While working at Regal-Beloit Corporation, William Davis participated in the company's long-term disability plan, which is administered by Unum Life Insurance Company of America. After Regal-Beloit fired Davis, he sought benefits under the plan. Unum awarded Davis benefits for a mental disability but denied benefits for a physical disability. Davis sued. The district court denied summary judgment to Unum and the plan and partially granted summary judgment to Davis, remanding the matter for further administrative proceedings. Dissatisfied with his partial victory, Davis appealed. Unum and the plan then cross-appealed. Concluding that the district court misapplied the governing arbitrary-and-capricious standard, we reverse and remand, with instructions for the district court to enter judgment in favor of Unum and the plan.

Regal-Beloit Corporation, an industrial machinery manufacturer, hired William Davis as a regional sales manager in February 1999. In January 2000, Regal-Beloit terminated him due to poor job performance. He was fifty-five at the time. During his brief term, Davis participated in the Regal-Beloit Corporation Long Term Disability Plan ("the plan"). The plan's administrator is Unum Life Insurance Company of America.

Under this plan, the maximum period for which a claimant can receive disability benefits varies based upon a number of factors, including whether the disability is physical or mental in nature. Generally, for physical disabilities arising before age sixty, benefit payments continue until the claimant turns sixty-five. To reach that maximum, the claimant must meet the plan's definition of "disabled" throughout the period. During the first twenty-four months of benefit payments, the term disabled means that a sickness or injury limits the claimant's performance of his "regular occupation." After twenty-four months of payments, however, the standard becomes more difficult for the claimant to meet. The plan then defines disabled as being "unable to perform the duties of any gainful occupation for which [the claimant is] reasonably fitted by education, training or experience." Further, separate from these provisions, there is a different time limitation for "disabilities due to mental illness." These benefits are capped at twenty-four months.

After Davis's termination, he searched for new positions and landed several interviews. Nevertheless, in July 2000, some six months after leaving Regal-Beloit, Davis remained unemployed and submitted a disability claim under the plan. On the initial form, Davis claimed to be disabled due to severe depression, memory loss, joint pain, and a stroke. He further stated that his condition prevented him from driving, lifting, and walking.

In a follow-up telephone interview, nonetheless, Davis stated that he was not restricted or limited as the result of any physical ailment; rather, he said the only reason that he was unable to work was his mental condition. According to Davis, he could not work because of depression, mood swings, forgetfulness, an inability to concentrate, and an inability to perform multiple tasks at the same time. Davis said he was homebound, adding that he did not drive at all, not even to the local market. Surveillance evidence obtained less than a month after this interview, however, showed Davis driving vehicles, refueling a vehicle, and engaging in routine outdoor chores.

Davis's doctors backed up his claims of a disabling mental illness. According to Davis's attending physician and internist, Alan Reich, M.D., the primary diagnosis was major depression. On the initial form submitted with Davis's claim, Dr. Reich listed the symptoms as indecisiveness, forgetfulness, and difficulty concentrating. Dr. Reich also noted that Davis had no cardiac limitations but had some back and hip pain. This summary from Dr. Reich came after several other doctors had examined Davis in 2000. These doctors largely gave Davis a clean bill of physical health, but some noted psychiatric problems. For instance, Davis's urologist, Gordon Gluckman, M.D., found no recurrence of Davis's 1992 bout with prostate cancer and said Davis was "doing well." Further, Davis's neurologist, Barry Levy, M.D., in conjunction with a psychologist, Joshua Barras, Ph.D., found that depression was the source of Davis's problems and recommended psychiatric treatment. It is worth noting that, according to Dr. Levy, Davis was "very unhappy" with Drs. Levy's and Barras's opinions. Further, while Davis's disability claim was pending, Davis believed that he had a stroke, but Drs. Reich and Levy had reasons to question that belief. Particularly, Dr. Levy, in discussing the "suspected stroke," stated that "nonphysiologic features are present and suggest a psychogenic component." In other words, Davis's problem was not physical but mental.

For its part, Unum responded to the claim by consulting Robert Buchanan, M.D., a psychiatrist who worked at Unum — i.e., an "in-house doctor." The importance of this phrase will become apparent below. After reviewing Davis's medical file, Dr. Buchanan generally concurred with the findings of Davis's doctors, stating that Davis had "significant impairment from depression." In December 2000, Unum then approved Davis's disability claim pursuant to the plan's mental disability coverage. Unum paid Davis benefits retroactive to July 2000 and continued to pay him benefits until Davis reached the plan's twenty-four month limitation for mental disability benefits in July 2002.

Meanwhile, Davis began seeing a new primary doctor in February 2001, Michael Raymond, M.D. Dr. Raymond, an internist, immediately keyed in on Davis's physical complaints, including problems walking and standing. Dr. Raymond also referred Davis to a new neurologist, Steven Meyers, M.D. Dr. Meyers found that Davis "cannot stand or walk for any distance" and concluded that Davis was "permanently disabled" due to lumbar spinal stenosis (a narrowing of the spinal canal in the lower back) and the aftereffects of childhood polio. However, according to Dr. Meyers, diagnostic testing revealed that Davis did not actually have a condition known as post-polio syndrome (progressive muscle weakness, muscle and joint pain, fatigue, etc.).

In May 2001, Davis then supplemented his existing file at Unum with the opinions of Drs. Raymond and Meyers. Davis thus began a sustained effort to obtain the more generous (i.e., longer lasting) physical disability benefits under the plan. Unum then had an orthopedic surgeon, Joseph Thomas, M.D., and an internist, Steven Feagin, M.D., review Davis's updated file. Both worked in-house at Unum. Based upon their assessments of the medical evidence, Unum notified Davis in December 2001 that his eligibility remained unchanged and that his benefits were still due to expire in July 2002.

Davis then retained counsel and submitted more information from Dr. Raymond related to spinal stenosis and other physical ailments. Despite Dr. Meyers ruling out post-polio syndrome, Dr. Raymond's papers still listed post-polio syndrome as a reason to support Davis's disability claim. Dr. Raymond further concluded Davis should be restricted from walking, lifting, bending, carrying, and climbing. It is noteworthy, however, that in this submission, Dr. Raymond stated that Davis had no functional limitations as the result of a cardiac condition.

On Unum's behalf, Dr. Thomas reviewed the new data and found that it supported a diagnosis of spinal stenosis. However, Dr. Thomas concluded that, although Davis could not stand or walk for prolonged periods, he was still able to perform sedentary work. George Fluter, M.D., a doctor of physical medicine and rehabilitation at Unum, concurred, stating that Davis was not precluded from sedentary activities. Dr. Feagin also revisited the file and found no limitations based upon Davis's diabetes, hypertension, sleep apnea, cardiac condition, and other physical complaints, all of which were successfully treated or being treated. Additionally, according to David Goldsmith, Ph.D., a clinical psychologist at Unum, Davis's depressive disorder remained and his file indicated that Davis's "perceived" cognitive problems were only psychiatric in nature. Finally, Kelly Marsiano, M.Ed., a vocational rehabilitation specialist at Unum, identified several sedentary occupations that Davis could, in light of his physical capacity, educational background (two years of college), and employment history, perform. Unum then determined that Davis failed to meet the standard for physical disability benefits beyond twenty-four months — that is, he was not unable to perform the duties of any gainful occupation for which he was reasonably fitted. Unum thus denied his physical disability claim in December 2002.

Davis appealed, supplementing his claim with a letter from Dr. Raymond. In the letter, Dr. Raymond opined that Davis was incapable of active employment, not even a sedentary position. To this end, Dr. Raymond added carpal tunnel syndrome (previously unmentioned) to Davis's growing list of physical complaints, and, despite earlier findings in the record, Dr. Raymond's list surprisingly still included prostate cancer, post-polio syndrome, and coronary artery disease. According to Dr. Raymond, it was "difficult, if not impossible, for Mr. Davis to sit, stand or walk for any prolonged timeframes [sic]." Dr. Raymond also cited to a hypoglycemic episode (decrease of blood sugar) related to Davis's diabetes, which necessitated hospitalization in February 2003. Dr. Raymond, however, did not attach any new medical data to support the assertions in his letter. Dr. Fluter reviewed the letter, noting the obvious absence of supporting documentation, and consequently held to his opinion that Davis could engage in sedentary work. Unum then affirmed its denial but offered Davis the opportunity to submit additional records.

Davis then supplied Unum with the hospital records of his hypoglycemic attack. The hospital records revealed that, the evening before the incident, Davis forgot to take his insulin at the scheduled time. In the morning, he then took two full doses of insulin. In addition, he did not eat breakfast. Dr. Fluter found that this occurrence of hypoglycemia was associated with the confluence of these missteps. Dr. Fluter concluded: "It would be unlikely for [Davis] to experience significant hypoglycemia with attention to proper dosing and administration of his diabetic treatment medications and to proper dietary management. This occurrence would not preclude performance of sedentary level activities." With that, the administrative case was closed. Unum issued its final decision, upholding its denial of physical disability benefits, in August 2003.

Within days, Davis sued Unum and the plan. The suit, brought under the Employee Retirement Income Security Act ("ERISA"), alleged wrongful denial of benefits pursuant to 29 U.S.C. § 1132(a)(1)(B). The parties filed cross-motions for summary judgment. The district court denied Unum and the plan summary judgment and partially granted Davis summary judgment. The district court faulted Unum for using only in-house doctors who perform "a mere paper review" of Davis's claims and who did not explain their conclusions to the district court's satisfaction. The district court, nevertheless, did not order Unum to award Davis physical disability benefits. Rather, the district court remanded the case back for Unum to correct the deficiencies identified by the district court in its opinion.1 Displeased, Davis appealed, seeking an outright reversal of Unum's denial and a full award of physical disability benefits. Unum and the plan then cross-appealed, contesting the district court's denial of summary judgment.

Our review of the district court's summary judgment decisions is de novo. See Sisto v. Ameritech Sickness & Accident Disability Benefit Plan, 429 F.3d 698, 700 (7th Cir.2005). Before addressing the denial of Davis's claim, we first turn to the standard under which that denial should be reviewed.

When, as here, the terms of an employee benefit plan afford the plan administrator broad discretion to interpret the plan and determine benefit eligibility, judicial review of the administrator's decision to deny benefits is limited to the arbitrary-and-capricious standard. See id. The district court ostensibly applied this deferential standard, but Davis challenges its applicability in this case. He, however, does not dispute the plan's language. Rather, he contends that, due to a conflict of interest, Unum was biased against him and should not be afforded any deference. This is a difficult road for Davis because the existence of potential bias, a potential conflict, is not enough to dislodge our ordinary arbitrary-and-capricious review. See Dougherty v. Ind. Bell Tel. Co., 440 F.3d 910, 2006 WL 647719, at *5 (7th Cir. Mar.16, 2006); Rud v. Liberty Life Assur. Co. of Boston, 438 F.3d 772, 777 (7th Cir. 2006); Mers v. Marriott Int'l Group Accidental Death & Dismemberment Plan, 144 F.3d 1014, 1020 (7th Cir.1998). We presume neutrality "unless a claimant shows by providing specific evidence of actual bias that there is a significant conflict." Kobs v. United Wis. Ins. Co., 400 F.3d 1036, 1039 (7th Cir.2005) (quoting Mers, 144 F.3d at 1020).

The source of Davis's argument is Unum's in-house doctors. However, whether a doctor is in-house or not is an irrelevant distinction in this context. To start, plan administrators have a duty to all plan participants and beneficiaries to investigate claims and make sure to avoid paying benefits to claimants who are not entitled to receive them. See Dougherty, 440 F.3d 910, 2006 WL 647719, at *6; Barnhart v. UNUM Life Ins. Co. of Am., 179 F.3d 583, 589 (8th Cir.1999). Further, an administrator's decision to "seek[ ] independent expert advice is evidence of a thorough investigation." Hightshue v. AIG Life Ins. Co., 135 F.3d 1144, 1148 (7th Cir.1998). When an administrator, like Unum here, opts to investigate a claim by obtaining an expert medical opinion — independent of its own lay opinion and that of the claimant's doctors — the administrator is going to pay a doctor one way or another. See Wallace v. Reliance Standard Life Ins. Co., 318 F.3d 723, 724 (7th Cir.2003). Thus, whether the administrator retains in-house doctors (arguably reducing overhead costs for the benefit of the plan's participants and beneficiaries) or pays for freelance doctors makes no difference in this conflict analysis. Paying for a legitimate and valuable service in order to evaluate a claim thoroughly does not create a review-altering conflict.

Had Unum given its doctors some specific stake in the outcome of Davis's case, such as paying the doctors more if Davis's claim were denied, then Davis would have an argument; however, Davis has not shown that the doctors in this case had any specific incentive to derail his claim. See Leipzig v. AIG Life Ins. Co., 362 F.3d 406, 409 (7th Cir.2004); Perlman, 195 F.3d at 981. Moreover, were there any legitimate doubts about the objectivity of Unum's in-house doctors in general, they would be dispelled by the actions of Dr. Buchanan, an Unum in-house doctor whose assessment, favorable to Davis, led to Unum approving mental disability benefits in the early stages of this case. In his argument here, Davis merely theorizes that in-house doctors have an inherent conflict in every case, including his. Such a theoretical argument is not enough to alter or otherwise diminish the customary arbitrary-and-capricious standard. See Mers, 144 F.3d at 1020; see also Rud, 438 F.3d at 777 (claimant must "demonstrate the existence of a real and not merely notional conflict of interest"); Kobs, 400 F.3d at 1039; Ruiz v. Cont'l Cas. Co., 400 F.3d 986, 991 n. 1 (7th Cir.2005).

Under this standard, then, we will uphold Unum's denial of benefits so long as that decision has "rational support in the record." Leipzig, 362 F.3d at 409. "`[Q]uestions of judgment are left to the plan administrator,' and `it is not our function to decide whether we would reach the same conclusion' as the administrator." Sisto, 429 F.3d at 701 (quoting Trombetta v. Cragin Fed. Bank for Sav. Employee Stock Ownership Plan, 102 F.3d 1435, 1438 (7th Cir.1996); Tegtmeier v. Midw. Operating Eng'rs Pension Trust Fund, 390 F.3d 1040, 1045 (7th Cir.2004)). Put simply, an administrator's decision will not be overturned unless it is "downright unreasonable." Sisto, 429 F.3d at 700 (quoting Tegtmeier, 390 F.3d at 1045).

The question before us, then, is whether there is rational support in the record for Unum's determination that Davis was not disabled to the extent that he was unable to perform the duties of any gainful occupation for which he was reasonably fitted. Based upon the extensive record, summarized in our background section above, we can easily answer this question in the affirmative. The expert opinions of Drs. Thomas, Feagin, Fluter, and Goldsmith adequately established that Davis could perform sedentary activities, and the vocational specialist identified several sedentary jobs for Davis. That is rational support. This evidence makes it "possible" for Unum "to offer a reasoned explanation" for its decision to deny benefits, and, under the arbitrary-and-capricious standard, we must respect Unum's judgment. Semien v. Life Ins. Co. of N. Am., 436 F.3d 805, 812 (7th Cir.2006) (quoting Trombetta, 102 F.3d at 1438); see also Dougherty, 440 F.3d 910, 2006 WL 647719, at *7; Kobs, 400 F.3d at 1040; Leipzig, 362 F.3d at 409.

Nevertheless, Davis argues, and the district court ruled, that Unum should not rely on this evidence for a variety of reasons, each of which we reject. Primarily, the district court (in a vein similar to that of Davis's conflict-of-interest argument above) penalized Unum for relying on in-house doctors, implying that in-house doctors inherently lack objectivity. However, again, Unum's in-house doctors were every bit as capable as outside doctors to evaluate the medical information in the file and provide independent expert medical opinions. The singular fact of working in-house does not disqualify a doctor from rendering an independent opinion any more than does paying an outside doctor to do the same, and neither the district court nor Davis go so far as to suggest that Unum could not hire a doctor to check the veracity and genuineness of Davis's claim and his doctors' opinions. Nor could they. See Dougherty, 440 F.3d 910, 2006 WL 647719, at *6; Leipzig, 362 F.3d at 409; Barnhart, 179 F.3d at 589; cf. Gaither v. Aetna Life Ins. Co., 394 F.3d 792, 807 (10th Cir.2004) ("fiduciary has a duty to protect the plan's assets against spurious claims"). Furthermore, to the extent that the district court would have preferred that Unum supplement and verify its doctors' opinions with that of an outside doctor, the district court went beyond the bounds of arbitrary-and-capricious review. Such questions of judgment and management are to be left to the administrator in this situation. The judicial task here is not to determine if the administrator's decision is correct, but only if it is reasonable.

The district court and Davis also fault Unum for relying on "a mere paper review," lamenting the fact that Unum's doctors did not personally examine Davis or speak with his doctors. However, neither the district court nor Davis has cited, and our research has not disclosed, any authority that generally prohibits the commonplace practice of doctors arriving at professional opinions after reviewing medical files. In such file reviews, doctors are fully able to evaluate medical information, balance the objective data against the subjective opinions of the treating physicians, and render an expert opinion without direct consultation. It is reasonable, therefore, for an administrator to rely on its doctors' assessments of the file and to save the plan the financial burden of conducting repetitive tests and examinations. See Dougherty, 440 F.3d 910, 2006 WL 647719, at *5 (reasonable for administrator to take fair-minded actions aimed at conserving plan assets for the benefit of all participants and beneficiaries).

This was not a situation, moreover, in which the administrator's doctors were completely at odds with the claimant's doctors and the medical evidence. For example, Unum's Dr. Feagin found that the results of medical testing indicated that Davis did not have post-polio syndrome. While Davis's Dr. Raymond clung to an opposite conclusion, Dr. Feagin's view was in line with the clinical findings of Dr. Meyers, Davis's second neurologist. Similarly, Davis's first neurologist, Dr. Levy, attributed Davis's perceived physical ailments to his psychiatric condition, and Dr. Reich, Davis's first treating physician, largely indicated that Davis was mentally disabled. The consensus view of Unum's doctors that Davis was not physically incapable of performing sedentary activities is not inconsistent with the views of Davis's first set of doctors (with whom Davis expressed dissatisfaction and replaced).

Interestingly, Dr. Raymond, who is the focal point of Davis's case, never took account of Dr. Levy's conclusion that psychiatric problems were the source of Davis's physical symptoms. Such an omission, not lost on Unum, weakens Dr. Raymond's reliability. See Shyman v. Unum Life Ins. Co., 427 F.3d 452, 456 (7th Cir.2005). There are additional reasons in the record to question the reliability of Dr. Raymond. For instance, in 2001 and 2002, Dr. Raymond's noted limitations for Davis were standing, walking, lifting, bending, carrying, and climbing, and Dr. Thomas at Unum agreed with Dr. Raymond in this regard, concluding that Davis could not walk or stand for prolonged periods. Nonetheless, after Dr. Thomas also stated that Davis could perform sedentary activities, Dr. Raymond then, in a 2003 letter, conveniently added sitting to Davis's list of limitations. Dr. Raymond's letter, in a similar after-the-fact approach, attempted to undermine the notion that Davis could perform sedentary activities by introducing carpal tunnel syndrome to Davis's growing list of complaints; however, Dr. Raymond provided nothing to substantiate this newfound ailment. Dr. Raymond further cited Davis's hypoglycemic episode as a source of disability. Nonetheless, Dr. Raymond ignored the fact, cogently identified by Unum's Dr. Fluter, that, if Davis adhered to his medication schedule and a proper diet, Davis would avoid such attacks. Moreover, Dr. Raymond continued to list prostate cancer as a factor contributing to Davis's disability, yet Dr. Raymond made no attempt to reconcile the fact that Davis's urologist, Dr. Gluckman, found that Davis's 1992 encounter with prostate cancer was successfully treated without recurrence. Further, Dr. Raymond's views on Davis's cardiac health were also inexplicably inconsistent. In 2001, Davis had a successful angioplasty to relieve a partially blocked artery, and, in 2002, Dr. Raymond affirmatively designated that Davis had no functional limitations on account of his cardiac condition. However, in 2003, without new data in the record, Dr. Raymond changed his position, listing coronary artery disease as a reason to consider Davis disabled.

In this light, these examples and others in the record show Dr. Raymond more as an advocate than a doctor rendering objective opinions. Cf. Leipzig, 362 F.3d at 409 ("Most of the time, physicians accept at face value what patients tell them about their symptoms; but [administrators] must consider the possibility that applicants are exaggerating in an effort to win benefits (or are sincere hypochondriacs not at serious medical risk)."). To a lesser extent, the record reveals similar questions about certain conclusions from Dr. Meyers. Suffice it to say that Davis presented a medical file of questionable reliability. It is not surprising, therefore, that Unum's doctors arrived at some opposite conclusions from those of Drs. Raymond and Meyers. What is more, given the vagaries and contradictions apparent on the face of Davis's claim, it was reasonable for Unum to rely on its doctors' file reviews in denying Davis's claim. See Black & Decker Disability Plan v. Nord, 538 U.S. 822, 834, 123 S. Ct. 1965, 155 L. Ed. 2d 1034 (2003) ("Plan administrators, of course, may not arbitrarily refuse to credit a claimant's reliable evidence, including the opinions of a treating physician. But, we hold, courts have no warrant to require administrators automatically to accord special weight to the opinions of a claimant's physician; nor may courts impose on plan administrators a discrete burden of explanation when they credit reliable evidence that conflicts with a treating physician's evaluation." (emphasis added)). Furthermore, reaching a decision amid such conflicting medical evidence is a question of judgment that should be left to Unum under the arbitrary-and-capricious standard. See id.; Ruiz, 400 F.3d at 992. Once again, the objective here is not to determine if Unum's decision is correct, but only if it is reasonable.

The final point to address is the explanations supporting the denial, which the district court and Davis conclude are insufficient. The dispute here is not about the adequacy of Unum's explanations, as the plan administrator, for the denial. Nor could it be. Unum's denial letters to Davis sufficiently detailed Unum's reasoning. See Herman v. Cent. States, Se. & Sw. Areas Pension Fund, 423 F.3d 684, 692-93 (7th Cir.2005) ("[T]his court will not substitute the conclusion it would have reached for the decision of the administrator, as long as the administrator makes an informed judgment and articulates an explanation for it that is satisfactory in light of the relevant facts." (quotation omitted)). Nevertheless, the district court and Davis take issue with the underlying explanations from Unum's doctors, criticizing the brevity of their medical writings. However, contrary to the district court's view, there is nothing in ERISA or our precedent requiring doctors to write like lawyers or plan administrators. Were we to arrive at such a decision today, we would unnecessarily and unwisely drive up the administrative costs of benefit plans, the negative affects of which would ultimately be borne by workers. ERISA counsels a different approach. See Gilbertson v. Allied Signal, Inc., 328 F.3d 625, 635 (10th Cir.2003) ("congressional purpose in enacting ERISA was `not to create a system that is so complex that administrative costs, or litigation expenses, unduly discourage employers from offering welfare benefit plans'" (quoting Varity Corp. v. Howe, 516 U.S. 489, 497, 116 S. Ct. 1065, 134 L. Ed. 2d 130 (1996))); Martin v. Ark. Blue Cross & Blue Shield, 299 F.3d 966, 972 (8th Cir.2002) (en banc).

It is enough, in situations such as this, for the doctors to review the file and render a professional, medical opinion. The record here unequivocally demonstrates that Unum's doctors reviewed the data supplied by Davis. The district court admitted as much in stating, "In their review of the file, Unum's in-house doctors did discuss [Davis's] medical testing and (implicitly) suggested that the objective results did not indicate that [Davis] cannot perform sedentary work." The Unum doctors' conclusions, however, were not simply "implicit."2 For example, there was nothing implicit or vague about Dr. Fluter's aforementioned conclusion regarding Davis's hypoglycemia. It was explicit and compelling. Further, to the extent that other conclusions were summary, we again see no grounds here for mandating that doctors draft lengthy, lawyer-like opinions.

Bottom line, the record contains rational support for Unum's denial of physical disability benefits. In concluding otherwise — and penalizing Unum for relying on in-house doctors who reviewed the file and gave doctor-like explanations for their conclusions — the district court went beyond the bounds of arbitrary-and-capricious review. On this record, it is not "downright unreasonable" to find that Davis was not unable to perform the duties of any gainful occupation for which he was reasonably fitted. We will thus uphold Unum's denial of long-term disability benefits under the arbitrary-and-capricious standard. Further, because Unum and the plan are entitled to summary judgment, there is no need for us to discuss Davis's challenge to the district court's remand for further administrative proceedings.

The correct standard to review Unum's denial of benefits in this case is the arbitrary-and-capricious standard. There is no evidence of conflict or bias that would necessitate a less deferential level of review. Furthermore, as with Unum's grant of mental disability benefits (which lasted twenty-four months), Unum's denial of long-term physical disability benefits has rational support in the record. Therefore, under the arbitrary-and-capricious standard, we will not disturb Unum's decision. Accordingly, the judgment of the district court is REVERSED, and the case is REMANDED with instructions for the district court to enter judgment in favor of Unum and the plan.