This case arrives at our doorstep after a procedural odyssey through both the state and federal court systems. It began as a suit in a Louisiana state court by an insured and his secured lender to recover policy proceeds from the insurance company, and has ended with a protracted multi-party dispute among the insured ("Hussain"), his attorneys (collectively "Rando"), the lender/mortgagee ("Hibernia"), the defendant insurance company (Boston Old Colony or "BOC") and the defendant-appellee United States of America (the "government") over the proper distribution of an insurance proceeds fund that is insufficient to satisfy all claims in full. Central to this appeal is Rando's challenge to federal subject matter jurisdiction, as well as the rulings of the district court on the priority and amounts of distribution of the insurance fund among the several claimants. We affirm the district court in all respects except for its assessment of costs for Hussain's expert fees.

This case was sparked by a 1991 fire that destroyed the inventory of Sheik's Oriental Rugs, Inc., owned by Hussain. He was insured up to $500,000 by BOC. Hibernia was named on the BOC insurance policy (the "Policy") as a loss payee, under a 1989 loan of $177,699.02 to Hussain who had secured his promissory note by a chattel mortgage on his inventory (the "covered property"). Although the Policy appears to contain two different provisions governing the relationship between Hussain and Hibernia, a close reading suggests that only one — the "Loss Payable" provision — applies: The other, by its terms, envisions Hibernia as the putative buyer of the covered property, which it does not appear ever to have been. The Policy's "Loss Payable" provision reads as follows:

For covered property in which both you and a Loss Payee ... have an insurable interest, we will:

...

2. Pay any claim for loss or damage jointly to you and the Loss Payee, as interests may appear.

In signing the Hibernia loan documents, Hussain warranted that he alone owned the mortgaged property, "free and clear from any adverse claim, mortgage, lien, security interest, privilege, or encumbrance."

After the fire, Hussain defaulted on the Hibernia promissory note and litigation commenced. In 1992, as holder of the secured note, Hibernia sued Hussain, and in 1994 obtained a state court default judgment against him which recognized Hibernia's continuing security interest in the covered property. Also in 1992, Hibernia, as loss payee under the Policy, filed a separate state court action against Hussain and BOC to recover a portion of the policy proceeds.

In 1993, Hussain sued BOC on the Policy in state court. This suit was consolidated with Hibernia's suit as loss payee. In 1995, Hussain retained Rando as new counsel, and in doing so, executed a contingency-fee agreement that assigned one-third of any recovery to his new attorneys. Rando asserted before the district court that he has since spent $368,449.72 in prosecuting this case.1

BOC denied liability under the Policy's arson exclusion. The state court rejected that defense in a directed verdict, ruling that BOC owed Hussain and Hibernia $500,000 in policy proceeds, plus interest, costs, and fees. BOC appealed to the higher Louisiana courts; but once the Louisiana Supreme Court denied certiorari on November 13, 2000, the judgment as to BOC's liability became final.

On December 8, 2000, the Internal Revenue Service of the Department of the Treasury ("IRS") notified BOC of federal tax liens against Hussain's property. On the same day, Hussain executed a new fee agreement giving Rando a 39% interest "in and to any gross recovery I/we may have in this matter." Rather than institute concursus (interpleader) proceedings, BOC — still in the context of the consolidated suits against it, and despite the finality of the judgment in that case — filed a motion to have the court determine the amounts and priority of distribution of funds. As part of that motion, BOC also secured and served on the IRS an order to show cause why it should not be excluded from any distribution. This prompted the IRS first to file a notice of its tax lien in the records of Orleans Parish on January 9, 2001, and then, on January 17, 2001 to remove the case to the Eastern District of Louisiana.2

After briefing, the district court ruled that (1) Hibernia takes first priority, as loss payee; (2) second in line, Rando takes as attorneys' fees pursuant to the state court litigation, one-third of the amount remaining after Hibernia's claim is fully paid; and (3) third in line, the government takes the remainder of the policy proceeds, ahead of all other creditors of Hussain.3 Also, on motion by Hussain, the court taxed costs of $5,000, which was one-fifth of the amount that Hussain had sought, in payment for his expert witness, John Theriot, CPA.4

On appeal Rando has raised several issues, in essence contending that he should be paid an additional $196,377.48, an amount determined by applying his claimed contingent fee percentage (39%) to the entire amount payable from the Policy and awarded by the court.

This court reviews de novo a district court's determination of its subject-matter jurisdiction.5 On the merits, statutory construction is reviewed de novo,6 with factual findings reviewed for clear error.7 An award of attorney's fees is reviewed for abuse of discretion.8

Section 2410(a) waives the sovereign immunity of the federal government, enabling private parties to hale the government into court to determine the priority of outstanding liens on real or personal property.9 As a trade off for the waiver of sovereign immunity, section 1444 permits the government to remove to federal district court any such case initiated in state court.10 In light of this conditional relationship between §§ 2410(a) and 1444, we have held that § 1444 "confers a substantive right to remove, independent of any other jurisdictional limitations."11

Thus, to find subject matter jurisdiction in this case we had to resolve initially whether § 2410(a) applies in this case. For us to conclude that jurisdiction was proper at the time of judgment, not only must we justify the presence of the United States in the dispute, but we must also demonstrate how § 2410(a)(5), which covers actions in interpleader or in the nature of interpleader, applies to the parties' actions in this dispute. Once the applicability of § 2410(a) is established, federal subject matter jurisdiction is present on the basis of § 1444.

The law is well settled that the government is not subject to suit unless it has waived its sovereign immunity.12 As noted, § 2410 was specifically passed to waive the sovereign immunity of the United States so that private parties could get the government into court when necessary to quiet title or resolve priority of liens or mortgages.13 Such waiver, however, must be narrowly construed to comport precisely with congressional intent.14 Thus, "no suit may be maintained against the United States unless the suit is brought in exact compliance with the terms of a statute under which the sovereign has consented to be sued."15

In conformity with this strict construction, we have found at least three instances in which waiver of sovereign immunity does not exist under § 2410(a). There is no waiver (1) when a taxpayer seeks to challenge the validity of any underlying tax assessment,16 (2) when the government is claiming a title interest in property rather than a lien interest,17 or (3) when the government no longer has a mortgage on, or other security right in, the property in dispute.18

Even so, such strict compliance with the statute does not imply that literal interpretation of § 2410's every word is required. Standing together, the cases noted above simply reiterate that § 2410(a) applies only when the issue concerns the priority of an existing government mortgage or other security interest. On this issue, the determination of sovereign immunity is strict. In contrast, we and other courts have taken a more inclusive approach to the types of underlying relief, such as quiet title and interpleader suits, to which § 2410(a) expressly applies. Specifically, when the priority of a presently existing lien interest of the government is in dispute, and the question is whether § 2410(a) applies to the type of relief sought, the statute has been read much more expansively.

In explaining the application of § 2410(a), we have found that "jurisdiction does not depend on the specific relief sought, [e.g.] foreclosure. Rather it rests on the existence of the traditional controversy in which a private party asserts an ownership which is superior to the claimed lien of the United States Government."19 Informing this approach is our recognition of Congress's conviction "that a means be available to determine such disputes lest the absence of judicial recourse depress the marketability of property subject to federal tax liens."20 Other courts have found the applicability of § 2410(a) to underlying actions equally capacious.21

In the instant case, the government maintains an outstanding tax lien on Hussain's property. Thus, § 2410(a) appears applicable. Although the applicability of § 2410(a)(5) to the suit as a whole remains to be discussed, our prior holdings and our understanding of congressional intent predispose us to accept the government's presence in this case despite its unique mode of entrance.

2. Breadth of Interpleader Actions under § 2410(a)(5)

Given the expansive approach to determining the kinds of relief for which § 2410(a) is available, we next consider whether § 2410(a)(5) covers the state court motion practice instituted by BOC after the judgment against it and in favor of Hussain and Hibernia had become final and no longer appealable. Even though establishing the applicability of this statute in terms of the government's interest and its waiver of sovereign immunity has been accomplished, we are nevertheless required to decide whether the post-judgment proceedings in this case come within the scope of § 2410(a)(5). Specifically, we must determine (1) whether BOC's Motion to Determine Amount and Distribution of Funds constitutes an action in interpleader within the contemplation of § 2410(a)(5), and (2) whether the government's presence as respondent under a show cause order satisfies the statutory requirement that the government be "named a party in any civil action or suit."22

a. Whether There Is an Action in Interpleader

Section 2410(a)(5) expressly covers both interpleader actions and actions in the nature of interpleader. A traditional interpleader suit is an equitable action available to a plaintiff-stakeholder who is, or may be, exposed to multiple liability or multiple litigation, usually when two or more claims are brought that are mutually inconsistent.23 The purpose of interpleader is to enable the plaintiff-stakeholder to avoid "the burden of unnecessary litigation or the risk of loss by the establishment of multiple liability when only a single obligation is owing."24 Thus, traditionally the claims of the defendant claimants must be mutually exclusive and adverse to one another such that one claimant's gain in the stake would be another claimant's loss.25 In contrast to the subsequently evolved bill in the nature of interpleader, the stakeholder in a strict bill of interpleader maintains no claim or interest in the stake.26

An "action in the nature of interpleader" is a term of art that refers to those actions in which an interpleading plaintiff asserts an interest in the subject matter of the dispute.27 In all other respects, actions in the nature of interpleader are identical to traditional interpleader suits. It is § 2410(a)(5)'s reference to suits in the nature of interpleader that the district court appears to have relied on in finding jurisdiction.28 The district court was incorrect in this reading, however, because BOC had already disclaimed all interest in the insurance proceeds when it filed its motion to determine distribution of funds. Thus, BOC did not have the requisite claim or interest in the fund that is required for the plaintiffs' action to constitute one in the nature of interpleader.

Even though we have not previously determined the precise scope of an interpleader action under § 2410(a)(5), two primary factors convince us that this provision applies to the facts of this case. First, as noted above, we have taken an expansive approach to determining the types of relief to which § 2410(a) applies. Second, the equitable purpose of interpleader to protect stakeholders from multiple liability and multiple litigation has led us to construe its requirements liberally.29 In addition to the more traditional interpleader requirement that two or more claimants have competing, mutually exclusive claims to the stake, we have recognized that a plaintiff-stakeholder may employ interpleader when its liability is limited and the combined claims are in excess of such limited liability, even though not mutually exclusive.30 Further, the Supreme Court has counseled that when removal has occurred and jurisdictional requirements otherwise have been met, any problems with party labels are immaterial because the parties could have been "realigned" by the court.31

In like manner, other courts at both the federal and state levels have interpreted broadly which actions constitute an interpleader. One older district court case in the Second Circuit addressed a factual situation similar to the instant case. In United States v. Webster Record Corp., a chemical company was a state court judgment creditor of Webster Record Corporation (Webster), which in turn was a depositor in (and thus a creditor of) Bankers Trust Company (Bankers).32 To collect its state judgment against Webster, the chemical company initiated supplemental proceedings in state court against Bankers to acquire the deposit that Bankers was holding for Webster.33 During these supplemental proceedings, the IRS served a notice of levy on Bankers, demanding the deposit to satisfy a tax lien against Webster.34 Instead of appearing in the supplemental proceedings to determine lien priority over this deposit, the government sued directly in federal district court to foreclose on the lien.35

The district court recognized that any order directed at Bankers to turn over the Webster deposit to the chemical company would not bind the government, and thus would lead to further litigation.36 Without using the label "interpleader," the court also recognized that there was a substantial dispute as to priority of the claimants' entitlement to the deposited funds. As such, the appropriate action would be to join the government as defendant in a state court suit to determine lien priorities; yet, recognized the court, such an action undoubtedly would provoke the government to remove the action to federal court under the authority of § 2410(a). Thus, concluded the court, "it would be an idle gesture to permit [the chemical company] to proceed in the State Court, since its action would eventually be removed to this Court."37

In sum, the equitable purpose of interpleader to protect the stakeholder from multiple litigation and liability (expressed in Webster as the desirability of avoiding unnecessary litigation later) and the purpose of § 2410 of resolving outstanding government liens, outweighed any defects in the procedural history. Louisiana courts have also recognized that a concursus proceeding (Louisiana's version of interpleader) "should be construed liberally and given a broad application."38 In short, the broad approach toward determination of the types of relief available under § 2410(a) generally, in combination with the equitable nature of interpleader to protect a stakeholder from multiple liability and vexatious litigation, indicates that § 2410(a)(5) encompasses BOC's actions.

It is true here that the motion practice of the parties did not use the same labels as actions taken to initiate an interpleader proceeding.39 Regardless of the misleading case caption, however, the substantive posture of the parties mirrored the substance of an action in interpleader. By the time of judgment in federal court, the parties had taken several procedural steps that produced a situation in which BOC had brought Hussain, Hibernia, Rando, and the government together to contest the priority of liens and distribution of payments from the insurance fund possessed by BOC.

Although BOC did not refer to "interpleader" or "concursus," and did not deposit the fund in the registry of the court, it possessed an insurance proceeds fund of $500,000 in which it claimed no interest and to which there were several claimants. BOC had been found liable to Hussain and Hibernia under the Policy; the IRS had served notice of a tax levy on BOC for the money; and Hussain's attorney, Rando, had asserted a right to recover his contingent fee from the proceeds. In addition, those four claimants sought satisfaction from the insurance proceeds and, in the aggregate, their claims (1) exceeded the total amount available in the fund, and (2) were adverse to each other.40 Consequently, BOC had a genuine fear of multiple and vexatious litigation.

To avoid that, BOC filed its Motion to Determine Amount and Distribution of Funds with the Louisiana trial court, and asked that court to order the IRS to show cause as to its interest in the insurance proceeds fund. Similarly, in the federal district court, BOC's Motion to Determine Amount and Distribution of Funds was an attempt to bring in all who claimed an interest in the stake so that BOC would not be liable for an amount greater than the limits of the insurance policy or have to defend multiple suits. Thus, in conformity with the expansive approach taken toward this form of the equitable relief, the actions of BOC were sufficient in fact to constitute interpleader against the government under the requirements of § 2410(a)(5).

b. Whether the Presence of the United States Pursuant to § 2410(a) is Satisfied by the Order to Show Cause

The second interpretation issue presented is whether the government's presence in this suit by virtue of the state court's order to show cause satisfies § 2410(a)'s requirement that the United States be "named a party in any civil action or suit."41 We are aware of no case law, either in this circuit or elsewhere, interpreting this particular phrase from § 2410(a). Furthermore, the statutory construction of other language in § 2410 that we have undertaken points in different directions. As described above, we have narrowly construed the language waiving sovereign immunity, but we have broadly interpreted the types of relief to which this statute applies. As the requirement that the government be "named a party in any civil action" relates more to the form of relief sought in the underlying state court proceedings than to the waiver of sovereign immunity, we interpret "any civil action or suit" broadly to include the instant orders to show cause because their purpose is to determine the priority of claims to money or property.

The cases narrowly interpreting the waiver of sovereign immunity are less applicable here because the nature of the government's interest, which triggers (or prevents) the waiver of immunity, is not at stake. Rather, at issue here is simply the procedure by which the government was brought into an action to determine the ranking of its lien.

Additionally, applying § 2410 to these orders to show cause comports with the statute's purpose. On two occasions, Congress has broadened the original section 2410(a), each time allowing waiver of sovereign immunity in additional types of cases, so as to enhance the ability of private parties to resolve issues of ranking or priority when government liens are involved. The 1942 amendment, which extended the statute's ambit to include quiet title actions, "was in response to the recognized need for a way to force disputes over government tax liens to resolution, rather than leaving the United States in complete control of the timing."42 Similarly, in 1966, Congress wanted to expand the waiver of sovereign immunity to cover more types of litigation so that private parties could bring the government into those additional kinds of cases.43 We discern a pattern over time in Congress's broadening of the applicability of § 2410(a) to include more and more instances in which it was "desirable for the Government to be a party in order to assert its interest."44 Congress has clearly wanted the government to be amenable to suit in an increasing variety of actions. In keeping with this trend toward inclusiveness, we conclude that the government's presence in a suit, by virtue of an order to show cause as to the ranking of its lien in a specific fund, comes within the statute's coverage; inclusion facilitates the statutory purpose of resolving the priorities of those liens on property in which the government has a security interest.

In combination, a more expansive approach to determining the kinds of relief covered by § 2410(a) and the intent of Congress to address the "practical problems facing owners whose property was encumbered by government liens" instead of the "niceties of common law pleading,"45 dictates § 2410(a)'s inclusion of an order to show cause that brings the government into court to answer a stakeholder's call to determine lien priorities.

In sum, the district had federal subject matter jurisdiction because this case met the requirements of § 2410(a) as well as those of § 1444. Section 2410(a) only waives sovereign immunity and does not create a basis for federal subject matter jurisdiction. But, once it is deemed applicable in a state court action, such as the one here, it makes available to the government § 1444, which we have held creates a substantive right of removal to federal court, regardless of other jurisdictional limitations.46

There is one remaining complication that we must resolve before addressing to the merits of this appeal. Unlike any of the cases discussed above, BOC's interpleader-like actions were not initiated until after the state court judgment (in a case that the government was not involved in, as a party or otherwise) that required BOC to pay the insurance proceeds to Hussain and Hibernia "as their interests appear in the policy" had become final and no longer subject to appeal. The Supreme Court has interpreted the Full Faith and Credit Act to require that federal courts grant the same preclusive effect to a state court judgment as the state court would have given to it.47 Therefore, whether the Louisiana court judgment is considered final, in that sense, is a matter of state law.

The Louisiana Code of Civil Procedure provides that "[n]o claimant may be impleaded in a concursus proceeding whose claim has been prosecuted to judgment."48 The purpose of this rule is to protect the claimant, "who has prosecuted his claim to judgment, and otherwise would be forced to relitigate the matter, not only with the obligor, but with all other adverse claimants."49 In conformity with this language, the Louisiana Supreme Court has held that when claims have been "established," they may not be impleaded in a later interpleader suit.50 Although the statutory language was adopted after the period during which the court tackled these cases, its decisions give contextual meaning to the provision. Review of the cases, however, demonstrates that the requirements to preclude subsequent interpleader by an "established" claim are neither completely clear nor met in this instance.

In Victor v. Lewis, Louisiana's highest court stated that the law is well-settled that a "claimant who has been put to the test of a trial by a surety, and has established his claim, may not be impleaded later by the surety in an interpleader suit, and compelled to prove his claim again with other adverse claimants."51 In Victor, a plaintiff had successfully sued a real estate broker and an insurance company, as the broker's surety, for damages.52 The state court of appeal affirmed the surety's liability, but reversed and remanded for a determination of quantum because the surety was entitled to offset its liability by the value to the plaintiff of the use of some land. The appellate court denied a rehearing requested by the surety, and the Louisiana Supreme Court denied certiorari. On remand in district court, the surety instituted a concursus (interpleader) action and deposited the full amount of the bond with the court. Rejecting this attempt at interpleader, the state supreme court found that the plaintiff's claim had been "finally and definitely established" because three courts had reviewed the issue;53 the remaining issue relating to the value of the land during the time occupied by plaintiff was not enough to alter the conclusion that the judgment was final.54

The same court rejected a plaintiff's attempt at interpleader under an almost identical set of facts in American Surety Co. of New York v. Brim.55 The court reasoned that a party who is aware of its ability to initiate concursus but chooses not to bring such an action cannot do so after losing the original suit. The court concluded that allowing interpleader under those conditions would increase the amount of litigation, thus working against the purpose of interpleader, and would force a successful claimant to "accept a pro rata part of a fund ... and thereby largely cause her to lose the benefits of her judgment."56

In American Surety Co. of New York v. Ryan, yet another factually similar case, the Louisiana Supreme Court rejected American Surety's attempt to recoup awards it had paid to individual claimants after it had instituted a concursus action.57 Apparently relying on the doctrine of novation, the court characterized those plaintiffs as judgment creditors of American Surety, no longer being claimants to a fund.58 The factors on which the court relied were (1) the claims had been reduced to judgment before American Surety made any deposit into the registry of the court, (2) the judgments were executory against any property or fund owned by American Surety, and (3) American Surety had paid the judgments from other funds and without reference to the court-deposited fund.59

In contrast to those conclusions, it was also in Ryan that the court finally allowed interpleader against one claimant whose claim the court determined was not "established."60 This claimant, Howland & Co. ("Howland"), had won a judgment against a different company, Home Accident Insurance Company, by suing its ancillary receiver in a process that was more summary than adversarial.61 Howland then successfully sued American Surety, as surety of Home Accident Insurance Company, for payment on the unpaid judgment.62 That judgment, however, was apparently never signed, and American Surety filed a motion for rehearing and an order for Howland to show cause why the rehearing should not be granted.63 The proceedings stopped there because American Surety filed its interpleader action at about the same time.64

In rejecting Howland's assertion that its claim was established, the court found that Howland's prior judgment was based on insufficient proof; that it was never signed; and that the motion for rehearing was never adjudicated.65 The court also based its conclusion on the fact that the prior judgment was executory only against American Surety for no more than the amount of the bond it held, thereby restricting recovery to the available proceeds from an insurance bond.66 The court surmised that these factors prevented Howland's claim from being "established," thereby freeing the stakeholding plaintiff to bring Howland into a subsequent concursus proceeding.67

These Louisiana cases suggest that any plaintiff who has prosecuted his case through the appellate levels and achieved a judgment for which there is no further avenue of litigation cannot later be brought into a suit in interpleader. And, at first glance, the facts of our case suggest that Hussain and Hibernia had indeed established their claims through a final, executory judgment, thus blocking their being interpled subsequently by BOC: Hussain and Hibernia had prevailed at the trial and appellate court levels, and BOC's writ application had been denied by the Louisiana Supreme Court, before the government ever asserted its tax lien and, necessarily, before BOC could have responded by filing its show cause motion. Thus, as described in Victor, "th[e] matter was passed upon by three courts in definitive judgments."68

Closer scrutiny reveals, however, that this case is distinguishable from prior decisions. The positions of the parties in this case are factually different from past examples of "established" claims such that this case does not come within the scope of article 4652 of the Louisiana Code of Civil Procedure prohibiting concursus (interpleader) against claimants who have already prosecuted their claims to judgment.69

Like Howland, the unestablished claimant in Ryan, Hussain and Hibernia's aggregate recovery is limited to the policy proceeds, and thus is not a general judgment against BOC itself. The initial state court judgment held BOC liable to Hussain and Hibernia for 500,000, "the policy limits under Business Owner's Xtra Policy... as their interests appear in the Policy." In fact, the trial court specifically rejected all other claims brought against BOC, including those for penalties for arbitrary failure to pay claims under the policy. Thus, as in Ryan, the trial court specifically restricted the judgment to the amount of the policy (plus interest and costs), and prohibited any recovery against the insurance company in excess of that amount.

Furthermore, this case is unlike Brim, because it is not clear that BOC had "knowledge of all the facts" such that it could have chosen to proceed in concursus originally.70 During the liability stage of this litigation, only Hussain and Hibernia had made claims to the insurance proceeds. Although a situation in which two parties sue for the same funds is often ripe for concursus, such is not the case when the combined claims fail to exceed the fund and are not mutually exclusive. As Hibernia's claim was derivative of Hussain's and as the cases were consolidated relatively early in the litigation, there was apparently no fear of multiple liability or multiple litigation to prompt BOC to file a concursus proceeding. In fact, notice of the IRS lien on Hussain's property and its threatened legal action did not appear until December 2000, which was subsequent to the Louisiana Supreme Court's denial of certiorari.

It is also evident that further proceedings in this instance likely would not increase litigation. Regardless of whether the government remains a party, the merits of this case indicate that Rando and Hibernia would continue to dispute the priority of their claims in the fund: A court would still have to sort out which contingency fee percentage Rando was entitled to receive, and whether the demand for expert fees was proper. Thus, even if a concursus or interpleader were not allowed in this case, additional litigation would still arise.

Neither would Hussain lose the benefits of his judgment if interpleader were allowed. State court litigation ensured that BOC was liable to Hussain and Hibernia for the full policy proceeds plus interest and costs. Further litigation would not alter the judgment against BOC. If future litigation should force Hussain to accept a reduced recovery, it will be because the government partially satisfies its tax lien against Hussain through the interpleader proceeding; but the government would have obtained this result regardless. Without interpleader, Hussain would have taken the remaining policy proceeds, and would have been forced to pay these proceeds in separate litigation with the government to satisfy tax arrearages. Allowing interpleader simply skips this unnecessary step, and enables the government to collect without engaging in duplicative litigation.

Finally, this case is unlike Victor v. Lewis because of the remaining substantive legal issues requiring resolution. The Victor court found that factual determination of the value of land occupancy did not upset the "established" character of the plaintiff's claim. Here, in contrast, the parties will continue to advance substantive legal disputes after the finding of liability. Such issues do not call into question the final judgment against BOC, but they are significant enough, in our view, to allow litigation to continue through interpleader.

In conclusion, even though Hussain and Hibernia's judgment against BOC is a final, unalterable state court judgment, it is not "established" in the sense of Louisiana jurisprudence and does not preclude a subsequent interpleader-like action under these unique facts. The state court judgment was specifically limited to the insurance policy proceeds and it was not practical, or expected, for BOC to attempt an interpleader-like action until it became aware of the government's lien. Furthermore, none of the dangers that accompany subsequent interpleader are present. The remaining issues concern substantive questions of priority, and would have arisen even in the absence of this interpleader-like action. If Hussain's award is reduced, it will be because of Hussain's own tax liabilities and not because of an independent claim to the fund. Allowing this action may actually bring a benefit to Hussain, because it preempts separate litigation between him and the government and, as a result, saves him attorney fees in such an action. Thus, because these proceedings do not threaten relitigation or the reconsideration of BOC's liability, but instead provide a potential benefit to Hussain, we see no reason why an interpleader-like action should not have proceeded in this particular case.

Having resolved jurisdictional and final judgment ambiguities, the first merits issue is whether Rando was entitled to have attorneys' fees calculated on Hussain's net recovery under the insurance policy, as the district court held, or on his gross recovery, as Rando argues. An alternative way of viewing the question is: Does Hibernia, as a named loss-payee mortgagee, prime the secured claim of Rando, the insured's attorney, in the distribution of insurance proceeds from BOC?

The district court gave priority to Hibernia over Rando, reasoning that "the loss payee-mortgagee (Hibernia) is entitled to the proceeds of the policy to the extent of the mortgage debt, with any surplus then payable to the insured-mortgagor (Hussain)."71 Rando argues on appeal that (1) Hibernia is a loss payee under a simple or open loss-payable clause, not a standard or union clause; (2) under an open loss-payable clause, the loss payee recovers only if the insured recovers; ergo (3) Hibernia's right to recover is contingent on, and derivative of, Hussain's right to recover. In Rando's view, this is the reason that the state court awarded judgment in the insurance suit for Hussain and Hibernia "jointly" — as the Policy's Loss Payable provision required. Hence, Rando concludes, the contingency-fee rate should apply to Hussain's gross recovery from BOC, not to his recovery net of BOC's payment to Hibernia.

Hibernia counters that (1) Louisiana law clearly ranks the loss-payee ahead of the insured; (2) there is no basis for applying Rando's contingency fee to Hibernia's recovery; and (3) Rando's distinction between types of loss payees is irrelevant in this case. Hibernia also emphasizes that it has independently prosecuted this suit, having been the first to sue, thereby interrupting prescription and commencing the accrual of judicial interest.

Rando's argument that he has priority over Hibernia has little support in either statutory or case law. The relevant state statute, La.Rev.Stat. § 9:5001(A), states:

A special privilege is hereby granted to attorneys at law for the amount of their professional fees on all judgments obtained by them, and on the property recovered thereby, either as plaintiff or defendant, to take rank as a first privilege thereon superior to all other privileges and security interests under Chapter 9 of the Louisiana Commercial Laws.72

Even from the statute alone, one might conclude that "judgment obtained by them," in a consolidated case, refers to the judgment an attorney obtains for his client, and not any amount obtained by another party. This view is reflected both by our decisions and those of the Louisiana courts.

On several occasions, we have recognized that "when a mortgagee is designated as the loss payee, the insured in effect appoints him to receive payment under the policy."73

It is well established as the law of Louisiana that where insurance is taken out by the mortgagor for the benefit of mortgagee, or is made payable to the mortgagee as his interest may appear, the mortgagee is entitled to the proceeds of the policy to the extent of his mortgage debt, holding the surplus, if any, after the extinguishment of his debt for the benefit of the mortgagor. Adams v. Allen, 19 So. 2d 578 (La.App. 1st Cir.1944).74

We have gone so far as to apply this principle to create an equitable lien in favor of the mortgagee, entitling it to recover insurance proceeds, even when the insurance policy did not name the mortgagee as a loss payee.75

Rando cites Lazlo v. State Farm Fire & Casualty Co.,76 for the proposition that he should recover on the gross amount BOC paid out, not merely the net amount. But, as Hibernia and the government assert, Lazlo is distinguishable. In that case, fire consumed a house that was insured and mortgaged; and, as here, the mortgagee was a named loss payee in the insurance policy.77 Unlike this case, however, the insurance company in Lazlo purchased the mortgage from the mortgagee.78 Litigation over the arson issue resulted in cancellation of the mortgage. We held that the mortgage became property "recovered" for the homeowner through the arson litigation; the value of the mortgage, under La.Rev.Stat. § 9:5001, was thus to be included in the attorneys' fees privilege.79

At first glance, it does appear difficult to distinguish Lazlo meaningfully from this case: Both plaintiffs had loss-payee provisions in their mortgages directing the insurance company to pay the mortgagee in the event of losses to the mortgaged property. The most significant difference is that the insurance company bought the mortgage from the mortgagee in Lazlo and cancelled it after litigation.80 This difference, however, in no way increased the insured's recovery beyond the surplus after satisfaction of the mortgage; neither does it reflect any greater professional effort on the part of Lazlo's attorney. Irrespective of whether the mortgage was bought and cancelled, the loss-payee provision made certain that the professional efforts of Lazlo's lawyer would never have produced a recovery of more than the surplus policy funds. If the mortgage had not been cancelled, then the insurance company likely would have paid itself the funds due under the loss-payee provision, thereby paying off Lazlo's mortgage. Thus, if Lazlo's reduced recovery was a foregone conclusion because of the loss-payee provision, it is hard to see why Rando should be deprived of the fees on the total policy when the loss payee provisions similarly reduced Hussain's recovery.

Nonetheless, Lazlo remains distinguishable. Assigning the mortgage to the insurer may seem like a procedural technicality as far as the insured's ultimate recovery is concerned, but it did have the practical effect of covering the mortgagee's interest in the mortgage. Unlike the mortgagee in Lazlo, the mortgagee in this case, Hibernia, did not convey or assign the mortgage to BOC. Instead, it pursued litigation on its own to protect the amount it was entitled to receive as loss-payee. Although Rando is correct that to recover Hibernia needed Hussain, that statement is only correct insofar as Hussain's action may have been a technical prerequisite to recovery from BOC.

The law is clear that even under an open or simple loss payable provision, such as the one in operation here, the mortgagee has a direct claim to the insurance proceeds up to the amount necessary to cover the outstanding balance on the obligation secured by that mortgage.81 The only drawback with this type of provision, as opposed to a standard or union clause, is that the mortgagee cannot recover if the insured is somehow at fault for the loss. Thus, if Hussain's presence in the suit was necessary at all, it was merely to allow a defense against arson allegations. As Hibernia notes, it initiated litigation itself a year before Hussain sued; and Hibernia obtained a judgment against Hussain which liquidated the amount due to it under the loss-payable provision.

This difference from Lazlo is material. Unlike Lazlo, the mortgagee's financial interest was not satisfied through an assignment. Instead, the mortgagee litigated to ensure its receipt of the amount outstanding on the note. Thus, Rando cannot alone claim credit for recovering the amount that was due directly to Hibernia as a result of the loss payee provision; neither was he the only lawyer — or even the first one — presumably working to refute the arson allegation. In light of the well-established character of Hibernia's right to the money, and the work of its counsel in the litigation, Rando is not entitled to an award of fees on that portion of the recovery.

Finally, despite any remaining similarities between Lazlo and this case, the weight of authority both in our precedent and under Louisiana's indicates that a mortgagee's rights under a loss-payee provision vest automatically when a loss-causing incident occurs.82 As a result, any amount ultimately recovered by Hussain was certain to be, from the outset, net of the outstanding balance on his mortgage to Hibernia.83

Simply stated, then, Rando's contentions that (1) he has priority over Hibernia, and (2) his fee should be calculated on the gross amount paid out by BOC rather than the net amount recovered by Hussain, are wrong. Thus, we affirm the district court's allowance of the net amount only.

Rando also contends that the district court erred in applying as his contractual contingency fee percentage the 33 1/3% figure from his 1995 fee agreement with Hussain rather than the 39% figure from their 2000 fee agreement. The district court concluded that Rando's super-priority for attorney's fees under 26 U.S.C. § 6323(b)(8) only applied to his success in having BOC held liable for the insurance proceeds, not his post-judgment effort to collect the monies.84 Thus, reasoned the court, Rando only deserved fees under this provision for the work done pursuant to the 1995 agreement, which specified a contingency fee of 33 1/3%. On appeal, Rando chiefly argues that under Louisiana law, "the highest reasonable contingency fee controls."

No party disputes that Rando's effort to achieve a judgment against BOC for the insurance fund entitles him to a super-priority ranking for his fees over the federal tax lien. Rather, the issue is whether Rando can substitute a higher contingency fee percentage after the adjudication of BOC's liability but before any of the funds are actually distributed. Stated differently, we must determine whether § 6323(b)(8)'s super-priority ranking attaches when BOC's is found liable, or when the proper distributions from the insurance fund are actually determined. To answer this question, we must first determine the scope of the term "judgment" as used in the relevant statute.

Section 6321 of the Internal Revenue Code establishes "a lien in favor of the United States upon all property and rights to property, whether real or personal" against a delinquent taxpayer.85 This lien attaches at the time of assessment and to all property owned or subsequently acquired by the taxpayer.86 Section 6323(b)(8) creates an exception to this federal tax lien priority. In part, it provides that a previously filed federal tax lien shall not be valid, "[w]ith respect to a judgment or other amount in settlement of a claim or of a cause of action, as against an attorney who, under local law, holds a lien upon or a contract enforcible [sic] against such judgment or amount, to the extent of his reasonable compensation for obtaining such judgment or procuring such settlement."87 The announced purpose of this provision is to provide an incentive to attorneys to enhance the value of a taxpayer's property, which would ultimately increase the government's revenue collection.88

We have already determined that, because of continuing disputes about the proper distributions from the insurance fund, the adjudication of BOC's liability constituted a final state court judgment, but did not preclude a subsequent interpleader-like action. This portion of the case asks, as a question of law, whether the "judgment" of which § 6323(b)(8) speaks requires actual distribution of the fund or merely a determination of liability. As such, the issue requires our de novo review.

We have not had the opportunity to interpret § 6323(b)(8) in the context of the specific issue we review today. Other courts have concluded that "obtaining such judgment"89 means that an attorney must have created a fund which increases the taxpayer's taxable property.90 Thus, it is fairly well-settled that the word "judgment" as used in § 6323(b)(8), must be read in light of the statute's purpose; that is, the creation of a fund that increases the taxpayer's property for the ultimate benefit of the IRS.

What is required to create such a fund is much less clear. Although to our knowledge no court has addressed the precise fact pattern that is now before us, other courts have confronted this general question in slightly different contexts. For instance, some courts have taken "a more expansive view of what constitutes the creation of such a fund."91 These fora have found that both recovery of seized funds from the government92 and efforts to obtain confirmation of an arbitration award93 constitute fund creation within the intendment of § 6323(b)(8). Another court has concluded that "judgment" in the statute "refers both to the judicial act and that which in whole or in part satisfies the `judgment'," suggesting that fees charged for efforts to collect a judgment deserve super-priority treatment as well.94

In contrast, appearing more restrictive, several courts have held that such other activities fail to create a fund and thus fail to constitute a judgment under the statute. One court has found, for example, that the filing of a suit against a party who responds by interpleading the plaintiff and depositing an amount into the court registry is not enough to constitute creating a fund and thereby a "judgment" from which attorney fees are warranted.95 Similarly, courts have concluded that interlocutory decisions, even if creating new funds, do not come within the statutory meaning of "judgment."96

These cases provide little clear guidance for us because they neither establish a uniform interpretation nor address the particular circumstances of this case. We face a situation here in which the issue of BOC's liability has been exhausted in state courts and accepted as final in federal court. The final resolution of this case, however, is still open because some of the parties continue to compete for priority and quantum of the limited insurance proceeds. Even though the judgment here may not be final in terms of all distribution issues, however, it is final in that, in the end, BOC cannot avoid or reduce its obligation to pay out all the insurance fund proceeds and interest.

If we were to conclude that creation of a fund, and thus the term "judgment," requires actual distribution, § 6323(b)(8) would give super-priority to all the fees charged, from the very beginning through the very last distribution from the fund. The renegotiation of the contingency fee, as a result, would appear to apply to whatever is the final distribution. The problem with this interpretation is that the outcome it dictates undermines the purpose of § 6323(b)(8), i.e., increasing the value of the taxpayer's property to better satisfy outstanding tax liens. Once BOC is found liable, the fund is created; any subsequent effort cannot increase the amount of the insurance proceeds or the government's recovery. Indeed, because there is a limited fund available for distribution, the outcome advocated by Rando would increase his fees but decrease Hussain's potential recovery. Reducing Hussain's potential recovery to add to Rando's take would, in turn, reduce the property available to satisfy the tax lien. Such a result is diametrically opposed to that intended by Congress in enacting § 6323(b)(8).

In contrast, applying § 6323(b)(8) only to Rando's effort in securing BOC's liability comports with the purpose of the statute. The point at which BOC is held liable is the point at which a fund is created in the taxpayer's favor. Any litigation subsequent to this point, albeit necessary to bring the case to its final conclusion, does nothing to increase the value of the taxpayer's property. Indeed, as indicated above, negotiation of a higher contingency fee would actually reduce Hussain's theoretical recovery and thereby diminish the value of the property available to satisfy the federal tax lien.

Furthermore, this conclusion is not likely to lessen the incentive of attorneys in cases of this nature as it does not preclude super-priority treatment of attorney fees. For his effort in finding BOC liable, Rando will still receive fees equaling 33-1/3% of Hussain's recovery. Thus, lawyers in future cases will still have an incentive to represent delinquent taxpayers.

We conclude that only the fees earned in litigating BOC's liability deserve super-priority under § 6323(b)(8), and these fees are assessed pursuant to the original 33 1/3% contingent fee agreement. As this is the approach taken by the district court, we affirm its ruling on the issue.

The final substantive issue concerns whether the district court erred in reducing the expert fee of John Theriot, CPA, from $24,509.50 — the amount he billed, which the court described as "obscene" — to $5,000, and awarding that reduced amount.97 On appeal Rando simply argues that the attorney's privilege to fees under La.Rev.Stat. 9:5001 extends to all court costs and litigation expenses advanced by the attorney for the client's benefit. The government, on the other hand, contends that Rando was entitled to no expert fee for Theriot because he never testified; and in the alternative, that the district court's assessment was reasonable.

The district court accepted Rando's contention that Theriot would have been called had the case gone to trial instead of being concluded by directed verdict, and therefore reasoned that some fee was appropriate.98 The court justified the reduced fee by applying Louisiana's balancing test for allocating experts' fees, which weighs such factors as (1) the amount of time the expert spends in preparing for trial; (2) the amount of time he spends in court; (3) his expertise; (4) the amount he charges; (5) the amount involved in the award; and (6) the degree to which the opinion of his experts aid the court.99

Our review of the district court's opinion is conducted under the very deferential abuse of discretion standard. Even under this standard, however, we must conclude that, in this instance, the district court abused its discretion by granting any fee whatsoever.100

It is well settled under Louisiana law that, because costs can be awarded pursuant to statute only, and because Louisiana has no statute that provides costs for the fees of experts who do not testify, any taxation for such fees as costs is an error of law and thus an abuse of discretion.101 This is consistent with Louisiana's rule that, if a deposition is not used as evidence at trial, the costs in procuring it cannot be taxed as costs.102

The district court acknowledged the rule that fees of experts who do not testify cannot be taxed as costs, but concluded that because Rando intended to call Theriot and was only prevented from doing so by the court's rendering of a directed verdict, some such costs should be awarded. Louisiana precedent indicates, however, that the reason an expert does not or is unable to testify is irrelevant. Courts have held in particular that such costs should be denied even when an expert's inability to testify arises involuntarily.103

Regardless of the court's role in preventing Theriot's testifying, the fact remains that he never testified in person and never had a report or other testimony, such as deposition or affidavit testimony, entered into evidence. Louisiana law dictates unequivocally that, under these circumstances, none of Theriot's charges can be taxed as costs.

For the reasons stated herein, we affirm the district court's holdings as to (1) subject matter jurisdiction, and (2) the priority and amount of BOC's distributions from the insurance proceeds fund to Hibernia, Rando, and the United States, including the court's ruling on the appropriate contingency fee percentage used in calculating Rando's recovery and the principal sum to which that percentage should be applied. We reverse, however, the district court's taxation of any part of Theriot's expert fee as costs. We therefore vacate the court's original judgment and remand with instructions for the district court to enter a new judgment with updated and corrected distribution amounts, and with no award of expert witness fee.

AFFIRMED in part; REVERSED in part; and REMANDED with instructions.