Duke Energy Fla. v. FirstEnergy Corp.

NOT RECOMMENDED FOR PUBLICATION File Name: 18a0187n.06 No. 17-3024 UNITED STATES COURT OF APPEALS FILED FOR THE SIXTH CIRCUIT Apr 10, 2018 DEBORAH S. HUNT, Clerk DUKE ENERGY FLORIDA, LLC, fka Florida ) Power Corporation, ) ) Plaintiff-Appellant ) ON APPEAL FROM THE Counter-Defendant, ) UNITED STATES DISTRICT ) COURT FOR THE v. ) NORTHERN DISTRICT OF ) OHIO FIRSTENERGY CORP., ) ) OPINION Defendant-Appellee ) Counter-Claimant. ) BEFORE: KEITH, McKEAGUE, and STRANCH, Circuit Judges. JANE B. STRANCH, Circuit Judge. This case asks what entity will be liable under the Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA), 42 U.S.C. § 9601 et seq., for the costs associated with cleaning up hazardous waste that was released at two manufactured gas plants in the early 1900s. In particular, at issue is whether the corporate successor to the local companies that operated those plants must bear that liability on its own or may force contribution from the successor to the former corporate parent. A corporate parent may be held indirectly liable under CERCLA only if the corporate veil separating parent and subsidiary may be pierced under the corporate law of the relevant state. Because Florida law does not permit piercing the veil on the facts of this case, we AFFIRM the district court’s decision granting Defendant FirstEnergy’s motion for summary judgment. No. 17-3024 Duke Energy, LLC v. FirstEnergy Corp. I. BACKGROUND A. Factual Background This case is a dispute over who may be held liable for hazardous waste cleanup costs under CERCLA. The hazardous waste at issue was released by two manufactured gas plants in Florida between 1929 and 1943. The processes used at the time to create gas for home consumption inevitably released harmful byproducts, including coal tar, into the local environment, causing groundwater contamination. N.Y. State Elec. & Gas Corp. v. FirstEnergy Corp. (NYSEG II), 766 F.3d 212, 217 (2d Cir. 2014). At the time the tar was released, two utility companies, Florida Public Service Company (FPSC) and Sanford Gas Company (Sanford) operated the plants. A large New York holding company, Associated Gas & Electric Company (AGECO), owned the controlling interest in both companies through its subsidiaries. Plaintiff- Appellant Duke Energy is the admitted corporate successor to FPSC and Sanford; Defendant- Appellee FirstEnergy is the stipulated corporate successor to AGECO. Beginning in 1998, Duke Energy and other previous owners of the gas plant sites entered into a series of agreements with the Environmental Protection Agency (EPA) to conduct remediation at the sites and reimburse the EPA for response costs it had incurred. Fla. Power Corp. v. FirstEnergy Corp., 810 F.3d 996, 998–99 (6th Cir. 2015). In the present case, Duke Energy asserts that FirstEnergy, as AGECO’s corporate successor, should be required to contribute to the cleanup costs based on the theory of indirect liability. 1. The AGECO Empire AGECO was a public utility holding company incorporated in New York in 1906. N.Y. State Elec. & Gas Corp. v. FirstEnergy Corp. (NYSEG I), 808 F. Supp. 2d 417, 430 (N.D.N.Y. 2011), aff’d in part and vacated in part, NYSEG II, 766 F.3d 212. By 1929, AGECO’s “empire” -2- No. 17-3024 Duke Energy, LLC v. FirstEnergy Corp. included more than 200 utility and transportation companies in twenty-seven states and the Philippines. Professor Jonathan Macey, Duke Energy’s expert, summarized the corporate structure: “AGECO’s assets consisted primarily of the stock of its single subsidiary, AGECORP. AGECORP’s assets consisted primarily of the stock of its subsidiaries, which included Gen Gas. Gen Gas’s assets consisted primarily of the stock of its operating companies, which included FPSC and Sanford Gas.” This pyramid ownership structure was typical within the AGECO empire. See NYSEG I, 808 F. Supp. 2d at 437–38 (listing nine companies that owned AGECO subsidiary NYSEG); Rochester Gas & Elec. Corp. v. GPU, Inc. (RG&E I), No. 00-cv-6369, 2008 WL 8912083, at *4 (W.D.N.Y. Aug. 8, 2008), aff’d, RG&E II, 355 F. App’x 547 (2d Cir. 2009) (listing five companies that owned AGECO subsidiary RG&E). AGECO has been dubbed the “‘poster child’ for the abusive practices of certain public utility holding companies” in the first half of the twentieth century. RG&E I, 2008 WL 8912083, at *2. From 1922 until 1940, the sprawling company was controlled by Howard Hopson (who Professor Macey dubs an “iconic felon”) and his associate John I. Mange. Hopson’s abuses “were both legion and well-documented.” NYSEG I, 808 F. Supp. 2d at 499. Proceedings at the Securities and Exchange Commission and other independent, contemporaneous investigations into AGECO’s operations confirmed Hopson’s criminal mismanagement of the company. Hopson “siphoned off large sums of money to finance his own personal ventures and interests.” Id. His myriad of techniques to accomplish this siphoning included, to name just a few, trading bonds back and forth among subsidiaries, collecting salaries and fees from the subsidiaries for his services and those of his immediate family members, and requiring employees to invest ten percent of their pay into the system’s holding companies as part of an employee welfare program. The Federal Power Commission (FPC) concluded that its investigation into the -3- No. 17-3024 Duke Energy, LLC v. FirstEnergy Corp. operation of six Pennsylvania utilities revealed “an extraordinary picture of the exploitation of an essential public service for which the holding-company device served as a cloak. Almost every possibility for plunder was exploited.” AGECO’s improper behavior inspired at least four government investigations over the course of the 1930s; one such investigation culminated in proceedings by the Securities and Exchange Commission to divest AGECO of its monopoly. In January 1940, AGECO voluntarily filed for reorganization under Chapter X of the Bankruptcy Act, after which the court appointed trustees to run the company. See In re Associated Gas & Elec. Co., 61 F. Supp. 11, 17 (S.D.N.Y. 1944), aff’d, 149 F.2d 996 (2d Cir. 1945). Upon filing, Hopson and Mange lost voting control. See id. at 24. In December of 1940, Hopson was convicted of mail fraud and sentenced to five years in prison. RG&E I, 2008 WL 8912083, at *4 n.6. Over the course of the next several years, the trustees worked to end the abuses that had been encouraged under Hopson and Mange. 2. AGECO’s Relationship with FPSC and Sanford FPSC owned and operated the Orlando gas plant from 1924 until 1943. FPSC was folded into the AGECO empire in 1929, when AGECO purchased its corporate parent. Sanford purchased the Sanford gas plant in 1928 and entered the AGECO system two years later, when all of its outstanding shares were purchased by an AGECO subsidiary. Though the parties have stipulated that there are no material factual disputes in this case, the exact nature and extent of AGECO’s interactions with FPSC and Sanford some eighty years ago remains uncertain. Among the constellation of subsidiaries in the AGECO empire, FPSC and Sanford were minor players, distant from the New York headquarters and not particularly profitable. They received no mention in the government investigations that sparked the bankruptcy filing. But as was the case at many other AGECO subsidiaries, FPSC and Sanford -4- No. 17-3024 Duke Energy, LLC v. FirstEnergy Corp. were run by an interlocking cast of officers and directors who staffed positions across the AGECO empire, with board meetings often held at AGECO offices in New York and much of the day-to-day operations of the companies controlled by AGECO-affiliated management service companies. In 1944, FPSC and Sanford merged into Florida Power Company; by 1946, Florida Power, now known as Duke Energy, separated from the AGECO empire. Meanwhile, AGECO was consolidated with its subsidiary AGECORP to form the General Public Utilities Corporation, which would eventually merge into FirstEnergy. The parties do not dispute their relative corporate relationships to FPSC, Sanford, and AGECO. B. Procedural Background Duke Energy filed this CERCLA action seeking contribution and response costs from FirstEnergy in the Middle District of Florida in 2011. Lacking personal jurisdiction over the defendant, that court transferred the case to the Northern District of Ohio. The Ohio district court eventually dismissed the case as barred by the statute of limitations; a divided panel of this court reversed and remanded. Fla. Power, 810 F.3d at 999. Upon remand, the parties stipulated that there were no material disputes of fact and filed cross motions for summary judgment on liability. In ruling on the summary judgment motions, the district court declined to pierce the corporate veils separating FPSC and Sanford from AGECO. Applying Florida’s law of corporations, the court determined that although Duke Energy had successfully shown that AGECO dominated and controlled FPSC and Sanford, it had not proven by clear and convincing evidence that AGECO had used the two subsidiaries for an improper or fraudulent purpose. Fla. Power Corp. v. FirstEnergy Corp., No. 1:12-cv-1839, 2016 WL 7178660, at *7–10 (N.D. Ohio -5- No. 17-3024 Duke Energy, LLC v. FirstEnergy Corp. Dec. 9, 2016). The district court concluded in the alternative that even if Duke Energy had mustered sufficient evidence to justify piercing the veil, such liability “would be cut off as of January 10, 1940, the date on which AGECO filed for bankruptcy.” Id. at *10. Duke Energy appeals. II. ANALYSIS We review a grant of summary judgment de novo. See Gillis v. Miller, 845 F.3d 677, 683 (6th Cir. 2017). A. Indirect Liability Under CERCLA CERCLA was enacted in 1980 “to abate and control the vast problems associated with abandoned and inactive hazardous waste disposal sites. . . . CERCLA was intended primarily to facilitate the prompt cleanup of hazardous waste sites by placing the ultimate financial responsibility for cleanup on those responsible for hazardous wastes.” United States v. R.W. Meyer, Inc., 889 F.2d 1497, 1500 (6th Cir. 1989) (citations and internal quotation marks omitted). To that end, CERCLA imposes strict liability on past and present owners and operators (called “potentially responsible parties,” or PRPs) of hazardous waste disposal sites. Burlington N. & Santa Fe Ry. Co. v. United States, 556 U.S. 599, 608 (2009). Because of the statute’s language, its remedial purpose, and its focus on abandoned and inactive sites, retroactive liability for pre-enactment conduct is permitted under CERCLA. Franklin Cty. Convention Facilities Auth. v. Am. Premier Underwriters, Inc., 240 F.3d 534, 552–53 (6th Cir. 2001); R.W. Meyer, 889 F.2d at 1506. Where multiple parties share responsibility for some harm, CERCLA contemplates two mechanisms for divvying costs: apportionment and contribution. “‘[A]pportionment . . . looks to whether defendants may avoid joint and several liability by establishing a fixed amount of -6- No. 17-3024 Duke Energy, LLC v. FirstEnergy Corp. damage for which they are liable,’ while contribution actions allow jointly and severally liable PRPs to recover from each other on the basis of equitable considerations.” Burlington N., 556 U.S. at 615 n.9 (quoting United States v. Burlington N. & Santa Fe Ry. Co., 520 F.3d 918, 939 (9th Cir. 2008)). The case before us seeks contribution. To establish a prima facie case for contribution under CERCLA, a plaintiff must show that: (1) the polluting site is a “facility” within the meaning of 42 U.S.C. § 9601(9); (2) the facility released or threatened to release a hazardous substance; (3) the release caused the plaintiff to incur necessary costs of response; and (4) the defendant falls within one of the four groups of potentially responsible parties described at 42 U.S.C. § 9607(a). Village of Milford v. K-H Holding Corp., 390 F.3d 926, 933 (6th Cir. 2004). In the present case, the parties have stipulated that the first two elements are satisfied, and FirstEnergy does not dispute that Duke Energy incurred the costs of responding to the hazardous waste through its agreements with the EPA. Thus, only the fourth element of this test—whether FirstEnergy is a potentially responsible party—is at issue. CERCLA imposes strict liability on four types of potentially responsible parties: (1) the owner or operator of a facility; (2) the owner or operator of a facility at the time a hazardous substance was released; (3) any person who arranged for disposal of a hazardous substance at a facility; and (4) any person who accepts hazardous waste for transport to disposal or treatment facilities. 42 U.S.C. § 9607(a)(1)–(4); see also Fla. Power, 810 F.3d at 1000. Thus, although Duke Energy’s predecessor sold the FPSC and Sanford gas plants decades ago, Duke Energy remains liable because its predecessors owned or operated the plants at the time this waste was released. -7- No. 17-3024 Duke Energy, LLC v. FirstEnergy Corp. Duke Energy argues that AGECO’s successor, FirstEnergy, should share in its liability. Supreme Court caselaw has clarified that such suits proceed under two distinct theories of recovery: direct and indirect liability. Direct liability is appropriate if the corporation “operates the facility, and that operation is evidenced by participation in the activities of the facility.” United States v. Bestfoods, 524 U.S. 51, 68 (1998) (quoting Lynda J. Oswald, Bifurcation of the Owner and Operator Analysis Under CERCLA, 72 Wash. U. L.Q. 223, 269 (1994)). The parties have stipulated that direct liability is not at issue in this case. Thus, Duke Energy exclusively pursues a theory of indirect liability. According to the Supreme Court, “[c]ontrol of the subsidiary, if extensive enough, gives rise to indirect liability under piercing doctrine.” Id. (quoting Oswald, supra, at 269). “[W]hen (but only when) the corporate veil may be pierced, may a parent corporate be charged with derivative CERCLA liability for its subsidiary’s actions.” Id. at 63–64 (footnote omitted). Thus, FirstEnergy may be held indirectly liable for the cleanup costs only if the corporate veils that separated FPSC and Sanford from AGECO may be pierced. New York federal courts have twice been presented with similar cases about CERCLA liability for cleanup costs incurred due to operation of gas plants in the AGECO empire. Those district courts concluded, and the Second Circuit affirmed, that veil-piercing was warranted under New York law and that AGECO’s successor, FirstEnergy, could be held indirectly liable for cleanup costs. See NYSEG I, 808 F. Supp. 2d 417, aff’d in part and vacated in part, NYSEG II, 766 F.3d 212; RG&E I, 2008 WL 8912083, aff’d, RG&E II, 355 F. App’x 547. The holdings and reasoning of these similar cases are instructive. Their outcomes, however, do not control this case, as we must apply a different state’s corporate law to analyze AGECO’s interactions with different subsidiaries. -8- No. 17-3024 Duke Energy, LLC v. FirstEnergy Corp. B. Piercing the Corporate Veil For indirect liability claims in CERCLA cases, state law governs the propriety of piercing the corporate veil. Carter-Jones Lumber Co. v. Dixie Distrib. Co., 166 F.3d 840, 847 (6th Cir. 1999). Because the parties dispute whether New York or Florida law applies, we apply the choice-of-law rules of the forum state—in this case, Ohio—to select the appropriate law. See Glennon v. Dean Witter Reynolds, Inc., 83 F.3d 132, 136 (6th Cir. 1996). For tort claims, the Ohio Supreme Court has adopted the most-significant-relationship approach. Morgan v. Biro Mfg. Co., 474 N.E.2d 286, 288–89 (Ohio 1984); see also Pilgrim v. Universal Health Card, L.L.C., 660 F.3d 943, 946 (6th Cir. 2011). Under that approach, a court presumes “that the law of the place of the injury controls unless another jurisdiction has a more significant relationship to the lawsuit.” Morgan, 474 N.E.2d at 289 (citing Restatement (Second) of Conflict of Laws § 146 (Am. Law Inst. 1971)). The main factors a court considers in making that determination are “(1) the place of the injury; (2) the place where the conduct causing the injury occurred; (3) the domicile, residence, nationality, place of incorporation, and place of business of the parties; [and] (4) the place where the relationship between the parties, if any, is located.” Id. The injury in this case—the release of toxic waste—occurred in Florida, as did the conduct causing the release of toxic waste—the operation of the gas plants. The contamination caused significant environmental harm in Florida, and the ongoing cleanup operations necessarily occur in Florida as well. These considerations weigh heavily in favor of finding that Florida has the most significant relationship to the case. On the other hand, AGECO was incorporated in New York, NYSEG I, 808 F. Supp. 2d at 430, and directed—perhaps even controlled—the operation of the Florida subsidiaries from a distance. On balance, those New -9- No. 17-3024 Duke Energy, LLC v. FirstEnergy Corp. York connections are insufficient to outweigh the presumption that the law of the place of injury controls. Florida law thus governs the veil-piercing inquiry. Florida courts are “reluctant to pierce the corporate veil” and take such action only in “severely limited” circumstances. Becherer v. Merrill Lynch, Pierce, Fenner & Smith, Inc., 43 F.3d 1054, 1064 (6th Cir. 1995). The party seeking to pierce the veil must prove three elements: (1) the shareholder dominated and controlled the corporation to such an extent that the corporation’s independent existence, was in fact non-existent and the shareholders were in fact alter egos of the corporation; (2) the corporate form must have been used fraudulently or for an improper purpose; and (3) the fraudulent or improper use of the corporate form caused injury to the claimant. Gasparini v. Pordomingo, 972 So. 2d 1053, 1055 (Fla. Dist. Ct. App. 2008) (quoting Seminole Boatyard, Inc. v. Christoph, 715 So. 2d 987, 990 (Fla. Dist. Ct. App. 1998)). Each of these three prongs of the test “must be proven by a preponderance of the evidence.” Seminole Boatyard, 715 So. 2d at 990; see also Bookworld Trade, Inc. v. Daughters of St. Paul, Inc., 532 F. Supp. 2d 1350, 1360 (M.D. Fla. 2007); Dania Jai-Alai Palace, Inc. v. Sykes, 450 So. 2d 1114, 1121 (Fla. 1984); Priskie v. Missry, 958 So. 2d 613, 614 (Fla. Dist. Ct. App. 2007). 1. Domination and Control To determine whether a shareholder or corporate parent exercised domination and control over its subsidiary, Florida courts do not apply a uniform test. One Florida appellate court listed relevant factors as follows: (1) The parent corporation owns all or majority of the capital stock of the subsidiary. (2) The parent and subsidiary corporations have common directors or officers. (3) The parent corporation finances the subsidiary. (4) The parent corporation subscribes to all the capital stock of the subsidiary or otherwise causes its incorporation. (5) The subsidiary has grossly inadequate capital. (6) The parent corporation pays the salaries or expenses or losses of the subsidiary. (7) The subsidiary has substantially no business except with the parent corporation or no assets except those conveyed to it by the parent -10- No. 17-3024 Duke Energy, LLC v. FirstEnergy Corp. corporation. (8) In the papers of the parent corporation, and in the statements of its officers, “the subsidiary” is referred to as such or as a department or division. (9) The directors or executives of the subsidiary do not act independently in the interest of the subsidiary but take direction from the parent corporation. (10) The formal legal requirements of the subsidiary as a separate and independent corporation are not observed. Dania Jai-Alai Palace, Inc. v. Sykes, 425 So. 2d 594, 599 (Fla. Dist. Ct. App. 1982), aff’d in part, quashed in part on other grounds, 450 So. 2d 1114 (Fla. 1984). But these elements are not mechanically applied and, indeed, are rarely cited. Instead, Florida courts appear to look to the totality of the circumstances to determine whether domination and control has been proven. See, e.g., 17315 Collins Ave., L.L.C. v. Fortune Dev. Sales Corp., 34 So. 3d 166, 168 (Fla. Dist. Ct. App. 2010) (applying no set test but finding two companies “operated as alter egos” where one owned the other but had no operations, employees, or bank accounts, and the two had the same purpose, then proceeding to find the improper purpose prong also satisfied); Priskie, 958 So. 2d at 615 (noting that there was no dispute that a shareholder had been personally funding a corporation, then proceeding to find the second prong not satisfied); Unijax, Inc. v. Factory Ins. Ass’n, 328 So. 2d 448, 454 (Fla. Dist. Ct. App. 1976) (dismissing on the first prong alone where the only evidence presented was ownership and common officers and directors). The district court relied on three of the ten listed factors to support its conclusion that a finding of domination and control was warranted: “AGECO owned the majority of FPSC and Sanford’s capital stock at the time in question, the officers and directors overlapped and FPSC and Sanford did not act independently from AGECO.” Fla. Power, 2016 WL 7178660, at *8. FirstEnergy argues that because the other seven factors are not satisfied, a finding of domination and control is not warranted—but it fails to point to any Florida cases requiring satisfaction of either a set number of the factors or any particular factors to meet this first prong of the three- part test. Based on Florida law, the appropriate approach is to look at those concerning -11- No. 17-3024 Duke Energy, LLC v. FirstEnergy Corp. circumstances that have been identified and decide whether, when viewed in their entirety, they support the inference that AGECO dominated or controlled its subsidiaries. AGECO owned the majority of FPSC and Sanford’s stock. This factor is not dispositive because “[a] parent corporation generally is not liable for the acts of its subsidiary, even if its subsidiary is wholly owned.” Corrigan v. U.S. Steel Corp., 478 F.3d 718, 724 (6th Cir. 2007). But the factor is one to be considered. AGECO and its subsidiaries shared common officers and directors. This factor alone also is not dispositive; “it is entirely appropriate for directors of a parent corporation to serve as directors of its subsidiary, and that fact alone may not serve to expose the parent corporation to liability for its subsidiary’s acts.” Bestfoods, 524 U.S. at 69 (quoting Am. Protein Corp. v. AB Volvo, 844 F.2d 56, 57 (2d Cir. 1988)). Overlap is nonetheless a proper factor to consider, as it may be so pervasive as to raise questions about a subsidiary’s independence. In this case, the overlap suggests AGECO may have directly controlled its subsidiaries’ actions. For example, Mange and Hopson themselves sat on FPSC’s Board for a period of time, and AGECO officers and directors sat on the Boards of FPSC and Sanford until 1938. To give just two examples from Professor Macey’s thirty-page list of overlapping officials, J.F. McKenna served as an officer or director in 147 AGECO companies, including FPSC and Sanford; C.A. Dougherty worked in 130, again including FPSC and Sanford. The sheer extent of overlap raises concerns. The greatest concern arises in the third factor identified by the district court, FPSC and Sanford’s inability to act independently. First, important decisions were made close to AGECO’s home base. Board meetings and shareholders’ meetings for both FPSC and Sanford were held at Hopson’s office in New York City from 1929 to 1937 and from 1930 to 1938, -12- No. 17-3024 Duke Energy, LLC v. FirstEnergy Corp. respectively. As of 1934, the majority of directors and officers for both FPSC and Sanford were located in New York City. Keeping the decision makers in New York appears to have facilitated AGECO’s central control of distant subsidiaries. For example, a special master investigating FPSC’s corporate structure during a shareholder derivative suit in 1930—discussed further below—noted that all of FPSC’s “financing was done through the New York office,” such that FPSC’s Florida office “had no record of the moneys borrowed by [FPSC] except as and when copies of notes already executed for money borrowed were sent from New York to the Florida office.” The subsidiaries’ officers and directors in New York, moreover, do not appear to have been careful about protecting FPSC and Sanford’s independence. When the special master asked one officer who served as a vice president of both FPSC and AGECO which role he assumed when loaning money between the two entities, the officer replied, “I don’t know whether a man can divide himself up that way or not.” Even when officers and directors were physically removed from New York, Hopson generally kept the subsidiaries on a tight leash. As the court recounted during AGECO’s bankruptcy proceeding, Hopson controlled the financial and accounting policies of [AGECO] and its subsidiaries throughout. He controlled their Boards of Directors and held their undated signed resignations. Hopson’s employees kept the minute books; some of the minutes were spurious. They also kept the books of account (irregularly maintained). Entries were changed and reinstated as Hopson directed; one item was changed 13 times. Alleged contracts for stock subscriptions of [AGECO] in subsidiaries, disappeared and reappeared as the occasion required. In re Associated Gas & Elec. Co., 61 F. Supp. at 24. The record does not show that FPSC and Sanford bucked this trend. In 1930, minutes from a meeting of Sanford’s Board of Directors declared that Sanford was “not in a position to maintain independently an organization that could purchase apparatus, supplies and materials required by the company.” Years later, in 1939, -13- No. 17-3024 Duke Energy, LLC v. FirstEnergy Corp. FPSC and Sanford’s Boards approved the transfer of their management service contracts from one AGECO service company to another on the same day, at the same location, and within thirty minutes of one another. The language in the two sets of minutes is identical. Problems with the use of AGECO-affiliated management service companies extended far beyond the manner in which the contracts were ratified. As the court summarized it in RG&E, The relationship between AGECO and the various service companies Hopson and Mange created and operated was incestuous. The threat to corporate independence posed by overlapping officers and directors within the holding company structure paled in comparison to corporate relationships in which the directors of the operating utility were also directors, officers or paid employees of the service company with whom the utility entered into multi-million dollar contracts that outsourced virtually all of the utility’s operating responsibilities. RG&E I, 2008 WL 8912083, at *7. One independent director in the AGECO system testified to the FPC’s examiner that the relationship between service companies and boards “gave to the service companies a dominance in the affairs of the System that . . . was not consistent with . . . the independent integrity of a Board of Directors.” After examining several northeastern AGECO subsidiaries’ service contracts, the FPC examiner concluded that “the respondents and other operating companies in the Associated System had no freedom of action whatever in the matter of arranging for the character of service or in determining whether such service was necessary or desirable.” As already described, the record suggests that FPSC and Sanford were included in this perilous outsourcing pattern, with the two companies entering almost simultaneously into contracts with AGECO-affiliated service companies in 1930. Recognizing the limitations of analysis conducted approximately eighty years after the events in question, we give substantial weight to conclusions made by courts at the time. The parties address only one case from the time that focused particularly on one of these subsidiaries, a shareholder derivative suit decided in 1931, Meiners v. Florida Public Service Co., No. 730 -14- No. 17-3024 Duke Energy, LLC v. FirstEnergy Corp. Equity (S.D. Fla. June 30, 1931). In Meiners, the shareholder alleged that FPSC had entered into exploitative management contracts with AGECO-affiliated companies. The relevant contract dated from 1924, before FPSC was acquired by AGECO—and before FPSC changed contracts in 1930, as previously discussed. The suit was dismissed for lack of jurisdiction because AGECO “was not . . . engaged in transacting business within the State of Florida, and had no agents or officers engaged in transacting business for it within said State of Florida.” The judge nonetheless determined, based on a special master report, that the fees FPSC paid to AGECO affiliated management and purchasing companies “were for services actually rendered and were reasonable in amount,” and the interest rate an AGECO subsidiary charged FPSC “was reasonable in amount and was neither exorbitant nor excessive.” And yet the report underlying those conclusions declared that AGECO “through interlocking directors and the same persons as officers in both companies, dominated and directed all the operations of [FPSC] from its office in New York.” Thus, while Meiners may support the argument that AGECO acted without a fraudulent or improper purpose, as discussed further below, it bolsters the conclusion that AGECO dominated and controlled FPSC. FirstEnergy’s expert concluded that FPSC and Sanford’s plants were operated by locally- based vice presidents, general managers, and plant superintendents who were not otherwise employed by AGECO or its holding companies. But local involvement in day-to-day operations does not undermine the conclusion that key decisions were made in New York. As with the other two factors considered, no single fact is necessarily dispositive in determining that FPSC and Sanford did not act independently. Considering all of the evidence in its entirety, Duke Energy has shown by a preponderance of the evidence that AGECO dominated -15- No. 17-3024 Duke Energy, LLC v. FirstEnergy Corp. and controlled key aspects of FPSC and Sanford’s operations in the period prior to the 1940 bankruptcy filing. We therefore turn to the second prong of Florida’s veil-piercing test. 2. Fraud or Improper Purpose The second prong, use of the corporate form for a fraudulent or improper purpose, lies at the heart of Florida’s veil-piercing test. Florida courts will not pierce the corporate veil “unless it be shown that the corporation is formed or used for some illegal, fraudulent or other unjust purpose.” Sykes, 450 So. 2d at 1121 (quoting Roberts’ Fish Farm v. Spencer, 153 So. 2d 718, 721 (Fla. 1963)). Contrary to the conclusion of the district court, the plaintiff may prove fraud or improper purpose by a simple preponderance. See Seminole Boatyard, 715 So. 2d at 990; Powerhouse, Inc. v. Walton, 557 So. 2d 186, 187 (Fla. Dist. Ct. App. 1990). A preponderance standard is consistent with Florida’s general rule that fraud need only be proven by a preponderance. Bacon & Bacon Mfg. Co. v. Bonsey Partners, 62 So. 3d 1285, 1287 (Fla. Dist. Ct. App. 2011) (explaining that the Florida Supreme Court changed the burden of proof for fraud from clear and convincing evidence to a preponderance in Wieczoreck v. H & H Builders, Inc., 475 So. 2d 227 (Fla. 1985)). Classic examples of improper conduct include the use of the corporate form “to mislead or defraud creditors, to hide assets, to evade the requirements of a statute or some analogous betrayal of trust.” Lipsig v. Ramlawi, 760 So. 2d 170, 187 (Fla. Dist. Ct. App. 2000). Florida courts do not apply this test mechanically. They carefully consider the “subjective motivation” of the corporate actors, In re Hillsborough Holdings Corp., 166 B.R. 461, 469 (Bankr. M.D. Fla. 1994), aff’d, 176 B.R. 223 (M.D. Fla. 1994), as they watch for “conduct that has a bad aroma,” Steinhardt v. Banks, 511 So. 2d 336, 338 (Fla. Dist. Ct. App. 1987) (per curiam). -16- No. 17-3024 Duke Energy, LLC v. FirstEnergy Corp. The second prong of the veil-piercing inquiry presents a close question. The difficulty is due in large part to the incomplete nature of the historical record. There is substantial evidence tending to show that AGECO dominated and controlled FPSC and Sanford, but there is much less evidence shedding light on the subjective motivations of AGECO leadership regarding these two relatively minor subsidiaries in the sprawling AGECO empire. We must nonetheless determine whether it is more likely than not that those leaders behaved with an improper or fraudulent purpose. Violating environmental laws can amount to an improper or fraudulent purpose. Thus, in Carter-Jones Lumber Co. v. LTV Steel Co., 237 F.3d 745 (6th Cir. 2001), we pierced the veil and held a company’s sole shareholder liable under CERCLA because of environmental violations. The shareholder, after signing an affidavit acknowledging the problem of disposing of transformers containing hazardous chemicals, was “personally involved in the scheme . . . to hide the transformers from the EPA.” Id. at 747–48. We determined that there was “no real argument” that the second prong of Ohio’s veil-piercing test, which is very similar to Florida’s, was satisfied. Id. at 748. By the same logic, if AGECO sought to dodge CERCLA or other environmental laws by spinning off undercapitalized subsidiaries to avoid liability, it would have attempted “to evade the requirements of a statute.” Lipsig, 760 So. 2d at 187; see also United States v. Carolina Transformer Co., 978 F.2d 832, 840 (4th Cir. 1992) (“We are unwilling to hold that merely by splitting off the particular part of its operations that resulted in its environmental problems and shifting the remainder of its assets, employees, management, customers, accounts and production methods to another corporation, an otherwise responsible corporation could all but completely wash its hands of its environmental liability.”); Ocala Breeders’ Sales Co. v. Hialeah, Inc., -17- No. 17-3024 Duke Energy, LLC v. FirstEnergy Corp. 735 So. 2d 542 (Fla. Dist. Ct. App. 1999) (piercing the veil where a subsidiary had never had sufficient assets to satisfy its obligations). Any such action would constitute an improper purpose, and piercing the veil might be appropriate in such a case. The district court, in analyzing this prong, asked whether “the parent company misled or worked a fraud upon its subsidiary.” Fla. Power Corp., 2016 WL 7178660, at *9. The court concluded that this record, unlike the records at issue in RG&E and NYSEG, did not support a conclusion that AGECO had “extracted exorbitant funds from” or otherwise abused FPSC or Sanford. Id. And the record does not show that either FPSC or Sanford was undercapitalized. On appeal, Duke Energy argues that “[t]he district court misunderstood the improper conduct at issue,” and does not pursue this theory. Duke Energy argues instead that the release of hazardous waste at the two plants was itself improper conduct. The Second Circuit appears to have accepted a version of this argument in RG&E, stating: The district court found that coal tar was an inevitable byproduct of RG&E’s manufactured gas production; that leakage and soil contamination were “inherent” in the storage methods used at the relevant time; and that AGECO so dominated RG&E that “the actions of RG&E were the actions of AGECO” during the period in question. Under these circumstances, a decision to produce gas was, in effect, a decision to pollute, and that decision to cause harm was effectively AGECO’s. RG&E II, 355 F. App’x at 550 (citation omitted). This argument would carry the day if the issue were whether the operators of these plants could be held liable under CERCLA for the release of this hazardous waste. CERCLA applies retroactively, and strict liability under CERCLA attaches and is passed on to corporate successors regardless of fault or foreseeability. See Franklin Cty. Conv. Facilities Auth., 240 F.3d at 550–53; Anspec Co. v. Johnson Controls, Inc., 922 F.2d 1240, 1245 (6th Cir. 1991). This argument might also prove successful if the issue before us were AGECO’s direct liability -18- No. 17-3024 Duke Energy, LLC v. FirstEnergy Corp. for the operations of its subsidiary. See United States v. Kayser-Roth Corp., 910 F.2d 24, 26–28 (1st Cir. 1990). Under such a theory, evidence showing that AGECO was an operator—for example, that it controlled FPSC and Sanford’s day-to-day operations, capital expenditures, or environmental responses—could lead to the conclusion that AGECO was liable under CERCLA simply because the waste was released. But we are not presented with that evidence, and the parties have stipulated that Duke Energy is not pursuing a direct liability theory. Under the indirect liability theory presented, we are asked to determine whether releasing this hazardous waste was using the corporate form for an “illegal, fraudulent or other unjust purpose” such that veil-piercing is permissible under Florida law. Sykes, 450 So. 2d at 1121 (quoting Roberts’ Fish Farm, 153 So. 2d at 721). The facts presented do not show that AGECO was purposefully avoiding environmental liability or cutting costs at the expense of the environment. This waste was released decades before most major environmental legislation, including CERCLA, was passed. The record does not contain evidence tending to show that AGECO officials were even aware of the environmental costs of their business model. As discussed above, the lapse of decades does not undermine the conclusion that CERCLA liability attaches to a site’s former operator—but the lapse of time and resulting sparsity of the record undermines the conclusion that AGECO officials’ subjective motivations in releasing the waste were fraudulent or improper. Duke Energy has not shown by a preponderance of the evidence that AGECO officials had a fraudulent or improper purpose in allowing the hazardous waste to be released. Thus, on the record before us, veil piercing is not warranted under Florida’s corporate law. We therefore need not reach the third prong of the veil-piercing inquiry or the question of whether veil piercing could continue after the bankruptcy filing. -19- No. 17-3024 Duke Energy, LLC v. FirstEnergy Corp. III. CONCLUSION For the foregoing reasons, we find that piercing the corporate veil and holding AGECO’s successor, FirstEnergy, indirectly liable for cleanup at the sites operated by its subsidiaries during the period between 1929 and 1943 is not warranted under Florida law. We therefore AFFIRM the decision of the district court. -20-