JUSTICE O'CONNOR delivered the opinion of the Court.
These consolidated cases present the question of the applicability of the tax benefit rule to two corporate tax situations: the repayment to the shareholders of taxes for which they were liable, but that were originally paid by the corporation, and the distribution of expensed assets in a corporate liquidation. We conclude that, unless a nonrecognition provision of the Internal Revenue Code prevents it, the tax benefit rule ordinarily applies to require the inclusion of income when events occur that are fundamentally inconsistent with an earlier deduction. Our examination of the provisions granting the deductions and governing the liquidation in these cases leads us to hold that the rule requires the recognition of income in the case of the liquidation, but not in the case of the tax refund.
I
In No. 81-485, Hillsboro National Bank v. Commissioner, the petitioner, Hillsboro National Bank, is an incorporated bank doing business in Illinois. Until 1970, Illinois imposed a property tax on shares held in incorporated banks. Ill.Rev.Stat., ch. 120, § 557 (1971). Banks, required to retain earnings sufficient to cover the taxes, § 558, customarily paid the taxes for the shareholders. Under § 164(e) of the Internal Revenue Code of 1954, 26 U.S.C. § 164(e), [ Footnote 1 ] the bank was allowed a deduction for the amount of the tax, but the shareholders were not. In 1970, Illinois amended its Constitution to prohibit ad valorem taxation of personal property owned by individuals, and the amendment was challenged as a violation of the Equal Protection Clause of the Federal Constitution. The Illinois courts held the amendment unconstitutional in Lake Shore Auto Parts Co. v. Korzen, 49 Ill.2d 137, 273 N.E.2d 592 (1971). We granted certiorari, 405 U.S. 1039 (1972), and, pending disposition of the case here, Illinois enacted a statute providing for the collection of the disputed taxes and the placement of the receipts in escrow. Ill.Rev.Stat., ch. 120, 676.01 (1979). Hillsboro paid the taxes for its shareholders in 1972, taking the deduction permitted by § 164(e), and the authorities placed the receipts in escrow. This Court upheld the state constitutional amendment in Lehnhausen v. Lake Shore Auto Parts Co., 410 U. S. 356 (1973). Accordingly, in 1973, the County Treasurer refunded the amounts in escrow that were attributable to shares held by individuals, along with accrued interest. The Illinois courts held that the refunds belonged to the shareholders, rather than to the banks. See Bank & Trust Co. of Arlington Heights v. Cullerton, 25 Ill.App.3d 721, 726, 324 N.E. 2d 29, 32 (1975) (alternative holding); Lincoln National Bank v. Cullerton, 18 Ill.App.3d 953, 310 N.E.2d 845 (1974). Without consulting Hillsboro, the Treasurer refunded the amounts directly to the individual shareholders. On its return for 1973, Hillsboro recognized no income from this sequence of events. [ Footnote 2 ] The Commissioner assessed a deficiency against Hillsboro, requiring it to include as income the amount paid its shareholders from the escrow. Hillsboro sought a redetermination in the Tax Court, which held that the refund of the taxes, but not the payment of accrued interest, was includible in Hillsboro's income. On appeal, relying on its earlier decision in First Trust and Savings Bank of Taylorville v. United States, 614 F.2d 1142 (1980), the Court of Appeals for the Seventh Circuit affirmed. 641 F.2d 529, 531 (1981).
In No. 81-930, United States v. Bliss Dairy, Inc., the respondent, Bliss Dairy, Inc., was a closely held corporation engaged in the business of operating a dairy. As a cash basis taxpayer, in the taxable year ending June 30, 1973, it deducted upon purchase the full cost of the cattle feed purchased for use in its operations, as permitted by § 162 of the Internal Revenue Code, 26 U.S.C. § 162. [ Footnote 3 ] A substantial portion of the feed was still on hand at the end of the taxable year. On July 2, 1973, two days into the next taxable year, Bliss adopted a plan of liquidation, and, during the month of July, it distributed its assets, including the remaining cattle feed, to the shareholders. Relying on § 336, which shields the corporation from the recognition of gain on the distribution of property to its shareholders on liquidation, [ Footnote 4 ] Bliss reported no income on the transaction. The shareholders continued to operate the dairy business in noncorporate form. They filed an election under § 333 to limit the gain recognized by them on the liquidation, [ Footnote 5 ] and they therefore calculated their basis in the assets received in the distribution as provided in § 334(c). [ Footnote 6 ] Under that provision, their basis in the assets was their basis in their stock in the liquidated corporation, decreased by the amount of money received, and increased by the amount of gain recognized on the transaction. They then allocated that total basis over the assets, as provided in the regulations, Treas.Reg. § 1.334-2, 26 CFR § 1.334-2 (1982), presumably taking a basis greater than zero in the feed, although the amount of the shareholders' basis is not in the record. They in turn deducted their basis in the feed as an expense of doing business under § 162. On audit, the Commissioner challenged the corporation's treatment of the transaction, asserting that Bliss should have taken into income the value of the grain distributed to the shareholders. He therefore increased Bliss' income by $60,000. Bliss paid the resulting assessment and sued for a refund in the District Court for the District of Arizona, where it was stipulated that the grain had a value of $56,565, see Pretrial Order, at 3. Relying on Commissioner v. South Lake Farms, Inc., 324 F.2d 837 (CA9 1963), the District Court rendered a judgment in favor of Bliss. While recognizing authority to the contrary, Tennessee-Carolina Transportation, Inc. v. Commissioner, 582 F.2d 378 (CA6 1978), cert. denied, 440 U.S. 909 (1979), the Court of Appeals saw South Lake Farms as controlling, and affirmed. 645 F.2d 19 (CA9 1981) (per curiam).
II
The Government [ Footnote 7 ] in each case relies solely on the tax benefit rule -a judicially developed principle [ Footnote 8 ] that allays some of the inflexibilities of the annual accounting system. An annual accounting system is a practical necessity if the federal income tax is to produce revenue ascertainable and payable at regular intervals. Burnet v. Sanford & Brooks Co., 282 U. S. 359, 282 U. S. 365 (1931). Nevertheless, strict adherence to an annual accounting system would create transactional inequities. Often an apparently completed transaction will reopen unexpectedly in a subsequent tax year, rendering the initial reporting improper. For instance, if a taxpayer held a note that became apparently uncollectible early in the taxable year, but the debtor made an unexpected financial recovery before the close of the year and paid the debt, the transaction would have no tax consequences for the taxpayer, for the repayment of the principal would be recovery of capital. If, however, the debtor's financial recovery and the resulting repayment took place after the close of the taxable year, the taxpayer would have a deduction for the apparently bad debt in the first year under § 166(a) of the Code, 26 U.S.C. § 166(a). Without the tax benefit rule, the repayment in the second year, representing a return of capital, would not be taxable. The second transaction, then, although economically identical to the first, could, because of the differences in accounting, yield drastically different tax consequences. The Government, by allowing a deduction that it could not have known to be improper at the time, would be foreclosed [ Footnote 9 ] from recouping any of the tax saved because of the improper deduction. [ Footnote 10 ] Recognizing and seeking to avoid the possible distortions of income, [ Footnote 11 ] the courts have long required the taxpayer to recognize the repayment in the second year as income. See, e.g., Estate of Block v. Commissioner, 39 B. T. A. 338 (1939), aff'd sub nom. Union Trust Co. v. Commissioner, 111 F.2d 60 (CA7), cert. denied, 311 U.S. 658 (1940); South Dakota Concrete Products Co. v. Commissioner, 26 B.T.A. 1429 (1932); Plumb, The Tax Benefit Rule Today, 57 Harv.L.Rev. 129, 176, 178, and n. 172 (1943) (hereinafter Plumb). [ Footnote 12 ]
The taxpayers and the Government in these cases propose different formulations of the tax benefit rule. The taxpayers contend that the rule requires the inclusion of amounts recovered in later years, and they do not view the events in these cases as "recoveries." The Government, on the other hand, urges that the tax benefit rule requires the inclusion of amounts previously deducted if later events are inconsistent with the deductions; it insists that no "recovery" is necessary to the application of the rule. Further, it asserts that the events in these cases are inconsistent with the deductions taken by the taxpayers. We are not in complete agreement with either view.
An examination of the purpose and accepted applications of the tax benefit rule reveals that a "recovery" will not always be necessary to invoke the tax benefit rule. The purpose of the rule is not simply to tax "recoveries." On the contrary, it is to approximate the results produced by a tax system based on transactional, rather than annual, accounting. See generally Bittker & Kanner 270; Byrne, The Tax Benefit Rule as Applied to Corporate Liquidations and Contributions to Capital: Recent Developments, 56 Notre Dame Law. 215, 221, 232, (1980); Tye, The Tax Benefit Doctrine Reexamined, 3 Tax L.Rev. 329 (1948) (hereinafter Tye). It has long been accepted that a taxpayer using accrual accounting who accrues and deducts an expense in a tax year before it becomes payable and who for some reason eventually does not have to pay the liability must then take into income the amount of the expense earlier deducted. See, e.g., Mayfair Minerals, Inc. v. Commissioner, 456 F.2d 622 (CA5 1972) (per curiam); Bear Manufacturing Co. v. United States, 430 F.2d 152 (CA7 1970), cert. denied, 400 U.S. 1021 (1971); Haynsworth v. Commissioner, 68 T.C. 703 (1977), affirmance order, 609 F.2d 1007 (CA5 1979); G. M. Standifer Construction Corp. v. Commissioner, 30 B.T.A. 184, 186-187 (1934), petition for review dism'd, 78 F.2d 285 (CA9 1935). The bookkeeping entry canceling the liability, though it increases the balance sheet net worth of the taxpayer, does not fit within any ordinary definition of "recovery." [ Footnote 13 ] Thus, the taxpayers' formulation of the rule neither serves the purposes of the rule nor accurately reflects the cases that establish the rule. Further, the taxpayers' proposal would introduce an undesirable formalism into the application of the tax benefit rule. Lower courts have been able to stretch the definition of "recovery" to include a great variety of events. For instance, in cases of corporate liquidations, courts have viewed the corporation's receipt of its own stock as a "recovery," reasoning that, even though, the instant that the corporation receives the stock, it becomes worthless, the stock has value as it is turned over to the corporation, and that ephemeral value represents a recovery for the corporation. See, e.g., Tennessee-Carolina Transportation, Inc. v. Commissioner, 582 F.2d at 382 (alternative holding). Or payment to another party may be imputed to the taxpayer, giving rise to a recovery. See First Trust and Saving Bank of Tayorville v. United States, 614 F.2d at 1146 (alternative holding). Imposition of a requirement that there be a recovery would, in many cases, simply require the Government to cast its argument in different and unnatural terminology, without adding anything to the analysis. [ Footnote 14 ]
The basic purpose of the tax benefit rule is to achieve rough transactional parity in tax, see n 12, supra, and to protect the Government and the taxpayer from the adverse effects of reporting a transaction on the basis of assumptions that an event in a subsequent year proves to have been erroneous. Such an event, unforeseen at the time of an earlier deduction, may in many cases require the application of the tax benefit rule. We do not, however, agree that this consequence invariably follows. Not every unforeseen event will require the taxpayer to report income in the amount of his earlier deduction. On the contrary, the tax benefit rule will "cancel out" an earlier deduction only when a careful examination shows that the later event is indeed fundamentally inconsistent with the premise on which the deduction was initially based. [ Footnote 15 ] That is, if that event had occurred within the same taxable year, it would have foreclosed the deduction. [ Footnote 16 ] In some cases, a subsequent recovery by the taxpayer will be the only event that would be fundamentally inconsistent with the provision granting the deduction. In such a case, only actual recovery by the taxpayer would justify application of the tax benefit rule. For example, if a calendar-year taxpayer made a rental payment on December 15 for a 30-day lease deductible in the current year under § 162(a)(3), see Treas.Reg. § 1.461-1(a)(1), 26 CFR § 1.461-1(a)(1) (1982); e.g., Zaninovich v. Commissioner, 616 F.2d 429 (CA9 1980), [ Footnote 17 ] the tax benefit rule would not require the recognition of income if the leased premises were destroyed by fire on January 10. The resulting inability of the taxpayer to occupy the building would be an event not fundamentally inconsistent with his prior deduction as an ordinary and necessary business expense under § 162(a). The loss is attributable to the business, [ Footnote 18 ] and therefore is consistent with the deduction of the rental payment as an ordinary and necessary business expense. On the other hand, had the premises not burned and, in January, the taxpayer decided to use them to house his family rather than to continue the operation of his business, he would have converted the leasehold to personal use. This would be an event fundamentally inconsistent with the business use on which the deduction was based. [ Footnote 19 ] In the case of the fire, only if the lessor -by virtue of some provision in the lease -had refunded the rental payment would the taxpayer be required under the tax benefit rule to recognize income on the subsequent destruction of the building. In other words, the subsequent recovery of the previously deducted rental payment would be the only event inconsistent with the provision allowing the deduction. It therefore is evident that the tax benefit rule must be applied on a case-by-case basis. A court must consider the facts and circumstances of each case in the light of the purpose and function of the provisions granting the deductions.
When the later event takes place in the context of a nonrecognition provision of the Code, there will be an inherent tension between the tax benefit rule and the nonrecognition provision. See Putoma Corp. v. Commissioner, 601 F.2d 734, 742 (CA5 1979); id. at 751 (Rubin, J., dissenting); cf. Helvering v. American Dental Co., 318 U. S. 322 (1943) (tension between exclusion of gifts from income and treatment of cancellation of indebtedness as income). We cannot resolve that tension with a blanket rule that the tax benefit rule will always prevail. Instead, we must focus on the particular provisions of the Code at issue in any case. [ Footnote 20 ]
The formulation that we endorse today follows clearly from the long development of the tax benefit rule. JUSTICE STEVENS' assertion that there is no suggestion in the early cases or from the early commentators that the rule could ever be applied in any case that did not involve a physical recovery, post at 460 U. S. 406 -408, is incorrect. The early cases frequently framed the rule in terms consistent with our view and irreconcilable with that of the dissent. See Barnett v. Commissioner, 39 B.T.A. 864, 867 (1939) ("Finally, the present case is analogous to a number of others, where… [w]hen some event occurs which is inconsistent with a deduction taken in a prior year, adjustment may have to be made by reporting a balancing item in income for the year in which the change occurs") (emphasis added); Estate of Block v. Commissioner, 39 B.T.A. at 341 ("When recovery or some other event which is inconsistent with what has been done in the past occurs, adjustment must be made in reporting income for the year in which the change occurs") (emphasis added); South Dakota Concrete Products Co. v. Commissioner, 26 B.T.A. at 1432 ("[W]hen an adjustment occurs which is inconsistent with what has been done in the past in the determination of tax liability, the adjustment should be reflected in reporting income for the year in which it occurs") (emphasis added). [ Footnote 21 ] The reliance of the dissent on the early commentators is equally misplaced, for the articles cited in the dissent, like the early cases, often stated the rule in terms of inconsistent events. [ Footnote 22 ]
Finally, JUSTICE STEVENS' dissent relies heavily on the codification in § 111 of the exclusionary aspect of the tax benefit rule, which requires the taxpayer to include in income only the amount of the deduction that gave rise to a tax benefit, see n 12, supra. That provision does, as the dissent observes, speak of a "recovery." By its terms, it only applies to bad debts, taxes, and delinquency amounts. Yet this Court has held, Dobson v. Commissioner, 320 U. S. 489, 320 U. S. 505 -506 (1943), and it has always been accepted since, [ Footnote 23 ] that § 111 does not limit the application of the exclusionary aspect of the tax benefit rule. On the contrary, it lists a few applications and represents a general endorsement of the exclusionary aspect of the tax benefit rule to other situations within the inclusionary part of the rule. The failure to mention inconsistent events in § 111 no more suggests that they do not trigger the application of the tax benefit rule than the failure to mention the recovery of a capital loss suggests that it does not, see Dobson, supra.
JUSTICE STEVENS also suggests that we err in recognizing transactional equity as the reason for the tax benefit rule. It is difficult to understand why even the clearest recovery should be taxed if not for the concern with transactional equity, see supra at 460 U. S. 377. Nor does the concern with transactional equity entail a change in our approach to the annual accounting system. Although the tax system relies basically on annual accounting, see Burnet v. Sanford & Brooks Co., 282 U.S. at 282 U. S. 365, the tax benefit rule eliminates some of the distortions that would otherwise arise from such a system. See, e.g., Bittker & Kanner 268-270; Tye 350; Plumb 178, and n. 172. The limited nature of the rule and its effect on the annual accounting principle bears repetition: only if the occurrence of the event in the earlier year would have resulted in the disallowance of the deduction can the Commissioner require a compensating recognition of income when the event occurs in the later year. [ Footnote 24 ]
Our approach today is consistent with our decision in Nash v. United States, 398 U. S. 1 (1970). There, we rejected the Government's argument that the tax benefit rule required a taxpayer who incorporated a partnership under § 351 to include in income the amount of the bad debt reserve of the partnership. The Government's theory was that, although § 351 provides that there will be no gain or loss on the transfer of assets to a controlled corporation in such a situation, the partnership had taken bad debt deductions to create the reserve, see § 166(c), and when the partnership terminated, it no longer needed the bad debt reserve. We noted that the receivables were transferred to the corporation along with the bad debt reserve. Id. at 398 U. S. 5, and n. 5. Not only was there no "recovery," id. at 398 U. S. 4, but there was no inconsistent event of any kind. That the fair market value of the receivables was equal to the face amount less the bad debt reserve, ibid., reflected that the reserve, and the deductions that constituted it, were still an accurate estimate of the debts that would ultimately prove uncollectible, and the deduction was therefore completely consistent with the later transfer of the receivables to the incorporated business. See Citizens' Acceptance Corp. v. United States, 320 F.Supp. 798 (Del.1971), rev'd on other grounds, 462 F.2d 751 (CA3 1972); Rev.Rul. 78-279, 1978-2 Cum.Bull. 135; Rev.Rul. 78278, 1978-2 Cum.Bull. 134; see generally O'Hare, Statutory Nonrecognition of Income and the Overriding Principle of the Tax Benefit Rule in the Taxation of Corporations and Shareholders, 27 Tax L.Rev. 215, 219-221 (1972). [ Footnote 25 ]
In the cases currently before us, then, we must undertake an examination of the particular provisions of the Code that govern these transactions to determine whether the deductions taken by the taxpayers were actually inconsistent with later events and whether specific nonrecognition provisions prevail over the principle of the tax benefit rule. [ Footnote 26 ]
III
In Hillsboro, the key provision is § 164(e). [ Footnote 27 ] That section grants the corporation a deduction for taxes imposed on its shareholders but paid by the corporation. It also denies the shareholders any deduction for the tax. In this case, the Commissioner has argued that the refund of the taxes by the State to the shareholders is the equivalent of the payment of a dividend from Hillsboro to its shareholders. If Hillsboro does not recognize income in the amount of the earlier deduction, it will have deducted a dividend. Since the general structure of the corporate tax provisions does not permit deduction of dividends, the Commissioner concludes that the payment to the shareholders must be inconsistent with the original deduction, and therefore requires the inclusion of the amount of the taxes as income under the tax benefit rule.
In evaluating this argument, it is instructive to consider what the tax consequences of the payment of a shareholder tax by the corporation would be without § 164(e), and compare them to the consequences under § 164(e). Without § 164(e), the corporation would not be entitled to a deduction, for the tax is not imposed on it. See Treas.Reg. § 1.164-1(a), 26 CFR § 1.164-1(a) (1982); Wisconsin Gas & Electric Co. v. United States, 322 U. S. 526, 322 U. S. 527 -530 (1944). If the corporation has earnings and profits, the shareholder would have to recognize income in the amount of the taxes, because a payment by a corporation for the benefit of its shareholders is a constructive dividend. See §§ 301(c), 316(a); e.g., Ireland v. United States, 621 F.2d 731, 735 (CA5 1980); B. Bittker & J. Eustice, Federal Income Taxation of Corporations and Shareholders 7.05 (4th ed.1979). The shareholder, however, would be entitled to a deduction, since the constructive dividend is used to satisfy his tax liability. § 164(a)(2). Thus, for the shareholder, the transaction would be a wash: he would recognize the amount of the tax as income, [ Footnote 28 ] but he would have an offsetting deduction for the tax. For the corporation, there would be no tax consequences, for the payment of a dividend gives rise to neither income nor a deduction. 26 U.S.C. § 311(a) (1976 ed., Supp. V).
Under § 164(e), the economics of the transaction of course remain unchanged: the corporation is still satisfying a liability of the shareholder, and is therefore paying a constructive dividend. The tax consequences are, however, significantly different, at least for the corporation. The transaction is still a wash for the shareholder; although § 164(e) denies him the deduction to which he would otherwise be entitled, he need not recognize income on the constructive dividend, Treas.Reg. § 1.164-7, 26 CFR § 1.164-7 (1982). But the corporation is entitled to a deduction that would not otherwise be available. In other words, the only effect of § 164(e) is to permit the corporation to deduct a dividend. Thus, we cannot agree with the Commissioner that, simply because the events here give rise to a deductible dividend, they cannot be consistent with the deduction. In at least some circumstances, a deductible dividend is within the contemplation of the Code. The question we must answer is whether § 164(e) permits a deductible dividend in these circumstances -when the money, though initially paid into the state treasury, ultimately reaches the shareholder -or whether the deductible dividend is available, as the Commissioner urges, only when the money remains in the state treasury, as properly assessed and collected tax revenue.
Rephrased, our question now is whether Congress, in granting this special favor to corporations that paid dividends by satisfying the liability of their shareholders, was concerned with the reason the money was paid out by the corporation or with the use to which it was ultimately put. Since § 164(e) represents a break with the usual rules governing corporate distributions, the structure of the Code does not provide any guidance on the reach of the provision. This Court has described the provision as
prompted by the plight of various banking corporations which paid and voluntarily absorbed the burden of certain local taxes imposed upon their shareholders, but were not permitted to deduct those payments from gross income.
Wisconsin Gas & Electric Co. v. United States, supra, at 322 U. S. 531 (footnote omitted). The section, in substantially similar form, has been part of the Code since the Revenue Act of 1921, 42 Stat. 227. The provision was added by the Senate, but its Committee Report merely mentions the deduction without discussing it, see S.Rep. No. 275, 67th Cong., 1st Sess., 19 (1921). The only discussion of the provision appears to be that between Dr. T. S. Adams and Senator Smoot at the Senate hearings. Dr. Adams' statement explains why the States imposed the property tax on the shareholders and collected it from the banks, but it does not cast much light on the reason for the deduction. Hearings on H.R. 8245 before the Committee on Finance, 67th Cong., 1st Sess., 250-251 (1921) (statement of Dr. T. S. Adams, tax advisor, Treasury Department). Senator Smoot's response, however, is more revealing:
I have been a director in a bank… for over 20 years. They have paid that tax ever since I have owned a share of stock in the bank…. I know nothing about it. I do not take 1 cent of credit for deductions, and the banks are entitled to it. They pay it out.
Id. at 251 (emphasis added).
The payment by the corporations of a liability that Congress knew was not a tax imposed on them [ Footnote 29 ] gave rise to the entitlement to a deduction; Congress was unconcerned that the corporations took a deduction for amounts that did not satisfy their tax liability. It apparently perceived the shareholders and the corporations as independent of one another, each "know[ing] nothing about" the payments by the other. In those circumstances, it is difficult to conclude that Congress intended that the corporation have no deduction if the State turned the tax revenues over to these independent parties. We conclude that the purpose of § 164(e) was to provide relief for corporations making these payments, and the focus of Congress was on the act of payment, rather than on the ultimate use of the funds by the State. As long as the payment itself was not negated by a refund to the corporation, the change in character of the funds in the hands of the State does not require the corporation to recognize income, and we reverse the judgment below. [ Footnote 30 ]
IV
The problem in Bliss is more complicated. Bliss took a deduction under § 162(a), so we must begin by examining that provision. Section 162(a) permits a deduction for the "ordinary and necessary expenses" of carrying on a trade or business. The deduction is predicated on the consumption of the asset in the trade or business. See Treas.Reg. § 1.162-3, 26 CFR § 1.162-3 (1982) ("Taxpayers… should include in expenses the charges for materials and supplies only in the amount that they are actually consumed and used in operation in the taxable year…") (emphasis added). If the taxpayer later sells the asset, rather than consuming it in furtherance of his trade or business, it is quite clear that he would lose his deduction, for the basis of the asset would be zero, see, e.g., Spitalny v. United States, 430 F.2d 195 (CA9 1970), so he would recognize the full amount of the proceeds on sale as gain. See §§ 1001(a), (c). In general, if the taxpayer converts the expensed asset to some other, nonbusiness use, that action is inconsistent with his earlier deduction, and the tax benefit rule would require inclusion in income of the amount of the unwarranted deduction. That nonbusiness use is inconsistent with a deduction for an ordinary and necessary business expense is clear from an examination of the Code. While § 162(a) permits a deduction for ordinary and necessary business expenses, § 262 explicitly denies a deduction for personal expenses. In the 1916 Act, the two provisions were a single section. See § 5(a)(First), 39 Stat. 756. The provision has been uniformly interpreted as providing a deduction only for those expenses attributable to the business of the taxpayer. See, e.g., Kornhauser v. United States, 276 U. S. 145 (1928); Hearings on Proposed Revision of Revenue Laws before the Subcommittee of the House Committee on Ways and Means, 75th Cong., 3d Sess., 54 (1938) ("a taxpayer should be granted a reasonable deduction for the direct expenses he has incurred in connection with his income ") (emphasis added); see generally 1 Bittker, supra, n 9, 20.2. Thus, if a corporation turns expensed assets to the analog of personal consumption, as Bliss did here -distribution to shareholders [ Footnote 31 ] -it would seem that it should take into income the amount of the earlier deduction. [ Footnote 32 ]
That conclusion, however, does not resolve this case, for the distribution by Bliss to its shareholders is governed by a provision of the Code that specifically shields the taxpayer from recognition of gain -§ 336. We must therefore proceed to inquire whether this is the sort of gain that goes unrecognized under § 336. Our examination of the background of § 336 and its place within the framework of tax law convinces us that it does not prevent the application of the tax benefit rule. [ Footnote 33 ]
Section 336 was enacted as part of the 1954 Code. It codified the doctrine of General Utilities Co. v. Helvering, 296 U. S. 200, 296 U. S. 206 (1935), that a corporation does not recognize gain on the distribution of appreciated property to its shareholders. Before the enactment of the statutory provision, the rule was expressed in the regulations, which provided that the corporation would not recognize gain or loss, "however [the assets] may have appreciated or depreciated in value since their acquisition." Treas. Regs. 118, § 39.22(a) 20 (1953) (emphasis added). The Senate Report recognized this regulation as the source of the new § 336, S.Rep. No. 1622, 83d Cong., 2d Sess., 258 (1954). The House Report explained its version of the provision:
Thus, the fact that the property distributed has appreciated or depreciated in value over its adjusted basis to the distributing corporation will in no way alter the application of subsection (a) [providing nonrecognition].
H.R.Rep. No. 1337, 83d Cong., 2d Sess., A90 (19