected it. More important, Sinclair involved an employment contract, where the use of promissory estoppel to get around the statute of frauds is particularly troublesome. Employment at will (i.e., without a contract of employment) remains the dominant type of employment relationship in this country, and would be seriously undermined if employees could use the doctrine of promissory estoppel to make alleged oral contracts of employment enforceable. Reliance is easily, perhaps too easily, shown in the employment setting. Agreeing to work for a particular employer, thereby giving up alternative opportunities for employment, can easily be described as reliance on the employer’s alleged oral promises concerning the terms of employment.
In Bennett, a case involving a contract for the sale of goods, we relied heavily on a then-recent decision of the Illinois Appellate Court, Jenkins & Boiler Co. v. Schmidt Iron Works, Inc., 36 Ill.App.3d 1044, 344 N.E.2d 275 (1976), but we failed to cite the even more recently decided case of Libby-Broadway Drive-In, Inc. v. McDonald’s System, Inc., 72 Ill.App.3d 806, 809-11, 28 Ill.Dec. 802, 805, 391 N.E.2d 1, 4 (1979), which both held that the statute of frauds is a defense to promissory estoppel and described Jenkins as a case outside of rather than rejecting the statute of frauds — and indeed the opinion in Jenkins is quite unclear in this respect. An even more recent case shows an evident reluctance to allow the statute of frauds to defeat a claim of promissory estoppel, but, as in Jenkins, it is unclear whether the court thought that the promise, even if contractual, would have been within the scope of the statute. See Hux v. Woodcock, 130 Ill.App.3d 721, 724, 86 Ill.Dec. 44, 47, 474 N.E.2d 958, 961 (1985). Many cases from other jurisdictions support Bennett. See Allen M. Campbell Co. v. Virginia Metal Industries, Inc., 708 F.2d 930 (4th Cir.1983); Farnsworth, supra, § 6.12. They reflect a larger trend: as Farnsworth remarks, the statute of frauds “has been the subject of constant erosion.” Id., § 6.1, at 373. Maybe the cases that erode it in the area of promissory estoppel exaggerate the ability of courts to distinguish true oral promises from fabricated promises and language misunderstood as promissory. Maybe they overlook the simple logical point that the promisee should know the promise is unenforceable because of the statute of frauds and therefore shouldn’t rely on it. But for what it is worth the logic of such cases is that the element of reliance that is required to make a promise enforceable on grounds of estop-pel takes the place of a writing in giving adequate assurance that the promise has not been fabricated, or misunderstood without fault on the part of the promisor. The “promisee” takes a great risk if he gives up something in purported reliance on a nonexistent promise, merely to bolster his claim that there was a promise. See Wisconsin Knife Works v. National Metal Crafters, 781 F.2d 1280, 1287 (7th Cir. 1986). There may, however, be settings where this is not true, as we have suggested might be the case in the employment setting, the setting of the Sinclair case.
The picture in Illinois is decidedly mixed, especially when we reflect that both Bennett and Jenkins involved the statute of frauds in the Uniform Commercial Code, though Hux did not. Ceres Illinois, Inc. v. Illinois Scrap Processing, Inc., 130 Ill. App.3d 798, 805-06, 86 Ill.Dec. 48, 52-53, 474 N.E.2d 1245, 1249-50 (1984), contains intimations — no stronger word is possible— that the statute of frauds will trump a claim of promissory estoppel in a case not governed by the Code. Ceres must give some pause to anyone who argues that the Illinois Supreme Court really would be willing to reconsider, in areas not governed by the Uniform Commercial Code, its old cases holding the statute of frauds applicable to claims of promissory estoppel. On the other hand, the movement away from the statute of frauds is not a development peculiar to the Code. It is reflected for example in section 139(1) of the Second Restatement, which provides that an oral promise can be *466the basis of a promissory estoppel “notwithstanding the Statute of Frauds if injustice can be avoided only by enforcement of the promise.” Subsection (2) gives substance to the concept of avoiding injustice by specifying the considerations relevant to whether the promise should be enforced despite the statute of frauds, considerations such as whether “the making and terms are otherwise established by clear and convincing evidence.” § 139(2)(c). The effect is to create a more flexible standard which nonetheless preserves some judicial power to keep out of the hands of a jury a case based solely on evidence of an oral promise and of reasonable reliance thereon. Would the Illinois Supreme Court embrace this standard? It really is not possible to say. We have found only two recent cases in which that court dealt with the statute of frauds, and though in both the defense was rejected, they do not provide strong evidence that the court has been swept up in the “constant erosion” that has eaten away the statute in other jurisdictions. See Sierens v. Clausen, 60 Ill.2d 585, 328 N.E.2d 559 (1975); Lee v. Central Nat’l Bank & Trust Co., 56 Ill.2d 394, 308 N.E.2d 605 (1974).
If forced to guess what the Illinois Supreme Court would do in a case like this we would guess (a word used deliberately) that it would hold the statute of frauds applicable to promissory estoppel (since to do otherwise would require the court to overrule two of its decisions), and thus would disapprove Bennett. But we do not think we should guess, at least without the benefit of the district judge’s views. Although the issue is a pure issue of law which ordinarily we would feel free to resolve without the benefit of the district court’s view, we follow the practice in diversity cases of giving substantial deference to the district judge’s interpretation of the law of the state in which the judge sits. See, e.g., American Motorists Ins. Co. v. Trane Co., 718 F.2d 842, 845 (7th Cir.1983); Morin Building Products Co. v. Baystone Construction, Inc., 717 F.2d 413, 417 (7th Cir.1983); Murphy v. White Hen Pantry Co., 691 F.2d 350, 354 (7th Cir.1982). This precept is especially applicable here, given Judge Shadur’s long experience in commercial practice in Illinois before his appointment to the bench. Ordinarily he would be bound by a controlling decision of this court, such as Bennett, but from what we have said it should be clear that we regard that decision as open to reexamination. Another reason for our refusal to bite the bullet is that the number of theories of liability and of defenses in this case makes it quite likely that the statute of frauds issue will wash out on remand — assuming the parties are not able to settle the case on the basis of the additional clarification of the merits provided by this opinion. We are not eager to decide difficult issues of state law, especially when the decision may not be necessary to the final resolution of the lawsuit, as it may not be here. Finally, and decisively, there is a good chance that the particular promise is outside of the statute of frauds. Cases such as Jenkins and Hux suggest that Illinois courts will strain to avoid allowing the statute of frauds to bar a claim based on promissory estoppel, and they could avoid it in a case like this by holding that the promise, being a mere promise to pay, could be performed within a year though the defendant had no duty to pay so quickly.
The last string on the plaintiffs’ bow is restitution. Assume that if Kusmiersky had broken his contract with the plaintiffs before Netzky came on the scene, they would have had no recourse against ICM because ICM had paid Kusmiersky for the benefits they had conferred; still, by the time Netzky came into the picture, everything had changed. Kusmiersky was on his way out and ICM was on the verge of abandoning the property. Netzky appeared as a possible savior but attached two conditions to putting new capital into the property: that the back taxes be paid by ICM and that the plaintiffs release their claim. The back taxes were lower as a result of the plaintiffs’ efforts, and while it is true that ICM had paid Kusmiersky to pay off the taxes he had not done so. They loomed as an obligation of ICM — an obli*467gation less onerous as a result of the plaintiffs’ efforts than it otherwise would have been. In addition, by executing the release the plaintiffs conferred a benefit on ICM by making it possible to close the deal with Netzky.
A person who confers benefits gratuitously — officiously—obtains no legal claim for compensation. If Goldstick had appeared at a meeting of ICM’s shareholders and serenaded them with his violin, and the shareholders had listened raptly, still Goldstick would have no claim of restitution against ICM for benefits conferred; he would have had to negotiate a contract with ICM in advance. But when a businessman confers benefits in circumstances where payment is normal and any inference of altruism can be rejected, a claim for restitution, or “quasi-contract,” arises. Industrial Lift Truck Service Corp. v. Mitsubishi Int’l Corp., 104 Ill.App.3d 357, 360, 60 Ill.Dec. 100, 103, 432 N.E.2d 999, 1002 (1982). That describes this case — or so a reasonable jury could find. By executing the release the plaintiffs conferred a benefit on ICM. The release allowed the deal with Netzky to go through, which gave ICM hope (though later dashed) of recovering its investment in Rolling Meadows. This is not the kind of benefit that one businessman or professional renders another without expectation of payment.
A second benefit was conferred but its legal significance is different. As a result of Kusmiersky’s departure from the scene without paying the back taxes, by the time of the negotiations with Netzky it was clear that the plaintiffs’ legal work had conferred substantial benefits on ICM by reducing the amount of money it would have to pay in back taxes in order to get Netzky to put in additional capital. In this way too the plaintiffs facilitated the transaction, a transaction that conferred benefits on ICM ex ante though not ex post. But they cannot recover these benefits on a theory of restitution. The reason is the original contract that the plaintiffs signed with Kusmiersky to do the legal work that resulted in the reduction of the taxes that were past due. If the contract was authorized or ratified by ICM as Küsmiersky’s principal, it defined the plaintiffs’ rights against ICM and they cannot rewrite the contract by suing for restitution rather than breach of contract. Industrial Lift Truck Service Corp. v. Mitsubishi Int’l Corp., supra, 104 Ill.App.3d at 360-61, 60 Ill.Dec. at 103, 432 N.E.2d at 1002.
If on the other hand the contract was just between the plaintiffs and Kusmier-sky, the plaintiffs cannot bring ICM in through the back door by pointing out that their contractual performance conferred benefits on ICM. To illustrate, if you do work pursuant to a contract with X, you don’t expect that Y, a nonparty, will pay you if X defaults, merely because Y was benefited by your work; and expectation of payment is an essential element of a claim for restitution. For the doctrine of restitution does not make altruism a paying proposition. As its synonym “contract implied in law” brings out, it allows damages to be recovered in settings where (unlike our example of the unsolicited serenade) the parties would have agreed that there was a contractual obligation had the point occurred to them. If Kusmiersky was merely a tenant of ICM and not an agent, ICM would not have expected to have to make good on a claim by the plaintiffs arising from the performance of their contract with him; and this, as the Illinois cases make abundantly clear, defeats a claim of restitution for the benefits conferred on ICM by that performance. See, e.g., Daley v. G’Sell, 102 Ill.App.3d 548, 551, 58 Ill. Dec. 524, 527, 430 N.E.2d 556, 559 (1981); Premier Electrical Construction Co. v. LaSalle Nat’l Bank, 132 Ill.App.3d 485, 496, 87 Ill.Dec. 721, 729-30, 477 N.E.2d 1249, 1257-58 (1984).
The district judge emphasized another point: that under a functional definition of ownership ICM was not the owner of the Rolling Meadows property, so much control had it given up to Kusmiersky and intended to give up to his successor, Netzky, and therefore it was not liable under Illinois law for payment of back taxes on the prop*468erty. We do not understand the significance of this point. Whether under Illinois law ICM had to pay the back taxes or Netzky would have had to pay them has no economic significance. Someone had to pay them — that, or the property would have to be abandoned. Netzky was unwilling to pay them. So ICM’s choice was either to pay them or abandon the property and with it all hope of recouping its investment. Since it wasn’t prepared to abandon the property it had to pay the back taxes, as a matter of economics if not law; and this being so it derived a real benefit from the fact that the taxes were lower thanks to the efforts of Goldstick and Smith. It just is not the kind of benefit for which compensation can be obtained under the law of restitution.
But concerning the separate benefit that the plaintiffs conferred on ICM by releasing their contract claims against it and thereby enabling the deal with Netzky to go through, the plaintiffs may be entitled to restitution, for they would expect to be paid for the release, and paid by ICM, the beneficiary. So far as liability is concerned, it makes no difference (it could make a difference with respect to damages) whether Goldstick and Smith could have slapped a lien on the property to enforce the payment of their legal fee by Kusmier-sky. Netzky refused — it doesn’t matter why — to go through with a deal that ICM very much wanted unless ICM procured a release of the lawyers' claim, and by executing that release Goldstick and Smith conferred a benefit on ICM for which they are entitled to be compensated. Of course it is possible that the only compensation they sought and received was contingent on the property’s yielding a profit; if so, though they are entitled to compensation, that compensation would be zero. But that is an issue in hot dispute, and a jury could find that Goldstick and Smith never accepted an offer of contingent payment.
If the plaintiffs conferred a benefit for which they reasonably expected to be compensated, and not on a contingent basis either, the next question is the “reasonable worth” of the benefit. In re Estate of Milborn, 122 Ill.App.3d 688, 690, 78 Ill.Dec. 241, 244, 461 N.E.2d 1075, 1078 (1984). It might seem to be zero, since in fact ICM never made a profit from the deal with Netzky. But that is to take an incorrect, ex post perspective. A lottery ticket is not worth zero when it is bought, merely because it later turns out not to be the winning ticket. The question is what ICM would have paid for the release to close the deal with Netzky — what the release was worth to it at the time that it obtained it from Goldstick and Smith. That will be an issue for trial.
Other grounds presented by the defendants for upholding the district judge’s decision are unpersuasive. To summarize, we agree with the district judge that the plaintiffs have no claim for breach of an express contract made at the time of the deal with Netzky. They may however have claims for promissory estoppel, restitution, and breach of a contract made by Kusmiersky but authorized or ratified by ICM. We do not say they do. We hold only that they have raised genuine issues of material fact which preclude summary judgment for the defendants.
REVERSED AND REMANDED.