POSNER, Circuit Judge.

This suit under ERISA for disability payments presents the recurring question whether an employee welfare benefits plan creates an entitlement to lifetime benefits rather than just to benefits that can be terminated by an amendment to the plan.

In 1997 Michael Marrs, an employee of Motorola, ceased working because of a psychiatric condition and began drawing disability benefits under Motorola’s Disability Income Plan. Six years later Motorola amended the plan to place a two-year limit on benefits for disability resulting from certain “Mental, Nervous, Alcohol, [or] Drug-Related” (MNAD) conditions, including Marrs’s. Such limitations on MNAD conditions are common in employee disability plans. Rogers v. Department of Health & Environmental Control, 174 F.3d 431, 435 (4th Cir.1999); see, e.g., Krolnik v. Prudential Ins. Co., 570 F.3d 841, 842 (7th Cir.2009); Hackner v. Long Term Disability Plan for Employees of Havi Group LP, 81 Fed.Appx. 589, 591 (7th Cir.2003); Kahane v. UNUM Life Ins. Co., 563 F.3d 1210, 1212 (11th Cir.2009); Steven J. Sacher et al., Employee Benefits Law 1003-05, 1147-48 (2d ed. Supp.2008); cf. Dana L. Kaplan, “Can Legislation Alone Solve America’s Mental Health Dilemma? Current State Legislative Schemes Cannot Achieve Mental Health Parity,” 8 Quinnipiac Health L.J. *785 325, 328-30 (2005). Previously, however, Motorola had imposed no time limit on the receipt of such benefits. Although Marrs had already received benefits for more than two years, he was given an additional two years of benefits, starting on the date of the amendment. That period has ended and the benefits have ceased.

His suit is on behalf of himself and others in the same position, and a class action has been certified. The district court granted summary judgment in Motorola’s favor. In an earlier stage of the litigation, the district court had dismissed two other claims under Rule 12(b)(6). Marrs renews them on appeal, but they are too clearly without merit to require us to discuss them.

Marrs contends that the application of the amendment to persons in his situation violates a provision of the disability-income plan that, like the provision in Article V of the U.S. Constitution according to which a state cannot by constitutional amendment be deprived of its right to two senators without its consent, limits Motorola’s power to amend the plan. The provision states that no amendment “shall adversely affect the rights of any Participant to receive benefits with respect to periods of Disability prior to the adoption date of the [amendment].” Marrs interprets “periods of Disability prior to the adoption date” to mean one or more periods of disability that began before the plan was amended but may not have ended before then.

That is a forced reading. The reference in the plan to “periods” rather than “period” suggests the segmentation of a period of disability, with some segments (“periods”) lying before and some after the amendment. It is true that the plan defines the term “Period of Disability” to mean “one or more periods of absence from Active Employment due to Disability,” and if we substituted “Period of Disability prior to the plan” for “periods of Disability prior to the plan” we would come closer to Marrs’s preferred interpretation. But the plan provides that only words the initial letters of which are capitalized are defined terms, and so “periods of Disability” cannot be equated to “Period of Disability.”

Marrs’s interpretation is further undermined by the disability plan that was in force in 1997 when he stopped working and that states that if a participant became disabled prior to 1994, the benefit levels specified in the disability income plan in force then would “continue in force until [the participant] returns to work for thirty (30) days.” This provision would be surplusage in the superseding plan (where it also appears) if the plan guaranteed the continuation of disability benefits without diminution even if, as in Marrs’s case, the disability did not arise before 1994.

Whether this interpretation of the plan, which is the plan administrator’s interpretation, is correct or not, it is reasonable; and we are inclined to stop with that observation. But Marrs asks us to give no weight to the administrator’s interpretation because the administrator labored under a conflict of interest: Motorola is both the plan administrator and the payor of benefits awarded under the plan.

Marrs cites the Supreme Court’s recent decision in Metropolitan Life Ins. Co. v. Glenn, — U.S. -, 128 S.Ct. 2343, 171 L.Ed.2d 299 (2008), which addresses the significance of a conflict of interest by the plan administrator when, as in that case and this one, the plan commits the decision whether to award benefits to the administrator’s discretion. (If the plan did not confer discretion on the administrator, his decision would be entitled to no deference by the reviewing court.)

Marrs reads Glenn to require “a more penetrating inquiry into the actions of a conflicted administrator” than the ear *786 lier case law, which was all over the map— see Kathryn J. Kennedy, “Judicial Standard of Review in ERISA Benefit Claim Cases,” 50 Am. U.L.Rev. 1083, 1135-62 (2001); see also Denmark v. Liberty Life Assurance Co., 566 F.3d 1, 6 (1st Cir.2009)—had required. Although normally an employer can amend a welfare benefits plan without regard to the employees’ interests, Hughes Aircraft Co. v. Jacobson, 525 U.S. 432, 443-44, 119 S.Ct. 755, 142 L.Ed.2d 881 (1999); Curtiss-Wright Corp. v. Schoonejongen, 514 U.S. 73, 78, 115 S.Ct. 1223, 131 L.Ed.2d 94 (1995), remember that Marrs’s claim is that the amendment curtailing his benefits violated the plan, and that is a claim of wrongful denial of benefits and so the Supreme Court’s concern with “conflicted” administrators would seem pertinent.

True, the issue in this case is the interpretation of a plan document rather than the application of the plan’s criteria for an award of benefits to particular facts; and the interpretation of a contract, unless extrinsic evidence is considered, is usually treated as an issue of law, which an appellate tribunal therefore resolves without deferring to the opinion of the first-line decision maker. But when an ERISA plan gives the plan administrator discretion to interpret its terms as well as to determine eligibility for benefits under terms the meaning of which is not questioned, the court can, as the parties to this case agree, reject the administrator’s interpretation only if it is unreasonable (“arbitrary and capricious”). Firestone Tire & Rubber Co. v. Bruch, 489 U.S. 101, 115, 109 S.Ct. 948, 103 L.Ed.2d 80 (1989); Carr v. Gates Health Care Plan, 195 F.3d 292, 294 (7th Cir.1999) (“under the arbitrary and capricious standard, the administrator’s decision will only be overturned if it is ‘downright unreasonable.’ A denial of benefits will not be set aside if the denial was based upon a reasonable interpretation of the plan documents”) (citations omitted); Call v. Ameritech Management Pension Plan, 475 F.3d 816, 822 (7th Cir.2007). And both the original and the amended plan in this case confer discretion on the plan administrator to “construe and interpret the Plan, decide all questions of fact and questions of eligibility and determine the amount, manner and time of payment of any benefits hereunder” (emphasis added).

The administrator is not by virtue of such a grant of authority free to disregard unambiguous language in the plan, id. at 822-23; Swaback v. American Information Technologies Corp., 103 F.3d 535, 540 (7th Cir.1996) (“if fiduciaries or administrators of an ERISA plan controvert the plain meaning of a plan, their actions are arbitrary and capricious”) — that would be unreasonable. But the administrator’s use of interpretive tools to disambiguate ambiguous language is, one would think, by the terms of the plan entitled to deferential consideration by a reviewing court.

Yet Vallone v. CNA Financial Corp., 375 F.3d 623 (7th Cir.2004), although noting that the plan in that case gave the plan administrator interpretive discretion, treats the question whether benefits claimed by the plaintiffs had vested as a straight issue of contract interpretation. The court seemed to give no weight to the plan administrator’s judgment, though in the end agreeing with the administrator’s denial of benefits. But the court was mirroring the parties’ emphasis, and it did not say that the grant of interpretive discretion to a plan administrator is entitled to no weight, although one passage in the opinion could be thought to imply this. See id. at 632. Any such implication would be in tension with the Firestone and Carr decisions — and the former is a Supreme Court decision, which is a trump.

Confusion may have been injected into the issue of deference to interpretive discretion by cases which say that the *787 interpretation of an ERISA plan is governed by the ordinary federal common law principles of contract interpretation, e.g., Ruttenberg v. United States Life Ins. Co., 413 F.3d 652, 659 (7th Cir.2005); Mathews v. Sears Pension Plan, 144 F.3d 461, 465 (7th Cir.1998); Dobson v. Hartford Financial Services Group, Inc., 389 F.3d 386, 399 (2d Cir.2004), and by cases in this circuit that say that welfare benefits plans are presumed not to create lifetime benefits. Hackett v. Xerox Corp. Long-Term Disability Income Plan, 315 F.3d 771, 774 (7th Cir.2003); Rossetto v. Pabst Brewing Co., 217 F.3d 539, 543 (7th Cir.2000); contra, Gibbs v. CIGNA Corp., 440 F.3d 571, 576 (2d Cir.2006). But these statements, so far as applicable to plans in which the administrator is given interpretive discretion, are properly understood as aids to determining whether the denial of benefits by the administrator is reasonable, rather than as warrants for a court’s resolving interpretive disputes without any deference to the administrator’s exercise of interpretive discretion. Thus, as explained in Ross v. Indiana State Teacher’s Ass’n Ins. Trust, 159 F.3d 1001, 1011 (7th Cir. 1998), “although, generally, ambiguities in an insurance policy are construed in favor of an insured, in the ERISA context in which a plan administrator has been empowered to interpret the terms of the plan, this rule does not obtain. See Morton v. Smith, 91 F.3d 867, 871 n. 1 (7th Cir.1996) (explaining that rule of contra proferentem applies only when courts undertake de novo review of an administrator’s interpretation of an ERISA plan).”

The play between principles of contract interpretation and the scope of review of a plan administrator’s decision is illustrated by Hackett. The court had to determine which version of a plan controlled the determination of entitlement to disability benefits. The earlier version did not grant interpretive authority to the plan administrator, but the later one did, so we had to determine as a matter of contract law which plan controlled because that would determine the standard of review of the plan administrator’s interpretation that we should use. 315 F.3d at 773-74. In this case both plans confer interpretive authority on the plan administrator.

But maybe the court in Vallone thought that an alleged misinterpretation of a plan’s eligibility criteria is less serious than an alleged misinterpretation of a plan provision that affects an employer’s right to amend it without trampling on vested rights, and therefore that appellate review of the latter claim should be less deferential. Neither Vallone nor, to our knowledge, any other case says that. Nor is it apparent why the interpretive principles (including the scope of judicial review of the first-line interpreter’s decision) that govern the two types of issue should differ. In any event we do not think that the Glenn decision has the significance that Marrs attaches to it even if it is fully applicable to cases involving the interpretation of plan amendments alleged to infringe on vested rights.

When a payment of benefits comes out of the plan administrator’s pocket, the administrator has an incentive to resolve a close case in favor of a denial of benefits. But that incentive may be outweighed by other incentives, such as an employer’s interest in maintaining a reputation among current and prospective employees for fair dealing — a reputation that may enable him to obtain a better-quality employee at a lower cost in wages and benefits. E.g., Williams v. Interpublic Severance Pay Plan, 523 F.3d 819, 821-22 (7th Cir.2008). Then too, the employees who actually decide benefits claims at the plan-administrator level may not be acutely concerned with the financial implications of a benefits award for their employer. Id. at 821; Perlman v. Swiss Bank Corp. Comprehen *788 sive Disability Protection Plan, 195 F.3d 975, 981 (7th Cir.1999). But especially when a firm is struggling (which may or may not be the case here — there is nothing in the record bearing on the question), an opportunity for short-run economies may dominate decision making by benefits officers. In any event, a majority of the Supreme Court Justices consider the potential conflict of interest of a plan administrator (or its staff) serious enough to be given weight in judicial review of the denial of benefits.

But how much weight should it be given? The nub of the Glenn opinion is the following passage:

[W]hen judges review the lawfulness of benefit denials, they will often take account of several different considerations of which a conflict of interest is one. This kind of review is no stranger to the judicial system. Not only trust law, but also administrative law, can ask judges to determine lawfulness by taking account of several different, often case-specific, factors, reaching a result by weighing all together. In such instances, any one factor will act as a tiebreaker when the other factors are closely balanced, the degree of closeness necessary depending upon the tiebreaking factor’s inherent or case-specific importance. The conflict of interest at issue here, for example, should prove more important (perhaps of great importance) where circumstances suggest a higher likelihood that it affected the benefits decision, including, but not limited to, cases where an insurance company administrator has a history of biased claims administration. It should prove less important (perhaps to the vanishing point) where the administrator has taken active steps to reduce potential bias and to promote accuracy, for example, by walling off claims administrators from those interested in firm finances, or by imposing management checks that penalize inaccurate decisionmaking irrespective of whom the inaccuracy benefits.

128 S.Ct. at 2351 (citations omitted). A dissent by Justice Scalia argued that a conflict of interest should prompt an inquiry only into the existence of improper motive that would render the plan administrator’s decision unreasonable. If the decision is reasonable, he argued, in the sense in which “a reasonable decision is one over which reasonable minds seeking the ‘best’ or ‘right’ answer could disagree,” the fact that the administrator had a conflict of interest is irrelevant, id. at 2360, “unless the conflict actually and improperly motivates the decision.” Id. at 2357 (emphasis in original).

There are two ways to read the majority opinion. One, which tracks its language and has been echoed in opinions in this and other circuits, e.g., Jenkins v. Price Waterhouse Long Term Disability Plan, 564 F.3d 856, 861-62 (7th Cir.2009); Holland v. Int’l Paper Co. Retirement Plan, 576 F.3d 240, 246-49 (5th Cir.2009), makes the existence of a conflict of interest one factor out of many in determining reasonableness. That sounds like a balancing test in which unweighted factors mysteriously are weighed. Such a test is not conducive to providing guidance to courts or plan administrators. “Multifactor tests with no weight assigned to any factor are bad enough from the standpoint of providing an objective basis for a judicial decision; multifactor tests when none of the factors is concrete are worse.” Menard, Inc. v. Commissioner, 560 F.3d 620, 622-23 (7th Cir.2009) (citations omitted); see also Sullivan v. William A. Randolph, Inc., 504 F.3d 665, 671 (7th Cir.2007); Short v. Belleville Shoe Mfg. Co., 908 F.2d 1385, 1394 (7th Cir.1990) (concurring opin *789 ion); Stevens v. Tillman, 855 F.2d 394, 399-400 (7th Cir.1988).

If that’s the test the Supreme Court has adopted, we must bow. But it is not clear that the rudderless balancing test suggested by the passage that we quoted was intended to be the last word on the standard that should guide decision in these cases. The test can be made more directive, without contradicting the Court’s opinion, by first recognizing that while a decision may look reasonable if one just reads the decision and the record, a decision that is “reasonable” rather than clearly correct is a decision that might just as well have gone the other way, United States v. Williams, 81 F.3d 1434, 1437-38 (7th Cir.1996), as when “reasonable” is used, as it often is, to mean that a ruling was not an abuse of discretion. Call v. Ameritech Management Pension Plan, supra, 475 F.3d at 822. If the circumstances indicate that probably the decision denying benefits was decisively influenced by the plan administrator’s conflict of interest, it must be set aside, just as a decision by a judge who should have recused himself must be set aside even if he might well have reached the same decision had there been no basis for recusal.

The likelihood that the conflict of interest influenced the decision is therefore the decisive consideration, as seems implicit in the majority opinion’s reference to indications of “procedural unreasonableness” in the plan administrator’s handling of the claim in issue, 128 S.Ct. at 2352, and its suggestion that efforts by a administrator to minimize a conflict of interest would weigh in favor of upholding his decision. Id. at 2351. It is thus not the existence of a conflict of interest — which is a given in almost all ERISA eases — but the gravity of the conflict, as inferred from the circumstances, that is critical. For example, the terms of employment of the staff that decides benefits claims might, consistent with the Williams and Perlman cases of this court, affect a determination of how likely it is that those employees would slant their decisions in favor of their employer’s short-term interest in minimizing his benefits expense.

Justice Scalia appears to believe that if the plan administrator has no improper motive, the existence of a conflict of interest cannot have affected the denial of benefits: “if one is to draw any inference about a fiduciary from the fact that he made an informed, reasonable, though apparently self-serving discretionary decision, it should be that he suppressed his selfish interest (as the settlor anticipated) in compliance with his duties of good faith and loyalty.” Id. at 2360 (emphasis in original). But if, as we have suggested, both a decision in favor of the applicant for benefits and a decision against may be reasonable, the plan administrator’s conflict of interest may cause him unconsciously to decide against the applicant and thus in favor of the plan’s financial welfare; hence the majority Justices’ interest in whether the plan establishes safeguards designed to minimize such a tendency. There are no indications in this case, however, that the plan administrator labored under a conflict of interest serious enough to influence his decision consciously or unconsciously — a decision that was otherwise entirely reasonable — decisively. The district court’s decision must therefore be

Affirmed.