Dresser Industries, Inc. v. United States

IN THE UNITED STATES COURT OF APPEALS FOR THE FIFTH CIRCUIT _________________ No. 99-11231 _________________ DRESSER INDUSTRIES, INCORPORATED, Plaintiff-Appellant, VERSUS UNITED STATES OF AMERICA, Defendant-Appellee. ________________________________ Appeal from the United States District Court for the Northern District of Texas ________________________________ January 10, 2001 Before SMITH and DENNIS, Circuit Judges, and HARMON, District Judge.1 MELINDA HARMON, District Judge: This is an appeal from a suit for tax refund in which the district court ruled against the taxpayer. Plaintiff-Appellant Dresser Industries, Inc. argues that the district court erred when it held on summary judgment that: (a) Treasury Regulation 1.861- 8(e) disallows “interest netting”; (b) interest liability exists on deficiencies later eliminated or reduced by foreign tax credit carrybacks; and (c) such interest accrues until the filing date of 1 District Judge of the Southern District of Texas, sitting by designation. the return of the tax year in which the foreign tax credit arises. Finding that the district court correctly ruled in favor of the United States in light of express statute and applicable case law, we affirm. I. Dresser Industries, Inc. (“Dresser”) is a worldwide supplier of technology, products, and services to industries involved in the development of energy and natural resources. Dresser is subject to the Coordinated Examination Program, and, as a result, is under continuous audit by the Internal Revenue Service. In 1972, Dresser established Dresser International Sales Corporation (“Dresser International”) as a wholly owned subsidiary. Dresser International qualified as a Domestic International Sales Corporation (“DISC”) to take advantage of Congress’s overall strategy to increase domestic exports by providing tax incentives to companies involved in export trade.2 A qualified DISC subsidiary is not taxed on income derived from the sale of exports; rather, its shareholders are taxed on a specified percentage of DISC taxable income as if a dividend distribution had been made at the end of the tax year. DISC taxable income, from which this constructive dividend distribution is calculated, is based on a complex statutory framework that establishes a “deemed” transfer 2 See Revenue Act of 1971, Pub.L. No. 92-178, 85 Stat. 535 (1971), codified as amended at 26 U.S.C. §§ 991-997. For a more complete discussion on the structure of DISCs, see Dresser Indus., Inc. v. United States, 73 F. Supp.2d 682, 684-85 (N.D. Tex. 1999). 2 price for export goods provided to the DISC by the parent supplier. The taxpayer calculates the deemed transfer price as 50% of the “combined taxable income” of the DISC and its parent. In 1984, when Congress replaced the DISC provisions of the tax code with the Foreign Sales Corporation (“FSC”), Dresser responded by incorporating Dresser Foreign Sales Corporation. The FSC serves essentially the same purpose as the DISC, except a taxpayer calculates combined taxable income using a 23% standard instead of a 50% one.3 While the Internal Revenue Code governs transfer prices applicable to DISCs and FSCs, the Treasury Regulations provide rules governing the allocation of expenses, losses, or deductions in computing the combined taxable income from those sources. Allocation of interest expenses in the instant case is governed by the 1977 version Treasury Regulation § 1.861-8(e). That Regulation provides that “the aggregate of deductions for interest shall be considered related to all income producing activities and properties of the taxpayer and, thus, allocable to all the gross income which the income producing activities and properties of the taxpayer generate, have generated, or could reasonably have been expected to generate.” Treas.Reg. 1.861-8(e)(2)(ii) (as amended in 1977). 3 For a thorough treatment of the FSC, see generally Note, The Making of a Subsidy, 1984: The Tax and International Trade Implication of the Foreign Sales Corporation Legislation, 38 STAN. L. REV. 1327, 1334-55 (1986). 3 The issues in this case arise from an audit of Dresser that ended in September 1993. At the conclusion of the audit, Dresser paid additional taxes and interest for taxable years 1980, 1981, 1982, 1984, 1986, and 1987, but subsequently filed formal claims for refund. In these claims, Dresser asserted that, in apportioning interest expenses between its DISC and FSC activities and its non-DISC and non-FSC activities, it had erroneously allocated gross income expense, and that it should have first offset interest expense against interest income, and then allocated only the net interest expense. The purpose of this practice, called “interest netting,” would be to maximize the income treated as included in the combined taxable income of Dresser and its DISC and FSC from exports and foreign trade; the advantage would thus be to increase the amount of income eligible for the favorable tax benefits conferred by Congress on the DISC and FSC. Dresser also asserted that use of interest netting in its 1983 taxable year gave rise to an additional foreign tax credit carryback to its 1981 taxable year in the amount of $257,236; that use of interest netting in 1985 resulted in an additional net operating loss carryback to its 1982 taxable year; and that use of interest netting in its 1984 taxable year resulted in an overpayment of its tax liability for that year. Accordingly, based on the technique of interest netting, Dresser sought tax refunds for its 1981, 1982, and 1984 taxable years. 4 In addition to its claims for refunds based on interest netting, Dresser sought refunds of interest it previously had paid on deficiencies in its 1981 and 1984 tax liabilities. Dresser had previously filed a petition in the Tax Court contesting a deficiency in its 1981 tax liability; as a result of Dresser’s execution of a Waiver of Restrictions on Assessment and Collection of Deficiency in Tax and Acceptance of Overassessment, and in light of the Tax Court’s determination of tax deficiencies for those years, Dresser was allowed to carry back excess foreign taxes from its 1983 taxable year to its 1981 taxable year. Dresser was also allowed to carry back excess foreign taxes from its 1986 taxable year to its 1984 taxable year. These carrybacks effectively reduced or eliminated Dresser’s tax deficiencies for 1981 and 1984. Dresser sought refunds of the interest that it had paid on the deficiencies that existed for those years because the foreign tax carrybacks reduced or eliminated the initial deficiencies. The Internal Revenue Service (“IRS”), in response, not only rejected Dresser’s argument of using foreign tax credit carrybacks to eliminate deficiency interest owed, but it also rejected Dresser’s claim of interest netting and the subsequent refunds arising from interest netting. Instead, the IRS maintained that only a ratable share of “gross interest” could be apportioned to the DISC and FSC. 5 Dresser eventually sued the United States (“Government”) for tax refund in the United States District Court for the Northern District of Texas, seeking, inter alia, refunds of the federal income taxes and interests that would arise from interest netting and elimination of Dresser's deficiency interest. Dresser's desired refunds totaled $2,585,776. On cross motions for partial summary judgment, the district court held in favor of the Government.4 While the court observed that an earlier version of Treas.Reg. § 1.861-8 allowed for interest netting, the court held that the version of the Regulation applicable to Dresser's case specifically forbade interest netting. The district court also held that Dresser was not entitled to a refund of the interest it had paid with respect to deficiencies in its 1981 and 1984 tax liabilities that were later reduced or eliminated as a result of the carryback to those years of excess foreign tax credits from Dresser’s 1983 and 1986 taxable years. The district court further held that Dresser’s liability for interest continued to accrue 4 The court first considered the Government’s procedural argument that Dresser’s claim for refund for its 1981 taxable year, based on a carryback of a foreign tax credit, was both untimely and barred by § 6512(a) of the Internal Revenue Code. Section 6512(a) prohibits obtaining tax refunds for taxable years previously before the Tax Court. The district court concluded that, although Dresser’s claim for refund was timely filed pursuant to § 6511(a) (which allows filing of a claim within two years from the date a tax was paid), Dresser’s claim was nonetheless barred by § 6512(a)’s prohibition. The district court also concluded that Dresser’s claim for refund for its 1982 taxable year was untimely, but that there was a genuine issue of material fact as to whether Dresser had made a timely “informal” claim for refund for that year. Because the court subsequently held that interest netting was not permissible, the court did not resolve the issue of “informal” claim. Dresser and the Government, in their briefs, both recognize that if this Court were to hold that the Treasury Regulations allow for interest netting, a remand would be required to permit the District Court to resolve the issue of whether Dresser made a timely informal claim for its 1982 taxable year. 6 until the due date for filing tax returns for the years in which the excess foreign tax credits arose. Dresser now appeals the district court's holding on these three substantive issues. II. This Court reviews the district court's granting of summary judgment for the Government de novo, applying the same standards as those applied by the district court in making its determination. Neff v. American Dairy Queen Corp., 58 F.3d 1063, 1065 (5th Cir. 1995) (citing McDaniel v. Anheuser-Busch, Inc., 987 F.2d 298, 301 (5th Cir. 1993)); Cavallini v. State Farm Mut. Auto Ins. Co., 44 F.3d 256, 266 (5th Cir. 1995) (citing LeJeune v. Shell Oil Co., 950 F.2d 267, 268 (5th Cir. 1992)). The first issue presented on appeal is whether Treasury Regulation § 1.861-8(e)(2) permits Dresser first to offset interest income against interest expense and then apportion only the “net” interest expense between its DISC/FSC activities and its non- DISC/FSC activities in computing combined taxable income attributable to qualified export receipts or foreign trading gross receipts. Dresser argues that § 1.861-8(e)(2) allows for interest netting, and, as a result, Dresser calculates that it is entitled to refunds for its 1981, 1982, and 1984 taxable years. In support of its position, Dresser invokes this Court's decision in Dresser Indus., Inc. v. Commissioner, 911 F.2d 1128 7 (5th Cir. 1990), which explicitly ruled in favor of interest netting. Specifically, we held in that case that “Dresser may offset its interest income against its interest expense and allocate the net interest expense to its DISC for purposes of calculating [combined taxable income].” Id. at 1136. Our 1990 Dresser opinion, however, analyzed the 1975 version of Treas.Reg. 1.861-8(a), which provided in pertinent part: From the items of gross income specified in §§ 1.861-2 to 1.861-7, inclusive, as being income from sources within the United States there shall be deducted the expenses, losses, and other deductions properly apportioned or allocated thereto and a ratable part of any other expenses, losses, and other deductions properly apportioned or allocated thereto and a ratable part of any other expenses, losses, or deductions which cannot definitely be allocated to some item or class of gross income. Treas.Reg. § 1.861-8(a) (as amended in 1975). While this regulation set out the method of allocating “expenses, losses, and other deductions,” it did not define the amount of deductions to be allocated. Id. Additionally, “the specific code sections pertaining to DISCs, the legislative history, and the applicable Treasury Regulations [were] silent with regard to the amount of interest expense to be allocated in calculating” combined taxable income. Dresser, 911 F.2d at 1135. Such ambiguity gave rise to our consideration of competing interpretations of the 1975 Regulation. On the one hand, the Commissioner in Dresser urged that “expenses” under the Regulation meant “specific itemized deductions set out” in the Tax Code, an 8 interpretation that would dictate that gross interest expense would be apportioned. Id. On the other hand, Dresser argued the “equally acceptable interpretation” that interest expense among related business operations was “the actual cost of financing those operations.” Id. We found the latter reading more consistent with the legislative intent of the DISC and more agreeable to “the realities of business of finance.” Id. The Treasury Regulations take the view that money is fungible. Id. (citing Treas.Reg. 1.861- 8(e)(2)(I) (as amended in 1988)). Given such fungibility, we appreciated that, in the normal business context, the amount and timing of business borrowing rarely, if ever, correlate to specific investments. Id. For example, a business may incur a debt in a single transaction, even though its cash requirements are spread out over the ensuing quarter. The business may then choose to reduce the cost of holding these as-yet unneeded funds by investing in short-term, interest-bearing instruments. Id. (citing Ideal Basic Indus. v. Commissioner, 82 T.C. 352, 400 (1984)); see also Portland General Cement Co. v. United States, 628 F.2d 321, 342-43 (5th Cir. 1980) (holding that the actual cost of borrowing was the amount properly allocable to mining operations so as not to allocate a disproportionate share of the business's financing costs to a specific phase of its operations). In such a case, “the total cost of the borrowing is the interest expense on the debt incurred, 9 reduced by the interest earned on the investment of any temporary cash surplus.” Dresser, 911 F.2d at 1135. We concluded in the 1990 Dresser case that requiring the allocation of gross interest expenses in this kind of business context “would burden the DISC with a disproportionate share of the actual borrowing costs attributable to all operations, not merely to export operations.” Id. at 1136. Therefore, with respect to the 1975 version of Treas.Reg. 1.861-8(e), we did “not believe that Congress contemplated or intended this result when it enacted the DISC legislation, and we [found] nothing in the statute, the legislative history, or the applicable Treasury Regulations that contradict[ed] our belief.” Id. Rather, under the 1975 version of the Regulation, “interest expense to be apportioned among related business operations [was] the actual cost of financing those operations.” Id. This Court recognized in Dresser, however, that Congress substantially revised the 1975 version of Treas.Reg. 1.861-8 in 1977. Id. at 1134, n.11. Moreover, we acknowledged that our reasoning in that case did “not address whether interest netting is consistent with the [1977] version of the Treasury Regulations, or whether the [1977] Regulations are consistent with Congressional intent underlying the original DISC legislation.” Id. Although it is the 1977 version of the Treas.Reg. 1.861-8(e) that is at issue in the instant case, Dresser presently argues that 10 the 1977 version does not call for any different result with regard to interest netting. Indeed, Dresser insists that the only difference between the 1975 and 1977 versions of Treas.Reg. § 1.861-8(e) merely involves the method of allocating and apportioning expense items, not the amount of an expense item. Dresser thus claims that the district court committed error when it failed to follow the 1990 Dresser opinion and instead held that the plain language of the revised Regulation disallows interest netting. In contrast, the Government argues that the 1977 version of Treas.Reg. § 1.861-8, by its explicit terms, does not permit interest netting, and instead requires allocation of gross interest expenses between the DISC/FSC and non-DISC/FSC activities. We agree with the Government. The plain language of the 1977 version, as noted above, specifically states that “the aggregate of deductions for interest shall be considered related to all income producing activities and properties of the taxpayer and, thus, allocable to all the gross income which the income producing activities and properties of the taxpayer generate, have generated, or could reasonably have been expected to generate.” Treas.Reg. § 1.861-8(e)(2)(ii) (emphasis added). The 1977 revision moreover provides that [t]he method of allocation and apportionment for interest set forth in this paragraph is based on the approach that money is fungible and that interest expense is attributable to all activities and property regardless of any specific purpose for incurring an obligation on which interest is paid. 11 This approach recognizes that all activities and property require funds and that management has a great deal of flexibility as to the source and use of funds. Treas.Reg. 1.861-8(e)(2)(I) (emphasis added). Thus, while the Regulation's adherence to the principle of money fungibility remains intact in the newer version of Treas.Reg. 1.861-8(e), the new Regulation nonetheless explicitly mandates that interest be calculated by “all income producing activities” and “all the gross income.” Treas.Reg. 1.861-8(e)(2). Cf. Bowater, Inc. and Subsidiaries v. C.I.R., 108 F.3d 12, 14 (2nd Cir. 1997) (finding it ironic that the Regulation, in mandating that interest expense be allocated to all income producing activities, and the taxpayer, in arguing for interest netting, both cited the principle of fungibility of money); Dresser Indus., Inc. v. United States, 73 F. Supp.2d 682, 693 (N.D. Tex. 1999) (“Indeed, the revised regulations use the concept of fungibility to reach the exact opposite conclusion as the [1990] Dresser court.”) We therefore hold that the plain language of the 1977 version of Treas.Reg. 1.861-8(e)(2) disallows interest netting and precludes taxpayers from offsetting interest expense against interest income with only the balance being attributable to other income producing activities. The two-tiered allocation mechanism urged by Dresser, where interest income and expense are netted and then allocated only for any remaining interest expense, finds no place in the revised language of the Regulation. Instead, the 12 unambiguous language of Treas.Reg. 1.861-8(e)(2) orders that gross interest expenses be ratably allocated to “all income producing activities.” Treas.Reg. 1.861-8(e)(2) (emphasis added). Although “the realities of business finance,” Dresser, 911 F.2d at 1135, may not have changed since our 1990 Dresser decision, the plain language of the applicable Regulation clearly has. Our interpretation of the 1977 language of Treas.Reg. 1.861- 8(e) parallels the Second Circuit's in Bowater, Inc. v. Commissioner, 108 F.3d 12 (2nd Cir. 1997).5 As in this case, the taxpayer in Bowater insisted that Treas.Reg. § 1.861-8(e) permitted the netting of interest income and expense, and apportionment of only net interest expense among the taxpayer’s other income producing activities. Id. at 14. The taxpayer, in essence, argued that the Regulation's “all income producing activities language” should be read as “all income producing activities except those that produce income in the form of interest.” Id. In rejecting the taxpayer's argument, the Bowater court found that there is no relevant difference between debt and equity investments under Treas.Reg. 1.861-8(e)(2) because “both involve 5 In Bowater, the Second Circuit considered the 1978 version of Treas.Reg. 1.861- 8(e)(2), which provides that “the aggregate of deductions for interest shall be considered related to all income producing activities and properties of the taxpayer and, thus, allocable to all the gross income which the income producing activities and properties of the taxpayer generate, have generated, or could reasonably have been expected to generate.” Bowater, Inc. v. Commissioner, 108 F.3d 12, 13 (2nd Cir. 1997) (quoting Treas.Reg. 1.861-8(e)(2) (as amended in 1978)) (emphasis added). Notably, the language of the 1978 version echoes exactly the language of the 1977 version relevant to this case. Because the language is the same, we find the Second Circuit’s analysis in Bowater illuminating to the case at bar. 13 the use of money to produce income.” Id. Because the revised Regulation provides that “deductions for interest expense be allocated to all income producing activities, with no distinction based on whether the income produced bears the label 'interest' or dividend' or any other appellation,” the Second Circuit reasoned that the newer language of the Regulation, by its explicit terms, disallowed interest netting. Id. Like the taxpayer in the instant case, the taxpayer in Bowater cited this Court's 1990 Dresser decision to argue in favor of interest netting. Id. at 15. The Second Circuit, however, correctly recognized that Dresser “arose under an earlier version of the regulations that was much less clear than the Regulation at issue” in Bowater. Id. at 15 n.6. Because the revision of Treas.Reg. 1.861-8(e) explicitly spoke in terms of “all income producing activities” and “items of gross income,” the Bowater court ultimately concluded that “the plain language of the Regulation [did] not give [the court] the latitude to interpret” the Regulation in light of business models or economic theories. Id. at 16 (questioning this Court's economic premises in Dresser, but holding that the plain language of the revised Regulation nevertheless rendered any disagreement moot). In the instant case, we agree that the unambiguous language of the Regulation provides for gross interest expenses to be ratably allocated to all income producing activities. 14 As a final argument in favor of interest netting, Dresser contends that the IRS took a contrary position and allowed interest netting in two prior Tax Court cases involving Dresser’s 1978 to 1982 taxable years. In both cases, the 1977 version of Treas.Reg. § 1.861-8(e)(2) was in effect; nevertheless, the IRS allowed interest netting. Dresser, in essence, argues that the IRS has waived any position against interest netting. We disagree. First, Dresser fails to demonstrate how the IRS's past allowance of interest netting constitutes a waiver on the part of the Government. Second, and more important, it is well established that the Commissioner may change an earlier interpretation of the law, even if such a change is made retroactive in effect. Dickman v. Commissioner, 465 U.S. 330, 343 (1984) (citing Dixon v. United States, 381 U.S. 68, 72-75 (1965); Automobile Club of Michigan v. Commissioner, 353 U.S. 180, 183-184 (1957)). This rule applies even though a taxpayer may have relied to his detriment upon the Commissioner's prior position. Id. (citing Dixon, 381 U.S. at 73). Additionally, the Commissioner is under no duty to assert a particular position as soon as a relevant statute authorizes such an interpretation. Id. (citing Bob Jones Univ. v. United States, 461 U.S. 574 (1983)). Therefore, the IRS's past allowance of Dresser's interest netting does not compromise the IRS's current position against interest netting, nor does it 15 preclude this Court from reading the plain language of Treas.Reg. 1.861-8(e)(2).6 III. The second issue raised on appeal is whether Dresser's deficiency interest was eliminated when foreign tax credit carrybacks extinguished the deficiencies in the years to which the foreign tax credits were carried. As discussed supra, Dresser was allowed to compute foreign tax credits for taxable years 1981 and 1984 which resulted in an elimination of tax owed for those tax years. Specifically, Dresser was allowed to carry back excess foreign taxes in the amount of $265,109 from its 1983 taxable year to its 1981 taxable year. Dresser was similarly allowed to carry back excess foreign taxes in the amount of $6,261,397 from its 1986 taxable year to its 1984 taxable year. The result of these foreign tax credit carrybacks was the reduction or elimination of deficiencies in Dresser's 1981 and 1984 income tax liabilities. Dresser subsequently sought refunds of the interest it had paid for the 1981 and 1984 deficiencies, insisting that the carrybacks eliminated not only the deficiency, but also the deficiency interest. 6 In light of our holding against interest netting under the 1977 version of Treas.Reg. 1.861-8(e)(2), we find some of Dresser's refund claim for taxable year 1981 moot. Dresser contends that its use of interest netting in 1983 resulted in the availability of excess foreign taxes that it was entitled to carry back to 1981 in the amount of $257,236. Both parties acknowledge in their briefs that this particular claim would be rendered moot if the Court were to hold against interest netting. Because we find that the 1977 language of Treas.Reg. 1.861-8(e)(2) plainly disallows interest netting, we further find Dresser's refund claim based on such a practice without merit. 16 Arguing this position on appeal, Dresser first contends the plain language of Internal Revenue Code § 904(c) clearly and conclusively provides that a foreign tax credit carryback eliminates both the deficiency and the interest on the deficiency. Section 904(c) states that excess foreign taxes shall be deemed taxes paid or accrued to foreign countries in or possessions of the United States in the second preceding taxable year . . . . Such amount deemed paid or accrued in any year may be availed of only as a tax credit and not as a deduction and only if the taxpayer for such year chooses to have the benefits of this subpart . . . . 26 U.S.C. § 904(c) (emphasis added). From this language, Dresser asserts that because its foreign taxes were “deemed paid or accrued” in 1981 and 1984, the result is as if the excess foreign taxes at issue had initially accrued in 1981 and 1984. Dresser contends that the IRS should essentially operate under the fiction that the taxes were indeed actually paid in 1981 and 1984. The result of this counterfactual construction would then be that no interest would exist because no predicate deficiency would have existed. We reject the fiction. Inherent in Dresser's argument is the notion that the “deemed paid” language of § 904(c) speaks not only to what year the credit will be applied, but also to when the reallocation of the tax credit will be deemed to occur. Finding no case law or legislative history to support such a proposition, we 17 hold that Dresser remained liable for interest on deficiencies that were later reduced or eliminated by foreign tax credit carrybacks. The Federal Circuit carefully addressed this precise issue in Fluor Corp. and Affiliates v. United States, 126 F.3d 1397 (Fed. Cir. 1997), reh’g granted, 132 F.3d 700 (Fed. Cir. 1997), cert. denied, 522 U.S. 1118 (1998). As that court explained, The word “deemed” is used interchangeably with the words “treated as” in a closely associated provision of the Code, 26 U.S.C. § 902(a), and the legislative history of section 904(c) confirms that the statute uses the word “deemed' in that sense. . . . . We thus can be confident that when Congress used the word “deemed,” it meant “treated as if,” and that a foreign tax paid in a particular year would be treated as if it were paid in another year. Indeed, that interpretation is necessary to effect the reduction of a deficiency in a carryback year. Id. at 1401. However, the Fluor court noted that “while interpreting the word 'deemed' to mean 'treated as if' answers the question of what year the credit will be applied to, it does not answer the question of when the reallocation of the foreign tax credit will be deemed to occur-–whether in the carryback year or at the time the carryback was generated, one or two years later.” Id. We recognize that if the reduction or elimination of the deficiency is treated as having occurred in the carryback year, “it does not make sense to assess interest on the deficiency.” Id. Conversely, if the reduction or elimination of the deficiency is considered to have occurred at the time the carryback was generated, “it makes sense that interest should be assessed on the 18 deficiency during the time it was outstanding.” Id. The language of § 904(c), however, is unclear as to whether the “deemed paid” language relates to the carryback year or the year in which the carryback arises. The district court therefore correctly appreciated that the text of § 904(c) does not provide “clear legislative expression” regarding whether Congress intended to eliminate deficiency interest on a tax deficiency that is subsequently extinguished by a foreign tax credit carryback. See Dresser, 73 F. Supp.2d at 695- 96. The Federal Circuit in Fluor and the Tax Court in Intel Corp. and Consol. Subsidiaries v. Commissioner, 111 T.C. 90 (1998), have also perceived as much. See Fluor, 126 F.3d at 1401-02 (“We are thus confronted with an ambiguity as to whether Congress meant the language of section 904(c) to forbid the assessment of interest on a previous tax deficiency that is erased as a result of the foreign tax carryback.”); Intel, 111 T.C. at 98 (describing § 904(c) as “ambiguous” and interpreting it “without the benefit of any legislative history directed to this ambiguity”).7 Moreover, 7 But see Fluor Corp. v. United States, 35 Fed. Cl. 520, 526 (1996), rev'd, 126 F.3d 1397 (Fed. Cir. 1997), cert. denied, 522 U.S. 1118 (1998). In Fluor, the Federal Circuit reversed the Court of Federal Claims, which had held that no ambiguity in § 904(c) exists. Rather, the Court of Federal Claims found that [t]he intent of Congress can be discerned from the plain language of the statute. Congress provided in § 904(c) that they carryback of the credit for excess foreign tax paid “shall be deemed taxes paid or accrued” in the earlier year. Under § 904(c), therefore, Fluor's tax obligation for 1982 was changed; it was reduced by the foreign tax carried back which was “deemed” paid in 1982. Fluor Corp. v. United States, 35 Fed. Cl. at 526 (1996). 19 nothing in the language of § 904(c) or its legislative history addresses this ambiguity. See Fluor, 126 F.3d at 1402; Intel, 111 T.C. at 98. Rather, § 904(c) conclusively answers only the question of to what year will the tax credit be applied. Fluor, 126 F.3d at 1401. Because § 904(c) does not answer the issue posed in this appeal, we look to 26 U.S.C. § 6601(a), the general provision for deficiency interest in the Tax Code. Section 6601(a) provides that when an underpayment in tax for a particular year is reduced or eliminated as a result of a carryback from a later year, the taxpayer remains liable for interest on its underpayment from the time the tax in question was due until the date the tax was satisfied by application of the carryback. 26 U.S.C. § 6601(a); see Fluor, 126 F.3d at 1402. The principle undergirding this rule is that the government is deprived of the money for the period between the original time the tax deficiency should have been paid and the time the deficiency was abated by the foreign tax credit. Manning v. Seeley Tube & Box Co., 338 U.S. 561, 566 (1950); Fluor, 126 F.3d at 1402; see also In re Rush-Hampton Indus., Inc., 98 F.3d 614, 616 (11th Cir. 1996). In the absence of an explicit countermand to § 6601(a)'s general rule, the statute “compels the conclusion that the government is entitled to interest for that interim period.” Fluor, 126 F.3d at 1402. Without clear language in § 904(c) to provide an exception to the § 6601(a) standard, we 20 find that § 6601(a) operates in the instant case to impose interest liability on Dresser's 1981 and 1984 deficiencies, even though the deficiencies were later reduced or eliminated by foreign tax credit carrybacks. Two closely analogous Supreme Court decisions, cited by the Government, support this conclusion. In the first, Manning v. Seeley Tube & Box Co., 338 U.S. 561 (1950), the government assessed a deficiency in the taxpayer's 1941 taxes, with interest from the date the taxes were due. The taxpayer subsequently filed a return in 1943, showing a net operating loss for that year. The net operating loss carryback, when applied to the taxpayer's 1941 taxes, was sufficient to eliminate its tax liability for that year. The Commissioner abated the deficiency, but the question then arose as to whether the taxpayer was entitled to refund of the interest that it had paid on the deficiency. Id. at 563-65. The Supreme Court held that the taxpayer was not relieved of its liability to pay interest on the 1941 deficiency despite the net operating loss carryback because “subsequent cancellation of the duty to pay [the] assessed deficiency does not cancel in like manner the duty to pay the interest on that deficiency.” Id. at 565. In coming to its conclusion, the Court reasoned that the taxpayer, by its failure to pay the taxes owed, had the use of funds which rightfully should have been in the possession of the United States. The fact that the statute permits the taxpayer subsequently to avoid the payment of that debt in no way indicates that the 21 taxpayer is to derive the benefits of the funds for the intervening period. Id. at 566. Rather, as the Court explained, the general proposition is that “Congress intended the United Stats to have the use of money due when it became due.” Id. Accordingly, when the Commissioner assesses a deficiency, “he also may assess interest on that deficiency from the date the tax was due to the assessment date.” Id. The Supreme Court in Seeley Tube saw nothing in the net operating loss carryback statute that altered this fundamental principle. Id. Moreover, because the Tax Code prohibits a taxpayer who pays a tax that is later abated by a carryback from claiming interest from the government in the intervening period, the Supreme Court inferred that Congress “did not intend to change the basic statutory policy: the United States is to have the possession and use of the lawful tax at the date it is properly due.” Id. at 568. The Court thus concluded that, “[i]n the absence of a clear legislative expression to the contrary, the question of who properly should possess the right of use of the money owed to the Government for the period it is owed must be answered in favor of the Government.” Id. at 566. In holding against the taxpayer's claim for interest, the Supreme Court also identified a policy justification against cancellation of interest on a deficiency subsequently abated. To allow a taxpayer to recoup interest paid on a later-extinguished deficiency, the Court reasoned, “would be to place a premium on 22 failure to conform diligently with the law.” Id. The undesirable result would thus be a delinquent taxpayer's receiving “ the full use of the tax funds for the intervening period,” id., while a diligent taxpayer's being “statutorily prohibited from having the use of the money for that period,” Fluor, 126 F.3d at 1400. As the Court concluded, “We cannot approve such a result.” Seeley Tube, 338 U.S. at 568. The Supreme Court issued a similar decision in the case of United States v. Koppers, Inc., 348 U.S. 254 (1955), which involved interest on a deficiency with respect to excess profits taxes. The Koppers Court held that an abatement of federal excess profits taxes that eliminated a taxpayer’s deficiency in that tax did not relieve the taxpayer from having to pay interest on the deficiency for the period between the tax's being due and the tax's being abated. Koppers, 348 U.S. at 269. In concluding that the taxpayers still had duties to pay interest on tax obligations subsequently abated, the Court held that to extinguish the interest obligation would be to “sustain the proposition that the tax relief granted under [the adjustment provision] is necessarily retroactive, extinguishing the deficiency as of the original due date of the tax and thus eliminating the interest charges for the corresponding period.” Id. at 263. Because no explicit provision evinced any intent by Congress to eliminate the interest charges on the former deficiency, the Court held that the taxpayer was liable 23 for the interest. Id. at 269. Finally, as in Seeley Tube, the Supreme Court in Koppers observed that there was “nothing to justify a greater tax advantage to any taxpayer that underpays its correct tax, over one that pays such tax in full when due.” Koppers, 348 U.S. at 262. In the instant case, the Government argues that the reasoning underlying Seeley Tube and Koppers bars Dresser from obtaining a refund of interest on its deficiencies that were later reduced or eliminated by the carryback of excess foreign taxes. Likewise, the Federal Circuit in Fluor found “powerful support” in Seeley Tube and Koppers when it held that a taxpayer remained liable for interest on a deficiency that was later eliminated by the carryback of excess foreign taxes. Fluor, 126 F.3d at 1400. We agree that Seeley Tube and Koppers inform our examination of interest liability in the context of foreign tax credit carrybacks. As the Fluor court recognized, foreign tax carrybacks operate in essentially the same manner as the net operating loss carrybacks of Seeley Tube or the adjustments to excess profits taxes of Koppers, because all three reduce or eliminate a tax deficiency in a previous year. See id. at 1399-1400. Moreover, as with the taxpayers in Seeley Tube, Koppers, and Fluor, there is no dispute that Dresser in the instant case is subject to the general provisions of § 6601(d). While no statute, including § 904(c), specifically addresses the issue of interest liability with respect 24 to a foreign tax credit carryback, § 6601(d) mandates that a taxpayer must pay interest on any deficiency from the time the deficiency arises until it is paid or otherwise abated. See id. at 1400 (referring to 26 U.S.C. § 6601(a), discussed supra). Thus, following the Supreme Court in Seeley Tube and the Federal Circuit in Fluor, we hold that any departure from that general rule “would require 'a clear legislative expression to the contrary.'” Id. (quoting Seeley Tube, 338 U.S. at 566); see Intel, 111 T.C. at 98- 100 (providing similar reasoning in holding that foreign tax credit carrybacks do not eliminate a taxpayer's interest liability). Because, as discussed above, such clear legislative expression is absent in this case, we agree with the Fluor court that the “question of who should possess the right of use of the money owed the Government for the period it is owed must be answered in favor of the Government.” Fluor, 126 F.3d at 1401 (quoting Seeley Tube, 338 U.S. at 566). Dresser nevertheless argues that the general rule of § 6601(d) is inapplicable to foreign tax credit carrybacks because the 1939 version of § 6601(d), under which the Supreme Court decided Seeley Tube and Koppers, included net operating loss carrybacks, but not foreign tax credit carrybacks. To Dresser, that difference defines the instant case. We disagree. The version of § 6601(d) relevant to this case authorizes the government to collect deficiency interest from 25 taxpayers whose deficiencies are eliminated by net operating loss carrybacks, capital loss carrybacks, and credit carrybacks. 26 U.S.C. § 6601(d)(1)-(2). While the statute does not mention foreign tax credit carrybacks, we disagree with Dresser's conclusion that such absence constitutes a “clear legislative expression” by Congress to carve out foreign tax credit carrybacks as an exception to the general rule of § 6601(d). Cf. Seeley Tube, 338 U.S. at 566 (requiring a “clear legislative expression” before finding that a taxpayer is entitled to money owed to the government for the period it is owed); Fluor, 126 F.3d at 1400 (citing Seeley Tube and holding same).8 Moreover, the legislative history of § 6601(d) indicates that Congress did not intend to prohibit the assessment of interest on deficiencies later eliminated by foreign tax credit carrybacks. As the Fluor court explained: When the foreign tax carryover was enacted in 1958, section 6601(d) (which was then section 6601(e)) addressed only one form of carryback--the net operating loss carryback. It was not until later that Congress converted section 6601(d) into a catchall provision imposing deficiency interest in the case of all subsequently enacted carryback statutes. Because no such 8 Subsequent to the tax years at issue, Congress, in the Taxpayer Relief Act of 1997, modified § 6601(d) to provide explicitly that interest must be paid even if the deficiency is eliminated by a foreign tax credit carryback. According to the Intel case, the legislative history behind the change in the statute makes it clear it was intended to overrule the decision of the Court of Federal Claims in Fluor Corp. & Affiliates v. United States, 35 Fed. Cl. 520 (1996), which allowed the foreign tax carryback to reduce an underpayment for purposes of computing interest, and that Congress believed that the rule should be the same for both underpayments and overpayments. Intel Corp. and Consol. Subsidiaries v. Commissioner, 111 T.C. 90, 101-04 (1998) (citing H. Conf. Rept. 105-220, 575-576 (1997); S. Rept. 105-33; 178-179 (1997); H. Rept. 105-148, 551-552 (1997)). 26 catchall provision existed in 1958, Congress's failure to alter the single-purpose section 6601(e) at that time to include a reference to foreign tax carrybacks does not compel the conclusion that Congress intended to prohibit the collection of deficiency interest in the case of foreign tax carrybacks. Fluor, 126 F.3d at 1404. We also appreciate, as did the district court, that Seeley Tube and Koppers had recently been decided when the foreign tax credit was created. Dresser, 73 F. Supp.2d at 696 (citing Fluor, 126 F.3d at 1404). It was thus “reasonable for Congress to assume that those cases would apply to analogous carryback provisions,” id., especially because the rule of Seeley Tube and Koppers “did not depend on specific legislation imposing deficiency interest,” Fluor, 126 F.3d at 1404. Because it was reasonable for Congress to make that assumption, no special legislation was needed to ensure that the principles of Seeley Tube and Koppers regarding net operating loss carrybacks (or adjustments due to excess profits taxes) would apply similarly to foreign tax credit carrybacks.9 Id. As a final argument on this issue, Dresser contends that the IRS took the position for approximately thirty-five years that deficiency interest under Internal Revenue Code § 6601(a) was not payable to the extent that income tax was eliminated by foreign 9 Like the Fluor court, we recognize a fundamental principle of statutory construction is “that Congress is presumed to be aware of judicial interpretations of the law, and that when Congress enacts a new statute incorporating provisions similar to those in prior law, it is assumed to have acted with awareness of judicial interpretations of prior law.” Fluor, 126 F.3d at 1404; see Merrill, Lynch, Pierce, Fenner & Smith, Inc. v. Curran, 456 U.S. 353, 382 n.66 (1982). 27 taxes which were deemed paid or accrued by the taxpayer pursuant to § 904(c). This stance, according to Dresser, is reflected by the fact that Treas.Reg. 301.6601-1 fails to provide for such interest.10 Dresser moreover notes that the Internal Revenue Manual was not amended to conform to the current IRS position with respect to deficiency interests vis-à-vis foreign tax credit carrybacks until December 1992. See I.R.M. § 8712.3. Dresser therefore concludes that “[t]he fact that IRS agents have consistently failed to collect deficiency interest [related to foreign tax credit carrybacks] for over 35 years supports [the] position that the law did not require that deficiency interest be charged . . . .” To Dresser, thirty-five years of “silence” by the IRS on this issue is, “in effect, a tacit agreement with [Dresser's] particular position.” We disagree with Dresser's inference and refuse to hold that the IRS's previous failure to impose interest on deficiencies eliminated by foreign tax credit carrybacks necessarily precludes it from imposing such interest now. “This is not a case . . . in which the Service is attempting to reverse a position it has long taken in construing a statute.” Fluor, 126 F.3d at 1405. Rather, 10 Treas.Reg. 301.6601-1 provides: The carryback of a net operating loss, net capital loss, investment credit, or work incentive program (WIN) credit shall not affect the computation of interest on any income tax for the period commencing with the last day prescribed for the payment of such tax and ending with the last day of the taxable year in which the loss or credit arises. Treas.Reg. 301.6601-1 (as amended in 1983). 28 the district court in this case observed that no practice regarding deficiency interest related to foreign tax credits was followed uniformly, and Dresser has presented no evidence showing how interest was imputed in other cases. See Dresser, 73 F. Supp.2d at 696. Similar to the Fluor court, we “are unwilling to treat as established administrative practice what amounts to no more than a failure to advert to the issue at a policymaking level.” Fluor, 126 F.3d at 1405.11 In sum, we find that the inconclusive language of § 904(c) fails to establish an exception to the mandate of § 6601(d) in the clear terms that the Supreme Court contemplated when it stated its rule in Seeley Tube and Koppers. We agree with the district court that the “deemed paid” language of § 904(c) can only be inferred to “relate[] . . . to the year in which the foreign tax credit will be applied.” Dresser, 73 F. Supp.2d at 695; see Fluor Corp., 126 F.3d at 1401. Accordingly, elimination or reduction of Dresser's deficiencies for 1981 and 1984 taxable years does not eliminate Dresser's interest on those deficiencies. IV. 11 Even assuming arguendo that the IRS's failure to impose deficiency interest related to foreign tax credit carrybacks constituted an administrative practice, the IRS is not precluded from departing from a prior administrative practice. See Dickman v. Commissioner, 465 U.S. 330, 343 (1984) (stating that the IRS Commissioner was “under no duty to assert a particular position as soon as the statute authorizes such interpretation” even though “a taxpayer may have relied to his detriment upon the Commissioner's prior position”). 29 Having decided that Dresser is liable for interest on deficiencies later eliminated by the carryback of excess foreign taxes from its 1983 and 1986 taxable years, we now address the issue of when the interest on those deficiencies ceases to accrue. The district court, following the rule enunciated by the Tax Court in Intel, held that “deficiency interest accrues until the filing date of the return for the tax year in which the foreign tax credit arises.” Dresser, 73 F. Supp.2d at 697; see Intel, 111 T.C. at 101-04; Hallmark Cards, Inc. v. Commissioner, 111 T.C. 266, 272 (1998). Dresser, however, appeals the district court's holding and instead argues that deficiency interest accrues only until the end of the taxable year in which the carryback was generated. Dresser's position is consistent with the Federal Circuit's decision in Fluor. See Fluor, 126 F.3d at 1406. In Fluor, the Federal Circuit held that deficiency interest should be calculated only through the end of the taxable year when the carryback arises because such was the language of § 6601(e) (as it related to net operating loss carrybacks) when the foreign tax credit was enacted in 1958. Id. (citing 26 U.S.C. § 6601(e) (1958)). The Fluor court presumed that “Congress . . . would have selected the same date for ending the accrual of deficiency interest with respect to foreign tax carrybacks if it had expressly addressed that subject in the 1958 legislation.” Id. Although the court in Fluor acknowledged that Congress amended the timing rules in 1982 (changing the 30 accrual date for carrybacks covered by § 6601(d) to the filing date for the taxable year in which the credit arises), the court nonetheless declined to treat the legislative change as affecting foreign tax credit carrybacks as well. Id. (“[W]e cannot attribute to Congress the intention to have the foreign tax carryover timing rules follow the 1982 legislative change in the rules applicable to other carryovers.”).12 We agree with the district court that the Fluor court's reasoning on the matter is “perplexing,” Dresser, 73 F. Supp.2d at 697, and we note that such reasoning has been criticized by the Tax Court in recent decisions. See Intel, 111 T.C. at 102-04; Hallmark Cards, 111 T.C. at 272 (1998). In 1982, when Congress changed the accrual period for underpayments with respect to tax credits specifically covered under § 6601(d), it also made similar changes to the accrual period for interest on overpayments, including overpayments generated by the allowance of foreign tax credits. See 26 U.S.C. § 6611(f)(1) & (2). In the Tax Equity and Fiscal Responsibility Act of 1982, Pub.L. 97-248, § 346(c), 96 Stat. 637, Congress changed the effective dates of carryback credits in all of the carryback interest provisions for both overpayments and deficiencies from the last day of the taxable year in which the credit arose to the due 12 The amendments promulgated by Congress were part of the Tax Equity and Fiscal Responsibility Act of 1982, Pub.L. No. 97-248, § 346(c), 96 Stat. 324 (1982). This change was effective for interest accruing after October 3, 1982. 31 date for filing the return for that year. The rule that the Fluor court effectively pronounced, and the argument that Dresser now advances, is that “Congress intended for the interest accrual rules to be the same with respect to overpayments and underpayments for every type of credit except foreign tax credits.” Dresser, 73 F. Supp.2d at 697. Such a proposition is illogical and inconsistent with the Internal Revenue Code's consistent policy of symmetrical treatment with respect to the period during which interest accrues on both underpayments of tax that are eliminated by carrybacks and overpayments of tax resulting from carrybacks. See Intel, 111 T.C. at 103. Thus, as the Tax Court opined in Intel, it would be contrary to well-ingrained tax policy, and indeed an “eccentric action by Congress,” for interest attributable to a deficiency that is reduced or eliminated by the carryback of a foreign tax credit to accrue for a different period than interest on an overpayment resulting from the same foreign tax carryback. Id. at 104. Indeed, Congress's amendments to § 6601(d) in 1997 indicate the intent to maintain symmetry between the interest accrual rules. See § 6601(d)(2) (specifying that interest related to carryback of a foreign tax credit continues to accrue until the filing date of the year in which the credit arises). Consequently, we hold that Dresser's deficiency interest accrued until the filing date of the returns for the tax years in which the foreign tax credits arose. 32 V. For the reasons assigned, the district court's judgment for the Government is AFFIRMED. 33