Riblet Products Corporation makes components of "site-built" homes and recreational vehicles. In 1981 Riblet signed a contract promising Ernest J. Nagy, who then owned 14 percent of Riblet's shares and was its President and CEO, that if Nagy were discharged for any reason other than theft, disclosure of trade secrets, or "similar dishonest acts," he would receive 60 percent of his regular salary until he turned 62. Five years later Riblet was acquired by a new group of investors, who put in $7 million of their own money and secured a $25 million loan, backed by their personal guarantees. Nagy received more than $3 million for his stock in "old Riblet" and purchased 15 percent of the stock in "new Riblet." The owners of the other 85 percent--David Bistricer, Nachum Stein, and their relatives--negotiated a new employment contract with Nagy. The 1986 contract gave Nagy a five-year term with increased compensation and substantial retirement, disability, and death benefits. Paragraph 15 of the 1986 contract adds that if Nagy is terminated for "cause" (which the agreement defines as "conviction of a felony, fraud, dishonesty, illegal use of federally controlled substances, and/or misappropriation of [Riblet's] funds") all benefits of the 1986 contract would be forfeited. If Riblet terminates employment "other than for cause as defined herein" Nagy receives "all salaries, benefits, bonuses, and other direct and indirect forms of compensation" for the remainder of his five-year term. Paragraph 6 of the 1986 contract provides that the 1981 contract "shall remain in full force and effect notwithstanding the execution of this Agreement."

In April 1990, with 17 months remaining on the 1986 contract, Riblet fired Nagy. Riblet believes that it had "cause" as defined in both the 1981 and the 1986 contracts, because Nagy engaged in a series of self-dealing transactions with the firm (for example, he was an undisclosed principal in the group buying Riblet's headquarters building) and refused to follow explicit instructions issued by the board of directors (he wrote checks without the required approval, and he kept his personal secretary on the payroll after the board discharged her). Nagy believes that he was a good chief executive, fired without cause. He admits defying the board but believes that this is not "cause" as the contracts define it, and he admits that he bought the building without revealing his identity but says that this is not "dishonesty" or "misappropriation" because he paid the market price, and Bistricer and Stein discovered his role before the transaction closed. Many other disputes of fact or characterization separate the parties on the question whether Riblet had "cause" to cashier Nagy; we need not discuss them.

Nagy demanded full payment under the 1981 and 1986 contracts; when Riblet refused, Nagy filed this suit, asserting federal jurisdiction under the Employee Retirement and Income Security Act (ERISA). The ERISA claim did not survive in the district court, and although Nagy argues it anew in this court it is insubstantial. Among the many documents Nagy signed is one recognizing that his post-employment benefits are governed by contract rather than by an ERISA plan. Contracts between an employer and a single worker do not become ERISA trusts or plans as 29 U.S.C. § 1002(3) defines them just because they deal with future payments. See Jervis v. Elerding, 504 F. Supp. 606, 608 (C.D.Cal.1980); Lackey v. Whitehall Corp., 704 F. Supp. 201, 204-05 (D.Kan.1988); cf. Bartholet v. Reishauer A.G. (Zurich), 953 F.2d 1073 (7th Cir.1992). Agreements between corporations and their chief executives are in a class by themselves under 29 U.S.C. §§ 1051(2), 1081(a)(3), and 1101(a)(1) (special rules for "top hat" programs). And Nagy ought to be happy with this conclusion, because ERISA preempts all state law relating to pension plans--including the tort and contract law on which Nagy pins his principal hopes, and was awarded more than $2 million by the jury. If these contracts establish "plans" governed by ERISA, then Nagy's verdict evaporates and the litigation begins again, without any opportunity to seek punitive damages. See Kleinhans v. Lisle Savings Profit Sharing Trust, 810 F.2d 618, 625-27 (7th Cir.1987). But ERISA does not apply, so we turn to the state-law claims. These depend on the supplemental jurisdiction under 28 U.S.C. § 1367. As all theories of recovery are part of a single case or controversy, and no one questions the district court's decision to decide the state claims, we do not inquire whether this was the best way to proceed. Myers v. County of Lake, 30 F.3d 847 (7th Cir.1994).

The jury concluded that Riblet broke its promise to Nagy and awarded him $1,267,747 in compensatory damages. The judge instructed the jury that Bistricer and Stein, as majority shareholders, owed Nagy, as minority shareholder, a fiduciary duty of loyalty. Their selfinterested conduct in firing Nagy (to improve corporate profits, and thus the value of their own investments) led the jury to hold Bistricer and Stein jointly and severally liable with Riblet for the compensatory award--and to tack on $375,000 in punitive damages against each of them. When instructing the jury, the district judge did not mention the clause of the 1981 contract (which passed unchanged through the 1986 deal) allowing Riblet to avoid all post-discharge payments if it had the kind of "cause" the parties specified. Defendants believe that this requires a new trial, because it put them in an untenable position before the jury--for if the "cause" provision in the 1986 pact was the sole document allowing Riblet to reduce its obligations to Nagy, then Riblet owed Nagy something for deferred compensation and was hard pressed to tell the jury why it had not paid one red cent.

In a post-verdict memorandum denying the motion for a new trial, the district judge justified his decision as the consequence of waiver. Riblet first raised the subject during the conference on jury instructions--too late, the judge held. Riblet argues to us that the effect of the 1981 contract did not have to be raised in the pleadings as an affirmative defense, but this is beside the point. By the time trial begins, pleadings have faded into history. The document controlling the course of trial is the pretrial order under Fed.R.Civ.P. 16. Each side plans for trial on the assumption that it needs to meet only the contentions its adversary identifies in that order. Unless issues omitted from the pretrial order are foreclosed, neither side could make that assumption, and the function of the pretrial order would be defeated. So we have enforced a stringent rule of forfeiture. See Gorlikowski v. Tolbert, 52 F.3d 1439, 1443-44 (7th Cir.1995) (collecting cases). There are exceptions: for example, to preserve the possibility of appeal, the parties need not list an issue that was resolved before trial, see Calderon v. Witvoet, 999 F.2d 1101, 1108 (7th Cir.1993). District judges may amend the pretrial order to prevent injustice, but appellate review of decisions either way is deferential. The district judge did not abuse his discretion in holding Riblet to the issues and contentions it listed.

From the vantage of hindsight, the "cause" provision of the 1981 contract does not seem important. The jury awarded Nagy everything he asked for. This means that the jury found that Riblet lacked "cause" to discharge Nagy even as the 1986 agreement defines that term. "Cause" in the 1981 agreement is a subset of "cause" in the 1986 agreement; to reject the employer's position under the 1986 agreement is necessarily to reject its position under the 1981 pact. Riblet replies that the jury's verdict is untrustworthy because the inability to explain why they sent Nagy packing empty-handed made Bistricer and Stein look like a pair of stinkers. Maybe so, but the legal system cannot afford to indulge the possibility that jurors disregard their instructions on emotional grounds. Jurors' ability and willingness to follow instructions is not a "fact" about juries; it is, rather, a premise of the jury system. See Gacy v. Welborn, 994 F.2d 305, 313 (7th Cir.1993). If the jury followed the instructions given it, then the verdict shows lack of "cause"; and the lack of cause makes Riblet's arguments about the 1981 contract academic.

Riblet must pay Nagy what the jury found it owes under the contracts. Whether Bistricer and Stein owe anything personally depends on the law of torts and the law of corporations. One of Nagy's theories is that Bistricer and Stein tortiously interfered with the Nagy-Riblet contract. The district court dismissed this theory, and properly so. The premise of applying this tort to intra-corporate affairs is that investors want to prevent their agents from injuring the corporation by firing good employees for personal reasons--for example, from getting rid of a good executive in order to find a sinecure for a crony. See Rice v. Nova Biomedical Corp., 38 F.3d 909 (7th Cir.1994); Kumpf v. Steinhaus, 779 F.2d 1323 (7th Cir.1985). Personnel management in the interests of supervisors is an agency cost that investors strive to curtail. But it is hard to conceive of Bistricer and Stein firing Nagy to promote their own interests at the expense of Riblet's; for practical purposes, their interests are Riblet's. If Riblet loses a good CEO, Bistricer and Stein not only are worse off as stockholders but also are exposed to liability on their guarantees of the bank debt. Nagy's own theory is that Bistricer and Stein fired him to reduce their risk under the guarantees. Bistricer and Stein then must have believed that Nagy's departure would make the firm better off. Nagy's understanding of the motivation for his discharge scuttles this theory of recovery; as the district court observed, no evidence at all suggests that Bistricer and Stein acted for any reason other than the proper one of increasing the corporation's profits and prospects.

Thus we come to the question whether corporate law requires controlling shareholders to act as fiduciaries toward minority shareholder-employees. The answer to that question in Indiana appears to be "yes." See W & W Equipment Co. v. Mink, 568 N.E.2d 564, 574 (Ind.App.1991), following Wilkes v. Springside Nursing Home, Inc., 370 Mass. 842, 353 N.E.2d 657 (1976). Although the Supreme Court of Indiana has not addressed the issue, it has not questioned W & W Equipment or Wilkes. But does Indiana's law govern? Riblet's principal place of business is in Indiana, but it is incorporated in Delaware--and the liability of corporate investors and directors for intra-corporate affairs almost invariably depends on the law of the place of incorporation. This is the "internal affairs doctrine." See CTS Corp. v. Dynamics Corp. of America, 481 U.S. 69, 89-93, 107 S. Ct. 1637, 1649-52, 95 L. Ed. 2d 67 (1987); First National City Bank v. Banco Para El Comercio Exterior de Cuba, 462 U.S. 611, 621, 103 S. Ct. 2591, 2597, 77 L. Ed. 2d 46 (1983); Edgar v. MITE Corp., 457 U.S. 624, 645, 102 S. Ct. 2629, 2642, 73 L. Ed. 2d 269 (1982); RTC v. Chapman, 29 F.3d 1120, 1122-24 (7th Cir.1994); Restatement (2d) of Conflict of Laws §§ 302, 306, 309 (1971); see also 4 Model Business Corporation Act Annotated 1573-88 (3d ed.1993) (collecting state authority). A single rule for each corporation's internal affairs reduces uncertainty and the prospect of inconsistent obligations; it also enables the corporate venturers to adjust the many variables of corporate life (including the contractual promises made to CEOs), confident that they can predict the legal effect of these choices.

Because the case was filed in a federal court in Indiana, that state's conflict-of-laws principles apply. Klaxon Co. v. Stentor Electric Manufacturing Co., 313 U.S. 487, 61 S. Ct. 1020, 85 L. Ed. 1477 (1941). Does Indiana follow the internal affairs doctrine? The district judge apparently thought not, for he charged the jury under Indiana law. When denying the post-trial motions, the judge did not defend this decision; instead he wrote that "[t]he court does not see the yawning difference described by the defendants between Delaware law and the jury instructions given in this case." Where Delaware stands is a question to which we return; for now it is important to determine which body of law to use.

Neither side found an Indiana case discussing the scope of the internal affairs doctrine; our own research has been equally fruitless. Perhaps Indiana's courts deem the application of the doctrine too obvious to require analysis. After all, the corporate charter is a species of contract, and selecting a state of incorporation then is no different from putting a choice-of-law clause in a complex commercial contract. States compete to provide better rules of corporate law, from which entrepreneurs may choose. Indiana enforces choice-of-law clauses and therefore should apply the internal affairs doctrine too. Certainly the state's legislature has embraced it. Indiana has adopted the Model Business Corporation Act, and § 15.05(c) of the Model Act, enacted as Ind.Code § 23-1-49-5(c), provides that the state's corporate code as a whole "does not authorize Indiana to regulate the organization or internal affairs of a foreign corporation authorized to transact business in Indiana." This statute does not in terms forbid the development of a common law doctrine that affects the internal affairs of foreign corporations, but it establishes that Indiana presumptively conforms to the norm in looking to the law of the state of incorporation for internal-affairs issues.

Nagy calls Riblet a "pseudo-foreign corporation" and appeals to the principle that states have the power to regulate matters that lack any connection to the state of incorporation. It relies on Mansfield Hardwood Lumber Co. v. Johnson, 268 F.2d 317 (5th Cir.1959). Let us suppose that Indiana has the power to override the venturers' choice of law; yet it does not do this for ordinary contracts, and it has not hinted at doing so for the special kind of contract known as the corporation. In a few states such a choice has been made by statute; in none has it been made by the judiciary. Mansfield is a derelict in American caselaw. California has been the most aggressive in asserting control over the internal affairs of foreign corporations located in that state; California subjects them to a list of rules selected from its domestic corporate statute. But even California does not assert this power unless the firm does more than half of its business within California, and "more than one-half of its outstanding voting securities are held of record by persons having addresses in this state." Cal. Corp.Code § 2115(a). Eightyfive percent of Riblet's voting securities, and substantially all of its debt, are held by investors from New York--a state whose highest court follows a path that diverges from Wilkes. See Gallagher v. Lambert, 74 N.Y.2d 562, 549 N.Y.S.2d 945, 549 N.E.2d 136 (1989); Ingle v. Glamore Motor Sales, Inc., 73 N.Y.2d 183, 538 N.Y.S.2d 771, 535 N.E.2d 1311 (1989). If even California would not apply its domestic corporate law to a firm like Riblet, then we are confident that Indiana would not do so. The corporate law of Delaware therefore supplies the rule of decision.

But what is that law? The Supreme Court of Delaware has never addressed the question. Ueltzhoffer v. Fox Fire Development Co., 1991 WL 271584 (Del.Ch. Dec.19, 1991) (Berger, V.C.), affirmed, 618 A.2d 90 (Del.1992), marks its closest approach to the subject, but in Ueltzhoffer it did no more than approve the Vice Chancellor's opinion, itself equivocal. No one supposes that Delaware would apply Wilkes to widely held corporations; a "fiduciary" duty to every baggage handler at United Air Lines just because the employee owns a share of stock would put employee-owned firms at such a competitive disadvantage that they would soon collapse. But closely held corporations may run best under special rules. Many closely held firms endeavor to show no profits (to minimize their taxes) and to distribute the real economic returns of the business to the investors as salary. When firms are organized in this way, firing an employee is little different from canceling his shares. A "fiduciary" duty to investor-employees, which protects the return on investment, then may approximate the terms the investors would have accepted had they bargained expressly. See Frank H. Easterbrook & Daniel R. Fischel, The Economic Structure of Corporate Law 243-48 (1991). But this understanding implies that the label "fiduciary" does not trump a real contract; it is a gap-filling approach, rather than one that prevents participants in a closely held firm from making their own arrangements. See generally John H. Langbein, The Contractarian Basis of the Law of Trusts, 105 Yale L.J. 625, 657-69 (1995). Compare Jordan v. Duff & Phelps, Inc., 815 F.2d 429, 436-39 (7th Cir.1987), with Industrial Representatives, Inc. v. CP Clare Corp., 74 F.3d 128 (7th Cir.1996). Nagy and Riblet negotiated elaborate contracts specifying the parties' rights and duties; neither Indiana nor Delaware uses punitive damages to redress breach of contract, and we question whether Delaware would use a fiduciary principle of corporate law to do so.

Unfortunately, however, Ueltzhoffer does not resolve the question. The Vice Chancellor's opinion states that "to the extent that there is some sort of 'business purpose' test" for whether a shareholder-employee may be discharged, the defendants established such a purpose; that enabled the court to step past the question whether, and when, such a purpose is essential. The Vice Chancellor added that "I am aware of no Delaware authority following Wilkes or applying Wilkes outside the context of a close corporation"--but our case involves a close corporation, and so did Ueltzhoffer. This leaves Delaware law in an unsettled condition. One district judge has predicted that Delaware will repudiate Wilkes when the time comes, see Byington v. Vega Biotechnologies, Inc., 869 F. Supp. 338 (D.Md.1994), but the district judge in this case had a different understanding of Delaware's proclivities.

Recognizing the nationwide application of Delaware corporate law, and the benefits of making that law more certain, we think the best way to resolve this debate is to ask the Supreme Court of Delaware. Only that court can give a definitive answer. Delaware Sup.Ct. R. 41(a)(ii) permits any United States Court of Appeals to certify a question of law to that court for resolution. We hereby do so. We respectfully request the Supreme Court of Delaware to answer this question:

Whether majority shareholders in a Delaware corporation have a fiduciary duty of loyalty to a minority shareholder, who is also an employee under a written contract, with respect to issues affecting that employment.

It may be that the Justices of the Supreme Court of Delaware will conclude that this is not the controlling question. In that event, the Justices should feel free to reformulate the question, just as they would when dealing with the issues posed in a case wholly within the state system. It is not our purpose to constrain the state court but only to suggest how we conceive of the issue.

We have endeavored to resolve all other questions in the case, so that this is the sole remaining issue, the answer to which controls the liability of Bistricer and Stein. One last potential issue, raised in Nagy's crossappeal, requires only brief comment. The district judge did not allow interest on the awards of punitive damages for the 28 months between return of the verdict and entry of judgment after the court had resolved the many post-trial motions. Nagy argues that he is entitled to such interest under Indiana law. We have held, however, that the punitive damages must stand or fall under Delaware law. Nagy does not contend that Delaware authorizes interest on punitive damages for the time between verdict and judgment. So every issue or dispute in the case has been wrapped up--except for the meaning of Delaware corporate law. The Clerk will transmit the briefs and appendices in this case to the Supreme Court of Delaware, together with this opinion, "substantially in the form set forth in Official Form K". Del. Sup.Ct. R. 41(c)(i).