concurring:
Although I agree that Texas courts would focus on the events giving rise to liability to determine the number of occurrences, I do not agree that the question under Texas law is “whether HEB’s negligent employment relationship with its pedophilic employee, rather than the two acts of sexual abuse, ‘caused’ the injuries to the two children and gave rise to HEB’s liability.” 150 F.3d at 530. I would hold that the appropriate test for counting occurrences under Texas law is a “liability-triggering event” test rather than the “immediate cause” test applied by Judge Garza. Applying a liabihty-triggering event test yields the same result reached by Judge Garza. There were two occurrences: the employee’s molestation of each child.
Both Goose Creek Consol. ISD v. Continental Cas. Co., 658 S.W.2d 338, 339 (Tex. App. — Houston [1st Dist.] 1983, no writ) and Maurice Pincoffs Co. v. St. Paul Fire & Marine Insurance Co., 447 F.2d 204 (5th Cir.1971), are best understood as applying a “liability-triggering event” test, rather than *536the “immediate cause” test adopted in the majority opinion.1 In Goose Creek, the court focused on the particular language of the policy at issue in that case, which defined occurrence in terms of a “single event,” and concluded that each of two fires was a separate event and thus that there were two occurrences under the policy. 658 S.W.2d at 339. Although the court noted that a majority of states had adopted a “cause” test for occurrence, the court did not in fact apply a cause test, but rather relied on its understanding of the term “single event” in determining how many occurrences there were. Id. at 340.
This circuit’s approach was similar in Pin-cojfs. The Pincojfs court viewed the occurrence as the “event” giving rise to liability from the insured’s point of view. Pincoffs involved the sale by Pincoffs of contaminated bird seed to dealers who in turn sold the seed to bird owners, whose birds then died. The liability-triggering event, from Pincoffs’s point of view, was Pineoffs’s sale of contaminated seed rather than the original contamination of the seed (apparently by a third party) or the subsequent sales by the dealers. This'holding was not based on a conclusion that the contamination of the seed or subsequent sales did not cause the bird owner’s injuries or that the sale was an “immediate” or “intervening” cause, but rather on the idea that the sales were “the events or incidents for which Pincoffs is liable.” 447 F.2d at 206. Indeed, the court clearly acknowledged that “the damage to the birds resulted from the contamination of the bird seed.” Id. at 207.
Similarly, in this case, the children’s injuries resulted from HEB’s negligent hiring of the pedophilic employee and from the employee’s acts. But the events that gave rise to liability, even from HEB’s point of view, were the employee’s molestations of each child. Notably, however, applying the “immediate cause” test to the facts in Pincoffs would produce a different result: the immediate cause of the bird owner’s injuries was not Pineoffs’s sale to the dealers but the dealers’ sale to the bird owners (or even more particularly, the feeding of the birds). Nonetheless, the liability-triggering event from Pincoffs’s point of view was its sale of the seed to the dealers. Thus, Pincoffs supports the result that Judge Garza reaches here, but in my view is inconsistent with the test he would adopt.
Moreover, the cause test set forth by the district court in Transport Insurance Co. v. Lee Way Motor Freight, Inc., 487 F.Supp. 1325, 1327 (N.D.Tex.1980), cited by Judge Garza, is inconsistent with the result reached in this case. Lee Way involved a company-wide racially discriminatory policy. The court held that the occurrence was not each individual instance of discrimination that resulted from the discriminatory policy, but rather the adoption of the policy itself. By analogy, the occurrence in this case would be the negligent hiring and supervision of a pedophilic employee instead of the two instances in which that pedophilic employee harmed the children of customers.
Indeed, the Nevada Supreme Court in Washoe County v. Transcontinental Insurance Co., 110 Nev. 798, 878 P.2d 306 (1994), applied the causal analysis employed in Lee Way and reached a conclusion contrary to that reached by Judge Garza. The Nevada court held that even though multiple children had been molested by an employee negligently hired by the county, there was only one occurrence from the county’s point of view (the county’s negligent hiring). See id. 878 P.2d at 308. The Washoe County court actually cites Lee Way in support of its conclusion that there was a single occurrence. See id. The Lee Way [Washoe County approach bears little resemblance to the approach taken in Pincoffs and Goose Creek.
In addition to being more consistent with Pincoffs and Goose Creek, a liability-triggering event test would also be more consistent *537with the specific policy language in this case, which defines “occurrence” in terms of an “event.” The liability-triggering event test also has the virtue of avoiding the acrobatics required to conclude that the molester’s actions rather than HEB’s negligence caused the injuries to the molested children. Judge Garza relies on Burlington v. Mexican Am. Unity Council, 905 S.W.2d 359 (Tex.App.— San Antonio 1995, no writ), and Commercial Union Ins. Co. v. Roberts, 7 F.3d 86 (5th Cir.1993), for the proposition that occurrences should be counted with reference to the immediate cause of the injuries. Both Burlington and Roberts, however, involved the applicability of intentional tort policy exclusions (in Burlington, an exclusion for assault and battery; in Roberts, for intentional torts generally). These cases, which deal with what happens when the covered cause of harm is not independent of the excluded cause of harm, do not shed any light on counting the number of occurrences under an insurance policy. Suppose, for example, in Pincoffs, that the bird seed dealer’s insurance policy had contained coverage for its liability arising out of sales to pet store owners, but excluded coverage for the death of livestock resulting from the ingestion of contaminated food. Under Burlington and Roberts, there would have been no coverage for the birds’ deaths under the policy because the ingestion of contaminated food (the immediate, but excluded cause) and the sales to pet store owners (the covered cause) were not independent causes of the birds’ deaths. That does not mean, however, that, absent the exclusion, the number of occurrences would be based on each bird’s ingestion of the contaminated seed rather than the sales as the Pincoffs court held. In other words, whether coverage is negated because the immediate cause of harm is excluded from coverage (and the covered cause is not independent of the excluded cause) has no necessary connection to the way occurrences should be counted under the policy.
Finally, I disagree with Judge Garza’s conclusion that this circuit in Society of the Roman Catholic Church v. Interstate Fire & Cas. Co., 26 F.3d 1359 (5th Cir.1994), found that the term “occurrence” was not ambiguous. 150 F.3d at 532 n. 4 (criticizing Preferred Risk Mut. Ins. Co. v. Watson, 937 S.W.2d 148, 150 (Tex.App. — Fort Worth 1997, writ denied)). To the extent, as Judge Garza suggests, that the Catholic Church panel held that the definition of occurrence unambiguously requires that occurrences be counted from the injured party’s point of view, Judge Garza’s opinion here conflicts with Catholic Church. The definition of occurrence he would adopt (keyed to “immediate cause”) is inconsistent with the Catholic Church panel’s effects test. This inconsistency can be demonstrated by applying the two definitions to the facts in Anchor Casualty Co. v. McCaleb, 178 F.2d 322 (5th Cir.1949). In McCaleb, an oil well exploded, injuring the property of four people who brought suit. Like the Catholic Church court, the McCaleb court adopted an “effects” test for counting the number of occurrences under Texas law, holding that number of occurrences must be determined based on the property damage suffered by each individual property owner as a result of the explosion. See McCaleb, 178 F.2d at 325. Applying the test adopted by Catholic Church and McCaleb and examining the claim from the point of view of the injured parties, there were four occurrences in McCaleb. Under Judge Garza’s definition, however, there was only one occurrence because there was only one' immediate cause of the injuries (the oil well explosion). Thus, Judge Garza cannot simultaneously conclude that the Catholic Church case found the definition of occurrence to be unambiguous.
Applying the liability-triggering event test to the facts of McCaleb also leads to a finding that there was but one occurrence. Unlike Judge Garza, however, I would conclude that the panel in the Catholic Church case did in fact believe that the definition of occurrence was subject to more than one interpretation. The court noted that the meaning of the phrases “a continuous or repeated exposure to conditions” and “substantially the same general conditions” is “malleable” and that the meaning of “occurrence” “can be perplexing in application.” Id. at 1364. The court further explained:
An “occurrence” could be the church’s continuous negligent supervision of a priest, *538the negligent supervision of a priest with respect to each child, the negligent supervision of a priest with respect to each molestation, or each time the Diocese became aware of a fact which should have led it to intervene, just to name a few possibilities.
Id. Nevertheless, the panel was bound by the Louisiana Supreme Court’s decision in Lombard v. Sewerage & Water Bd. of New Orleans, 284 So.2d 905 (La.1973), in which the court held that the number of occurrences under Louisiana law must be determined from the point of view of the injured parties (ie., applied an “effects” test).
In sum, because the liability-triggering event test is more consistent with Texas law and with the language of policy in this case and is more easily applied, I respectfully concur.
. See Dow Chem. Co. v. Associated Indem. Corp., 727 F.Supp. 1524, 1528 (E.D.Mich.1989); Comment, Tung Yin, Nailing Jello to a Wall: A Uniform Approach for Adjudicating Insurance Coverage Disputes in Products Liability Cases with Delayed Manifestation Injuries and Damages, 83 Cal.L.Rev. 1243, 1254 (1995). I recognize that there is room for disagreement on this point. See Michael J. Murphy & Robert E. Wilder, The "Event” Debate in Asbestos-Related Excess of Loss Reinsurance Disputes, 31 Tort & Ins.L.J. 687, 702 (1996).