Four law firms representing class plaintiffs in this case appeal from the district court’s order allowing fees to be paid out of a fund created by the settlement of the class action suit.^[1]

Participants in and beneficiaries of the Simmons Mattress Company’s employee stock ownership plan (“ESOP”), brought this action alleging that the defendants^[2] violated their fiduciary obligations under the Employee Retirement Income Security Act (“ERISA”) by causing or permitting shares of Simmons stock to be sold to the ESOP at a price above fair market value. The class plaintiffs sought to recover damages from the sale of stock. In March 1993, the parties settled the suit for a total of $15,448,304.66. A settlement fund was created for the benefit of the plaintiff class, and the parties agreed that class counsel would seek its fees and expenses from the fund.

Plaintiffs’ counsel then petitioned the district court for an award of fees and expenses from the settlement fund. Counsel requested attorney’s fees of $2.85 million, which represented 18.45% of the fund, or the lodestar figure^[3] of $1,863,838.75 enhanced by a risk multiplier of 1.53.^[4]

The district court first determined that the rates and hours documented by the attorneys for use in calculating the lodestar were “fair and reasonable.” Thus, it approved counsel’s suggested lodestar of $1,863,838.75. The court then turned to decide whether it would increase the award with a multiplier to reflect the risk the plaintiffs lawyers undertook when they decided to pursue the case. The court stated that it was “of the opinion that counsel have billed at their usual hourly rates and have been recompensed with this fee award for the full fees for services expended in this matter,” and found “no compelling reason to provide that multiplier which has been requested in this matter.” Having denied counsel the multiplier, the district court accordingly awarded attorney’s fees only in the amount of the lodestar.

Standard of Review

We review a district court’s award of attorney’s fees for abuse of discretion. Harman v. Lyphomed, Inc., 945 F.2d 969, 973 (7th Cir.1991). We will find an abuse of discretion when the district court reaches an erroneous conclusion of law. Id.; see also Skelton v. General Motors Corp., 860 F.2d 250, 257 (7th Cir.1988), cert. denied, 493 U.S. 810, 110 S.Ct. 53, 107 L.Ed.2d 22 (1989). We review de novo the court’s method of determining fees, to determine that it comports with approved procedures for calculating awards. Harman, 945 F.2d at 973.

Analysis

In the United States, parties to a lawsuit usually bear their own expenses, regardless of which party prevails. This is sometimes called the “American Rule.” See Alyeska Pipeline Serv. Co. v. Wilderness Soc’y, 421 U.S. 240, 247, 95 S.Ct. 1612, 1616, 44 L.Ed.2d 141 (1975). Congress, however, has created exceptions to the American Rule, by inserting fee-shifting provisions in certain statutes, such as the one in ERISA, at 29 U.S.C. § 1132(g). These fee-shifting statutes were enacted for the purpose of encouraging the private prosecution of certain favored actions, by requiring defendants who have violated plaintiffs’ rights to compensate plain[*563]tiffs for the costs they incurred to enforce those rights. County of Suffolk v. Long Island Lighting Co., 907 F.2d 1295, 1327 (2d Cir.1990); Skelton, 860 F.2d at 253.

Courts may also make an exception to the American rule based on equitable doctrines, such as the “common fund” or “equitable fund” doctrine.^[5] When a case results in the creation of a common fund for the benefit of the plaintiff class, the common fund doctrine allows plaintiffs’ attorneys to petition the court to recover its fees out of the fund. In such a case, the defendant typically pays a specific sum into the court, in exchange for a release of its liability. The court then determines the amount of attorney’s fees that plaintiffs’ counsel may recover from this fund, thereby diminishing the amount of money that ultimately will be distributed to the plaintiff class. The common fund doctrine is based on the notion that not one plaintiff, but all “ ‘those who have benefited from litigation should share its costs.’ ” Skelton, 860 F.2d at 252 (citations omitted).

I.

The issue in this case is whether the district court abused its discretion by failing to award appellants a multiplier for risk. However, we must first determine whether this case should be governed by common fund principles or by statutory fee-shifting principles, for “when a case is initiated under a statute with a fee-shifting provision [but] is settled with the creation of common fund, the question may arise whether statutory fee principles should govern in whole or in part the attorney fee award.” Skelton, 860 F.2d at 254. In such a situation, “the statute must control and the doctrine must be deemed abrogated to the extent necessary to give full effect to the statute.” Long Island Lighting, 907 F.2d at 1327.

In Skelton v. General Motors Corp., plaintiffs brought a class action against GM under the Magnuson-Moss Warranty-Federal Trade Commission Improvement Act. The Magnuson-Moss Act contained a fee-shifting provision; however, the plaintiff class settled the case through the creation of a common fund. We held there that “when a settlement fund is created in exchange for release of the defendant’s liability both for damages and for statutory attorney’s fees, equitable fund principles must govern the court’s award of the attorney’s fees.” Skelton, 860 F.2d at 256 (citing In re Fine Paper, 751 F.2d 562, 582-84 (3d Cir.1984); Lindy Bros. Builders, Inc. v. American Radiator & Standard Sanitary Corp., 487 F.2d 161, 164-65 (3d Cir.1973)).

Our holding in Skelton applies to this ease. Even more so than the fee-shifting provision in Skelton, the terms of ERISA’s fee-shifting provision do not purport to control fee awards in cases settled with the creation of a common fund, nor does the operation of common fund principles in this case conflict with the provision’s intended purpose. Section 1132(g)(1) provides simply, “... [i]n any action under this title by a participant, beneficiary, or fiduciary, the court in its discretion may allow a reasonable attorney’s fee and costs of action to either party.” This fee-shifting provision is unusual because it does not specify that prevailing plaintiffs may be awarded attorney’s fees, but instead states that attorney’s fees may be awarded to either party.^[6] However, we have interpreted the provision to mean that although winning defendants may, in rare cases, collect fees from losing plaintiffs, the ERISA fee-shifting statute primarily benefits winning plaintiffs. See generally Chicago Painters & Decorators Pension, Health and Welfare, and Deferred Savings Plan Trust Funds v. Karr Bros., 755 F.2d 1285 (7th Cir.1985); Bittner v. Sadoff & Rudoy Indus., 728 F.2d 820 (7th[*564] Cir.1984); Marquardt v. North Amer. Car Corp., 652 F.2d 715 (7th Cir.1981).

The settlement agreement approved by the court provides that the defendants are released from potential liability for statutory attorney’s fees and that class counsel may instead petition the court for an award of fees from the settlement fund. Thus, the settlement agreement seems to anticipate that the amount paid by the defendants into the fund includes an unspecified sum for class counsel’s fees.

An award of attorney’s fees from the fund would therefore be consistent with the goal of the fee-shifting provision to allow “‘the offending party [to] bear the costs of the award rather than ... plan participants.’ ” Bowen v. Southtrust Bank of Alabama, 760 F.Supp. 889, 894 (M.D.Ala.1991) (citing Eaves v. Penn, 587 F.2d 453, 464 (10th Cir.1978)). Furthermore, an award of fees from the settlement fund comports with the fee-shifting policy of enabling meritorious plaintiffs who would not otherwise be able to afford to bring a lawsuit under ERISA, to pursue their claims. See Berlind, supra note 6, at 1060-61 (“Applying the common benefit doctrine reflects the statute’s purpose.... Nothing in ERISA indicates that Congress intended to preempt the common benefit doctrine.”).

II.

Our determination that common fund principles control this case is crucial in the risk multiplier context, because in City of Burlington v. Dague, — U.S. -, 112 S.Ct. 2638, 120 L.Ed.2d 449 (1992), the Supreme Court has held that when granting attorney’s fees under the fee-shifting provisions of the Solid Waste Disposal Act and the Clean Water Act, courts may not enhance a fee award above the lodestar amount to reflect risk of loss or contingency. The fee-shifting provisions in the statutes in Dague are “typical of most federal fee-shifting provisions,” and the decision has been interpreted to preclude generally the use of risk multipliers in fee-shifting cases. See, e.g., Gusman v. Unisys Corp., 986 F.2d 1146, 1150 (7th Cir.1993).

Although the district court in this case awarded class counsel an amount of fees from the settlement fund, the court seemed to regard this case as governed at least in part by statutory fee-shifting principles. To support its denial of class counsel’s risk multiplier, it referred to Johnson v. Georgia Highway Express, Inc., 488 F.2d 714 (5th Cir.1974) and Gusman v. Unisys Corp., 986 F.2d at 1146. Johnson and Gusman are both statutory fee-shifting cases, and neither case involves a settlement fund created for the benefit of a plaintiff class. Furthermore, in Gusman, we noted parenthetically that Dague would have precluded the granting of a risk multiplier under a fee-shifting statute. 986 F.2d at 1150.

In any event, we conclude that the holding in Dague should not extend to this case. As we have explained, common fund principles properly control a case which is initiated under a statute with a fee-shifting provision, but is settled with the creation of a common fund. Dague, by its terms, applies only to statutory fee-shifting cases, and its reasoning is largely based on the statutory language of fee-shifting provisions.

Moreover, the policy considerations in Da-gue that militate for forbidding risk multiples in statutory fee-shifting eases have little force in common fund cases. Unlike in fee-shifting cases, the fee award in a common fund case is ultimately charged against the plaintiffs’ fund, rather than directly against the defendant. Thus, the defendant’s liability is fixed — it cannot exceed the amount of the common fund the defendant has agreed to pay. Because of this feature of common fund cases, “there is no direct or immediate danger of unduly burdening the defendant” with a multiplier to compensate for risk of loss. Skelton, 860 F.2d at 254. The awarding of a multiplier simply diminishes the amount ultimately recovered by each member of the plaintiff class.

For example, in a common fund case, there is a diminished fear that by awarding a multiplier to plaintiffs’ counsel, a defendant “would in effect pay for the attorney’s time (or anticipated time) in cases where his client does not prevail.” Dague, — U.S. at-, 112 S.Ct. at 2643. The Dague Court ex[*565]plained that “[a]n attorney operating on a contingency-fee basis pools the risks presented by his various cases: cases that turn out to be successful pay for the time he gambled on those that did not.” Id. Thus, by requiring defendants to compensate plaintiffs’ counsel for their risk of loss, the Court concluded, defendants would essentially be subsidizing plaintiffs’ attorneys for unsuccessful lawsuits against other defendants. In a common fund case, however, because compensation for risk is charged against the plaintiff class, defendants would not be forced to subsidize directly plaintiffs’ attorneys’ losing endeavors. Further, there is no injustice in requiring plaintiff class members to shoulder the burden of compensating counsel for prosecuting the class’ case without any assurance of compensation.

Having determined that risk multipliers remain available in common fund cases after Dague, we further note that we have held, in a pre-Dague case, that a risk multiplier is not merely available in a common fund case but mandated, if the court finds that counsel “had no sure source of compensation for their services.” In re Continental Illinois Securities Litigation, 962 F.2d 566, 569 (7th Cir.1992); see also Harman, 945 F.2d at 975-76; Skelton, 860 F.2d at 257-58. Moreover, we have observed that “the need for such an adjustment is particularly acute in class action suits. The lawyers for the class receive no fee if the suit fails, so their entitlement to fees is inescapably contingent.” In re Continental Illinois Securities Litigation, 962 F.2d at 569. Hence, because we conclude that this case is properly controlled by common fund principles and that Dague has no application to epmmon fund cases, we must vacate the district court’s award.

III.

On remand, the district court must consider how much compensation class counsel should receive for incurring the risk of nonpayment when it took the suit. Because the district court has already computed and approved of counsel’s lodestar, the court may, under the “lodestar method,” need only determine what multiplier to use to adequately compensate counsel for contingency.

Although this determination is inevitably somewhat subjective, we have set some guidelines for district courts to follow. “The district court must balance the competing goals of fairly compensating attorneys for their services rendered on behalf of the class and of protecting the interests of the class members in the fund.” Skelton, 860 F.2d at 258 (citation omitted). A court must assess the riskiness of the litigation by measuring the probability of success of this type of case at the outset of the litigation. Harman, 945 F.2d at 976; Skelton, 860 F.2d at 258. Because class counsel have requested a multiplier of 1.53, the district court need not worry about exceeding what we have suggested is a sensible ceiling of double the lodestar. See Skelton, 860 F.2d at 258.

Class counsel have urged us to declare a rule in this case that would compel district courts to use the “percentage-of-reeovery method” to award attorney’s fees in all common fund cases. They point out that at least two other circuits have done so, citing Swedish Hospital Corp. v. Shalala, 1 F.3d 1261 (D.C.Cir.1993) and Camden I Condominium Ass’n v. Dunkle, 946 F.2d 768 (11th Cir.1991). We, however, have already considered this suggestion and rejected it in Harman v. Lyphomed, Inc. There, we stated:

Although the lodestar approach certainly has problems, we think it premature to banish it now. First, the lodestar process initially responded to concerns that a percentage approach resulted in over-compensation for attorneys. Requiring an explicit accounting of hours and rates provides greater accountability. Second, the separation of multiplier and lodestar provides — in theory at least — a means of accounting both for the hours and rate reasonable in the type of case and for the risk an attorney assumes in undertaking a ease.

In addition, the lodestar encourages attorneys to assess the marginal value of continuing to work on the case. Assured of recovering a “reasonable fee” if victorious, an attorney will not find it worthwhile to spend time that does not add to the[*566] value of the suit, since there is significant risk that the court will in the end disapprove such time. A frequently cited advantage to the percentage method — its alignment of class attorney’s interests with those of the client — may, as the district court noted, encourage holding out or early settlement on the basis of the expected value of the recovery. These incentive effects ... indicate at least that the door should not be closed on lodestar awards.

Harman, 945 F.2d at 974.

It bears reiterating here that we do not believe that the lodestar approach is so flawed that it should be abandoned. Instead, we are of the opinion that both the lodestar approach and the percentage approach may be appropriate in determining attorney’s fee awards, depending on the circumstances. We therefore restate the law of this circuit that in common fund cases, the decision whether to use a percentage method or a lodestar method remains in the discretion of the district court. We recognize here, as we did in In re Continental Illinois Securities Litigation, that there are advantages to utilizing the percentage method in common fund cases because of its relative simplicity of administration. However, because of the district court’s familiarity with this litigation, we leave the decision as to which method is the most efficient and suitable to this case up to the district court.

Conclusion

For the foregoing reasons, the district court’s award of attorney’s fees is VACATED, and the matter is Remanded for proceedings consistent with this opinion.