Eckelberry v. ReliaStar Life Insurance

402 F. Supp. 2d 704, 36 Employee Benefits Cas. (BNA) 1449, 2005 U.S. Dist. LEXIS 30452, 2005 WL 3254500

• Court: District Court, S.D. West Virginia
• Decided: December 1, 2005
• Docket: 6:04 CV 01185
• Status: Published

Opinion

MEMORANDUM OPINION AND ORDER

GOODWIN, District Judge.

Pending before the court are cross motions for summary judgment. The court GRANTS the plaintiffs motion for summary judgment [Docket 16] and DENIES the defendant’s motion for summary judgment [Docket 13].

I. Background

Around 3:49 a.m. on March 19, 2004, Earl Eckelberry died in an automobile accident when he struck a tractor trailer parked on the side of the road. (R. at 8.) The presence of alcohol was confirmed by a toxicology report, which measured Mr. Eckelberry’s blood-alcohol level to be 0.15%, an amount 0.05% higher than the legal limit. (Id. at 9-12.) At the time of Mr. Eckelberry’s death, he was employed by Ames True Temper, Inc. in Parkers-burg, West Virginia. (Id. at 8.) Mr. Eckel-berry had accidental death and dismemberment (AD & D) insurance of $86,000 through his employer’s employee benefit plan. (Pl.’s Mot. for Summ. J. 1.) The defendant, ReliaStar Life Insurance Company, is the exclusive administrator of the AD & D plan. (Mem. Supp. PL’s Mot. for Summ. J. 3.) The plaintiff, Michele Eckel-berry, the beneficiary under her husband’s policy, filed a claim for death benefits with ReliaStar on March 31, 2004. (R. at 2-3.)

On May 18, 2004, ReliaStar sent a letter to the plaintiff denying the claim because her husband’s death was not an “accident” as defined by the policy. (Id. at 18-20.) The plaintiff appealed the decision, which ReliaStar’s ERISA Appeals Committee affirmed in a July 20, 2004 letter. (Id. at 25-28.) Following this final claim determination, the plaintiff brought suit in the Circuit Court of Wood County, West Virginia. ReliaStar removed the action pursuant to 28 U.S.C. § 1441(b) based on this court’s federal question jurisdiction. The parties have stipulated that ERISA governs the case. (Stipulations ¶ 1.) This opinion follows the court’s determination that the sole legal issue is whether the defendant properly denied the plaintiffs application for death benefits. (Scheduling Order, Jan. 19, 2005.)

II. Standard of Review

According to the plan, the defendant has discretionary authority to determine whether a claimant is entitled to benefits. Specifically, ReliaStar “has final discretionary authority to determine all questions of eligibility and status and to interpret and construe the terms of this policyfies) of insurance.” (Stipulations ¶ 2.) Because of this language, the court ordinarily would review ReliaStar’s decision to deny the plaintiffs application for abuse of discretion. (Stipulations ¶ 3); Baker v. Provident Life & Accident Ins. Co., 171 F.3d 939, 941 (4th Cir.1999) (citing Firestone Tire & Rubber Co. v. Bruch, 489 U.S. 101, 115, 109 S.Ct. 948, 103 L.Ed.2d 80 (1989)). Under that standard, the court will not “disturb any reasonable interpretation by the administrator.” Doe v. Group Hospitalization & Med. Servs., 3 F.3d 80, 85 (4th Cir.1993). As long as the administrator uses a deliberate, principled reasoning process and the decision is supported by substantial evidence, the decision will not be disturbed. Martin v. Am. Bancorporation Ret. Plan, 407 F.3d 643, 655 (4th Cir.2005); Elliott v. Sara Lee Corp., 190 F.3d 601, 605 (4th Cir.1999). When the plan administrator, however, also has a financial stake in the outcome of the benefits decision, “this deference will be lessened to the degree necessary to neutralize any untoward influence resulting from the conflict.” Doe, 3 F.3d at 87. Because ReliaStar both insures and administers the payment of benefits under the AD & D plan, its decision to deny benefits will be reviewed using the less deferential standard.

III. Analysis

A. The Reasonableness of the Interpretation

The AD & D plan provides that “ReliaStar Life pays this benefit if you lose your life ... due to an accident.” (R. at 50.) The plan defines an accident as “an unexpected and sudden event which the insured does not foresee.” (Id. at 65.) ReliaStar denied the plaintiffs claim on the grounds that her husband’s death was not an “accident” under the policy. In the letter denying the plaintiffs claim, ReliaS-tar explained that Mr. Eckelberry’s death was “not unexpected” and therefore was not an accident. (R. at 18-20.) When ReliaStar’s ERISA Appeals Board denied the plaintiffs appeal, the Board noted that Mr. Eckelberry “put himself in a position in which he should have known serious injury or death could occur.” (Id. at 25-28.)

The plaintiff does not argue that the defendant lacked substantial evidence to conclude Mr. Eckelberry was under the influence of alcohol. The only issue is whether ReliaStar’s interpretation of the plan’s definition of accident was reasonable. This court must examine whether the interpretation was the result of a principled reasoning process and whether it was supported by substantial evidence. In determining whether the plan administrator made a reasonable interpretation, the court can examine multiple factors, including but not limited to a review of (1) the language of the plan; (2) whether' the interpretation is at odds with ERISA’s requirements; (3) the reasoning process used by the administrator and the extent of the evidence used to make the determination; (4) whether the interpretation is consistent with the plan’s goals; (5) whether the interpretation is consistent with the rest of the plan’s language; and (6) the fiduciary’s motives and any potential conflict of interest. Booth v. Wal-Mart Stores, Inc., 201 F.3d 335, 342-43 (4th Cir.2000). Thoroughly examining these factors reveals that ReliaStar’s interpretation of the policy’s definition of “accident” was not reasonable.

1. The Language of the Plan

What is an accident? Courts and underwriters have attempted to answer this apparently simple question for a century and a half. Adam F. Scales, Man, God, and the Serbonian Bog: The Evolution of Accidental Death Insurance, 86 Iowa L. Rev. 173, 175 (2000). One court described the question as “one of the more philosophically complex simple questions” in the law. Fegan v. State Mut. Life Assurance Co., 945 F.Supp. 396, 399 (D.N.H.1996). Whenever an insurer derives a seemingly impenetrable definition, “the endless cup-chain of events has dipped down into the bottomless well of evolution and brought to the surface a new and unthought-of combination of circumstances, smashing the formula into little bits.” Scales, supra, at 190 (quoting Edson S. Lott, Accident Insurance, 26 Am. Acad. Pol. & Soc. Sci. Annuals 483, 483 (1905)). Attempts to create a perfect definition for an accident have led to a horde of cases, which to review properly would require the writing of a treatise. One scholar who courageously attempted this endeavor explained, “[t]he jumbled mass of precedent which has steadily accumulated on the judicial shores over, the past 150 years attests the Scylla or Charybdis 1 nature of the task.” Scales, supra, at 191.

In reality, few accident insurers attempt the creation of a definition for accident. 2 Id. at 192. “The occasional effort typically results in failure.” Id. at 192-93. In this case, ReliaStar defines “accident” and consequently attempts to navigate the treacherous strait that has doomed other accident insurers. According to ReliaStar’s policy, an accident is “an unexpected and sudden event which the insured does not foresee.” (R. at 50.) To qualify as an accident under this definition, something must be 1) unexpected, 2) constitute a sudden event, and 3) not be foreseen by the insured. According to the language of this definition, the third element of foreseeability is entirely subjective. No component of objective foreseeability is present in this definition. ReliaStar easily could have written the definition to include an objective component by defining accident as an unexpected, sudden event that the insured should not have foreseen. The drafters of the policy, however, chose to use the words “does not foresee” instead of “should not have foreseen.”

Was the crash that killed Mr. Eckelber-ry unexpected? According to the letter initially denying benefits, ReliaStar believes the accident was not unexpected. To say, however, that Mr. Eckelberry expected to crash his car defies logic. Even though he was intoxicated, Mr. Eckelberry expected to drive home safely, not to hit a parked tractor trailer. To sustain ReliaS-tar’s contention that Mr. Eckelberry’s crash was not unexpected, Mr. Eckelberry would have had to expect the crash. Arguing, however, that he expected the crash and crashed anyway is analogous to arguing Mr. Eckelberry committed suicide. There is no evidence in the record suggesting Mr. Eckelberry expected his life to end when he chose to drive his car.

Was the crash a sudden event? Mr. Eckelberry lost control of his car and hit a parked tractor trailer. It is beyond doubt that this event was sudden.

Was the crash foreseen by the insured? ReliaStar believes the crash was foreseen by the insured because Mr. Eckelberry was driving drunk. If he foresaw the accident, then how did the accident occur? Merriam-Webster defines “foresee” as “knowing beforehand.” If Mr. Eckelberry knew beforehand that he was going to crash his car into a parked tractor trailer and die as a result, he would have avoided the collision. There is no doubt that Mr. Eckelberry’s impaired condition could have contributed to the cause of the collision. One of the chief goals of accident insurance, however, is to protect insureds from the effects of their own acts. Providence Life Ins. & Inv. Co. v. Martin, 32 Md. 310, 1870 WL 3948, at *2 (1870); Scales, supra, at 243. Even if an accident results because of the insured’s own fault, the insured still expects to receive coverage. Scales, supra, at 247. Otherwise, insurers could deny almost any claim under any accident insurance policy on the grounds that the insured contributed to the resulting accident. Having accident insurance would become as useful as socks for fish.

The court finds that Mr. Eckelberr/s crash was unexpected, constituted a sudden event, and was- not foreseen by Mr. Eckelberry. Therefore, Mr. Eckelberry’s crash was an accident as defined by the clear-cut language of ReliaStar’s policy. ReliaStar, however, by reasoning that Mr. Eckelberry “put himself in a position in which he should have known serious injury or death could occur” reads an objective component of foreseeability into the definition. The actual definition in the policy does not contain this objective component. Because ReliaStar’s interpretation requires implying a requirement that is not specified by the plain, ordinary meaning of the words used in the plan’s definition, the first factor weighs in favor of a finding that ReliaStar’s interpretation of the plan was not reasonable.

2. ERISA’s Requirements

To consider whether the plan administrator’s interpretation conflicts with the substantive requirements of ERISA, the court must first determine whether the interpretation of the plan’s definition is consistent with the term’s plain, ordinary meaning. 3 There is near-universal agreement among federal courts that the word “accident” lacks an ordinary meaning. West v. Aetna Life Ins. Co., 171 F.Supp.2d 856, 880 (N.D.Iowa 2001). The First Circuit explained that describing what constitutes an accident is a “nietaphysical conundrum.” Wickman v. Northwestern Nat’l Ins. Co., 908 F.2d 1077, 1084 (1st Cir.1990). The seminal decision regarding the federal common-law meaning of “accident” is Wickman v. Northwestern National Insurance Company, which has been described as the “primary embodiment of federal common law in the area of accidental death benefits.” Bevans v. Iron Workers’ Tri-State Welfare Plan, 971 F.Supp. 357, 361 (C.D.Ill.1997). In Wickman, the ERISA plan’s accidental death provisions defined accident as “an unexpected, external, violent, and sudden event.” Wickman, 908 F.2d at 1080.

The Wickman court began its analysis by outlining the approach most early federal courts used to determine what constituted an accident. Id. at 1085. This approach distinguished accidental means from accidental results. Id. Courts employing this analysis required the means causing death to be completely unintentional if the insurance contract insured against “accidental means.” Id. If the means were in any way intentional, then the result was not an accident, even if the result was not specifically intended Id. Although this approach “may” appear simple enough on paper, 4 it has proven to be entirely artificial in practice. Saying Justice Cardozo was psychic when he predicted that adhering to this artificial distinction would “plunge this branch of law into a. Serbonian Bog” 5 is a vast understate ment. See Landress v. Phoenix Mut. Life Ins. Co., 291 U.S. 491, 499, 54 S.Ct. 461, 78 L.Ed. 934 (1934) (Cardozo, J., dissenting) (“When a man has died in such a way that his death is spoken of as an accident, he has died because of an accident, and hence by accidental means.”). The Wickman court traced the problems that courts have had with this distinction and rejected this approach in favor of an analysis that “safely circumvents the ‘Serbonian Bog.’ ” Wickman, 908 F.2d at 1086.

To determine whether a death was accidental, the court focused its analysis on the reasonable expectations of the insured when the accidental death policy was purchased. Id. at 1088. Although the court found that the actual, subjective expectations of the insured are important, it reasoned they should not guide the analysis exclusively because of the difficulty in ascertaining an insured’s actual expectations and because an insured’s expectations occasionally could be patently unreasonable. Id. at 1087-88. To accommodate these concerns, the analysis should proceed in two steps: First, attempt to determine the insured’s actual expectations. If there is insufficient evidence to determine the insured’s subjective expectations, then determine whether a reasonable person similarly situated 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 (emphasis added). Courts addressing whether deaths or injuries involving intoxicated drivers constitute accidents almost always state they are following Wickman’s analysis. These courts, however, routinely fail to apply the two-step test properly, either by skipping step one entirely or by misstating step two. 6 When reviewing a plan that defines “accident” not in accordance with its generally understood meaning, the court will use the Wickman two-step analysis to determine if a death or injury results from an accident. West, 171 F.Supp.2d at 886.

If ReliaStar’s interpretation of “accident” was based on reasonable foreseeability in a manner at odds with the Wickman analysis, then ReliaStar’s interpretation is at odds with the substantive requirements of ERISA. ReliaStar’s definition of accident definitely focuses on an event, not on an injury. The Wickman court explained that when an accident is defined in terms of an event, this definition “would prompt courts recognizing the distinction between ‘means’ and ‘results’ to look at the ‘means,’ because only the means can be termed an event.” Wickman, 908 F.2d at 1085. Only if a contract spoke of unexpected injury would a court look at the “results.” Id. In Wickman, the court identified the type of definition that embodied the antiquated accidental means/results distinction: “To constitute an accident under this [‘accidental means’] standard, the cause of the injury ... must be ‘unforeseen, unexpect ed, and unusual; happening or coming by chance without design, that is casual or fortuitous, as opposed to designed or intended.’ ” Id. (citing 10 Couch on Insurance 2d § 41:28, 40 (1982)). The definition ReliaStar uses is precisely the kind of definition that Wickman targeted as employing the artificial accidental means/results distinction. See West, 171 F.Supp.2d at 886 (finding a plan’s definition of accident as “an event which happens by chance, or fortuitously, without intention or design, and which is unexpected, unusual and unforeseen” to embody the accidental means test).

ReliaStar’s denial of Mrs. Eckelberry’s claim states that Mr. Eckelberry “put himself in a position in which he should have known serious injury or death could occur.” (R. at 25-28.) This analysis does not follow Wickman’s two-step test. Reli-aStar skipped step one entirely by failing to examine foreseeability through the eyes of the insured as the common law requires. Instead, ReliaStar examined foreseeability only through the eyes of ReliaStar. In the letter denying Mrs. Eckelberry’s appeal. ReliaStar explained, “We believe that an individual driving with a blood alcohol level above the legal limit knowingly puts themselves at risk for serious injury or death.” (Id.) This interpretation fails to comport with the federal common-law definition of “accident” as formulated by Wickman. 7 Therefore, this interpretation is at odds with the substantive requirements of ERISA. Accordingly, the second factor weighs in favor of a finding that ReliaS-tar’s interpretation of the plan was not reasonable. ■

3. ReliaStar’s Reasoning Process

As stated above, ReliaStar reasoned that because Mr. Eckelberry was intoxicated, his death was not an accident. 8 (R. at 18.) ReliaStar explains the decision not to provide coverage for Mr. Eckelberry’s death was made after reviewing the West Virginia Traffic Crash Report, the medical examiner’s report, the death certificate, the death claim form, and the insurance policy. (Id.) At no point does ReliaStar explain how it concluded that being intoxicated makes an automobile accident something other than an actual accident. ReliaStar does not identify any evidence suggesting that someone with a blood-alcohol content of 0.15% who decides to drive should foresee his death, nor is there any evidence identified to suggest that a death occurring under those circumstances is expected.

No evidence was.identified because the evidence clearly shows that a person who drives while intoxicated should not foresee his death as a result. Does driving while intoxicated increase a person’s chances of crashing and possibly dying? Absolutely, but so do many other actions that hinder a driver’s concentration, such as talking on a cell-phone, applying lipstick, trying to subdue screaming children, or attempting to locate the correct station on a satellite radio. Even though the person performing these actions increases the likelihood of dying in a crash, it defies reason to suggest that the resulting crash is not an accident. Likewise, although a person that gets behind the wheel of a car while intoxicated increases the likelihood of dying in a crash, it is illogical to suggest that the driver reasonably foresaw that result.

According to the National Highway Traffic Safety Administration, 17,419 persons died as a result of alcohol-related traffic deaths in 2002. 9 Although a high number, it is remarkably low when compared to the number of times a person impaired by alcohol got behind the wheel of a car during the same year. The website operated by Mothers Against Drunk Driving states there were more than 159 million alcohol-impaired trips taken during 2002. 10 Based upon these numbers, one out of every 9,128 alcohol-impaired trips results in a crash that causes a fatality. The court is unaware of any mathematical formula to calculate whether a result is reasonably foreseeable, but assumes that if a result has a l-in-9,128 chance of occurring, it is not reasonably foreseeable. Realizing that the number of alcohol-impaired trips taken is an estimate, assume that the number was overestimated by 100 million trips. 11 Even then, only one out of every 3,387 alcohol-impaired trips results in a fatality, a percentage still lower than the chances of being struck by lightning sometime during a person’s lifetime. 12 ReliaS-tar’s decision finding death reasonably foreseeable from driving while under the influence is not supported by substantial evidence and was not made through a principled reasoning process. Accordingly, the third factor weighs in favor of a finding that ReliaStar’s interpretation of the plan was not reasonable.

4. The Consistency of the Interpretation with the Goals of the Plan

The goal of the AD & D plan is to provide insurance benefits for an insured’s loss of life, limb, or sight due to an accident except in cases where an express exclusion is applicable. (R. at 50.) Provisions exist under the policy expressly excluding coverage for suicide, intentionally self-inflicted injury, physical or mental illness, bacterial infection, airplane crashes, armed conflict, injury suffered in military service, injury occurring in the commission of a felony, or injury caused by the use of any illegal drug, narcotic, or hallucinogenic agent. (Id. at 50-51.) Based upon ReliaStar’s correspondence with Mrs. Eckelberry, it appears that ReliaStar would deny coverage for any claim made by an insured who was drunk driving. There is no evidence suggesting that ReliaStar considers factors such as how high a driver’s blood-alcohol content was, how far the driver was attempting to travel, how familiar the driver was with the roads, or the weather conditions. Instead, ReliaStar simply denies coverage as soon as it determines the insured’s blood-alcohol content exceeded the legal limit.

By maintaining this blanket policy, Reli-aStar is reading another exclusion into the policy. If ReliaStar wants to exclude accidental death coverage for insureds who are driving while intoxicated, ReliaStar should provide an express exclusion for these situations. In fact, most accident insurance policies expressly exclude this situation. Instead, ReliaStar implies an exclusion not in the policy that frustrates the policy’s goal.

By excluding any claim based on an intoxicated driver’s death, it is easy to conceive of situations in which accidental death benefits are denied even though the driver failed to contribute at all to the resulting accident. For example, assume a person has a blood-alcohol level of .01 higher than the legal limit and gets in his car to drive two blocks home. While driving below the speed limit, he passes under a raised highway. As he begins to pass, a reckless tractor trailer barrels off the raised highway and smashes onto his vehicle, killing him instantly. The exact scenario would have occurred whether he was completely sober or legally intoxicated, yet under the reasoning that ReliaStar used in this ease, coverage would be denied because of his blood-alcohol level.

ReliaStar denied Mrs. Eckelberry’s claim for her husband’s death solely because he was driving while intoxicated. Although ReliaStar stated that Mr. Eckel-berry’s death was not unexpected as a result of drinking and driving, there is no evidence that death should' be expected when a person drinks and drives. As explained above, there is a major difference between a result being expected and a result’s chances of occurring being increased. Just because the chances of a result occurring are increased certainly does not necessarily mean that the result is to be expected.

By finding Mr. Eckelberry’s death “not unexpected,” ReliaStar opens the door to exclude coverage for any accident in which an insured engages in some activity that increases the chances of dying. Essentially, ReliaStar is taking the “accident” out of accident. Assume a tourist is walking along the rim of the Grand Canyon. There are two paved trails available for tourists, Trail 1 and Trail 2. Even though each trail is safely located at least ten feet from the edge, Trail 1 is a foot closer to the edge than Trail 2. The tourist, feeling adventurous, chooses Trail 1. A few steps into the trail, the tourist decides to take a picture of a passing ram but while trying to get the perfect angle on his digital camera, he slips on a piece of limestone, awkwardly rolls ten feet, and tumbles off the cliff to his death. .Obviously, the tourist increased his chances of dying when he decided to get out of his car and walk along a trail. He then further increased his chances of dying when he adventurously chose Trail! instead of Trail 2, which would have given him one additional foot of safety. Is the tourist’s death an accident? Yes. Is the tourist’s death an accident under ReliaStar’s definition of accident? Yes, for the tourist’s death was unexpected and he did not foresee falling to his death when he decided to take a picture of the passing ram. Under the reasoning ReliaStar used in Mr. Eckelber-ry’s case, however, ReliaStar could deny coverage because the tourist’s decisions to get out of his car and walk Trail 1 increased his chances of dying. 13

Alternatively, assume a middle-aged woman decides to unwind after a hard day’s work by reading Tolstoy’s literary classic War and Peace. To relax the emotional roller-coaster that the passionate relationship between Natasha Rostova and Pierre Bezukhov generates, she decides to read while in her hot tub. Unfortunately, somewhere around the end of Chapter XXXIV of Book Eleven she dozes off. As a result, she drowns and dies. An accident? Yes. An accident under ReliaStar’s definition of accident? Yes. Under ReliaS-tar’s narrow reading of the term, however, ReliaStar could deny coverage because she increased her chances of dying by deciding to read War and Peace while in her hot tub. 14

Almost any accident imaginable involves a degree of negligence or a decision that somehow increases the chance of the accident’s occurrence. See Scales, supra, at 298 (arguing that many ERISA courts “evidently do not understand that virtually every car crash may be traced to some failure of judgment that fully reveals its dangers only when it is too late. That is precisely why they are accidents.”). The fact that negligence or a decision that makes an accident more likely to occur exists does not mean that once the accident occurs, that it was not really an accident. “[Ijnsureds expect coverage even for accidents that are their own fault.” Id. at 247. If a person gets in his car and begins driving home on icy roads, it is obvious that sthe chances of an accident occurring are enhanced. This increased probability, however, does not in turn make the resulting crash anything other than an accident. Likewise, Mr. Eckelber-ry obviously made a bad decision when he got in his car to drive home, but the fact he was intoxicated does not mean that the resulting car crash was expected or foreseen. Even though he increased the probability of an accident occurring, the crash was still an accident. ReliaStar’s interpretation of accident frustrates the goals of the policy and sets up a slippery slope in which ReliaStar could deny coverage for any alleged “accident.” Accordingly, the fourth factor weighs heavily in favor of a finding that ReliaStar’s interpretation of the plan was not reasonable.

5. Internal Inconsistencies in the Plan

This factor concerns whether ReliaStar’s interpretation renders any other language in the plan meaningless or internally inconsistent. Finley v. Special Agents Mut. Benefit Ass’n, Inc., 957 F.2d 617, 621 (8th Cir.1992). By interpreting the definition of accident to include any foreseeable risk of. death, the interpretation renders meaningless most of the express exclusions in the policy. For example, why have an explicit exclusion for suicidal behavior if ReliaStar is going to find that conduct carrying a foreseeable risk of death is not an accident? The same can be said for the exclusions relating to intentionally self-inflicted injury, armed conflict, and the use of illegal drugs. The reason for the list of exclusions is to cover situations in which the conduct would normally constitute an accident but is not treated as such under the policy. The insured can read the list of exclusions when purchasing the policy to see what it covers and what is excluded. By reading in an exclusion not expressly listed, the policy becomes internally inconsistent. Although this factor weighs against ReliaStar, the court finds that it is largely just an extension of the fourth factor and is not independently helpful to the analysis.

6. Fiduciary’s Motives and Potential Conflict of Interest

As stated above when considering the level of deference given to ReliaStar’s interpretation, ReliaStar served as both the insurer and the entity' that decides whether to pay benefits on asserted claims. Naturally, this creates a conflict of interest as ReliaStar will have to lose money on any claims it considers viable. When Reli-aStar denies coverage on claims, such as the claim in this case, ReliaStar saves money. Because ReliaStar’s interpretation of accident in this case prevented the company from having to pay money on a claim, the interpretation positively affected ReliaStar. Like the fifth factor, however, although it weighs against ReliaStar, it is not extremely helpful to the analysis.

7. The Weighing of the Factors

Carefully reviewing the factors necessary to determine whether ReliaStar’s interpretation of the plan’s definition of accident was reasonable, the court concludes that it was not. The interpretation was at odds with the ordinary meaning of the words that ReliaStar used to define accident; it was at odds with ERISA’s substantive provisions as developed through federal common law; the reasoning process used to arrive at the decision was not supported by substantial evidence; and the interpretation was inconsistent with the goals of the AD & D plan. ReliaStar’s interpretation of the definition it uses to define accident was unreasonable in this case. Mr. Eckelberry died in an accident and his family should be entitled to accidental death benefits. '

B. Common Sense

The list of factors the court must consider in determining whether ReliaStar’s interpretation was reasonable is nonexhaustive. See Booth, 201 F.3d at 342-43 (“a court may consider, but is not limited to, such factors”). The court finds that more important than any of the factors already considered is genuine common sense. 15 It defies reason to argue that Mr. Eckelberry expected to die when he got behind the wheel of his car. Likewise, it defies reason to argue that he foresaw his death, or even that he should have foreseen it.

The defendant contends it is well settled in the Fourth- Circuit that an intoxicated driver’s death can never be an accident. (Def.’s Mem. Supp. Mot. for Summ. J. 1.) The Fourth Circuit discussed this issue in Baker v. Provident Life & Accident Insurance Company. In that case, Thomas Baker was denied health care benefits by Provident Life and Accident Insurance Company. Baker, 171 F.3d at 940. Mr. Baker, who had a blood-alcohol content of 0.286, attempted to pass a car by crossing over the double yellow line on the highway. Id. at 941. When attempting the pass, he crashed into an oncoming vehicle and killed that vehicle’s driver. Id. Mr. Baker pled guilty to involuntary manslaughter, which was a felony under North Carolina law. Id. His health care benefit plan explicitly excluded coverage for injuries sustained in the voluntary participation of a felony. Id. The court held that because Mr. Baker’s drinking and driving were voluntary, so was his commission of involuntary manslaughter. Id. As a result, the court affirmed a grant of summary judgment in favor of the insurance company. Id.

The court analogized the issue of Mr. Baker’s voluntariness with accidental death benefits cases. Id. at 942. The court first looked to North Carolina state law, but the case it reviewed was decided in 1957 and embodies the accidental means/result distinction that has been criticized heavily by other federal courts. See Allred v. Prudential Ins. Co. of Am., 247 N.C. 105, 100 S.E.2d 226 (1957) (“nor is it produced by accidental means”).

The Baker court then looked to other federal courts that have dealt with drunk driving accidents. Id. It stated that a majority of federal courts have found that an intoxicated driver’s death is not an accident. Id. The court then used this rule and analogized it to Mr. Baker’s voluntariness. Id. The federal cases relied on by Baker, however, are those cases cited in footnote six of this opinion that assert they are following the two-step test laid out in Wickman, but actually skip step one entirely or misstate the second step. 16 At no point does the Baker court actually step back and conduct a Wickman analysis of its own to determine whether an intoxicated driver’s death is an accident, for this analysis would have been misplaced in Baker as the issue was not critical to the holding. 17 Because any discussion of this issue in Baker was dicta, this court is not bound by that discussion.

When Wickman’s two-step test 18 is applied to the case at bar, it becomes obvious that Mr. Eckelberry’s death was an accident. First, the court must attempt to determine the insured’s actual expectations. This step is completely ignored by most federal courts, including those cases that the Baker court cited. In this case, it is simple to determine the insured’s actual expectations. Undoubtedly, Mr. Eckelberry expected to drive home. His blood-alcohol level was 0.15 and he did not fore see himself dying on the way. 19 If there is insufficient evidence to determine the subjective expectations of the insured, only then should the court consider whether a reasonable person similarly situated to the insured “would have viewed the injury as highly likely to occur as a result of the insured’s intentional conduct.” Wickman, 908 F.2d at 1088 (emphasis added). Reviewing the evidence regarding the odds of dying when drinking and driving, it cannot be said that a reasonable person can view drinking and driving as reasonably foreseeable, let alone “highly likely,” to cause death. Under the Wickman analysis, intoxicated drivers’ deaths are accidents. See Metropolitan Life Ins. Co. v. Potter, 992 F.Supp. 717, 730 (D.N.J.1998) (“The court cannot find as a matter of law that death is ‘highly likely to occur’ as a result of drunk driving and that ‘any other expectation would be unreasonable’ where many drunken drivers survive to be prosecuted or perhaps repeat their risky conduct.”).

To hold otherwise is to impute society’s moral concerns about drunk drivers ahead of sound reason. See Scales, supra, at 298 n. 531 (asserting courts that find intoxicated driver deaths not to be accidents interpret the term “accident” to embrace “an unwritten morals clause”). The true definition of accident should not be discarded depending on whether the conduct involved is morally condemned. Professor Adam F. Scales, Associate Professor of Law at Washington and Lee University, explains the problem well: “Drunk driving accidents are simply more senseless, more unforgivable, and altogether more deserving of moral disapprobation than other accidents. But they are accidents nonetheless.” Id. at 299. Automobile crashes, regardless of how they are caused, whether by icy road conditions or intoxicated drivers, are accidents. 20 “ ‘The common speech of men’ is both unanimous and correct in describing such crashes as accidental. To hold otherwise is to sanction the deception of the public.” Id. at 302.

IV. Conclusion

ReliaStar’s interpretation of the definition of accident as it appears in the AD & D plan issued to Mr. Eckelberry was not reasonable. As a result, the court finds that Mrs. Eckelberry’s claim for death benefits was not properly denied. Accordingly, the court GRANTS the plaintiffs motion for summary judgment [Docket 16] and DENIES the defendant’s motion for summary judgment [Docket 13].

The court DIRECTS the Clerk to send a copy of this Order to counsel of record and any unrepresented party.

1 . In Greek mythology, the monster Scylla and the whirlpool Charybdis occupy opposite sides of a treacherous strait. Scylla has "twelve mis-shapen feet, and six necks of the most prodigious length; and at the end of each neck she has a frightful head with three rows of teeth in each, all set very close together, so that they would crunch anyone to death in a moment .... No ship ever yet got past her without losing some men, for she shoots out all her heads at once, and carries off a man in each mouth.” Homer, The Odyssey, Book XII. Charybdis, however, is such a powerful whirlpool that ships are forced to hug the side of the strait where Scylla lurks. Id. Ulysses himself lost his "six best men” to the dreadful monster. Id.

2 . According to Professor Scales, "[ajccidents exist in the region between two diametrically opposed phenomena. At one end of the continuum lie events that the law has historically, and confusingly, referred to as 'acts of God.' At the other are events resulting from the deliberate and effective exercise of human will. An accident occurs when the gravitational pull of neither pole is strong enough to dominate.” Scales, supra, at 269.

3 . In the previous section, the court examined whether the interpretation was at odds with the actual words used in ReliaStar's definition. This section focuses on whether the definition used in ReliaStar’s plan is consistent with the true definition of an accident.

4 . This court does not believe that the distinction is clear on paper either; hence the qualifier “may.”

5 .A “Serbonian Bog” was believed to be a lake between Egypt and Palestine in ancient times. According to lore, blowing sand often disguised the lake as a solid-land mass, which lead to the deaths of many unfortunate travelers. In Paradise Lost, Milton exclaims, "A gulf profound as that Serbonian Bog, betwixt Damiata and Mount Cassius old, where armies whole have been sunk.” The bog sup posedly received its name from Sebaket-Bar-doil, a former king of Jerusalem who died in the bog on a return from Egypt. In modern parlance, the phrase “Serbonian Bog” is used to refer to a "mess from which there is no way of extricating oneself.” E. Cobham Brewer, The Dictionary of Phrase and Fable (1894).

6 . For examples of cases that have found intoxicated drivers’ deaths not to be accidents, but have misapplied Wickman in the process, see Jones v. Metropolitan Life Ins. Co., 385 F.3d 654, 656 (6th Cir.2004); Schadler v. Anthem Life Ins. Co., 147 F.3d 388, 397 n. 10 (5th Cir.1998); Cozzie v. Metropolitan Life Ins. Co., 140 F.3d 1104, 1110 (7th Cir.1998); Clark v. Metropolitan Life Ins. Co., 369 F.Supp.2d 770, 777 (E.D.Va.2005); Mullaney v. Aetna U.S. Healthcare, 103 F.Supp.2d 486, 492-94 (D.R.I.2000); Metropolitan Life Ins. Co. v. Potter, 992 F.Supp. 717, 728-29 (D.N.J.1998); Walker v. Metropolitan Life Ins. Co., 24 F.Supp.2d 775, 780-82 (E.D.Mich.1997); Lincoln Nat'l Life Ins. Co. v. Evans, 943 F.Supp. 564, 568-71 (D.Md.1996).

7 . Although the court must look to federal common law when analyzing the terms of an ERISA plan, the court may look to state law for guidance to the extent state law does not conflict with ERISA or its policies. Griggs v. E. I. DuPont de Nemours & Co., 385 F.3d 440, 441 n. 4 (4th Cir.2004); Singer v. Black & Decker Corp., 964 F.2d 1449, 1452 (4th Cir.1992). West Virginia courts, however, continue to recognize the distinction between accidental means and accidental results'. Floyd v. The Equitable Life Assurance Soc’y of the United States, 164 W.Va. 661, 264 S.E.2d 648, 650 (1980). Because the state law approach is inconsistent with the federal approach, state law in this case does not provide further guidance on whether ReliaStar's interpretation was reasonable.

8 . It is difficult to imagine how an insurance company can reasonably believe that Mr. Eck-elberry’s death was not a true accident. One scholar points out that insurance companies should just acknowledge the fact that drunk driving deaths are accidents, yet are caused by "particularly stupid” behavior. Scales, supra, at 301. This acknowledgment would avoid "the tortured jujitsu necessary to characterize such events as nonaccidents.” Id.

9 .Alcohol Alert, 2002 Drunk Driving Statistics, http://www.alcoh.ol— alert.com/drunk-driv ing-statistics-2002 .htm (last visited Nov. 28, 2005). The court notes that this number includes fatalities that resulted from crashes in which any driver had a blood-alcohol content of more than 0.01, which means that the number of fatalities from crashes in which a driver was legally intoxicated actually would be lower. Of these 17,419 fatalities, only 8,474 were to drivers. Michael E. Gardner, Accidental Death Insurance Coverage of Drunk Drivers, 69 Mo. L. Rev. 235, 249 (2004).

10 . Mothers Against Drunk Driving, Stats & Resources, http://madd.com/stats/ (last visited Nov. 28, 2005).

11 . Obviously, suggesting that this number was overestimated by 100 million trips is an extreme example.

12 . National Weather Service, Lightning Safety, http://www.lightning safety.noaa.gov/ med ical.htm (last visited Nov. 28, 2005) (estimating the chances of a person being struck by lightning sometime in an eighty-year lifetime is 3,000-to-l).

13 . A district judge in Georgia recently wrote the following in an opinion when questioning an insurer’s argument that any increase of hazard beyond "normal” behavior was not accidental: "[I] enjoy[] standing in the middle of big rivers casting bits of hair and fur at fish. This certainly increases the risk of drowning as compared with staying at home and sitting on the couch. Is this not normal behavior? Is the judge's widow to be deprived of protection because of his piscatorial pursuits?” Buce v. Nat'l Serv. Indus., Inc., 74 F.Supp.2d 1272, 1278 (N.D.Ga.1999).

14 . Despite this example, the court is sure that many people find Tolstoy's extravagant descriptions of wolf hunts and grand balls riveting.

15 . After all, according to John Rawls, a "liberal democracy, in order to meet its definitional requirements of placing sovereignty in free and equal citizens, must base its policies on reasons that are equally available to all citizens, whatever their underlying beliefs.” Suzanna Sherry, The Sleep of Reason, 84 Geo. L.J. 453, 471 (1996) (discussing John Rawls' Political Liberalism).

16 .See Michael E. Gardner, Accidental Death Insurance Coverage of Drunk Drivers, 69 Mo. L. Rev. 235, 245-46 (2004) ("A recent decision, West v. Aetna Life Insurance Co., expounded on ... the way federal courts have applied the Wickman test in the drunk driving context. The West court criticized other courts' reliance on blood alcohol content levels to conclude that death is 'highly likely' to result from driving drunk. The court acknowledged that these things show that drunk driving is ‘irresponsible and dangerous,' but maintained that driving drunk is not 'highly likely' to result in the driver's death.''); see also Scales, supra, at 297 ("[S]everal ERISA decisions have misused Wickman's quasi-objective test in ways that may appeal greatly to moralists, but that advance not at all the task of determining whether an accident has occurred.”).

17 . Not having thoroughly analyzed an issue hardly makes it well settled. In Poeppel v. Hartford Insurance Company, the Fourth Circuit issued an unpublished opinion erroneously stating that Baker "adopted the rule” that an intoxicated driver's death cannot be an accident. Poeppel v. Hartford Ins. Co., 87 Fed.Appx. 885, 2004 WL 298414, *1 (4th Cir.2004) (unpublished). To the extent that the unpublished Poeppel decision finds this issue well settled, this court disagrees. This court also notes that unpublished opinions are not binding authority. United States v. Ruhe, 191 F.3d 376, 392 (4th Cir.1999).

18 . The Baker court actually cites Wickman on two occasions. Baker, 171 F.3d at 942.

19 . In Baker, Mr. Baker’s blood-alcohol content was 0.286, which obviously increased the chances for a fatality dramatically more than Mr. Eckelberiy's case.

20 . Unless, of course, there is an intent to commit suicide present.