Nearly two years into the COVID-19 pandemic, the battles over threshold business interruption coverage issues like the presence of physical loss or damage, causation, and the applicability of policy exclusions continue to rage.  Results have been mixed, with insurers notching wins in federal courts, and policyholders faring better in state courts and in certain jurisdictions.

But scores of policyholders who have avoided pre-discovery dismissal are now grappling with how other policy terms will impact their recovery.

Chief among these – particularly for policyholders with diverse and widespread physical locations and operations – is the impact of a classic question typically dealt with in handling catastrophic liability claims: how many occurrences are there?

The answer has profound implications for the amount of recovery, as it affects not only how deductibles may (or may not) apply but also how primary and excess coverage might respond to a large loss.

Three “tests” have emerged in case law to deal with this problem in the liability context. The “cause” test (adopted by the majority of jurisdictions) determines the number of “occurrences” by looking to the number of causes of injury or loss. The “effect” test determines the number of “occurrences” by looking to the number of resulting injuries or losses. And the “continuous process” test determines the number of occurrences by looking at the number of processes resulting in damage that were continuous, repetitive, and interrelated. See, e.g., Unigard Ins. Co. v. United States Fid. & Guar. Co., 728 P.2d 780 (Idaho Ct. App. 1986).

In property policies, “occurrence” might typically be defined (when it’s defined at all) as “a happening” or “series of happenings, … arising out of an event.” This is a broad definition, suggesting that a property policy should respond separately where losses have separate causal origins and be unconcerned with the effects or numbers of injuries that result. Francis F. Maloney III, The Application of ‘Per-Occurrence’ Deductible Provisions in First Party Property Claims, 37 Tort & Ins. L.J. 921, 924 (2002). In other words, apply the cause test and look for the “immediate act” that gave rise to the loss. If there is more than one such “immediate act,” then there should be more than one occurrence. See Koikos v. Travelers Ins. Co., 849 So. 2d 263, 271-73 (Fla. 2003) (“occurrence” looks to the immediate injury-causing act, not the underlying tortious omission, finding separate gunshots by a single shooter to be multiple occurrences).

An example of how the test works where property damage at different sites stems from different causes of loss can be found in Doe Run Res. Corp. v. Certain Underwriters at Lloyd’s London, 400 S.W.3d 463, 475-76 (Mo. App. 2013). The case dealt with a liability claim for property damage under an environmental policy. The insured sought excess coverage to pay remediation costs brought about by three distinct causes (mining, chat piles, and tailing ponds) at six separate sites. The Missouri Court of Appeal – reversing a lower court decision after a seven-day trial – found three occurrences at each active operation site, one for each cause of the environmental damage. In the court’s view, it did not matter that the effect – environmental contamination – was singular. Because the causes were distinct, involving physically distinct sites with different pollution migration profiles, separate causes – and therefore multiple occurrences – existed.

More recently, a Missouri-based federal court distinguished Doe Run on the basis that it involved property damage, not bodily injury, and the presence of “separate locations and physical distinctions” in finding multiple occurrences. Fluor Corp. v. Zurich Am. Ins. Co., No. 4:16CV00429 ERW, 2021 U.S. Dist. LEXIS 138047 at *35 (E.D. Mo. July 25, 2021). While the propriety of that decision is soon to be determined by the Eighth Circuit, the basis to make a similar argument remains.

Thus far, no court appears to have dealt with this problem in the context of widespread COVID-19-related business interruption losses, where the actual (or presumed) presence of the virus in turn led to a government-issued shutdown order, resulting in a loss of business income. In such a case, of course, the immediate “act” giving rise to the loss of income would not be the presence of the virus, but the shutdown order itself. If a business operates in multiple jurisdictions, thereby becoming subject to multiple shutdown orders, a strong argument may exist for multiple occurrences. Presuming per-occurrence deductibles do not swallow each individual loss, this would have the effect of increasing the available amount of coverage.