Seemingly in anticipation of the expected deluge of coverage disputes arising from Katrina, Rita, and Wilma, the Second Circuit released a careful opinion in a case where rain damage resulted from wind-caused openings in a building. The precise issue concerned application of a “wind deductible,” but the decision sweeps more broadly in addressing the methodology for interpreting policies in determining covered causes of loss and the relevance of insurance-industry practice. Turner Constr. Co. v. ACE Prop. & Cas. Ins. Co. (2d Cir. Oct. 28, 2005).

ACE Property & Casualty issued to Turner Construction a policy covering a Texas construction project. Turner had purchased a builder’s risk policy for the project, and the project was later damaged when rain entered openings that were caused by high winds. The policy included a large deductible for wind losses, which if applicable would wipe out coverage for the loss.
The builder’s risk policy provided first-party property coverage for covered causes of loss. In Turner, one question was what were the covered causes of loss. The policy covered risks of direct physical loss except those causes of loss listed in the exclusions. Wind is a covered cause of loss; certain risks were identified by name in the exclusions section. Here’s where things start to get tricky. One exclusion takes the form of “A, unless B.” Let’s call the subset of all circumstances where “A” is involved and where B also occurs “A!”. Formally, then, if (i) all risks are covered except those excluded and (ii) A is excluded except where B occurs, then “A!” is a covered cause of loss. The particular exclusion reads:

Rain . . . whether driven by wind or not, to Covered Property, unless located within a fully enclosed structure and then only for such loss that is caused by or results from rain . . . entering through an opening caused by a Covered Cause of Loss.

In this instance, the opening was caused by wind, a covered cause. Accordingly, while Rain in general was not a covered cause (“A”), rain-caused damage that results from an opening caused by wind (“A!”) would be a covered loss.
The issue in Turner is the application of the “wind deductible.” For the loss in question, “[t]he damage in this case . . . was directly caused by rain, and only indirectly caused by wind.” Slip op. at 5. The insurer urged that because wind was the covered cause of loss that created the opening, the wind deductible necessarily applies, even if the damage was from the entering rain.
The Second Circuit majority reasoned, however, that the wind deductible applied only where wind was the covered cause of the loss; given how the rain exclusion was structured and how formally that defined rain from a wind-caused opening as a covered peril (A!), the wind deductible did not apply. After citing a treatise on modern symbolic logic, the court elaborated:

The policy language says, in effect, “Rain is not a covered cause of loss, unless it enters a building as the result of a different covered cause of loss.” This is logically equivalent to saying, “if rain enters a building as the result of a different covered cause of loss, rain is a covered cause of loss.”

Slip op. at 7. Because the application of a deductible is a limitation of coverage, the insurer could not show that the “wind” deductible plainly and unambiguously applied to the rain-as-a-result-of-wind-caused-opening peril.
The dissent took a different tack. Judge Straub first contextualized the dispute against the backdrop of “windstorm” insurance policies, commonly understood as a thing in itself in the “Texas insurance industry.” Consequently, while implicitly acknowledging that the majority’s formal construction was linguistically permissible, the dissent believed that that construction was not a reasonable one (and therefore the policy was not ambiguous).
Relying on the maxim that exclusions do not create coverage, the dissent reasoned that the exception to the rain exclusion served only to

limit[] the force of the rain exclusion. That is, the clause gives the insured protection from a covered event (in this case, wind) and all the damages flowing from it, despite the fact that some of the damage has also been caused by an excluded event (rain). . . . In sum, my wholly logical reading is that rain is never a Covered Cause of Loss as defined under the policy, and even where rain in fact causes damage, that damage is attributed to the Covered Cause of Loss that caused the opening and allowed in the rain.

Slip op. at 11.
Responding to the majority’s rejoinder that this construction means that rain-caused damage can be subject to differing deductibles depending on the circumstances that let in the rain (wind, earthquake, or fire), the dissent conceded the point but was untroubled by it since the dissent assumed the conclusion that rain was never a separate cause of loss; consequently, applying different deductibles results from how the antecedent peril is covered (wind, earthquake, fire). In other words, whatever covered loss is caused by, e.g., earthquake is subject to a particular deductible, regardless whether the earthquake also let in rain.
The dissent’s rationale in part is revealed in a footnote where it embraces the remote cause of loss as being determinative (the wind that caused the opening) rather than the nearer cause of loss, the water. Slip op. at 12 n.1. More typically, courts apply the immediate cause at least where it will result in coverage (the policy generally being silent on the point of remote or immediate causes). E.g., Maples v. Aetna Cas. & Surety Co., 83 Cal.App.3d 641, 647-48 (1978) (“the term ‘accident’ unambiguously refers to the event causing damage, not the earlier event creating the potential for future injury”) (citing cases).
But the dissent mainly relied on what it characterized as Texas insurance industry practice (even though the policyholder is a New York company and the insurer is a Pennsylvania one). Judge Straub contended that “windstorm” coverage was a known quantity in Texas, and it would have covered this loss; reasoning back from the existence of this other coverage that had not been purchased, Judge Straub ruled that it was unreasonable to construe the builder’s risk policy to afford this separate coverage. (There was no evidence that Turner had been offered windstorm coverage separately and turned it down, though the salience of that is disputable anyway inasmuch as (i) the policyholder’s pure, undisclosed subjective intent is not relevant to limiting coverage and (ii) the policyholder could have turned down the offer of windstorm coverage thinking that the risk of that loss was covered at least in part by the builder’s risk policy.) The dissent noted that Texas windstorm coverage insured against loss from “wind-driven rain damage, regardless of whether an opening is made by the wind.” Slip op. at 14.
As the dissent concluded:

In light of this context, I would find that a reasonable observer knowledgeable of insurance practices in Texas would view damage caused by rain entering through a wind-blown opening as windstorm damage and would view a “wind deductible” as applying to the windstorm-protection aspects of the policy.

Slip op. at 14. Judge Straub does not analyze the choice-of-law question implicit in his analysis, though there certainly is a sound basis to conclude that Texas law would apply generally. But there is a separate question whether custom and practices of Texas are properly chargeable to a New York company and a Pennsylvania company that between them lawfully insure a project in Texas. Generally, custom and practice may be considered to give context to what the speaker or writer must have meant by his use of a particular term, see Frigaliment Importing Co. v. B.N.S. Int’l Sales Corp., 190 F. Supp. 116 (S.D.N.Y. 1960) (Friendly, J.); the dissent however is using custom not to interpret the meaning of wind as expressed in the contract but rather to presume that something that was unexpressed – windstorm – should limit the words that were expressed. This is not an appropriate basis to interpret an insurance policy to restrict coverage that otherwise may be found under a construction of the words used themselves. In other words, the dissent errs in relying on an implied, subjective-intent theory of contract, in derogation of the rule that establishes a contract “not on the parties’ having meant the same thing but on their having said the same thing.” Frigaliment, 190 F. Supp. at 117, citing Holmes, The Path of the Law at 178.
The Turner case will be relevant to the various losses associated with the 2005 hurricane season as policyholders and insurers seek to sort out which losses are covered or not. The majority’s rejection of the dissent’s approach – everybody knows that such and such is not covered – in favor of an approach based on construing the policy language where doing so will afford coverage is consistent with bedrock principles of insurance law (and contract law, too). It also makes it easier for insurers to administer the coverage and handle claims, for the claims handler can rely on the contract as written rather than on secret knowledge that lawyers debate that everyone knows. Turner also shows that policyholders need to closely read and analyze the policy language, for it is not obvious to most people that what is excepted from an exclusion is presupposed to be covered in the first instance by the insuring agreement.