EVANSTON INSURANCE COMPANY, Plaintiff-Appellee, v. HOUSING AUTHORITY OF SOMERSET; Kentucky Housing Authorities Self-Insurance Fund, Inc., Defendants-Appellees, Jason Steele; Robin Mullins; Rhonda Kay Griffin, Defendants-Appellants, 867 F.3d 653

OPINION SUTTON, Circuit Judge. In 2009, an old tree broke low on its trunk and fell on cousins Kaitlyn Griffin and Joshua Thacker. The tree killed Kait-lyn and severely injured Joshua. Kaitlyn was pregnant, and the doctors tried to save the baby, but he died an hour after being born. Their families filed—and won—a state court lawsuit for nearly $4 million against the Housing Authority of Somerset, which was found responsible for failing to maintain the area where the tree had been. At stake is how much of this judgment will be paid by insurance.

The Housing Authority belonged to a self-insurance Fund and enjoyed liability coverage through a policy the Fund purchased from the Evanston Insurance Company. Evans-ton denied coverage for the full judgment. It then filed a complaint for declaratory judgment in federal court against the families, the Housing Authority, and the Fund seeking to limit its liability to $1 million. The district court agreed with Evanston.

On appeal, we remanded the case to permit the district court to ensure that subject matter jurisdiction existed, to determine more specifically whether complete diversity existed between the parties. On remand, the district court determined that there was complete diversity and reinstated its original judgment. We affirm both rulings. I. On December 9, 2009, a large tree fell on Kaitlyn and Joshua. Kaitlyn died within minutes. She was pregnant at the time. Doctors delivered the baby, but Nicholas Ayden Steele died an hour after his mother. Joshua survived but suffered serious injury.

In December 2013, a state court jury found the Housing Authority liable for the accident and awarded $3,736,278 in damages. The Housing Authority did not face the prospect of liability alone. The Authority belonged to the Kentucky Housing Authorities Self-Insurance Fund along with dozens of other municipal housing authorities. The Fund’s policy with Evanston came with a series of caps and conditions, all detailed in a few pages. In 2014, Evanston filed a complaint in the Eastern District of Kentucky under the court’s diversity jurisdiction.

It sought a declaratory judgment limiting its liability under the Fund’s policy to $1 million. The complaint named as defendants all of the individual plaintiffs in the state court lawsuit along with the Housing Authority and the Fund. Meanwhile, through mediation of the state court case, Evanston agreed to pay the individual defendants the “policy limits” in return for the individual defendants’ agreement to dismiss the state court action and release the Housing Authority from further liability.

All parties stipulated that the “sole ... issue ... to be resolved by [the federal declaratory judgment] litigation, is whether policy limits are $1,000,000.00” or some larger amount. R. 34 at 2. Evanston took the position that $1 million was the coverage cap while the individual defendants claimed it was between $2 and $4 million. The district court ruled for Evanston. Evanston Ins. Co. v. Hous. Auth. of Somerset, No. 6:14-027 DCR, 2015 WL 8373945, at *1 (E.D. Ky. Dec. 8, 2015).

The individual defendants appealed, claiming for the first time that the interests of the Housing Authority and the Fund (both Kentucky entities) aligned with those of Evanston (an Illinois entity) and that the district court should have placed all three of them on the same side of the lawsuit and the individual Kentucky plaintiffs on the other side of the lawsuit. That party alignment would have destroyed complete diversity by placing Kentucky citizens on either side of the lawsuit. The panel remanded the case to permit the trial court to examine the proper alignment of. the parties in the first instance.

Evanston Ins. Co. v. Hous. Auth. of Somerset, 658 Fed.Appx. 799, 804 (6th Cir. 2016). On remand, the district court held that it had properly aligned the parties given their respective. interests in the primary dispute at the time of filing. Evanston Ins. Co. v. Hous. Auth. of Somerset, No. 6:14-027-DCR, 2016 WL 6650843, at *3 (E.D. Ky. Nov. 9, 2016). It reinstated the judgment. This appeal followed. II. Jurisdiction. Evanston invoked our diversity jurisdiction in filing this lawsuit. See 28 U.S.C. § 1332.

A federal court has such jurisdiction only if complete diversity exists, only if each of the plaintiffs comes from a different State from each of the defendants. See Exxon Mobil Corp. v. Allapattah Servs., Inc., 545 U.S. 546, 553-54, 125 S.Ct. 2611, 162 L.Ed.2d 502 (2005); Strawbridge v. Curtiss, 7 U.S. (3 Cranch) 267, 2. L.Ed. 435 (1806). In multiparty lawsuits, satisfaction of this requirement hinges on the alignment of the parties. Party alignment is not solely a function of how the parties align themselves; courts may realign the parties to .reflect their actual “interests in, the litigation.” Cleveland Hous.

Renewal Project v. Deutsche Bank Tr. Co., 621 F.3d 554, 559 (6th Cir. 2010). After looking to the “face of the pleadings!? and “the nature of the controversy,” courts must group parties with similar interests on the same side of the suit and pit them against parties with competing interests. Smith v. Sperling, 354 U.S. 91, 96, 77 S.Ct. 1112, 1 L.Ed.2d 1205 (1957). Two other considerations enter the mix.

Party interests-are not static, meaning we must determine when to assess these issues, It is “hornbook law” that subject matter jurisdiction founded in diversity “is governed by that condition, as it was at the commencement of the suit,” Grupo Dataflux v. Atlas Global Grp., L.P., 541 U.S. 567, 571, 124 S.Ct. 1920, 158 L.Ed.2d 866 (2004) (quotation omitted). What about parties who have multiple interests in a lawsuit? A party might be adverse to one party over one issue but share an interest with another.

A court has to pluck out the “primary dispute in the controversy” from the tangle of interésts that the lawsuit implicates. It then must align the parties according to their interests in the primary dispute “even where a different, legitimate dispute between the parties supports”. a different alignment, Cleveland Hous. Renewal Project, 621 F.3d at 559 (quotation omitted). Complete diversity exists in this case, as gauged by these requirements, and as indeed the district court concluded.

The parties’ formal alignment at the time Evanston filed this declaratory judgment action honors'the fiscal, legal, and practical realities of the dispute. Self-interest prompts most insurers to minimize how much they have to pay and prompts most insureds and injured parties to maximize what the insurer will cover. For that reason, as one of our sister circuits has pointed out, “the normal alignment of parties in a suit seeking a declaratory judgment of non-coverage is Insurer versus Insured and Injured Party.” Home Ins. Co. of Ill. v. Adco Oil Co., 154 F.3d 739, 741 (7th Cir. 1998).

The interests of today’s parties fit the normal pattern. Evanston is incorporated in Illinois and operates its principal place of business there, while the Housing Authority, the Fund, and the Individual Defendants all hale from Kentucky. Evanston must be adverse to the other parties. It is. Early in the litigation, but after Evans-ton filed its complaint, all of the parties acknowledged that the only dispute was “whether [the] ' policy limits are $1,000,000.00, as contended by Evanston ..., or whether policy limits are some amount greater than $1,000,000.00, as contended by Defendants.” R. 34 at 2.

Evans-ton’s interest in the primary dispute thus diverged from the interests of all of the Kentucky defendants—the individuals, the Housing Authority, the Fund—at the time of filing. Start with the individuals. They stood to benefit from more insurance rather than less. No one challenges that conclusion, either below or on appeal. Consider next the Housing Authority. It owed a $3.7 million judgment to the individuals at the time Evanston filed its complaint and hoped that Evanston would’cover as much of the judgment as possible.

That’s why the Housing Authority denied the part of the Evanston complaint asking the district court to issue a declaratory judgment limiting Evanston’s liability to $1 million. Evanston stood to lose a dollar for every dollar that the Housing Authority saved. Zero-sum relationships tend to be adverse. Same conclusion for the Fund. The Housing Authority was a member of the Fund at the time of filing. The Fund had an interest in helping its members avoid a large, liability. That’s one of the reasons it exists.

That aligned the Fund with the Housing Authority against Evanston. -The pleadings show as much, as the Fund, like the Housing Authority, denied that Evans-ton owed only $1 million on the policy. Trying' to counter this conclusion, the individuals claim that the Fund had a financial interest in an Evanston victory. The theory has several steps to it: (1) the more Evanston had to pay on the policy, the riskier the Fund would seem to insurers;. (2) higher risk policyholders have to pay higher premiums; and (3) so the Fund had an interest in minimizing Evanston’s liability. That does not follow Occam’s Razor.

The most direct interest of the Fund—and its fiduciary duty to its members—was to limit the liability of its members. Only if Evanston paid more, not less, could that happen. The Fund protects a lot of member housing authorities, and its duty is to avoid, not incur, judgments that might cripple one of those members. Think about the point another way. What if all of its members faced such a judgment? Would it not be in the Fund’s interest to- limit the liability of each member? We think so.

Whether one member, some members, or all members faced such a judgment, the Fund had an interest in lowering, .not increasing, their,liability exposure. Ample .adversity .existed at the time the complaint was filed, both as. a matter of law and of common sense. The individuals next point out that the Housing Authority and the Fund resisted adding Evanston to the state court case, suggesting that their interests align with the insurance company. But the primary dispute in state court concerned the Housing Authority’s liability and the amount of any such liability.

At that point, the individuals all wanted to establish liability and damages for as much as possible, while Evanston, the Housing Authority, and the Fund all wanted to avoid or minimize .liability. Once that ease ended and the federal case began, their interests diverged. Faced with a $3.7 million judgment, the Housing Authority and Fund had an interest in obtaining as much coverage as possi-Me from Evanston. And Evanston had an interest in establishing a $1 million cap. Different phases of a case sometimes lead to different adversarial relationships.

The individuals claim that an affidavit by Fund president Kent Latham leads to a different conclusion. Nine months after Evanston filed this declaratory judgment action, Latham said that the Fund “never” thought the policy would cover more than $1 million. R. 47-1 at 2. But this affidavit is not the Rosetta Stone the individuals think it is. Latham’s comment might suggest that, even though diversity jurisdiction existed at the time of filing nine months earlier, a live dispute no longer existed between these two parties. But that does not tell us how to align the parties at the outset.

No less importantly, a lawyer’s or client’s candid answer about the meaning of an insurance policy does not tell us how the case should be aligned for diversity purposes. It just tells us what that lawyer or client thought the policy said at that time, in this instance months after filing, not whether to respect the parties’ alignment in February 2014 and not what their true financial interests were. That the Housing Authority and the Fund have not filed briefs on the merits in this appeal means little. The brief of the individuals serves the same interests.

And no one has pointed to any merits argument that these entities might have made but that the individuals did not. All in all, diversity jurisdiction exists over this case. III. Turning to the merits, the language of the insurance policy is a good place to start. The Fund self-insures for up to $150,000 to cover general liability claims against any of its members. Members of the Fund also have coverage for claims that exceed $150,000 through a policy that the Fund bought from Evanston.

Part A of the policy insures Fund members for general liability stemming from “[bjodily injury, personal and advertising injury, or property damage caused by an occurrence that takes place in the coverage territory.” R. 1-4 at 14. The policy defines “occurrence” as “an accident, including continuous or repeated exposure to substantially the same general harmful conditions.” Id. at 31. It limits coverage to $1,000,000 for each “occurrence,” and $2,000,000 in the aggregate. Id. at 9.

Part B of the policy insures members for “wrongful act[s]” stemming from “any actual or alleged error, misstatement or misleading statement, act or omission, neglect, negligence, or breach of duty” committed by the insured parties. Id. at 48. It provides insurance based on “claims,” which it defines as demands “containing an allegation of wrongful act(

s) committed by and seeking damages against an Insured.” Id. at 9, 46. Like Part A, Part B covers up to $1 million per claim and $2 million in the aggregate. Both parts limit Evanston’s liability for events that qualify as occurrences and claims. Part A. Does Part A of the contract limit coverage to $1 million because the accident and injuries arose from a single occurrence? Yes, Part A obligates Evanston to provide a maximum of $1 million of coverage per “occurrence,” with an aggregate limit of $2 million should there be more than one occurrence. Id. at 9.

The contract defines “occurrence” to mean “an accident, including continuous or repeated exposure to substantially the same general harmful conditions.” Id. at 31. In common parlance, when one tree falls at one time, that is one occurrence and one accident. One dictionary describes an occurrence as “[a]ny incident or event, especially] one that happens without being designed or expected.” Webster’s New International Dictionary 1684 (2d ed. 1934).

Another describes it as an “accident, event or continuing condition that results in a personal injury or property damage that is neither expected nor intended from the standpoint of an insured party.” Black’s Law Dictionary (10th ed. 2014). The

definitions look to the number of causes, not the number of effects. Here we have one unfortunate cause (the fallen tree), and three tragic effects (two deaths and one serious injury). Case law leads to the same place. Davis v. Kentucky Farm Bureau Mutual Insurance Co. interpreted “occurrence,” defined the same way in that insurance contract, to mean the same thing. 495 S.W.3d 159, 161 (Ky. Ct. App. 2016). The Kentucky Supreme Court, it reasoned, had already “adopted the cause approach.” Id. at 163 (citing Continental Ins. Cos. v.

Hancock, 507 S.W.2d 146 (Ky. 1973)). “Under the prevailing cause approach,” it added, “the number of occurrences is determined by whether there is ‘but one proximate, uninterrupted and continuing cause which resulted in all of the injuries and damages.’ ” Id. (quotation omitted). A single tree, in a sudden moment, caused all of the injuries that gave rise to the Housing Authority’s liability. Just one occurrence, the untimely fall of the tree, triggered liability under the Fund’s policy, and the policy caps Ev-anston’s liability at $1 million per occurrence. An analogy supports this conclusion.

When one car collides with another, no one would refer to the event as several accidents even if the driver broke a rib on the steering wheel, the person in the passenger seat suffered a concussion after hitting the dash, and the person in the back seat suffered a psychological trauma. In day-today conversation, we would say a single car accident, a single occurrence, caused many injuries. That’s what happened here. Cf. Flemming v. Air Sunshine, Inc., 311 F.3d 282 (3d Cir. 2002). Not just Kentucky uses this- approach by the way; so do a majority of other jurisdictions. See Michael P.

Sullivan, What Constitutes Single Accident or Occurrence Within Liability Policy Limiting Insurer’s Liability to a Specified Amount Per Accident or Occurrence, 64 A.L.R. 4th 668 § 3 (2017). The individuals claim that another Kentucky case, Fryman v. Pilot Life Insurance Co., uses a different approach. 704 S.W.2d 205 (Ky. 1986). In Fryman, the Kentucky Supreme Court found that “[a]n accident is generally understood as an unfortunate consequence” or “result” of an unintended event. Id. at 206.

By defining accident using the singular “consequence” and “result” rather than the plural “consequences” and “results,” the individuals, say, Fryman adopted the effects approach to this problem. Appellants’ Br. 24-26. That reads, more into Fryman than it deserves. Nothing in the opinion indicates that it meant to make this sea change in state law or for that matter to overrule the State High Court’s prior decision in Hancock. No less importantly, Fryman (and the cases cited in it) do not directly také on the cause-versus-effects question.

They addressed whether an activity that a party carries out intentionally could count as an accident in the first place. Fryman itself considered whether a man’s death counted as “accidental” when he died while riding a motorcyclé under the influence of alcohol he voluntarily ingested. 704 S.W.2d at 205-06. That’s not remotely this case or this issue. The same can be said of the other cited cases. See Westfield Ins. Co. v. Tech Dry, Inc., 336 F.3d 503, 504 (6th Cir. 2003); Thompson v. W. Am. Ins. Co., 839 S.W.2d 579, 580-81 (Ky. Ct. App. 1992).

None of the cases asked how to count the number of occurrences for which an insurer is responsible after it is clear that something accidental occurred. The individuals cannot solve this problem by claiming that different branches of the same tree caused the different injuries, creating three accidents, not one. That slices things too finely. You can’t “cut a plank so many times that it ha[s] just one side.” Am. Energy Corp. v. Rockies Express Pipeline LLC, 622 F.3d 602, 606 (6th Cir. 2010).

The individuals separately argue that the deaths arose from multiple causes— cardiac arrest in one casé, blunt force in another—arid thus that each qualifies as ah occurrence under the cause approach. But that is a variation oh a defeated theme, defeated indeed by the Kentucky Supreme Court in Hancock itself. In that case, three men sued a nightclub owner for injuries sustained during a fight vdth several employees. Hancock, 507 S.W.2d at 147. Much like the individuals here, each man in Hancock suffered individual injuries.

Yet the court found that the, whole altercation amounted to one occurrence because the injuries arose out of “continuous exposure to substantially the same general conditions.” Id. at 152; see also Davis, 495 S.W.3d at 163-64 (describing Hancock). That was the harder case, as many employees of the bar caused the many injuries. Here just one tree in one fall caused all of the harm. Part B. The individuals next seek to expand Evanston’s coverage beyond $1 million by invoking Part B of the policy. That does not work either. ■ This part of the agreement covers liability stemming from “wrongful act[s].” R. 1-4 at 36.

The phrase covers “any actual-or alleged ... neglect, negligence, or breach of duty” by a Housing Authority employee. Id. at 48. Because a single event could qualify as both an “occurrence” under Part A and a “wrongful act” under Part B, the contract has parallel language in both parts to limit Evanston’s exposure.

Section 111(

E) of Part A explains the limit first: If any occurrence covered in whole or in part under Coverage Part A of this policy ... also constitutes a wrongful act(

s) covered in whole or in part under Coverage Part B of this policy ..,, then only the Coverage Part with the higher limits either for the each occurrence (Coverage Part

A) or for each claim(s) (Coverage Part

B) as listed on the Declarations Page or any Endorsement (and its corresponding retained limit) will apply to the entire claim or suit. If the each occurrence (Coverage Part

A) and the each claim(s) (Coverage Part

B) Limit(

s) of Insurance as listed' on the Declarations Page or any Endorsement are equal, only one limit will still apply to the entire claim or suit and it will be the each occurrence (Coverage Part

A) Limit(

s) of Insurance and its corresponding retained limit. Id. at 27-28.

Section III(

E) of Part B explains the corresponding limit in a similar way: If any occurrence covered in whole or in part under Coverage Part A of this policy ... also constitutes a wrongful aet(

s) covered in whole or in part under Coverage Part B of this policy ..., then only the Coverage Part with the higher limits (and its corresponding retained limit) will apply to the entire claim(s). If the applicable limits of each coverage part are equal, only one limit, the each occurrence Limit(

s) of Insurance of Coverage Part A (and its corresponding retained limit),- will apply to the entire claim(s). Id. at 44. The policy includes these provisions to address today’s scenario. Neither party disputes that a coverable event could constitute an occurrence and a claim. Both parts provide for a maximum of $1 million in coverage per event. Because the applicable limits of each coverage part are equal, the text of each says that only Part A’s limit applies. Evanston thus does not owe anything under Part B of the contract. “All matters of

interpretation begin with the text; some end there.” Sexton v. Panel Processing, Inc., 754 F.3d 332, 335 (6th Cir. 2014). This one ends there. The individuals disagree. They argue/ that the canon against superfluity demands that Part A’s § III(

E) and Part B’s § III(

E) be read differently. Otherwise, they say (with some superfluity of their own), one of the two sections would be “totally redundant and superfluous.” Appellants’ Br. 44. But they never close the deal on why that is so. The redundancy is necessary .to make sure the insured collects under just one of the two options. Yes, the drafters might have put this policy provision in one place, applicable to both parts, but it is hardly irrational to restate the point in each of the relevant sections. The minor differences in language between the two Parts simply reflect the different types of coverage in each section.

Contrary to their concern, there is no material ambiguity to resolve. For these reasons, we affirm.