Moro v. State of Oregon

384 P.3d 504, 360 Or. 467

Opinion

BALMER, C. J.

Attorney fees and costs awarded.

BALMER, C. J.

This matter is before us on petitions for attorney fees and costs brought by a law firm and three individuals (claimants) who participated in the underlying litigation. In that litigation, claimants were petitioners or represented petitioners who challenged legislation passed in 2013 that changed the pension benefits paid to certain members of the Public Employee Retirement System (PERS) by limiting the statutory cost-of-living adjustment (COLA) and eliminating a PERS income-tax offset for out-of-state retirees. In Moro v. State of Oregon, 357 Or 167, 351 P3d 1 (2015) (Moro I), this court largely agreed with petitioners’ argument that modifications to the COLA formula impaired petitioners’ contractual rights, thus violating Article I, section 21, of the Oregon Constitution. But the court rejected petitioners’ similar challenge to the elimination of the income-tax offset. Petitioners, who were active and retired members of PERS, were the prevailing parties.

Following the decision in Moro I, claimants petitioned for attorney fees and costs. State respondents and county/school district respondents filed objections. 1 We referred those petitions to a special master for recommended findings of fact and conclusions of law. Moro v. State of Oregon, 358 Or 375, 381, 364 P3d 325 (2015) (Moro II). The special master reported his recommendations to this court, and the parties subsequently filed objections and responses to those recommendations. The issues raised in those filings include which legal doctrines justify an award of attorney fees in this case; whether self-represented attorneys are eligible to receive an award of attorney fees; whether the fees sought by claimants are reasonable; and how to pay for an award of fees and costs.

After reviewing those filings, and for the reasons described below, we conclude that fees should be awarded based on the common-fund and substantial-benefit doctrines; that the self-represented attorneys are eligible to receive a fee award under those doctrines; that a reasonable fee award under the lodestar approach must be based on reasonable hourly rates and reflect reductions to account for duplicative work and work on unsuccessful claims; and that an award in this case should be paid for as determined by the Public Employees Retirement Board (PERB) in a manner that is consistent with its statutory authority and fiduciary obligations.

Four claimants seek compensation here. One claimant is the law firm Bennett, Hartman, Morris & Kaplan, LLP (“Bennett Hartman”), which represented the Moro group of petitioners. The three additional claimants — Reynolds, Riemer, and Jones — are PERS members who acted as pro se petitioners in the underlying litigation. Reynolds and Riemer, although pro se petitioners, also are attorneys and seek both attorney fees and costs. Jones seeks only his costs.

Claimants who seek attorney fees have calculated their fees using the lodestar method. Under the lodestar method, a court determines a reasonable attorney fee award by multiplying the reasonable hours expended by a reasonable hourly rate and, when appropriate, enhancing the lodestar amount with a fee multiplier. See Strawn v. Farmers Ins. Co., 353 Or 210, 217, 297 P3d 439 (2013) (describing the lodestar method). Bennett Hartman seeks $1,401,040 in fees, based on 1,693.8 hours of attorney time at between $150 and $500 per hour and a fee multiplier of 2.0. Reynolds seeks $562,000 in fees, based on 562 hours of attorney time at $500 per hour and a fee multiplier of 2.0. And Riemer seeks $397,500 in fees, based on 265 hours of attorney time at $500 per hour and a fee multiplier of 3.0.

As it relates to costs, Bennett Hartman seeks $62,066.13; Reynolds seeks $1,214.48; Riemer seeks $1,159.15; and Jones seeks $1,479.24. Bennett Hartman’s cost request is substantially higher because it includes the costs of an expert witness who testified in support of petitioners in the underlying litigation.

State respondents and county/school district respondents filed objections with this court asserting various reasons to deny or reduce the fees claimed. As an initial matter, respondents dispute what legal grounds are available to justify attorney fees. Bennett Hartman and Reynolds rely on the common-fund doctrine, while Riemer relies on both the common-fund doctrine and on this court’s decision in Deras v. Myers, 272 Or 47, 535 P2d 541 (1975). Although respondents agree that a fee award may be justified under the common-fund doctrine, they dispute the applicability of Deras, and county/school district respondents additionally argue that a portion of the fee award should be justified under the substantial-benefit doctrine. Respondents also contend that, regardless of which doctrine justifies a fee award, no fees should be awarded to Reynolds and Riemer because of their status as pro se petitioners, rather than attorneys serving in a representative capacity.

If fees are awarded, the parties agree that any fee award allowed in this case must be reasonable. Respondents object to the reasonableness of the fees sought— specifically, whether claimants are using appropriate hourly rates and fee multipliers and whether fees should be reduced to account for duplicative work and work on the unsuccessful tax-offset claim.

Finally, the parties dispute how to pay for any award of costs and fees, namely, how to collect the money from the beneficiaries of the litigation. Those beneficiaries consist of active, inactive, and retired PERS members falling within different tiers of membership. The assets of those beneficiaries are therefore spread out among different accounts held within the Public Employees Retirement Fund (PERF). That raises the question of whether the money for any awards should come from, for example, payments being made to retirees, PERS’s contingency reserve account, or individual PERS accounts.

All those disputes were presented to the special master, whose report contained recommended findings of fact and conclusions of law. The special master concluded that the common-fund and substantial-benefit doctrines applied, but that Deras fees should not be allowed. The special master further concluded that the common-fund and substantial-benefit doctrines largely justified the fees sought by Bennett Hartman, although he recommended using a 1.5 fee multiplier rather than the 2.0 fee multiplier that Bennett Hartman requested.

The special master recommended no award of attorney fees to Reynolds and Riemer, because they were acting as pro se litigants rather than as attorneys. He also made the alternative recommendation that, if this court were to determine that Reynolds and Riemer were entitled to attorney fees despite their status as pro se litigants, any fee award should be adjusted based on his determination that only 20 percent of their work contributed to the success of the litigation. According to the special master, 80 percent of their work either went to the losing tax-offset claim or was duplicative of that performed by Bennett Hartman. The special master did not recommend that any multiplier be applied to fees awarded to Reynolds and Riemer.

Separate from attorney fees, Reynolds and Jones seek their costs under ORAP 13.05; Bennett Hartman seeks its litigation expenses as costs that should be awarded under the common-fund doctrine; and Riemer seeks his costs under both. The special master recommended an award of costs to Bennett Hartman under the common-fund doctrine in the amount of $62,066.13, to be paid in the same manner as the fee award — that is, out of PERS funds held on behalf of the PERS members who had benefitted from the litigation. And, based on ORAP 13.05, the special master recommended granting the costs sought by Reynolds and recommended small downward adjustments to the costs sought by Riemer and Jones — leading to a total cost award of $548.05 for Riemer and $1,379.24 for Jones. With the exception of the deduction applied to Riemer, we accept the special master’s recommendations on costs. 2

The parties dispute numerous issues regarding the special master’s recommended attorney-fee awards. We address each of those issues below.

A. Grounds for Attorney-Fee Recovery

The parties assert three grounds for attorney-fee recovery: the common-fund doctrine, the substantial-benefit doctrine, and this court’s decision in Deras. We address that issue first because the rationales underlying different grounds for recovering fees may influence how we resolve other issues related to the determination of fee amounts.

Riemer is the only claimant seeking fees based on Deras. Under Deras, a court has the discretion to award fees if “the parties who request attorney fees prevailed [and] those prevailing parties vindicated an important constitutional right applying to all citizens” rather than “gaining] something peculiar to themselves.” Lehman v. Bradbury, 334 Or 579, 583, 54 P3d 591 (2002).

In previous PERS litigation, we denied a claimant’s request for Deras fees. Strunk v. PERB, 341 Or 175, 181, 139 P3d 956 (2006) (Strunk II). The special master in this case recommends following that precedent. We agree. The litigants in this case, including Riemer, were attempting to gain something peculiar to themselves and other PERS members. Although the number of people affected is large, it is, nevertheless, a discrete group of people and a group that is easily distinguished from the public as a whole. For that reason, we deny Riemer’s request to award attorney fees under Deras.

The remaining grounds offered by the parties are the common-fund and substantial-benefit doctrines. Both doctrines rest on the same equitable and restitutionary grounds: to avoid unjust enrichment by spreading the litigation costs among those who benefited from a legal action brought by a plaintiff. Crandon Capital Partners v. Shelk, 342 Or 555, 566, 157 P3d 176 (2007). The common-fund doctrine applies when a plaintiffs “legal efforts create, discover, increase, or preserve a fund of money to which others also have a claim.” Strunk II, 341 Or at 181. A party who litigates such a case may recover the costs of those legal efforts, including attorney fees, from the created or preserved fund. Id. The substantial-benefit doctrine, on the other hand, ordinarily applies when a party’s legal efforts create nonliqui-dated benefits — whether or not pecuniary — that are held in common with others, such as when a union member’s lawsuit benefits other union members, Gilbert v. Hoisting & Port. Engrs., 237 Or 130,137,384 P2d 136 (1963), or a shareholder’s lawsuit benefits the corporation and thus other shareholders, Crandon Capital Partners, 342 Or at 562-63.

In common-fund and substantial-benefit cases, a party pursuing its litigation objectives necessarily confers benefits on nonparties, because the litigation implicates interests that they share. See Restatement (Third) of Restitution and Unjust Enrichment § 29 comment a (2011) (“The underlying premise of all such claims in restitution is that — by reason of the parties’ interconnected interests — the claimant cannot pursue justifiable, self-interested objectives without benefiting the [nonparties] as well.”). In Strunk II, this court observed that the common-fund doctrine is “used to spread litigation expenses among all beneficiaries of a preserved fund so that litigant-beneficiaries are not required to bear the entire financial burden of the litigation while [nonparty-] beneficiaries receive the benefits at no cost.” 341 Or at 181. Similarly, in Crandon Capital, we linked the common-fund and the substantial-benefit doctrines, noting that, for both, “fees are awarded not, as in a ‘prevailing party’ case, to make the plaintiff whole by shifting all costs to the wrongdoer, but instead to spread the costs among those on whose behalf the case was brought and who benefitted from the plaintiffs efforts.” 342 Or at 566.

Thus, under both doctrines, nonparty beneficiaries of the litigation may be required to contribute to the legal expenses of the litigation that secured the benefits. A claim for a contribution to those legal expenses may be brought by the party who incurred the expenses or by the attorney who provided the legal work. Strunk II, 341 Or at 183-84. 3 Requiring nonparties to pay the fee award is in contrast to fee-shifting provisions, which generally require an adverse party to pay the legal expenses of the prevailing party. See Restatement § 29 comment c (“[T]here can be no common-fund recovery against an adverse party.”).

We agree with the special master that both the substantial-benefit and common-fund doctrines provide a basis for attorney fees here. It is beyond dispute that the litigation undertaken by petitioners conferred very substantial benefits on other PERS members. Some of those members are retirees who, following enactment of the 2013 amendments, received smaller COLAs than they were entitled to and, as a result of that litigation, are now recovering that shortfall. Other PERS members, including retired and active members, will benefit substantially over many years from our decision that PERB cannot reduce the COLA rights applied to PERS benefits earned before the legislature modified those COLA rights in 2013. See Moro I, 357 Or at 184-87 (describing pre-amendment and post-amendment COLA benefits).

According to the special master, the present combined value of benefits of the COLA holding to all nonparty beneficiaries is about $4.5 billion, and no party seriously disputes that figure. As to the application of the common-fund theory, there is disagreement over the size of the common fund established by the litigation. Claimants assert that the litigation created a “fund” in the amount of the benefit that the nonparty beneficiaries will receive over time— essentially, the entire $4.5 billion of benefits identified by the special master. County/school district respondents, however, argue that the only common fund created by the litigation is the $66 million in restored COLA adjustments for retired PERS members who had received lesser amounts between the effective date of the 2013 COLA amendment and this court’s decision invalidating those changes. The remainder of the $4.5 billion, they argue, relates to amounts that PERS members will receive over time, but is not a liquidated “fund” of money subject to the common-fund doctrine.

We need not resolve that dispute over the size of the common fund in order to determine attorney fees in this case, however. As discussed, both the substantial-benefit and common-fund doctrines provide a basis for an award of fees here, and the difference between them is primarily in the nature of the benefits created by the litigation. Under both theories, the amount of the fee award is based on the same equitable and restitutionary considerations. Indeed, the Restatement treats the substantial-benefit doctrine as an application of the common-fund doctrine, where the common fund is an entity in which the party and other beneficiaries have interconnected interests. See id. at § 29 comment f (“The standard example of a fund of this second type involves corporate stock: in the terminology of Section 29, the corporation itself is then the Tund’ and its shareholders the beneficiaries.”). Accordingly, we turn to a consideration of the appropriate fee awards in this case, based on the common-fund and substantial-benefit doctrines.

B. Fees for Self-Represented Attorneys

The next question is whether self-represented attorneys may receive an attorney-fee award under the common-fund and substantial-benefit doctrines. Claimants Reynolds and Riemer contend that they are each entitled to an attorney-fee award because they are both attorneys licensed to practice law in other states — Reynolds in Washington and Riemer in Arizona — and they performed legal work in this litigation. Respondents object, arguing that neither has the authority to practice law in Oregon and that the equitable and restitutionary grounds for an award under the common-fund and substantial-benefit doctrines do not justify an award for self-represented attorneys. The special master agreed with respondents and recommended awarding Reynolds and Riemer no attorney fees.

The special master reached that conclusion based on both narrow technical grounds and broad policy grounds. The narrow technical grounds are statutes and rules setting out requirements for out-of-state attorneys to practice law in Oregon courts, such as being admitted pro hac vice. 4 Neither Reynolds nor Riemer complied with those requirements. The special master therefore reasoned that neither Reynolds nor Riemer was “practicing law” when they participated in the litigation. Instead, both were allowed to participate only because a statute, ORS 9.320, allows all parties to a lawsuit, whether or not they are attorneys, to “prosecute [] and defend[]” an action without violating the prohibition on unauthorized practice of law. See also ORS 9.160(2) (establishing that self-represented parties do not violate ban on the unauthorized practice of law).

For support, the special master contrasted the facts presented by Reynolds and Riemer with the facts of Colby v. Gunson, 349 Or 1, 238 P3d 374 (2010), a case in which this court held that a self-represented attorney could obtain fees under a statutory fee-shifting provision. In Colby, this court noted that the self-represented attorney in that case “is an attorney, in the ordinary sense of the word. He graduated from law school, is a member of the Oregon State Bar, and is authorized to practice law in this state. Throughout the proceedings below, he was subject to the Oregon Rules of Professional Conduct, along with other statutory provisions that govern the conduct of attorneys.” Id. at 8. Because Reynolds and Riemer were pro se litigants who had not complied with the rules on pro hac vice admission, the special master concluded that they were not authorized to practice law in Oregon and, therefore, were not eligible for a fee award.

The problem with relying on those narrow technical grounds is that they do not correspond to the equitable goal served by a restitutionary attorney-fee award under the common-fund and substantial-benefit doctrines — namely, avoiding the unjust enrichment that would result from allowing nonparties to enjoy the benefits of the litigation without contributing to the costs of the litigation. A restitu-tionary award for the services of another is generally limited to professional services. See Matter of Cont’l Illinois Sec. Litig., 962 F2d 566, 571 (7th Cir 1992), as amended on denial of reh’g (May 22, 1992) (Posner, J.) (“The basis for an award of fees in a common-fund case is, as we said, restitutionary, and the law of restitution (excepting salvage in admiralty) generally confines the right to restitution to professionals, such as doctors and lawyers.”) (citing 2 George E. Palmer, The Law of Restitution, ch 10 (1978)). Therefore, the underlying question in deciding on a restitutionary fee award is whether Reynolds and Riemer were lawyers who performed work normally performed by a lawyer.

The rules on pro hac vice admission and the unauthorized practice of law do not answer that question and are not directed at unjust enrichment. Instead, those rules are directed at consumer protection and prohibit a nonlawyer, or an out-of-state lawyer who has not complied with the applicable rules, from representing or advising another person as a lawyer in Oregon. See Johnson v. Premo, 355 Or 866, 872, 333 P3d 288 (2014) (“The prohibition against nonlawyer legal practice serves the dual purpose of protecting the public interest and the rights of individual litigants”). Those rules do not apply to individuals representing themselves, because, as noted above, parties to a lawsuit may generally prosecute or defend themselves under ORS 9.320, regardless of whether they are lawyers.

Reynolds and Riemer failed to comply with the rules of pro hac vice admission not because they could not satisfy those standards, but because, as self-represented parties, they were not required to comply with those rules. Their failure to comply with those rules means only that they did not represent or advise others as lawyers; it does not answer the question of whether they are lawyers and whether the work that they performed was legal work, such that they may be entitled to attorney fees under the substantial-benefit and common-fund doctrines.

Further, our statement in Colby noting that the attorney in that case was authorized to practice law in Oregon, 349 Or at 8, should not be read as creating a standard for determining whether someone is a lawyer who performed legal work. Instead, we stated only that someone who is authorized to practice law in Oregon is an attorney, not that an attorney is only someone who is authorized to practice law in Oregon.

Rather, like the attorney in Colby, Reynolds and Riemer are lawyers. They went to law school and are active members of bars in other states. And the work that they performed in this case was legal work — namely, researching the law, developing legal arguments, and presenting those arguments in briefs to this court. Therefore, Reynolds and Riemer were lawyers who performed legal work in this case.

The special master also denied Reynolds and Riemer attorney fees based on broad policy grounds, which would preclude fee awards even to self-represented attorneys who were authorized to represent or advise others as a lawyer, either as active members of the Oregon bar or as attorneys admitted pro hac vice. The policy grounds that the special master relied on are set out in Zucker v. Westinghouse Elec., 374 F3d 221 (3d Cir 2004), where the court refused to award fees under the common-fund doctrine to a self-represented attorney who successfully raised objections to a proposed class-action settlement. The court reasoned that “awarding [a self-represented attorney] attorney’s fees potentially could ‘tempt’ other lawyer-shareholders to ‘advance garden variety objections because of the prospect of an award of attorney fees for their personal service.’” Id. at 226. According to the court,

“We note that [the self-represented attorney] did not incur any financial liabilities for his work on this case. Failure to award [the self-represented attorney] fees should not discourage other shareholders from raising meritorious objections in the future; it will only ensure that they pursue objections with the assistance of third-party counsel.”

Id. 5

There are two difficulties with applying those grounds in this case. First, as to the concern that allowing fees would tempt self-represented attorneys to bring specious claims, this court rejected a similar concern raised in Colby:

“Although not necessary to our decision here, we note that the legislature has addressed that concern by permitting attorney fees only to parties who ‘prevail [] in the suit’ and by requiring that the attorney fee award be ‘reasonable.’ ORS 192.490(3). Further, the legislature has provided a list of factors that a court must consider in determining the amount of any attorney fee award, several of which protect against the abusive fee generation potential that the Court of Appeals feared. See ORS 20.075(1), (2) (listing factors to be considered in determining amount of any attorney fee award).”

349 Or at 8-9. Like the statutory fees at issue in Colby, common-fund and substantial-benefit fees are allowed only to prevailing parties and must be reasonable. Further, common-fund and substantial-benefit fees include additional protections, because the fees cannot exceed the value of the benefit conferred by the litigation. See Restatement § 29(3)(b) (permitting an attorney-fee award from a common fund only if “the measurable value added to the beneficiary’s interest in the common fund by the claimant’s intervention exceeds the beneficiary’s liability to the claimant”).

Second, as noted above, the purpose of a fee award under the common-fund and substantial-benefit doctrines is to avoid unjust enrichment. Whether the nonparty beneficiaries are enriched by the litigation, and whether that enrichment is unjust, does not turn on whether that litigation was brought by an attorney in a representative capacity or a nonrepresentative capacity. The California Supreme Court has allowed fees to self-represented attorneys in common-fund cases on that rationale:

“It would be inconsistent with the common fund theory to deny [the self-represented attorney] compensation for his services in these circumstances. The rationale of that theory is that fees should be awarded to the person who creates such a fund because all who will benefit from it should bear equally the burdens of its creation or preservation, and this result is best achieved by taxing the fund itself.”

Consumers Lobby Against Monopolies v. Pub. Utilities Com., 25 Cal 3d 891, 914-15, 603 P2d 41 (1979), overruled on other grounds by Kowis v. Howard, 3 Cal 4th 888, 838 P2d 250 (1992).

In this case, the litigation benefited nonparty PERS members by invalidating COLA reductions, and both Reynolds and Riemer participated in the litigation by performing legal work as lawyers. That remains true even though Reynolds and Riemer performed that work in a self-represented capacity. As a result, Reynolds and Riemer are entitled to an attorney-fee award necessary to avoid unjust enrichment.

C. Reasonableness of the Fees Requested

Determining the amount needed to avoid unjust enrichment requires assessing the reasonableness of the fees that claimants have requested. Under the common-fund and substantial-benefit doctrines, an attorney-fee award is limited to a reasonable fee. Claimants have the burden of establishing the reasonableness of the fees that they are requesting. Strawn, 353 Or at 225. There are three issues in this case related to the reasonableness of the fees requested: the hourly rates; the extent to which the legal work bene-fitted the nonparty PERS members, including whether the fees requested are duplicative of work by other claimants or related to unsuccessful claims; and the fee multiplier, if any.

The special master found that the hourly rates sought by Bennett Hartman were appropriate and that all of Bennett Hartman’s work benefitted nonparty PERS members. But the special master determined that Bennett Hartman was entitled to a fee multiplier of 1.5, rather than the 2.0 fee multiplier that Bennett Hartman sought. The special master also recommended reasonable attorney fees for Reynolds and Riemer — if this court determines, as we have, that their status as pro se litigants does not prevent them from receiving a fee award. In those alternative recommendations, the special master calculated the fee award using the hourly rates sought by Reynolds and Riemer, but he excluded work that went to the losing tax-offset claim and work that was duplicative of that performed by Bennett and Hartman. He concluded that only 20 percent of the work that Reynolds and Riemer performed benefitted the nonparty PERS members. Further, the special master recommended that Reynolds and Riemer receive no fee multiplier.

1. Hourly rates

Bennett Hartman seeks different hourly rates for different attorneys, with the maximum rate of $500 per hour for its lead counsel, Greg Hartman. Reynolds and Riemer also seek fees based on an hourly rate of $500 per hour. The special master concluded that Bennett Hartman’s rates were reasonable. He did not address the reasonableness of those rates for Reynolds and Riemer, but his alternative recommendations were based on the $500-per-hour rate sought by Reynolds and Riemer.

The $500-per-hour rate exceeds Hartman’s normal labor/employment rate of $315 per hour for work that is not contracted to unions, which receive a lower rate because they are long-standing clients. But the market for PERS-related work is not the same as the market for normal labor/ employment work. Reynolds and Riemer offer no evidence of their normal market rate. Reynolds appears to be retired without an active legal practice. And Riemer is the staff director of the Arizona Judicial Ethics Advisory Committee without an active private practice.

A court should not rubberstamp hourly rates, particularly when an attorney seeks rates beyond what he or she ordinarily would receive from paying clients and when the rates sought are at the very top of the market, such as those in this case. For context, in the 2012 Oregon State Bar Economic Survey, which is the last one available, the hourly billing rate for the 95th percentile of private practice attorneys in Portland was $450, and the 95th percentile for the entire state was $405.

Nevertheless, the rates charged by Hartman and Reynolds are justified, because they have substantial experience in appellate matters and both played a substantial role in earlier PERS litigation. 6 Hartman has represented petitioners in each of the major PERS cases discussed in the Moro I opinion. And Reynolds, while working as Assistant Attorney General, briefed and argued one of those cases, Oregon State Police Officers’ Assn. v. State of Oregon, 323 Or 356, 918 P2d 765 (1996), and participated in other PERS-related litigation. As a result, they are uniquely knowledgeable about the mechanics of PERS benefits and the relevant legal arguments. And PERS cases are generally high stakes and are both factually and legally complicated. So a rate at the top of the market is not unreasonable for those skills. We therefore conclude that the hourly rates requested by Bennett Hartman and Reynolds are reasonable.

It is difficult, however, to justify Riemer’s work at that same rate. Although he has appellate experience briefing and arguing attorney disciplinary cases before this court in his previous role as General Counsel of the Oregon State Bar, he has no particular expertise in PERS litigation. Based on his appellate experience but his lack of PERS experience, we conclude that a reasonable hourly rate for Riemer’s work is at the 75th percentile from the 2012 report, which was $350 for Portland.

2. Extent of the benefit provided by the work

“The cases are unanimous that simply doing work on behalf of the class does not create a right to compensation; the focus is on whether that work provided a benefit to the class.” In re Cendant Corp. Sec. Litig., 404 F3d 173, 191 (3d Cir 2005) (emphasis in original). For example, this court has previously refused to award substantial-benefit fees to parties that “gave no attention in their briefing and argument to the statute and rule on which the court’s ultimate disposition turned.” Leo v. Keisling, 329 Or 273, 280, 986 P2d 562 (1999) (refusing equitable fees where the parties made constitutional arguments but the court relied on statutory grounds to reach its result).

The principle that compensable legal work must benefit the nonparty beneficiaries has been applied by other courts to deny or reduce fees to account for work on unsuccessful claims. See, e.g., In re Enron Corp. Sec., Derivative & ERISA Litig., 586 F Supp 2d 732, 822 (SD Tex 2008) (considering work on unsuccessful claims). It has also been applied to account for duplication of effort. See, e.g., Reynolds v. Beneficial Nat. Bank, 288 F3d 277, 288-89 (7th Cir 2002) (denying fees to settlement objectors who “added nothing” because lead counsel made same objections). The parties dispute the extent to which the requested fees should be reduced to account for work on unsuccessful claims or for duplicative work. We address those issues separately.

In this case, petitioners presented numerous arguments related to two broad categories of PERS benefits: the COLA and the income-tax offsets for out-of-state retirees. Petitioners prevailed on the COLA claim but not on the tax-offset claim. To determine any reasonable amount of fees, we must consider whether a fee award should include fees for work advancing the unsuccessful claim for tax offsets. The special master did not reduce Bennett Hartman’s fees to account for unsuccessful claims, but, in his alternative fee award for Reynolds and Riemer, the special master reduced the award to account for work on unsuccessful claims. Reynolds and Riemer object to that reduction. And state respondents object to the special master’s failure to similarly reduce Bennett Hartman’s fees to account for work on the tax-offset claim.

In Strunk v. PERB, 343 Or 226, 169 P3d 1242 (2007) (Strunk III), this court did not reduce attorney-fee awards to account for work on unsuccessful claims, but it is unclear to what extent that issue was considered by the court or previously presented to the special master in that case. Nevertheless, while considering other proposed reductions, the court announced the applicable standard for determining whether work contributed to the benefits in that case:

“[A]fter examining respondents’ other billing-related objections and carefully scrutinizing petitioners’ billing records, we conclude that the requisite nexus between the benefits provided in this case and the fees sought as a result is missing for some items that petitioners’ seek compensation for.”

Id. at 240. As a result, the court applied a standard requiring a “nexus between the benefits provided * * * and the fees sought.” Id.

That standard is similar to a frequently used standard in other courts: “whether the successful and unsuccessful claims are based upon the same facts and legal theories, i.e., whether the claims are related.” In re Enron Corp. Sec., Derivative & ERISA Litig., 586 F Supp 2d at 822 (quotations omitted). Under that standard, “[w]hen the successful and unsuccessful claims involve a ‘common core of facts’ or ‘are based on related legal theories,’ then attorney fees incurred in the presentation of unsuccessful claims are recoverable on the theory that they contributed to the plaintiffs ultimate success.” Id.

In this case, claimants argue that there was a sufficient nexus between the successful COLA claim and the unsuccessful tax-offset claim because the two claims overlapped — that is, both were premised on constitutional rights against the impairment of contracts, and the factual record largely addressed the respondent’s public policy defenses related to economic necessity.

That overlap fails to establish a sufficient nexus, however. To the extent that there is overlapping work, that work would be treated as if it went solely to the fee-generating COLA claim, because fees for that work would have been incurred regardless of the non-fee-generating claim. See Estate of Smith v. Ware, 307 Or 478, 481-82, 769 P2d 773 (1989) (holding, in a statutory fee case, that overlapping work is treated as going to the fee-generating claim). So the question is the extent to which nonoverlapping work on the tax-offset claim contributed to the benefits.

On that question, claimants — including Bennett Hartman — fail to present grounds for concluding that non-overlapping work on the tax-offset claim benefitted the non-party PERS members. The lack of such grounds is acute in this case, because the COLA claim and the tax-offset claim, if successful, would have benefitted two different classes of potential nonparty beneficiaries. The COLA claim benefits all PERS members who earned PERS benefits before the 2013 legislative modifications. And the tax-offset claim would have benefitted only those PERS members who earned PERS benefits before October 1991 — and only those who now or in the future will reside outside of Oregon. It is inconsistent with the restitutionary rationale justifying the common-fund and substantial-benefit awards to require PERS members who did not earn benefits before October 1991 to pay for legal work on the tax-offset claim, because that work was never going to benefit them in the first place. As a result, all fees awarded to Reynolds, Riemer, and Bennett Hartman must be reduced to exclude work on the tax-offset claim.

Respondents further argue that claimants’ fees should be reduced to account for duplication of effort. When two attorneys duplicate their efforts, they are generally not both benefiting the nonparty beneficiaries of the litigation, because the same benefit would result even if one of the attorneys had not performed the duplicative work. The special master recommended finding that, in this case, Reynolds, Riemer, and Bennett Hartman duplicated their efforts because each presented substantially similar legal arguments with regards to the prevailing claim — namely, that the COLA adjustments violated the right against impairment of contract. The special master recommended accounting for that duplication by reducing the fees awarded to Reynolds and Riemer, but not the fees awarded to Bennett Hartman.

Reynolds and Riemer object to the reductions recommended by the special master. They argue that there are no grounds in this case to determine that their work duplicated Bennett Hartman instead of determining that Bennett Hartman duplicated their work. In Strunk III, for example, the court stated that “the fact that petitioners’ lawyers briefed some of the same issues in the course of bringing their cases to this court * * * without more, is insufficient to support” reducing the awards for duplication of effort. 343 Or at 239. The court went on to say that the record in that case did not allow the court to determine which attorneys duplicated the efforts of which other attorneys. Id. (“[T]hat argument assumes that the efforts of the Strunk petitioners’ and their lawyers was the sine qua non of the fund preserved here, while the work product of the other parties named as petitioners in this case derived solely from that effort.”).

This case, however, is distinguishable from Strunk III. In that case, the special master made no findings relevant to the issue of duplication. Moro II, 358 Or at 381 (“In Strunk III, the record created by the parties before the special master did not allow this court to determine the extent to which those attorneys had duplicated their efforts or directed their efforts toward unsuccessful claims”). In this case, however, by recommending that we reduce Reynolds’ and Riemer’s fees to account for duplication of Bennett Hartman, the special master found, in effect, that Bennett Hartman acted as the lead counsel. That conclusion is in accord with the apparent role of the attorneys in the case, with Bennett Hartman taking the lead role in litigating each stage of the case, from the factfinding proceedings before the special master to oral arguments before this court.

Because Bennett Hartman clearly acted as lead counsel, Reynolds and Riemer had the burden to demonstrate that the contributions of their work went beyond that performed by Bennett Hartman. See, e.g., In re Cendant Corp. Sec. Litig., 404 F3d at 191 (“In the ordinary case, most work that lead counsel does will typically advance the class’s interests, but the inquiry into non-lead counsel’s work must be more detailed. Non-lead counsel will have to demonstrate that their work conferred a benefit on the class beyond that conferred by lead counsel.” (Emphasis in original.)). Reynolds and Riemer have not satisfied that burden beyond the fact that their duplicative work likely added some marginal persuasive force to the argument. As a result, we conclude that Reynolds’ and Riemer’s fee awards should be reduced to account for their duplication of effort.

Having established that claimants should not receive fees for work on the unsuccessful tax-offset claim and that Reynolds and Riemer should not receive fees for work that duplicated Bennett Hartman’s work on the COLA claim, we must determine how much that work amounts to. The difficulty with making that determination is that claimants provided this court with billing records that largely omit any reference to which claim they were working on. That omission weighs against claimants, because it is their burden to establish the reasonableness of the fees they are requesting. Strawn, 353 Or at 225. Claimants chose not to update their billing records even after this court instructed the special master to make findings of fact on those issues and even after the special master allowed claimants the opportunity to do so. See Moro II, 358 Or at 381 (“[W]e instruct the special master to make findings of fact, if possible, on the extent to which the attorneys duplicated their efforts or directed their efforts toward unsuccessful claims”).

Without updated billing records, the special master determined that Reynolds and Riemer should receive 20 percent of their requested fees and made no findings with regard to Bennett Hartman’s work on the tax-offset claim. As noted, Reynolds and Riemer object to the special master’s reduction, and state respondents again argue that those individuals should receive no fees and that Bennett Hartman’s award should be reduced to account for duplication, as well as for work on the tax-offset claim.

We agree with some of respondents’ arguments regarding the special master’s determinations. Moreover, because claimants have the burden of establishing their entitlement to fees and the reasonableness of their request, the lack of more specific billing records weighs against them. Although we have some, often nonspecific, time-keeping records, as well as the transcript of the special master’s hearing on the merits and the briefing before the special master and in this court, adjustments to the fee requests to account for duplication and work on the unsuccessful claims are admittedly rough. Nevertheless, we agree with the special master that adjustments are appropriate for time spent on the unsuccessful tax-offset claim and for duplication of work. We also agree with the special master’s implicit finding that Bennett Hartman acted as lead counsel throughout the proceeding.

Our review of the record and the proceedings in the litigation suggests substantial duplication by Reynolds and Riemer of the arguments of lead counsel concerning the COLA and substantial time spent on the tax-offset issue. The latter, of course, is not surprising, as Reynolds and Riemer, PERS retirees now living outside the state, are directly affected by that change in PERS benefits. As to the duplication of effort on the COLA, Reynolds’ and Riemer’s submissions do not identify any specific value that their work added to petitioners’ case or any novel legal or factual argument that Bennett Hartman did not make and that this court relied on in its decision. Respondents assert that Reynolds and Riemer should receive no fee award because all their work essentially duplicated Bennett Hartman’s work. In our view, as noted, their work likely added some marginal persuasive force to petitioners’ arguments. In these circumstances, we conclude that probably 90 percent of the time incurred by Reynolds and Riemer essentially duplicated Bennett Hartman’s work on the COLA issue or was spent on the tax-offset issue. Therefore, Reynolds and Riemer are entitled to 10 percent of their requested time. That entitles Reynolds to compensation for 56.2 hours, which, at $500 per hour, results in a lodestar amount of $28,100. Riemer is entitled to compensation for 26.5 hours, which, at $350 per hour, results in a lodestar of $9,275.

As noted, the special master made no adjustment to the Bennett Hartman request for work on the tax-offset claim. Based on our review of the record, and again considering the burden on the claimants to prove their claim for fees, we find that 20 percent of the time submitted by Bennett Hartman went to work on the tax-offset claim that did not overlap with the COLA claim. Therefore, Bennett Hartman is entitled to 80 percent of its requested time. That entitles Bennett Hartman to compensation for 1,355 hours, which results in a lodestar of $560,399.46.

3. Fee multiplier

The final component in determining a reasonable fee award is the multiplier. The special master recommended awarding Bennett Hartman a 1.5 fee multiplier for the exceptional success of the litigation and, in his alternative recommendations, recommended no fee multiplier for Reynolds and Riemer. Bennett Hartman does not object to the recommended fee multiplier of 1.5. Reynolds and Riemer argue that they are entitled to the same fee multiplier as Bennett Hartman because they shared in the same exceptional success of the litigation.

The grounds for a fee multiplier have been stated differently in different cases. In Strawn, this court noted that a fee multiplier may be justified to account for the risk of nonpayment in a contingency fee case. 353 Or at 226. In

Strunk III, the court justified a fee multiplier based on the “exceptional success” of the litigation, securing over $1 billion in benefits for PERS members. 343 Or at 246. The court then stated that “factors such as the difficulty and complexity of the issues involved in this case, the value of the interests at stake, as well as the skill and professional standing of lawyers involved also support an enhancement of fees.” Id.

In this case, we place greater weight on the risk of nonpayment, because the additional factors discussed in Strunk III were already considered when determining reasonable hourly rates for claimants’ work. As noted, none of the claimants rely on a market rate that reflects a negotiated rate with a client. So the hourly rates are a construction based on the same type of factors noted in Strunk III to calculate a fee multiplier. That is particularly true because, as noted above, the stakes of the litigation are a factor in determining the attorneys’ hourly rates. Although the monetary impact of this case will, over time, be larger than the monetary impact of the Strunk litigation, that impact reflects the higher dollar value of the COLA change that the court invalidated, rather than a more exceptional success by the attorneys as compared to the attorneys in the Strunk litigation. 7 The justification for a higher hourly rate is that we would have expected another attorney who could also command that rate to reach a similar result in this case.

Nevertheless, a fee multiplier may be justified when the attorney’s payment is based on a contingency-fee arrangement or there is otherwise a delay in getting paid. Bennett Hartman was on a quasi-contingency-fee arrangement, because, although it was paid its normal union rates, it retained the right to seek to recover and keep market rates in a fee request, if it prevailed. We therefore find that Bennett Hartman is entitled to the fee multiplier of 1.5 recommended by the special master.

Reynolds and Riemer, however, were not working based on a contingency-fee or even quasi-contingency-fee arrangement. Instead, as self-represented parties, they did not have a fee arrangement at all. Therefore, the grounds that entitle Bennett Hartman to a 1.5 fee multiplier do not apply to Reynolds and Riemer. And the other factors justifying a fee multiplier in Strunk were already used to calculate their reasonable hourly rates. Therefore, consistent with the special master’s recommendation, we find that Reynolds and Riemer are not entitled to a fee multiplier.

Based on those findings, and our findings above establishing the reasonable rates and reasonable time for claimants, we conclude that Bennett Hartman is entitled to $840,599.19 in attorney fees, Reynolds is entitled to $28,100 in attorney fees, and Riemer is entitled to $9,275 in attorney fees. When combined with the cost awards noted above, we conclude that each claimant is entitled to the following awards: Bennett Hartman, $902,665.32; Reynolds, $29,314.48; Riemer, $10,434.15; Jones, $1,379.24.

D. Funding the Fee Award

Having calculated the amounts to which claimants are entitled, we turn to the issue of the appropriate source of payment. The restitutionary principles underlying an award of fees from a common fund suggest that the award be paid from the PERF in a manner that affects only the PERS members who benefit from the litigation — and not PERS members who do not benefit from the litigation or PERS employers. And, ideally, PERS members would contribute to the fee award in proportion to the benefits that they receive. See Restatement § 29(2) (beneficiaries may be required to pay from the “common fund for their benefit, in proportion to their respective interests therein”).

But there are administrative obstacles to reaching those ideals — namely, there is no segregated fund consisting of the increased COLAs that PERS members who will benefit from the litigation will receive over time. (Indeed, because of the different terms of service and retirement status of those members, there could not be such a fund.) Instead, the PERF is made up of numerous accounts that are designated for different purposes. According to a declaration submitted by PERS Assistant Chief Administrative Officer Mary Dunn, there are three accounts (or categories of accounts) containing money belonging to PERS members. Retired members have money in the “benefits-in-force” reserve. Nonretired Tier One and Tier Two members have money in their “member accounts.” And all PERS members who earned benefits after the 2001 creation of the Oregon Public Service Retirement Plan (OPSRP), which includes Tier One, Tier Two, and OPSRP members, have money in the Individual Account Program.

Money in those accounts is invested, and the investment income is credited back into the accounts. According to Dunn, PERB can take money out of the investment income before it is credited back into the specific accounts and use that money to pay for the award of fees and costs. And, because the credits are made proportional to each member’s interest, withdrawing money from the investment income is automatically ratable (in the same way that requiring a corporation to pay a fee award in a common-fund or substantial-benefit case is automatically ratable to each shareholder’s interest in ownership).

The problem with that approach is that the fee award must be paid now, while the actual value of the COLA benefits to specific PERS members will depend on which tier the member is in, as well as on future events. Tier One and Tier Two members, but not OPSRP members, are entitled to their COLA “bank” established prior to the amendments at issue, which increases the value of the benefits. Moro I, 357 Or at 186-87 (describing COLA bank). Further, the earlier a member retires and begins to receive benefits, the more years he or she will receive COLA adjustments. As a result, the COLA benefits will comprise a larger portion of those members’ expected PERS benefits.

As an alternative, Bennett Hartman suggests that at least some of the money for the fee award could be drawn from the PERS contingency reserve. The contingency reserve is funded through money that PERB sets aside from investment income in certain years. ORS 238.670(l)(a). One of the statutory purposes for the contingency reserve is “[t]o pay any legal expenses or judgments that do not arise in the ordinary course of adjudicating an individual member’s benefits or an individual employer’s liabilities.” ORS 238.670(l)(b). This fee award fits that statutory description. Respondents object, however, noting that the contingency reserve is funded through investment income from the entire PERF, which includes funds containing employer contributions that have not yet been allocated to other accounts. If the fee award is paid from the contingency reserve, they point out, then PERB may have to make larger future payments into the contingency reserve, which will be taken in part from employer contributions. The effect of that would be to marginally increase the amount that employers will need to contribute in the future.

The special master did not make any finding as to which PERF accounts would be appropriate to use and how much money should be used from each account. Instead, he recommended ordering PERB to pay the award and to resolve later disputes that arise, if any, about the manner in which it had paid for the award.

We agree that the better course is to allow PERB to determine, consistent with its statutory authority and fiduciary obligations, how best to allocate the burdens of the fee award among the accounts, including the contingency reserve, held within the PERF. Our cases recognize that trust law principles and applicable statutes give PERB discretion to make reasonable decisions in operating PERS. See White v. Public Employees Retirement Board, 351 Or 426, 440-41, 268 P3d 600 (2011) (discussing PERB’s duties and authority). That authority to operate the system includes making decisions related to litigation, and, as long as PERB acts reasonably and consistently with statutory requirements and its fiduciary duty to members, such decisions ordinarily will be upheld, unless PERB has abused its discretion. Id. at 442-45 (discussing PERB’s authority to conduct and settle litigation). We direct PERB to pay the amounts ordered from such PERF accounts as it deems appropriate, in the exercise of its discretion and subject to its statutory and fiduciary obligations and the principles discussed in this opinion.

To summarize: We conclude that Bennett Hartman is entitled to $902,665.32 in costs and attorney fees; Reynolds is entitled to $29,314.48 in costs and attorney fees; Riemer is entitled to $10,434.15 in costs and attorney fees; and Jones is entitled to $1,379.24 in costs. PERB shall determine, consistently with this opinion, how to pay the award from the accounts held in PERF and shall pay the amounts awarded.

5 Those broad policy grounds have been applied in other federal cases to deny fees to self-represented attorneys. See, e.g., In re Currency Conversion Fee Antitrust Litig., 263 FRD 110, 132 (SDNY 2009), aff’d sub nom Priceline.com, Inc. v. Silberman, 405 Fed Appx 532 (2d Cir 2010); In re Texaco Inc. S’holder Derivative Litig., 123 F Supp 2d 169, 173 (SDNY 2000), aff’d, 28 Fed Appx 83 (2d Cir 2002).

1 “State respondents” are the State of Oregon, Governor Kate Brown, Attorney General Ellen Rosenblnm, the Public Employees Retirement Board, and the Public Employees Retirement System. “County/school district respondents” are Linn County, Estacada School District, Oregon City School District, Ontario School District, West Linn School District, Beaverton School District, and Bend School District as well as intervenors Oregon School Boards Association and Association of Oregon Counties. Further, respondent Tualatin Valley Fire and Rescue joined in the objections filed by state respondents and county/school district respondents.

2 Riemer is the only party disputing the special master’s recommendation with regard to costs. He argues that he is entitled to costs under the common-fund doctrine, rather than the more limited costs available under ORAP 13.05, as applied by the special master. For the reasons explained below, 360 Or at 478-83, we agree that Riemer is entitled to costs under the common-fund and substantial-benefit doctrines and, therefore, award him the costs that he requested: $1,159.15.

3 In this case, as in the Strunk litigation, the Bennett Hartman firm’s contract with the union that paid its fees provided that the firm would, on its own behalf, seek an award of attorney fees if the petitioners prevailed. Thus, the award that the firm seeks here is on its own behalf, not on behalf of the individual clients or the union that paid its bills. Under the contract, the firm is obligated to repay from any fee award the amount that it was paid by the union, but it is entitled to retain any additional fees that are awarded.

4 See ORS 9.160(1) (providing that, with exceptions, only active members of the Oregon bar may “practice law” in Oregon); ORS 9.241(1) (allowing out-of-state attorney to practice in Oregon courts only if “the attorney is associated with an active member of the Oregon State Bar”); ORAP 8.10(4) (allowing out-of-state attorney to “appear by brief and argue the cause in a proceeding before an appellate court” if the attorney complies with UTCR 3.170); UTCR 3.170 (requiring out-of-state attorney to be admitted pro hac vice).

6 The rates attributed to other attorneys at Bennett Hartman were all within normal ranges for their experience levels.

7 In Strunk III, the court awarded the fee multipliers sought by the claimants. 343 Or at 246. For some, that was a 2.0 fee multiplier; and for others (including Bennett Hartman), that was a 1.5 fee multiplier. Id. at 233.