Part I. Rulings — Internal Revenue Bulletin 1996-34

tax years ended on, or with reference to, June 30, 1996.

The Department Store Inventory Price Indexes are prepared on a national basis and include (a) 23 major groups of departments, (b) three special combinations of the major groups—soft goods, durable goods, and miscellaneous goods, and (c) a store total, which covers all departments, including some not listed separately, except for the following: candy, foods, liquor, tobacco, and contract departments.

Section 472.—Last-in, First-out Inventories

26 CFR 1.472–1: Last-in, first-out inventories.

LIFO; price indexes; department stores. The June 1996 Bureau of Labor Statistics price indexes are accepted for use by department stores employing the retail inventory and last-in, first-out inventory methods for valuing inventories for tax years ended on, or with reference to, June 30, 1996.

Rev. Rul. 96–39

The following Department Store Inventory Price Indexes for June 1996 were issued by the Bureau of Labor Statistics on July 16, 1996. The indexes are accepted by the Internal Revenue Service, under § 1.472–1(k) of the Income Tax Regulations and Rev. Proc. 86–46, 1986–2 C.B. 739, for appropriate application to inventories of department stores employing the retail inventory and last-in, first-out inventory methods for

BUREAU OF LABOR STATISTICS, DEPARTMENT STORE INVENTORY PRICE INDEXES BY DEPARTMENT GROUPS

(January 1941 = 100, unless otherwise noted)

Percent Change from

June 1996 ¹

Groups

June June 1995 1996

June 1995 to

Piece Goods. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 522.9 551.1 5.4
Domestics and Draperies . . . . . . . . . . . . . . . . . . . . . . . . 646.7 641.0 �0.9
Women’s and Children’s Shoes . . . . . . . . . . . . . . . . . . . 624.2 649.3 4.0
Men’s Shoes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 919.3 895.4 �2.6
Infants’ Wear . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 587.9 627.1 6.7
Women’s Underwear . . . . . . . . . . . . . . . . . . . . . . . . . . . . 515.1 535.4 3.9
Women’s Hosiery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283.3 288.0 1.7
Women’s and Girls’ Accessories . . . . . . . . . . . . . . . . . . 549.5 545.5 �0.7
Women’s Outerwear and Girls’ Wear. . . . . . . . . . . . . . . 416.3 401.1 �3.7
Men’s Clothing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 597.5 612.2 2.5
Men’s Furnishings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 562.4 584.5 3.9
Boys’ Clothing and Furnishings . . . . . . . . . . . . . . . . . . . 477.8 485.7 1.7
Jewelry. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1004.9 1011.5 0.7
Notions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 758.7 774.1 2.0
Toilet Articles and Drugs . . . . . . . . . . . . . . . . . . . . . . . . 859.9 877.8 2.1
Furniture and Bedding . . . . . . . . . . . . . . . . . . . . . . . . . . 663.1 673.6 1.6
Floor Coverings. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 577.0 576.4 �0.1
Housewares . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 771.8 808.7 4.8
Major Appliances . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247.2 245.5 �0.7
Radio and Television. . . . . . . . . . . . . . . . . . . . . . . . . . . . 82.1 79.3 �3.4
Recreation and Education ² . . . . . . . . . . . . . . . . . . . . . . . 114.0 112.8 �1.1
Home Improvements ² . . . . . . . . . . . . . . . . . . . . . . . . . . . 122.6 127.4 3.9
Auto Accessories ² . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106.8 107.5 0.7

Groups 1–15: Soft Goods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 587.8 592.4 0.8

Groups 16–20: Durable Goods . . . . . . . . . . . . . . . . . . . . . . . . . . 462.8 469.7 1.5

Groups 21–23: Misc. Goods ² . . . . . . . . . . . . . . . . . . . . . . . . . . . 113.9 113.7 �0.2

Store Total ³ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 545.8 550.3 0.8

1Absence of a minus sign before percentage change in this column signifies price increase. 2Indexes on a January 1986=100 base. 3The store total index covers all departments, including some not listed separately, except for the following: candy, foods, liquor, tobacco, and contract departments.

DRAFTING INFORMATION

The principal author of this revenue ruling is Stan Michaels of the Office of

Assistant Chief Counsel (Income Tax and Accounting). For further information regarding this revenue ruling, contact

Mr. Michaels on (202) 622–4970 (not a toll-free call).

Business Machines Corporation (IBM) ships products that it manufactures in the United States to numerous foreign subsidiaries and insures those shipments against loss. When the foreign subsidiary makes the shipping arrangements, the subsidiary often places the insurance with a foreign carrier. When it does, both IBM and the subsidiary are listed as beneficiaries in the policy.

IBM filed federal excise tax returns for the years 1975 through 1984, but reported no liability under § 4371. The IRS audited IBM and determined that the premiums paid to foreign insurers were taxable under § 4371 and that IBM—as a named beneficiary of the insurance policies—was liable for the tax. The IRS assessed a tax against IBM for each of those years.

IBM paid the assessments and filed refund claims, which the IRS denied. IBM then commenced suit in the Court of Federal Claims, contending that application of § 4371 to policies insuring its export shipments violated the Export Clause. The focus of the suit was this Court’s decision in Thames & Mersey Marine Ins. Co. v. United States, 237 U.S. 19 (1915), in which we held that a federal stamp tax on policies insuring marine risks could not, under the Export Clause, be constitutionally applied to policies covering export shipments. The United States argued that the analysis of Thames & Mersey is no longer valid, having been superseded by subsequent decisions interpreting the Import-Export Clause—specifically, Michelin Tire Corp. v. Wages, 423 U. S. 276 (1976), and Department of Revenue of Wash. v. Association of Wash. Stevedoring Cos., 435 U. S. 734 (1978). The Court of Federal Claims noted that this Court has never overruled Thames & Mersey and ruled that application of § 4371 to policies insuring goods in export transit violates the Export Clause. 31 Fed. Cl. 500 (1994). The Court of Appeals for the Federal Circuit affirmed. 59 F. 3d 1234 (1995). We agreed to hear this case to decide whether we should overrule Thames & Mersey . 516 U.S. __ (1995).

The Export Clause states simply and directly: ‘‘No Tax or Duty shall be laid on Articles exported from any State.’’ U.S. Const., Art. I, § 9, cl. 5. We have had few occasions to interpret the lan District of Columbia, within which such insurer is authorized to do business.’’ 26 U. S. C. § 4373(1) (1982 ed.).

Section 4371.—Imposition of Tax

Ct.D. 2060

SUPREME COURT OF THE UNITED

STATES

No. 95–591

UNITED STATES, PETITIONER v.

INTERNATIONAL BUSINESS

MACHINES CORPORATION

[517 U.S.—]

ON WRIT OF CERTIORARI TO THE

UNITED STATES COURT OF APPEALS FOR THE FEDERAL

CIRCUIT

June 10, 1996

Syllabus

Pursuant to § 4371 of the Internal Revenue Code, respondent International Business Machines Corporation (IBM) paid a tax on insurance premiums remitted to foreign insurers to cover shipments of goods to its foreign subsidiaries. When its refund claims were denied, IBM filed suit in the Court of Federal Claims, contending that § 4371’s application to policies insuring export shipments violated the Export Clause, which states that ‘‘[n]o Tax or Duty shall be laid on Articles exported from any State.’’ The court agreed, rejecting the Government’s argument that Thames & Mersey Marine Ins. Co. v. United States, 237 U. S. 19—in which this Court held that a federal stamp tax on policies insuring marine risks could not, under the Export Clause, be constitutionally applied to policies covering export shipments— had been superseded by subsequent decisions interpreting the Import-Export Clause, which states in relevant part, ‘‘No State shall . . . lay any Imposts or Duties on Imports or Exports.’’ The Court of Appeals affirmed.

Held: The Export Clause prohibits assessment of nondiscriminatory federal taxes on goods in export transit.

(a) While this Court has strictly enforced the Export Clause’s prohibition against federal taxation of goods in export transit and certain closely related services and activities, see, e.g., Thames & Mersey, supra, it has not exempted pre-export goods and services from ordinary tax burdens or exempted from federal taxation various services and activities only tangentially related to the export process, see, e.g., Cornell v. Coyne, 192 U.S. 418. Conceding that the tax assessed here violates the Export Clause under Thames & Mersey, the Government asks that the case be overruled because its underlying theory has been rejected in the context of the Commerce and Import-Export Clauses and those Clauses have historically been interpreted in harmony with the Export Clause.

(b) When this Court expressly disavowed its early view that the dormant Commerce Clause required a strict ban on state taxation of interstate commerce, Complete Auto Transit, Inc. v. Brady, 430 U.S. 274, 288–289, it resolved a long struggle over the meaning of the nontextual negative command of that Clause. The Export Clause, on the other hand, expressly prohibits Congress from laying any tax or duty on exports. These textual disparities strongly suggest that shifts in the

Court’s view of the dormant Commerce Clause’s scope cannot govern Export Clause interpretation. Cf. Richfield Oil Corp. v. State Bd. of Equaliza- tion, 329 U. S. 69, 75–76.

(c) While one may question Thames & Mersey ’s finding that a tax on policies insuring exports is functionally the same as a tax on exportation itself, the Government apparently has chosen not to do so here. Under the principles that animate the policy of stare decisis, the Court declines to overrule Thames & Mersey ’s long-standing precedent, which has caused no uncertainty in commercial export transactions, on a theory not argued by the parties.

(d) This Court’s recent Import-Export Clause cases do not require that Thames & Mersey be overruled. Meaningful textual differences that should not be overlooked exist between the Export Clause and the Import-Export Clause. In finding the assessments in Michelin Tire Corp. v. Wages, 423 U.S. 276, and Department of Revenue of Wash. v. Association of Wash. Stevedoring Cos., 435 U. S. 734, valid, the Court recognized that the Import-Export Clause’s absolute ban on ‘‘Imposts or Duties’’ is not a ban on every tax. Because impost and duty are thus narrower terms than tax, a particular state assessment might be beyond the Import-Export Clause’s reach, while an identical federal assessment might be subject to the Export Clause. The word ‘‘Tax’’ has a common, and usually expansive, meaning that should not be ignored. The Clauses were also intended to serve different goals. The Government’s policy argument—that the Framers intended the Export Clause to narrowly alleviate the fear of northern repression through taxation of southern exports by prohibiting only discriminatory taxes—cannot be squared with the Clause’s broad language. The better reading is that the Framers sought to alleviate their concerns by completely denying to Congress the power to tax exports at all. See Fairbank v. United States, 181 U. S. 283.

(e) Even assuming that Michelin and Washing- ton Stevedoring govern the Export Clause inquiry here, those holdings do not interpret the ImportExport Clause to permit assessment of nondiscriminatory taxes on imports and exports in transit. 59 F. 3d 1234, affirmed. THOMAS, J., delivered the opinion of the Court, in which REHNQUIST, C. J., and O’CONNOR, SCALIA, SOUTER, and BREYER, JJ., joined. KENNEDY, J., filed a dissenting opinion, in which GINSBURG, J., joined. STEVENS, J., took no part in the consideration or decision of the case.

JUSTICE THOMAS delivered the opinion of the Court.

We resolve in this case whether the Export Clause of the Constitution permits the imposition of a generally applicable, nondiscriminatory federal tax on goods in export transit. We hold that it does not.

Section 4371 of the Internal Revenue Code imposes a tax on insurance premiums paid to foreign insurers that are not subject to the federal income tax. ¹ 26 U. S. C. § 4371 (1982 ed.). International

1 The tax does not apply if a policy issued by a foreign insurer is ‘‘signed or countersigned by an officer or agent of the insurer in a State, or in the

guage of the Export Clause, but our cases have broadly exempted from federal taxation not only export goods, but also services and activities closely related to the export process. At the same time, we have attempted to limit the term ‘‘Articles exported’’ to permit federal taxation of pre-export goods and services.

Our early cases upheld federal assessments on the manufacture of particular products ultimately intended for export by finding that pre-export products are not ‘‘Articles exported.’’ See Pace v. Burgess, 92 U. S. 372 (1876); Turpin v. Burgess, 117 U. S. 504 (1886); Cornell v. Coyne, 192 U. S. 418 (1904). Pace and Turpin both involved a federal excise tax on tobacco products. In Pace, though tobacco intended for export was exempted from the tax, the exemption itself was subject to a per-package stamp charge of 25 cents. When a tobacco manufacturer challenged the stamp charge, we upheld the charge on the basis that the stamps were designed to prevent fraud in the export exemption from the excise tax and did not, therefore, represent a tax on exports. 92 U.S., at 375. When Congress later repealed the 25-cent charge for the exemption stamp in a statute that referred to the stamp as an ‘‘export tax,’’ another manufacturer sued to recover the money it had paid for the exemption stamps. See Turpin, supra . Without disturbing the prior ruling in Pace that the stamp charge was not a tax on exports, 117 U.S., at 505, we explained that the prohibition of the Export Clause ‘‘has reference to the imposition of duties on goods by reason or because of their exportation or intended exportation, or whilst they are being exported,’’ id., at 507. We said that the plaintiffs would have had no Export Clause claim even if there had been no exemption from the excise because the goods were not in the course of exportation and might never be exported. Ibid. Turpin broadly suggested that the Export Clause prohibits both taxes levied on goods in the course of exportation and taxes directed specifically at exports.

In Cornell, the Court addressed whether the Export Clause prohibited application of a federal excise tax on filled cheese manufactured under contract for export. Looking to the analysis set out in Turpin, we rejected the contention that the Export Clause bars application of a nondiscriminatory tax imposed before the product entered the course of exportation. ‘‘The true con

struction of the constitutional provision is that no burden by way of tax or duty can be cast upon the exportation of articles, and does not mean that articles exported are relieved from the prior ordinary burdens of taxation which rest upon all property similarly situated.’’ Cornell, supra, at 427. Pace, Turpin, and Cornell made clear that nondiscriminatory pre-exportation assessments do not violate the Export Clause, even if the goods are eventually exported.

At the same time we were defining a domain within which nondiscriminatory taxes could permissibly be imposed on goods intended for export, we were also making clear that the Export Clause strictly prohibits any tax or duty, discriminatory or not, that falls on exports during the course of exportation. See Fairbank v. United States, 181 U. S. 283 (1901); United States v. Hvoslef, 237 U. S. 1 (1915); Thames & Mersey Marine Ins. Co. v. United States, supra. In Fairbank, for example, we addressed a federal stamp tax on bills of lading for export shipments imposed by the War Revenue Act of 1898. The Court found that the tax was facially discriminatory, Fairbank, supra, at 290, and, though not directly imposed on the goods being exported, the tax was nevertheless ‘‘in effect a duty on the article transported,’’ 181 U. S., at 294. Consequently, the tax fell directly into the category of forbidden taxes on exports defined in Turpin . In striking down the tax, we said:

‘‘The requirement of the Constitution is that exports should be free from any governmental burden. The language is ‘no tax or duty.’ Whether such provision is or is not wise is a question of policy with which the courts have nothing to do. We know historically that it was one of the compromises which entered into and made possible the adoption of the Constitution. It is a restriction on the power of Congress . . . .’’ 181 U. S., at 290. Hvoslef and Thames & Mersey differed from Fairbank in that the taxes imposed in those cases—on ship charters and marine insurance, respectively—did not facially discriminate against exports. The Court nonetheless prohibited the application of those generally applicable, nondiscriminatory taxes to the transactions at issue because each tax was, in effect, a tax on exports. The type of charter contract at issue in Hvoslef was ‘‘in contemplation of law a mere contract of affreightment,’’ 237

U.S., at 16, and we found that the tax, as applied to charters for exportation, ‘‘was in substance a tax on the exportation; and a tax on the exportation is a tax on the exports,’’ id., at 17. Likewise, in Thames & Mersey, we found that ‘‘proper insurance during the voyage is one of the necessities of exportation’’ and that ‘‘the taxation of policies insuring cargoes during their transit to foreign ports is as much a burden on exporting as if it were laid on the charter parties, the bills of lading, or the goods themselves.’’ 237 U. S., at 27.

Shortly after Hvoslef and Thames & Mersey, the Court rejected an attempt to shield from taxation the net income of a company engaged in the export business. William E. Peck & Co. v. Lowe, 247 U. S. 165 (1918). In accordance with the analysis set out in Turpin, we found both that the tax was nondiscriminatory and that ‘‘[i]t is not laid on articles in course of exportation or on anything which inherently or by the usages of commerce is embraced in exportation or any of its processes.’’ 247 U. S., at 174.

Only a few years later the Court struck down the application of a tax on the export sale of certain baseball equipment. See A. G. Spalding & Bros. v. Edwards, 262 U. S. 66 (1923). Although the tax was clearly nondiscriminatory, we explained that the goods being taxed had entered the course of exportation when they were delivered to the export carrier. Id., at 70. Because the taxable event, the transfer of title, occurred at the same moment the goods entered the course of exportation, we held that the tax could not constitutionally be applied to the export sale. Id., at 69–70.

The Court has strictly enforced the Export Clause’s prohibition against federal taxation of goods in export transit, and we have extended that protection to certain services and activities closely related to the export process. We have not, however, exempted pre-export goods and services from ordinary tax burdens; nor have we exempted from federal taxation various services and activities only tangentially related to the export process.

III

The Government concedes, as it did below, that this case is largely indistinguishable from Thames & Mersey and that, if Thames & Mersey is still good law, the tax assessed against IBM under § 4371 violates the Export Clause. See

Tr. of Oral Arg. 5; 59 F. 3d, at 1237. The parties apparently agree that there is no legally significant distinction between the insurance policies at issue in this case and those at issue in Thames & Mersey, and, accordingly, the Government asks that we overrule Thames & Mersey.

The Government asserts that the Export Clause permits the imposition of generally applicable, nondiscriminatory taxes, even on goods in export transit. The Government urges that we have historically interpreted the Commerce, Import-Export, and Export Clauses in harmony and that we have rejected the theory underlying Thames & Mersey in the context of the Commerce and Import-Export Clauses. Accordingly, the Government contends that our Export Clause jurisprudence, symbolized by Thames & Mersey, has become an anachronism in need of modernization. The Government asks us to reinterpret the Export Clause to permit the imposition of generally applicable, nondiscriminatory taxes as we have under the Commerce Clause and, it argues, under the Import-Export Clause.

The Government contends that our dormant Commerce Clause jurisprudence has shifted dramatically and that our traditional understanding of the Export Clause, which is based partly on an outmoded view of the Commerce Clause, can no longer be justified. It is true that some of our early Export Clause cases relied on an interpretation of the Commerce Clause that we have since rejected. In Fairbank, 181 U. S., at 298–300, for example, we analogized to Robbins v. Shelby County Taxing Dist., 120 U. S. 489, 497 (1887), in which we held that ‘‘[i]nterstate commerce cannot be taxed at all [by the States], even though the same amount of tax should be laid on domestic commerce, or that which is carried on solely within the state.’’ Referring to the categorical ban on taxation of interstate commerce declared in Robbins, we likened the scope of the Commerce Clause’s ban on state taxation of interstate commerce to the Export Clause’s ban on federal taxation of exports. Fairbank, supra, at 300; see also Hvoslef, 237 U. S., at 15 (‘‘The court

[in Fairbank ] found an analogy in the construction which had been given to the commerce clause in protecting interstate commerce from state legislation

imposing direct burdens’’). After Thames & Mersey, the Commerce Clause construction espoused in Robbins fell out of favor, see Western Live Stock v. Bureau of Revenue, 303 U. S. 250, 254 (1938) (‘‘It was not the purpose of the commerce clause to relieve those engaged in interstate commerce from their just share of state tax burden even though it increases the cost of doing the business’’), and we expressly disavowed that view in Complete Auto Transit, Inc. v. Brady, 430 U. S. 274, 288—289 (1977).

Our rejection in Complete Auto of much of our early dormant Commerce Clause jurisprudence did not, however, signal a similar rejection of our Export Clause cases. Our decades-long struggle over the meaning of the nontextual negative command of the dormant Commerce Clause does not lead to the conclusion that our interpretation of the textual command of the Export Clause is equally fluid. At one time, the Court may have thought that the dormant Commerce Clause required a strict ban on state taxation of interstate commerce, but the text did not require that view. ²

The text of the Export Clause, on the other hand, expressly prohibits Congress from laying any tax or duty on exports. These textual disparities strongly suggest that shifts in the Court’s view of the scope of the dormant Commerce Clause should not, and indeed cannot, govern our interpretation of the Export Clause. Cf. Richfield Oil Corp. v. State Bd. of Equalization, 329 U. S. 69, 75–76 (1946) (distinguishing accommodations made under the Commerce Clause from the express textual prohibition of the Import-Export Clause).

The Government’s primary assertion is that modifications in our ImportExport Clause jurisprudence require parallel modifications in the Export Clause context. More specifically, the Government argues that our decisions in Michelin Tire Corp. v. Wages, 423 U. S. 276 (1976), and Department of Revenue of Wash. v. Association of Wash. Steve-

2 The Commerce Clause is an express grant of power to Congress to ‘‘regulate Commerce . . . among the several States.’’ U. S. Const., Art. I, § 8, cl. 3. It does not expressly prohibit the States from doing anything, though we have long recognized negative implications of the Clause that prevent certain state taxation even when Congress has failed to legislate. See Fulton Corp. v. Faulkner, 516 U. S. __, __ (1996) (slip op., at 4–5); Quill Corp. v. North Dakota, 504 U. S. 298, 309 (1992).

doring Cos., 435 U. S. 734 (1978), establish that States may impose generally applicable, nondiscriminatory taxes even if those taxes fall on imports or exports. The Export Clause, the Government contends, is no more restrictive.

The Import-Export Clause, which is textually similar to the Export Clause, says in relevant part, ‘‘No State shall . . . lay any Imposts or Duties on Imports or Exports.’’ U. S. Const., Art. I, § 10, cl. 2. Though minor textual differences exist and the Clauses are directed at different sovereigns, historically both have been treated as broad bans on taxation of exports, and in several cases the Court has interpreted the provisions of the two Clauses in tandem. For instance, in the Court’s first decision interpreting the Import-Export Clause, Chief Justice Marshall said:

‘‘The States are forbidden to lay a duty on exports, and the United States are forbidden to lay a tax or duty on articles exported from any State. There is some diversity in language, but none is perceivable in the act which is prohibited.’’ Brown v. Maryland, 12 Wheat. 419, 445 (1827). See also Kosydar v. National Cash Register Co., 417 U. S. 62, 67, n. 5 (1974); Hvoslef, supra, at 13–14; Cornell, 192 U. S., at 427–428; Turpin, 117 U. S., at 506–507. The Government argues that our longstanding parallel interpretations of the two Clauses require judgment in its favor. We disagree.

In Michelin, we addressed whether a State could impose a nondiscriminatory ad valorem property tax on imported goods that were no longer in import transit. Michelin, which imported tires from Canada and France and stored them in a warehouse, argued that Georgia could not constitutionally assess ad valorem property taxes against its imported tires. We explained that ‘‘[t]he Framers of the Constitution . . . sought to alleviate three main concerns’’: (i) ensuring that the Federal Government speaks with one voice when regulating foreign commerce; (ii) preserving import revenues as a major source of federal revenue; and (iii) preventing disharmony likely to be caused if seaboard States taxed goods coming through their ports. Michelin, supra, at 285–286. The Court found that nondiscriminatory ad valorem taxes violate none of these policies. A century earlier, however, the Court had ruled that, under the ‘‘original package doctrine,’’ a State could not impose such a tax until the goods had lost their

character as imports and had been incorporated into the mass of property in the State. Low v. Austin, 13 Wall. 29, 34 (1872). The Michelin Court overruled Low and held that the nondiscriminatory property tax levied on Michelin’s inventory of imported tires did not violate the Import-Export Clause because it was not an impost or duty on imports. 423 U. S., at 301. See also Limbach v. Hooven & Allison Co., 466 U. S. 353 (1984) (reaffirming that Michelin expressly overruled the original package doctrine altogether and not merely Low on its facts).

Two years later, in Washington Steve- doring, we upheld against an ImportExport Clause challenge a nondiscriminatory state tax assessed against the compensation received by stevedoring companies for services performed within the State. The Court found that Washington’s stevedoring tax did not violate the policies underlying the ImportExport Clause. Unlike the property tax at issue in Michelin, the activity taxed by Washington occurred while imports and exports were in transit. That fact was not dispositive, however, because the tax did not fall on the goods themselves:

‘‘The levy reaches only the business of loading and unloading ships or, in other words, the business of transporting cargo within the State of Washington. Despite the existence of the first distinction, the presence of the second leads to the conclusion that the Washington tax is not a prohibited ‘Impost or Duty’ when it violates none of the policies [that animate the Import-Export Clause].’’ Wash- ington Stevedoring, supra, at 755. Relying on Canton R. Co. v. Rogan, 340 U. S. 511 (1951), which upheld a tax on the gross receipts of a railroad that operated a marine terminal and transported imports and exports, we ruled in Washington Stevedoring that taxation of transportation services, whether by railroad on the docks or by stevedores loading and unloading ships, did not relate to the value of the goods and could not be considered imposts or duties on the goods themselves. 435 U.S., at 757.

A tax on policies insuring exports is not, precisely speaking, the same as a tax on exports, but Thames & Mersey held that they were functionally the

same under the Export Clause. We noted in Washington Stevedoring that one may question the finding in Thames & Mersey that the tax was essentially a tax upon the exportation itself. 435 U. S., at 756, n. 21. We expressed concern that ‘‘[t]he basis for distinguishing Thames & Mersey is less clear’’ than for Fairbank or Richfield Oil, because the marine insurance policies in Thames & Mersey arguably ‘‘had a value apart from the value of the goods.’’ 435 U. S., at 756, n. 21. Nevertheless, the Government apparently has chosen not to challenge that aspect of Thames & Mersey in this case. Tr. of Oral Arg. 5, 8–9, 40. When questioned on that implicit concession at oral argument, the Government admitted that it ‘‘chose not to’’ argue that § 4371 does not impose a tax on the goods themselves. Id., at 9. It would be inappropriate for us to reexamine in this case, without the benefit of the parties’ briefing, whether the policies on which § 4371 is assessed are so closely connected to the goods that the tax is, in essence, a tax on exports. ³ See, e.g., id., at 27–28 (‘‘[T]he record doesn’t reveal the sort of statistical information Justice Breyer was suggesting might be relevant’’ to determine ‘‘whether this is sufficiently indirect that it’s not a tax on exports, . . . because the Government has conceded throughout that they are not disputing that this tax, if discriminatory, is in violation of the Constitution’’).

Stare decisis is a ‘‘principle of policy,’’ Helvering v. Hallock, 309 U. S. 106, 119 (1940), and not ‘‘an inexorable

3 The Court has never held that the Export Clause prohibits only direct taxation of goods in export transit. In Brown v. Maryland, 12 Wheat. 419 (1827), Chief Justice Marshall expressed in dicta his skepticism that a federal occupational tax on exporters could pass scrutiny under the Export Clause. Id., at 445 (‘‘[W]ould government be permitted to shield itself from the just censure to which this attempt to evade the prohibitions of the constitution would expose it, by saying that this was a tax on the person, not on the article, and that the legislature had a right to tax occupations?’’). In Fairbank, Hvoslef, and Thames & Mersey, we struck down taxes that were not assessed directly on goods in export transit, but which the Court found to be so closely related as to be effectively a tax on the goods themselves. We have never repudiated that principle, but neither have we ever carefully defined how we decide whether a particular federal tax is sufficiently related to the goods or their value to violate the Export Clause. To the extent the issue was raised in the petition for certiorari, the Government failed to address the issue in its brief on the merits and therefore has abandoned it. See Posters ‘N’ Things, Ltd. v. United States, 511 U. S. ___, ___ (1994) (slip op., at 15); Russell v. United States, 369 U. S. 749, 754, n. 7 (1962).

command,’’ Payne v. Tennessee, 501 U. S. 808, 828 (1991). Applying that policy, we frequently have declined to overrule cases in appropriate circumstances because stare decisis ‘‘promotes the evenhanded, predictable, and consistent development of legal principles, fosters reliance on judicial decisions, and contributes to the actual and perceived integrity of the judicial process.’’ Id., at 827. ‘‘[E]ven in constitutional cases, the doctrine carries such persuasive force that we have always required a departure from precedent to be supported by some ‘special justification.’’’ Id., at 842 (SOUTER, J., concurring) (quoting Arizona v. Rumsey, 467 U. S. 203, 212 (1984)).

Though from time to time we have overruled governing decisions that are ‘‘unworkable or are badly reasoned,’’ Payne, supra, at 827; see Smith v. Allwright, 321 U. S. 649, 665 (1944), we have rarely done so on grounds not advanced by the parties. Thames & Mersey has been controlling precedent for over 80 years, and the Government does not, indeed could not, argue that the rule established there is ‘‘unworkable.’’ Despite the dissent’s speculative protestations to the contrary, post, at 9–11, there is simply no evidence that Thames & Mersey has caused or will cause uncertainty in commercial export transactions. The principles that animate our policy of stare decisis caution against overruling a long-standing precedent on a theory not argued by the parties, and we decline to do so in this case. ⁴

What the Government does argue is that our Import-Export Clause cases require us to overrule Thames & Mersey ⁵ . We have good reason to hesitate before adopting the analysis of our recent Import-Export Clause cases into our Export Clause jurisprudence. Though we have frequently interpreted the Clauses together, see supra, at 9–10, our more

4 The dissent suggests that ‘‘the Court assumes the statute to be invalid rather than deciding it to be so.’’ Post, at 2. We make no such assumptions. Rather, we begin with a longstanding decision that, by all accounts, controls this case. Even the Government agrees that Congress enacted a law whose application in this case directly contravenes our holding in Thames & Mersey . We sit not to condemn § 4371, but rather to determine whether it is to be saved by overruling binding precedent. 5 The dissent suggests that we make a ‘‘serious mistake’’ in deciding whether a nondiscriminatory tax on goods violates the Export Clause, post, at 19. We do not agree that it is a mistake to address the arguments actually advanced by the parties.

recent Import-Export Clause cases, on which the Government relies, caution that meaningful textual differences exist and should not be overlooked. The Export Clause prohibits Congress from laying any ‘‘Tax or Duty’’ on exports, while the Import-Export Clause prevents the States from laying any ‘‘Imposts or Duties’’ on imports or exports. In both Michelin and Washington Stevedoring, we left open the possibility that a particular state assessment might not properly be called an impost or duty, and thus would be beyond the reach of the Import-Export Clause, while an identical federal assessment might properly be called a tax and would be subject to the Export Clause. Though we found in Michelin that a nondiscriminatory state property tax does not transgress the policy dictates of the Import-Export Clause, we also recognized that the Import-Export Clause is ‘‘not written in terms of a broad prohibition of every ‘tax,’ ’’ and that impost and duty are narrower terms than tax. 423 U. S., at 290–293. In Washington Stevedoring, we likewise rejected the assertion that the Import-Export Clause absolutely prohibits all taxation of imports and exports. 435 U. S., at 759. We said that ‘‘the term ‘Impost or Duty’ is not selfdefining and does not necessarily encompass all taxes’’ and that the respondents’ argument to the contrary ignored ‘‘the central holding of Michelin that the absolute ban is only of ‘Imposts or Duties’ and not of all taxes.’’ Ibid.

The distinction between imposts or duties and taxes is especially pertinent in light of the peculiar definitional analysis we chose in Michelin . Finding substantial ambiguity in the phrase ‘‘Imposts or Duties,’’ we ‘‘decline[d] to presume it was intended to embrace taxation that does not create the evils the Clause was specifically intended to eliminate.’’ Michelin, supra, at 293–294. We entirely bypassed the etymological inquiry into the proper meaning of the terms ‘‘impost’’ and ‘‘duty,’’ and instead created a regime in which those terms are conclusions to be drawn from an examination into whether a particular assessment ‘‘was the type of exaction that was regarded as objectionable by the Framers of the Constitution.’’ 423 U. S., at 286. We are not prepared to say that the word ‘‘Tax’’ is ‘‘sufficiently ambiguous,’’ id., at 293, that we may ignore its common, and usually expan

sive, ⁶ meaning in favor of an Export Clause decisional rule in which a tax is not a ‘‘Tax’’ unless it discriminates against exports. Consequently, Michelin and Washington Stevedoring, which held that the assessments in question were not ‘‘Imposts or Duties’’ at all, do not logically validate the assessment at issue in this case, which, by all accounts, remains a ‘‘Tax.’’

It is not intuitively obvious that Michelin ’s three-pronged analysis of the Framers’ concerns is really just another way of stating a nondiscrimination principle. But even if it were, the Government cannot reasonably rely on Michelin to govern the Export Clause because Michelin drew its analysis around the phrase ‘‘Imposts or Duties’’ and expressly excluded the broader term ‘‘Tax’’ that appears in the Export Clause. Michelin marked a more permissive approach to state taxation under the Import-Export Clause only by distinguishing the presumptively stricter language of the Export Clause. We agree with the Government that Michelin informs our decision in this case, but not in a way that supports the Government’s position. It is simply no longer true that the Court perceives no substantive difference between the two Clauses.

We are similarly hesitant to adopt the Import-Export Clause’s policy-based analysis without some indication that the Export Clause was intended to alleviate the same ‘‘evils’’ to which the ImportExport Clause was directed. Unlike the Import-Export Clause, which was intended to protect federal supremacy in international commerce, to preserve federal revenue from import duties and imposts, and to prevent coastal States with ports from taking unfair advantage of inland States, see Michelin, supra, at 285–286, the Export Clause serves none of those goals. Indeed, textually, the Export Clause does quite the opposite. It specifically prohibits Congress from regulating international commerce through export taxes, disallows any attempt to raise federal revenue from exports, and has no direct effect on the way the States treat imports and exports.

As a purely historical matter, the Export Clause was originally proposed

6 Though Michelin discusses ‘‘taxes’’ in terms of ‘‘every exaction,’’ 423 U. S., at 290, it also suggests that at the time of the Founding ‘‘probably only capitation, land, and general property exactions were known by the term ‘tax’ rather than the term ‘duty,’ ’’ id., at 291. In any event, the Michelin Court understood that the terms used in the Export Clause were broader than those used in the Import-Export Clause.

by delegates to the Federal Convention from the Southern States, who feared that the Northern States would control Congress and would use taxes and duties on exports to raise a disproportionate share of federal revenues from the South. See 2 M. Farrand, The Records of the Federal Convention of 1787, pp. 95, 305–308, 359–363 (rev. ed. 1966). The Government argues that this ‘‘narrow historical purpose’’ justifies a narrow interpretation of the text and that application of § 4371 to policies insuring exports does not conflict with the policies embodied in the Clause. Brief for United States 32–34. While the original impetus may have had a narrow focus, the remedial provision that ultimately became the Export Clause does not, and there is substantial evidence from the Debates that proponents of the Clause fully intended the breadth of scope that is evident in the language. See, e. g., 2 Farrand, Records of the Federal Convention, at 220 (Mr. King: ‘‘In two great points the hands of the Legislature were absolutely tied. The importation of slaves could not be prohibited—exports could not be taxed’’); id., at 305 (‘‘Mr. Mason urged the necessity of connecting with the power of levying taxes . . . that no tax should be laid on exports’’); id., at 360 (Mr. Elseworth [sic]: ‘‘There are solid reasons agst. Congs taxing exports’’); ibid . (‘‘Mr. Butler was strenuously opposed to a power over exports’’); id., at 361 (Mr. Sherman: ‘‘It is best to prohibit the National legislature in all cases’’); id., at 362 (‘‘Mr. Gerry was strenuously opposed to the power over exports’’).

The Government argued for a different narrow interpretation of the Export Clause in Fairbank . See 181 U. S., at 292–293. Arguing that the Debates expressed a primary interest in diffusing sectional conflicts, the Government urged the Fairbank Court to interpret the Export Clause to permit taxation of ‘‘the act of exportation or the document evidencing the receipt of goods for export, for these exist with substantial uniformity throughout the country.’’ Id., at 292. We rejected that argument:

‘‘If mere discrimination between the States was all that was contemplated, it would seem to follow that an ad valorem tax upon all exports would not be obnoxious to this constitutional prohibition. But surely under this limitation Congress can impose an export tax neither on one article of export, nor on all articles of export.’’ Ibid.

As in Fairbank, we think the text of the constitutional provision provides a better decisional guide than that offered by the Government. The Government’s policy argument—that the Framers intended the Export Clause to narrowly alleviate the fear of northern repression through taxation of southern exports by prohibiting only discriminatory taxes—cannot be squared with the broad language of the Clause. The better reading, that adopted by our earlier cases, is that the Framers sought to alleviate their concerns by completely denying to Congress the power to tax exports at all.

3 Even assuming that Michelin and Washington Stevedoring govern our Export Clause inquiry in this case, the Government’s argument falls short of its goal. Our holdings in Michelin and Washington Stevedoring do not reach the facts of this case and, more importantly, do not interpret the Import-Export Clause to permit assessment of nondiscriminatory taxes on imports and exports in transit. Michelin involved a tax on goods, but the goods were no longer in transit. The tax in Washington Steve- doring burdened imports and exports while they were still in transit, but it did not fall directly on the goods themselves. This case, as it comes to us, is a hybrid in which the tax both burdens exports during transit and—as the Government concedes and our earlier cases held—is essentially a tax on the goods themselves. The Government argues that Michelin and Washington Stevedoring by analogy permit Congress to impose generally applicable, nondiscriminatory taxes that fall directly on exports in transit. Brief for United States 32 ( Michelin and Washington Stevedoring ‘‘demonstrate that, when a generally applicable, nondiscriminatory tax is at issue, the mere fact that the tax applies also to goods that are in the export or import process does not provide a constitutional immunity from taxation’’). If this contention is to succeed, the Government at the very least must show that our Import-Export Clause jurisprudence now permits a State to impose a nondiscriminatory tax directly on goods in import or export transit. We think the Government has failed to make that showing.

The Court has never upheld a state tax assessed directly on goods in import or export transit. In Michelin, we suggested that the Import-Export Clause would invalidate application of a non

discriminatory property tax to goods still in import or export transit. 423 U. S., at 290 (compliance with the Import-Export Clause may be secured ‘‘by prohibiting the assessment of even nondiscriminatory property taxes on [import or export] goods which are merely in transit through the State when the tax is assessed’’). See also Virginia Indonesia Co. v. Harris County Appraisal Dist., 910 S. W. 2d 905, 915 (Tex. 1995) (invalidating application of a nondiscriminatory ad valorem property tax to goods in export transit).

We also declined to endorse the Government’s theory in Washington Steve- doring . After reciting that the Court in Canton R. Co. had distinguished Thames & Mersey, Fairbank, and Richfield Oil, we pointed out that in those cases ‘‘the State [or Federal Government] had taxed either the goods or activity so connected with the goods that the levy amounted to a tax on the goods themselves.’’ Washington Stevedoring, 435 U. S., at 756, n. 21. We expressly declined to ‘‘reach the question of the applicability of the Michelin approach when a State directly taxes imports or exports in transit,’’ id., at 757, n. 23, because, although the goods in that case were in transit, the tax fell on ‘‘a service distinct from the goods and their value,’’ id., at 757. Thus, contrary to the Government’s contention, this Court’s Import-Export Clause cases have not upheld the validity of generally applicable, nondiscriminatory taxes that fall on imports or exports in transit. We think those cases leave us free to follow the express textual command of the Export Clause to prohibit the application of any tax ‘‘laid on Articles exported from any State.’’

- - - - We conclude that the Export Clause

does not permit assessment of nondiscriminatory federal taxes on goods in export transit. Reexamination of the question whether a particular assessment on an activity or service is so closely connected to the goods as to amount to a tax on the goods themselves must await another day. We decline to overrule Thames & Mersey . The judgment of the Court of Appeals for the Federal Circuit is affirmed.

It is so ordered.

Section 6323.—Validity and Priority Against Certain Persons

Ct.D. 2059

SUPREME COURT OF THE

UNITED STATES

No. 95–323

UNITED STATES, PETITIONER v. THOMAS R. NOLAND, TRUSTEE

FOR DEBTOR FIRST TRUCK

LINES, INC.

517 U.S.—

ON WRIT OF CERTIORARI TO THE

UNITED STATES COURT OF APPEALS FOR THE SIXTH CIRCUIT

May 13, 1996

Syllabus

The Internal Revenue Service filed claims in the Bankruptcy Court for taxes, interest, and penalties that accrued after debtor First Truck Lines, Inc., sought relief under Chapter 11 of the Bankruptcy Code but before the case was converted to a Chapter 7 bankruptcy. The court found that all of the IRS’s claims were entitled to first priority as administrative expenses under 11 U. S. C. §§ 503(b)(1)(C) and 507(a)(1), but held that the penalty claim was subject to ‘‘equitable subordination’’ under § 510(c), which the court interpreted as giving it authority not only to deal with inequitable Government conduct, but also to adjust a statutory priority of a category of claims. The court’s decision to subordinate the penalty claim to the claims of the general unsecured creditors was affirmed by the District Court and the Sixth Circuit, which concluded that postpetition, nonpecuniary loss tax penalty claims are susceptible to subordination by their very nature.

Held : A bankruptcy court may not equitably subordinate claims on a categorical basis in derogation of Congress’s priorities scheme. The language of § 510(c), principles of statutory construction, and legislative history clearly indicate Congress’s intent in its 1978 revision of the Code to use the existing judge-made doctrine of equitable subordination as the starting point for deciding when subordination is appropriate. By adopting ‘‘principles of equitable subordination,’’ § 510(c) allows a bankruptcy court to reorder a tax penalty when justified by particular facts. It is also clear that Congress meant to give courts some leeway to develop the doctrine. However, a reading of the statute that would give courts leeway broad enough to allow subordination at odds with the congressional ordering of priorities by category is improbable in the extreme. The statute would then empower a court to modify the priority provision’s operation at the same level at which Congress operated when it made its characteristically general judgment to establish the hierarchy of claims in the first place, thus delegating legislative revision, not authorizing equitable exception. Nonetheless, just such a legislative type of decision underlies the reordering of priorities here. The Sixth Circuit’s decision runs directly counter to Congress’s policy judgment that a postpetition tax penalty should receive the priority of an administrative expense. Since the Sixth Circuit’s rationale was inappropriately categorical

in nature, this Court need not decide whether a bankruptcy court must always find creditor misconduct before a claim may be equitably subordinated. 48 F. 3d 210, reversed and remanded. SOUTER, J., delivered the opinion for a unanimous Court.

JUSTICE SOUTER delivered the opinion of the Court.

The issue in this case is the scope of a bankruptcy court’s power of equitable subordination under 11 U.S.C. § 510(c). Here, in the absence of any finding of inequitable conduct on the part of the Government, the Bankruptcy Court subordinated the Government’s claim for a postpetition, noncompensatory tax penalty, which would normally receive first priority in bankruptcy as an ‘‘administrative expense,’’ §§ 503(b)(1)(C), 507(a)(1). We hold that the bankruptcy court may not equitably subordinate claims on a categorical basis in derogation of Congress’s scheme of priorities.

In April 1986, First Truck Lines, Inc., voluntarily filed for relief under Chapter 11 of the Bankruptcy Code, and in the subsequent operation of its business as a debtor-in-possession incurred, but failed to discharge, tax liabilities to the Internal Revenue Service. First Truck moved to convert the case to a Chapter 7 liquidation in June 1988, and in August 1988 the Bankruptcy Court granted that motion and appointed respondent Thomas R. Noland as trustee. The liquidation of the estate’s assets raised insufficient funds to pay all of the creditors.

After the conversion, the IRS filed claims for taxes, interest, and penalties that accrued after the Chapter 11 filing but before the Chapter 7 conversion, and although the parties agreed that the claims for taxes and interest were entitled to priority as administrative expenses, §§ 503(b), 507(a)(1), and 726(a)(1), ¹ they disagreed about the priority to be given tax penalties. The Bankruptcy Court determined that the penalties (like the taxes and interest) were administrative expenses under § 503(b) but held them to be subject to

1 Section 507(a)(1) provides, in relevant part: ‘‘(a) The following expenses and claims have priority in the following order: (1) First, administrative expenses allowed under section 503(b) of this title . . . .’’ Under § 503(b)(1), administrative expenses include ‘‘any tax . . . incurred by the estate’’ (with certain exceptions not relevant here), as well as ‘‘any fine [or] penalty . . . relating to [such] a tax . . . .’’ Section 726(a)(1) adopts the order of payment specified in § 507 for Chapter 7 proceedings.

equitable subordination under § 510(c). ²

In so doing, the Court read that section to provide authority not only to deal with inequitable conduct on the Government’s part, but also to adjust a statutory priority of a category of claims. The Bankruptcy Court accordingly weighed the relative equities that seemed to flow from what it described as ‘‘the Code’s preference for compensating actual loss claims,’’ and subordinated the tax penalty claim to those of the general unsecured creditors. In re First Truck Lines, Inc., 141 B. R. 621, 629 (SD Ohio 1992). The District Court affirmed. In- ternal Revenue Service v. Noland, 190 B. R. 827 (SD Ohio 1993).

After reviewing the legislative history of the 1978 revision to the Bankruptcy Code and several recent appeals cases on equitable subordination of tax penalties, the Sixth Circuit affirmed, as well. In re First Truck Lines, Inc., 48 F. 3d 210 (1995). The Sixth Circuit stated that it did

‘‘not see the fairness or the justice in permitting the Commissioner’s claim for tax penalties, which are not being assessed because of pecuniary losses to the Internal Revenue Service, to enjoy an equal or higher priority with claims based on the extension of value to the debtor, whether secured or not. Further, assessing tax penalties against the estate of a debtor no longer in existence serves no punitive purpose. Because of the nature of postpetition, nonpecuniary loss tax penalty claims in a Chapter 7 case, we believe such claims are susceptible to subordination. To hold otherwise would be to allow creditors who have supported the business during its attempt to reorganize to be penalized once that effort has failed and there is not enough to go around.’’ Id., at 218. See also Burden v. United States, 917 F. 2d 115, 120 (CA3 1990); Schultz Broad- way Inn v. United States, 912 F. 2d 230, 234 (CA8 1990); In re Virtual Network Services Corp., 902 F. 2d 1246, 1250 (CA7 1990). We granted certiorari to determine the appropriate scope of the power under the Bankruptcy Code to subordinate a tax penalty, 516 U. S. ___ (1995), and we now reverse.

2 Section 510(c) provides that ‘‘the court may . . . under principles of equitable subordination, subordinate for purposes of distribution all or part of an allowed claim . . . .’’

The judge-made doctrine of equitable subordination predates Congress’s revision of the Code in 1978. Relying in part on our earlier cases, see, e.g., Comstock v. Group of Institutional Investors, 335 U. S. 211 (1948); Pepper v. Litton, 308 U. S. 295 (1939); Taylor v. Standard Gas & Elec. Co., 306 U. S. 307 (1939), the Fifth Circuit, in its influential opinion in In re Mobile Steel Co., 563 F. 2d 692, 700 (CA5 1977), observed that the application of the doctrine was generally triggered by a showing that the creditor had engaged in ‘‘some type of inequitable conduct.’’ Mobile Steel discussed two further conditions relating to the application of the doctrine: that the misconduct have ‘‘resulted in injury to the creditors of the bankrupt or conferred an unfair advantage on the claimant,’’ and that the subordination ‘‘not be inconsistent with the provisions of the Bankruptcy Act.’’ Ibid . This last requirement has been read as a ‘‘reminder to the bankruptcy court that although it is a court of equity, it is not free to adjust the legally valid claim of an innocent party who asserts the claim in good faith merely because the court perceives that the result is inequitable.’’ DeNatale & Abram, The Doctrine of Equitable Subordination as Applied to Nonmanagement Creditors, 40 Bus. Law. 417, 428 (1985). The district courts and courts of appeals have generally followed the Mobile Steel formulation, In re Baker & Getty Financial Services, Inc., 974 F. 2d 712, 717 (CA6 1992). Although Congress included no explicit criteria for equitable subordination when it enacted § 510(c)(1), the reference in § 510(c) to ‘‘principles of equitable subordination,’’ clearly indicates congressional intent at least to start with existing doctrine. This conclusion is confirmed both by principles of statutory construction, see Midlantic Nat. Bank v. New Jersey Dept. of Environ- mental Protection, 474 U. S. 494, 501 (1986) (‘‘The normal rule of statutory construction is that if Congress intends for legislation to change the interpretation of a judicially created concept, it makes that intent specific. The Court has followed this rule with particular care in construing the scope of bankruptcy codifications’’) (citation omitted), and by statements in the legislative history that Congress ‘‘intended that the term ‘principles of equitable subordination’ follow existing case law and leave to the courts development of this principle,’’ 124 Cong. Rec. 32398 (1978)

(Rep. Edwards); see also id., at 33998 (Sen. DeConcini). In keeping with pre1978 doctrine, many Courts of Appeals have continued to require inequitable conduct before allowing the equitable subordination of most claims, see, e.g., In re Fabricators, Inc., 926 F. 2d 1458, 1464 (CA5 1991); In re Bellanca Air- craft Corp., 850 F. 2d 1275, 1282–1283 (CA8 1988), although several have done away with the requirement when the claim in question was a tax penalty. See, e.g., Burden, supra, at 120; Schultz, supra, at 234; In re Virtual Network, supra, at 1250.

Section 510(c) may of course be applied to subordinate a tax penalty, since the Code’s requirement that a Chapter 7 trustee must distribute assets ‘‘in the order specified in . . . section 507,’’ (which gives a first priority to administrative expense tax penalties) is subject to the qualification, ‘‘[e]xcept as provided in section 510 of this title . . . .’’ 11 U.S.C. § 726(a). Thus, ‘‘principles of equitable subordination’’ may allow a bankruptcy court to reorder a tax penalty in a given case. It is almost as clear that Congress meant to give courts some leeway to develop the doctrine, 124 Cong. Rec. 33998 (1978), rather than to freeze the pre-1978 law in place. The question is whether that leeway is broad enough to allow subordination at odds with the congressional ordering of priorities by category.

The answer turns on Congress’s probable intent to preserve the distinction between the relative levels of generality at which trial courts and legislatures respectively function in the normal course. Hence, the adoption in § 510(c) of ‘‘principles of equitable subordination’’ permits a court to make exceptions to a general rule when justified by particular facts, cf. Hecht Co. v. Bowles, 321 U. S. 321, 329 (1944) (‘‘The essence of equity jurisdiction has been the power of the Chancellor to do equity and to mould each decree to the necessities of the particular case’’). But if the provision also authorized a court to conclude on a general, categorical level that tax penalties should not be treated as administrative expenses to be paid first, it would empower a court to modify the operation of the priority statute at the same level at which Congress operated when it made its characteristically general judgment to establish the hierarchy of claims in the first place. That is, the distinction between characteristic legislative and trial court functions would simply be swept away, and

the statute would delegate legislative revision, not authorize equitable exception. We find such a reading improbable in the extreme. ‘‘Decisions about the treatment of categories of claims in bankruptcy proceedings . . . are not dictated or illuminated by principles of equity and do not fall within the judicial power of equitable subordination . . . .’’ Burden, 917 F. 2d, at 122 (Alito, J., concurring in part and dissenting in part).

Just such a legislative type of decision, however, underlies the Bankruptcy Court’s reordering of priorities in question here, as approved by the District Court and the Court of Appeals. Despite language in its opinion about requiring a balancing of the equities in individual cases, the Court of Appeals actually concluded that ‘‘postpetition, nonpecuniary loss tax penalty claims’’ are ‘‘susceptible to subordination’’ by their very ‘‘nature.’’ 48 F. 3d, at 218. And although the court said that not every tax penalty would be equitably subordinated, ibid., that would be the inevitable result of consistent applications of the rule employed here, which depends not on individual equities but on the supposedly general unfairness of satisfying ‘‘postpetition, nonpecuniary loss tax penalty claims’’ before the claims of a general creditor.

The Court of Appeals’s decision thus runs directly counter to Congress’s policy judgment that a postpetition tax penalty should receive the priority of an administrative expense, 11 U.S.C. §§ 503(b)(1)(C), 507(a)(1), and 726(a)(1). This is true regardless of Noland’s argument that the Bankruptcy Court made a distinction between compensatory and noncompensatory tax penalties, for this was itself a categorical distinction at a legislative level of generality. Indeed, Congress recognized and employed that distinction elsewhere in the priority provisions: Congress specifically assigned 8th priority to certain compensatory tax penalties, see § 507(a)(8)(G), and 12th priority to prepetition, noncompensatory penalties, see § 726(a)(1), and (4). ³

3 Noland argues that ‘‘although the penalties at issue arose postpetition,’’ this claim should be viewed as a prepetition penalty because a ‘‘reorganized debtor is in many respects similar to a prepetition debtor . . . [and] the conversion of

[this] case to chapter 7 was tantamount to the filing of a new petition.’’ Brief for Respondent 16, n. 7. But we agree with the Sixth Circuit, see In re First Truck Lines, Inc., 48 F. 3d 210, 214 (1995), that the penalties at issue here are postpetition administrative expenses pursuant to 11 U. S. C.

The Sixth Circuit, to be sure, invoked a more modest authority than legislative revision when it relied on statements by the congressional leaders of the 1978 Code revisions, see 48 F. 3d, at 215, 217–218, and it is true that Representative Edwards and Senator DeConcini stated that ‘‘under existing law, a claim is generally subordinated only if [the] holder of such claim is guilty of inequitable conduct, or the claim itself is of a status susceptible to subordination, such as a penalty or a claim for damages arising from the purchase or sale of a security of the debtor.’’ 124 Cong. Rec. 32398 (1978) (Rep. Edwards); see also id., at 33998 (Sen. DeConcini). But their remarks were not statements of existing law and the Sixth Circuit’s reliance on the unexplained reference to subordinated penalties ran counter to this Court’s previous endorsement of priority treatment for postpetition tax penalties. See Nicholas v. United States, 384 U. S. 678, 692–695 (1966). More fundamentally, statements in legislative history cannot be read to convert statutory leeway for judicial development of a rule on particularized exceptions into delegated authority to revise statutory categorization, untethered to any obligation to preserve the coherence of substantive congressional judgments.

Given our conclusion that the Sixth Circuit’s rationale was inappropriately categorical in nature, we need not decide today whether a bankruptcy court must always find creditor misconduct before a claim may be equitably subordinated. We do hold that (in the absence of a need to reconcile conflicting congressional choices) the circumstances that prompt a court to order equitable subordination must not occur at the level of policy choice at which Congress itself operated in drafting the Bankruptcy Code. Cf. In re Ahlswede, 516 F. 2d 784, 787 (CA9) (‘‘[T]he [equity] chancellor never did, and does not now, exercise unrestricted power to contradict statutory or common law when he feels

§§ 348(d), 503(b)(1). Although § 348(d) provides that a ‘‘claim against the estate or the debtor that arises after the order for relief but before conversion in a case that is converted under section 1112, 1208, or 1307 of this title, other than a claim specified in section 503(b) of this title, shall be treated for all purposes as if such claim had arisen immediately before the date of the filing of the petition,’’ the claim for priority here is ‘‘specified in section 503(b)’’ and Congress has already determined that it is not to be treated like prepetition penalties. Noland may or may not have a valid policy argument, but it is up to Congress, not this Court, to revise the determination if it so chooses.

a fairer result may be obtained by application of a different rule’’), cert. denied sub nom. Stebbins v. Crocker Citizens Nat. Bank, 423 U.S. 913 (1975); In re Columbia Ribbon Co., 117 F. 2d 999, 1002 (CA3 1941) (court cannot ‘‘set up a subclassification of claims . . . and fix an order of priority for the sub-classes according to its theory of equity’’).

In this instance, Congress could have, but did not, deny noncompensatory, postpetition tax penalties the first priority given to other administrative expenses, and bankruptcy courts may not take it upon themselves to make that categorical determination under the guise of equitable subordination. The judgment of the Court of Appeals is reversed, and the case is remanded for further proceedings consistent with this opinion.

It is so ordered.

Section 6512.—Limitations in Case of Petition to Tax Court

Ct.D. 2058

SUPREME COURT OF THE

UNITED STATES

No. 94–1785

COMMISSIONER OF INTERNAL REVENUE, PETITIONER v. ROBERT

F. LUNDY

516 U.S.—

Court under 26 U.S.C. § 6512(b)(3)(B), and decide whether the Tax Court can award a refund of taxes paid more than two years prior to the date on which the Commissioner of Internal Revenue mailed the taxpayer a notice of deficiency, when, on the date the notice of deficiency was mailed, the taxpayer had not yet filed a return. We hold that in these circumstances the 2-year lookback period set forth in § 6512(b)(3)(B) applies, and the Tax Court lacks jurisdiction to award a refund.

During 1987, respondent Robert F. Lundy and his wife had $10,131 in federal income taxes withheld from their wages. This amount was substantially more than the $6,594 the Lundys actually owed in taxes for that year, but the Lundys did not file their 1987 tax return when it was due, nor did they file a return or claim a refund of the overpaid taxes in the succeeding two and a half years. On September 26, 1990, the Commissioner of Internal Revenue mailed Lundy a notice of deficiency, informing him that he owed $7,672 in additional taxes and interest for 1987 and that he was liable for substantial penalties for delinquent filing and negligent underpayment of taxes, see 26 U. S. C. §§ 6651(a)(1) and 6653(1).

Lundy and his wife mailed their joint tax return for 1987 to the Internal Revenue Service (IRS) on December 22, 1990. This return indicated that the Lundys had overpaid their income taxes for 1987 by $3,537 and claimed a refund in that amount. Two days after the return was mailed, Lundy filed a timely petition in the Tax Court seeking a redetermination of the claimed deficiency and a refund of the couple’s overpaid taxes. The Commissioner filed an answer generally denying the allegations in Lundy’s petition. Thereafter, the parties negotiated towards a settlement of the claimed deficiency and refund claim. On March 17, 1992, the Commissioner filed an amended answer acknowledging that Lundy had filed a tax return and that Lundy claimed to have overpaid his 1987 taxes by $3,537.

The Commissioner contended in this amended pleading that the Tax Court lacked jurisdiction to award Lundy a refund. The Commissioner argued that if a taxpayer does not file a tax return before the IRS mails the taxpayer a notice of deficiency, the Tax Court can only award the taxpayer a refund of

ON WRIT OF CERTIORARI TO THE

UNITED STATES COURT OF APPEALS FOR THE FOURTH

CIRCUIT

January 17, 1996

Syllabus

Respondent Lundy and his wife withheld from their 1987 wages substantially more in federal income taxes than they actually owed for that year, but they did not file their 1987 tax return when it was due, nor did they file a return or claim a refund of the overpaid taxes in the succeeding 2½ years. On September 26, 1990, the Commissioner of Internal Revenue mailed Lundy a notice of deficiency for 1987. Some three months later, the Lundys filed their joint 1987 tax return, which claimed a refund of their overpaid taxes, and Lundy filed a timely petition in the Tax Court seeking a redetermination of the claimed deficiency and a refund. The Tax Court held that where, as here, a taxpayer has not filed a tax return by the time a notice of deficiency is mailed, and the notice is mailed more than two years after the date on which the taxes are paid, a 2-year ‘‘look-back’’ period applies under 26 U. S. C. § 6512(b)(3)(B), and the court lacks jurisdiction to award a refund. The Fourth Circuit reversed, finding that the applicable look-back period in

these circumstances is three years and that the Tax Court had jurisdiction to award a refund.

Held: The Tax Court lacks jurisdiction to award a refund of taxes paid more than two years prior to the date on which the Commissioner mailed the taxpayer a notice of deficiency, if, on the date that the notice was mailed, the taxpayer had not yet filed a return. In these circumstances, the applicable look-back period under § 6512(b)(3)(B) is two years.

(a) Section 6512(b)(3)(B) forbids the Tax Court to award a refund unless it first determines that the taxes were paid ‘‘within the [look-back] period which would be applicable under section 6511(b)(2) . . . if on the date of the mailing of the notice of deficiency a claim [for refund] had been filed.’’ Section § 6511(b)(2)(A) in turn instructs the court to apply a 3-year look-back period if a refund claim is filed, as required by § 6511(a), ‘‘within 3 years from the time the return was filed,’’ while § 6511(b)(2)(B) specifies a 2-year look-back period if the refund claim is not filed within that 3-year period. The Tax Court properly applied the 2-year look-back period to Lundy’s case because, as of September 26, 1990 (the date the notice of deficiency was mailed), Lundy had not filed a tax return, and, consequently, a claim filed on that date would not be filed within the 3-year period described in § 6511(a). Lundy’s taxes were withheld from his wages, so they are deemed paid on the date his 1987 tax return was due (April 15, 1988), which is more than two years prior to the date the notice of deficiency was mailed. Lundy is therefore seeking a refund of taxes paid outside the applicable look-back period, and the Tax Court lacks jurisdiction to award a refund.

(b) Lundy suggests two alternative interpretations of § 6512(b)(3)(B), neither of which is persuasive. Lundy first adopts the Fourth Circuit’s view, which is that the applicable look-back period is determined by reference to the date that the taxpayer actually filed a claim for refund, and argues that he is entitled to a 3-year look-back period because his late-filed 1987 tax return contained a refund claim that was filed within three years from the filing of the return itself. This interpretation is contrary to the requirements of the statute and leads to a result that Congress could not have intended, as it in some circumstances subjects a timely filer of a return to a shorter limitations period in Tax Court than a delinquent filer. Lundy’s second argument, that the ‘‘claim’’ contemplated by § 6512(b)(3)(B) can only be a claim filed on a tax return, such that a uniform 3-year look-back period applies under that section, is similarly contrary to the language of the statute.

(c) This Court is bound by § 6512(b)(3)(B)’s language as it is written, and even if the Court were persuaded by Lundy’s policy-based arguments for applying a 3-year look-back period, the Court is not free to rewrite the statute simply because its effects might be susceptible of improvement. 45 F. 3d 856, reversed. O’CONNOR, J., delivered the opinion of the Court, in which REHNQUIST, C. J., and SCALIA, KENNEDY, SOUTER, GINSBURG, and BREYER, JJ., joined. STEVENS, J., filed a dissenting opinion. THOMAS, J., filed a dissenting opinion, in which STEVENS, J., joined.

JUSTICE O’CONNOR delivered the opinion of the Court.

In this case, we consider the ‘‘lookback’’ period for obtaining a refund of overpaid taxes in the United States Tax

taxes paid within two years prior to the date the notice of deficiency was mailed. See 26 U.S.C. § 6512(b)(3)(B). Under the Commissioner’s interpretation of § 6512(b)(3)(B), the Tax Court lacked jurisdiction to award Lundy a refund because Lundy’s withheld taxes were deemed paid on the date that his 1987 tax return was due (April 15, 1988), see § 6513(b)(1), which is more than two years before the date the notice was mailed (September 26, 1990).

The Tax Court agreed with the position taken by the Commissioner and denied Lundy’s refund claim. Citing an unbroken line of Tax Court cases adopting a similar interpretation of § 6512(b)(3)(B), e.g. Allen v. Commissioner, 99 T. C. 475, 479–480 (1992); Galuska v. Commissioner, 98 T. C. 661, 665 (1992); Berry v. Commissioner, 97 T. C. 339, 344–345 (1991); White v. Commis- sioner, 72 T. C. 1126, 1131–1133 (1979) (renumbered statute); Hosking v. Com- missioner, 62 T. C. 635, 642–643 (1974) (renumbered statute), the Tax Court held that if a taxpayer has not filed a tax return by the time the notice of deficiency is mailed, and the notice is mailed more than two years after the date on which the taxes are paid, the look-back period under § 6512(b)(3)(B) is two years and the Tax Court lacks jurisdiction to award a refund. 65 TCM 3011, 3014–3015, RIA TC memo ¶93, 278 (1993).

The Court of Appeals for the Fourth Circuit reversed, finding that the applicable look-back period in these circumstances is three years and that the Tax Court had jurisdiction to award Lundy a refund. 45 F. 3d 856, 861 (1995). Every other Court of Appeals to have addressed the question has affirmed the Tax Court’s interpretation of § 6512(b)(3)(B), see Davison v. Com- missioner, 9 F. 3d 1538 (CA2 1993) (unpublished disposition); Allen v. Com- missioner, 23 F. 3d 406 (CA6 1994) (unpublished disposition); Galuska v. Commissioner, 5 F. 3d 195, 196 (CA7 1993); Richards v. Commissioner, 37 F. 3d 587, 589 (CA10 1994); see also Rossman v. Commissioner, 46 F. 3d 1144 (CA9 1995) (unpublished disposition) (affirming on other grounds). We granted certiorari to resolve the conflict, 515 U. S. (1995), and now reverse.

A taxpayer seeking a refund of overpaid taxes ordinarily must file a timely claim for a refund with the Internal

Revenue Service (IRS) under 26 U. S. C. § 6511. ¹ That section contains two separate provisions for determining the timeliness of a refund claim. It first establishes a filing deadline: The taxpayer must file a claim for a refund ‘‘within 3 years from the time the return was filed or 2 years from the time the tax was paid, whichever of such periods expires the later, or if no return was filed by the taxpayer, within 2 years from the time the tax was paid.’’ § 6511(b)(1) (incorporating by reference § 6511(a)). It also defines two ‘‘look- back’’ periods: If the claim is filed ‘‘within 3 years from the time the return was filed,’’ ibid., then the taxpayer is entitled to a refund of ‘‘the portion of the tax paid within the 3 years immediately preceding the filing of the claim.’’ § 6511(b)(2)(A) (incorporating by refer 1 In relevant part, 26 U. S. C. § 6511 provides: ‘‘(a) Period of limitation on filing claim Claim for credit or refund of an overpayment of any tax imposed by this title in respect of which tax the taxpayer is required to file a return shall be filed by the taxpayer within 3 years from the time the return was filed or 2 years from the time the tax was paid, whichever of such periods expires the later, or if no return was filed by the taxpayer, within 2 years from the time the tax was paid. Claim for credit or refund of an overpayment of any tax imposed by this title which is required to be paid by means of a stamp shall be filed by the taxpayer within 3 years from the time the tax was paid.

‘‘(b) Limitation on allowance of credits and refunds

‘‘(1) Filing of claim within prescribed period No credit or refund shall be allowed or made after the expiration of the period of limitation prescribed in subsection (a) for the filing of a claim for credit or refund, unless a claim for credit or refund is filed by the taxpayer within such period.

‘‘(2) Limit on amount of credit or refund ‘‘(A) Limit where claim filed within 3-year period

If the claim was filed by the taxpayer during the 3-year period prescribed in subsection (a), the amount of the credit or refund shall not exceed the portion of the tax paid within the period, immediately preceding the filing of the claim, equal to 3 years plus the period of any extension of time for filing the return. If the tax was required to be paid by means of a stamp, the amount of the credit or refund shall not exceed the portion of the tax paid within the 3 years immediately preceding the filing of the claim.

‘‘(B) Limit where claim not filed within 3-year period

If the claim was not filed within such 3-year period, the amount of the credit or refund shall not exceed the portion of the tax paid during the 2 years immediately preceding the filing of the claim.

‘‘(C) Limit if no claim filed If no claim was filed, the credit or refund shall not exceed the amount which would be allowable under subparagraph (A) or (B), as the case may be, if claim was filed on the date the credit or refund is allowed.’’

ence § 6511(a)). If the claim is not filed within that 3-year period, then the taxpayer is entitled to a refund of only that ‘‘portion of the tax paid during the 2 years immediately preceding the filing of the claim.’’ § 6511(b)(2)(B) (incorporating by reference § 6511(a)).

Unlike the provisions governing refund suits in United States District Court or the United States Court of Federal Claims, which make timely filing of a refund claim a jurisdictional prerequisite to bringing suit, see 26 U.S.C. § 7422(a); Martin v. United States, 833 F. 2d 655, 658–659 (CA7 1987), the restrictions governing the Tax Court’s authority to award a refund of overpaid taxes incorporate only the look-back period and not the filing deadline from § 6511. See 26 U.S.C. § 6512(b)(3). ² Consequently, a taxpayer who seeks a refund in the Tax Court, like respondent, does not need to actually file a claim for refund with the IRS; the taxpayer need only show that the tax to be refunded was paid during the applicable look-back period.

2 In relevant part, 26 U. S. C. § 6512(b) provides: ‘‘(1) Jurisdiction to determine Except as provided by paragraph (3) and by section 7463, if the Tax Court finds that there is no deficiency and further finds that the taxpayer has made an overpayment of income tax for the same taxable year . . . in respect of which the Secretary determined the deficiency, or finds that there is a deficiency but that the taxpayer has made an overpayment of such tax, the Tax Court shall have jurisdiction to determine the amount of such overpayment, and such amount shall, when the decision of the Tax Court has become final, be credited or refunded to the taxpayer.

. . . . . ‘‘(3) Limit on amount of credit or refund No such credit or refund shall be allowed or made of any portion of the tax unless the Tax Court determines as part of its decision that such portion was paid—

‘‘(A) after the mailing of the notice of deficiency,

‘‘(B) within the period which would be applicable under section 6511(b)(2), (c), or (d), if on the date of the mailing of the notice of deficiency a claim had been filed (whether or not filed) stating the grounds upon which the Tax Court finds that there is an overpayment, or

‘‘(C) within the period which would be applicable under section 6511(b)(2), (c), or (d), in respect of any claim for refund filed within the applicable period specified in section 6511 and before the date of the mailing of the notice of deficiency’’

‘‘(i) which had not been disallowed before that date,

‘‘(ii) which had been disallowed before that date and in respect of which a timely suit for refund could have been commenced as of that date, or

‘‘(iii) in respect of which a suit for refund had been commenced before that date and within the period specified in section 6532.’’

In this case, the applicable look-back period is set forth in § 6512(b)(3)(B), which provides that the Tax Court cannot award a refund of any overpaid taxes unless it first determines that the taxes were paid:

‘‘within the period which would be applicable under section 6511(b)(2) . . . if on the date of the mailing of the notice of deficiency a claim had been filed (whether or not filed) stating the grounds upon which the Tax Court finds that there is an overpayment.’’ The analysis dictated by § 6512(b)(3)(B) is not elegant, but it is straightforward. Though some courts have adverted to the filing of a ‘‘deemed claim,’’ see Galuska, 5 F. 3d, at 196; Richards, 37 F. 3d, at 589, all that matters for the proper application of § 6512(b)(3)(B) is that the ‘‘claim’’ contemplated in that section be treated as the only mechanism for determining whether a taxpayer can recover a refund. Section 6512(b)(3)(B) defines the lookback period that applies in Tax Court by incorporating the look-back provisions from § 6511(b)(2), and directs the Tax Court to determine the applicable period by inquiring into the timeliness of a hypothetical claim for refund filed ‘‘on the date of the mailing of the notice of deficiency.’’

To this end, § 6512(b)(3)(B) directs the Tax Court’s attention to § 6511(b)(2), which in turn instructs the court to apply either a 3-year or a 2-year lookback period. See §§ 6511(b)(2)(A) and (B) (incorporating by reference § 6511(a)); see supra, at 5. To decide which of these look-back periods to apply, the Tax Court must consult the filing provisions of § 6511(a) and ask whether the claim described by § 6512(b)(3)(B)—a claim filed ‘‘on the date of the mailing of the notice of deficiency’’—would be filed ‘‘within 3 years from the time the return was filed.’’ See § 6511(b)(2)(A) (incorporating by reference § 6511(a)). If a claim filed on the date of the mailing of the notice of deficiency would be filed within that 3-year period, then the lookback period is also three years and the Tax Court has jurisdiction to award a refund of any taxes paid within three years prior to the date of the mailing of the notice of deficiency. §§ 6511(b)(2)(A) and 6512(b)(3)(B). If the claim would not be filed within that 3-year period, then the period for awarding a

refund is only two years. §§ 6511(b)(2)(B) and 6512(b)(3)(B).

In this case, we must determine which of these two look-back periods to apply when the taxpayer fails to file a tax return when it is due, and the Commissioner mails the taxpayer a notice of deficiency before the taxpayer gets around to filing a late return. The Fourth Circuit held that a taxpayer in this situation is entitled to a 3-year lookback period if the taxpayer actually files a timely claim at some point in the litigation, see infra, at 10–11, and respondent offers additional reasons for applying a 3-year look-back period, see infra, at 13–17. We think the proper application of § 6512(b)(3)(B) instead requires that a 2-year look-back period be applied.

We reach this conclusion by following the instructions set out in § 6512(b)(3)(B). The operative question is whether a claim filed ‘‘on the date of the mailing of the notice of deficiency’’ would be filed ‘‘within 3 years from the time the return was filed.’’ See supra, at 7; § 6512(b)(3)(B) (incorporating §§ 6511(b)(2) and 6511(a)). In the case of a taxpayer who does not file a return before the notice of deficiency is mailed, the claim described in § 6512(b)(3)(B) could not be filed ‘‘within 3 years from the time the return was filed.’’ No return having been filed, there is no date from which to measure the 3-year filing period described in § 6511(a). Consequently, the claim contemplated in § 6512(b)(3)(B) would not be filed within the 3-year window described in § 6511(a), and the 3-year look-back period set out in § 6511(b)(2)(A) would not apply. The applicable look-back period is instead the default 2-year period described in § 6511(b)(2)(B), which is measured from the date of the mailing of the notice of deficiency, see § 6512(b)(3)(B). The taxpayer is entitled to a refund of any taxes paid within two years prior to the date of the mailing of the notice of deficiency.

Special rules might apply in some cases, see e.g., § 6511(c) (extension of time by agreement); § 6511(d) (special limitations periods for designated items), but in the case where the taxpayer has filed a timely tax return and the IRS is claiming a deficiency in taxes from that return, the interplay of §§ 6512(b)(3)(B) and 6511(b)(2) generally ensures that the taxpayer can obtain a refund of any taxes against which the IRS is asserting a deficiency. In most cases, the notice of

deficiency must be mailed within three years from the date the tax return is filed. See 26 U. S. C. §§ 6501(a) and 6503(a)(1); Badaracco v. Commissioner, 464 U. S. 386, 389, 392 (1984). Therefore, if the taxpayer has already filed a return (albeit perhaps a faulty one), any claim filed ‘‘on the date of the mailing of the notice of deficiency’’ would necessarily be filed within three years from the date the return is filed. In these circumstances, the applicable look-back period under § 6512(b)(3)(B) would be the 3-year period defined in § 6511(b)(2)(A), and the Tax Court would have jurisdiction to award a refund.

Therefore, in the case of a taxpayer who files a timely tax return, § 6512(b)(3)(B) usually operates to toll the filing period that might otherwise deprive the taxpayer of the opportunity to seek a refund. If a taxpayer contesting the accuracy of a previously filed tax return in Tax Court discovers for the first time during the course of litigation that he is entitled to a refund, the taxpayer can obtain a refund from the Tax Court without first filing a timely claim for refund with the IRS. It does not matter, as it would in district court, see § 7422 (incorporating §§ 6511), that the taxpayer has discovered the entitlement to a refund well after the period for filing a timely refund claim with the IRS has passed, because § 6512(b)(3)(B) applies ‘‘whether or not

[a claim is] filed,’’ and the look-back period is measured from the date of the mailing of the notice of deficiency. Ibid. Nor does it matter, as it might in a refund suit, see 26 CFR § 301.6402– 2(b)(1) (1995), whether the taxpayer has previously apprised the IRS of the precise basis for the refund claim, because 26 U. S. C. § 6512(b)(3)(B) posits the filing of a hypothetical claim ‘‘stating the grounds upon which the Tax Court finds that there is an overpayment,’’ § 6512(b)(3)(B).

Section 6512(b)(3)(B) treats delinquent filers of income tax returns less charitably. Whereas timely filers are virtually assured the opportunity to seek a refund in the event they are drawn into Tax Court litigation, a delinquent filer’s entitlement to a refund in Tax Court depends on the date of the mailing of the notice of deficiency. Section 6512(b)(3)(B) tolls the limitations period, in that it directs the Tax Court to measure the look-back period from the date on which the notice of deficiency is mailed and not the date on which the taxpayer actually files a claim for re

fund. But in the case of delinquent filers, § 6512(b)(3)(B) establishes only a 2-year look-back period, so the delinquent filer is not assured the opportunity to seek a refund in Tax Court: If the notice of deficiency is mailed more than two years after the taxes were paid, the Tax Court lacks jurisdiction to award the taxpayer a refund.

The Tax Court properly applied this 2-year look-back period to Lundy’s case. As of September 26, 1990 (the date the notice was mailed), Lundy had not filed a tax return. Consequently, a claim filed on that date would not be filed within the 3-year period described in § 6511(a), and the 2-year period from § 6511(b)(2)(B) applies. Lundy’s taxes were withheld from his wages, so they are deemed paid on the date his 1987 tax return was due (April 15, 1988), see 26 U. S. C. § 6513(b)(1), which is more than two years prior to the date the notice of deficiency was mailed (September 26, 1990). Lundy is therefore seeking a refund of taxes paid outside the applicable look-back period, and the Tax Court lacks jurisdiction to award such a refund.

III

In deciding Lundy’s case, the Fourth Circuit adopted a different approach to interpreting § 6512(b)(3)(B) and applied a 3-year look-back period. Respondent supports the Fourth Circuit’s rationale, but also offers an argument for applying a uniform 3-year look-back period under § 6512(b)(3)(B). We find neither position persuasive. p1The Fourth Circuit held that:

‘‘[T]he Tax Court, when applying the limitation provision of § 6511(b)(2) in light of § 6512(b)(3)(B), should substitute the date of the mailing of the notice of deficiency for the date on which the taxpayer filed the claim for refund, but only for the purpose of determining the benchmark date for measuring the limitation period and not for the purpose of determining whether the two-year or three-year limitation period applies.’’ 45 F. 3d, at 861. In other words, the Fourth Circuit held that the look-back period is measured from the date of the mailing of the notice of deficiency ( i.e., the taxpayer is entitled to a refund of any taxes paid within either two or three years prior to that date), but that that date is irrelevant in calculating the length of the lookback period itself. The look-back period,

the Fourth Circuit held, must be defined in terms of the date that the taxpayer actually filed a claim for refund. Ibid. (‘‘[T]he three-year limitation period applies because Lundy filed his claim for refund . . . within three years of filing his tax return’’). Thus, under the Fourth Circuit’s view, Lundy was entitled to a 3-year look-back period because Lundy’s late-filed 1987 tax return contained a claim for refund, and that claim was filed within three years from the filing of the return. Ibid. (taxpayer entitled to same look-back period that would apply in district court). Contrary to the Fourth Circuit’s interpretation, the fact that Lundy actually filed a claim for a refund after the date on which the Commissioner mailed the notice of deficiency has no bearing in determining whether the Tax Court has jurisdiction to award Lundy a refund. See supra, at 6. Once a taxpayer files a petition with the Tax Court, the Tax Court has exclusive jurisdiction to determine the existence of a deficiency or to award a refund, see 26 U. S. C. § 6512(a), and the Tax Court’s jurisdiction to award a refund is limited to those circumstances delineated in § 6512(b)(3). Section 6512(b)(3)(C) is the only provision that measures the look-back period based on a refund claim that is actually filed by the taxpayer, and that provision is inapplicable here because it only applies to refund claims filed ‘‘before the date of the mailing of the notice of deficiency.’’ § 6512(b)(3)(C). Under § 6512(b)(3)(B), which is the provision that does apply, the Tax Court is instructed to consider only the timeliness of a claim filed ‘‘on the date of the mailing of the notice of deficiency,’’ not the timeliness of any claim that the taxpayer might actually file. The Fourth Circuit’s rule also leads to a result that Congress could not have intended, in that it subjects the timely, not the delinquent, filer to a shorter limitations period in Tax Court. Under the Fourth Circuit’s rule, the availability of a refund turns entirely on whether the taxpayer has in fact filed a claim for refund with the IRS, because it is the date of actual filing that determines the applicable look-back period under § 6511(b)(2) (and, by incorporation, § 6512(b)(3)(B)). See 45 F. 3d, at 861; see supra, at 11. This rule might ‘‘eliminate[] the inequities resulting’’ from adhering to the 2-year look-back period, 45 F. 3d, at 863, but it creates an even greater inequity in the case of a tax

payer who dutifully files a tax return when it is due, but does not initially claim a refund. We think our interpretation of the statute achieves an appropriate and reasonable result in this case: The taxpayer who files a timely income tax return could obtain a refund in the Tax Court under § 6512(b)(3)(B), without regard to whether the taxpayer has actually filed a timely claim for refund. See supra, at 8–9. If it is the actual filing of a refund claim that determines the length of the lookback period, as the Fourth Circuit held, the filer of a timely income tax return might be out of luck. If the taxpayer does not file a claim for refund with his tax return, and the notice of deficiency arrives shortly before the 3-year period for filing a timely claim expires, see 26 U. S. C. §§ 6511(a) and (b)(1), the taxpayer might not discover his entitlement to a refund until well after the commencement of litigation in the Tax Court. But having filed a timely return, the taxpayer would be precluded by the passage of time from filing an actual claim for refund ‘‘within 3 years from the time the return was filed,’’ as § 6511(b)(2)(A) requires. § 6511(b)(2)(A) (incorporating by reference § 6511(a)). The taxpayer would therefore be entitled only to a refund of taxes paid within two years prior to the mailing of the notice of deficiency. See § 6511(b)(2)(B); 45 F. 3d, at 861–862 (taxpayer entitled to same look-back period as would apply in district court, and look-back period is determined based on date of actual filing). It is unlikely that Congress intended for a taxpayer in Tax Court to be worse off for having filed a timely return, but that result would be compelled under the Fourth Circuit’s approach. Lundy offers an alternative reading of the statute that avoids this unreasonable result, but Lundy’s approach is similarly defective. The main thrust of Lundy’s argument is that the ‘‘claim’’ contemplated in § 6512(b)(3)(B) could be filed ‘‘within 3 years from the time the return was filed,’’ such that the applicable look-back period under § 6512(b)(3)(B) would be three years, if the claim were itself filed on a tax return. Lundy in fact argues that Congress must have intended the claim described in § 6512(b)(3)(B) to be a claim filed on a return, because there is no other way to file a claim for refund with the IRS. Brief for Respondent 28, 30 (citing 26 CFR § 301.6402– 3(a)(1) (1995). Lundy therefore argues that § 6512(b)(3)(B) incorporates a uni

form 3-year look-back period for Tax Court cases: If the taxpayer files a timely return, the notice of deficiency (and the ‘‘claim’’ under § 6512(b)(3)(B)) will necessarily be filed within three years of the return and the lookback period is three years; if the taxpayer does not file a return, then the claim contemplated in § 6512(b)(3)(B) is deemed to be a claim filed with, and thus within three years of, a return and the look-back period is again three years. Like the Fourth Circuit’s approach, Lundy’s reading of the statute has the convenient effect of ensuring that taxpayers in Lundy’s position can almost always obtain a refund if they file in Tax Court, but we are bound by the terms Congress chose to use when it drafted the statute, and we do not think that the term ‘‘claim’’ as it is used in § 6512(b)(3)(B) is susceptible of the interpretation Lundy has given it. The Internal Revenue Code does not define the term ‘‘claim for refund’’ as it is used in § 6512(b)(3)(B), cf. 26 U.S.C. § 6696(e)(2) (‘‘For purposes of section 6694 and 6695 . . . [t]he term ‘claim for refund’ means a claim for refund of, or credit against, any tax imposed by subtitle A’’), but it is apparent from the language of § 6512(b)(3)(B) and the statute as a whole that a claim for refund can be filed separately from a return. Section 6512(b)(3)(B) provides that the Tax Court has jurisdiction to award a refund to the extent the taxpayer would be entitled to a refund ‘‘if on the date of the mailing of the notice of deficiency a claim had been filed.’’ (Emphasis added.) It does not state, as Lundy would have it, that a taxpayer is entitled to a refund if on that date ‘‘a claim and a return had been filed.’’ Perhaps the most compelling evidence that Congress did not intend the term ‘‘claim’’ in § 6512 to mean a ‘‘claim filed on a return’’ is the parallel use of the term ‘‘claim’’ in § 6511(a). Section 6511(a) indicates that a claim for refund is timely if it is ‘‘filed by the taxpayer within 3 years from the time the return was filed,’’ and it plainly contemplates that a claim can be filed even ‘‘if no return was filed.’’ 26 U.S.C. § 6511(a). If a claim could only be filed with a return, as Lundy contends, these provisions of the statute would be senseless, cf. 26 U. S. C. § 6696 (separately defining ‘‘claim for refund’’ and ‘‘return’’), and we have been given no reason to believe that Congress meant the term ‘‘claim’’ to mean one thing in § 6511 but to mean something else

altogether in the very next section of the statute. The interrelationship and close proximity of these provisions of the statute ‘‘presents a classic case for application of the ‘normal rule of statutory construction that identical words used in different parts of the same act are intended to have the same meaning.’ ’’ Sullivan v. Stroop, 496 U. S. 478, 484 (1990) (quoting Sorenson v. Secretary of Treasury, 475 U. S. 851, 860 (1986) (internal quotation marks omitted). The regulation Lundy cites in support of his interpretation, 26 CFR § 301.6402– 3(a)(1) (1995), is consistent with our interpretation of the statute. That regulation states only that a claim must ‘‘[i]n general’’ be filed on a return, ibid., inviting the obvious conclusion that there are some circumstances in which a claim and a return can be filed separately. We have previously recognized that even a claim that does not comply with federal regulations might suffice to toll the limitations periods under the Tax Code, see, e.g., United States v. Kales, 314 U. S. 186, 194 (1941) (‘‘notice fairly advising the Commissioner of the nature of the taxpayer’s claim’’ tolls the limitations period, even if ‘‘it does not comply with formal requirements of the statute and regulations’’), and we must assume that if Congress had intended to require that the ‘‘claim’’ described in § 6512(b)(3)(B) be a ‘‘claim filed on a return,’’ it would have said so explicitly.

Lundy offers two policy-based arguments for applying a 3-year look-back period under § 6512(b)(3)(B). He argues that the application of a 2-year period is contrary to Congress’ broad intent in drafting § 6512(b)(3)(B), which was to preserve, not defeat, a taxpayer’s claim to a refund in Tax Court, and he claims that our interpretation creates an incongruity between the limitations period that applies in Tax Court litigation and the period that would apply in a refund suit filed in district court or the Court of Federal Claims. Even if we were inclined to depart from the plain language of the statute, we would find neither of these arguments persuasive.

Lundy correctly argues that Congress intended § 6512(b)(3)(B) to permit taxpayers to seek a refund in Tax Court in circumstances in which they might otherwise be barred from filing an administrative claim for refund with the IRS. This is in fact the way § 6512(b)(3)(B) operates in a large number of cases. See

supra, at 8–9. But that does not mean that Congress intended that § 6512(b)(3)(B) would always preserve taxpayers’ ability to seek a refund. Indeed, it is apparent from the face of the statute that Congress also intended § 6512(b)(3)(B) to act sometimes as a bar to recovery. To this end, the section incorporates both the 2-year and the 3-year look-back periods from § 6511(b)(2), and we must assume (contrary to Lundy’s reading, which provides a uniform 3-year period, see supra, at 13–14) that Congress intended for both those look-back periods to have some effect. Cf. Badaracco, 464 U. S., at 405 (Stevens, J., dissenting) (‘‘Whatever the correct standard for construing a statute of limitations . . . surely the presumption ought to be that some limitations period is applicable’’). (Emphasis deleted.)

Lundy also suggests that our interpretation of the statute creates a disparity between the limitations period that applies in Tax Court and the periods that apply in refund suits filed in district court or the Court of Federal Claims. In this regard, Lundy argues that the claim for refund he filed with his tax return on December 28 would have been timely for purposes of district court litigation because it was filed ‘‘within three years from the time the return was filed,’’ § 6511(b)(1) (incorporating by reference § 6511(a)); see also Rev. Rul. 76–511, 1976–2 Cum. Bull. 428, and within the 3-year look-back period that would apply under § 6511(b)(2)(A). Petitioner disagrees that there is any disparity, arguing that Lundy’s interpretation of the statute is wrong and that Lundy’s claim for refund would not have been considered timely in district court. See Brief for Petitioner 12, 29–30 and n. 11 (citing Miller v. United States, 38 F. 3d 473, 475 (1994)). We assume without deciding that Lundy is correct, and that a different limitations period would apply in district court, but nonetheless find in this disparity no excuse to change the limitations scheme that Congress has crafted. The rules governing litigation in Tax Court differ in many ways from the rules governing litigation in the district court and the Court of Federal Claims. Some of these differences might make the Tax Court a more favorable forum, while others may not. Compare 26 U. S. C. § 6213(a) (taxpayer can seek relief in Tax Court without first paying an assessment of taxes) with Flora v. United States, 362 U.S. 145, 177 (1960) (28 U.S.C. § 1346(a)(1) requires full

payment of the tax assessment before taxpayer can file a refund suit in district court); and compare 26 U.S.C. § 6512(b)(3)(B) (Tax Court must assume that the taxpayer has filed a claim ‘‘stating the grounds upon which the Tax Court’’ intends to award a refund) with 26 CFR § 301.6402–2(b)(1) (1995) (claim for refund in district court must state grounds for refund with specificity). To the extent our interpretation of § 6512(b)(3)(B) reveals a further distinction between the rules that apply in these fora, it is a distinction compelled

by the statutory language, and it is a distinction Congress could rationally make. As our discussion of § 6512(b)(3)(B) demonstrates, see supra, at 8–9, all a taxpayer need do to preserve the ability to seek a refund in the Tax Court is comply with the law and file a timely return.

We are bound by the language of the statute as it is written, and even if the rule Lundy advocates might ‘‘accor[d] with good policy,’’ we are not at liberty ‘‘to rewrite [the] statute because [we] might deem its effects susceptible of

improvement.’’ Badaracco, 464 U. S., at 398. Applying § 6512(b)(3)(B) as Congress drafted it, we find that the applicable look-back period in this case is two years, measured from the date of the mailing of the notice of deficiency. Accordingly, we find that the Tax Court lacked jurisdiction to award Lundy a refund of his overwithheld taxes. The judgment is reversed.

It is so ordered.