MEMORANDUM OPINION

GABRIEL W. GORENSTEIN, Magistrate Judge.

Plaintiff Liyan He brings this action under the Employee Retirement Income Security Act, 29 U.S.C. § 1001 et seq. ("ERISA"), seeking long-term disability benefits under a plan administered by Cigna Life Insurance Company of New York ("Cigna"). Plaintiff was employed by Cornell University. See Complaint, filed Mar. 27, 2014 (Docket # 2), ¶ 7. Cigna issued a group long-term disability insurance policy to Cornell to benefit eligible Cornell employees, including plaintiff. Id. ¶¶ 8-10. Plaintiff filed a claim with Cigna for long-term disability benefits. See id. ¶¶ 12, 14. Cigna denied the claim, plaintiff appealed, and Cigna upheld its decision to deny her claim. See id. ¶¶ 17-19.

Plaintiff filed a motion seeking the depositions of three Cigna individuals who were involved in the denial of her claim as well as a deposition of a Cigna representative pursuant to Fed. R. Civ. P. 30(b)(6) on a list of sixteen topics. See Memorandum of Law in Support of Motion to Compel, filed Oct. 29, 2014 (Docket # 26) ("Pl. Mem."), at 1; Ex. 1 to Pl. Mem. (notices of depositions). She also sought to compel Cigna to produce documents relating to its reserves, employee evaluations, and employee compensation. See Pl. Mem. at 15-16. This request was the subject of a hearing on January 7, 2015, at which the Court made certain discovery rulings. This Memorandum Opinion explains the Court's reasoning in making those rulings.

Under ERISA, decisions of benefit plan administrators are reviewed under either an "arbitrary and capricious" or a "de novo" standard. See, e.g., O'Shea v. First Manhattan Co. Thrift Plan & Trust, 55 F.3d 109, 111-12 (2d Cir. 1995) (quoting Firestone Tire & Rubber Co. v. Bruch, 489 U.S. 101, 115 (1989)). Here, the parties have stipulated that the standard of review is de novo.

To support her motion for discovery, plaintiff cites to cases where discovery was allowed on the issues of the claims administrator's conflict of interest and procedural irregularities in the treatment of the plaintiff's claim. See Pl. Mem. at 4-8 (citing, inter alia, Puri v. Hartford Life & Accident Ins. Co., 784 F.Supp.2d 103, 105 (D. Conn. 2011); Mergel v. Prudential Life Ins. Co. of Am., 2009 WL 2849084, at *2 n.3 (S.D.N.Y. Sept. 1, 2009); Burgio v. Prudential Life Ins. Co. of Am., 253 F.R.D. 219, 227-29 (E.D.N.Y. 2008); Reittinger v. Verizon Commc'ns Inc., 2006 WL 3327676, at *2-3 (N.D.N.Y. Nov. 15, 2006); Mitchell v. First Reliance Standard Life Ins. Co., 237 F.R.D. 50, 53 (S.D.N.Y. 2006); Nagele v. Elec. Data Sys. Corp., 193 F.R.D. 94, 102 (W.D.N.Y. 2000)). But in many of these cases, including this Court's decision in Mitchell, there was the potential that the case would ultimately be reviewed under the "arbitrary and capricious" standard. See, e.g., Mitchell, 237 F.R.D. at 53. None of these cases is relevant because, on arbitrary and capricious review, whether the claims administrator has a conflict of interest and whether there were procedural irregularities are both relevant to the disposition on the merits. See Metro. Life Ins. Co. v. Glenn, 554 U.S. 105, 111 ("If a benefit plan gives discretion to an administrator or fiduciary who is operating under a conflict of interest, that conflict must be weighed as a factor in determining whether there is an abuse of discretion.") (quoting Firestone, 489 U.S. at 115) (internal quotation marks and emphasis omitted); id. at 117 ("[W]hen judges review the lawfulness of benefit denials, they will often take account of several different considerations of which conflict of interest is one."); Durakovic v. Bldg. Serv. 32 BJ Pension Fund, 609 F.3d 133, 138 (2d Cir. 2010) (requiring courts to consider any conflict of interest and whether "procedural safeguards are in place" which mitigate that conflict). That is, in cases potentially involving arbitrary and capricious review, the district court may be called upon to address the questions of conflict of interest and procedural regularities in making its merits determination. In de novo review cases, by contrast, any conflict of interest and procedural irregularities are not per se relevant to the merits determination because the district court conducts its own review of the evidence without deferring to the administrator's prior determinations.

That being said, the existence of a conflict of interest and procedural irregularities still has some relevance in a case involving de novo review. This is because the Second Circuit has held that a district court may consider evidence outside of the administrative record on its merits review only where "good cause" is shown, see Paese v. Hartford Life & Accident Ins. Co., 449 F.3d 435, 441 (2d Cir. 2006) (citation omitted), and that such "good cause" is evaluated based on whether there is a conflict of interest and procedural irregularities, see Locher v. Unum Life Ins. Co. of Am., 389 F.3d 288, 294-96 (2d Cir. 2004) (discussing DeFelice v. Am. Int'l Life Assurance Co. of N.Y., 112 F.3d 61, 66-67 (2d Cir. 1997)).

The rationale behind expanding the record based on the existence of a conflict of interest and procedural irregularities is obvious where the court reviews the claims administrator's decision under arbitrary and capricious review. After all, these matters will have to be addressed in the merits review under ERISA to evaluate the degree of deference that will be given to the administrator's decision, as described in cases such as Glenn and Durakovic. But the relevance of any conflicts and procedural irregularities is much less obvious in cases that will ultimately be subject to de novo review. Nonetheless, the Second Circuit made clear in two de novo review cases that the decision whether to consider material outside the administrative record turns on the existence of a conflict of interest and procedural regularities. See Locher, 389 F.3d at 293 n.2, 296; Paese, 449 F.3d at 441-42. As a matter of logic, the relevance of conflicts and procedural irregularities in de novo review cases should derive from their potential to impact the development of a proper administrative record. We thus conclude that discovery in cases that will ultimately be subject to de novo review should normally be designed to cast light on conflicts and procedural irregularities that might have affected the completeness of the administrative record.

In considering what specific discovery should be ordered here, we recognize that there are a number of cases holding that there is a special test for obtaining discovery in ERISA benefit determination cases: that is, whether the plaintiff has shown a "reasonable chance that the requested discovery will satisfy the good cause requirement." E.g., Garban v. Cigna Life Ins. Co. of N.Y., 2011 WL 3586070, at *2 (S.D.N.Y. Aug. 11, 2011) (citation omitted). This standard derives from some phrasing used in the case of Anderson v. Sotheby's Inc. Severance Plan, 2005 WL 6567123 (S.D.N.Y. May 13, 2005), in which the court was trying to explain why it would not allow a "fishing expedition" to conduct discovery on the factual bases for a potential "good cause" showing. Id. at *7. We agree with Joyner v. Continental Casualty Co., 837 F.Supp.2d 233 (S.D.N.Y. 2011), however, that the use of this phrasing as a special standard to govern ERISA cases is "unwarranted." Id. at 242. Instead, Fed. R. Civ. P. 26(b)(1) — limiting the scope of discovery to material that is "relevant to any party's claim or defense" — applies. Nonetheless, we must limit discovery if "the burden or expense of the proposed discovery outweighs its likely benefit." Fed. R. Civ. P. 26(b)(2)(C)(iii). In addition, we must recognize the "significant ERISA policy interests of minimizing costs of claim disputes and ensuring prompt claims-resolution procedures." Locher, 389 F.3d at 295.

With these principles in mind, we now turn to the question of what discovery should reasonably be afforded to plaintiff to probe Cigna's conflict of interest and any alleged procedural irregularities in its decision on plaintiff's claim. We recognize that some courts have granted expansive discovery on the issues of conflict and procedural irregularities in de novo review cases by allowing multiple depositions and document requests. See, e.g., Garban, 2011 WL 3586070, at *3; Allison v. Unum Life Ins. Co., 2005 WL 1457636, at *8, *13 (E.D.N.Y. Feb. 11, 2005). At the same time, other courts have refused to allow any discovery on these issues. See Yasinoski v. Conn. Gen. Life Ins. Co., 2009 WL 3254929, at *4, *13 (E.D.N.Y. Sept. 30, 2009); see also Quninones v. First Unum Life Ins. Co., 2011 WL 797456, at *1, *3 (S.D.N.Y. Mar. 4, 2011) (court declined to determine the applicable standard of review and denied plaintiff's request for discovery). Others still have allowed only a limited amount of discovery. See Durham v. Prudential Life Ins. Co. of Am., 890 F.Supp.2d 390, 396-98 (S.D.N.Y. 2012) (allowing only a Rule 30(b)(6) deposition and plaintiff's "reasonable document requests").

In this case, Cigna does not dispute that it has an inherent conflict of interest, inasmuch as it "administer[s] a plan and pay[s] benefits out of its own funds." Zuckerbrod v. Phx. Mut. Life Ins. Co., 78 F.3d 46, 49 (2d Cir. 1996). Thus, there is no need for discovery into the issue of Cigna's structural conflict of interest. We also believe that the existing record provides a sound basis on which to evaluate the existence of procedural irregularities and other non-structural conflicts that might have affected the completeness of the record. We are thus skeptical that plaintiff needs much more discovery beyond what is already contained in the administrative record in order to attempt to make her "good cause" showing. In light of this, and considering Rule 26(b)(2)(C)(iii) and the policy interests articulated in Locher, 389 F.3d at 295, the Court at the January 7, 2015 conference limited plaintiff to a single deposition of a Cigna employee as to the procedural administration of plaintiffs claim and denied all document discovery with the exception of the examination of the written evaluations of some key employees involved in the decisionmaking on plaintiffs claim. All other discovery has been denied. Again, the goal of this discovery will be only to determine whether there are conflicts or procedural irregularities that bear on the question of whether the existing administrative record is incomplete.