SodexoMAGIC LLC v. Drexel University

PRECEDENTIAL UNITED STATES COURT OF APPEALS FOR THE THIRD CIRCUIT _______________ Nos. 19-1028 & 19-1107 _______________ SODEXOMAGIC, LLC v. DREXEL UNIVERSITY SodexoMAGIC, LLC, Appellant in No. 19-1028 Drexel University, Appellant in No. 19-1107 _______________ On Appeal from the United States District Court for the Eastern District of Pennsylvania (D.C. No. 2:16-cv-05144) District Judge: Honorable Michael M. Baylson _______________ Argued: March 16, 2020 Before: AMBRO, BIBAS, and PHIPPS, Circuit Judges. (Filed: January 20, 2022) _______________ Shay Dvoretzky [ARGUED] Skadden Arps Slate Meagher & Flom 1440 New York Avenue, N.W. Washington, DC 20005 Attorney for SodexoMAGIC, LLC Jared D. Bayer Stephen A. Cozen [ARGUED] Robert W. Hayes Cozen O’Connor 1650 Mark Street One Liberty Place, Suite 2800 Philadelphia, PA 19103 Attorneys for Drexel University __________ OPINION OF THE COURT __________ PHIPPS, Circuit Judge. TABLE OF CONTENTS I. Introduction ................................................................... 5 II. Factual Background ...................................................... 6 A. The Competition to Provide Food Services at Drexel University...................................................................... 6 B. Negotiation of the Management Agreement................. 8 C. The Short Unhappy Life of the Management Agreement ..................................................................... 9 III. Procedural History and Jurisdictional Analysis .......... 12 2 A. The Parties’ Claims and Counterclaims ..................... 12 B. The District Court’s Jurisdiction over the Dispute ..... 13 C. The District Court’s Resolution of All Claims and Counterclaims ............................................................. 15 D. Appellate Jurisdiction ................................................. 16 IV. Discussion ................................................................... 17 A. The Parties’ Competing Fraudulent Inducement Claims Have Different Fates: SodexoMAGIC’s Claim Survives; Drexel’s Counterclaim Does Not. .... 19 1. Common-Law Fraud Claims in Pennsylvania ..... 20 2. SodexoMAGIC’s Fraudulent Inducement Claim for Compensatory Damages Survives Summary Judgment............................................................... 23 a. SodexoMAGIC Presents Sufficient Evidence of a Misrepresentation as Well as Concealment. ...... 23 b. The District Court Did Not Abuse Its Discretion in Denying Drexel’s Motion to Strike Declarations by Three SodexoMAGIC Witnesses. ................... 28 c. Drexel’s Remaining Counterarguments for Upholding Summary Judgment in Its Favor Also Fail. .............................................................. 34 3. The Parol Evidence Rule Does Not Bar Sodexo- MAGIC’s Claim for Fraudulent Inducement. ...... 36 a. Integration Clauses, the Parol Evidence Rule, and Fraudulent Inducement Claims Under Pennsylvania Law. ................................................ 37 3 b. The Management Agreement Lacks Fraud- Insulating Provisions, so the Parol Evidence Rule Does Not Preclude SodexoMAGIC’s Fraudulent Inducement Claim. ............................................... 42 4. Pennsylvania’s Gist of the Action Doctrine Does Not Bar SodexoMAGIC’s Fraud Claim...... 44 5. Drexel’s Fraudulent-Inducement Counterclaim Fails ...................................................................... 47 B. SodexoMAGIC’s Breach-of-Contract Claim for Failure to Renegotiate in Good Faith Survives Summary Judgment. ................................................... 51 1. A Promise to Renegotiate in Good Faith May Be Enforceable Under Pennsylvania Law, and the Promise Between These Parties Is Enforceable. .. 51 2. A Genuine Dispute of Material Fact Remains as to Whether Drexel Renegotiated in Good Faith. .. 59 C. SodexoMAGIC’s Claim for Enhanced Payments for Fall 2016 Survives Summary Judgment. .................... 65 1. There Was Consideration for a Separate Contract for the Fall 2016 Semester. .................... 65 2. A Reasonable Jury Could Find that SodexoMAGIC Accepted Drexel’s Offer. ........... 66 D. Drexel’s Challenge to SodexoMAGIC’s Catering- Shortfall Claim Fails. .................................................. 68 E. SodexoMAGIC’s Claims for Unjust Enrichment Fail. ............................................................................. 70 V. Conclusion .................................................................. 73 4 I. INTRODUCTION After a long-standing business relationship went bad, this, the ensuing litigation, went big. For years, a vendor provided food services at a private university, but in 2014 the university announced that it would competitively bid the contract for on- campus dining. Although the same vendor ultimately won that competition, the process of bidding, negotiating, and finalizing that new contract fractured the relationship beyond repair. About two years into the contract’s ten-year period of performance, the vendor sued the university for fraud, multiple breaches of contract, and alternatively for unjust enrichment. The university responded with fraud and breach-of-contract counterclaims. In resolving cross-motions for summary judgment and attendant motions to strike, the District Court rejected the bulk of both parties’ claims. All that survived summary judgment were relatively small pieces of the vendor’s breach-of-contract claims and portions of the university’s breach-of-contract claim. Rather than proceed to trial on the fragments of their respective cases, the parties referred the remaining claims and counterclaims to arbitration and jointly moved to dismiss them. The District Court granted that motion and entered final judgment, which the parties now appeal, primarily to dispute the summary judgment ruling. In reviewing the District Court’s summary judgment rulings de novo, see Cranbury Brick Yard, LLC v. United States, 943 F.3d 701, 708 (3d Cir. 2019), and the motion-to- strike order for an abuse of discretion, see Daubert v. NRA Grp., LLC, 861 F.3d 382, 389 (3d Cir. 2017), the District Court 5 correctly resolved much of this controversy, but it erred in rejecting the vendor’s fraud and breach-of-contract claims. Thus, we will affirm the judgment in part, vacate in part, and remand this case for further proceedings. II. FACTUAL BACKGROUND A. The Competition to Provide Food Services at Drexel University For nearly twenty years, either directly or through one of its predecessor companies, SodexoMAGIC, LLC (‘SDM’) provided on-campus dining services at Drexel University, a private university in Philadelphia. But in early 2014 one of SDM’s rivals, Aramark Corporation, which has its global headquarters in Philadelphia, made an unsolicited offer to Drexel to take over campus dining services. SDM responded with its own unsolicited offer to continue providing food services. Rather than select one of those offers, Drexel announced a two-phase competitive bidding process for the on- campus dining contract. In the first phase, Drexel sought to identify qualified bidders. It proposed a scope of work that involved providing “world class dining and catering services” at its campuses. Drexel Univ., Request for Proposal for a Campus Dining Strategic Partnership (SA133). Drexel also expected the food- service provider to make capital investments of $20 million over the span of the ten-year contract. While Drexel’s phase- one solicitation did not contain many details about on-campus dining, it did represent that Drexel would share “as much detail about the desired relationship and facts around the current program as can be reasonably provided.” Id. (SA131). The 6 solicitation also included an estimate that the contract, over its full term, would have a value between $275 and $300 million. At the end of the first phase, Drexel determined that SDM and Aramark, along with two other bidders, qualified to advance to the second phase. The second phase took place during the summer of 2014, and it involved competitive negotiations. In its more detailed phase-two solicitation, Drexel provided an overview of its on- campus dining and food-service operations. That overview pointed out that Drexel’s Strategic Plan called for an increase of its then-current enrollment of 26,132 students to a student population of 34,000 by 2021. Drexel noted, however, that not all students had to have all-inclusive meal plans – that requirement applied only to first-year undergraduates. And an appendix to the solicitation informed bidders that Drexel was projecting an incoming first-year class of 3,100 students for the 2014–15 academic year. But for purposes of its own internal budget, Drexel estimated that class size at 2,800 students. As succinctly stated in internal correspondence from one of Drexel’s senior associate vice presidents, “we gave the bidders participant numbers based on a freshman class size of 3,100 and the FY15 budget that was loaded by Finance is based on 2,800.” Email from Joe Campbell, Senior Associate Vice President, Drexel, to Rita LaRue, Senior Associate Vice President, Drexel (Aug. 4, 2014) (SA182). Three of the remaining bidders submitted proposals. Drexel selected two of them – SDM and Aramark – as finalists and requested that they each submit a best and final offer, referred to as a ‘BAFO.’ In its BAFO, SDM included an additional term: an offer to increase its capital contributions by $4 million beginning in year six of the contract. After 7 considering both BAFOs, Drexel determined that SDM’s BAFO had more favorable financial terms and was superior in other respects. On August 15, 2014, Drexel selected SDM for the dining services contract and broke off negotiations with Aramark. Having won the competition, SDM continued to provide on-campus dining services for the 2014–15 academic year while it worked with Drexel to finalize a new contract. B. Negotiation of the Management Agreement Finalizing the new contract was not a smooth process. Problems started in September 2014 when SDM proposed an initial draft that omitted the additional $4 million capital contribution that it had proposed in its BAFO. After several months of negotiations, the parties agreed to condition that investment on certain sales levels, evaluated before year six of the contract. But that was not the only point of contention. SDM also proposed that Drexel guarantee annual increases in its first- year student enrollment. Instead of a guarantee, the parties agreed that SDM had relied on Drexel’s representations of its future student enrollment and that they would renegotiate in good faith if first-year student enrollment did not increase annually by at least two percent. Having overcome those and other obstacles, the parties executed the contract, known as the ‘Management Agreement,’ on May 28, 2015. Although it was signed then, the Management Agreement specified a commencement date of August 25, 2014, so that it covered the services SDM provided after it had won the competitive bidding process. The Management Agreement also included an integration clause, 8 which stated that it contained all agreements between the parties on the subject matter. C. The Short Unhappy Life of the Management Agreement The Management Agreement may have been doomed from the start. The day it was executed, The Philadelphia Inquirer reported that Drexel’s first-year class had nearly 200 fewer students than the prior year. See Susan Snyder, That Big Enrollment Change at Drexel: How Did it Go?, Phila. Inquirer (May 28, 2015) (SA1553–54). Soon afterward, through a ‘Dear Colleague’ letter, Drexel’s President informed university stakeholders, including SDM, that the university had “focused on attracting a smaller pool of exceptionally qualified applicants,” with the result being that the “enrolled class will be smaller than in previous years,” causing the university to operate “with less revenue than [it] historically experienced.” Letter from John A. Fry, President, Drexel, to Colleagues of Drexel (June 8, 2015) (SA519–20). According to SDM, that news of reduced student enrollment came as “a complete shock.” Email from Nancy C. Arnett, Vice President, Sodexo, to Greg Klose, Vice President, Sodexo (June 12, 2015) (SA524). And when SDM asked about that development, Drexel’s lead negotiator for the Management Agreement forwarded the inquiry to one of Drexel’s executive directors, who remarked, “I guess they were going to find out sooner or later.” Email from Donald Liberati, Executive Director, Drexel, to Rita LaRue, Senior Associate Vice President, Drexel (June 9, 2015) (SA526). Consistent with those revelations, Drexel enrolled 2,720 students in the first-year class for the 2015–16 academic year – 9 a seven percent decrease from the year before. Even with that diminution, SDM made an initial capital contribution of $9.3 million in July 2015, which funded construction of a new dining facility, the Urban Eatery at Lancaster Avenue. But the reduced first-year student enrollment had an immediate effect on SDM’s revenues. In November 2015, only six months after executing the contract, revenue from catering and meal plans fell below the assumptions referenced in the Management Agreement. And SDM requested compensation for that shortfall. For several months, the parties engaged in discussions to restructure certain aspects of the Management Agreement, including reducing the additional capital contributions that SDM had promised. As the parties’ relationship soured, the termination provisions of the Management Agreement took on greater significance. Those terms provided two mechanisms for termination before full term: for cause and for convenience. To terminate for cause, a party had to notify the other of a breach, and that would commence a cure period. If the breach was not cured or otherwise resolved by the end of the cure period, then the complaining party could terminate the Management Agreement. The termination-for-convenience provision required 60 days’ notice that a party would terminate, and upon expiration of that period, the Management Agreement would terminate. After several months without resolution, on July 25, 2016, SDM began the process to terminate the Management Agreement for cause. It notified Drexel that it believed the shortfalls were material breaches on the theory that sales fell 10 below the amounts listed as assumptions in the Management Agreement. SDM then tolled the cure period, and the parties continued discussions through August 2016. Upon receiving SDM’s notice of breach, Drexel also began pursuing other options. Two days after SDM’s notice, Drexel reconnected with other bidders to explore their interest in taking over the on-campus food-services operations. Using five-year projections for an incoming first-year class of 2,400 students, Drexel and Aramark discussed the possibility that Aramark would start providing on-campus food services in January 2017. As that possibility became more likely, and without reconciling completely with SDM, Drexel exercised its option to terminate the Management Agreement for convenience. Although the Management Agreement required only 60 days’ notice to terminate for convenience, by letter dated September 19, 2016, Drexel provided 82 days’ notice: it stated its intention to terminate the Management Agreement for convenience on December 10, 2016, the last day of the fall semester. By that same correspondence, Drexel offered SDM various financial incentives to remain on campus and provide food services for the rest of the Fall 2016 Semester. Those incentives included an increased daily rate for certain meal plans, a reduced commission due to Drexel, and a release of SDM’s obligation to make additional capital contributions under the Management Agreement. The days following Drexel’s notice of termination were eventful. On September 23, 2016, senior staff at Drexel met with Aramark and conditionally agreed to contract with Aramark for on-campus dining services starting in January 11 2017. On September 26, 2016, SDM responded to Drexel’s notice of termination. In that letter, SDM gave notice that it was ending the cure period, and it purported to terminate the Management Agreement for cause effective at the close of business that day. But through that same communication, SDM also agreed to remain on campus for the remainder of the Fall 2016 Semester, explaining that it “does not leave students in mid-term.” Letter from Timothy J. Fazio, Manion Gaynor & Manning, LLP, to Stephen A. Cozen, Cozen O’Connor P.C. (Sept. 26, 2016) (JA158). The next day, SDM filed this lawsuit in the Eastern District of Pennsylvania. Despite suing Drexel, SDM remained on campus and provided services through December 10, 2016. Drexel did not accept a reduced commission or make enhanced payments for dining services provided by SDM, and SDM did not make any further capital contributions. III. PROCEDURAL HISTORY AND JURISDICTIONAL ANALYSIS A. The Parties’ Claims and Counterclaims In its original complaint, SDM pursued three causes of action. First, it alleged that Drexel fraudulently induced it to enter the Management Agreement, and it sought compensatory and punitive damages. Second, SDM alleged that Drexel breached its contractual duty to renegotiate in good faith. Third, as a claim in the alternative, SDM alleged that Drexel was unjustly enriched when SDM assumed certain liabilities and provided dining services at reduced rates. 12 Drexel also sought to vindicate its grievances against SDM. It counterclaimed that SDM breached the Management Agreement by failing to perform specific obligations, including filling key management positions, making financial contributions to certain Drexel initiatives, meeting certain performance standards, and securing campus appearances from Earvin “Magic” Johnson. Both parties later revised their initial pleadings. SDM amended and supplemented its complaint to add another breach-of-contract claim. In that count, SDM alleged that Drexel refused to pay invoices due under the Management Agreement and that Drexel owed SDM enhanced payments for dining services provided during the Fall 2016 Semester. After discovery, on September 12, 2017, Drexel supplemented its counterclaim to allege that SDM’s offer of additional capital contributions in its BAFO was insincere and that through that term SDM fraudulently induced Drexel to award the contract to SDM instead of Aramark. B. The District Court’s Jurisdiction over the Dispute Under the diversity statute, the District Court had subject matter jurisdiction over this case. That statute requires complete diversity of citizenship between opposing parties and an amount in controversy in excess of $75,000. See 28 U.S.C. § 1332(a); Exxon Mobil Corp. v. Allapattah Servs., Inc., 545 U.S. 546, 553 (2005) (explaining that the presence of a single plaintiff from the same state as a single defendant deprives a district court of original diversity jurisdiction); Strawbridge v. Curtiss, 7 U.S. (3 Cranch) 267, 267 (1806). Both of those jurisdictional requirements are satisfied here. 13 There is complete diversity because SDM and Drexel are not citizens of the same state. As a limited liability corporation, SDM takes on the citizenship of its members and submembers. See Zambelli Fireworks Mfg. Co. v. Wood, 592 F.3d 412, 420 (3d Cir. 2010) (explaining that determining an LLC’s citizenship requires drilling down “through however many layers of partners or members there may be” to evaluate the citizenship of each (internal quotation marks omitted)); cf. Lincoln Benefit Life Co. v. AEI Life, LLC, 800 F.3d 99, 111 (Ambro, J., concurring) (“There is no good reason to treat LLCs differently from corporations for diversity-of-citizenship purposes.”). Accounting for those members and submembers, SDM is a citizen of Delaware, Maryland, and California for purposes of the diversity statute.1 As a non-profit corporation, Drexel has citizenship for purposes of the diversity statute in the state of its incorporation and in the state of its principal place of business – Pennsylvania in both instances. See 28 U.S.C. § 1332(c)(1); see also Zambelli, 592 F.3d at 419. Without any overlap in the citizenship of SDM and Drexel, the complete diversity requirement is satisfied. 1 SDM has two members: Sodexo Operations, LLC and Magic Food Provision, LLC. Sodexo Operations, LLC has one member, Sodexo, Inc., which, as a corporation, has citizenship in Delaware as its state of incorporation and in Maryland through its principal place of business. See 28 U.S.C. § 1332(c)(1); see also Zambelli, 592 F.3d at 419. SDM’s other member, Magic Food Provision, LLC, is a wholly owned subsidiary of Magic Johnson Enterprises, Inc., that is both incorporated and has its principal place of business in California. 14 The amount in controversy also exceeds the $75,000 threshold in the diversity statute. See 28 U.S.C. § 1332(a). The multi-million-dollar value of the Management Agreement coupled with the nature of SDM’s claims for compensatory damages, expectation damages, and punitive damages precludes a finding that to “a legal certainty” SDM seeks a recovery less than the jurisdictional amount. Auto-Owners Ins. Co. v. Stevens & Ricci Inc., 835 F.3d 388, 395 (3d Cir. 2016) (quoting St. Paul Mercury Indem. Co. v. Red Cab Co., 303 U.S. 283, 288–89 (1938)). Accordingly, this case also clears the $75,000 amount-in-controversy threshold. C. The District Court’s Resolution of All Claims and Counterclaims The District Court resolved much of this dispute by partially granting and partially denying the parties’ cross- motions for summary judgment. That ruling decimated SDM’s case. It lost its fraudulent inducement claim, its breach-of- contract claims (which were based on Drexel’s alleged failure to renegotiate in good faith and failure to pay enhanced compensation for services during the Fall 2016 Semester), and its alternative claim for unjust enrichment. The District Court reached those outcomes despite denying Drexel’s motion to strike three declarations by SDM employees under the sham affidavit rule. But Drexel lost more than that motion: the District Court also rejected its fraudulent inducement counterclaim. Shortly before the scheduled trial on the remaining counts, SDM moved for clarification on whether its breach-of-contract claim for failure to pay a catering-shortfall invoice survived 15 summary judgment. Over Drexel’s objection, the District Court permitted that claim. With that issue resolved, both parties agreed to arbitrate the remaining claims and counterclaims. To that end, they jointly filed a motion to dismiss and attached exhibits identifying the remaining claims and counterclaims. The District Court granted that motion and, through a separate document, entered final judgment on December 6, 2018. See Fed. R. Civ. P. 58(a). Both parties then appealed. D. Appellate Jurisdiction As timely challenges to a final order, both appeals come within this Court’s appellate jurisdiction. See 28 U.S.C. § 1291. The District Court’s judgment on December 6, 2018, was final and appealable because all outstanding claims and counterclaims had been resolved through the combination of the summary judgment ruling and the subsequent order dismissing the remaining claims. See Aluminum Co. of Am. v. Beazer E., Inc., 124 F.3d 551, 559–62 (3d Cir. 1997) (holding that an order dismissing and referring to private arbitration all unresolved claims achieved finality). SDM’s notice of appeal, filed on January 2, 2019, was within the time permitted by rule. See Fed. R. App. P. 4(a)(1)(A) (requiring parties in a civil case to file a notice of appeal within 30 days after entry of the judgment). Drexel’s notice of cross-appeal, filed on January 11, 2019, was also timely. See Fed. R. App. P. 4(a)(3) (providing that a notice of cross-appeal must be filed within 14 days of another party’s notice of appeal or within the time otherwise remaining for the other party to appeal, whichever is longer). 16 IV. DISCUSSION The summary judgment standard has not substantively changed since a trilogy of Supreme Court cases on the topic in 1986.2 By the text of Rule 56, summary judgment is appropriate “if the movant shows that there is no genuine dispute as to any material fact and the movant is entitled to judgment as a matter of law.” Fed. R. Civ. P. 56(a); see also Brooks v. Kyler, 204 F.3d 102, 105 n.5 (3d Cir. 2000). As explained by the Supreme Court, for a factual dispute to be material, its resolution must have the potential to affect the outcome of the suit. See Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 248 (1986). A dispute is genuine “if the evidence is such that a reasonable jury could return a verdict for the nonmoving party,” id., but “the mere existence of a scintilla of evidence” favoring the non-moving party will not prevent summary judgment, id. at 252. See also Jutrowski v. Twp. of Riverdale, 904 F.3d 280, 288–89 (3d Cir. 2018). Still, in assessing the genuineness of a potential factual dispute, inferences from the underlying facts should be drawn in favor 2 See Celotex Corp. v. Catrett, 477 U.S. 317 (1986); Anderson v. Liberty Lobby, Inc., 477 U.S. 242 (1986); Matsushita Elec. Indus. Co. v. Zenith Radio Corp., 475 U.S. 574 (1986); see also Fed. R. Civ. P. 56 advisory committee’s note to 1987 amendment (explaining that “[n]o substantive change is intended”); id. advisory committee’s note to 2007 amendment (explaining that “[t]hese changes are intended to be stylistic only”); id. advisory committee’s note to 2009 amendment (consolidating and revising the timing provisions for summary judgment, but not affecting the standard); id. advisory committee’s note to 2010 Amendment (“The standard for granting summary judgment remains unchanged.”). 17 of the nonmoving party. See Matsushita Elec. Indus. Co. v. Zenith Radio Corp., 475 U.S. 574, 587 (1986); In re IKON Office Solutions, Inc., 277 F.3d 658, 666 (3d Cir. 2002). But if the nonmoving party “fails to make a showing sufficient to establish the existence of an element essential to [its] case, and on which [it] will bear the burden of proof at trial,” then summary judgment is appropriate for the moving party. Celotex Corp. v. Catrett, 477 U.S. 317, 322 (1986). In exercising diversity jurisdiction, a federal court employs the choice-of-law principles of its forum state to determine which substantive law governs whether a party is entitled to judgment as a matter of law. See Klaxon Co. v. Stentor Elec. Mfg. Co., 313 U.S. 487, 496–97 (1941). Under Pennsylvania choice-of-law rules, the first step involves assessing whether a conflict exists between the substantive law of multiple jurisdictions. See Auto-Owners, 835 F.3d at 404. But here, the parties have identified no such conflict and have litigated this case under Pennsylvania substantive law. Thus, substantive issues should be decided as the Pennsylvania Supreme Court “would rule if it were deciding this case.” Id. at 403 (quoting Norfolk S. Ry. Co. v. Basell USA Inc., 512 F.3d 86, 91–92 (3d Cir. 2008)); see generally Hanna v. Plumer, 380 U.S. 460, 471 (1965) (providing “rough[]” guidance that in diversity cases “federal courts are to apply state ‘substantive’ law and federal ‘procedural’ law”). In making that prediction, the decisions of intermediate Pennsylvania appellate courts receive “significant weight in the absence of an indication that the highest state court would rule otherwise.” Polselli v. Nationwide Mut. Fire Ins. Co., 126 F.3d 524, 528 n.3 (3d Cir. 1997) (quoting City of Phila. v. Lead Indus. Ass’n, 994 F.2d 112, 123 (3d Cir. 1993)); see also Pac. Emps. Ins. Co. v. Glob. Reinsurance Corp. of Am., 693 F.3d 417, 433 (3d Cir. 2012). 18 A. The Parties’ Competing Fraudulent Inducement Claims Have Different Fates: SodexoMAGIC’s Claim Survives; Drexel’s Counterclaim Does Not. The parties both sued each other for fraudulent inducement. In its fraud claim against Drexel, SDM alleged that Drexel misrepresented and concealed its future student-enrollment projections, which led SDM to bid more favorably on the Management Agreement. Drexel countersued for fraudulent inducement on allegations that SDM’s BAFO included capital contributions that SDM never intended to make and that such deceit prompted Drexel to negotiate exclusively with SDM, not Aramark. Both parties moved for summary judgment against their opponent’s fraud claims. At the prompting of the District Court, they each argued that the parol evidence rule did not bar their fraud claim. They also asserted that the gist of the action doctrine precluded the other’s fraud claim. SDM additionally argued that Pennsylvania’s statute of limitations foreclosed Drexel’s counterclaim. The District Court rejected the parties’ competing fraud claims at summary judgment. See SodexoMAGIC, LLC v. Drexel Univ., 333 F. Supp. 3d 426, 456, 466–67 (E.D. Pa. 2018). In so doing, the District Court found that SDM produced evidence of a prima facie fraud claim but entered summary judgment for Drexel due to the parol evidence rule and the gist of the action doctrine. Id. at 445–57. The District Court relied on those same affirmative defenses to reject Drexel’s fraud claim. Id. at 466–67. In ruling for SDM on 19 those grounds, the District Court did not address SDM’s statute-of-limitations defense. On appeal, each party challenges the District Court’s fraud rulings. SDM argues that the District Court misapplied the parol evidence rule and the gist of the action doctrine to its own fraud claim. Consistent with that position, SDM does not defend the District Court’s reliance on those principles to bar Drexel’s fraud claim. Instead, it offers two alternative bases for sustaining the judgment against Drexel’s fraud claim: it renews the statute-of-limitations defense that the District Court did not consider, and it argues for the first time on appeal that Drexel has not produced evidence needed for the justifiable- reliance and causation elements of a prima facie fraud case. Drexel takes a more nuanced approach. It contends that the parol evidence rule and the gist of the action doctrine do not apply to its fraud counterclaim but that they do bar SDM’s fraud claim. As an alternative, Drexel argues for the first time on appeal that SDM did not demonstrate a misrepresentation needed for a prima facie fraud case. For the reasons below, SDM’s fraud claim survives summary judgment, but Drexel’s fraud counterclaim does not. 1. Common-Law Fraud Claims in Pennsylvania To redress injuries caused by intentional falsehoods – either false statements or deliberately concealed facts – Pennsylvania recognizes a civil action for fraud. See Moser v. DeSetta, 589 A.2d 679, 682 (Pa. 1991) (“The concealment of a material fact can amount to a culpable misrepresentation no less than does an intentional false statement.”). As developed in Pennsylvania common law, a fraud claim consists of six elements: 20 (1) (a) A misrepresentation or (b) A concealment; (2) Which is material to the transaction at hand; (3) (a) Made with knowledge of its falsity or recklessness as to whether it is true or false (for a misrepresentation), or (b) Calculated to deceive (for a conceal- ment); (4) With the intent of misleading another into relying on it; (5) Justifiable reliance on the misrepresenta- tion; and (6) A resulting injury proximately caused by such reliance. See Gibbs v. Ernst, 647 A.2d 882, 889 & n.12 (Pa. 1994); Youndt v. First Nat’l Bank of Port Allegany, 868 A.2d 539, 545 (Pa. Super. Ct. 2005). To prevail, a fraud plaintiff must prove each of those elements by clear and convincing evidence. See Moser, 589 A.2d at 682; Gerfin v. Colonial Smelting & Refin. Co., 97 A.2d 71, 72 (Pa. 1953). Upon such proof, a fraud plaintiff may recover compensatory damages. See Neuman v. Corn Exch. Nat’l Bank & Tr. Co., 51 A.2d 759, 766 (Pa. 1947) (explaining that compensatory damages are recoverable for fraudulent misrepresentation claims); Smith v. Renault, 564 A.2d 188, 193 (Pa. Super. Ct. 1989) (same); see also McShea v. City of Phila., 995 A.2d 334, 347 n.5 (Pa. 2010) (Baer, J., dissenting) (explaining that a “tort plaintiff usually recovers the actual damages or compensatory damages that she suffered because of the tort”). 21 A successful fraud plaintiff may also recover punitive damages. But to do so requires the additional showing “that the defendant has acted in an outrageous fashion due to either the defendant’s evil motive or [its] reckless indifference to the rights of others.” Phillips v. Cricket Lighters, 883 A.2d 439, 445 (Pa. 2005) (internal quotation marks omitted); see also Feld v. Merriam, 485 A.2d 742, 747–48 (Pa. 1984) (“Punitive damages must be based on conduct which is malicious, wanton, reckless, willful, or oppressive.” (internal quotation marks omitted)). Thus, proof of the six elements of fraud without additional evidence of outrageous conduct enables a fraud plaintiff to recover compensatory damages but not punitive damages. See Pittsburgh Live, Inc. v. Servov, 615 A.2d 438, 442 (Pa. Super. Ct. 1992) (explaining that even when fraud supports an award of compensatory damages, “the same fraudulent conduct is not sufficient to [support] an award of punitive damages without more”). The six elements of fraud apply to the two subspecies of fraud claims related to the formation of contracts: fraudulent inducement and fraud in the execution. A claim for fraudulent inducement requires proof of the six elements and is available when a person under no duty to enter a contract was deceived into doing so. See Coll. Watercolor Grp., Inc. v. William H. Newbauer, Inc., 360 A.2d 200, 206 (Pa. 1976) (quoting Restatement (Second) of Contracts § 476 (1932)); see also Justin Sweet, Promissory Fraud and the Parol Evidence Rule, 49 Calif. L. Rev. 877, 888 (1961) (“Fraud in the inducement occurs when one party, by means of false statements of fact, warranties, or promises, misleads another into contracting.”). A claim for fraud in the execution consists of the same six elements but with a different focus: proof that a party “was 22 mistaken as to the terms and actual contents of the agreement [it] executed due to the other’s fraud.” Toy v. Metro. Life Ins. Co., 928 A.2d 186, 206 (Pa. 2007); see also Sweet, Promissory Fraud, supra, at 888 (defining “fraud in the execution” as “deception by one party about the contents of the written instrument”). Here, both parties’ claims are for fraudulent inducement, not fraud in the execution. 2. SodexoMAGIC’s Fraudulent Inducement Claim for Compensatory Damages Survives Summary Judg- ment. a. SodexoMAGIC Presents Sufficient Evidence of a Misrepresentation as Well as Concealment. In attacking SDM’s fraudulent inducement claim, Drexel argues that SDM cannot prove the misrepresentation element. But SDM produces evidence that would allow a jury to find a misrepresentation or concealment by clear and convincing evidence. The theme of SDM’s fraudulent inducement claim is that Drexel provided it with one set of projections for future student enrollment but then used another set of projections for its internal purposes. SDM cites Drexel’s phase-two solicitation from July 2014 that included an estimate of an incoming first- year class of 3,137 students for the 2014–15 academic year. And SDM also produces other evidence that, less than a month later, in August 2014, Drexel calculated its budget based on an estimated first-year class of 2,800 students, roughly 300 less students than the figure it provided SDM. Ultimately, 2,926 first-year students enrolled at Drexel for the 2014–15 academic year. 23 Those figures show a disconnect between what Drexel told bidders during the summer of 2014 and its actual student enrollment projections. But SDM did not execute the Management Agreement until May 2015 – after it provided on- campus dining services for the 2014–15 academic year. Thus, even if Drexel provided a false projection of its first-year students for the 2014–15 academic year, SDM would struggle mightily to prove that it justifiably relied on that figure in May 2015, when it finalized the Management Agreement. See Toy, 928 A.2d at 207 (explaining that a person “is not justified in relying upon the truth of an allegedly fraudulent misrepresentation if he knows it to be false or if its falsity is obvious”). But that is not SDM’s only evidence of Drexel’s deception regarding its long-term student-enrollment projections. In its initial solicitation for bids, Drexel represented that it would share “as much detail about the desired relationship and facts around the current program as can be reasonably provided.” Request for Proposal (SA131). And in providing information to the bidders, Drexel noted that its Strategic Plan called for an enrollment increase from 26,132 students to 34,000 students by 2021. Similarly, three SDM employees, who were involved in various phases of the bidding and negotiation process for the Management Agreement, signed declarations that Drexel “repeatedly represented . . . that freshman enrollment would continue to grow over the ten-year life of the contract, or at the very least, enrollment would stay flat for the first three years.” Sherman Decl. ¶ 23 (JA247); Riccio Decl. ¶ 23 (JA254–55); Arnett Decl. ¶ 9 (JA259). Those same employees also averred that Drexel never told them during negotiations “that there was a risk of enrollment decline” or “that enrollment was already 24 declining, and that Drexel was budgeting for that decline.” Sherman Decl. ¶ 22 (JA247); Riccio Decl. ¶ 22 (JA254); Arnett Decl. ¶ 9 (JA259). But at their prior depositions those same employees testified somewhat differently, with one of them recounting that Drexel represented that student enrollment “might go up” or “it might go down.” Sherman Dep. 225:17–226:1 (JA200–01). Some of SDM’s strongest evidence of deceit comes not from its own witnesses but from Drexel’s employees. In email correspondence, Drexel’s Executive Vice President recounted that “[w]hen [Drexel] contracted with Sodexo two years ago, we never told them we were dialing back our freshman enrollment.” Email from Helen Bowman, Executive Vice President, Drexel, to Robert Francis, Drexel (Apr. 22, 2016) (SA529). Another internal memorandum prepared in early May 2015 reveals that Drexel expected that its first-year enrollment would decline in 2015–16, leading to an even lower budgetary forecast of a 2,600 student first-year class. But Drexel was not entirely secretive about its plan to decrease first-year student enrollment while it negotiated the Management Agreement. A March 2015 article from a leading industry news provider reported that Drexel was “making major changes in admissions.” Ry Rivard, Drexel U. Charts A New Course For Itself, Inside Higher Ed (Mar. 27, 2015) (SA1546). The article explained that after it received a report from a consulting firm in 2013, Drexel decided to shift approaches away from its enrollment target of 34,000 students by 2021, and it “was now looking to slow growth.” SA1546– 47. 25 Nonetheless, SDM alleges that it did not gain insight into the full extent of Drexel’s student enrollment decline until after it signed the Management Agreement. The same day that the parties executed the Management Agreement, The Philadelphia Inquirer published an article in which Drexel’s Senior Vice President for Enrollment Management reported that in the Spring, Drexel had “overhauled its admission process” and now anticipated a decline in freshman enrollment for Fall 2015. Snyder, Enrollment Change at Drexel, supra (SA1553–54). Shortly afterwards, in June 2015, Drexel’s President emailed the Drexel community, including SDM, with notice not only that it was now expecting a decline in first-year student enrollment but also that Drexel had “focused on attracting a smaller pool of exceptionally qualified applicants.” Email from John A. Fry, President, Drexel, to Drexel Listserv (June 8, 2015) (SA527). In response to the shock of that news, SDM emailed a request for an update on the precise enrollment for the Fall 2015 Semester first-year class, and the Executive Director of one of Drexel’s student centers remarked, “I guess they were going to find out sooner or later.” Email from Donald Liberati, Executive Director, Drexel, to Rita LaRue, Senior Associate Vice President, Drexel (June 9, 2015) (SA526). Ultimately, less than a month after executing the Management Agreement, Drexel informed SDM that its incoming class size would be lower than the year before – projecting it to be between 2,600 and 2,700 students. Considered in aggregate, while not airtight, this evidence would allow a reasonable jury to conclude that Drexel misled SDM or concealed its true intention from SDM until after finalization of the Management Agreement. If a jury resolves this genuine dispute of material fact in SDM’s favor, then SDM 26 could recover compensatory damages. See Neuman, 51 A.2d at 766. But SDM’s evidence does not permit recovery of punitive damages. Under Pennsylvania law, the same evidence used to establish fraud cannot be the sole support for an award of punitive damages. See Smith, 564 A.2d at 193 (“If the rule were otherwise, punitive damages could be awarded in all fraud cases. This is not the law.”); see also Pittsburgh Live, 615 A.2d at 442. Something more is required: outrageous conduct. See Phillips, 883 A.2d at 445. And SDM does not come forth with sufficient evidence3 that Drexel acted with an 3 The standard of proof required for punitive damages for a fraud claim is not clear under Pennsylvania law. See, e.g., Weston v. Northampton Pers. Care, Inc., 62 A.3d 947, 960–61 (2013) (explaining that fraud claims require clear-and- convincing proof but not addressing standard of proof for punitive damages); Pittsburgh Live, 615 A.2d at 441–42 (same). The likely answer is that the standard of proof for punitive damages mirrors the standard of proof for the underlying claim. Compare Martin v. Johns-Manville Corp., 494 A.2d 1088, 1098 (Pa. 1985) (applying a preponderance standard for punitive damages in a product liability case, which otherwise requires such proof), abrogated on other grounds by Kirkbride v. Lisbon Contractors, Inc., 555 A.2d 800 (Pa. 1989), with Hepps v. Phila. Newspapers, Inc., 485 A.2d 374, 389 (1984) (applying a clear-and-convincing standard for punitive damages for a defamation claim, which otherwise requires such proof), rev’d on other grounds, 475 U.S. 767 (1986), and Polselli v. Nationwide Mut. Fire Ins. Co., 23 F.3d 747, 750 (3d Cir. 1994) (applying a clear-and-convincing standard for punitive damages for a claim of bad-faith conduct 27 evil motive or that Drexel’s conduct was malicious, vindictive, or evidenced a wholly wanton disregard for SDM’s rights. See id. Without such evidence, SDM’s claim for punitive damages fails at summary judgment. See Celotex, 477 U.S. at 322. b. The District Court Did Not Abuse Its Discretion in Denying Drexel’s Motion to Strike Declara- tions by Three SodexoMAGIC Witnesses. By way of counterargument, Drexel invokes the sham affidavit rule to contend that the District Court abused its discretion by considering three declarations by SDM employees. See SodexoMAGIC, 333 F. Supp. 3d at 443 (reasoning that the witnesses’ declarations could be reconciled with their prior deposition testimony). The sham affidavit rule allows a court to disregard a later statement by a deponent on two conditions: the later statement contradicts the witness’s deposition testimony, and the discrepancy between the two statements is neither supported by record evidence nor otherwise satisfactorily explained. See Daubert, 861 F.3d at 391–92; Jiminez v. All Am. Rathskeller, Inc., 503 F.3d 247, 254 (3d Cir. 2007) (explaining that “courts generally have refused to disregard the affidavit” where there is “independent evidence in the record to bolster an otherwise questionable affidavit” (internal quotation marks omitted)); see generally 10A Charles Alan Wright, Arthur R. Miller & Mary K. Kane, by insurers, which otherwise requires such proof). If so, then SDM would need to supply clear-and-convincing evidence of outrageous conduct, and it does not. But due to the paucity of its evidence of such conduct, SDM’s punitive damages claim would similarly fail at summary judgment under the preponderance standard. 28 Federal Practice and Procedure § 2726.1 (4th ed. 2021) (“[A]n interested witness who has given clear answers to unambiguous questions cannot create a conflict and resist summary judgment with an affidavit that is clearly contradictory, without providing a satisfactory explanation of why the testimony is changed.”). On abuse-of-discretion review, the District Court’s decision not to apply the sham affidavit rule holds up – by a narrow margin. Three SDM witnesses involved in negotiating the Management Agreement testified at their depositions about representations that Drexel made about future student enrollment. One SDM witness testified that Drexel could not predict future student enrollment and that Drexel was noncommittal about future student enrollment: Q. In every meeting that you had with [Drexel’s representative], did she not say that nobody could predict what the freshman enrollment would be going forward? A. Every time – Q. Well, answer my question. A. Yes. *** Q. It could go up, it could go down. They didn’t know exactly where it would be. A. What I recollect is a noncommittal, they didn’t know exactly where it would be. 29 Riccio Dep. 101:10–15; 141:20–23 (JA206–07); see also Hunt Dep. 90:16–91:3 (JA211–12) (adopting that testimony through SDM’s designated Rule 30(b)(6) corporate witness as SDM’s official position). A second witness testified similarly – that Drexel was noncommittal and stated that future student enrollment might increase or it might decrease: Q. [Drexel’s representative] told you she couldn’t make any commitment as to what the enrollment would be, correct? A. Correct. Q. One way or the other, correct? A. Correct. Q. It might go up, it might go down? A. Correct. Sherman Dep. 225:17–226:1 (JA200–01). A third witness testified that she knew in March 2015 that Drexel’s growth initiatives did not include increasing the first-year class size: Q. The second sentence you say, “Jim needs to know that the financial model did not hurdle at the new understanding of [Drexel’s] growth initiatives (no new students – just growing retention, grad students and online students).” Do you see that? A. Yes. Q. So that’s what Drexel told you, right? 30 A. That was the knowledge that I had on 3/18. Arnett Dep. 218:12–23 (SA1744). Later, each of those witnesses signed declarations under penalty of perjury with statements about the negotiation of the Management Agreement. See Sherman Decl. (JA244–49); Riccio Decl. (JA251–56); Arnett Decl. (JA258–61); see also 28 U.S.C. § 1746. Most of those averments do not contradict the witnesses’ prior deposition testimony. For example, the declarations state that Drexel never told SDM that, following recommendations from a consulting firm, it implemented an initiative that could affect student enrollment and revenues for the first few years of its implementation. See Sherman Decl. ¶¶ 6–9, 11 (JA245–46); Riccio Decl. ¶¶ 6–9, 11 (JA252– 53); Arnett Decl. ¶¶ 6–7 (JA259). The sham affidavit rule does not apply to those noncontradictory portions of the declarations. Because summary judgment does not present an occasion to make credibility assessments, see Jiminez, 503 F.3d at 253, the sham affidavit rule does not permit striking the entirety of a later- provided affidavit or declaration since doing so would require a broader assessment of the witness’s credibility. Thus, the sham affidavit rule permits striking only contradictory statements in later-provided affidavits or declarations. Here, the witnesses’ declarations also contained statements in tension with their deposition testimony. All three witnesses averred that Drexel repeatedly represented during the bidding and negotiation process “that freshman enrollment would continue to grow over the ten-year life of the contract, or at the 31 very least, enrollment would stay flat for the first three years.” Riccio Decl. ¶ 23 (JA254–55); Sherman Decl. ¶ 23 (JA247); Arnett Decl. ¶ 9 (JA259). The declarations also state that the witnesses first learned that Drexel expected a significant decline in first-year student enrollment through the ‘Dear Colleague’ letter sent after SDM had executed the Management Agreement. See Riccio Decl. ¶ 31 (JA256); Sherman Decl. ¶ 31 (JA249); Arnett Decl. ¶ 18 (JA261). For the first two witnesses, it is hard to reconcile their later statements with their prior deposition testimony. At their depositions, they testified that Drexel was non-committal about future student enrollment. See Riccio Dep. 101:10–15; 141:20–23 (JA206–07); Sherman Dep. 225:17–226:1 (JA200– 01). That is different from their later statements that Drexel consistently predicted increased or steady student enrollment. See Riccio Decl. ¶ 23 (JA254–55); Sherman Decl. ¶ 23 (JA247). But the sham affidavit rule does not reject later statements solely because they conflict with prior deposition testimony. See Baer v. Chase, 392 F.3d 609, 624 (3d Cir. 2004) (“[M]erely because there is a discrepancy between deposition testimony and the deponent’s later affidavit a district court is not required in all cases to disregard the affidavit.”). When a later statement is supported by independent record evidence, courts generally refuse to strike those statements. See id. at 625. And record evidence indicates that in its initial phase-one solicitation Drexel predicted that student enrollment would increase through 2021. Permitting consideration of the third witness’s seemingly contradictory statements also requires a fine line. That witness testified that she had a “new understanding” of Drexel’s student enrollment in March 2015. Arnett Dep. 218:12–23 32 (SA1744). But in her later declaration she stated that she first learned that Drexel expected a “significant decline” in first- year student enrollment from the ‘Dear Colleague’ letter in June 2015. Arnett Decl. ¶ 18 (JA261). Most generously, those statements can be harmonized by reading them as reflecting emerging knowledge about Drexel’s plans for future enrollment: the witness had knowledge in March 2015 that enrollment would not increase and learned later in June 2015 that enrollment would decline significantly. That construction of these statements, coupled with the lenient standard of appellate review, leads to the conclusion that the District Court did not abuse its discretion in refusing to apply the sham affidavit rule. For greater repose, SDM’s fraud claim would survive summary judgment even if the sham affidavit rule did exclude the challenged statements in these three declarations. The sham affidavit rule does not permit a court to reject a witness’s prior deposition testimony. And here, one of SDM’s declarants testified at his earlier deposition that Drexel held open the possibility that first-year student enrollment “might go up.” Sherman Dep. 225:23–226:1 (JA200–01). But if Drexel were following a plan to dramatically reduce first-year enrollment, a reasonable jury could find that it was a misrepresentation for Drexel to state that student enrollment could still increase. Most significantly, SDM’s evidence of a misrepresentation or concealment does not come solely from these three witnesses’ testimony or declarations. As explained above, statements by Drexel employees also suggest that Drexel deceived SDM about its plans for future student enrollment.4 4 This decision does not preclude a jury from considering these three witnesses’ declarations in assessing their credibility – if 33 c. Drexel’s Remaining Counterarguments for Upholding Summary Judgment in Its Favor Also Fail. Drexel attacks SDM’s prima facie case of fraudulent inducement on three additional grounds. None of those have merit. Drexel argues that SDM’s fraud claim fails as a matter of law. According to Drexel, its student-enrollment projections were forward-looking statements, and an entity cannot be liable in fraud for incorrectly predicting the future. In general, that is correct: due to the known unreliability of predicting the future, forward-looking statements typically cannot constitute misrepresentations for purposes of a fraud claim. See Coll. Watercolor, 360 A.2d at 206. But that principle does not apply when an entity misrepresents its true future projections. And that is the premise of SDM’s fraud claim. It does not sue because Drexel’s projections did not come true but rather because it believes that Drexel provided false projections. Such conduct qualifies as “a misrepresentation of an existing fact” that could support a fraud claim. Id. (citing Rose v. Rose, 123 A.2d 693, 697 (Pa. 1956)); see also Nat’l Data Payment Sys., Inc. v. Meridian Bank, 212 F.3d 849, 858 (3d Cir. 2000) (citing College Watercolor, 360 A.2d at 206, for the proposition that, under Pennsylvania law, statements about plans or projections may be fraudulent if they “knowingly misstate[]the speaker’s true state of mind when made”). otherwise permissible under the Federal Rules of Evidence. See, e.g., Fed. R. Evid. 613, 801(d)(1)(A). 34 Because false future projections qualify as misrepresentations, SDM’s fraud claim is not barred as a matter of law. Drexel also submits that it did not commit fraud because it had no duty to inform SDM that it would miss its future enrollment targets. That is incorrect. Drexel had a duty not to misrepresent its student enrollment projections, and even if it provided accurate initial projections, when those facts materially changed, it had a duty to not intentionally conceal that development. See Neuman, 51 A.2d at 764 (“The deliberate nondisclosure of a material fact amounts to culpable misrepresentation no less than does an intentional affirmation of a material falsity.” (citing Restatement (Second) of Torts § 551)); see also Restatement (Second) of Torts § 551(2)(c) cmt. f (explaining that a person who makes a representation initially believed to have been true but “remains silent after he has learned that it is untrue and that the person to whom it is made is relying upon it in a transaction with him is morally and legally in the same position as if he knew that his statement was false when made”). Drexel’s final counterargument has undertones of concession. Drexel asserts that SDM cannot sue for fraud, despite any misrepresentations or concealments that Drexel may have made, because SDM learned the true enrollment numbers shortly after the Management Agreement was finalized and SDM should have terminated the contract for convenience rather than perform. But Pennsylvania law is not so unkind to the defrauded. Instead, it allows for an election of remedy: a defrauded party may affirm the contract and sue for tort damages or pursue contract damages. See Tilghman v. Dollenberg, 213 A.2d 324, 327 (Pa. 1965) (“The affirmance of a contract induced by fraud of the seller does not extinguish the 35 right of the purchaser, and it is not a waiver of the fraud, nor does it bar the right of the purchaser to recover damages for the fraud.”); see also Allied Erecting & Dismantling, Co. v. USX Corp., 249 F.3d 191, 199 (3d Cir. 2001) (“[A] party can affirm a contract over a period of time without waiving a claim to fraudulent inducement.”). Thus, SDM was not required to terminate the Management Agreement once it suspected fraud; it was allowed to perform and seek remedies in tort, which it now does. For these reasons, Drexel’s counterarguments lack merit, and the evidence produced by SDM would allow a reasonable jury the option of concluding by clear and convincing evidence that Drexel misrepresented or concealed its own projections for student enrollment. See Moser, 589 A.2d at 682 (“[A] party alleging fraud has the burden of proving the same by clear and convincing evidence.”). But without additional evidence of outrageous conduct by Drexel, SDM cannot proceed with its claim for punitive damages. See Phillips, 883 A.2d at 445 (explaining that punitive damages claims require evidence of outrageous conduct); Pittsburgh Live, 615 A.2d at 442 (explaining that fraud plaintiff cannot rely solely on evidence supporting underlying fraud claim). 3. The Parol Evidence Rule Does Not Bar SodexoMAGIC’s Claim for Fraudulent Inducement. In another effort to defeat SDM’s fraud claim, Drexel invokes Pennsylvania’s parol evidence rule. Drexel argues that because the Management Agreement contains an integration clause, the parol evidence rule prevents the use of extrinsic evidence to establish the alleged misrepresentations at the heart of SDM’s fraud claim. For the reasons below, Drexel’s 36 argument overextends Pennsylvania’s parol evidence rule, which does not preclude SDM’s fraud claim. a. Integration Clauses, the Parol Evidence Rule, and Fraudulent Inducement Claims Under Penn- sylvania Law. The parol evidence rule attaches legal consequences to an integrated contract – a contract that is the only agreement between the parties on a specific topic. See Restatement (Second) Contracts § 209(1) (“An integrated agreement is a writing or writings constituting a final expression of one or more terms of an agreement.”). To demonstrate integration, parties commonly include an integration clause in a contract. Such a clause typically states that the contract contains the final expression of all the terms of the parties’ agreement and that it supersedes all prior agreements on the subject matter. For integrated contracts, the parol evidence rule prevents the use of extrinsic evidence to add to or modify the contract’s terms: Once a writing is determined to be the parties’ entire contract, the parol evidence rule applies and evidence of any previous oral or written negotiations or agreements involving the same subject matter as the contract is almost always inadmissible to explain or vary the terms of the contract. Yocca v. Pittsburgh Steelers Sports, Inc., 854 A.2d 425, 436– 37 (Pa. 2004) (emphasis added).5 5 The parol evidence rule is not absolute; Pennsylvania recognizes four exceptions. Extrinsic evidence may be used to 37 Under that formulation, the parol evidence rule is a substantive rule of contract law that prevents the use of extrinsic evidence to nullify, modify, or augment the terms of a contract. See id.; see also Rose v. Food Fair Stores, Inc., 262 A.2d 851, 854 (Pa. 1970) (holding that the parol evidence rule precludes proof of oral representations that would “directly contradict the clear meaning of the written contract”); Nicolella v. Palmer, 248 A.2d 20, 22–23 (Pa. 1968) (holding that for an integrated contract, the parol evidence rule prevents extrinsic evidence to establish an additional contract term). Importantly, the rule does not prevent the use of extrinsic evidence for other purposes. See, e.g., Berger v. Pittsburgh Auto Equip. Co., 127 A.2d 334, 335 (Pa. 1956) (allowing evidence of a misrepresentation “not to alter or vary” the terms of an integrated contract, but to rescind it). And for fraudulent inducement claims, the purpose of extrinsic evidence is to prove a precontractual misrepresentation or concealment – not to alter or vary the terms of the contract. See Blumenstock v. Gibson, 811 A.2d 1029, 1036 (Pa. Super. Ct. 2002) (explaining that, for fraudulent inducement claims, the party seeks to offer modify contract terms when an integrated contract is ambiguous or to add contract terms omitted by fraud in the execution, by accident, or by mistake. See Yocca, 854 A.2d at 437 (stating that parol evidence is admissible to explain or clarify an ambiguous term in a contract “whether the ambiguity is created by the language of the instrument or by extrinsic or collateral circumstances” (quoting In re Est. of Herr, 161 A.2d 32, 34 (Pa. 1960))); Nicolella v. Palmer, 248 A.2d 20, 22 (Pa. 1968) (stating that extrinsic evidence to vary terms of an integrated agreement is barred “in the absence of fraud, accident, or mistake”). None of those are at issue in this case. 38 extrinsic evidence to show not “that the representations were omitted from the written agreement, but, rather, . . . that the representations were fraudulently made”); 1726 Cherry St. P’Ship v. Bell Atl. Props., Inc., 653 A.2d 663, 666 (Pa. Super. Ct. 1995) (same). Thus, the parol evidence rule acting alone does not prevent fraudulent inducement claims arising out of integrated contracts. See generally Sweet, Promissory Fraud, supra, at 877 (“It is generally agreed that the parol evidence rule does not prevent admission of evidence of fraud.”). A contract may, however, contain fraud-insulating clauses. Those provisions take many forms, but they often have a common objective: to prevent a party from satisfying the justifiable-reliance element of a fraudulent inducement claim. One such fraud-insulating term is a no-reliance clause, through which a party expressly disclaims reliance on another party’s precontractual representations.6 Other similar fraud-insulating provisions are clauses that assume joint responsibility for 6 See Glenn D. West & Kim M. Shah, Debunking the Myth of the Sandbagging Buyer: When Sellers Ask Buyers to Agree to Anti-Sandbagging Clauses, Who Is Sandbagging Whom?, 11 The M&A Lawyer, no. 1, Jan. 2007, at 4 (explaining that no-reliance clauses “relieve the seller of extra-contractual tort claims” based on precontractual representations because “the existence of such a clause makes the buyer’s claim of reliance on such statements to its detriment unjustified and unreasonable” (citation omitted)); see also Allen Blair, A Matter of Trust: Should No-Reliance Clauses Bar Claims for Fraudulent Inducement of Contract?, 92 Marq. L. Rev. 423, 468–75 (2009) (explaining that sophisticated commercial parties bargain for no-reliance clauses in their agreements for many reasons). 39 precontractual representations, see, e.g., Bardwell v. Willis Co., 100 A.2d 102, 103–04 (Pa. 1953), or that state that the representations in the contract either supersede all prior representations, see, e.g., Yocca, 854 A.2d at 431, or are the only representations made, see, e.g., 1726 Cherry St., 653 A.2d at 670. When an integrated contract includes a fraud-insulating term – to form what may be called an ‘integration-plus’ contract – that extends the reach of the parol evidence rule. In that circumstance, the parol evidence rule prevents the use of extrinsic evidence to vary the fraud-insulating term. And without such evidence, it is virtually impossible to establish the justifiable-reliance element needed for a fraud claim. As the Pennsylvania Supreme Court has explained for integrated contracts, “due to the parol evidence rule’s operation, a party cannot be said to have justifiably relied on prior representations that he has superseded and disclaimed.” Toy, 928 A.2d at 207 (emphasis added). For example, consider an integrated contract between sophisticated parties with a no-reliance clause. There, the parol evidence rule prevents the use of extrinsic evidence to vary the terms of the contract, including the no-reliance clause. Left undisturbed by extrinsic evidence, a no-reliance clause, through which a party disclaims reliance on any prior representations, makes it legally impossible for a party to establish that it justifiably relied on a precontractual representation. See Glenn D. West & W. Benton Lewis, Jr., Contracting to Avoid Extra-Contractual Liability – Can Your Contractual Deal Ever Really Be the “Entire” Deal?, 64 Bus. Law. 999, 1018 (2009) (explaining that no-reliance clauses “purport to preclude proof of the mandatory reliance element 40 of extra-contractual misrepresentation actions” (internal quotation marks omitted)); Joseph Wylie, Using No-Reliance Clauses to Prevent Fraud-in-the-Inducement Claims, 92 Ill. Bar J. 536, 539 (2004) (“[A] party agreeing to [a no-reliance] clause in essence agrees in advance to waive any causes of action it may have for fraud.”). Thus, through the operation of the parol evidence rule, a no-reliance clause in an integrated contract precludes fraudulent inducement claims that depend on a precontractual misrepresentation. Pennsylvania caselaw abounds with similar examples. A seminal application of this principle occurred in Bardwell v. Willis Co., when the Pennsylvania Supreme Court rejected a tort claim for deceit arising out of an integrated contract with joint diligence and joint representation terms. 100 A.2d at 104–05. There, the Court reasoned that the parol evidence rule prevented extrinsic evidence from nullifying those terms, and with those provisions unaltered, the complaining party had no redress in tort. See id. The Pennsylvania Supreme Court reached a similar conclusion for an integrated contract that contained a clause stating that the contract superseded “any representations or agreements previously made or entered into by the parties hereto.” Yocca, 854 A.2d at 431 (emphasis added). Because the parol evidence rule prevented modification of that term through extrinsic evidence, the complaining party could not demonstrate justifiable reliance on prior disclaimed representations. See id. at 439. Pennsylvania courts treat integrated contracts with no-additional- representations clauses similarly: the parol evidence rule prevents fraudulent inducement claims based on precontractual misrepresentations. See 1726 Cherry St., 653 A.2d at 670 (holding that the parol evidence rule prevented the use of evidence of a prior representation when the integrated contract 41 stated that there were “no other representations or understandings” between the parties); see also Blumenstock, 811 A.2d at 1036 (“[T]he case law clearly holds that a party cannot justifiably rely upon prior oral representations yet sign a contract denying the existence of those representations.”); McGuire v. Schneider, Inc., 534 A.2d 115, 119 (Pa. Super. Ct. 1987) (same). In sum, Pennsylvania’s parol evidence rule does not prevent fraudulent inducement claims for all integrated contracts, but the rule may preclude such claims based on misrepresentations for ‘integration-plus’ contracts – integrated contracts with fraud-insulating provisions. See Youndt, 868 A.2d at 546 (explaining that “parol evidence is inadmissible where the contract contains terms that deny the existence of representations regarding the subject matter of the alleged fraud,” but “when the contract contains no such term denying the existence of such representations, parol evidence is admissible to show fraud in the inducement”). b. The Management Agreement Lacks Fraud-Insu- lating Provisions, so the Parol Evidence Rule Does Not Preclude SodexoMAGIC’s Fraudulent Inducement Claim. The parol evidence rule does not apply to SDM’s fraud claim because, although the Management Agreement contains an integration clause, it lacks fraud-insulating provisions. The Management Agreement’s integration clause references prior agreements; it does not mention, much less disclaim, prior representations: 42 This Agreement contains all agreements of the parties with respect to matters covered herein, superseding any prior agreements, and may not be changed other than by an agreement in writing signed by the parties hereto. Management Agreement § 10.11 (emphasis added) (SA66). And lest any uncertainty remain, the Management Agreement also includes an express reliance clause that states explicitly that the parties relied on certain categories of prior assumptions and representations: The financial terms set forth in this Agreement and other obligations assumed by SodexoMAGIC hereunder are based on conditions in existence on the date SodexoMAGIC commences operations, including by way of example [Drexel’s] student population; labor, food and supply costs; and federal, state and local sales, use and excise tax. In addition, each party has relied on representations regarding existing and future conditions and projections made by the other in connection with the negotiation and execution of this Agreement. Id. § 9.1 (emphases added) (SA57). One category of those recognized representations – projections about future student population – is the basis for SDM’s fraud claim. Under these circumstances, the parol evidence rule does not preclude SDM’s fraudulent inducement claim. SDM seeks to introduce extrinsic evidence of Drexel’s precontractual 43 misrepresentations of its student enrollment projections. But it does not proffer that evidence for the purpose prohibited by the parol evidence rule: to nullify, vary, or supplement a contractual term. See Yocca, 854 A.2d at 436–37. Rather, SDM accepts the terms of the Management Agreement as they are – in particular, those related to its obligations to make investments and provide dining services – and it seeks to use the extrinsic evidence to prove that Drexel fraudulently induced it to enter into the Management Agreement. Nor does the Management Agreement contain a fraud-insulating provision. It does not disclaim reliance on precontractual representations, and it does not state that representations in the Management Agreement are exclusive or supersede all prior representations. Thus, here, the parol evidence rule does not prevent the use of extrinsic evidence to prove precontractual misrepresentations. See Youndt, 868 A.2d at 546.7 4. Pennsylvania’s Gist of the Action Doctrine Does Not Bar SodexoMAGIC’s Fraud Claim. The District Court also barred SDM’s fraud claim under Pennsylvania’s gist of the action doctrine. It did so on the theory that Drexel owed SDM no duty apart from those imposed by the Management Agreement and thus SDM’s fraud claim was prohibitively duplicative of its breach-of-contract claim. See SodexoMAGIC, 333 F. Supp. 3d at 454–56. That is mistaken. 7 Because the Management Agreement does not contain any fraud-insulating provisions regarding intentional concealments of material facts, the parol evidence rule likewise does not bar SDM from pursuing a fraudulent inducement claim based on concealment. 44 Under Pennsylvania law, the gist of the action doctrine prevents a purely contractual duty from serving as the basis for a tort claim. See Bruno v. Erie Ins. Co., 106 A.3d 48, 65 (Pa. 2014). Tort actions arise from the breach of a duty owed to another as a matter of social policy, while breach-of-contract actions arise from the breach of a duty created by contract. See id. at 68 (explaining that tort claims involve a “violation of a broader social duty owed to all individuals” and thus exist “regardless of the contract”); eToll, Inc. v. Elias/Savion Advert., Inc., 811 A.2d 10, 14 (Pa. Super. Ct. 2002); Bohler- Uddeholm Am., Inc. v. Ellwood Grp., Inc., 247 F.3d 79, 103– 04 (3d Cir. 2001). When a duty is created by contract, the gist of the action doctrine requires that a claim for a breach of that duty be brought in contract, not tort. See Bruno, 106 A.3d at 68 (“[If] the duty breached is one created by the parties by the terms of their contract . . . then the claim is to be viewed as one for breach of contract.”). The doctrine has a long lineage in Pennsylvania, having its roots in precedents from the first half of the nineteenth century. See generally id. at 60–64 (tracing this principle from its origins to the present). Examples of its application include the case of a bailor who was alleged to have negligently misplaced a bailment, but who under the doctrine could not be sued in tort. See McCahan v. Hirst, 7 Watts 175, 179 (Pa. 1838). As the Pennsylvania Supreme Court more recently explained, without the bailment agreement, the bailor had no duty to properly handle the property, and the gist of the action doctrine barred the tort claim. Bruno, 106 A.3d at 63 (citing McCahan, 7 Watts at 179). Similarly, a party who contractually promised to safely keep and pasture horses was not liable in tort for negligently keeping and pasturing the horses because, absent 45 the contract, that party had no duty to keep and pasture those horses. See id. at 64 (citing Cook v. Haggarty, 36 Pa. 67, 69 (Pa. 1859)). In sum, under the gist of the action doctrine, “the nature of the duty alleged to have been breached . . . [is] the critical determinative factor,” with the doctrine permitting tort claims that exist “regardless of the contract.” Id. at 68. Under the doctrine, it is still possible for the same act to breach both a duty under tort law and a contractual duty. One such example occurred when a party contractually promised to maintain a fence but then removed a portion of the fence near a quarry and a horse owned by the other party to the contract tumbled to its death in the unfenced quarry. See id. at 64 (citing Krum v. Anthony, 8 A. 598 (Pa. 1887)). Because the party who removed a portion of the fence near the quarry would be liable for the horse’s death regardless of the contractual duty to maintain the fence, the gist of the action doctrine did not bar a tort claim. See id. As another example, a landlord who promised to repair a defective porch could be subject to tort claims after the porch had collapsed on the tenant because the landlord had a duty to invitees even absent the promise to repair the porch. See id. at 65–66 (citing Reitmeyer v. Sprecher, 243 A.2d 395 (Pa. 1968)). The gist of the action doctrine does not apply here because SDM’s fraudulent inducement claim does not depend on the breach of a contractual duty. SDM alleges that Drexel misrepresented and intentionally concealed its internal student enrollment projections while the parties were negotiating the Management Agreement. At that time, however, the parties had not executed the Management Agreement. And without a binding contract, any duty Drexel owed SDM during negotiations was grounded only in tort. 46 In reaching a contrary conclusion, the District Court attributed Drexel’s duty to the terms of the Management Agreement. See SodexoMAGIC, 333 F. Supp. 3d at 455–56. But a precontractual duty not to deceive through misrepresentation or concealment exists independently of a later-created contract. See Moser, 589 A.2d at 682 (explaining that parties are under a duty to avoid “intentional false statement[s]” or “concealment[s] of material fact[s]” (citing Commonwealth v. Monumental Props., Inc., 329 A.2d 812, 829 (Pa. 1974))). And even if such a duty could be retroactively incorporated into a contract, that would still not foreclose a tort action. When a contractual duty duplicates an obligation generally owed to another in society, the gist of the action doctrine does not bar a tort claim; it prevents only the contractual duty from serving as a basis for a tort claim. See Bruno, 106 A.3d at 62–65; see also, e.g., Reitmeyer, 243 A.2d at 398; Krum, 8 A. at 600. Because SDM’s tort-based fraud claim would exist with or without a later-in-time contract, the gist of the action doctrine does not bar such a claim here. 5. Drexel’s Fraudulent-Inducement Counterclaim Fails. In counterclaiming for fraudulent inducement, Drexel asserts that SDM had no intention of making the $4 million capital contribution that it promised in its BAFO. According to Drexel, SDM’s misrepresentation caused Drexel to award the contract to SDM, not Aramark. Typically, a fraudulent inducement claim involves deception that leads a person to enter a contract. See Maguire v. Wheeler, 150 A. 882, 884 (Pa. 1930) (“Where the person 47 making the promise under such circumstances intended at the time not to perform it, thus fraudulently making use of the promise as a device to procure the contract or deed, equity will grant relief . . . .”). Yet Drexel makes no such allegation here. Instead, it claims that SDM’s misrepresentation induced it to forgo contracting with Aramark. While a nontraditional application, the Pennsylvania Supreme Court would likely recognize such a fraudulent inducement cause of action because a promise without an intention to perform may form the basis of a fraud claim. See Rose, 123 A.2d at 697 (explaining that promises made without the intention to carry out the promise support an action for fraud); Sweet, Promissory Fraud, supra, at 888 (“The general rule is that a promise made without the intent to perform it is fraud.”). But this is an unusual context to advance such a theory. The evidence here suggests that Drexel misrepresented its predictions for future student enrollment not just to SDM but to all bidders, including Aramark. From that perspective, Drexel is claiming that it lost the opportunity to contract with Aramark on the basis of misrepresentations that it made to Aramark. Under these circumstances, it is doubtful that Pennsylvania would permit a fraudulent-inducement claim. But it is unnecessary to reach that novel question of state law here. That is so because Drexel’s fraud claim is untimely. Under the relevant statute of limitations, a party has two years to sue for fraud. See 42 Pa. Stat. and Cons. Stat. § 5524(7). That period commences when a cause of action accrues. See Rice v. Diocese of Altoona-Johnstown, 255 A.3d 237, 246 (Pa. 2021). Working backwards, Drexel first filed its fraud counterclaim on July 13, 2017. Thus, to be timely, its claim must have 48 accrued on or after July 13, 2015 – a little less than two months after the parties finalized the Management Agreement. Yet Drexel’s fraud claim accrued well before that date. See Gleason v. Borough of Moosic, 15 A.3d 479, 484 (Pa. 2011) (explaining that a cause of action accrues “when an injury is inflicted and the corresponding right to institute a suit for damages arises”). SDM proposed its BAFO on August 8, 2014, and Drexel awarded the contract to SDM a week later, on August 15, 2014. As of that date, Drexel relied on SDM’s alleged false promise and discontinued competitive negotiations with Aramark and could have sued SDM for fraud. And because Drexel did not initiate its counterclaim for over two years – not until July 13, 2017 – Pennsylvania’s statute of limitations, through its normal operation, would bar Drexel’s fraud claim. Drexel contends that the statute of limitations should not operate normally in this case. Instead, according to Drexel, the statute of limitations should be tolled by either the discovery rule or the fraudulent-concealment exception. Drexel argues first that the discovery rule should toll the statute of limitations until April 28, 2017. That is the date on which Drexel received discovery materials from SDM suggesting that SDM never intended to fulfill the commitments it made in its final offer. See In re Risperdall Litig., 223 A.3d 633, 640 (Pa. 2019) (explaining that the discovery rule tolls the statute of limitations for the period that “an injury or its cause [are] not reasonably knowable”). It is undisputed that Drexel had constructive knowledge of SDM’s allegedly false intentions by that date. And such knowledge stops tolling under the discovery rule. See Gleason, 15 A.3d at 484. 49 But discovery-rule tolling ceases once a party has “actual or constructive knowledge of at least some form of significant harm and of a factual cause linked to another’s conduct, without the necessity of notice of the full extent of the injury . . . or precise cause.” Id. (quoting Wilson v. El-Daief, 964 A.2d 354, 364 (Pa. 2009)); see also Fine v. Checcio, 870 A.2d 850, 858 (Pa. 2005) (explaining that the discovery rule still requires reasonable diligence, measured under an objective standard of whether a plaintiff used “the means of information within [its] reach, with the vigilance the law requires” (citation omitted)). And here, Drexel had constructive knowledge of a potential material misrepresentation at an earlier point in time, when SDM circulated an initial draft of the Management Agreement on September 12, 2014, without the $4 million capital contribution. As alleged by Drexel, that term was “one of the principal financial advantages distinguishing Sodexo’s BAFO from Aramark’s,” Am. Countercl. ¶ 54 (SA13), and Drexel’s lead negotiator suspected that SDM was pulling a bait and switch. Thus, as of September 12, 2014, the discovery rule no longer tolled the statute of limitations. Drexel still had two years to investigate and sue, but it did not file its counterclaim within that period. The related but distinct doctrine of fraudulent-concealment tolling does not salvage Drexel’s counterclaim. That doctrine allows tolling of the statute of limitations for the period in which the opposing party, through fraud or concealment, causes another party to “relax [its] vigilance or deviate from [its] right of inquiry into the facts.” Fine, 870 A.2d at 860. But here, although Drexel alleges that SDM’s BAFO misrepresented its intention to make an additional $4 million 50 capital contribution, that alleged misrepresentation was not concealed after SDM circulated its first draft of the Management Agreement. At that point, because the draft did not include that previously promised term, SDM was no longer concealing its intention to not make that additional investment. Thus, fraudulent-concealment tolling ceased on September 12, 2014, as well. And because Drexel did not file its fraud counterclaim until over two years after that date, the statute of limitations bars it. B. SodexoMAGIC’s Breach-of-Contract Claim for Failure to Renegotiate in Good Faith Survives Summary Judgment. SDM also sued Drexel for breaching the Management Agreement for failing to renegotiate in good faith after annual first-year student enrollment did not increase by two percent. The District Court granted summary judgment to Drexel on this count, reasoning primarily that promises to renegotiate in good faith are too indefinite to be enforced. See SodexoMAGIC, 333 F. Supp. 3d at 457–59. As an alternative ground, the District Court found that Drexel did renegotiate in good faith. See id. at 459 n.6. Both conclusions are incorrect: the first as a matter of law; the second due to a genuine dispute of material fact. 1. A Promise to Renegotiate in Good Faith May Be Enforceable Under Pennsylvania Law, and the Promise Between These Parties Is Enforceable. Pennsylvania affords parties broad latitude in fashioning their agreements. See Cent. Dauphin Sch. Dist. v. Am. Cas. Co., 426 A.2d 94, 96 (Pa. 1981); see also Restatement 51 (Second) of Contracts, ch. 8, intro. note (“In general, parties may contract as they wish, and courts will enforce their agreements without passing on their substance.”). As part of that flexibility, a voluntarily-agreed-to contract term is enforceable unless a statute or the common law specifically prevents enforcement of that term. See Greene v. Oliver Realty, Inc., 526 A.2d 1192, 1194 (Pa. Super. Ct. 1987) (“Contemporary contract law generally provides that a contract is enforceable when the parties reach mutual agreement, exchange consideration and have outlined the terms of their bargain with sufficient clarity.”); Stephan v. Waldron Elec. Heating & Cooling LLC, 100 A.3d 660, 665 (Pa. Super. Ct. 2014) (same). Statutorily prohibited contracts include oral contracts subject to the statute of frauds8 and contracts for wages below the minimum wage.9 The common law prevents enforcement of other types of promises, including promises 8 See 13 Pa. Stat. and Cons. Stat. § 2201(a); 33 Pa. Stat. and Cons. Stat. § 1. 9 See 43 Pa. Stat. and Cons. Stat. § 333.104; Chevalier v. Gen. Nutrition Ctrs., Inc., 220 A.3d 1038, 1055 (Pa. 2019) (recognizing the Pennsylvania General Assembly’s declaration of policy that, in the fair wage context, “‘freedom of contract’ as applied to [employees’] relations with their employers is illusory” (quoting 43 Pa. Stat. and Cons. Stat. § 333.101) (alteration in original)). 52 that are illegal;10 that otherwise violate public policy;11 or that are indeterminate, indefinite, or vague.12 10 Dippel v. Brunozzi, 74 A.2d 112, 114 (Pa. 1950) (“[A]n agreement which violates a provision of a statute, or which cannot be performed without violation of such a provision, is illegal and void.”); see Am. Ass’n of Meat Processors v. Cas. Reciprocal Exch., 588 A.2d 491, 495 (Pa. 1991) (“What we consider controlling, however, . . . is that the alleged contract is illegal under a statute enacted in aid of significant public policies identified by the Pennsylvania legislature.”); Rounick v. Neducsin, 231 A.3d 994, 1000 n.6 (Pa. Super. Ct. 2020) (explaining that “the courts of this Commonwealth will not be used to enforce contracts that are illegal pursuant to a statute”). 11 Pittsburgh Logistics Sys., Inc. v. Beemac Trucking, LLC, 249 A.3d 918, 930 (Pa. 2021) (“Generally, a clear and unambiguous contract provision must be given its plain meaning unless to do so would be contrary to a clearly expressed public policy.” (quoting Eichelman v. Nationwide Ins. Co., 711 A.2d 1006, 1008 (Pa. 1998))); Tayar v. Camelback Ski Corp., 47 A.3d 1190, 1199 (Pa. 2012); Ferguson v. McKiernan, 940 A.2d 1236, 1244 (Pa. 2007). 12 See Devlin v. City of Phila., 862 A.2d 1234, 1244 n.9 (Pa. 2004) (“[T]o create an enforceable contract, parties must delineate the terms of their bargain with sufficient clarity.” (citation and internal quotation marks omitted)); id. (“[A]n agreement cannot be enforced if its terms are indefinite.” (citation and internal quotation marks omitted)); Seiss v. McClintic-Marshall Corp., 188 A. 109, 110 (Pa. 1936) (“The contract here set up is so lacking in precision, so indefinite and vague, that nothing certain about it can be formulated.”); Edgcomb v. Clough, 118 A. 610, 614 (Pa. 1922) (“In order that 53 Those categories do not specifically include promises to renegotiate in good faith. Nonetheless, such a promise could still be unenforceable if it shares attributes with a prohibited category. The only prohibited category potentially implicated by a promise to renegotiate in good faith is the rule against indeterminate promises. But as a general matter, promises to renegotiate are not too indefinite to be enforceable. Indeed, the Pennsylvania Supreme Court long ago relied on a right-to- renegotiate-in-good-faith clause as a basis for its decision. See Cosgrove v. Kappel, 168 A.2d 319, 320 (Pa. 1961). More recently, three Justices of the Pennsylvania Supreme Court favored resolving a dispute through the enforceability of a similar agreement to negotiate in good faith. See Dep’t of Gen. Servs. v. On-Point Tech. Sys., 870 A.2d 873, 874 (Pa. 2005) (Saylor, J., joined by Nigro and Baer, JJ., concurring in part and dissenting in part) (“I would also specifically confirm the Third Circuit’s prediction that this Court would cognize a contract-based cause of action for breach of an agreement to negotiate in good faith.”). And several intermediate Pennsylvania courts, along with this Court, have concluded a contract may be enforceable, its terms must be certain and explicit and not vague or indefinite.”); Ingrassia Constr. Co. v. Walsh, 486 A.2d 478, 484 (Pa. Super. Ct. 1984) (“A court cannot enforce a contract unless it can determine what it is.” (quoting 1 A. Corbin, Corbin on Contracts § 95 (1963))); Reed v. Pittsburgh Bd. of Pub. Educ., 862 A.2d 131, 135 (Pa. Commw. Ct. 2004) (“If a court, due to indefiniteness or incompleteness, is unable to determine if a contract was performed, the court must find no contract existed in the first place.”). 54 that the Pennsylvania Supreme Court would recognize a contract-based cause of action for breach of a promise to negotiate in good faith. See GMH Assocs. Inc. v. Prudential Realty Grp., 752 A.2d 889, 903–04 (Pa. Super. Ct. 2000); Jenkins v. Cnty. of Schuylkill, 658 A.2d 380, 385 (Pa. Super. Ct. 1995); see also Flight Sys., Inc. v. Elec. Data Sys. Corp., 112 F.3d 124, 130 (3d Cir. 1997); Channel Home Ctrs. v. Grossman, 795 F.2d 291, 299 (3d Cir. 1986). Thus, under Pennsylvania law, a promise to renegotiate in good faith is likely not – as a category – too indeterminate to be enforceable. See Flight Sys., 112 F.3d at 130; Channel Home Ctrs., 795 F.2d at 299. Still, an individual promise to renegotiate may be too vague or indefinite to be enforceable. Courts decline to enforce promises to renegotiate that lack a framework for evaluating the parties’ good-faith obligations. See, e.g., Jenkins, 658 A.2d at 385 (rejecting a claim for breach of a duty to negotiate in good faith where the language of the letter did “not reveal that the parties intended to be bound by any terms of the original specifications” and “no specific terms were even agreed upon”).13 But here, two attributes of the Management 13 See also A/S Apothekernes Laboratorium for Specialpraeparater v. I.M.C. Chem. Grp., Inc., 873 F.2d 155, 158–59 (7th Cir. 1989); (rejecting a similar claim where “the letter [of intent] did not set forth any previously agreed upon terms much less provide a general framework within which the parties intended to conduct their negotiations”); Tchrs. Ins. & Annuity Ass’n of Am. v. Tribune Co., 670 F. Supp. 491, 498 (S.D.N.Y. 1987) (explaining that agreements to negotiate open terms in good faith commit parties “to the obligation to 55 Agreement’s promise to renegotiate in good faith give it the structure needed to be enforceable: the trigger for the obligation to renegotiate and the parameters for the proposed renegotiation. First, the Management Agreement has a sufficiently definite trigger for renegotiation obligations. Drexel and SDM agreed to “renegotiate[] on a mutually agreeable basis” upon certain changes to “representations regarding existing and future conditions” that they made to each other in negotiating the Management Agreement: [E]ach party has relied on representations regarding existing and future conditions and projections made by the other in connection with the negotiation and execution of this Agreement. In the event of a change in the conditions or the inaccuracy or breach of, or the failure to fulfill, any such representation by a party, the financial terms and other obligations assumed by the other party shall be renegotiated on a mutually agreeable basis to reflect such change, inaccuracy or breach. Management Agreement § 9.1 (SA57) (emphases added). Those representations include the statement that Drexel’s growth was a critical factor in determining certain contractually defined “Investments” that SDM promised.14 negotiate the open issues in good faith in an attempt to reach the alternate objective within the agreed framework”). 14 The Investments consist of specific classes of capital contributions. See id. § 8.5 (SA49–52). Most of those 56 Management Agreement § 8.5 (SA49). They also include SDM’s projection that future first-year student enrollment would increase annually by two percent: Parties agree that the University’s growth is a critical factor in calculating the Investments afforded under this Agreement and it has been projected by SodexoMAGIC that this growth will realize an increase of 2% per year in the freshman class year over year. Id. § 9.2 (SA58) (emphasis added). And recognizing that changes to the assumed future student population could have an adverse economic effect on SDM, the parties agreed to work “in good faith to mutually agree upon solutions in an effort to counter such impact”: contributions occurred in the past, such as when SDM or one of its affiliates made improvements to Drexel’s dining facilities. See id. § 8.5(A)–(B) (SA49–50). For those past contributions, the Management Agreement provided for a straight-line amortization schedule for Drexel to reimburse SDM over time, typically a ten-year period. See id. One class of Investments related to construction of an urban eatery. See id. § 8.5(C) (SA50). Another class of Investments was for promised enhancements related to a food truck, a dining center, and the urban eatery. See id. § 8.5(D) (SA51–52). The parties budgeted those future Investments at $10.7 million, an amount that Drexel would repay through a straight-line amortization schedule beginning on the date the funds are paid out and continuing through June 30, 2025. See id. 57 [Drexel] recognizes that Sodexo MAGIC [sic] made certain assumptions in preparing the financial package offered in this Agreement and understands that changes to the financial assumptions below may have an adverse economic impact on SodexoMAGIC; in such cases [Drexel] shall work with SodexoMAGIC in good faith to mutually agree upon solutions in an effort to counter such impact. SodexoMAGIC acknowledges its responsibility to respond quickly and expertly to factors under their control that may affect these outcomes. Id. § 9.2 (emphasis added). Together, these terms identify a trigger to renegotiate as an annual future first-year student enrollment increase of less than two percent. In light of these terms, no one disputes that the duty to renegotiate in good faith was in fact triggered. Second, the Management Agreement defines the scope of renegotiations. Future student enrollment was expressly linked to the Investments that SDM promised. See id. (stating that student “growth is a critical factor in calculating the Investments afforded under this Agreement”). Through renegotiation of the promises related to the Investments, the parties could agree to reduce future capital contributions for those Investments, see id. § 8.5(C)–(D), accelerate the amortization schedule for those Investments, see id. § 8.5(A)– (B), or provide some other setoff. By limiting the topics subject to renegotiation, the Management Agreement enables evaluation of whether good-faith renegotiations occurred. That, along with the definite trigger for renegotiation, removes 58 the promise to renegotiate in good faith from the realm of indeterminacy. 2. A Genuine Dispute of Material Fact Remains as to Whether Drexel Renegotiated in Good Faith. The District Court alternatively concluded that Drexel did renegotiate in good faith. See SodexoMAGIC, 333 F. Supp. 3d at 459 n.6. Evaluating that issue requires an understanding of the concept of good faith, which varies by context. See Stamerro v. Stamerro, 889 A.2d 1251, 1259 (Pa. Super. Ct. 2005) (explaining that “[t]he obligation to act in good faith in the performance of contractual duties varies somewhat with the context” (quoting Somers v. Somers, 613 A.2d 1211, 1213 (Pa. Super. Ct. 1992))). The Management Agreement did not specifically define the term, and the Pennsylvania Supreme Court has not given meaning to good faith in the context of a promise to renegotiate. In the abstract, however, the duty of good faith is ambiguous: it could require nothing more than that the parties avoid acting in bad faith, or it could impose an affirmative obligation on the parties. The first view is more prevalent in the context of an implied duty of good faith. The Restatement (Second) of Contracts, in discussing an implied duty of good faith, explains that good faith “excludes a variety of types of conduct characterized as involving ‘bad faith’ because they violate community standards of decency, fairness or reasonableness.” Restatement (Second) of Contracts § 205 cmt. a; see also Robert S. Summers, “Good Faith” in General Contract Law and the Sales Provisions of the Uniform Commercial Code, 54 Va. L. Rev. 195, 196 (1968) (arguing that “good faith, as used in the case law, is best understood as an ‘excluder’ – it is 59 a phrase which has no general meaning or meanings of its own, but which serves to exclude many heterogeneous forms of bad faith”). And some Pennsylvania cases have picked up on the notion that good faith means only the absence of bad-faith conduct. See, e.g., Stamerro, 889 A.2d at 1259 (describing good faith through express reference to recognized forms of bad-faith conduct); Herzog v. Herzog, 887 A.2d 313, 317 (Pa. Super. Ct. 2005) (finding a violation of an implied duty of good faith in part because one party’s “conduct [wa]s unfair and unreasonable”); Somers, 613 A.2d at 1213 (explaining that an implied duty of good faith excludes various types of bad faith, including, among other things, “evasion of the spirit of the bargain, lack of diligence and slacking off, [and] willful rendering of imperfect performance”). Alternatively, a duty of good faith could impose an affirmative obligation. In the commentary, the Restatement leaves room for that view by referencing two definitions of good faith from the Uniform Commercial Code in which the duty is affirmative. See Restatement (Second) of Contracts § 205 cmt. a (explaining that good faith is defined as “honesty in fact in the conduct or transaction concerned” in U.C.C. § 1- 201(19) and as “honesty in fact and the observance of reasonable commercial standards of fair dealing in the trade” in U.C.C. § 2-103(1)(b)). Consistent with an interpretation that good faith imposes an affirmative duty in the renegotiation context, Pennsylvania courts have examined the sincerity of the parties’ intention to reach an agreement pursuant to an obligation to negotiate in good faith in the labor contexts and for contracts governed by the U.C.C. See, e.g., Markham v. Wolf, 190 A.3d 1175, 1188 (Pa. 2018) (explaining that, in the labor context, good-faith bargaining entails “evincing an intent to bargain in an attempt to reach an agreement if possible”); 60 Eighth North-Val, Inc. v. William L. Parkinson, D.D.S., P.C., Pension Tr., 773 A.2d 1248, 1254 (Pa. Super. Ct. 2001) (requiring for good faith in contract modifications under U.C.C. § 2-209 “an honest desire to compensate for commercial exigencies”). In those cases, in evaluating good- faith negotiations, Pennsylvania courts have examined the reasonableness of the negotiation process. See, e.g., Markham, 190 A.3d at 1188 (explaining, in the labor context, that bargaining in good faith “includes a wide range of legally enforceable responsibilities, including meeting at certain times, places, with certain frequency”); Eighth North-Val, 773 A.2d at 1254 (“In analyzing [the good-faith requirement for contract modification under U.C.C. § 2-209], the trier of fact must determine whether the means used to obtain the modification constitute extortion or overreaching.”). Thus, an affirmative duty to renegotiate in good faith could be viewed as requiring a sincere intention to reach an agreement and the use of reasonable negotiation tactics. Both formulations – good faith as an excluder and good faith as an affirmative duty – have a core commonality: they each relate to the renegotiation process, not the outcome. For that reason, a promise to renegotiate in good faith does not compel the parties to agree. See Burbach Broad. Co. v. Elkins Radio Corp., 278 F.3d 401, 407 n.2 (4th Cir. 2002) (“Th[e] obligation [to negotiate in good faith] does not guarantee that the final contract will be concluded if both parties comport with their obligation, because good faith differences in the negotiation of the open issues may prevent the parties from reaching a final contract.”). Moreover, such a promise does not prevent one party from out-negotiating the other – that may be done as long as a party avoids bad faith (under the excluder formulation) or acts with a sincere intention to reach an 61 agreement and employs reasonable bargaining tactics (under an affirmative duty formulation). See Creeger Brick & Bldg. Supply, Inc. v. Mid-State Bank & Tr. Co., 560 A.2d 151, 154 (Pa. Super. Ct. 1989) (“[I]t cannot be said that a lender has violated a duty of good faith merely because it has negotiated terms of a loan which are favorable to itself.”). Those general principles alone do not resolve which understanding of the term ‘good faith’ should apply here. On the one hand, the Management Agreement makes express the duty to renegotiate in good faith, and that strains the applicability of the good-faith-as-an-excluder approach, which has been applied most commonly when the duty of good faith is implied. On the other hand, the Management Agreement falls outside the contexts in which Pennsylvania courts have imposed an affirmative duty of good faith: labor disputes and contracts governed by the U.C.C. Ultimately, it is not necessary to select one meaning of good faith over the other because either definition generates the same outcome here. As a baseline, without considering the evidence of fraud, the record evidence indicates that Drexel did renegotiate in good faith with SDM. Under the view that good faith functions only to exclude bad-faith conduct, Drexel met its obligation because the record lacks evidence of bad faith during the renegotiation correspondence. Alternatively, if good faith imposes an affirmative duty, both components of that duty appear to have been satisfied. Drexel negotiated with a sincere intention to reach an agreement: SDM’s own internal documents characterized Drexel as “forthcoming” and recognized that, although the numbers did not match up, the “meeting went extremely well,” with both parties showing determination to reach a deal. See Email from Nancy C. 62 Arnett, Vice President, Sodexo, to Leonard Riccio, Senior Vice President, Sodexo (May 20, 2016) (SA1572–73). Similarly, evidence suggests that Drexel employed reasonable negotiation tactics. At first, SDM proposed reducing its future capital contributions by $12.7 million, for a total future contribution of $2 million. In response, Drexel also proposed reducing SDM’s obligation for future capital contributions, but by less, $9 million. That counterproposal represented a 60% discount off the original promise to invest $14.7 million. The parties went back and forth several times on key terms, and Drexel provided specific reasons for its counteroffer. See Email from Helen Bowman, Executive Vice President, Drexel, to John A. Fry, President, Drexel (July 13, 2016) (SA533–36). The outstanding $3.7 million separating the parties from agreement, while an obstacle, amounted to about 14% of the initially promised future capital contribution. Although the parties could not close that gap, that alone does not prevent a finding of good faith, and here Drexel appears to have used reasonable tactics in negotiating with SDM. In short, looking only at the renegotiation correspondence, Drexel appears to have acted in good faith. But due to SDM’s fraudulent-inducement evidence, the context for evaluating good faith is broader. Under either formulation of a good-faith duty, a party cannot fraudulently induce a promise and also in good faith renegotiate that obligation as if the fraud never occurred. Knowingly gaining the benefit of one’s own prior uncured fraud would likely qualify as bad faith under Pennsylvania law. And renegotiating without correcting and offsetting prior deceit undermines sincerity and is not a reasonable negotiation tactic. Thus, the question of whether Drexel renegotiated in good faith hinges on whether Drexel fraudulently induced SDM to enter 63 the Management Agreement. Because that underlying issue is subject to genuine dispute, so is the question of whether Drexel renegotiated in good faith. In sum, it is inconsequential which formulation of good faith the Pennsylvania Supreme Court would apply to a contractual duty to renegotiate because, under either, SDM’s prima facie evidence of fraudulent inducement carries its good-faith renegotiation claim past summary judgment.15 15 Although both SDM’s fraud claim and SDM’s breach-of- contract claim for failure to renegotiate in good faith survive summary judgment, that does not mean that SDM is entitled as a matter of law to recover on both claims. It may be that SDM will have to elect one of those claims over the other. See Gamesa Energy USA, LLC v. Ten Penn Ctr. Assocs., L.P., 217 A.3d 1227, 1239 (Pa. 2019) (explaining that although “inconsistent remedies may be pleaded and pursued in litigation, damages calculated pursuant to only one theory may be recovered”); see also Schwartz v. Rockey, 932 A.2d 885, 893 n.10 (Pa. 2007) (citing Olympia Hotels Corp. v. Johnson Wax Dev. Corp., 908 F.2d 1363, 1371 (7th Cir. 1990), for its discussion of the timing of the choice between inconsistent claims for fraud and breach of contract). But assessing and applying Pennsylvania election-of-remedies law is left for the District Court in the first instance on remand. See generally Dan B. Dobbs & Caprice L. Roberts, Law of Remedies § 9.3, at 740 (3d ed. 2018) (recognizing that some courts “suggest that the election of remedies doctrine is based on a policy to avoid duplication of relief,” while others justify the doctrine on the ground that a party cannot seek recovery for remedies that are “logically inconsistent”). 64 C. SodexoMAGIC’s Claim for Enhanced Payments for Fall 2016 Survives Summary Judgment. Through an amended and supplemental count, SDM sought to recover for additional alleged breaches of contract. Beyond requesting compensation allegedly due under the Management Agreement, SDM also claimed that Drexel owed enhanced payments under a separate contract governing the Fall 2016 Semester. Finding a lack of consideration and acceptance, the District Court concluded that the parties did not enter a separate contract and rejected SDM’s claim for enhanced payments for food services that semester. See SodexoMAGIC, 333 F. Supp. 3d at 461–64. That was incorrect, and SDM’s breach-of- contract claim for Fall Semester 2016 survives summary judgment. 1. There Was Consideration for a Separate Contract for the Fall 2016 Semester. SDM provided consideration for a separate contract for the Fall 2016 Semester through forbearance. In its notice terminating the Management Agreement for convenience, Drexel specified a termination date 82 days later. But the Management Agreement required only 60 days’ notice to terminate for convenience. So once it received Drexel’s notice, SDM could have responded with its own notice of termination for convenience with an accelerated termination date. For example, SDM could have immediately given a 60-day notice of termination for convenience, which would have terminated the Management Agreement 22 days before Drexel’s proposed date. Thus, to accept Drexel’s offer, SDM would have to refrain from issuing its own accelerated notice to terminate for convenience and leaving campus as a result. And under deeply 65 entrenched Pennsylvania law, a party may provide consideration by agreeing to “refrain[] from doing anything which [it] has a right to do.” York Metal & Alloys Co. v. Cyclops Steel Co., 124 A. 752, 754 (Pa. 1924) (citation omitted); see generally Stelmack v. Glen Alden Coal Co., 14 A.2d 127, 128 (Pa. 1940) (explaining that, under Pennsylvania law, consideration must be “bargained for as the exchange for the promise,” and it must confer some “benefit to the party promising, or a loss or detriment to the party to whom the promise is made” (internal quotation marks omitted)); but cf. Chatham Commc’ns, Inc. v. Gen. Press. Corp., 344 A.2d 837, 840 (Pa. 1975) (“[T]he performance of an act which one party is legally bound to render to the other party is not legal consideration.”). By remaining on campus despite its ability to leave early, SDM provided the requisite consideration for a separate contract governing food services for the Fall Semester 2016.16 2. A Reasonable Jury Could Find that SodexoMAGIC Accepted Drexel’s Offer. The District Court separately rejected SDM’s claim for breach of contract for the Fall 2016 Semester due to a lack of a meeting of the minds. See SodexoMAGIC, 333 F. Supp. 3d at 463 (“There was no meeting of the minds as to the terms which Drexel proposed in its September 19 letter for Sodexo’s rates and commissions. Sodexo may not accept some terms, reject others, and then assert that an agreement conclusively existed.”). But under Pennsylvania law a party may accept a 16 Similarly, if SDM’s termination for cause were valid, then SDM’s continued provision of food services would also constitute consideration. 66 contract offer by promise or performance. See Herman v. Stern, 213 A.2d 594, 600 (Pa. 1965) (acceptance through promise); Ross v. Leberman, 148 A. 858, 859 (Pa. 1930) (acceptance through performance); Hartman v. Baker, 766 A.2d 347, 351 (Pa. Super. Ct. 2000) (same); see also Restatement (Second) of Contracts § 50 cmt. a (1981) (“In case of doubt, the offeree may choose to accept either by promising or by rendering the requested performance.”). And here, a reasonable jury could conclude that SDM accepted Drexel’s offer either by promise or, if it finds that SDM did not reject Drexel’s offer, through performance. The key evidence that SDM accepted through promise is SDM’s letter in response to Drexel’s offer. In that correspondence, SDM stated that it “agree[d] to remain on campus through December 10, 2016.” Letter from Timothy J. Fazio, Manion Gaynor & Manning, LLP, to Stephen A. Cozen, Cozen O’Connor P.C. (Sept. 26, 2016) (JA158). Even though the same letter also purports to terminate the Management Agreement for cause, a reasonable jury could still conclude that SDM separately promised to remain on campus through the specified date. Drexel disputes that SDM’s letter could qualify as an acceptance. It argues that SDM’s letter was not an acceptance because it did not reference Drexel’s offer and because it provided a different reason for staying – a commitment not to leave students mid-term. But Pennsylvania law does not require that an acceptance letter specifically reference the offer being accepted. See Ingrassia, 486 A.2d at 483 (“Offer and acceptance need not be identifiable and the moment of formation need not be pinpointed.”). Nor does an accepting party need to share the same motive as the offering party to 67 enter a contract. See generally 1 Corbin on Contracts § 3.4 (2021) (“[I]t is not necessary that the sole motive of the offeree must be the desire for the offered reward. It need not even be the offeree’s principal or prevailing motive.”). Drexel next contends that because the parties disagreed about whether SDM terminated the Management Agreement for cause, they could not have the meeting of the minds needed to form a contract for the Fall 2016 Semester. But a reasonable jury could find that, regardless of a broader disagreement between the parties, they still agreed to the “the material and necessary” terms on which SDM would remain on campus and provide dining services through December 10, 2016. Lombardo v. Gasparini Excavating Co., 123 A.2d 663, 666 (Pa. 1956). For these same reasons, a jury could reasonably conclude that SDM did not reject Drexel’s offer. With that finding, a jury could also conclude that SDM accepted Drexel’s offer through performance. It is undisputed that SDM remained on campus and provided the requested dining services – all with Drexel’s knowledge. And under Pennsylvania law, post-offer performance provides strong evidence of acceptance. See Selig v. Phila. Title Ins. Co., 111 A.2d 147, 151 (Pa. 1955); Hartman, 766 A.2d at 351; Accu-Weather, Inc. v. Thomas Broad. Co., 625 A.2d 75, 78–80 (Pa. Super. Ct. 1993). Thus, even if SDM’s written response to Drexel’s offer was ambiguous, a reasonable jury could still find that SDM accepted Drexel’s offer through performance. D. Drexel’s Challenge to SodexoMAGIC’s Catering- Shortfall Claim Fails. After the District Court’s summary judgment order, a dispute arose between the parties over which claims survived 68 summary judgment. They disagreed about whether SDM’s claim for catering shortfalls remained. SDM calculated those shortfalls based on a $3.4 million benchmark referenced in Section 9.2 of the Management Agreement.17 But through a later clarifying order, the District Court specified that “[c]laims based on Section 9.2 that are not related to any breach of a duty to renegotiate in good faith are still in the case.” Order re: Trial on Count V Issues (Nov. 1, 2018) (JA85). Following that order, the parties referred all remaining claims and counterclaims to arbitration and jointly moved to dismiss them. The granting of the parties’ joint motion achieved finality, which permitted the parties to appeal the District Court’s summary judgment ruling. See 28 U.S.C. § 1291. Now, after the joint motion dismissing all remaining claims, Drexel challenges the District Court’s clarifying order. It argues that the catering-shortfall claim did not survive summary judgment. That argument is meritless, nearing frivolous. In jointly moving to dismiss all remaining claims, Drexel represented to the District Court that the catering-shortfall claim remained. The arbitration agreement, which was an exhibit to the joint 17 The legal premise for any catering shortfall is vulnerable to the extent that the Management Agreement does not guarantee minimum catering net sales of $3.4 million the first year but provides only an assumption in that respect. Thus, if SDM had only an expectation but not a right to recover, its claim would be improperly premised on false billing – the submission of invoices for which there was no obligation to pay. 69 motion, stated that the only remaining claims were those in the parties’ mediation statements. And SDM’s mediation statement, which was an exhibit to the arbitration agreement, specifically identified the catering-shortfall claim. By those representations to the District Court, Drexel waived the right to argue the already-tenuous proposition that the District Court’s clarifying order impermissibly conflicted with its prior summary judgment ruling.18 E. SodexoMAGIC’s Claims for Unjust Enrichment Fail. On appeal, SDM also pursues unjust enrichment as an alternative remedy for two of its breach-of-contract claims. First, it seeks unjust enrichment if the duty to renegotiate in good faith in the Management Agreement is unenforceable. Second, SDM resorts to unjust enrichment to recover enhanced payments if the parties did not form a separate contract governing dining services for the Fall 2016 Semester. The District Court rejected SDM’s unjust enrichment claim due to pleading deficiencies, but it also concluded that the claim would fail as a matter of law. See SodexoMAGIC, 333 F. Supp. 3d at 473–74. For the reasons below, unjust enrichment is unavailable as a matter of law. 18 See In re Fine Paper Antitrust Litig., 685 F.2d 810, 817 (3d Cir. 1982) (“We will not interfere with a trial court’s control of its docket except upon the clearest showing that the procedures have resulted in actual and substantial prejudice to the complaining litigant.” (internal quotation marks omitted)); see also In re Asbestos Prods. Liab. Litig. (No. VI), 921 F.3d 98, 109 (3d Cir. 2019); United States v. Wecht, 484 F.3d 194, 217 (3d Cir. 2007). 70 SDM’s first argument for unjust enrichment rests on a novel legal theory. It argues that if the promise to renegotiate cannot be enforced, then despite the remainder of the Management Agreement, SDM should be able to seek compensation for the benefits Drexel received as a result of the unenforceability of that promise. Although it has not addressed that specific question, the Pennsylvania Supreme Court has left no doubt that “unjust enrichment is inapplicable when the relationship between parties is founded upon a written agreement or express contract, regardless of how ‘harsh the provisions of such contracts may seem in the light of subsequent happenings.’” Wilson Area Sch. Dist. v. Skepton, 895 A.2d 1250, 1254 (Pa. 2006) (quoting Third Nat’l Bank & Tr. Co. of Scranton v. Lehigh Valley Coal Co., 44 A.2d 571, 574 (Pa. 1945)).19 From the breadth and force of that expression, it may well be that the Pennsylvania Supreme Court would disallow an unjust enrichment claim even when a contract governing the parties’ relationship contains an unenforceable promise. But it is unnecessary to reach that issue. As explained above, the promise in the Management 19 See also Hershey Foods Corp. v. Ralph Chapek, Inc., 828 F.2d 989, 999 (3d Cir. 1987) (“Quantum meruit will not be awarded when there is an express agreement.” (citing Murphy v. Haws & Burke, 344 A.2d 543, 546 (Pa. Super. Ct. 1975))); Dan B. Dobbs & Caprice L. Roberts, Law of Remedies § 4.1(2), at 379 & n.64 (3d ed. 2018) (“A valid contract defines the obligations of the parties as to matters within its scope, displacing to that extent any inquiry into unjust enrichment.” (quoting Restatement (Third) of Restitution and Unjust Enrichment § 2(2) (2011) (emphasis omitted))). 71 Agreement to renegotiate in good faith is enforceable, and with a fully enforceable contract governing the parties’ relationship, the quasi-contractual remedy of unjust enrichment is unavailable. SDM also invokes unjust enrichment to recover enhanced compensation for providing on-campus dining services during the Fall 2016 Semester. At the outset, SDM’s claim for unjust enrichment is actionable only to the extent that no contract governs the parties’ relationship for that time period. See Wilson Area Sch. Dist., 895 A.2d at 1254. But here, if a jury finds that no contract governs the provision of Fall 2016 dining services, that would have to be because SDM did not accept Drexel’s offer. In that circumstance – the only one in which an unjust enrichment claim could proceed – the obstacle to SDM’s receiving enhanced compensation would be its own failure to accept Drexel’s offer. Yet an unjust-enrichment claim requires the “acceptance and retention of . . . benefits under such circumstances that it would be inequitable.” Meyer, Darragh, Buckler, Bebenek & Eck, P.L.L.C. v. Law Firm of Malone Middleman, P.C., 179 A.3d 1093, 1102 (Pa. 2018) (internal quotation marks omitted); see also Shafer Elec. & Constr. v. Mantia, 96 A.3d 989, 993 (Pa. 2014). And it is neither unjust nor inequitable to deny compensation to an entity that conferred a benefit but did not accept an offer for such compensation. Similarly, the Pennsylvania Supreme Court has rejected a claim of unjust enrichment when a benefit that a party confers upon another also protects its own interest. See Am. & Foreign Ins. Co. v. Jerry’s Sport Ctr., Inc., 2 A.3d 526, 546 (Pa. 2010). Here, by continuing to provide dining services, SDM served its own interest by maintaining its professed commitment to not 72 leave students without dining services mid-term. Accordingly, it would not be unjust or inequitable if SDM does not receive enhanced compensation for the dining services it provided for the Fall 2016 Semester. V. CONCLUSION For the foregoing reasons, we will affirm in part and vacate in part the District Court’s judgment and remand the remaining claims. We will affirm summary judgment in Drexel’s favor on SDM’s unjust enrichment and punitive damages claims as well as summary judgment in SDM’s favor on Drexel’s fraudulent inducement claim. Likewise, we will affirm the District Court’s clarifying order and its decision to deny Drexel’s motion to strike declarations by SDM witnesses under the sham affidavit rule. We will vacate the District Court’s order granting summary judgment to Drexel on SDM’s claims for fraudulent inducement, breach of contract for failure to renegotiate in good faith, and breach of a supplemental agreement for the Fall 2016 Semester. Those surviving claims will be remanded to the District Court. 73