dissenting
My colleagues see ambiguity in this case; I do not. John Dowling’s complaint is simple: Union Pacific’s Pension Plan provides a straightforward method for calculating the pension benefits of disabled former employees that the Plan administrator didn’t follow. If Dowling is correct, no amount of deference can justify the administrator’s decision. I believe he is.
As the majority notes, the amount of a Plan Participant’s1 pension benefits depends on two key figures: 1) the Participant’s Qredited Service; and 2) his Final Average Compensation. Plan § 5.01(a) (setting out the “Benefit Formula”). Dowl-ing does not contest the Plan administrator’s calculation of his Credited Service. He argues only that the administrator miscalculated his Final Average Compensation and thus arrived at an incorrect pension benefit amount when he applied the Benefit Formula. Accordingly, this case turns on a single question: Should the Plan administrator have calculated Dowling’s Final Average Compensation by looking to the pay Dowling received during the ten years before he became disabled or the pay he received during the ten years before he retired? Section 2.35 makes clear that the answer is the former.
The Plan provides that “ ‘Final Average Compensation’ shall mean the average of the Participant’s monthly Compensation for the 36 consecutive calendar months of the highest Compensation within the 120-calendar month period immediately preceding ... the last date on which he is a Covered Employee .... ” Plan § 2.35 (underscore in original). Put even more simply, Final Average Compensation is the best three consecutive years of pay an employee received in the ten years before he ceased to be a “Covered Employee.” So, to answer our question above, we need to know whether Dowling’s last day as a Covered Employee was the day before he started his period of disability on February 1, 1997, or just before he formally retired on October 1, 2012.
Again the answer is simple: Dowling was no longer a Covered Employee once his disability (multiple sclerosis) caused him to leave permanently in 1997. Admittedly, arriving at this answer requires working through a few of the Plan’s key terms. This requires attention at each step, but the steps are not hard to follow.
All agree that Dowling became a Disabled Participant on February 1, 1997. See J.A. 37; Plan § 2.25. A Disabled Participant is a Participant who suffers from a Total Disability and has had a “Separation from Service due to such Total Disability.” Plan §§ 2.25, 2.26. The day these events occur is called the “Disability Date,” and no one disputes that Dowling’s Disability Date is February 1, 1997. Plan § 2.25; J.A. 37. A Separation from Service occurs when an “Employee[’s] ... Total Disability ... causes him to cease to be an Employee!.]” Plan § 2.67, so there’s no question Dowling was no longer an Employee after his Separation from Service on his Disability Date in 1997. And a Covered Employee must be, at the very least, an “Employee.” See Plan § 2.21(“ ‘Covered Employee’ shall mean each Employee in the employ of an Employer....”).
Paring things down, we’re left with the following: Dowling became a Disabled Participant when his Total Disability and Separation from Service terminated his status as an Employee; only an Employee may be a Covered Employee, so he ceased to be a Covered Employee at the same time; and all of this happened on February 1,1997.
With this information in hand, we can return to Section 2.35’s definition of Final Average Compensation. It instructs us to look to the “120-calendar month period immediately preceding ... the last date on which [the Participant] is a Covered Employee ... [,] ” Plan § 2.35 (emphasis added)—i.e., the ten years between February 1, 1987 and February 1, 1997. The Plan administrator failed to follow these instructions, looking instead to the ten years preceding October 1, 2012, when Dowling was no. longer working and had ceased to be a Covered Employee. Accordingly, the administrator’s calculation of Dowling’s Final Average Compensation was incorrect.
My colleagues, however, don’t see it this way. They focus on whether Dowling’s period of disability (from February 1997 to October 2012) counted as an “absence” per § 2.18(a)(3)(C). In my view, their construction of that word, though creative, is beside the point.
Section 2.18 provides that “for a period of absence immediately prior to which [a Participant] is a Covered Employee ... [, he] shall be deemed to have received Compensation at the base pay in effect for him” before the period of absence began. Plan § 2.18(a)(3)(C)(i). Because Dowling’s period of disability could arguably qualify as an “absence,” the contention goes, he is deemed to have received Compensation in the amount of his base pay from February 1,1997 to October 1, 2012.
But as I show above, whatever Compensation Dowling was deemed to have received after February 1997 per § 2.18 is irrelevant to the calculation of his Final Average Compensation. This is because § 2.35 tells us to look to the period “preceding ... the last date on which [the Participant] is a Covered Employee[,]”— February 1, 1997. Plan § 2.35 (emphasis added). While, Section 2.18 tells us what Dowling was “deemed” to have received after he left work in 1997, § 2.35 is clear that Final Average Compensation depends on what he was paid before he became disabled that year.2
Section 2.18 thus does not justify the Plan administrator’s calculation of Dowl-ing’s Final Average Compensation by looking to the period between October 2002 and October 2012. One justification Union Pacific offers (but on which the majority rightly declines to rely) lies' in § 6.05. Under that section, certain Plan provisions apply to a Disabled Participant “as if his Separation from Service occurred on the date he ceases to be a Disabled Partici-pante.]” Plan § 6.05. If this language applied to the sections relevant to calculating the Final Average Compensation, which all appear in the Plan’s Article II, the Plan administrator’s choice would be vindicated. But § 6.05 makes clear that it only applies to “the other provisions of this Article [ie., Article VI] (or Article VII)[,]” leaving unaffected the sections of Article II discussed above.
How my colleagues get around this is by providing a more imaginative explanation for why the Plan administrator was not bound by § 2.35: The Plan is silent on whether there exists a special rule for calculating the Final Average Compensation of Disabled Participants, so the Plan administrator was free to craft one. Maj. Op. at 247-48. To justify this innovative approach, they note that the Plan has special rules to calculate Disabled Participants’ years of Credited Service as well as the date their benefits become available that differ from the rules applicable to other Participants. Id. at 246-47. Thus they conclude that “[t]he plan ... leaves a gaping hole as to whether a default-and-exception pattern continues for calculation of the Final Average Compensation for disabled participants.” Id. at 248. I find this conclusion dubious for two reasons.
First, I see no gaping hole. The special rule for Disabled Participants the majority seeks is provided in the sections discussed above. For the typical employee, calculate ing the Final Average Compensation is easy because the ten years preceding retirement will be his last ten years of employment. Id. at 242-43. But this is not so for a Disabled Participant. His Separation from Service occurs not at retirement but on his Disability Date, see §§ 2.25, 2.26; formal retirement might come years later. For this reason, the Plan carefully describes when an employee who becomes disabled ceases to be a Covered Employee. See Plan §§ 2.21, 2.25, 2.26 & 2.67. Because the rule for disabled former employees is contained in the definitions of Disabled Participant, Covered Employee, and other terms discussed above, there is no need for an alternative definition of Final Average Compensation for Disabled Participants in § 2.35. All the work is done by the Plan’s definitions establishing who is a Covered Employee and when.
Second, my colleagues’ approach proves too much. In their view, the Plan failed to specify how to calculate a Disabled Participant’s Final Average Compensation. So rather than follow the default rule provided for all Participants, the Plan administrator was free to make one up. That can’t be right. If a Plan administrator may depart from a general rule whenever a more specific one might have been, but is not, provided, what’s to stop him from simply denying Disabled Participants their pensions entirely? Given that Dowling had not been working for over ten years, why deem him to have received any Compensation at all? Couldn’t the Plan administrator have decided that Dowling’s Final Average Compensation was zero dollars?
Presumably the response would be no, as such a rule would not be reasonable. And here, I believe, is the heart of the majority’s mistake. The Plan administrator credited Dowling with years of service during his period of disability and calculated his Final Average Compensation with respect to those same years. This, my colleagues believe, is a reasonable way to design a pension program: looking to the same years to calculate a Participant’s Credited Service and Final Average Compensation is “good policy.”3
But we are not asked to opine whether the administrator has imagined a reasonable way to allocate pension benefits. Instead, we must decide whether his calculation of Dowling’s Final Average Compensation was “reasonably consistent with [the Plan’s] unambiguous text[.]” See Fleisher v. Standard Ins. Co., 679 F.3d 116, 121 (3d Cir. 2012). It was not, and thus I respectfully dissent.
. A careful reader might ask why § 2.18(a)(3)(C) would provide a rule for Compensation "deemed” received during an absence when § 2.35 calculates Final Average Compensation on the basis of the Compensation received before a Disabled Participant stops working. The answer is that the question before us is only one of many addressed by the Plan’s 277 pages and 19 articles. Compensation deemed received by Disabled Participants per § 2.18(a)(3)(C) may be relevant to any number of other issues not before us. Indeed, the parties direct us to one: § 5.01(c)(2)(B) relies on the Compensation a Participant is deemed to have received during a period of disability to offset the Participant's pension benefits against his Social Security benefits. Plan § 5.01(c)(2)(B).
. Although it should not bear on the outcome of this appeal, I am also not convinced the administrator’s decision necessarily represented good policy. Despite the majority's skepticism, it makes sense that Disabled Participants continue to receive Credited Service even after their Final Average Compensation is fixed the day they leave work. A significant portion of Dowling's pre-disability compensation was incentive and merit pay. (His average total annual compensation from 1993 to 1995 was $365,848, but his annual base pay during that period was $208,000.) Once disabled, of course Dowling received no incentive or merit pay; instead, the administrator "deemed” him to have received only his base pay. Accordingly, if the Plan looked to any period of disability in order to calculate the Final Average Compensation, pension benefits would not reflect the true economic value of employees during their working years. Participants like Dowling who received a large portion of their compensation in the form of merit and incentive pay would be shortchanged. At the same time, because pension benefits often take many years to vest, if a Disabled Participant received no Credited Service during a period of disability, he could forfeit substantial pension benefits through no fault of his own. Thus it is reasonable to calculate a Disabled Participant’s Final Average Compensation based on what he earned when actually working and yet award him Credited Service during his disability to avoid a forfeiture of benefits.