In re PAOLI RAILROAD YARD PCB LITIGATION. Mabel Brown; George Burrell; Wallace D. Cummins; Sylvan C. Cohen; K. Louise Jones; James Lament; Christopher S. Brown; Cathlene Brown; Craig Brown; Margherita Barbetta; Mary Retta Johnson; Celeste Brown; Clemmon Brown; Cloyd Brown; Curtis Brown; John Ingram; Mary Alice Knight; William Butler; Matthew Cunningham, Appellants, 221 F.3d 449

OPINION OF THE COURT BECKER, Chief Judge. These consolidated appeals from an order of the District Court upholding a large award of costs against the losing parties after trial in a long-running toxic tort case present a number of issues pertaining to costs awards. Most of them are quite technical. Of especial interest and importance are questions dealing with the extent to which the court may consider the relative indigence of the losing party in determining the amount of the costs award, and the ability of the court to impose liability jointly and severally among losing parties.

The prevailing party in a federal civil action is entitled to costs, “as of course,” except in cases in which either a federal rule or statute otherwise provides, or in cases in which the district court “otherwise directs.” FED. R.CIV.P. 54(d)(1). The clerk of court is charged with taxing such costs, but that decision “may be reviewed by” the district court, as it was here. Id.

This appeal by the Plaintiffs arises from a district court order that reviewed the Clerk of Court’s decision and awarded $154,129.30 in costs to the Defendants after the Defendants prevailed in the underlying litigation against the Plaintiffs. If the case caption seems familiar, it is because this is the fourth published opinion that this court has filed in the litigation. We reach several separate holdings on our way to affirming in part and vacating in part the costs order of the District Court. First, although we conclude that Eastern District of Pennsylvania Civil Procedure Local Rule 54.1(

b) is invalid insofar as its limitations provision conflicts with that of Federal Rule of Civil Procedure 54(d)(1), we hold that the District Court properly exercised its discretion under Rule 54(d)(1) to hear the Plaintiffs’ challenge, which was untimely under the federal rule, but timely under the local rule. Second, we hold that the District Court erred in concluding that it did not have the power to consider evidence that the Plaintiffs submitted to the Court in support of their motion, but did not present to the Clerk of Court.

A district court reviewing a clerk’s taxation of costs has the power to consider such evidence. In the costs award context, the district court is conducting a de novo review of an essentially ministerial act of the clerk of court. It is important that the district court have access to all evidence relevant to help it insure that the imposition of a costs award is equitable. Third, while we reaffirm that a district court may not consider the disparity in wealth between the prevailing and non-prevailing parties in imposing costs, see Smith v. Southeastern Pa. Transp.

Auth., 47 F.3d 97, 99 (3d Cir.1995) (per curiam), we hold that a district court may consider the non-prevailing party’s indigency (or inability to pay costs) in fashioning a costs award. In this case, it appears the District Court misapprehended its power when it declined to consider the Plaintiffs’ indigency in imposing the large costs award that it did. We also conclude that a district court may not consider the good faith of the losing party or the closeness or complexity of the underlying case in awarding costs.

These factors are only relevant insofar as they relate to the prevailing party’s bad faith or unclean hands, a separate costs award factor described herein. Finally, although a district court generally has broad discretion in imposing costs, we hold that the District Court abused its discretion in one aspect of its imposition of joint and several liability for the costs award among the nineteen Plaintiffs. As our prior opinions in this case reflect, see infra note 1 (collecting these opinions), this toxic tort case has primarily involved a host of personal injury and medical monitoring claims.

Plaintiffs Sylvan Cohen and James Lament sued only for property damages, not personal injuries or medical monitoring, and it was in defending against the latter two classes of claims that the Defendants incurred close to 94% of their costs. We conclude that it was inequitable to saddle Cohen and Lament with one hundred percent of the Defendants’ costs (or $154,129.30 in joint and several liability), when they were responsible for only a discrete and recognizable fraction of that sum, and that to do so amounted to an abuse of discretion.

As for the other seventeen (personal injury/medical monitoring) Plaintiffs, we hold that the District Court did not abuse its discretion in imposing liability jointly and severally. Even though these Plaintiffs’ cases were consolidated for efficiency’s sake and several of the seventeen Plaintiffs appear unable to pay any portion of the costs award, the costs that the Defendants incurred in defending against these Plaintiffs’ actions appear so intertwined with each other that the District Court did not abuse its discretion by not disaggregating them.

In accordance with these holdings, the order of the District Court will be affirmed in part and vacated in part, and the case remanded for still further proceedings. I. Factual and Procedural History The Plaintiffs are nineteen individuals who either live, formerly lived, or owned property near the Paoli Railroad Yard (the “Rail Yard”).

The named Defendants in this appeal (Monsanto Co., Westinghouse Electric Corp., and General Electric Co.), as well as several other defendants, were accused by the Plaintiffs of releasing large quantities of toxic chemicals, including po-lychlorinated biphenyls, or PCBs, at the Rail Yard. The Plaintiffs alleged that these toxic chemicals injured them personally and/or harmed their property. The Plaintiffs’ cases were consolidated for trial, but depending on their harm — personal injuries or property damages — involved requests for distinctly different remedies.

Those that suffered personal injuries asked, inter alia, for medical monitoring and money damages to compensate them for their personal injuries. Those that did not live near the site and suffered only injuries to their property, such as Cohen and Lament, asked only for money damages to compensate for their property losses. A. As we noted above, this case has been before this court a number of times. After some thirteen years of litigation, the Defendants ultimately prevailed at trial, and we affirmed the judgment entered on that verdict. See Paoli IV, 113 F.3d at 464.

Thus, the federal case is at an end except for this costs award dispute. The same cannot be said for Paoli Rail Yard litigation pending in the Pennsylvania state courts; several individual actions and a class action are still in the pre-trial stage in the Chester County Court of Common Pleas. Having prevailed at trial in the District Court, the Defendants filed their bill of costs with the Clerk of Court, as was their right under Fed.R.CivP. 54(d)(1). After the Plaintiffs lost their appeal to this court challenging the outcome of the trial, they filed objections to the bill of costs.

On the following day, the Clerk held a telephonic conference regarding the taxing of costs. Briefing followed, and the Clerk entered an order taxing costs in favor of Defendants in the amount of $184,675.12. The Clerk stated that the Plaintiffs were jointly and severally liable for this amount. B. Both sides filed motions with the District Court seeking review of the Clerk’s decision. The Defendants sought review of the Clerk’s failure to address five categories of costs.

The District Court rejected some of the Defendants’ arguments and accepted others, but, in total, adjusted upward the Clerk’s suggested award amount by several thousand dollars. Though the Defendants were denied certain costs by the District Court, they do not challenge those denials on appeal. The Plaintiffs had less success on review.

In their motion to the District Court, they set forth the following five arguments in support of their challenge of the Clerk of Court’s decision: (1) it would be inequitable to tax any costs against any of the Plaintiffs in light of their financial status, the magnitude of the costs at issue, and the merit of their claims; (2) it would be inequitable to tax costs against one Plaintiff — e.g., Sylvan Cohen — that were incurred in the other Plaintiffs’ lawsuits; (3) costs should not be taxed that were incurred by one or more settling defendants that had waived any such claim-i.e., those defendants other than the named Defendants in the present appeal; (4) certain of the costs that the Defendants sought were inappropriate; and (5) the Defendants cannot be said to have prevailed against Cohen, who voluntarily dismissed his individual action so that he could participate as an absent class member in any class action certified in state court, and, thus, no costs should be taxed against him.

A preliminary dispute arose as to whether the Plaintiffs’ challenge was timely. The Defendants argued that the Plaintiffs’ claims need not be addressed because the Plaintiffs filed their challenge four days late. The District Court agreed that the Plaintiffs’ motion was filed late, but nevertheless went on to rule on each of the Plaintiffs’ claims. In fact, the Court ruled favorably to the Plaintiffs with respect to their third argument.

The Court reasoned that, because the Clerk’s award was calculated based on the aggregated costs of five defendants, and two defendants had settled with the Plaintiffs before the costs award order was entered, the costs award should be reduced to three-fifths of the Defendants’ original estimates. The District Court went on, however, to reject each of the Plaintiffs’ additional contentions. First, it held that there was no equitable reason to reduce the Defendants’ total award based on the Plaintiffs’ conduct in the course of litigation or any of the Plaintiffs individual economic circumstances.

Second, in making this determination, it refused to consider several affidavits from many of the Plaintiffs and one of the paralegals working on their case. The paralegal’s affidavit stated that those Plaintiffs could not afford to pay the costs award because of individual economic circumstances detailed in the Plaintiffs’ affidavits. The Court ruled that it would not consider the Plaintiffs’ affidavits because they were not introduced before the Clerk of Court; the Court did not explain why it refused to address the paralegal’s affidavit.

Third, the District Court rejected the contention that the Plaintiffs should not be held jointly and severally liable, even though certain Plaintiffs were deceased (Matthew Cunningham, Wallace Cummins, and Mary Alice Knight), had not advanced personal injury or medical monitoring claims (Cohen and Lament), or had voluntarily dismissed their actions well before the case proceeded to the

summary judgment stage or trial (Cohen). Lastly, the District Court rejected the Plaintiffs’ numerous challenges to specific items of costs allowed in the Clerk of Court’s order. After adjusting the costs award upward for costs that the Clerk errantly omitted, the District Court reduced the award by two-fifths to reflect the fact that two of the original defendants (not involved in this appeal) had settled with the Plaintiffs and were no longer seeking costs. The Court tabulated the final costs award at $154,-129.30, and entered an order on August 3, 1999 that the Clerk of Court tax costs of this amount.

The Defendants have continually represented that they will not attempt to execute on this award unless the Plaintiffs receive compensation from the pending state court claims against the Defendants. See supra note 2 (discussing these state claims). C. The Plaintiffs timely appealed the District Court’s order, and we have jurisdiction under 28 U.S.C. § 1291. The District Court originally had jurisdiction over this case pursuant to 28 U.S.C. § 1331.

After the Plaintiffs’ federal claims were dismissed, the District Court exercised its discretion under the supplemental jurisdiction statute, 28 U.S.C. § 1367, to retain the Plaintiffs’ pendent state law claims.

The appeals are limited to. five issues: (1) whether the Plaintiffs’ motion for review of the Clerk of Court’s taxation of costs was timely; (2) whether the District Court should have considered the additional affidavits that the Plaintiffs’ filed; (3) whether the Court inappropriately failed to weigh certain factors in deciding whether the costs award was equitable; (4) whether the Court erred in holding all of the Plaintiffs jointly and severally liable; and (5) whether the Defendants’ stated intention not to execute the award, unless the Plaintiffs receive compensation from the pending state court claims against the Defendants, has any bearing on the Court’s ability to impose a costs award.

We discuss the arguments pertaining to each of these issues in turn, see infra Parts III-VI—except for those relating to the fifth one, which we now reject summarily as wholly without merit. Before doing so, however, we turn first to a brief history of Rule 54(d)(1), and then to our standard of review in Rule 54(d)(1) cases. II. Legal Background A. The History of Rule 54(d)

(1) Although this case turns on an

interpretation of the text of Fed.R.CivP. 54(d)(1), discussion of the meaning of the rule and its provisions is aided by “a page of history.” New York Trust Co. v. Eisner, 256 U.S. 345, 349, 41 S.Ct. 506, 65 L.Ed. 963 (1921) (“Upon this point a page of history is worth a volume of logic.”) (Holmes, J.). We find this page of history (and the source of the three paragraphs that follow) in Professor Wright’s and his colleagues’ treatise. See 10 Charles Alan Wright, Arthur R. Miller & Mary Kay Kane, Federal Practice And Procedure § 2665, at 199-200 (3d ed.1998).

At early common law, the taxation of costs, in the modern sense of that concept, was unknown. Through a series of statutes, costs in actions at law became available in England first to successful plaintiffs and then to successful defendants. In the period between the enactment of these statutes and the latter part of the nineteenth century, the English law courts awarded costs to the prevailing party. At that latter time, the existing equity practice of treating costs as a matter of judicial discretion was adopted for all of the English courts.

The cardinal principle of the English system has nonetheless been and continues to be one of total reimbursement; that is, if equity does not demand otherwise. In the United States, the English practice generally was adopted by the states at a relatively early time. This meant that during the early part of the nineteenth century, total reimbursement, including attorney’s fees, typically was given to the prevailing litigant. As time progressed, however, the trend in the states was to abolish attorneys’ fees as a reimbursable item by statute or to so limit the amount recoverable as to make it meaningless.

Before the adoption of the federal rules, the federal courts tended to allow costs by treating costs awards as within their inherent power in the absence of express statutory guidance. In practice, this meant that costs were awarded to the prevailing party automatically in cases at law. In cases at equity, however, the federal courts considered themselves free to exercise discretion. At an early date, the federal courts, as had the states, departed from the English practice of attempting to provide the successful litigant with total reimbursement and developed principles limiting the scope of taxable costs.

This admixture of American practices was for the most

part codified in the Federal Rules of Civil Procedure. Today, Rule 54(d)(1) defines the power of a federal court to tax costs. As is discussed below, in terms of defining judicial discretion, the rule is a hybrid of the traditional costs rules that were once employed at law and at equity; in terms of the types of costs allowed, the rule is in keeping with the American tradition of not providing total reimbursement. Rule 54(d)(1) provides, in full: Costs Other than Attorneys’ Fees.

Except when express provision therefor is made either in a statute of the United States or in these rules, costs other than attorneys’ fees shall be allowed as of course to the prevailing party unless the court otherwise directs; but costs against the United States, its officers, and agencies shall be imposed only to the extent permitted by law. Such costs may be taxed by the clerk on one day’s notice. On motion served within 5 days thereafter, the action of the clerk may be reviewed by the court. Fed.R.Civ.P. 54(d)(1) (italicization added).

The emphasized phrase “unless the court otherwise directs” makes the allowance of costs discretionary and thus adopts the practice formerly followed in courts of equity rather than at law; The emphasized phrase “costs ... shall be allowed as of course” reflects the rule at law that costs be awarded to the prevailing party automatically. Thus, “[mjelding these two [traditions], Rule 54(

d) provided a new standard for use of taxing of costs in all [federal] cases.” Smith v. Southeastern Pa. Transp. Auth., 47 F.3d 97, 99 (3d Cir.1995) (per curiam) (footnote omitted). In keeping with the American, as opposed to the British, policy judgment that the imposition of costs should not act as a bar to meritorious litigation, the types of costs recoverable under Rule 54(d)(1) are circumscribed. The “costs” capable of re-coupment under Rule 54(d)(1) are listed in 28 U.S.C. § 1920. See id. at 99 n. 1.

As we have explained, these costs “do not include such litigation expenses as attorney’s fees and expert witness fees in excess of the standard daily witness fee.” Friedman v. Ganassi, 853 F.2d 207, 209 (3d Cir.1988). This means that while a prevailing party is awarded its costs, those costs often fall well short of the party’s actual litigation expenses.

See 10 Wright et al., supra, § 2666, at 202-04; see also 10 Moore’s Federal Practice § 54.103, at 54-174 to 54-197 (3d ed.1999) (discussing exactly what types of “costs,” according to the text of the federal rules, case law, and statute, are recoverable under Rule 54(d)(1), which itself uses the word “costs” as a term of art, rather than to refer to all expenses a prevailing party may incur in a given action).

Expounding on the policy origins and recovery limits inherent in Rule 54(d)(1), Professors Wright, Miller, and Kane observe that [t]here continues to be a debate over the respective merits of the English and American approaches to costs. The English system is asserted to be more equitable inasmuch as it makes the successful party completely whole. It also is said to have the virtue of discouraging the institution of baseless claims and defenses because it imposes the litigation expenses of both parties on an unsuccessful party.

However, this argument probably has little weight in the context of litigants who are extremely wealthy or who are financially incapable of paying costs. The arguments in favor of the American approach are founded on the egalitarian concept of providing relatively easy access to the courts to all citizens and reducing the threat of liability for litigation expenses as an obstacle to the commencement of a lawsuit or the assertion of a defense that might have some merit. 10 Wright et al., supra, § 2665, at 201-02. B. Our Standard of Review in Rule 54(d)

(1) Cases Given the district court’s discretionary equitable power to award costs under Rule 54(d)(1), taxation of costs is reviewed only for abuse of discretion. See Smith v. Southeastern Pa. Transp. Auth., 47 F.3d 97, 99 (3d Cir.1995) (per curiam); see also 10 Moore’s, supra, § 54.100[4][b], at 54-147 to 54-148. As with any exercise of discretion by the district court, when it is within the bounds of the discretion conferred, we may not question it. The definition of those bounds, however, is a question of law. When a court exceeds or misapprehends those bounds, it abuses its discretion. Cf. Koon v.

United States, 518 U.S. 81, 100, 116 S.Ct. 2035, 135 L.Ed.2d 392 (1996) (“A district court by definition abuses its discretion when it makes an error of law.”). Therefore, in reviewing the District Court’s decision to impose costs in this case, we exercise plenary review as to legal questions pertaining to Rule 54(d)(1). In reviewing the District Court’s application of those legal precepts, we reverse only if that application exceeded the bounds of discretion. III. The Timeliness Issue On March 10, 1999, the Clerk of Court entered an order taxing costs in favor of the Defendants.

Rule 54 provides that “costs may be taxed by the clerk on one day’s notice. On motion served within 5 days thereafter, the action of the clerk may be reviewed by the court.” Fed.R.CivP. 54(d)(1) (emphasis added). The Defendants timely filed their challenge on March 18, 1999 — five days after the Clerk “acted” and entered his decision, excluding intermediate Saturdays and Sundays. The Plaintiffs’ counsel did not receive notice of the Clerk’s decision until March 15, 1999 and did not file their challenge to the costs award until March 22, 1999.

When the case came before the District Court, the Defendants argued that the Plaintiffs’ challenge was untimely by four days, and that the Plaintiffs had therefore waived their right to judicial review of the Clerk’s imposition of costs. In response, the Plaintiffs pointed to the Eastern District’s Local Rule 54.1(b). Local Rule 54.1(

b) allows a party to file objections to an order taxing costs “within five (5) days after notice of such taxation.... ” E.D.Pa. Civ.Pro.R. 54.1(b) (emphasis added). Allowing for weekend days, the Plaintiffs filed their challenge within five days of receiving notice that the Clerk had awarded costs. The District Court noted that Local Rule 54.1(b)’s focus on the day on which the parties receive notice of the Clerk’s action is inconsistent with the language of Rule 54(d)(1), which renders determinative the point at which the Clerk acts to tax costs. See In re Paoli R.R.

Yard PCB Litig., No. 86-2229 et al., 1999 WL 569435, at *2-3 (E.D.Pa. Aug.2, 1999). The Court therefore held that insofar as Local Rule 54.1(

b) conflicted with Rule 54(d)(1), the local rule was a nullity. See id. at *3 (citing Lasky v. Continental Prods. Corp., 804 F.2d 250, 255 (3d Cir.1986) (“The primacy of the Federal Rules of Civil Procedure with regard to the procedural aspects of litigation in federal courts is well-settled.”)). Accordingly, the District Court ruled that the Plaintiffs’ motion was untimely, but proceeded to rule on the motion nevertheless. The Plaintiffs challenge the District Court’s reading of Rule 54(d)(1). Because this issue raises a question of the

interpretation of the Federal Rules, our review is plenary. See Kleissler v. United States Forest Serv., 183 F.3d 196, 198 (3d Cir.1999). We agree with the District Court that the Local Rule is inconsistent with the Federal Rule and hence is void. The District Court’s reading of Rule 54(d)(1) was consistent with the plain language of the rule. Moreover, the District Court’s reading of Rule 54(d)(1) comports with that of the only courts of appeals to have addressed the issue, see, e.g., Lorenz v. Valley Forge Ins.

Co., 23 F.3d 1259, 1261 (7th Cir.1994), numerous district court decisions, see In re Paoli, 1999 WL 569435, at *2 n. 4 (collecting cases), and the opinion expressed by the authors of Moore’s Federal Practice treatise, see 10 Moore’s, supra, § 54.100[3], at 54-144 to 54-155 (collecting cases). A number of district courts have expressed disagreement with the District Court’s reading of the rule, see, e.g., Nalco Chem. Co. v. Hydro Techs., Inc., 149 F.R.D. 686, 699 (E.D.Wis.1993); Raio v. American Airlines, Inc., 102 F.R.D. 608, 610 (E.D.Pa.1984); Wright v. McDonald’s Corp., Civ. A.

No. 91-2061, 1995 WL 284216, at *1 (E.D.Pa. May 8, 1995), but these courts seem to ignore the language of Rule 54(d)(1) in order to prevent inequities to parties that do not receive notice of the clerk’s- decision until too late, even though there is a way to prevent these inequities without misreading Rule 54(d)(1). Rule 54(d)(l)’s five-day limitations period is not jurisdictional, and courts may, in their discretion, consider untimely objections.

See 10 Moore’s, supra, § 54.100[3], at 54-144 to 54-145 (collecting cases); see id. at 54-144 (“If a party fails to file a timely objection to the clerk’s taxation of costs, any objection is waived, and the district court is not obliged to review the taxation of costs.”) (collecting cases) (emphasis added).

Instead of distorting the language of Rule 54(d)(1) to insure “fair and reasonable application of the rule,” Wright, 1995 WL 284216, at *1, district courts need only do what the District Court did in this case: exercise their discretion to hear untimely objections when it appears that a delay in the mails or a potentially misleading local rule might have affected the petitioning parties’ right to receive judicial review. The District Court here properly exercised this discretion.

After ruling that the Plaintiffs’ motion was untimely by four days, the Court decided to address the Plaintiffs’ motion and granted, in part, the relief they sought. This approach guards against unduly harsh results without subverting the plain language of the rule. In sum, we hold that the District Court (1) did not err as a matter of law in concluding that the Plaintiffs’ motion was untimely or that Local Rule 54.1(

b) was invalid insofar as it conflicted with Rule 54(d)(1); and (2) did not abuse its discretion nor commit legal error in deciding to hear the Plaintiffs’ motion. IV. The Power of the District Court to Expand the Record in Costs Cases In pressing their costs award defense before the District Court, the Plaintiffs submitted, inter alia, that they could not afford to pay the costs in question. To demonstrate their lack of ability to pay, the Plaintiffs submitted materials, including numerous affidavits, which are summarized in the margin, that describe many of the Plaintiffs’ limited financial means.

However, none of these affidavits had been submitted to the Clerk of Court, though the Plaintiffs had provided the Clerk with a similar affidavit signed by a paralegal to one of their attorneys. The paralegal’s affidavit attested to the fact that several of the Plaintiffs were in dire financial straits, were deceased, or were in frail health. See App. at 280-81. The Defendants objected to the introduction of Plaintiffs’ affidavits because they were not part of the record before the Clerk of Court.

They argued that it was improper for the District Court to consider such evidence in reviewing the Clerk’s award of costs. The District Court accepted this argument, reasoning that “[wjhile the court has the discretion to consider the losing party’s inability to pay when reviewing the taxation of costs, the party asserting the lack of funds must demonstrate his indigence.” That Plaintiffs failed to present the above affidavits to the Clerk to support their purported financial status leaves this Court with no suitable basis for reducing the costs sought by Defendants. Cf. Samar Fashions, Inc. v.

Private Line, Inc., 116 B.R. 417, 420 (E.D.Pa.1990) (district court’s review of bankruptcy appeal is limited to the record before the bankruptcy court). In re Paoli R.R. Yard PCB Litig., No. 86-2229 et al., 1999 WL 569435, at *4 (E.D.Pa. Aug.2, 1999) (citation and footnote omitted). The District Court further opined that the Plaintiffs “belated affidavits” were “sketchy.” Id. at *4 n. 5. The District Court apparently did not consider the paralegal’s affidavit at all. A. The Plaintiffs challenge the District Court’s

interpretation of its own review power under Rule 54(d)(1), which the Court interpreted as precluding it from hearing evidence beyond that presented to the Clerk of Court. Our standard of review in addressing the merits of this

interpretation is plenary because it involves a construction of a Federal Rule. See Kleissler v. United States Forest Serv., 183 F.3d 196, 198 (3d Cir.1999). The. Defendants assert that the District Court understood the bounds of its power and exercised its discretion to ignore the Plaintiffs’ “belated” and “sketchy” affidavits. If this were the case, our standard of review would be for abuse of discretion. The Defendants’ argument, however, is too facile.

The District Court stated that, because the Plaintiffs failed to present their affidavits to the Clerk of Court, the Court had “no suitable basis” on which to reduce the costs award based on the Plaintiffs’ indigency or modest means. In re Paoli, 1999 WL 569435, at *4. But if the Court were to have credited the Plaintiffs’ statements in their affidavits about their inability to pay a $154,129.30 costs award, the Court could have reduced the award under our precedent and that of other courts of appeals. See infra

Section V.B. Therefore, from our reading of the District Court’s opinion, it was the Plaintiffs’ failure timely to perfect the evidentiary record before the Clerk of Court rather than the inadequacy of their evidence that the District Court understood to bind its hands. The District Court’s reasoning and invocation of precedent bears out this conclusion.

In opining that there was no suitable basis on which to assess the Plaintiffs’ respective financial statuses, the District Court analogized its power to hear evidence that was not before the Clerk of Court to that of a district court reviewing the decision of a bankruptcy court. The Court cited Samar Fashions, Inc. v. Private Line, Inc., 116 B.R. 417, 420 (E.D.Pa.1990), as defining the district court’s power to expand the record in a bankruptcy case. Samar Fashions holds that a district court’s review of a bankruptcy decision is limited to the record before the bankruptcy court. See id.

If the rule in Samar Fashions were to apply in the Rule 54(d)(1) context, a losing party could never expand the record before the District Court, no matter how thorough or compelling their new affidavits might be. Accordingly, we are forced to conclude that the District Court was interpreting the limits of its discretion under Rule 54(d)(1) in refusing to consider the Plaintiffs’ affidavits, not exercising its discretion not to do so. As such, our review of that decision is plenary. See Kleissler, 183 F.3d at 198. B.

The District Court’s legal conclusion that it could not hear evidence that was not first presented to the Clerk of Court is unsound for several reasons. Under the well-established Rule 54(d)(1) case law, the district court is charged with making a de novo review of the clerk’s determination of the costs issue. See 10 Moore’s, supra, § 54.100[3], at 54-145 (citing Farmer v. Arabian Am Oil Co., 379 U.S. 227, 233, 85 S.Ct. 411, 13 L.Ed.2d 248 (1964)).

In Farmer, 379 U.S. at 233, 85 S.Ct. 411, the Supreme Court acknowledged that conducting a de novo review in a costs award case properly involved reviewing new evidence and circumstances that militated in favor of reducing the earlier imposed costs award. The District Court’s analogies to the appellate review and bankruptcy cases that it cited were thus inapposite. In Fassett v. Delta Kappa Epsilon (New York), 807 F.2d 1150, 1165 (3d Cir.1986), we explained that the proper function of an appellate court is to review decisions on the basis of the record before the district court.

But as Farmer makes clear, 379 U.S. at 232-34, 85 S.Ct. 411, the rules governing de novo review — not appellate review or bankruptcy review, which is a variation thereof, see In re Columbia Gas Sys., Inc., 997 F.2d 1039, 1059 (3d Cir.1993)—govern Rule 54(d)(1) challenges. As this court has held in similar cases involving courts reviewing decisions of non-judicial officials, “ ‘De novo means here, as it ordinarily does, [that] ... the court’s inquiry is not limited to or constricted by the ... record, nor is any deference due the ... conclusion [under review].’” Luby v.

Teamsters Health, Welfare, & Pension Trust Funds, 944 F.2d 1176, 1184 (3d Cir.1991) (emphasis and alteration in original); see id. at 1184-85 (holding that, in an ERISA case, a district court conducting a de novo review of a plan administrator’s findings of fact could expand the factual record) (citing cases involving a similar

interpretation of the de novo review standard when a district court reviews the decisions of administrative agencies). There is a cognate basis why the District Court’s legal conclusion — i.e., that it could not expand the record before the Clerk of Court — is incorrect. The clerk of court’s role in taxing costs awards, while quasi-judicial, is essentially ministerial.

This fact not only explains why certain evidence might not be presented to the clerk of court, but underscores why it is important that district courts retain the discretion to assess independently the factual record, whether the record consists of new evidence or old. The Clerk in this case acknowledged in his written order that he lacked the discretion to consider certain equities in taxing cost against the Plaintiffs.

The Plaintiffs would therefore have had no reason then to submit evidence supporting their equitable arguments — e.g., their indigency argument or their argument that it would be unfair to award joint and several liability — until they sought review by the District Court. Moreover, in fulfilling its responsibility to insure that a costs award is not “inequitable,” Friedman v. Ganassi, 853 F.2d 207, 211 (3d Cir.1988), the district court not only can, but should, consider evidence that completes the factual record or sheds light on the equities in a given case. Doing so may involve expanding the evidentiary record.

Another reason for our conclusion is that “costs may be taxed by the clerk on one day’s notice,” Fed.R.Civ.P. 54(d)(1), and without adversarial proceedings. This short time frame and the attending limited opportunity to develop a factual record calls for careful supervision by district courts of costs awards. Recognizing this important distinction, several district courts have permitted parties challenging costs awards to introduce or rely upon evidence not presented before the clerk of court. See, e.g., Burks v. City of Philadelphia, No. Civ. A. 95-1636, 1999 WL 163636, at *1 (E.D.Pa.

Mar.22, 1999) (considering “copies of [the plaintiffs’] personal financial statements, bills and other information which they claim[ed] demonstrate[d] their inability to pay the assessed costs,” even though such records were presented for the first time in the plaintiffs’ motion to reconsider the clerk’s award). For these reasons, we hold that District Court erred as a matter of law in ruling that it did not have the power under Rule 54(d)(1) to consider the Plaintiffs’ affidavits not filed with the Clerk of Court.

Because courts “[t]ypieally ... will accept the allegations concerning the applicant’s financial circumstances [when their indigency is in question] at face value,” 10 Weight et al., supra, § 2673, at 307, the District Court also should not have ignored the affidavit filed by the paralegal with the Clerk of Court. That is not to say, however, that the Court could not have discredited the representations made in these affidavits or ultimately refused to reduce the costs award that it imposed in this case.

Moreover, even though district courts have the power to expand the record in Rule 54(d)(1) cases, they retain the discretion not to do so. Therefore, counsel for losing parties would be well advised, when it is feasible, to submit to the clerk of court all relevant evidence bearing on a potential costs award before the clerk proceeds to tax costs. V. What Factors a District Court May Consider in Weighing the Equities of a Costs Award A.

Rule 54(d)(1) provides in pertinent part that “[e]xcept when express provision therefor is made either in a statute of the United States or in these rules, costs other than attorneys’ fees shall be allowed as of course to the prevailing party unless the court otherwise directs....” FED.R.CIV.P. 54(d)(1). By mandating that, subject to court intervention, costs be allowed to a prevailing party as of course, the rule creates the “strong presumption” that costs are to be awarded to the prevailing party. 10 Moore’s, supra, § 54.101, at 54-149 (collecting cases); see also Delta Air Lines, Inc. v.

August, 450 U.S. 346, 352, 101 S.Ct. 1146, 67 L.Ed.2d 287 (1981); Institutionalized Juveniles v. Secretary of Pub. Welfare, 758 F.2d 897, 926 (3d Cir.1985). To overcome the presumption favoring the prevailing party and to deny that party costs, a district court must “ ‘support[ ] that determination with an explanation.’ ” Institutionalized Juveniles, 758 F.2d at 926 (quoting Samuel v. University of Pittsburgh, 538 F.2d 991, 999 (3d Cir.1976)). Moreover, the losing party bears the burden of making the showing that an award is inequitable under the circumstances.

See 10 Moore’s, supra, § 54.101, at 54-151 & n. 7 (collecting cases). Reported cases have discussed a number of equitable factors, most advanced by the parties herein, that a district court may consider in determining a costs award. The parties, as well as Amicus — the Pennsylvania Trial Lawyer’s Association — have advanced a number of factors as candidates for reviewing a costs award imposed by the clerk of court.

These factors include: (1) the unclean hands, or bad faith or dilatory tactics, of the prevailing party; (2) the good faith of the losing party and the closeness and difficulty of the issues they raised; (3) the relative disparity of wealth between the parties; and (4) the indigence or inability to pay a costs award by a losing party. While the first of these factors is not implicated in this case, the propriety of a court considering each of the remaining factors is in dispute.

The “unclean hands” factor is not in dispute because the Plaintiffs do not assert that the named Defendants engaged in any acts that would implicate this factor. We reaffirm, however, that in an appropriate case, the unclean hands factor is a relevant one for district courts to consider. See Smith v. Southeastern Pa. Transp. Auth., 47 F.3d 97, 99 (3d Cir.1995) (per curiam) (recognizing this factor). We turn to the remaining factors. B. The most important of these factors is the losing party’s indigency or inability to pay “the full measure” of a costs award against it. Smith, 47 F.3d at 99.

Several courts of appeals have held that indigency, or modest means, is a factor that a district court may consider in awarding costs. See, e.g., Cherry v. Champion Int’l Corp., 186 F.3d 442, 447 (4th Cir.1999) (evaluating whether a non-indigent losing plaintiff had the “effective ability to satisfy[the defendant’s] bill of costs” or was “of such modest means that it would be unjust or inequitable to enforce Rule 54(d)(1) against her”); Weeks v. Samsung Heavy Indus. Co., Ltd., 126 F.3d 926, 945 (7th Cir.1997) (“[T]he losing party’s inability to pay will suffice to justify denying costs.”).

Other courts that have adopted this approach also caution that a losing party’s indigency or an inability to pay costs does not automatically mean that a costs award levied against that party is inequitable. See, e.g., Weaver v. Toombs, 948 F.2d 1004, 1008 (6th Cir.1991) (holding that indigency may be a shield to imposition of costs, but that it is not an absolute shield). This case-by-case approach to the “indigency” factor has also been expressly or implicitly endorsed by noted commentators on the subject.

See, e.g., 10 Moore’s, supra, §§ 54.101[l][b], at 54-153, 54.104[l][a]-[c], at 54-198 to 54-201; 10 Wright et al., supra, § 2673, at 305-09. We note too that we have suggested the same. In Smith v. Southeastern Pennsylvania Transportation Authority, 47 F.3d 97, 100 (3d Cir.1995) (per curiam), we held that the “disparity in the parties’ financial resources seems to us to be irrelevant for purposes of Rule 54(d),” but we also stated that [i]f the losing party cannot afford to payfcosts], that party is not automatically exempted from the taxation of costs. On the contrary, 28 U.S.C. § 1915(

e) and cases decided thereunder make clear that costs may be taxed against a party who is permitted to proceed in forma pauperis. See, e.g., Washington v. Patlis, 916 F.2d 1036, 1039 (5th Cir.1990); Harris v. Forsyth, 742 F.2d 1277, 1278 (11th Cir.1984); Flint v. Haynes, 651 F.2d 970, 973 (4th Cir.1981), cert. denied, 454 U.S. 1151, 102 S.Ct. 1018, 71 L.Ed.2d 306 (1982).

While these cases recognize that a district court may consider a losing party’s indigency in applying Rule 54(d), the losing party in this case does not claim to be indigent, and the record does not establish that she is unable to pay the full measure of costs. Id. The negative pregnant contained in this statement is that a party may be exempted from costs if he is in fact indigent, if he has adduced evidence that he is indigent, and if the district court sees fit to reduce the costs award imposed for reasons of equity.

Policy considerations underpinning the American costs rule support the consideration of the “indigency or inability to pay” factor. As Professors Wright, Miller, and Kane have observed, Rule 54(d)(1) is “founded on the egalitarian concept of providing relatively easy access to the courts to all citizens and reducing the threat of liability for litigation expenses as an obstacle to the commencement of a lawsuit or the assertion of a defense that might have some merit.” 10 Wright et al., supra, § 2665, at 202.

Persuaded by the authority from other circuits and these policy considerations, we hold that the District Court could have considered this factor, given the affidavits that the Plaintiffs presented regarding their indigency. Therefore, the District Court’s order imposing costs must be set aside. Stating our position more precisely, we hold that if a losing party is indigent or unable to pay the full measure of costs, a district court may, but need not automatically, exempt the losing party from paying costs.

Such an approach is somewhat at odds with the traditional rule at law that the prevailing party was automatically entitled to its costs, but it is consistent with the rule at equity that the district court exercise its discretion to insure that the award be equitable. See supra

Section II.A. Allowing for the indigency factor in certain cases is also in keeping with the American tradition of not providing total reimbursement. See id. As was noted above, see id., the types of costs recoverable under Rule 54(d)(1) are quite circumscribed. These costs “do not include such litigation expenses as attorney’s fees and expert witness fees in excess of the standard daily witness fee,” Friedman v.

Ganassi, 853 F.2d 207, 209 (3d Cir.1988), and as a result, while a prevailing party is awarded its Rule 54(d)(1) costs, those costs often fall well short of the party’s actual litigation expenses, see 10 Wright et al., supra, § 2666, at 202-04; see also 10 Moore’s, supra, § 54.103, at 54-174 to 54-197. The plaintiffs have introduced evidence that they are indigent or unable to pay the large costs award imposed in this case. See supra

Part IV. We will therefore reverse the order of the District Court on this point. On remand, the District Court should consider this evidence and the “indigency or inability to pay” factor. C. The next disputed factor is the good faith of the losing party (here the Plaintiffs) and the closeness and difficulty of the issues that they raised. We have no doubt that the Plaintiffs acted in good faith, and our prior opinions in this matter reflect the seriousness, closeness, and difficulty of the many issues that they raised. This was a classic close case, brought in good faith by the Plaintiffs.

This court, in fact, expressed the sentiment that some may have been surprised by the fact that the Plaintiffs had lost at trial. See In re Paoli R.R. Yard PCB Litig., 113 F.3d 444, 464 (3d Cir.1997). However, despite the fervent advocacy of the Plaintiffs’ counsel, we do not consider these factors, in and of themselves, to be appropriate criteria in determining whether a costs award is equitable.

Amicus argues for the applicability of the “good faith” and “closeness or complexity” factors based on a complicated reading of our case law and this court’s statements, which are quoted in block form in footnote 4 above, in Smith v. Southeastern Pennsylvania Transportation Authority, 47 F.3d 97 (3d Cir.1995) (per curiam). In Smith, we referenced our previous opinion in ADM Corp. v. Speedmaster Packaging Corp., 525 F.2d 662 (3d Cir.1975), which quoted the Seventh Circuit Court of Appeals’s opinion in Chicago Sugar Co. v.

American Sugar Refining Co., 176 F.2d 1, 11 (7th Cir.1949), for the proposition that a district court may consider certain equitable factors in arriving at a costs award. See Smith, 47 F.3d at 99. Both Smith and ADM, however, quoted only one equitable factor enumerated in Chicago Sugar — that relating to the prevailing party’s bad faith. See supra note 4.

Chicago Sugar also held that “where it is clear that the action was brought in good faith, [and] involv[ed] issues as to which the law is in doubt, the court may in its discretion require each party to bear its own costs although the decision is adverse” to the prevailing party. Chicago Sugar, 176 F.2d at 11. Amicus argues that we must apply this latter set of factors, even though they were not referenced in Smith or ADM. To do so, Amicus must first explain why Smith and ADM quoted Chicago Sugar so selectively.

Amicus first argues that the reason Smith “did not follow Chicago Sugar fully” is that perhaps the Smith court “did not have to read Chicago Sugar” when it res-cribed the quote from ADM. Amicus Br. at 13. The conduct of the losing party factor was before the court in Smith, however, and this court’s failure to recognize that factor, when certain other courts have done so (as detailed in the margin), seems strong evidence that Smith’s failure to quote the second set of Chicago Sugar factors was not merely an errant omission, but purposeful.

The district court in Smith had taken into account the fact that the “ ‘[p]laintiff pursued a legitimate claim in good faith and raised a serious legal issue’ ” when reducing the defendant’s costs award. Smith, 47 F.3d at 99 (quoting District Ct. Op. at 3). In reversing this reduction for an abuse of discretion, we held that any denial of costs is a penalty and that only the prevailing party’s conduct should be taken into account in weighing the equities. See id.

Though not explicitly rejecting the good-faith or the complexity-of-the-issues factors mentioned in Chicago Sugar, we seem to have been faced with at least the “good faith” exception and implicitly to have rejected it. Smith’s rejection of the “good faith” exception in Chicago Sugar makes sense when one considers the rationale under-girding Smith. Smith stated plainly that any denial of costs penalizes the prevailing party. See 47 F.3d at 99.

Therefore, good citizen or not, the losing party’s strong or justified belief that her claim had merit would not support levying a penalty against the opposing party who acted in good faith and prevailed in defending itself. The only possible exception to this rule that Smith noted was when the losing party was indigent. See id. at 99-100. The Court of Appeals for the Ninth Circuit has squarely addressed both the “good faith” and the “complexity of the issues” factors mentioned in Chicago Sugar, and flatly rejected them. See National Info. Serv. v. TRW, Inc., 51 F.3d 1470, 1472-73 (9th Cir.1995).

In doing so, the court reasoned: We specifically reject Plaintiffs’ argument that their good faith in prosecuting the underlying antitrust action should defeat the presumption in favor of awarding costs. All parties to a federal action have an obligation to act in good faith and with proper purpose. It follows that noble intentions alone do not relieve an unsuccessful litigant of the obligation under Rule 54(

d) to compensate his opponent for reasonable costs. If the awarding of costs could be thwarted every time the unsuccessful party is a normal, average party and not a knave, Rule 54(d)(1) would have little substance remaining.... Plaintiffs offer an alternative basis on which to let the district court order stand. They ask us to affirm the denial of costs on the ground that their antitrust action raised close and difficult legal issues.... [W]e hold that difficulty alone does not justify penalizing the prevailing parties.

As we explained, the district court generally must award costs unless equity demands otherwise due to some impropriety on the part of the prevailing party during the course of the litigation. Generally speaking, neither side bears responsibility for the length or complexity of a particular case. Moreover, the difficulty of a case hardly constitutes an impropriety that calls out for punishment. Id. (citations and internal quote marks omitted).

Though some courts hold to the contrary, see 10 Wright et al., supra, § 2668, at 238-39 & n. 19 (collecting cases), we agree with the Ninth Circuit and the majority of courts that have rejected the losing party’s good faith as a relevant factor, see 10 Moore’s, supra, § 54.101[b], at 54-151 to 54-152 (collecting cases). The only time the losing party’s conduct should be at issue is if the losing party acted in bad faith during the course of litigation and then requested a costs award reduction.

In such a circumstance, we agree with the courts that hold that a district court may deny a losing party’s request to reduce costs if the losing party acted in bad faith. See, e.g., Cherry v. Champion Int’l Corp., 186 F.3d 442, 447 (4th Cir.1999) (recognizing that the losing party’s good faith is “is a prerequisite for her claim to relief from the presumptive application on Rule 54(d)(1)”) (emphasis in original).

Moreover, though some courts of appeals permit district courts to consider the complexity or closeness of the legal issues involved in a case when imposing a costs award, see supra note 7 (collecting cases), we hold that this factor is not an appropriate one for district courts to consider, see Klein v. Grynberg, 44 F.3d 1497, 1507 (10th Cir.1995) (holding similarly). Rather, we believe that to deny a prevailing party costs because it was successful in prosecuting or defending against a complex legal theory unduly punishes that party.

Effecting such a penalty, absent punishable conduct on the part of the prevailing party, is contrary to this court’s long line of precedent interpreting Rule 54(d)(1). See Smith, 47 F.3d at 99; see also Moore’s, supra, § 54.101[b], at 54-151 to 54-152 (“[T]he fact that the losing party’s position had merit and that the case presented close and difficult issues is not a sufficient ground to deny costs to the prevailing party.”)' (footnote and citations omitted).

In reaching this conclusion, however, we reiterate that a district court may reduce a costs award, on equitable grounds, if the prevailing party, through bad faith or dilatory tactics, has turned a relatively simple case into a complex morass. This analysis, however, should be done pursuant to the “unclean hands” factor discussed above, see supra

Section V.A, or in assessing which costs are actually recoverable under the statute, see 28 U.S.C. § 1920. Complexity, in and of itself, however, may not be considered as a relevant equitable factor. This is what we take the Ninth Circuit to mean when it writes: [W]e hold that “difficulty” alone does not justify penalizing the prevailing parties .... [T]he district court generally must award costs unless equity demands otherwise due to some impropriety on the part of the prevailing party.... Generally speaking, neither side bears responsibility for the length or complexity of a particular case. National Info.

Serv., 51 F.3d at 1473 (citations omitted). Similarly, when the Fourth Circuit speaks of the “excessiveness” of a costs award as a relevant factor for denying the prevailing party its costs or reducing them, we understand the court to be referring to the possibility that a prevailing party’s bad conduct may be responsible for excessive costs; not that the sheer size of a costs award is a relevant factor in and of itself. Cherry v. Champion Int’l Corp., 186 F.3d 442, 446 (4th Cir.1999).

If district courts in this circuit are to cognize this factor, they should do so under the rubric of their “unclean hands” analysis. However, for the reasons noted above — i.e., that there are no allegations of bad faith on the part of the Defendants in this case, see supra

Section V.A — the “unclean hands” factor, as described in this

Section and above, should be not considered on remand. D. The last Rule 54(d)(1) “equity factor” potentially implicated in this case is the relative wealth of the parties. The Plaintiffs and Amicus maintain that while many of the Plaintiffs were disabled, poor individuals living on fixed incomes with little or no assets, the Defendants-General Electric, Monsanto, and Westinghouse (CBS) — were “billion dollar corporations that are among the largest and wealthiest in the world.” Appellants’ Br. at 11.

The Plaintiffs and Amicus therefore contend that the District Court should have considered the parties’ relative ability to bear the Defendants’ costs in weighing the equities in this case. In Smith, 47 F.3d at 99-100, we rejected this potential Rule 54(d)(1) factor; so too have the clear majority of courts and commentators, see, e.g., Cherry v. Champion Int’l Corp., 186 F.3d 442, 448 (4th Cir.1999); Reed v. International Union of United Auto., Aerospace & Agric. Implement Workers, Local Union No. 663, 945 F.2d 198, 204 (7th Cir.1991); Moore’s, supra, § 54.101[b], at 54-153.

Thus, the District Court rightly declined to accept the Plaintiffs’ invitation to consider this factor. E. In sum, we hold that a district court may consider the following factors in reviewing a clerk of court’s costs award: (1) the prevailing party’s unclean hands, bad faith, dilatory tactics, or failures to comply with process during the course of the instant litigation or the costs award proceedings; and (2) each of the losing parties’ potential indigency or inability to pay the full measure of a costs award levied against them.

In contrast, a district court may not consider (1) the losing parties’ good faith in pursuing the instant litigation (although a finding of bad faith on their part would be a reason not to reduce costs); (2) the complexity or closeness of the issues — in and of themselves— in the underlying litigation; or (3) the relative disparities in wealth between the parties. Again, the presumption is that costs, as defined by the relevant statutes and case law, will be awarded in full measure.

Only if the losing party can introduce evidence, and the district court can articulate reasons within the bounds of its equitable power, should costs be reduced or denied to the prevailing party. For the reasons set forth above, we reverse and remand so that the District Court can reconsider the “indigency” factor described herein. VI. Apportioning the Costs Award Among the Plaintiffs The final issue is whether the District Court erred in imposing liability for costs jointly and severally among the Plaintiffs.

The general rule in this circuit and others is that a district court, in exercising its equitable discretion, may apportion costs between the prevailing and non-prevailing parties as it sees fit. See Croker v. Boeing Co. (Vertol Div.), 662 F.2d 975, 998 (3d Cir.1981) (en banc), overruled on other grounds, Halderman v. Pennhurst State Sch. & Hosp., 673 F.2d 628, 644 (3d Cir.1982) (en banc); Wright et al., supra, § 2668, at 243 & n. 31 (collecting cases).

This power to apportion between parties also includes the power to divide costs among the losing and winning parties in a case involving multiple defendants and plaintiffs such as this one, which involves nineteen plaintiffs and three defendants whose cases were all consolidated for efficiency’s sake by an unopposed order of the District Court. Cf. ADM Corp. v.

Speedmaster Packaging Corp., 525 F.2d 662, 665 (3d Cir.1975) (instructing the district court to consider whether and how the prevailing party’s costs should be apportioned between the two non-prevailing parties, given the non-prevailing parties’ different responsibilities in forcing the prevailing party to incur costs). In dividing costs among multiple parties on one side of the bar — either prevailing or non-prevailing— the court may choose to impose costs jointly and severally or to disaggregate costs and to impose them individually.

We read the sparse case law surrounding this issue as placing the burden on the losing parties to introduce evidence and persuade the court that costs should be apportioned, and in the event that they fail to do so, the default rule is that costs may be imposed jointly and severally. Cf. Electronic Specialty Co. v.

International Controls Corp., 47 F.R.D. 158, 160 (S.D.N.Y.1969) (holding that “despite the general rule that where there are several plaintiffs who join together each is individually liable for all the costs and not merely for his pro rata share,” “it is in the interest of justice for [the] defendant to seek no more than one-third of the awarded costs from” plaintiffs #1 or #2, because the evidence introduced by plaintiff # 1 and # 2 established that the defendant would not seek to collect the proportionate share from plaintiff # 3).

The Plaintiffs contend that they introduced ample evidence justifying an apportionment, and that the District Court erred in not dividing the Defendants’ costs among the Plaintiffs. Because the District Court’s decision to impose joint and several liability and not to apportion costs involved an application of Rule 54(d)(1), and not an

interpretation of the rule’s boundaries, we review this determination for abuse of discretion. See supra

Section II.B. A. The Plaintiffs present a relatively strong case for apportioning costs among all nineteen of them. Three of the nineteen Plaintiffs were deceased and left behind no assets against which the Defendants could collect. This meant that sixteen of the nineteen Plaintiffs would bear a slightly disproportionate share of the Defendants’ costs. Absent other circumstances, this is not enough, however, to require setting aside the District Court’s decision to impose costs jointly and severally. Cf. Southern Agency Co., 393 F.2d at 915 & n. 7 (discussed at note 9, supra ).

While a court may decide to reduce a costs award in such a circumstance, one must remember that our standard of review is limited and that there are only zero-sum gains in costs awards cases: Prevailing parties pay for every dollar that a costs award is reduced. It is within the discretion of the district court, in which equitable discretion is vested according to the terms of Rule 54(d)(1), to decide whether the losing or prevailing party should bear the brunt of such an inequity. We must defer to that judgment absent an abuse of discretion.

The Plaintiffs have pointed to a number of other circumstances in arguing that the District Court abused its discretion in not apportioning costs. The Plaintiffs brought their claims as individuals and the District Court consolidated their cases only for efficiency’s sake. Therefore, the Plaintiffs submit that they should only be taxed for a one-nineteenth share of the Defendants’ costs. Some case law supports this proposition. See, e.g., Davis v.

Parkman, 71 F. 961, 964 (1st Cir.1895) (holding that where the parties and court “saw fit, for economical reasons ... to consolidate the records,” in determining costs, “the costs should,” nevertheless, “be justly apportioned between the suits”); Modick v. Carvel Stores of New York, Inc., 209 F.Supp. 361, 363 (S.D.N.Y.1962) (holding that in a case in which several parties’ actions were consolidated for. trial-for the convenience of the parties and the court— “each case must stand on its own feet so far as costs are concerned”).

But the litigation world has vastly changed not only since 1895 when Davis was decided, but also since Modick was filed in 1962. The proliferation of complex eases, burgeoning dockets, and scarce judicial resources have made separate trials in cases like this, which are essentially mass tort cases, less a luxury than a near impossibility. This means that both defendants and plaintiffs are forced to expend a great deal of time and money in prosecuting and defending these actions. The advantage in this scheme for plaintiffs is that they are able to share their costs.

Defendants do not always have this luxury, and it is often the case, when defending against a mass tort claim of this sort, the defendant will not (and perhaps cannot) earmark each dollar it spends in defending against each of the plaintiffs’ individual claims. Moreover, multiple defendants in such cases are more likely than plaintiffs to have separate counsel.

Given the abuse of discretion standard of review, and the general similarity of most of the Plaintiffs’ claims and costs incurred in response thereto, we hold that the District Court did not abuse its discretion in not apportioning costs among the personal injury plaintiffs. Each of these plaintiffs advanced toxic tort claims that prompted the Defendants to hire experts, conduct studies, and prepare trial exhibits and documents aimed at defending against this class of claims.

The Plaintiffs have adduced no evidence suggesting that these costs could be connected — to a greater or lesser degree — with any of the particular Plaintiffs. That is not to say that a district court could not reach the opposite conclusion in such a circumstance, but merely that the decision in this case was not an abuse of discretion. Moreover, our holding should not suggest that there may not be cases in which a district court could find that an individual dragged into the web of a large litigation should not be responsible for the defendants’s costs no matter how hard it is disaggregate them precisely.

A rough division of costs liability among plaintiffs in such a circumstance may be the most equitable result. At all events, we hold that the District Court did not abuse its discretion in fashioning the costs award liability of the nineteen Plaintiffs generally here. B.

Notwithstanding the general presumption that costs be granted in the prevailing parties’ favor and the default rule that they may be imposed jointly and severally, we hold that the District Court abused its discretion in imposing joint and several liability against Plaintiffs James Lament and Sylvan Cohen for the entirety of the Defendants’ acceptable costs. Our reasons for so holding are as follows. First, these two Plaintiffs’ cases were simply different in kind from the rest of the Plaintiffs’.

Lament and Cohen brought diminished value of rental property claims, which were only accountable for a fraction of the Defendants’ $154,129.30 in costs. They did not live near the site of the Rail Yard, and hence, they never pursued any of the personal injury or medical monitoring claims that forced the Defendants to expend a great deal of money copying medical records, preparing elaborate trial exhibits, hiring medical experts, and taking scores of depositions.

The complexity and centrality to the litigation of these latter claims is demonstrated by a reading of Paoli I, II, and IV, and Lament and Cohen, unlike the other seventeen Plaintiffs, are in no way responsible for these costs. Second, the Plaintiffs have been able to disaggregate the amount of money the Defendants spent on diminution of property value claims ($9,284.59), see Appellants’ Br. at 18 n. 7, as compared to defense costs in total ($154,129.30).

Unlike Plaintiffs’ general contention that costs should be disaggregated among the nineteen Plaintiffs, this demonstration enabled the District Court to avoid imposing specific costs on Plaintiffs whose causes of action did not give rise to these costs. The Defendants have not disputed this representation, and the District Court did not focus on this evidence in issuing its costs award order. We find the Plaintiffs’ representations compelling.

In sum, given the huge disparity of the costs that the Defendants incurred in defending against property claims as compared to personal injury and medical monitoring claims, and the ability easily to disaggregate the costs expended in defending against the different groups of claims, we conclude that the joint and several costs award served only to penalize Lament and Cohen, not to award justifiably the Defendants their costs. Rule 54(d)(1) was not intended to serve such purposes. See supra

Section II.A. We will therefore vacate the District Court’s costs award order insofar as it affects Cohen and Lamont. On remand Cohen’s and Lamont’s costs liability must be disaggregated from their fellow Plaintiffs’ liability arising out of the personal injury and medical monitoring claims. The Court should exercise its discretion in determining both whether Cohen’s liability should be reduced further for his lack of involvement in the case during the last several years, see supra note 10, and whether the Plaintiffs’ property-based awards should be imposed in an apportioned or in a joint and several manner.

On remand, and as noted above, the Court should also determine whether any of the Plaintiffs’ indigency militates in favor of reducing his or her respective costs award liability. VII. Conclusion The order of the District Court will be affirmed in part and vacated in part, and the cases remanded for further proceedings consistent with this opinion. Parties to bear their own costs on appeal. . See In re Paoli R.R. Yard PCB Litig., 916 F.2d 829 ("Paoli I”) (3d Cir.1990); In re Paoli R.R. Yard PCB Litig., 980 F.2d 724 ("Paoli II”) (3d Cir.1992); In re Paoli R.R.

Yard PCB Litig., 35 F.3d 717 (3d Cir.1994) ("Paoli III"); In re Paoli R.R. Yard PCB Litig., 113 F.3d 444 (3d Cir.1997) ("Paoli IV”). For a more thorough discussion of the underlying facts and legal issues in this case, see Paoli I, III, and IV. . Approximately 290 actions alleging injury due to exposure to PCBs at the Rail Yard are pending in the Court of Common Pleas of Chester County, Pennsylvania.

The Pennsylvania Commonwealth Court recently upheld a Chester County judge’s decision to grant class certification on a medical monitoring claim to residents and workers who allege that they were exposed to above-normal levels of PCBs at the Rail Yard. Individual suits for recovery of personal injuries, emotional distress, real property, and punitive damages are also currently pending. See Lori Litchman, Commonwealth Court OKs Paoli Rail Yard Class, Legal Intelligencer, June 30, 2000, at 1, 10.

The parties disagree as to whether the federal plaintiffs in this case are entitled to recover in the pending state court actions. If the Plaintiffs recover from the Defendants in state court, the Plaintiffs would potentially be in a better position to pay the costs award imposed in this case. The Plaintiffs represent, in supplemental briefing, that they will not secure any such future recovery. The District Court will need to resolve this dispute, on remand, when determining whether the Plaintiffs are of too modest means to pay the costs award that the District Court imposes. See infra Sections V.B; V.E; VLB. .

According to the affidavits, several of the Plaintiffs were deceased and left behind little or no assets, and thus, their estates could not afford to pay for even a fraction of the award. See, e.g., App. at 419-20, 422, 424. Additionally, several of the Plaintiffs averred that they were on fixed incomes, had little or no savings, or were sick or disabled, and therefore had no discretionary income with which to pay the award. See, e.g., id. at 441-42.

Finally, still other Plaintiffs that had certain assets — such as a car, or a home on which they still owed a mortgage, or even those that owned unencumbered property — stated that there was no way that they could pay $185,000 in costs. See, e.g., id. at 426-27, 429-30, 435-36, 438-49. . The rationale for the unclean hands factor was described in Smith as follows: '[T]he denial of costs to the prevailing party ... is in the nature of a penalty for some defection on his

part in the course of the litigation.’ ” ADM Corp. v. Speedmaster Packaging Corp., 525 F.2d 662, 665 (3d Cir.1975) (quoting Chicago Sugar Co. v. American Sugar [Ref.] Co., 176 F.2d 1, 11 (7th Cir.1949)).

The Chicago Sugar case provides the following examples of a “defection” that would warrant denying costs to a prevailing party: "calling unnecessary witnesses, bringing in unnecessary issues or otherwise encumbering the record, or ... delaying in raising objection fatal to the plaintiff’s case....” Smith, 47 F.3d at 99 (quoting Institutionalized Juveniles, 758 F.2d at 926); see also Moore's, supra, § 54.101[b], at 54-153 to 54-154 (listing several examples of improper conduct that would implicate the unclean hands factor). .

In assessing the Plaintiffs' indigency or modest means, on remand, the Court should measure each Plaintiff's financial condition as it compares to whatever award the Court decides to tax against him or her. And if the Court decides to apportion the costs award among the Plaintiffs, it should measure each Plaintiff's individual ability to pay his or her portion of the Defendants' costs. If the Court imposes joint and several liability — in a manner that we hold that it may in

Part VI — it should measure each Plaintiff’s individual ability to pay against the entire joint and several costs award. Once this initial parsing is done, however, there are no hard and fast rules for assessing a losing party's indigency ar inability to pay; district courts should use their common sense in making this determination. . Before addressing whether good faith and complexity are indeed relevant Rule 54(d)(1) factors, we must first consider whether the issue is even properly before us.

The Plaintiffs advanced the good faith, close case argument before the District Court, where the Defendants resisted it, and the District Court rejected it. The Plaintiffs do not advance the argument on appeal, and the Defendants continue to maintain that the only two relevant factors in a Rule 54(d)(1) costs award adjustment are the prevailing party's unclean hands and the losing party’s indigency. It is Amicus therefore that has pressed the "good faith” and "complexity of the issues” factors here.

Though we need not necessarily reach issues advanced only by amici, courts have consistently noted that amici may raise arguments that were presented by the parties before the lower courts. See United Parcel Serv., Inc. v. Mitchell, 451 U.S. 56, 60 n. 2, 101 S.Ct. 1559, 67 L.Ed.2d 732 (1981); Bell v. Wolfish, 441 U.S. 520, 531 n. 13, 99 S.Ct. 1861, 60 L.Ed.2d 447 (1979); Knetsch v. United States, 364 U.S. 361, 370, 81 S.Ct. 132, 5 L.Ed.2d 128 (1960); Toussaint v. McCarthy, 801 F.2d 1080, 1106 n. 27 (9th Cir.1986).

The issue raised by Amicus has potential relevance in this case, because if we were to accept Amicus’s reading of our case law, the application of the “good faith” and "complexity of the issues” factors could have significant effects on a costs award imposed on remand. Therefore, we can, and will, reach this issue, which has been briefed by both Amicus and the Defendants. . See, e.g., Cherry v.

Champion Int’l Corp., 186 F.3d 442, 446, 447 (4th Cir.1999) (recognizing, among other factors, the "closeness and difficulty of the issues decided;” and noting that while the losing party’s "good faith” is a factor, it “is a prerequisite for her claim to relief," as well as an insufficient basis, standing alone, for refusing to assess costs against that party) (emphasis in original); White & White, Inc. v. American Hosp.

Supply Corp., 786 F.2d 728, 731, 733 (6th Cir.1986) (holding that, among other factors, the fact that case presented a close and difficult question was properly considered in denying costs to the prevailing defendants, and noting, in addition, that the good faith of the losing party, alone, was insufficient to overcome the presumption of an award); Chicago Sugar, 176 F.2d at 11; see also 10 Wright et al., supra, § 2668, at 238 ("Costs also have been denied ... on the ground that the losing party prosecuted the action in good faith, the issues presented were difficult, novel, or complex, and the winning party could absorb the costs.”); id. at 238 n. 19 (collecting cases). .

Amicus advances the circuitous argument that if Smith so implicitly ruled, it violated this court’s rules regarding a panel’s ability to overrule a prior panel’s decision. Amicus cites an en banc decision from this court that stands for the relatively banal proposition that "[t]he particular circumstances of case may permit a district court to refuse to award costs altogether or to apportion them between the parties.” Croker v. Boeing Co. (Vertol Div.), 662 F.2d 975, 998 (3d Cir.1981) (en banc), overruled on other grounds, Halderman v. Pennhurst State Sch. & Hosp., 673 F.2d 628, 644 (3d Cir .1982) (en banc).

Croker cited to a case and treatise that stand for similarly uncontroversial propositions regarding the scope of a district court’s discretion in Rule 54(d)(1) cases. The treatise, Wright & Miller’s, Federal Practice and Procedure (1st ed.1973), also noted that certain courts consider the conduct of the losing party and the complexity of the case in weighing the equities of costs award. From this trail of citations, Amicus argues that Croker adopted, in toto and sub silentio, the "good faith” and the "complexity of the issues” factors cognized in Chicago Sugar. This argument is unavailing.

Not only were the above factors not before the Croker court, the citation to Wright & Miller did not endorse the list of factors enumerated therein. It merely noted them. Moreover, Wright and Miller recognized, and continue to recognize, that courts have held that the conduct of the losing party is irrelevant in taxing costs, as are the complexity of the legal issues or the closeness of the case. See 10 Wright et al., supra, § 2668, at 239 n. 19 (citing Klein v. Grynberg, 44 F.3d 1497, 1507 (10th Cir.1995)). Other commentators have weighed in on the issue and have rejected both of these factors.

See 10 Moore's, supra, § 54.101 [b], at 54-151 to 54-152 (stating that the general rule is that “[t]he mere fact that the losing party litigated the action in good faith is not sufficient ground for denying costs to the prevailing party;” opining that "the fact that the losing party’s position had merit and that the case presented close and difficult issues is not a sufficient ground to deny costs to the prevailing party,” but citing authority holding to the contrary). Additionally, neither Wright & Miller nor Moore's cite Croker for the broad proposition for which Amicus say that it stands. .

See, e.g., In re Air Crash Disaster at John F.

Kennedy Int’l Airport on June 24, 1975, 687 F.2d 626, 629-30 (2d Cir.1982) (holding that where costs of litigation were incurred on behalf of all, not just the non-settling plaintiffs bringing an action against the defendant, and the non-settling plaintiffs did not object to a provision that the settling plaintiffs expressly waived costs to which they might have been entitled, the non-settling plaintiffs were entitled only to their proportionate share of allowable expenses incurred on behalf of all the plaintiffs, plus any items incurred specially on behalf of nonsettlors); Southern Agency Co. v. LaSalle Cas.

Co., 393 F.2d 907, 915 & n. 7 (8th Cir.1968) (holding that the district court did not abuse its discretion in apportioning costs between the two losing parties even though one of the losing parties was insolvent and the prevailing party would therefore bear one of the losing parties’ costs). . We note that Cohen voluntarily dismissed his action well before Paoli I was decided in 1990, and well before most of the Defendants’ costs were even incurred.

To charge Cohen with costs incurred during the more than ten years in which he had no part of the case might be inequitable, except perhaps insofar as the Defendants briefly had to defend against his claim. Cohen also argues that because his motion to voluntarily dismiss his case without prejudice was granted, the Defendants were not the "prevailing party” for Rule 54(d)(1) purposes. This argument runs contrary to the majority rule that defendants can be "prevailing parties” when a plaintiff voluntarily dismisses his action without prejudice. See, e.g., AeroTech, Inc. v.

Estes, 110 F.3d 1523, 1527 (10th Cir.1997).