One of the most difficult areas of contract law concerns the enforceability of letters of intent and other preliminary agreements, and in particular the subset of such agreements that consists of agreements to negotiate toward a final contract. See Steven J. Burton & Eric G. Andersen, Contractual Good Faith: Formation, Performance, Breach, Enforcement §§ 8.4-8.5 (1995); 1 E. Allan Farnsworth, Farnsworth on Contractsss 3.8a-3.8c, 3.26b-3.26c, pp. 186-207, 328-352 (1990); 1 Corbin on Contracts §§ 2.8-2.9, pp. 131-162 (Joseph M. Perillo ed., rev. ed.1993). When if ever are such agreements enforceable as contracts? If they are enforceable, how is a breach to be determined? Is "breach" even the right word? Or is the proper rubric "bad faith"? Could the duty of good faith negotiation that a letter of intent creates be a tort duty rather than a contract duty, even though created by a contract? And can the victim of bad faith ever get more than his reliance damages? These questions lurk on or just beneath the surface of the principal appeal, which is from a judgment by the district court, after a bench trial, finding that the defendant had not acted in bad faith and was not liable for any damages to the plaintiff. Federal jurisdiction is based on diversity of citizenship, and the substantive issues are governed by Illinois law.

The defendant, Zenith Data Systems Corporation (ZDS), owned Heath Company, the manufacturer of "Heathkits"--do-it-yourself kits for building stereos, computers, and other electronic systems. Heath was losing money, and in 1990 ZDS decided to sell it together with a related subsidiary, Prokit, but we shall generally use "Heath" to denote both Heath and Prokit. ZDS hired an investment banker to find someone who would buy Heath at a price, then estimated at $11 million, at which ZDS would lose no more than $6 million on the sale, the loss being calculated with reference to the net asset value shown for Heath on the books of ZDS. One of the prospects that the investment banker found was the plaintiff, Venture Associates Corporation. Apparently the investment banker did not conduct a credit check of Venture. Instead he relied on a representation by Venture that its most recent acquisitions had been of companies with revenues of $55 and $97 million.

On May 31, 1991, Venture sent a letter to the investment banker, for forwarding to ZDS, proposing to form a new company to acquire Heath for $5 million in cash, a $4 million promissory note, and $2 million in preferred stock of the new company--a total of $11 million, the price ZDS was seeking. The letter stated that it was "merely a letter of intent subject to the execution by Seller and Buyer of a definitive Purchase Agreement (except for the following paragraph of this letter, which shall be binding ... ) [and] does not constitute a binding obligation on either of us." The following paragraph stated that "this letter is intended to evidence the preliminary understandings which we have reached regarding the proposed transaction and our mutual intent to negotiate in good faith to enter into a definitive Purchase Agreement, and [ZDS] hereby agree[s] that, pending execution of a definitive Purchase Agreement and as long as the parties thereto continue to negotiate in good faith," ZDS shall not "solicit, entertain, or encourage" other offers for Heath or "engage in any transaction not in the ordinary course of business which adversely affects" Heath's value.

The letter invited ZDS to sign it. ZDS refused, but did write Venture on June 11 stating that "we are willing to begin negotiations with Venture Associates for the acquisition of the Heath Business based in principle on the terms and conditions outlined in" Venture's May 31 letter. The next day, Venture wrote ZDS accepting the proposal in the June 11 letter.

Let us pause here and ask what if any enforceable obligations were created by this correspondence. The use of the words "in principle" showed that ZDS had not agreed to any of the terms in Venture's offer. This is the usual force of "in principle" and is the meaning we gave the term in Skycom Corp. v. Telstar Corp., 813 F.2d 810, 814 (7th Cir. 1987); cf. Arcadian Phosphates, Inc. v. Arcadian Corp., 884 F.2d 69, 71-73 (2d Cir.1989); Itek Corp. v. Chicago Aerial Industries, Inc., 248 A.2d 625, 627 (Del.1968). That construal is reinforced by Venture's statement in its letter that the only binding obligation that ZDS's signing the letter would create would be an obligation of both parties to negotiate in good faith and, on ZDS's part, a further obligation not to entertain other offers or strip Heath of its assets--obligations that might be thought in any event entailed by the concept of good faith but which Venture thought useful to spell out. When last this case was before us, on appeal from an earlier decision (812 F. Supp. 788 (N.D.Ill.1992)) by a different district judge dismissing Venture's suit, we held that the exchange of letters had established a binding agreement to negotiate in good faith toward the formation of a contract of sale. The doctrine of the law of the case makes this determination binding in this second round of appeals in the absence of exceptional circumstances not here shown.

We therefore have no occasion to revisit the determination on this appeal, though we are mindful of the powerful argument that the parties' undertakings were too vague to be judicially enforceable--since courts, unlike the National Labor Relations Board or labor arbitrators, are not well equipped to determine whether people are negotiating with each other in good faith--or, if not, that the proper rubric for determining enforceability in such a case is not contract but promissory estoppel. See, e.g., Quake Construction, Inc. v. American Airlines, Inc., 141 Ill. 2d 281, 152 Ill. Dec. 308, 322-23, 565 N.E.2d 990, 1004-05 (1990); Skycom Corp. v. Telstar Corp., supra, 813 F.2d at 817; Arcadian Phosphates, Inc. v. Arcadian Corp., supra, 884 F.2d at 73-74. This is the approach taken in some states. 1 Farnsworth, supra, § 3.26b, pp. 329-34. But interpreting Illinois law, we have held that agreements to negotiate toward the formation of a contract are themselves enforceable as contracts if the parties intended to be legally bound. See, besides the previous panel opinion in this case (987 F.2d 429, 432 (7th Cir.1993)), A/S Apothekernes Laboratorium v. I.M.C. Chemical Group, Inc., 873 F.2d 155, 158 (7th Cir.1989). Unfortunately, this is one of those issues of state common law on which the only decisions are either decisions by a federal court, or decisions (in this case one decision, Itek Corp. v. Chicago Aerial Industries, Inc., supra, 248 A.2d at 627, 629) by the courts of another state. The two federal cases that we have cited (the previous panel decision and the Apothekernes decision) cite between them only one Illinois decision, Inland Real Estate Corp. v. Christoph, 107 Ill.App.3d 183, 63 Ill. Dec. 9, 11, 437 N.E.2d 658, 660 (1981), and it does not contain a clear statement of the proposition. Nor have we found any Illinois decisions that do.

ZDS has not asked us to reexamine our decisions, so we shall let them stand, quite apart from the force exerted by the doctrine of the law of the case. Right or wrong, the position they take cannot be said to be unreasonable. The process of negotiating multimillion dollar transactions, like the performance of a complex commercial contract, often is costly and time-consuming. The parties may want assurance that their investments in time and money and effort will not be wiped out by the other party's footdragging or change of heart or taking advantage of a vulnerable position created by the negotiation. Feldman v. Allegheny Int'l, Inc., 850 F.2d 1217, 1221 (7th Cir.1988); Runnemede Owners, Inc. v. Crest Mortgage Corp., 861 F.2d 1053, 1056 (7th Cir.1988). Suppose the prospective buyer spends $100,000 on research, planning, and consultants during the negotiation, money that will have bought nothing of value if the negotiation falls through, while the seller has spent nothing and at the end of the negotiation demands an extra $50,000, threatening to cancel the deal unless the buyer consents. This would be an extortionate demand, and, as it is profoundly unclear whether it would be independently tortious, E. Allan Farnsworth, "Precontractual Liability and Preliminary Agreements: Fair Dealing and Failed Negotiations," 87 Colum.L.Rev. 217, 221-22 (1987), parties to a negotiation would want a contractual remedy. But they might prefer to create one in the form of a deposit or drop fee (what in publishing is called a "kill fee"), rather than rely on a vague duty to bargain in good faith. That is one reason why the notion of a legally enforceable duty to negotiate in good faith toward the formation of a contract rests on somewhat shaky foundations, though some contracts do create such a duty, see, e.g., PSI Energy, Inc. v. Exxon Coal USA, Inc., 17 F.3d 969, 972 (7th Cir.1994), which shows that some business people want it. Anyway we have crossed these bridges, for the time being anyway.

In the prior round of appeals Venture's principal argument, which this court rejected, was that the parties' exchange of nonidentical contract drafts during the negotiation period had created a binding contract on the terms, specifically the $11 million sale price, in Venture's letter of May 31. The argument failed because of the "mirror image" rule; neither party had accepted the terms in the other party's offer. Venture continues to insist that it should be awarded damages equal to the difference between the $11 million price in the letter of May 31 and the current value of Heath (which was sold in 1995 to another purchaser, though we have not discovered at what price). Does this mean that Venture is treating the letter of intent as the contract of sale? It does not. Damages for breach of an agreement to negotiate may be, although they are unlikely to be, the same as the damages for breach of the final contract that the parties would have signed had it not been for the defendant's bad faith. If, quite apart from any bad faith, the negotiations would have broken down, the party led on by the other party's bad faith to persist in futile negotiations can recover only his reliance damages--the expenses he incurred by being misled, in violation of the parties' agreement to negotiate in good faith, into continuing to negotiate futilely. But if the plaintiff can prove that had it not been for the defendant's bad faith the parties would have made a final contract, then the loss of the benefit of the contract is a consequence of the defendant's bad faith, and, provided that it is a foreseeable consequence, the defendant is liable for that loss--liable, that is, for the plaintiff's consequential damages. Burton & Andersen, supra, § 8.4.2.3, pp. 364-66. The difficulty, which may well be insuperable, is that since by hypothesis the parties had not agreed on any of the terms of their contract, it may be impossible to determine what those terms would have been and hence what profit the victim of bad faith would have had. See 1 Farnsworth, supra, § 3.26b, p. 332. But this goes to the practicality of the remedy, not the principle of it. Bad faith is deliberate misconduct, whereas many breaches of "final" contracts are involuntary--liability for breach of contract being, in general, strict liability. It would be a paradox to place a lower ceiling on damages for bad faith than on damages for a perfectly innocent breach, though a paradox that the practicalities of proof may require the courts in many or even all cases to accept.

After Venture's confirmatory letter of June 12, the parties negotiated for six months. At the end of that time, with no sale contract signed, ZDS broke off the negotiations on the ground that Venture was refusing to furnish third-party guaranties of its post-closing financial obligations (namely to pay the $4 million promissory note and to honor the terms of the preferred stock) and agree to certain post-closing price adjustments. Early in the negotiations ZDS had asked Venture for financial information, but it had then abandoned the demand, for a time at least, content it seemed with the representations of acquisition prowess that Venture had made to the investment banker. The district judge found that as the negotiations looking to a final sale contract continued, ZDS became anxious about Venture's financial solidity and renewed its request for financial information. Venture did not comply with the request, so ZDS decided it would need a third-party guaranty.

The price adjustments that ZDS demanded were of two types: adjustments reflecting an increase in the value of Heath's inventory since the onset of negotiations, and adjustments reflecting changes in the estimated size of a debt that Prokit owed ZDS and that the buyer, Venture, was to pay. ZDS's estimate that it would lose $6 million on the sale of Heath (including Prokit) had been based on a valuation of the company at $17 million and a sale price of $11 million. It wanted the contract for the sale of Heath to Venture to adjust the price so that the difference between the value of Heath (including the Prokit receivable) at the time of sale and the purchase price would not exceed $6 million no matter what happened to that value. This would have entailed, under ZDS's calculations, an increase in the contract price from $11 million to $14 million.

Venture argues that since the letter of intent--its May 31 letter, which ZDS accepted in principle--made no reference to third-party guaranties or contract-price adjustments, ZDS exhibited bad faith by insisting on these terms to the point of impasse. This argument overlooks the difference between an agreement to negotiate a contract and the contract to be thrashed out in those negotiations. The agreement to negotiate does not contain the terms of the final agreement. Otherwise it would be the final agreement. A preliminary agreement might contain closed terms (terms as to which a final agreement had been reached) as well as open terms, and thus be preliminary solely by virtue of having some open terms. 1 Farnsworth, supra, § 3.8a, pp. 187-88. The parties would be bound by the closed terms. There were no such terms here.

Venture has another argument--that ZDS decided to pull out after the negotiations began because Heath's fortunes began to improve, and to this end imposed new conditions that it knew Venture would not accept. There is a fair amount of evidence that Heath's fortunes were improving, but only a bit of evidence, which the district judge was entitled to find outweighed by contrary evidence, that this improvement made ZDS reconsider its decision to sell the company. Since ZDS had not agreed on the sale price, it remained free to demand a higher price in order to reflect the market value of the company at the time of actual sale. Feldman v. Allegheny Int'l, Inc., supra, 850 F.2d at 1223. Self-interest is not bad faith. Id. Not having locked itself into the $11 million price, ZDS was free to demand as high a price as it thought the market would bear, provided that it was not trying to scuttle the deal, Chase v. Consolidated Foods Corp., 744 F.2d 566, 571 (7th Cir.1984), or to take advantage of costs sunk by Venture in the negotiating process. The qualification is vital. If the market value of Heath rose, say, to $25 million, ZDS would not be acting in bad faith to demand that amount from Venture even if it knew that Venture would not go so high. ZDS would be acting in bad faith only if its purpose in charging more than Venture would pay was to induce Venture to back out of the deal. The analogy to constructive discharge in employment law, or, for theology buffs, the Catholic doctrine of double effect, should be apparent.

Given that the price term was open, moreover, it is hard to see why ZDS would want to bust the deal by insisting on a ridiculously high price or on third-party guaranties that it didn't really want. It could, as we have just pointed out, set a price that reflected Heath's current market value without violating the agreement to negotiate in good faith. Having done so it would presumably be delighted to sell to Venture--because, so far as the record discloses, no other seriously interested prospective purchaser for Heath had appeared during the period of negotiations, and (the district judge could find without committing a clear error) nothing happened during the period of negotiations to make ZDS decide to keep Heath-provided, of course, that Venture was financially responsible. As Venture's financial responsibility was in question, and was an important consideration because this was not a cash purchase, cf. Miller v. LeSea Broadcasting, Inc., 87 F.3d 224, 228 (7th Cir.1996), ZDS did not exhibit bad faith in insisting on the protection that a guaranty by a financially responsible third party would provide. Indeed, the investment banker testified that he had raised the question of a guaranty with Venture before the letter agreement. Venture argues that since ZDS did not demand a guaranty in its "in principle" letter, it agreed to do without one. But this argument, if accepted, would turn the agreement to negotiate into an agreement with some closed terms (such as, no guaranties) and some open ones. It would no longer be a simple agreement to negotiate; it would be an agreement to the closed terms plus an agreement to negotiate toward a resolution of the open ones. Compare 1 Corbin on Contracts, supra, § 2.8(a), p. 133, with id., § 2.8(b), p. 142. An interpretation of the agreement in the present case as containing any closed terms would change the scope of the remand, in which the only issue was whether ZDS was acting in bad faith; for it is only through bad faith that an agreement to negotiate in good faith can be violated.

A finding of good faith, like a finding of negligence or possession, is treated as a finding of fact, and is therefore reviewed under the deferential "clear error" standard. Covey v. Commercial National Bank, 960 F.2d 657, 662 (7th Cir.1992). Venture has made little effort to show clear error. It does point out that the district judge appears to have overlooked internal memoranda of ZDS suggesting that ZDS wanted to raise the price of Heath when it discovered that Heath was worth more than it had thought when it accepted Venture's proposal in principle. But we have explained that this evidence is not evidence of bad faith. Venture argues that ZDS deliberately exaggerated the value of the Prokit receivable in order to sabotage the deal. But the judge's contrary finding was not clearly erroneous, and for the reasons explained earlier we cannot understand why ZDS would want to sabotage the deal unless it could get a better price from someone else or decided it could operate Heath more efficiently than anyone else, after all--and the evidence did not require the judge to find either of these things.

Venture filed a separate appeal challenging the amount of costs awarded by the district court to ZDS as the winning party, but later filed a Rule 60(b) motion in the district court seeking to challenge the award on the basis of newly discovered evidence. The district judge has indicated his willingness to revisit the issue, so in accordance with our usual practice we remand that portion of the case. Brown v. United States, 976 F.2d 1104, 1110-11 (7th Cir.1992). The judgment of the district court is otherwise

AFFIRMED.

I note only one significant area of difference from the majority's otherwise admirable explication of this elusive subject.

The majority states thatif the plaintiff can prove that had it not been for the defendant's bad faith the parties would have made a final contract, then the loss of the benefit of the contract is a consequence of the defendant's bad faith, and, provided it is a foreseeable consequence, the defendant is liable for that loss--liable, that is, for the plaintiff's consequential damages.

Maj. Op. at 278. The majority concedes that this theoretical measure of damages may be impossible of proof because of the obvious difficulty of showing what the terms of the eventual contract would have been absent a breach of the obligation of good faith. The majority, however, defends the principle, if not its practicality, by pointing to the "paradox" of allowing consequential damages for an innocent contract breach but not for the deliberate misconduct of bad faith.

By adopting this principle, the majority certainly does create a disincentive for acting in bad faith in contract negotiations; but the adoption of the principle has another effect which is not so salutary. For we cannot lose sight of the fact that a contract is a deliberate agreement of the parties, frequently marked by certain formalities, establishing their own framework of rules to govern their relations. It seems to create a different paradox to foist the peculiar and special consequences of an agreement on parties who have not in fact agreed. In principle, it cannot be reasonably foreseeable that the parties will ever reach agreement--let alone what the terms of an agreement will be. Good faith is a necessary but not a sufficient condition of agreement. There are many perfectly legitimate reasons for negotiations to fail, even if good faith prevails.

As a matter of policy, I think it is undesirable to force agreement on parties under threat of a bad faith finding and subsequent imposition of consequential damages, the same sanction as would issue from actual agreement. Freedom not to contract should be protected as stringently as freedom to contract. The present case is an excellent example of how preliminary negotiations may be pyramided into a demand indistinguishable from a claim for breach of contract.

Reliance damages should be an adequate sanction for breach of an agreement to negotiate in good faith. Presumably, punitive damages could be assessed in egregious cases. With those sanctions, a good faith obligation is more likely to be enforced than if the matter could be escalated into what appears to be a breach of contract suit.