The only Sixth Circuit opinion cited in the denial letter, Jones v. Metropolitan Life, did not even involve intoxication, and in fact supports the Kovaches. There, the court concluded that MetLife acted arbitrarily and capriciously by interpreting the term
"accident" in a manner that excluded coverage where the plaintiff, a nurse, injured her knee while bending down to administer first aid to a patient. It thus "added an eligibility requirement [that the injury be caused by an external force or event] under the guise of interpreting the term ‘accident' that does not exist in either the Plan documents or federal
common law." 385 F.3d at 659, 665. Further, in Walker v. Metropolitan Life, the insured had a BAC of .22-substantially higher than Mr. Kovach's .148-and was driving at a high rate of speed when he crashed into a wall. 24 F. Supp. 2d at 777. And
the injured driver in Miller v. Auto-Alliance had a BAC of .29, 953 F. Supp. at 173, while the insured in Fowler v. Metropolitan Life had a BAC of .26, 938 F. Supp. at 478. Finally, the policy in Cates v. Metropolitan Life excluded injuries resulting from "the use of any drug or medicine." There, the administrator concluded in its denial letter that "alcohol is considered by the medical community to be a drug." 14 F. Supp. 2d at 1025. No such exclusion is present in the Kovaches' AD&D policy. The district court in Cates, without providing any analysis, upheld the insurer's denial of the claim on the basis that the insured's injuries were reasonably foreseeable at a BAC of .18 and therefore nonaccidental. Id. at 1027.

Zurich is essentially left with only two drunk-driving decisions involving facts generally analogous to Mr. Kovach's collision-Nelson v. Sun Life, 962 F. Supp. at 1013 (upholding the denial of benefits to an insured who was involved in a wreck while
driving with a BAC of .18), and Cates v. Metropolitan Life. But having to place all of its eggs in the basket of these two lower court decisions (one of which, Cates, is factually distinguishable because of the drug exclusion) to deny the Kovaches' claim
was unreasonable in light of the decision in Harrell v. Metropolitan Life Insurance Co., 401 F. Supp. 2d 802 (E.D. Mich. 2005), which reached the opposite result. There, the insured died in a wreck and was found to have a BAC of .17. Id. at 805. The district court rejected the plan administrator's argument that the death was not "accidental," holding that, "to allow denial of benefits when a bad result is ‘reasonably foreseeable' undermines a common conception of ‘accidental injuries,' and therefore could violate ERISA's requirement that benefit plans be ‘written in a manner calculated to be understood by the average plan participant.'" Id. at 812-13 (quoting 29 U.S.C. § 1022(a)).

3. The foreseeability of injury

Zurich has offered no evidence whatsoever about the foreseeability of harm to Mr. Kovach based on his BAC of .148. Perhaps the reason that it did not do so is that its outright exclusion of alcohol-related collisions from its definition of "accidental" is not born out by the statistics, as aptly summarized in the Lennon dissent:

[A]ccording to the National Highway Traffic Safety Administration,
17,105 people died in alcohol-related motor vehicle crashes in 2003, a
figure that accounts for 40 percent of all traffic-related deaths that year.
The Federal Bureau of Investigation's 2003 Uniform Crime Report
indicates that an estimated 1,448,148 motorists were arrested for driving
under the influence that year. Finally, the Substance Abuse and Mental
Health Services Administration, in a report issued September 2005,
found an estimated 30.7 million persons nationwide took
alcohol-impaired trips during 2003. Even assuming only 10 million
alcohol-impaired trips occurred that year, a mere 14.4 percent of
impaired motorists were arrested, while 0.17 percent died in
alcohol-related incidents. Thus, injury or death most certainly cannot be
deemed a "highly likely" consequence of driving while intoxicated. See
West v. Aetna Life Ins. Co., 171 F. Supp. 2d 856, 904 (N.D. Iowa 2001)
("What ‘common knowledge' should actually tell a person driving while
intoxicated is that he or she is far more likely to be arrested for driving
while intoxicated than to die or be injured in an alcohol related
automobile crash, and far more likely to arrive home than to be either
arrested, injured, or killed.").
Lennon, 504 F.3d at 630 (Clay, J., dissenting).

The Lennon dissent employed a conservative estimate of the number of drunk drivers per year that was more than three times less than the actual estimate provided by the Substance Abuse and Mental Health Services Administration. Using the full 30.7 million figure results in a death rate of only one-twentieth of one percent. These statistics should not be taken as a dismissal of the tragic consequences of drunk-driving wrecks; instead, they are intended to show only that the likelihood of serious injury or death for each person who drives while intoxicated is something far less than "reasonably foreseeable"-the standard used by Zurich in denying the Kovach's claim.

4. Other risky activities

We also note that Zurich's definition of "accidental," which excludes activities that render the risk of serious injury "reasonably foreseeable," would bar accidental injury coverage in numerous situations in which the typical policyholder would expect to be covered. For example, a driver who runs off the road while typing a text message into a mobile phone would certainly consider himself or herself to have been involved in an "accident," and would therefore expect to be entitled to benefits under an accidental-injury policy. But the New York Times recently reported on a Virginia Tech study showing that text messaging increased a driver's risk of collision by 23 times. Matt Richtel, In Study, Texting Lifts Crash Risk by Large Margin, N.Y. Times, July 27, 2009. Moreover, a study of young drivers in England found that reaction times of young drivers were reduced by text messaging three times more than by drinking alcohol to the legal limit. Helen Nugent, Texting While Driving Is More Dangerous than Drink-Driving, London Times, September 18, 2008.

A driver also substantially increases his or her risk of a collision by driving while fatigued, see Lori Yerdon, Fatigued Driving Comparable to Drunk Driving, U.S. Army Combat Readiness/Safety Center News Release, June 16, 2008, available at; driving at excessive speeds, see Synthesis of Safety Research Related to Speed and Speed Limits,; and driving after taking certain over-the-counter medications, notably antihistamines, opioids, and muscarinic antagonists, see U.K. Dept. for Transport, Road Safety Report No. 24: Over-the-Counter Medicines and the Potential for Unwanted Sleepiness in Drivers, available at (Note that we do not endorse the results of the cited studies; these citations are included only to bolster the intuitive notion that the risk of driving can be substantially increased when drivers engage in common activities that do not carry the moral stigma of drunk driving.)

If Zurich denied coverage for injuries involving the above factors on the basis that they were not caused by an "accident," we would likely conclude that its decision was arbitrary and capricious in light of a typical policyholder's expectations. We see no
reason not to reach the same conclusion based on the Kovaches' legitimate expectations in the present case.

5. Lack of a specific exclusion

Zurich could have easily added an exclusion in the Plan for driving while intoxicated if it had wished to do so, but it did not. The sheer number of court cases nationwide involving disputes over claims by drunk drivers certainly would have put it
on notice that it would likely face claims under its AD&D policies based on injuries sustained in alcohol-related collisions. Zurich did, however, include specific exceptions for "skydiving, parasailing, hangglinding [sic], bungee-jumping, or any similar activity." But under the "reasonably foreseeable injury" standard that Zurich employed to deny the Kovaches' claim, these exclusions would be unnecessary and redundant because all of the listed activities involve a reasonably foreseeable risk of injury. See Restatement (Second) of Contracts § 203(a) (1981) ("[A]n interpretation which gives a reasonable,
lawful, and effective meaning to all the terms is preferred to an interpretation which leaves a part unreasonable, unlawful, or of no effect.").

In effect, Zurich's definition of the term "accidental" has added an eligibility requirement (i.e., an exclusion) that is not in the Plan. Based on the facts of the current case, that renders its interpretation of the Plan arbitrary and capricious. See Jones v.
Metro. Life. Ins. Co., 385 F.3d 654, 665 (6th Cir. 2004) ("In this case, MetLife added an eligibility requirement under the guise of interpreting the term "accident" that does not exist in either the Plan documents or federal common law; therefore, MetLife's interpretation of the Plan is arbitrary and capricious."). We also note that to the extent Zurich relied on the presence of drugs in Mr. Kovach's blood samples to deny coverage, where the record provides no basis to find that these drugs were present before the accident, that reliance is a further indication of an arbitrary and capricious decision.

6. Wickman standard

Given the varying interpretations of the word "accidental" that have been employed by the district courts in this circuit, compare Nelson v. Sun Life Assurance Co. of Canada, 962 F. Supp. 1010, 1013 (W.D. Mich. 1997) (applying a "reasonable
foreseeability" standard) with Harrell v. Metropolitan Life Insurance Co., 401 F. Supp. 2d 802, 812-13 (E.D. Mich. 2005) (applying a "highly likely to occur" standard), the time is ripe for this court to adopt a uniform standard for determining whether an injury is "accidental" in ERISA cases where the word is not otherwise defined in the applicable policy. After giving this matter considerable thought, we adopt the key holding announced in the seminal First Circuit case of Wickman v. Northwestern National Insurance Co., 908 F.2d 1077 (1st Cir. 1990). This key holding consists of asking "whether a reasonable person, with background and characteristics similar to the insured, would have viewed the injury as highly likely to occur as a result of the insured's intentional conduct." Id. at 1088.

As stated by the concurring opinion in Lennon,

Wickman's standard . . . is a high bar, and arguably many collisions
involving a drunk driver would not meet it: as a number of courts have
noted, the number of drunk driving arrests swamps the number of drunk
driving injuries or deaths, making it difficult to conclude that an injurious
collision is "highly likely to occur as a result" of driving while
intoxicated. 504 F.3d at 625.

The dissent, however, asserts that "[t]he combination of alcohol and Vicodin made it even more likely that a traffic collision would occur, rendering Kovach's behavior all the more reckless." (Dissenting Op. at 29) (emphasis added). Driving while
intoxicated concededly makes a collision "more likely" to occur than if the driver is not intoxicated, but this does not mean that a collision under such circumstances is "highly likely." Solely for the purpose of illustration, and without any pretense of being statistically accurate, let us hypothetically assume that a sober driver has a 1% probability of a collision. By contrast, assume that driving while intoxicated results in a 10% probability of a collision, leading to the conclusion that a collision is "more likely" to occur while driving intoxicated.

But to reach the level of "more likely than not," the probability of an accident would, by definition, have to be in excess of 50%. And because "highly likely," which is necessary under the Wickman standard to obviate Zurich's liability for an accidental
injury, obviously means a good bit more probable than simply "more likely than not," one might contemplate a 75% or higher probability before the average person would be persuaded that a collision was "highly likely" to occur. So even if the
motorcycle/Vicodin factors relied on by the dissent made Mr. Kovach's accident more likely, that is not sufficient to meet the Wickman standard and thereby relieve Zurich of its policy obligations.

The dissent criticizes this analysis by citing Stamp v. Metropolitan Life Insurance Co., 531 F.3d 84 (1st Cir. 2008), for the proposition that the "highly likely" standard is not wedded to the "more likely than not" standard. (Dissenting Op. at 29 n.4) Stamp is distinguishable from the present case, however, because the insured in Stamp, like the insured in Lennon, was driving with a BAC of more than three times the legal limit. Stamp, 531 F.3d. at 86. Thus, Stamp's behavior was materially closer to the extreme alcohol-related behavior at issue in Lennon than Mr. Kovach's. This material distinction in circumstances precisely meshes with a key point stressed by the majority opinion in Stamp that

courts have emphasized the decedent's level of intoxication when
determining that a plan administrator's denial of benefits was
reasonable. . . . We endorse this approach. The Wickman analysis does
not require a categorical determination that all alcohol-related deaths are
per se accidental or nonaccidental. Rather, it leads us to consider the
circumstances of the fatal event in question. Id. at 90-91 (internal citations omitted).

We further believe that the Russian roulette example cited in Stamp and by the dissent is inapposite because playing Russian roulette has zero social utility, whereas using motorized transportation to move about has a very high level of social utility. In sum, riding a motorcycle while "somewhat impaired" (as opposed to being "grossly intoxicated") may make an accident more likely to occur, but the risk of such an accident is not so "overwhelmingly disproportionate" when compared to the social utility involved, see id. at 93 (citation omitted), that an ordinary person would consider a collision involving the impaired driver as nonaccidental.

The solution for insurance companies like Zurich is simple: add an express exclusion in policies covering accidental injuries for driving while under the influence of alcohol, or for any other risky activity that the company wishes to exclude.
Policyholders would thus be able to form reasonable expectations about what type of coverage they are purchasing without having to make sense of conflicting bodies of caselaw that deal with obscure issues of contractual interpretation. Had Zurich included such an exclusion in first place, this litigation would have been entirely unnecessary.

7. Conclusion

Zurich's decision to deny benefits to the Kovaches was contrary to the everyday meaning of the word "accidental" as it would be understood by a typical policyholder, and was based almost entirely on a body of largely distinguishable district court cases. Further, Zurich's estimation of the foreseeability of Mr. Kovach's injuries is undermined by statistics showing that the likelihood of a wreck for an individual driving at less than extreme levels of intoxication is exceedingly small as a percentage of all such driving expeditions, and arguably represents Zurich's improper insertion of moral judgments into its analysis of the Plan's provisions. This observation is born out by the fact that insureds would almost certainly expect to be covered for injuries resulting from many other activities that are at least as risky as driving while intoxicated, such as driving while text messaging or while sleep-deprived.

In sum, Mr. Kovach's injuries were not "highly likely to occur" as a result of his intoxication, in contrast to the injuries that were highly likely to occur under the facts in Lennon. Zurich's interpretation of the Plan's provisions amounts to an additional,
unwritten exclusion for all drunk-driving injuries, which is not permitted under even the most deferential standard of review. For all of these reasons, we conclude that Zurich's denial of coverage based on Mr. Kovach's injuries being nonaccidental was unreasonable in light of the Plan's provisions, and thus arbitrary and capricious. See Morrison v. Marsh & McLennan Cos., 439 F.3d 295, 300 (6th Cir. 2006).

F. The self-inflicted-wound exclusion

The alternate explanation offered by Zurich for its denial of coverage to the Kovaches was that Mr. Kovach's injuries fell under the Plan's self-inflicted-wound provision. The relevant language provides that no benefits are payable if the loss is "caused by or connected with . . . a purposeful self-inflicted wound." Zurich reasons that Mr. Kovach's injury resulted from his intentional act of becoming drunk and operating a motorcycle, and was therefore self-inflicted.

The Eighth Circuit, sitting en banc, rejected an identical argument in King v. Hartford Life & Accident Insurance Co., 414 F.3d 994 (8th Cir. 2004) (en banc). There, the insured had died in a motorcycle accident with a BAC of .19. Id. at 997. That court
held, and we agree, that "[t]he most natural reading of the exclusion for injuries contributed to by ‘intentionally self-inflicted injury . . .' does not include injuries that were unintended by the participant, but which were contributed to by alcohol
intoxication." Id. at 1004. Although Mr. Kovach acted intentionally in drinking to excess and then riding his motorcycle, nothing in the record indicates that he did so with a mind towards harming himself. Zurich's interpretation of the exclusion in question conflates intentional actions with intentional results. Mr. Kovach's injuries, in other words, were the result of the collision, not simply a consequence of his acts of drinking and driving. After all, "[o]ne rarely thinks of a drunk driver who arrives home safely as an ‘injured' party." Id. Mr Kovach's intoxication likely contributed to the collision, but to define his excessive drinking as a "purposely self-inflicted wound" would be an illogical and "startling construction." Id. Zurich's denial of the Kovaches' claim on this basis was therefore arbitrary and capricious.

G. Appropriate proceedings on remand

This leaves us with a choice as to the appropriate proceedings on remand. In ERISA cases, appellate courts have the power either to send the matter back to the plan administrator for further consideration or to make a final decision on the merits. Cook v. Liberty Life Assurance Co., 320 F.3d 11, 24 (1st Cir. 2003) ("Once a court finds that an administrator has acted arbitrarily and capriciously in denying a claim for benefits, the court can either remand the case to the administrator for a renewed evaluation of the claimant's case, or it can award a retroactive reinstatement of benefits."). This court has frequently awarded benefits outright in lieu of a remand for reconsideration by the plan administrator. See, e.g., Cooper v. Life Ins. Co. of N. Am., 486 F.3d 157, 171 (6th Cir. 2007) ("[W]e find no need to remand this matter for additional consideration by LINA
because of our conclusion that Cooper has clearly established that she is disabled under the Plan."); Kalish v. Liberty Mut./Liberty Life Assurance Co., 419 F.3d 501, 513 (6th Cir. 2005) (concluding that the appropriate remedy was an immediate award of benefits rather than a remand to allow the plan administrator to consider evidence that it had previously ignored).

The case before us has no unresolved factual issues; instead, its resolution revolves around the proper interpretation of the Plan provisions-a question of law that we have answered in the Kovaches' favor. Having determined that Mr. Kovach's
collision falls within the definition of "accidental" and is not subject to the self-inflicted wound exclusion, we thus conclude that the proper disposition of this case is an award of benefits to the Kovaches. See Canseco v. Constr. Laborers Pension Trust, 93 F.3d 600, 609 (9th Cir. 1996) (concluding that no remand is necessary where no new factual determinations remain).


For all of the reasons set forth above, we REVERSE the judgment of the district court and REMAND the case for the entry of a judgment in favor of the Kovaches.

Judge David McKeague disagreed and his dissent can be read here.