Melissa A. McIntyre, Plaintiff - Appellee v. Reliance Standard Life Insurance Company, Defendant - Appellant
No. 19-2367
United States Court of Appeals For the Eighth Circuit
Submitted: June 18, 2020 Filed: August 25, 2020
Before GRUENDER, WOLLMAN, and KOBES, Circuit Judges.
Appeal from United States District Court for the District of Minnesota
GRUENDER,
In this action arising under the Employee Retirement Income Security Act (“ERISA“),
I.
All her life, McIntyre has suffered from Charcot Marie Tooth Syndrome (“CMT“), a degenerative neurological condition affecting peripheral nerves such as those in the hands and feet. From 2003 to 2011, she worked as a nurse employed by the Mayo
Starting in 2011, she applied for a series of benefits from the Reliance plan. The Reliance policy paid a “Monthly Benefit” if “an Insured” was “Totally Disabled” within the meaning of the policy. The policy contained two definitions of “Totally Disabled.” The definition applicable for the first twenty-four months in which a benefit was payable (“short-term disability definition“) provided that “Totally Disabled” means “an Insured cannot perform the material duties of his/her Regular Occupation,” meaning “the occupation the Insured is routinely performing when Total Disability begins.” The definition after that period (“long-term disability definition“) provided that “Totally Disabled” means “an Insured cannot perform the material duties of Any Occupation,” meaning “an occupation normally performed in the national economy for which an Insured is reasonably suited based upon his/her education, training or experience.” The policy also stated that Reliance was the “claims review fiduciary” and had “the discretionary authority to interpret the Plan and the insurance policy and to determine eligibility for benefits.”
In 2011, McIntyre applied for and received benefits under the short-term disability definition. In 2013, Reliance began evaluating whether McIntyre would remain eligible for benefits under the long-term disability definition. This evaluation process continued until December 2015, when Reliance concluded McIntyre was no longer eligible for benefits because she was capable of working in “Any Occupation.” In February 2016, Reliance notified McIntyre she would no longer receive benefits under the Reliance plan because she was ineligible under the long-term disability definition.
On May 31, 2016, McIntyre filed an appeal with Reliance, challenging its eligibility determination under the long-term disability definition. After delays caused by both McIntyre and Reliance, she eventually underwent an independent medical examination in December 2016. The doctor who performed that examination concluded McIntyre was “capable of working full time in a sedentary position.” Following this examination, Reliance upheld its original determination that McIntyre was not eligible for benefits under the long-term disability definition and informed her of the same in late December 2016.
McIntyre then filed suit against Reliance. See
II.
We review de novo the district court‘s determination of the appropriate standard of review under ERISA. Barnhart v. UNUM Life Ins. Co. of Am., 179 F.3d 583, 587 (8th Cir. 1999). Generally, “a denial of benefits challenged under
The district court nevertheless decided that a de novo standard applied on the basis of our caselaw providing that “less deferential review” applies despite a grant of such discretionary authority if (1) either the administrator faces a “palpable conflict of interest” or a “serious procedural irregularity” arose in the review process, and (2) either the conflict or the procedural irregularity “caused a serious breach of the plan administrator‘s fiduciary duty” to the claimant. See Woo v. Deluxe Corp., 144 F.3d 1157, 1160 (8th Cir. 1998), abrogated in part by Metro. Life Ins. v. Glenn, 554 U.S. 105, 115-16 (2008). Specifically, the district court found a palpable conflict of interest present insofar as Reliance “both determines and pays claims” and ostensibly has a “history of biased claims administration.” The district court also found a serious procedural irregularity; namely, Reliance‘s “long delay in deciding McIntyre‘s appeal.” It then concluded that both of these caused Reliance to breach its fiduciary duty owed to McIntyre and therefore decided to “review McIntyre‘s benefits claim de novo.” McIntyre defends the district court‘s application of de novo review under our caselaw and argues in the alternative that a de novo standard should apply in light of authorities from other circuits whose approach she invites us to adopt.
A.
The key precedent underlying the district court‘s decision to apply a de novo standard is Woo, 144 F.3d 1157. In Woo, at 1160, we recognized a court could apply “less deferential review” in certain circumstances even when the administrator possesses discretionary authority to interpret and apply the policy. The district court found these circumstances present and evidently read Woo‘s endorsement of “less deferential review” to mean de novo review. This reading was error.
To begin, the parties agree that the Supreme Court‘s decision in Metropolitan Life Insurance v. Glenn, 554 U.S. 105 (2008), abrogated at least the conflict-of-interest component of Woo, as we have repeatedly recognized. See, e.g., Boyd v. ConAgra Foods, Inc., 879 F.3d 314, 320 (8th Cir. 2018) (noting that Glenn “abrogated Woo to the extent Woo allowed a less deferential standard of review based on merely a conflict of interest” (citing Wrenn v. Principal Life Ins., 636 F.3d 921, 924 n.6 (8th Cir. 2011)). It is thus undisputed that the district court erred in relying on the presence of a conflict of interest to justify de novo review.
The district court, however, also relied on the presence of “serious procedural irregularities” to justify de novo review. We have repeatedly avoided deciding whether Glenn “abrogated the ‘procedural irregularity’ component of the Woo sliding-scale approach.” Leirer, 910 F.3d at 396 (“We need not decide that issue here . . . .“); Boyd, 879 F.3d at 320 (“We need not address the validity of th[e procedural-irregularities] component . . . .“); Waldoch v. Medtronic, Inc., 757 F.3d 822, 830 n.3 (8th Cir. 2014) (“Our circuit has not definitively resolved the impact of Glenn on the ‘procedural irregularity component of the Woo sliding scale approach.’ See Wrenn, 636 F.3d at 924 n.6]. We need not resolve this
In Woo, the ERISA beneficiary argued that we should not review the administrator‘s benefits denial “under the traditional abuse of discretion standard because of procedural irregularities” in the benefits determination and because the administrator had a “conflict of interest.” 144 F.3d at 1160. We identified a “two-part gateway” by which the beneficiary could “obtain a less deferential review“: the beneficiary had to “present material, probative evidence demonstrating that (1) a palpable conflict of interest or a serious procedural irregularity existed, which (2) caused a serious breach of the plan administrator‘s fiduciary duty to her.” Id. at 1160-61. We concluded the beneficiary had satisfied this two-part test, “warranting a less deferential standard of review.” Id. at 1161.
We then turned to what this standard might look like, considering two approaches taken by other circuits in similar circumstances. Id. On the one hand, some courts “always review[ed]” a benefits denial “for an abuse of discretion” but adopted a “‘sliding scale’ approach” to the standard of proof the plan administrator had to satisfy to show the denial was not an abuse of discretion. Id. (citing Chambers v. Family Health Plan Corp., 100 F.3d 818, 825-27 (10th Cir. 1996)). On the other hand, other courts in effect reviewed a benefits denial de novo. Id. (citing Brown v. Blue Cross & Blue Shield, 898 F.2d 1556, 1566-67 (11th Cir. 1990)).
We “adopt[ed] the ‘sliding scale’ approach,” and in doing so we recognized that it still “requires the courts to apply an abuse of discretion analysis.” Id. We then applied this “modified abuse of discretion standard,” reviewing the benefits denial for an abuse of discretion but requiring the administrator to support its decision with “substantial evidence bordering on a preponderance” rather than merely “substantial evidence.” Id. at 1162-63. At no point did we state that the sliding scale approach would warrant a change in the standard of review from abuse of discretion to de novo—in fact, some of the authorities on which we relied clearly disclaimed such an understanding of that approach. See id. at 1161 (citing Layes v. Mead Corp., 132 F.3d 1246, 1250 (8th Cir. 1998) (noting how, while we had previously “contemplated the application of a less stringent standard of review in situations involving substantial procedural irregularities,” “[w]e did not suggest ... that a de novo standard was appropriate” but rather suggested “a less deferential abuse of discretion standard” might apply
Notwithstanding the fact that the Woo sliding scale approach did not authorize departure from the abuse-of-discretion standard of review, Woo came to be read by us as providing a gateway to de novo review. See, e.g., Johnson v. United of Omaha Life Ins., 775 F.3d 983, 987 (8th Cir. 2014); Hackett v. Standard Ins., 559 F.3d 825, 830 (8th Cir. 2009); Smith v. United Television, Inc. Special Severance Plan, 474 F.3d 1033, 1035 n.1 (8th Cir. 2007); Parkman v. Prudential Ins. Co. of Am., 439 F.3d 767, 772 n.5 (8th Cir. 2006) (per curiam).
The confusion appears to stem from a sentence in Woo in which we noted that “the ‘sliding scale’ analysis adheres to our decision in Armstrong v. Aetna Life Ins. Co., 128 F.3d 1263, 1265 (8th Cir. 1997), where the egregious circumstances essentially required the court to give no deference to the administrator‘s decision.” Woo, 144 F.3d at 1161-62. In Armstrong, we concluded that the “continuing conflict” present when the administrator of an ERISA plan is also its insurer permitted a court to review de novo the conflicted administrator‘s benefits decision. 128 F.3d at 1265, abrogated by Glenn, 554 U.S. at 112, 115 (holding that this sort of “dual role” conflict did not permit “a change in the standard of review, say, from deferential to de novo review,” but instead was simply to be weighed as a factor in the abuse-of-discretion analysis).
Later panels appeared to read Woo‘s purported “adhere[nce]” to Armstrong as an indication that de novo review was permitted if Woo‘s two-part test was satisfied, whether by procedural irregularities or conflicts of interest. In Morgan v. Contractors, Laborers, Teamsters & Engineers Pension Plan, for example, we cited Woo for the proposition that “a less deferential standard of review” may apply when serious procedural irregularities are present, found them present, concluded the second part of the Woo test was also satisfied, and then, citing to Armstrong, held “that these circumstances require us to review the [administrator‘s] decision to deny benefits de novo.” 287 F.3d 716, 722-23 (8th Cir. 2002); see also Schatz v. Mut. of Omaha Ins., 220 F.3d 944, 948 n.7. (8th Cir. 2000) (reading Woo and Armstrong together to mean that “if there are ‘egregious circumstances,’ Woo, 144 F.3d at 1162, then our review is de novo“); cf. House v. Paul Revere Life Ins., 241 F.3d 1045, 1048 (8th Cir. 2001) (citing Woo and Armstrong together for the proposition that “conflicts of interest or procedural irregularities ... may prompt a more searching review“).
Three data points, however, should have tipped off future panels that Woo‘s dictum about adherence to Armstrong was not a cue that Woo‘s sliding scale approach was a gateway to de novo review. See Pinto v. Reliance Standard Life Ins., 214 F.3d 377, 391-92 (3d Cir. 2000) (recognizing the discrepancy between Armstrong and Woo), abrogated on other grounds by Glenn, 554 U.S. 105; accord West v. Aetna Life Ins., 171 F. Supp. 2d 856, 874-75 (N.D. Iowa 2001). First, Woo and Armstrong are facially irreconcilable. See Roger C. Siske et al., What‘s New in Employee Benefits: A Summary of Current Case and Other Developments 121 (ALI-ABA Course of Study 1998) (“Although the Woo court attempted to harmonize its decision and that of the Armstrong court, the two decisions really are at odds.“). Woo held that the presence of a “palpable conflict of interest” could trigger “the ‘sliding scale’ approach,” which “requires the courts to apply an abuse of discretion analysis,” simply “taking into consideration the conflict,” 144 F.3d at 1160-61 (emphasis added); while Armstrong held that de novo review was authorized when the administrator “faces a continuing conflict in playing the dual role of administrator and insurer of the health benefits plan,” 128 F.3d at 1265. Second, the Woo panel followed a Tenth Circuit decision to adopt the sliding scale approach, see 144 F.3d at 1161 (citing and following Chambers, 100 F.3d at 825-27), while the Armstrong panel explicitly declined to follow that same decision‘s “sliding scale’ approach,” see 128 F.3d at 1265 (citing and declining to follow Chambers, 100 F.3d at 824-27). Third, we found “egregious conduct” present in Woo yet still applied only a “modified abuse of discretion standard” to review the administrator‘s decision, undermining any notion that “egregious circumstances” could trigger de novo review under Woo. 144 F.3d at 1162-63.2
Unfortunately, this misunderstanding of Woo persisted, albeit uniformly in dicta. In some cases, we overtly suggested, or at least operated under the premise, that Woo sanctioned de novo review. See, e.g., Ingram, 812 F.3d at 631 (acknowledging the beneficiary‘s argument that “we should . . . apply a less deferential de novo standard of review” under Woo‘s two-part test but declining to do so because “none of the conflicts of interest and procedural irregularities . . . warranted departure from the abuse-of-discretion standard of review“); Johnson, 775 F.3d at 987 & n.1 (noting that “application of the de novo standard might apply” under Woo but declining to address whether the procedural-irregularities component of Woo was still good law post-Glenn because the beneficiary failed to establish that procedural irregularities existed); Wade v. Aetna Life Ins., 684 F.3d 1360, 1362 & n.2 (8th Cir. 2012) (assuming that the procedural-irregularities component of ”Woo still applies” post-Glenn but concluding that “the de novo standard [the beneficiary] suggests is not warranted in situations such as this“); Hackett, 559 F.3d at 829-31 (discussing what “a claimant seeking de novo review” under Woo had to show but reversing and remanding on the ground that the district court failed to take the administrator‘s conflict of interest into account when reviewing its decision for an abuse of discretion); Smith, 474 F.3d at 1035 n.1 (commenting that “[a] plan administrator‘s decision may be subject to de novo review” under Woo but noting that the beneficiary on appeal “does not contest that the applicable standard of review is the abuse-of-discretion standard“); Parkman, 439 F.3d at 772 n.5 (suggesting that the beneficiary would have been entitled “to have her claim reviewed de novo” under Woo but concluding Woo was not triggered).
In other cases, while we did not explicitly treat Woo as if it authorized de novo review, we still suggested Woo permitted review under a standard other than abuse of discretion. See, e.g., Waldoch, 757 F.3d at 830 n.3 (suggesting that Woo authorized “changing the standard of review from abuse of discretion to a less deferential standard” but concluding no such change was permitted in that case because the beneficiary “failed to establish that any procedural irregularities exist“); Wrenn, 636 F.3d at 924 n.6 (”Woo held a less deferential standard of review than abuse of discretion applied .... Because we conclude [the administrator] abused its discretion, we do not address the extent to which Glenn may have changed the procedural
Regardless, none of these later decisions control to the extent they treat Woo as providing a gateway to de novo review or some other heightened form of review other than abuse of discretion—notions at odds with Woo itself. First, as previously noted, in most of the cases in which we treated Woo as permitting departure from the abuse-of-discretion standard, we did so in dicta. “[W]hen an issue is not squarely addressed in prior case law, we are not bound by precedent through stare decisis,” and we “need not follow dicta.” Passmore v. Astrue, 533 F.3d 658, 660-61 (8th Cir. 2008). Second, under Mader v. United States, 654 F.3d 794, 800 (8th Cir. 2011) (en banc), when we are “faced with conflicting panel opinions,” we must follow the “earliest opinion.” Obviously, all of our precedents mischaracterizing the holding in Woo postdate Woo, and under Mader we must follow Woo rather than subsequent cases misstating the holding in Woo and even, as in Morgan, misapplying Woo. See, e.g., NLRB v. Leiferman Enters., LLC, 649 F.3d 873, 879 n.1 (8th Cir. 2011) (declining to follow a subsequent panel precedent that “misread[]” an earlier panel precedent and so was “squarely at odds with what that [earlier] case actually held“); United States v. Gaines, 639 F.3d 423, 428 n.4 (8th Cir. 2011) (following the rule of a prior panel precedent even though, “over the course of several iterations” in subsequent precedential decisions, the rule “inadvertently evolved” into something other than what the prior panel held).
In sum, the district court erred in treating a conflict of interest as a trigger for de novo review rather than simply as a factor in determining whether Reliance abused its discretion. See Glenn, 554 U.S. at 115. The district court also erred in treating an ostensibly serious procedural irregularity as a trigger for de novo review rather than for Woo‘s “‘sliding scale’ approach,” which “require[d] the [district] court[] to apply an abuse of discretion analysis,” simply “taking into consideration the . . . procedural irregularity.” See Woo, 144 F.3d at 1161.3
B.
Defending the district court‘s decision to apply de novo review on alternative grounds, McIntyre invites us to adopt other circuits’ approach permitting district courts to review benefits denials de novo in the face of an administrator‘s “decisional delay” beyond the deadline prescribed in ERISA‘s implementing regulations or the plan itself.4 For instance, McIntyre calls
Such a rule, however, is not the law of our circuit. Consider Johnson, in which the district court found procedural irregularities present and concluded (mistakenly, see supra Section II.A) that de novo review thus was warranted under Woo. 775 F.3d at 988. As relevant here, one of the procedural irregularities was the administrator‘s “failure to timely process the [beneficiary‘s] claims.” Id. (brackets omitted). Specifically, the beneficiary appealed the initial denial of her long-term disability claim on August 27, 2010. See Johnson v. United of Omaha Life Ins., No. 8:11CV296, 2013 WL 942511, at *9-10 (D. Neb. Mar. 11, 2013). The administrator denied the appeal on January 28, 2011, meaning the appeal was pending for 154 days. See id. at *11.5 Under both the terms of the policy at issue there as well as the governing regulations at the time, the administrator was required to issue a decision within forty-five days of receipt of the appeal, with only one forty-five-day extension of that initial forty-five-day period permitted, for a maximum total of ninety days. See
Whatever may be the law in other circuits, it is apparent in light of Johnson that, in our circuit, the administrator‘s
McIntyre suggests we should extend McGarrah and Seman so that both a wholesale failure to act on an appeal and a decisional delay on appeal are treated as triggers for de novo review, arguing that a decisional delay “is just a different degree of failure to act.” This contention misunderstands the rationale underpinning the rule of McGarrah and Seman and thus blurs the material distinction between a wholesale failure to act and a decisional delay.
As we explained in Seman, the administrator‘s wholesale failure to act on an appeal can trigger de novo review because, in certain circumstances—when the initial denial was based on an incomplete record, the beneficiary developed the record more fully on appeal, but the administrator failed to reconsider the matter in light of the developed record—the failure to act on an appeal is “equivalent” to failing “to render a decision” at all concerning the claim. See 334 F.3d at 733 (citing Mansker v. TMG Life Ins., 54 F.3d 1322, 1328 (8th Cir. 1995) (holding that the administrator‘s “failure to render a decision on certain issues” means that “the district court could ... decide the issues de novo” by reasoning, in essence, that if the administrator‘s discretion is not exercised, there can be no review for an abuse of that discretion)). A delayed decision on appeal made after the beneficiary has fully developed the record is different in kind than the wholesale failure to act on an appeal we found equivalent to no decision at all in Seman. Delayed decisionmaking thus does not come within the rule of McGarrah and Seman.
In short, under circuit law, the administrator‘s delay in deciding an appeal is not a trigger for de novo review. It is, rather, a factor to be considered by the district court when reviewing the administrator‘s decision for an abuse of discretion. See Woo, 144 F.3d at 1161. Other circuits may opt for a different approach, see Nichols, 406 F.3d at 109 (recognizing Seman as part of a three-way circuit split but declining to follow it), but we are bound by our circuit‘s law on this point.
III.
Reliance asks us to apply the abuse-of-discretion standard of review and reverse the district court by finding that Reliance did not abuse its discretion in denying McIntyre‘s claim. Ordinarily, remand is appropriate when the district court errs by reviewing the administrator‘s benefits decision under the wrong standard of review. See Woo, 144 F.3d at 1162 n.3. We see no reason to depart from that protocol here, cf. id., particularly given that “review of the plan administrator‘s denial of benefits” in this case “is a highly fact-intensive inquiry,” see Seman, 334 F.3d at 734.