Before JOLLY, HIGGINSON, and ENGELHARDT, Circuit Judges. PER CURIAM: Appellant Advanced Indicator and Manufacturing, Inc. claims its building was damaged by Hurricane Harvey’s winds. Advanced’s insurer, Acadia Insurance Company, determined that the damage to the building was caused by poor maintenance and routine wear and tear. When Acadia denied Advanced’s claim, Advanced sued. From this seemingly commonplace insurance dispute, we are faced with jurisdictional questions that have deeply divided district courts and a question of whether summary judgment was properly granted. I. Appellant, Advanced Indicator and Manufacturing, Inc., owned property at 1463 Brittmore Road in Houston. The property was insured by a policy issued by Appellee Acadia Insurance Company. The policy covered wind damage but did not cover damage from wear and tear or lack of maintenance. Hurricane Harvey struck southern Texas in 2017. After the hurricane, Advanced submitted a claim to Acadia regarding the damage to 1463 Brittmore Road, which it claimed was caused by the hurricane’s winds. Acadia acknowledged the claim and sent an adjuster, Nick Warren, as well as an engineer, Jason Watson, to assess the building. After inspecting the building, Watson determined that pre-existing conditions—including ongoing leaks from deterioration and poor workmanship—caused the damage, rather than winds from Hurricane Harvey. Warren adopted these conclusions in his recommendations to Acadia. Acadia then denied Advanced’s claim based on Watson’s conclusions and Warren’s recommendation. On August 7, 2018, Advanced sued Acadia and Warren in state court, alleging various claims, including breach of contract, common law bad faith, and violations of the Texas Prompt Payment of Claims Act. On August 30, 2018, Acadia elected to accept responsibility for Warren under § 542A.006 of the Texas Insurance Code, which provides that an insurer may accept liability for its agents. The next day, Acadia removed the case to federal court. One week later, Warren filed a motion to dismiss, arguing in part that Advanced could no longer state a claim against him. Advanced filed a motion to remand the case to state court, arguing that Warren was not improperly joined notwithstanding Acadia’s § 542A.006 election. In a management order, the district court denied the motion to remand and ruled without analysis that “Nicholas Warren is struck as improvidently joined.” Acadia later moved for summary judgment, arguing that it did not breach the policy and that Advanced could not segregate any damages caused by the hurricane from pre-existing damage, as required by Texas law. The district court granted Acadia’s motion, finding that Acadia’s denial of Advanced’s claim was based on “extensive consideration of the evidence” and that Advanced failed to carry its burden of showing that covered and non-covered damages could be segregated as required by Texas’s concurrent causation doctrine. In doing so, the district court struck the declaration of Peter de la Mora and Art Boudin, two of Advanced’s experts. The district court further granted summary judgment on Advanced’s extra-contractual claims. Advanced promptly appealed. II. We begin by considering whether the district court erred in denying Advanced’s motion to remand. The parties agree that when Advanced filed suit against Acadia (an out-of-state resident) and Warren (an in-state resident), Advanced had valid claims against both defendants. Because Advanced and Warren are both Texas residents, there was not complete diversity at the outset of the suit, and the matter could not be removed. Acadia then elected to accept liability for Warren pursuant to Texas Insurance Code § 542A.006, which provides that should an insurer accept responsibility for its agent after suit is filed, “the court shall dismiss the action against the agent with prejudice.” TEX. INS. CODE § 542A.006(c). Acadia filed a notice of removal the next day on the grounds that Advanced could no longer state a claim against Warren. We must determine whether Acadia’s § 542A.006 election made this matter removable. Advanced offers two arguments for why remand was proper. First, it contends Acadia’s removal in this case violates the involuntary-voluntary rule, which states that a case may only be made removable by a voluntary act of a plaintiff. Second, it argues that Warren was properly joined under the specific language of § 542A.006 because Acadia only elected to accept liability for him after suit was filed. We address each argument in turn. “The federal courts may exercise diversity jurisdiction over a civil action between citizens of different States if the amount in controversy exceeds $75,000.” Flagg v. Stryker Corp., 819 F.3d 132, 135 (5th Cir. 2016) (en banc) (citing 28 U.S.C. § 1332(a)(1)). Federal law provides that federal courts have removal jurisdiction over suits that could have originally been filed in federal court. See 28 U.S.C. § 1441(a). Further, if it later becomes clear that diversity jurisdiction exists even when it was not clear from the face of the initial pleading, the case can often be removed to federal court. See 28 U.S.C. § 1446(b)(3). “Ordinarily, diversity jurisdiction requires complete diversity—if any plaintiff is a citizen of the same State as any defendant, then diversity jurisdiction does not exist.” Flagg, 819 F.3d at 136. “However, if the plaintiff improperly joins a non-diverse defendant, then the court may disregard the citizenship of that defendant, dismiss the non-diverse defendant from the case, and exercise subject matter jurisdiction over the remaining diverse defendant.” Id. We set forth the standard for improper joinder in Smallwood v. Illinois Central Railroad Co., 385 F.3d 568 (5th Cir. 2004) (en banc).[1] A defendant may establish improper joinder in two ways: “(1) actual fraud in the pleading of jurisdictional facts, or (2) inability of the plaintiff to establish a cause of action against the non-diverse party in state court.” Id. at 573 (quoting Travis v. Irby, 326 F.3d 644, 646–47 (5th Cir. 2003)) (internal quotation marks omitted). Only the second method of proving improper joinder is at issue here. To show that a plaintiff cannot establish a cause of action against the non-diverse party in state court, a defendant must show “that there is no possibility of recovery by the plaintiff against an in-state defendant, which stated differently means that there is no reasonable basis for the district court to predict that the plaintiff might be able to recover against an in-state defendant.” Id. “The court may conduct a Rule 12(b)(6)-type analysis, looking initially at the allegations of the complaint to determine whether the complaint states a claim under state law against the in-state defendant.” Id. But when a plaintiff “has misstated or omitted discrete facts that would determine the propriety of joinder” then “the district court may, in its discretion, pierce the pleadings and conduct a summary inquiry.” Id. Importantly, “to determine whether a plaintiff has improperly joined a non-diverse defendant, the district court must examine the plaintiff’s possibility of recovery against that defendant at the time of removal.” Flagg, 819 F.3d at 137 (emphasis in original). Advanced argues that removal of this case based on Acadia’s post-suit, pre-removal § 542A.006 election violates the voluntary-involuntary rule. This judicially created rule dictates that “an action nonremovable when commenced may become removable thereafter only by the voluntary act of the plaintiff.” Weems v. Louis Dreyfus Corp., 380 F.2d 545, 547 (5th Cir. 1967). Advanced contends that because the § 542A.006 election was an action of a defendant, rather than the plaintiff, it cannot make the case removable. This question has deeply divided district courts. Some courts have held that the voluntary-involuntary rule bars removal when an insurer makes a § 542A.006 election after the filing of suit. See, e.g., Morgan v. Chubb Lloyds Ins. Co. of Tex., 541 F. Supp. 3d 754 (N.D. Tex. 2021). Others have held that the voluntary-involuntary rule is inapplicable if the agent is improperly joined at the time of removal. See, e.g., Ramirez v. Allstate Vehicle & Prop. Ins. Co., 490 F. Supp. 3d 1092 (S.D. Tex. 2020). Today we adopt the latter approach, which is a natural extension of our precedent. Indeed, “courts have long recognized an exception to the voluntary-involuntary rule where a claim against a nondiverse or in-state defendant is dismissed on account of fraudulent joinder.” Crockett v. R.J. Reynolds Tobacco Co., 436 F.3d 529, 532 (5th Cir. 2006). Moreover, our en banc court stressed that “to determine whether a plaintiff has improperly joined a non-diverse defendant, the district court must examine the plaintiff’s possibility of recovery against that defendant at the time of removal.” Flagg, 819 F.3d at 137 (emphasis in original). In this case, Warren was improperly joined after Acadia’s election because § 542A.006′s mandate that an agent be dismissed with prejudice dictates that Advanced had no possibility of recovery against him. Taking our holdings in Crockett and Flagg together, the answer to this case becomes clear: because Warren was improperly joined at the time of removal, Acadia’s removal was proper. Our recent decision in Hoyt v. Lane Construction Corp. confirms this analysis. 927 F.3d 287 (5th Cir. 2019). In that case, the Texas plaintiffs sued three defendants, two of whom were Texas residents and one of whom was not, in state court. Id. at 291. The plaintiffs voluntarily dismissed one in-state defendant, and the state court granted another in-state defendant’s motion for summary judgment. Id. at 292. The out-of-state defendants then removed to federal court. Id. The plaintiffs moved to remand the case to state court, arguing (as relevant here) that the state-court’s grant of an in-state defendant’s motion for summary judgment could not serve as the basis for removal because it violated the voluntary-involuntary rule. Id. On appeal, we held that the in-state defendant who was dismissed was improperly joined, as clearly demonstrated by the state court’s grant of summary judgment. Id. at 296–97. We further held that because improper joinder is an exception to the voluntary-involuntary rule, that rule did not bar removal in this case. Id. If the state court’s post-filing, pre-removal ruling dismissing an in-state defendant can make a case removable, so too can a § 542A.006 election, which eviscerates any claim against an agent. Advanced attempts to distinguish Hoyt by arguing that the state court’s ruling only confirmed that the in-state defendant had been improperly joined at the time suit was filed. But as we stated in Flagg, “the district court must examine the plaintiff’s possibility of recovery against that defendant at the time of removal.” 819 F.3d at 137 (emphasis in original). Advanced makes a second, and related, argument based on the language of § 542A.006.[2] The statute provides that when an insurer elects to accept responsibility for an agent before an action is filed, “no cause of action exists against the agent related to the claimant’s claim, and, if the claimant files an action against the agent, the court shall dismiss that action with prejudice.” TEX. INS. CODE § 542A.006(b). But when an insurer elects to accept responsibility for an agent after an action is filed, the statute provides that “the court shall dismiss the action against the agent with prejudice.” Id. § 542A.006(c). Advanced argues that because Acadia elected to accept responsibility for Warren after suit was filed, Warren was properly joined as § 542A.006(c), rather than § 542A.006(b), applies, and § 542A.006(c) excludes the language “no cause of action exists against the agent related to the claimant’s claim.” This dispute has likewise divided district courts. Some courts have agreed with Advanced that difference between the language of the subsections is significant. See, e.g., Stephens v. Safeco Ins. Co. of Ind., No. 4:18-00595, 2019 WL 109395, at *6–7 (E.D. Tex. Jan. 4, 2019). Others have found the difference insignificant as § 542A.006(c) still requires a court to dismiss the claim against the insurer with prejudice. See, e.g., Electro Grafix, Corp. v. Acadia Ins. Co., No. SA-18-CA-589-XR, 2018 WL 3865416, at *3-4 (W.D. Tex. Aug. 14, 2018). We hold that Advanced’s argument is unavailing because it fails to consider Flagg‘s command that “the district court must examine the plaintiff’s possibility of recovery against that defendant at the time of removal.” 819 F.3d at 137 (emphasis in original). Because Acadia elected to accept liability for Warren before removal, even though suit had already been filed, the statute required that “the court shall dismiss the action against the agent with prejudice.” TEX. INS. CODE § 542A.006(c). At the time of removal, then, it would have been proper for the district court to find that “there is no possibility of recovery by [Advanced] against an in-state defendant”—here, Warren. Smallwood, 385 F.3d at 573. Accordingly, the differences between §§ 542A.006(b) and 542.006(c) are not material as long as the insurer elects to accept liability for the agent before removal. III. We now consider whether the district court erred in granting summary judgment.[3] We review a grant of a motion for summary judgment de novo and apply the same standard as the district court, viewing the evidence in the light most favorable to the nonmovant. First Am. Title Ins. Co. v. Cont’l Cas. Co., 709 F.3d 1170, 1173 (5th Cir. 2013). Summary judgment is appropriate where “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). “A dispute about a material fact is genuine if the evidence is such that a reasonable jury could return a verdict for the nonmoving party.” Stults Conoco Inc., 76 F.3d 651, 654 (5th Cir. 1996) (quoting Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 248 (1986)) (cleaned up). “Courts do not disfavor summary judgment, but, rather, look upon it as an important process through which parties can obtain a ‘just, speedy and inexpensive determination of every action.’” Goldring v. United States, 15 F.4th 639, 644 (5th Cir. 2021) (quoting Celotex Corp. v. Catrett, 477 U.S. 317, 327 (1986)). The district court essentially granted summary judgment on Advanced’s policy breach claim for two reasons. First, the district court credited Acadia’s investigation over Advanced’s investigation, essentially finding that Hurricane Harvey did not cause Advanced’s damages. Second, the district court found that because Advanced could not differentiate its pre-existing losses from its Hurricane Harvey losses, its claim fails under Texas’s concurrent causation doctrine. We address each argument in turn. Because we have diversity jurisdiction over this suit, we apply Texas law. Greenwich Ins. Co. v. Capsco Indus., Inc., 934 F.3d 419, 422 (5th Cir. 2019). In a claim for breach of an insurance contract, Texas law requires the insured to prove: “(1) the existence of a valid contract; (2) performance or tendered performance by the plaintiff; (3) breach of the contract by the defendant; and (4) damages sustained by the plaintiff as a result of the breach.” Certain Underwriters at Lloyd’s of London v. Lowen Valley View, L.L.C., 892 F.3d 167, 170 (5th Cir. 2018) (quoting Smith Int’l., Inc. v. Egle Grp., LLC, 490 F.3d 380, 387 (5th Cir. 2007)) (internal quotation marks omitted). The view taken by the district court seems to have been that Acadia did not breach the contract because it properly denied Advanced’s claim based on Watson’s report and Warren’s recommendation. The question before us is whether Advanced has put forth sufficient summary judgment evidence to create a disputed issue of material fact regarding whether the damage to the building came from a covered cause, here, wind from Hurricane Harvey. We hold that it has. Specifically, Advanced points to the testimony of Thomas Ross, a public adjuster, who stated that the roofing system “completely failed,” and that the damage was “absolutely” caused only by the hurricane. De la Mora, Advanced’s expert, also testified at his deposition that the damage was caused by Hurricane Harvey. Finally, the record contains previous reports demonstrating that the building was in good shape. Viewing this evidence in the light most favorable to Advanced, a reasonable jury could find that Hurricane Harvey’s winds were the cause of the damage to the building. To be sure, Acadia offers evidence to the contrary. But this evidence serves only to create a factual dispute. We next turn to the district court’s alternative holding: that the concurrent causation doctrine bars Advanced’s claim because it cannot segregate covered losses from non-covered losses.[4] Under this doctrine, 4 This Court has recognized the substantial gaps in the concurrent causation doctrine and, as a result, twice certified questions to the Supreme Court of Texas, including: Whether the concurrent cause doctrine applies where there is any non-covered damage, including “wear and tear” to an insured property, but such damage does not directly cause the particular loss eventually experienced by plaintiffs; If so, whether plaintiffs alleging that their loss was entirely caused by a single, covered peril bear the burden of attributing losses between that peril and other, non-covered or excluded perils that plaintiffs contend did not cause the particular loss; and “when covered and non-covered perils combine to create a loss, the insured is entitled to recover that portion of the damage caused solely by the covered peril.” Dallas Nat’l Ins. Co. v. Calitex Corp., 458 S.W.3d 210, 222 (Tex. App. 2015) (collecting cases). “Because an insured can recover only for covered events, the burden of segregating the damage attributable solely to the covered event is a coverage issue for which the insured carries the burden of proof.” Id. (citing Wallis v. United Servs. Auto. Ass’n, 2 S.W.3d 300, 303 (Tex. App. 1999)). “Failure to segregate covered and noncovered perils is fatal to recovery.” Id. at 223 (quoting Comsys Info. Tech. Servs., Inc. v. Twin City Fire Ins. Co., 130 S.W.3d 181, 198 (Tex. App. 2003)) (internal quotation marks and alternation omitted). An insured may carry its burden by putting forth evidence demonstrating that the loss came solely from a covered cause or by putting forth evidence by which a jury may reasonably segregate covered and non-covered losses. See Travelers Indem. Co. v. McKillip, 469 S.W.2d 160, 162 (Tex. 1971). Here, the same evidence that supports Advanced’s argument that Hurricane Harvey caused some of its damage supports its argument that Hurricane Harvey caused all of the damage. Indeed, both Ross and De la Mora testified that the hurricane was the sole cause of Advanced’s loss. Accordingly, because a jury could reasonably find that all of Advanced’s loss comes from a covered cause, the concurrent causation doctrine does not bar recovery. See id. Finally, we turn to Advanced’s bad faith and TPPCA claims. The district court dismissed Advanced’s bad faith claims because it dismissed Advanced’s breach of contract claim. The Texas Supreme Court has described breach of contract and bad faith claims as “largely interwoven.” USAA Tex. Lloyds Co. v. Menchaca, 545 S.W.3d 479, 489 (Tex. 2018). Accordingly, because we reverse the district court’s dismissal of Advanced’s breach of contract claim, we find it appropriate to reverse the dismissal of Advanced’s bad faith claims as well. For similar reasons, we reverse the district court’s dismissal of Advanced’s TPPCA claim. To prevail on a TPPCA claim for damages and interest, an insured must demonstrate the insurer’s (1) liability under the policy, and (2) failure to comply with the TPPCA. Barbara Techs. Corp. v. State Farm Lloyds, 589 S.W.3d 806, 812–13 (Tex. 2019). A wrongful rejection of a claim, if reduced to judgment, can be considered a delay in payment under the TPPCA. See Higginbotham v. State Farm Mut. Auto. Ins. Co., 103 F.3d 456, 461 (5th Cir. 1997); see also Barbara Techs., 589 S.W.3d at 819–20 (explaining that the TPPCA can apply in cases where “there is a judgment that the insurer wrongfully rejected the claim”). Accordingly, should Advanced succeed on its breach claim, it may succeed on its TPPCA claim as well. We therefore reverse the dismissal of this claim. In sum, we find that summary judgment was not warranted on Advanced’s breach of contract claim given the evidence Advanced has put forth. This conclusion requires reversal of the district court’s dismissal of Advanced’s other claims. IV. For the foregoing reasons, we AFFIRM the district court’s denial of the motion to remand, REVERSE the grant of summary judgment on Advanced’s claims, and REMAND the matter to the district court for proceedings consistent with this opinion. KURT D. ENGELHARDT, Circuit Judge, concurring: I join the panel’s opinion in full because it properly applies this court’s binding precedents. I write separately to highlight an important implication of this decision. Specifically, our holding all but eviscerates the voluntary-involuntary rule. As our opinion explains, that rule holds that “‘an action nonremovable when commenced may become removable thereafter only by the voluntary act of the plaintiff.’” Crockett v. R.J. Reynolds Tobacco Co., 436 F.3d 529, 532 (5th Cir. 2006) (quoting Weems v. Louis Dreyfus Corp., 380 F.2d 545, 547 (5th Cir. 1967)) (emphasis added). The voluntary-involuntary rule is derived from century-old Supreme Court caselaw. See Powers v. Chesapeake & Ohio Ry. Co., 169 U.S. 92 (1898); Am. Car & Foundry Co. v. Kettelhake, 236 U.S. 311 (1915). And it has been firmly established in the Fifth Circuit at least since Weems, 380 F.2d at 547, where we held that the voluntary-involuntary rule survived the enactment of 28 U.S.C. § 1446(b). Despite this storied history, for practical purposes the voluntary-involuntary rule does not survive our ruling today. If, as here, a defendant can take a post-filing, pre-removal action which the negates the claim against the in-state defendant, thereby triggering the improper joinder “exception,” then the voluntary-involuntary rule does not exist in practice.[5] As our opinion explains, this outcome is dictated by our intervening decisions in Crockett, 436 F.3d at 532, which held that improper joinder is an exception to voluntary-involuntary rule, and Flagg v. Stryker Corp., 819 F.3d 132, 137 (5th Cir. 2016) (en banc), which emphasized the need to focus our improper joinder analysis on the facts as they stood at the time of removal. Neither case explicitly overruled the voluntary-involuntary rule, but reading those holdings together requires effectively overruling it in this case, even though the rule existed decades before either decision. Moreover, both Crockett and Flagg have been the subject of criticism, particularly for how their broad holdings apply to cases like this one. District courts have complained that Crockett‘s characterization of improper joinder as an “exception” to the voluntary-involuntary rule is misleading, as they should be viewed as separate but sometimes competing doctrines. See, e.g., Morgan v. Chubb Lloyds Ins. Co. of Tex., 541 F. Supp. 3d 754, 763–64 (N.D. Tex. 2021). Indeed, the sole authority Crockett cited to hold that the characterization of improper joinder as an exception to the voluntary-involuntary rule is “settled law” is the Eleventh Circuit case Insinga. See Crockett, 436 F.3d at 532 n.2. But that case dealt with a different situation involving a defendant entitled to sovereign immunity. The Insinga court therefore emphasized that the case “involve[d] a determination by the [state] court that the resident defendant was never properly before the court, rather than a determination that the court had jurisdiction of that defendant but that the case against him, although not frivolous, was not meritorious.” 845 F.3d at 254–55 (emphasis added). District courts have also implicitly criticized Flagg‘s requirement that we focus on jurisdictional facts as they stood at the time of removal in improper joinder cases. See Stephens v. Safeco Ins. Co. of Ind., No. 4:18-cv-595, 2019 WL 109395, at *5 (E.D. Tex. Jan. 4, 2019) (“Whether a non-diverse defendant is improperly joined is a binary question; the defendant is either a proper party when joined to [the] suit or the defendant is an improper party when joined to the suit. . . . It does not follow that a non-diverse defendant that is initially properly joined may become initially improperly joined.”). And despite the Flagg court’s broad language, that case’s focus was on whether post-removal events could revive a claim against an in-state defendant, not whether pre-removal (but post-filing) events could mandate removal. But whatever the faults of Crockett and Flagg, they bind this court today. Our precedents in this field have flummoxed parties and district courts alike. I hope that our decision today offers some clarity. But I recognize that, given the wicked web our caselaw weaves, these concerns may merit further consideration in a future appropriate case.
