Donald E. Baker and Barbara M. Baker, Appellants, v. United States of America, Appellee, 460 F.2d 827 (8th Cir. 1972)

The sole issue on appeal concerns the tax treatment of ordinary distributions to corporate shareholders when a corporation with no accumulated earnings has current earnings and profits that are insufficient to cover both ordinary distributions and selective redemption distributions paid out during the fiscal year. The question of first impression to be decided here is whether either distribution is to be given priority as a charge against the current earnings and profits.

The district court held that ordinary dividends are to receive preferential treatment in reducing current earnings and profits. This court is in full agreement that the district court was correct in rejecting the taxpayers' argument that would result in a priority for redemption distributions in reducing current earnings and profits. Judges Gibson and Bright, for differing reasons, find that the Internal Revenue Code requires priority treatment be given ordinary distributions. Their separate opinions are appended. Judge Lay finds that neither distribution should be given a preference and that a pro rata reduction of the current earnings and profits account is required. The trial court is thus affirmed by a divided vote.

Taxpayers, husband and wife, reported as ordinary income the full amount of ordinary distributions ($4,025) with respect to their no par common stock received in January and May 1961 from Peter Kiewit Sons, Inc. The corporation subsequently notified shareholders that a portion of the fiscal 1961 ordinary distributions was from sources other than earnings and profits. After the government denied their claim for refund, plaintiffs instituted the present refund suit. The stipulated facts show that in fiscal 1961 Peter Kiewit Sons, Inc., redeemed part of its issued and outstanding no par common stock from certain shareholders (not including taxpayers) at a cost of $1,651,561.70.¹  During fiscal 1961, Kiewit also distributed $1,880,179.75 cash to its shareholders as distributions with respect to stock. This sum included the distribution of $4,025 to plaintiffs. During 1961 the corporation had no accumulated earnings and had current earnings of only $1,553,636.34. It is stipulated that of the $1,651,561.70 paid out as redemptive distributions $157,002.57 is properly chargeable to Kiewit's capital account.²  The remainder, $1,494,550.13, was charged by Kiewit to its earnings and profits account. On this basis taxpayers urge that the balance of the earnings and profits account, i. e. $59,077.21, is all that remained available for the ordinary distribution since the redemption distributions had otherwise absorbed the earnings. Thus, it is urged the portion of their distribution which is taxable as ordinary income is limited to their share of that remainder. The government urges that the ordinary distributions ($1,880,179.75) must be given priority and charged first against the earnings and profits account ($1,553,636.34) leaving only the difference, if any, available for the redemption distributions.³  The district court, the Honorable Richard E. Robinson, affirmed the government's position and denied plaintiff's relief. D.C., 308 F. Supp. 1129. This appeal followed.

Taxpayers urge three basic arguments to support reversal of the district court: (1) that the court erred in its statutory construction of Sec. 316(a) (2) I.R.C., (2) that the legislative history of Sec. 316(a) (2) supports taxpayers' interpretation of the statute, and (3) that the government's theory of priority destroys the harmony between Sec. 312(a) and (e) and Sec. 316. We will discuss these arguments seriatim.

Section 316 reads:

"(a) General rule.-For purposes of this subtitle, the term 'dividend' means any distribution of property made by a corporation to its shareholders-

"(1) out of its earnings and profits accumulated after February 28, 1913, or

"(2) out of its earnings and profits of the taxable year (computed as of the close of the taxable year without diminution by reason of any distributions made during the taxable year), without regard to the amount of the earnings and profits at the time the distribution was made." (Emphasis added).

Taxpayers urge that the words "any distributions" as used in the parenthetical clause of Sec. 316(a) (2) are limited to ordinary distributions and do not encompass redemptive distributions which allegedly reduce earnings and profits under Sec. 312(a) on the date of distribution. Taxpayers urge that Hellmich v. Hellman, 276 U.S. 233, 48 S. Ct. 244, 72 L. Ed. 544 (1928), controls the present construction of Sec. 316(a) (2), excluding redemptive distributions from the statutory language. It is urged that Hellmich found the words "any distribution" as used within the definition clause of the forerunner of Sec. 316(a) do not include redemptive distributions. Consequently, the words cannot take on a broader and different meaning in the parenthetical clause of Sec. 316(a) (2).

Prior to 1918 both ordinary distributions with respect to stock and liquidating or redemption distributions were included within the broad statutory definition of a "dividend" which encompassed "any distribution" from accumulated earnings and profits. The absence of any separate treatment of liquidating distributions resulted in the same tax treatment for all distributions from earnings.

In the Revenue Act of 1918 this situation was altered. While maintaining the same apparently all-inclusive definition of a dividend, Congress for the first time changed the statutory scheme by establishing a separate subsection which entitled a liquidating distribution to different tax treatment. At the same time, Congress altered the taxing provisions by imposing both a normal tax and a surtax on taxable income. However, a credit against the normal tax was provided for dividends received so that distributions qualifying as "dividends" would be subject only to the surtax. Thus, it was advantageous for taxpayers receiving liquidating distributions to claim that these distributions were entitled to the dividend received credit since they were clearly encompassed within the meaning of "any distribution" as used in the dividend definition.

In Hellmich v. Hellman, 276 U.S. 233, 48 S. Ct. 244, 72 L. Ed. 544 (1928), the Supreme Court was faced with the question of whether the dividend definition included liquidating distributions. The Court found that the existence of a separate section which dealt specifically with liquidating distributions removed this type of distribution from the broad general definition of a dividend. The language used by the Court is as follows:

"The controlling question is whether the amounts distributed to the stockholders out of the earnings and profits accumulated by the corporation since February 28, 1913, were to be treated under Sec. 201(a) as 'dividends,' which were exempt from the normal tax; or, under Sec. 201(c) as payments made by the corporation in exchange for its stock, which were taxable 'as other gains or profits.'

"It is true that if Sec. 201(a) stood alone its broad definition of the term 'dividend' would apparently include distributions made to stockholders in the liquidation of a corporation-although this term, as generally understood and used, refers to the recurrent return upon stock paid to stockholders by a going corporation in the ordinary course of business, which does not reduce their stock holdings and leaves them in a position to enjoy future returns upon the same stock. See Lynch v. Hornby, 247 U.S. 339, 344-346 [, 38 S. Ct. 543, 62 L. Ed. 1149]; and Langstaff v. Lucas (D.C.), 9 F.2d 691, 694.

"However, when Sec. 201(a) and Sec. 201 (c) are read together, under the longestablished rule that the intention of the lawmaker is to be deduced from a view of every material part of the statute, Kohlsaat v. Murphy, 96 U.S. 153, 159, [24 L. Ed. 844], we think it clear that the general definition of a dividend in Sec. 201(a) was not intended to apply to distributions made to stockholders in the liquidation of a corporation, but that it was intended that such distributions should be governed by Sec. 201(c), which, dealing specifically with such liquidation, provided that the amounts distributed should 'be treated as payments in exchange for stock' and that any gain realized thereby should be taxed to the stockholders 'as other gains or profits.' This brings the two sections into entire harmony, and gives to each its natural meaning and due effect." 276 U.S. at 236-237, 48 S. Ct. at 245 (emphasis added).

The taxpayers point out that the Hellmich reasoning is grounded on the fact that the existence of a specific section dealing with liquidating distributons removes these types of distributions from the scope of the general definition of a dividend. Thus, it is argued, when Congress in the Revenue Act of 1924 and in all subsequent acts, including the present Internal Revenue Code of 1954, set out a separate section dealing with redemption distributions, this type of distribution was not intended to fall within the meaning of a dividend as defined in the section dealing with that subject. We agree that the Hellmich decision must still be regarded as effective in compelling the conclusion that a redemption distribution is not within the meaning of a dividend. The cases following Hellmich bear out this conclusion. See Foster v. United States, 303 U.S. 118, 58 S. Ct. 424, 82 L. Ed. 700 (1938); Henry B. Babson, 27 BTA 859 (1933), aff'd, 70 F.2d 304 (7th Cir. 1934), cert. denied, Helvering v. Babson, 293 U.S. 571, 55 S. Ct. 107, 79 L. Ed. 669; Central & Southwest Corp. v. Brown, 249 F. Supp. 787 (D. Del. 1965); Atlantic City Electric Co. v. United States, 161 F. Supp. 811, 142 Ct. Cl. 519, cert. denied, 358 U.S. 834, 79 S. Ct. 56, 3 L. Ed. 2d 71 (1958); cf. Associated Tel. & Tel. Co. v. United States, 306 F.2d 824, 832 (2d Cir. 1962), cert. denied, 371 U.S. 950, 83 S. Ct. 504, 9 L. Ed. 2d 498 (1963); and Cummins Diesel Sales Corp. v. United States, 323 F. Supp. 1114, 1118 (S.D. Ind. 1971).

It is argued further, however, that when Hellmich determined that liquidating distributions were outside the dividend definition, the Court actually determined that the words "any distribution" as used in the dividend definition meant "any ordinary distribution." Carrying this argument one step further, the taxpayer claims that thereafter, in 1936, when Congress enacted the parenthetical clause of Sec. 316(a) (2)⁴  as part of the same sentence using these same two words, the same meaning was intended. As a result, according to taxpayers, the proper meaning of the parenthetical clause is that in computing current earnings available for dividend distributions, the total earnings for the taxable year are to be considered without diminution by any ordinary distribution, but reductions for distributions which are not ordinary are to be taken into account.

We disagree that Hellmich is controlling authority for the issue confronting us. The Supreme Court in Hellmich did not give judicial construction to the generic term "distribution" within the Internal Revenue Code. The Court simply held that the definition of a dividend under the forerunner to Sec. 316(a) did not include redemptive and liquidating distributions. The reason was plain enough: another special section governed their treatment. So it is here. Section 302 specifically governs the tax treatment of redemption distributions and removes these distributions from the definition of a dividend under Sec. 316(a). However, it does not follow that this separate treatment gives a special meaning to the generic words "any distribution" used elsewhere in Sec. 316(a) (2). First, and foremost, the words "any distribution" are plain and do not convey any technical meaning. It would be a strained rule of statutory construction to construe words of general meaning as specifying a particular meaning. The effect of such a construction would be to include language within the statute which Congress failed to insert in it. We literally would be amending the act and usurping the function of the legislature. It is axiomatic courts are not at liberty to modify plain words by judicial construction.

Furthermore, if we accept taxpayer's interpretation and read these words consistently throughout the entire section, the third sentence of Sec. 316(a) would in effect negate the purpose of Sec. 302(a). Section 302(a) accords "exchange" treatment to only those redemption distributions which meet certain qualifications. Section 302(d) subjects those redemptions which fail to qualify under (a) to Sec. 301 treatment. The third sentence of Sec. 316(a) extends dividend treatment to these non-qualifying redemptions by concluding: "To the extent that any distribution is, under any provision of this subchapter, treated as a distribution of property to which section 301 applies, such distribution shall be treated as a distribution of property for purposes of this subsection." (Emphasis added.) A redemption distribution, whether qualifying or non-qualifying, is not an ordinary distribution. Thus, non-qualifying redemptions are given ordinary dividend treatment only by reason of this sentence. If the above words "any distribution" were read to mean "any ordinary distribution," then the dividend treatment which is accorded to non-qualifying redemption distributions would be totally negated. As a result of the non-qualifying redemption would still be given Sec. 301 treatment because of the dictates of Sec. 302(d), but because dividend status is denied it must be accorded the treatment set out in Sec. 301(c) (2) and (3) which is "exchange" treatment. Hence, the purpose of Sec. 302 (a) in extending "exchange" treatment to only those redemptions which meet certain qualifications would thereby be entirely frustrated.

This court has uniformly refused to interpret taxing statutes by finding they include language by implication. In Helvering v. Alworth Trust, 136 F.2d 812, 814 (8th Cir.), cert. denied, Alworth v. Commissioner of Internal Revenue, 320 U.S. 784, 64 S. Ct. 193, 88 L. Ed. 471 (1943), we stated:

"The taxpayer's theory grafts upon the statute a condition or limitation based upon the assumption that Congress in providing that a taxable dividend 'means any distribution made by corporation to its shareholders * * * out of the earnings or profits of the taxable year' implied the condition: 'if there are such earnings or profits remaining after deducting federal taxes for the taxable year.' To read into the statute such an implied condition is equivalent to amending it; and neither the Board nor this court can so amend the Act. Congress alone has power to do so."

See also, Foley Securities Corp. v. Commissioner of Internal Revenue, 106 F.2d 731, 734 (8th Cir. 1939).

We conclude that Congress intended all types of distributions to be included within the computation set forth in Sec. 316(a) (2). We should be persuaded otherwise only if legislative history convinces us that the clause must be read with a more restricted meaning. Gellman v. United States, 235 F.2d 87 (8th Cir. 1956); Kelm v. Chicago, St. P., M. & O. Ry. Co., 206 F.2d 831, 834 (8th Cir. 1953);⁵  Helvering v. Rebsamen Motors, Inc., 128 F.2d 584, 587 (8th Cir. 1942).

II. LEGISLATIVE HISTORY.

The forerunner of Sec. 316(a) (2) was enacted as Sec. 115(a) (2) in the Revenue Act of 1936. In that year Congress introduced the undistributed profits tax which was intended to assist the nation's economy by inducing the distribution of corporate earnings. The inducement was created by allowing a corporation to reduce the new tax on its profits by a credit for dividends paid throughout the taxable year. "Dividends," however, as formerly defined could only be distributed from accumulated earnings and profits. Earnings were not deemed to be "accumulated" until the impairment of capital caused by past operating losses had been restored by subsequent earnings. Hence, a corporation whose current earnings did not exceed its accumulated losses had no "accumulated earnings" and could not receive the dividends paid credit even though it may have distributed the whole of its current earnings. To prevent this inequity the 1936 Act enlarged the dividend definition to include distributions from current earnings without regard to the state of the past earnings account. This change enabled the deficit corporation to obtain the credit for ordinary distributions made from its current earnings despite the fact that its existing deficit exceeded the amount of those earnings thereby preventing the existence of "accumulated earnings."

The change, however, did not affect liquidating distributions since such distributions were not within the scope of the dividend section. Thus, in order for liquidating distributions to qualify for a dividends paid credit specific statutory authority was needed. Section 27(f) of the 1936 Act dealt specifically with this credit:

"In the case of amounts distributed in liquidation the part of such distribution which is properly chargeable to the earnings or profits accumulated after February 28, 1913, shall, for the purposes of computing the dividends paid credit under this section, be treated as a taxable dividend paid."

See Helvering v. Credit Alliance Co., 316 U.S. 107, 110-111, 62 S. Ct. 989, 86 L. Ed. 1307 (1941).

Thereafter a number of cases dealt with the question of whether a dividends paid credit under Sec. 27(f) could be allowed for a liquidating distribution which was made in a year when a corporation had an accumulated deficit that was not absorbed by current earnings. Although some courts allowed the credit even though the deficit was not absorbed by current earnings, cf. Pembroke Realty & Sec. Corp. v. Commissioner of Internal Revenue, 122 F.2d 252 (2d Cir. 1941), the vast majority of decisions agreed with the Internal Revenue's position that the portion of such distribution which qualified for the credit was limited to the amount by which current earnings exceeded the existing deficit. These excess earnings were the only earnings which qualified as "accumulated." See e. g., St. Louis Co. v. United States, 237 F.2d 151 (3d Cir. 1956), cert. denied, 352 U.S. 1001, 77 S. Ct. 558, 1 L. Ed. 2d 546 (1957). See also, Van Norman Co. v. Welch, 141 F.2d 99 (1st Cir. 1944); Shellabarger Grain Products Co. v. Commissioner of Internal Revenue, 146 F.2d 177 (7th Cir. 1944).

Taxpayers rely upon this history to support their argument that the words "any distribution" under the parenthetical clause of Sec. 316(a) (2) do not contemplate redemptive distributions. Taxpayers urge (1) the forerunner of Sec. 316(a) (2) was enacted to enable deficit corporations to obtain the dividends paid credit by extending the definition of "dividend" to include distributions from current earnings and profits; (2) the method provided in this new section for computing current earnings required that the total earnings for the year be considered without diminution by "any distribution" made during the year; (3) the St. Louis case, supra, proves that a deficit corporation could never diminish current earnings by liquidating or redemption distributions since current earnings were "simply not available as a source" for such distributions; (4) therefore, since the only distributions which a deficit corporation could make from current earnings were ordinary distributions, Congress could not have been referring to liquidating distributions when it used the words "any distribution" in the parenthetical clause.

We disagree with this analysis. The very issue as stated in the St. Louis case was whether "a distribution out of such income [the deficit corporation's current earnings], * * * in complete liquidation" qualified for the credit. 237 F.2d at 152. The fact that the distribution was made out of current earnings was not denied; the Court's only concern was whether such a distribution met the express standards of Sec. 27(f). It is clear that these cases do not hold that current earnings of a deficit corporation are unavailable for liquidating distributions, as the taxpayers claim. These cases only stand for the proposition that the requirements of Sec. 27(f) are not met if the current earnings from which such a distribution may have been made do not exceed the deficit.

To say that current earnings cannot be utilized as a source of liquidating distributions when a corporation is in a deficit status ignores the fact that these earnings, while the deficit exists, are actually considered for all purposes except dividends to be capital⁶  and that a liquidating distribution must be charged to capital in an amount which is "properly chargeable to capital." (Sec. 312(e), formerly the third sentence of Sec. 115(c) of the 1939 Code.) In Helvering v. Jarvis, 123 F.2d 742 (4th Cir. 1941), it was held that the amount "properly chargeable to capital" is the original amount received for the corporation's capital stock, comprising both par value and the paid-in surplus. When a corporation is in a deficit position, both the par value and the paid-in surplus are impaired by the accumulated losses. Thus, so long as current earnings are treated as restoring the impairment, they are actually serving as capital. Therefore when a liquidating distribution is made which must be allocated to capital, it is not at all improper for such a distribution to be charged against the current earnings in compliance with Sec. 312(e). Cf. Shellabarger Grain Products Co. v. Commissioner of Internal Revenue, 2 T.C. 75, 82, aff'd, 146 F.2d 177 (7th Cir. 1944).

Considering taxpayers' argument in light of the above, it can only be concluded that in legislating the parenthetical clause of Sec. 316(a) (2) Congress was fully aware that a redemption distribution could be charged to current earnings of a deficit corporation and thus, when it used the phrase "any distribution" Congress intended to prevent both liquidating distributions and ordinary distributions from diminishing current earnings in the year-end computation.

We also note that Sec. 316(a) (2) clearly applies to all types of corporations, deficit and non-deficit. Thus, even if taxpayers were correct in their claim that a deficit corporation cannot make a liquidating distribution from current earnings, the meaning of the phrase "any distribution" must still be taken to embrace all those types of distributions which a non-deficit corporation may make from current earnings. To limit the interpretation of this phrase to only those types of distributions which a deficit corporation can make while at the same time recognizing that the entire clause has application to non-deficit corporations would place a strained construction upon the statute. Thus, since Sec. 316(a) (2) applies to all types of corporations the only logical conclusion is that Congress intended the phrase to apply to all types of distributions.

The Senate report accompanying the enactment of the forerunner of Sec. 316(a) (2) clearly states two purposes for enacting this particular clause: (1) to make available a new source for dividend distribution and (2) to simplify the computation of the available source. S.Rep. 2156, 74th Cong. 2d Sess., p. 18 (1936).

We conclude these two purposes can only be effectuated by reading Sec. 316 (a) (2) to mean that all distributions which diminish current earnings during the year are to be disregarded in computing the amount of earnings available for dividend distributions. As the district court observed, the computation would not be simple if redemption distributions are to be excluded from this computation. This is so since the portion of a redemptive distribution which would be allocated to earnings could only be known by making a preliminary determination of the portion of such distribution which under Sec. 312(e) must first be charged to capital,⁷  and then prorating those earnings up to the dates of redemptive distributions. These additional calculations only serve to complicate the application of this section. We thus conclude that taxpayers' construction was not only not intended but would in fact frustrate this legislative purpose.

Having resolved that under Sec. 316(a) (2) current earnings and profits are to be computed without diminution by reason of redemption distributions, we now face the question of whether the description of such a computation in the dividend section necessarily gives dividends a priority in the absorption of these earnings. The taxpayers argue that if such a priority is recognized, a disharmony will be created between Sec. 316 and Sec. 312(a) as it [Sec. 312(a)] applies to redemption distributions. The government, however, contends that once the above interpretation of the parenthetical clause has been accepted, a priority for dividends is the necessary consequence. This contention is correct if redemption distributions were not intended to play an equal role with dividends in reducing current earnings at the time of distribution.

As indicated, Judges Gibson and Bright with differing reasons, which will now be separately set forth, find that priority treatment is to be given ordinary distributions in reducing current earnings and profits. The judgment of the district court is therefore affirmed.

GIBSON, Circuit Judge.

That this case presents an extremely difficult issue is demonstrated by the fact that none of the six participants in this suit-taxpayers, the Government, the district court, and three Court of Appeals judges-can agree on an appropriate solution to the problems raised. The case vividly demonstrates the necessity for serious Congressional attention to the problem, and it is hoped that it may soon be forthcoming.

The difficulties of this case are compounded by the inadequacies of the stipulation upon which the case was tried. According to the stipulation, the corporation made total distributions-redemptive and ordinary-of more than $3,530,000. It had earnings and profits of only $1,553,000, and the distributions were "properly chargeable to capital" in only the amount of $157,000. Thus we are left to speculate as to the source of more than $1,820,000 used to make these distributions. Assuming that the corporation's directors were not unlawfully invading the capital of the corporation, it is clear that some funds were available for these distributions which are not revealed by the record. Thus both the taxpayer and the Government have put this Court in the position of deciding a question of considerable importance to the federal income tax program as a the-oretical, abstract question. This question should not be decided in the absence of adequate knowledge of the facts of the transactions as they actually occurred; the proper solution to the problem depends at least in part on the realities of corporate financing and accounting. Nevertheless, feeling constrained by the stipulation agreed to by both the parties and the time already spent at both the trial and appellate level on this case, I turn to the merits as I perceive them on the basis of this record.

In analyzing this problem, the practical result of the taxpayer's position should be kept in mind. According to the taxpayer's analysis of these transactions, the redemptive distributions exhausted all but $59,000 of the corporation's earnings and profits. The redemptive distributions are admittedly taxable only at capital gains rates. Only $59,000 of the $1,880,000 ordinary distribution would be deemed to be out of earnings and profits, the rest being considered a return of capital, and thus 97 per cent of the ordinary distributions would escape any taxation whatsoever. I cannot ascribe to such an absurd result in the absence of any indication that Congress intended such a result. The result under Judge Lay's approach would be that roughly 50 per cent of the ordinary distributions would escape taxation, and I view this result as equally objectionable.

In my view, the principal difficulty in this case is caused by the attempt to transform what is essentially a capital transaction-i. e., the redemptive distributions-into an ordinary income transaction of the corporation. Under ordinarily accepted accounting procedures, a corporation's transactions in its own stock are not reflected in its income statement, which reflects the potential availability of dividends to stockholders. While admittedly the income statement as a matter of accepted corporate accounting does not correspond exactly to the earnings and profits account for purposes of tax accounting, the two are closely enough related that there should be some semblance of similarity in their treatment. Under this view, I cannot accept the proposition that redemptive distributions can be allowed to so exhaust the corporation's earnings and profits as to render subsequent dividend distributions nontaxable.

I find support for this general approach in the Supreme Court case of Foster v. United States, 303 U.S. 118, 58 S. Ct. 424, 82 L. Ed. 700 (1938), the most recent Supreme Court case to deal with any aspect of the problem before us. In that case, the taxpayers argued that redemptive distributions had exhausted post-1913 earnings and profits, so that subsequent dividend distributions must have been deemed to have come from pre-1913 earnings and profits, and thus were tax-free. The Supreme Court disposed of this argument with the following observation:

"We have previously said that subsections (a) and (b), supra, [Sec. 115 (a) and (b) of the Revenue Act of 1928, now Sec. 316(a) of the Internal Revenue Code] construed together, disclose legislative purpose that pre-1913 accumulations shall not be distributed 'in such a fashion as to permit profits accumulated after that date to escape taxation.' Petitioners ask that we now construe these provisions in a way which would facilitate the escape of such profits from taxation and thereby defeat the undoubted purpose of Congress. We are urged so to expand and broaden an exemption granted by Congress as a 'concession to the equity of stockholders' that such concession would in reality serve to nullify and defeat the tax on corporate profits earned after 1913. Courts should construe laws in harmony with the legislative intent and seek to carry out legislative purpose. With respect to the tax provisions under consideration, there is no uncertainty as to the legislative purpose to tax post-1913 corporate earnings. We must not give effect to any contrivance which would defeat a tax Congress plainly intended to impose. The use of bookkeeping terms and accounting forms and devices cannot be permitted to devitalize valid tax laws." 303 U.S. at 120-121, 58 S. Ct. at 425 (footnotes omitted) (emphasis added).

While the facts here are somewhat different, I find this language peculiarly appropriate to the instant case. Adopting the taxpayer's position would result in the use of bookkeeping terms and accounting forms and devices which would devitalize the tax laws and permit the escape of dividends from a tax Congress clearly intended to impose. With this in mind, I turn to the specific issues in the case.

The first question is the applicability of Sec. 312(a). Both the taxpayer and the government have assumed that Sec. 312(a) is applicable to redemption distributions and have presented various arguments relevant to its interpretation. The only reported case I have found to have considered this issue, Bennett v. United States, 427 F.2d 1202 (Ct. Cl. 1970), also assumed that Sec. 312(a) was applicable to redemption distributions. However, that assumption was not necessary to the result in that case, for after application of the rule of Foster v. United States, 303 U.S. 118, 58 S. Ct. 424, 82 L. Ed. 700 (1938), there was no excess of the redemption distribution to be applied to earnings and profits. After consideration of the language of the section and its legislative history, I have concluded that Sec. 312(a) applies only to ordinary distributions and not to redemptive distributions, and that only Sec. 312(e) applies to the instant case.

The pertinent provisions of 26 U.S.C. § 312(a) are:

"General rule.-Except as otherwise provided in this section, on the distribution of property by a corporation with respect to its stock, the earnings and profits of the corporation (to the extent thereof) shall be decreased by the sum of-

"(1) the amount of money,

"(2) the principal amount of the obligations of such corporation, and

"(3) the adjusted basis of the other property, so distributed." (emphasis added.)

It is to be specifically noted that the section applies to a "distribution with respect to its stock." If this language is given its usual meaning, it applies only to ordinary or dividend distributions; a redemptive distribution is not usually referred to as a distribution with respect to stock. Actually, a redemptive liquidation is a repurchase of an equity interest and is in no sense a dividend.

The legislative history seems to bear out this interpretation of Sec. 312(a) as applying only to ordinary distributions. While the section does refer to decreasing earnings and profits by the "amount of money * * * so distributed," it is clear from the congressional reports that the section was primarily aimed at defining the tax accounting consequences to be given to distributions of property as dividends. Thus, in explaining the effect of the new section, the House report gives examples of corporate distributions of property having a fair market value both above and below its adjusted basis in the hands of the corporation. The report then states:

"In either case, the determination of the amount of earnings and profits, for the purpose of ascertaining the extent to which such distribution would constitute a dividend * * * shall be made by reference to the fair market value of the property received by the shareholder and the earnings and profits of the corporation at the time of the distribution." 3 U.S.Code Cong. & Admin.News, 83d Cong., 2d Sess., 1954, p. 4232 (emphasis added).

Furthermore, when the House proposed what has become Sec. 312 of the Act, it included a specific section which was intended to prescribe the rules for corporate accounting in redemptive and liquidating distributions. This seems clearly to imply that since a special section was included for redemptive distributions, Sec. 312(a) was not intended to apply to them. This section proposed by the House was rejected by the Senate in favor of what became Sec. 312(e), which was merely a re-enactment of existing law, with the following comment:

"The House bill supplied an additional rule for the determination of the manner in which earnings and profits should be allocated where there was a partial liquidation, a corporate separation, or a redemption. Your committee strikes this rule since it is believed that existing administrative practice, making these determinations as the facts of each case may indicate, has been successful in achieving correct results." Id. at 4678.

This comment also seems to imply that the Senate did not assume that Sec. 312(a) had application in determining the accounting results of a redemptive distribution.

The commentators also seem to agree that Sec. 312(a) does not apply to liquidating distributions. Thus, Mertens, in discussing Sec. 312 in general, points out that, "It should be noted that the various 'earnings and profits' adjustments, indicated in Section 312, involve entirely unrelated items having in common the single factor that each may have special 'earnings and profits' consequences." Mertens, Law of Federal Income Taxation, Code Commentary, Ch. 1, Subch. C-p. 95 (1968) (emphasis added). In discussing specifically Sec. 312 (a), Mertens notes:

"The 1954 Code introduced into the revenue laws detailed rules for adjustments to 'earnings and profits' where there is a nonliquidating distribution of property by a corporation with respect to its stock." Id. at 96 (emphasis added).

An article which contains an exhaustive analysis of the impact of liquidating distributions on earnings and profits also suggests that only Sec. 312(e) is applicable to the problem, and not Sec. 312(a):

"Since the earnings and profits account serves as a determinant of the tax consequences arising upon receipt of a corporate distribution, it is logical to suppose that once this account has carried out its function of categorizing a particular distribution with the taint of 'dividend,' it is to be reduced pro tanto. This is, in fact, the result prescribed by the general rule found in section 312(a).

"But what is the proper tax accounting to follow when by force of statutory exception the sting of dividend treatment has been removed from the receipt of a distribution on the aforementioned ground of its 'terminal' character? * * *

"* * * If a terminal distribution were treated in its entirety as reducing earnings and profits, then to the extent of the reduction the remaining shareholders could thereafter bail out corporate earnings by way of non-terminal distributions without the imposition of a dividend tax. On the other hand, if a terminal distribution were treated as having no effect at all on earnings and profits, then to that extent it would enhance the likelihood of the remaining shareholders bearing a dividend tax on subsequent non-terminal distributions.

"The statutory solution to this problem is found in section 312(e) * * *." Edelstein & Korbel, "The Impact of Redemption and Liquidation Distributions on Earnings and Profits: Tax Accounting Aberrations Under Section 312(e)," 20 Tax L.Rev. 479, 480-481 (1965) (footnotes omitted) (emphasis added).

I also note that Congress appears to have been consistent throughout the Revenue Act in referring to ordinary distributions as "distributions with respect to stock"-see, e. g., Sec. 305(a) (stock dividends)-while referring to terminal distributions as distributions in liquidation, partial liquidation, or redemption-see, e. g., Secs. 303, 304. On this point the legislative history of Sec. 311(a) is interesting. Section 311(a) of the 1954 Act provided the long-standing rule that "no gain or loss shall be recognized to a corporation on the distribution, with respect to its stock" of its stock or property. In implementing this section, the Treasury provided in its regulations that, "The term 'distributions with respect to its stock' includes distributions made in redemption of stock * * *." Regs., Sec. 1.311-1(a). The very fact that the Treasury found it necessary to specify such a definition of the term suggests that that is not its ordinary meaning. Moreover, when the Treasury moved to implement its regulation, by Revenue Ruling 68-21 which provided that in a redemption of stock with appreciated property no gain or loss was recognized to the redeeming corporation, Congress moved promptly in 1969 to amend the statute and reverse the ruling,¹  Thus indicating congressional disapproval of that administrative interpretation. Also, I have serious doubts as to the correctness of this regulation, since it appears to be in clear conflict with the statutory language and it is a well-recognized principle that the Treasury's regulations cannot amend the statute. However, in any event, the regulation applies only to Sec. 311(a) and has no applicability to Sec. 312(a).² 

To summarize then, I conclude that Sec. 312(a) has no application to the problem before us, and the case must be resolved on the basis of the interplay between Sec. 316(a) (2) and Sec. 312(e).

I now consider the legislative history of Sec. 312(e). This section originated as the last sentence of Sec. 201(c) of the Revenue Act of 1924; Sec. 201(c) accorded exchange treatment to liquidating distributions, which previously had been given dividend treatment. The last sentence was a new provision which read in pertinent part as follows:

"In the case of amounts distributed in partial liquidation * * * the part of such distribution which is properly chargeable to capital account shall not be considered a distribution of earnings or profits within the meaning of subdivision (b) of this section [the presumption that every distribution is out of earnings and profits] for the purpose of determining the taxability of subsequent distributions by the corporation."

The committee reports accompanying this section provide an example of a corporation which has $100,000 of capital, represented by $50,000 of preferred stock and $50,000 of common stock, $25,000 of pre-1913 earnings and profits, and $50,000 of post-1913 earnings and profits. It retires its preferred stock of $50,000 at par and then distributes a dividend of $25,000. The committee report then notes:

"In the absence of an express statutory provision it could be contended that the distribution of $50,000 in retirement of the preferred stock constituted under the presumption contained in subdivision (b), a distribution of the most recently accumulated earnings and profits of the corporation, with the result that such earnings and profits accumulated since March 1, 1913, were all distributed. It could be argued that the subsequent distribution of $25,000 was out of earnings and profits accumulated prior to March 1, 1913, and therefore tax exempt. The provisions of the bill make it clear that the distribution of the $50,000 in retirement of the preferred stock constitutes a capital transaction and does not affect the earnings and profits of the corporation on hand for subsequent distribution."

This provision was re-enacted as Sec. 115 (c) of the Revenue Act of 1928 and continued unchanged in the revenue law until 1936. In the Revenue Act of 1936, the provision was again re-enacted as the last sentence of Sec. 115(c), but the last half of the sentence was omitted, so that the new sentence read:

"In the case of amounts distributed * * * in partial liquidation * * the part of such distribution which is properly chargeable to capital account shall not be considered a distribution of earnings or profits."

It is this language which was re-enacted as Sec. 312(e) of the Revenue Act of 1954. There is no legislative history in connection with the 1936 Act which sheds any light on the reasons for this change in language. However, when this change is considered in connection with other changes made at the same time, the conclusion seems reasonable that ordinary distributions were intended to take priority over terminal distributions in the disposition of current earnings.

Prior to 1936 current earnings were prorated throughout the year in order to determine the amount of current earnings available on a particular date of distribution. Edwards v. Douglas, 269 U.S. 204, 46 S. Ct. 85, 70 L. Ed. 235 (1925); Mason v. Routzahn, 275 U.S. 175, 48 S. Ct. 50, 72 L. Ed. 223 (1927); Donahue v. United States, 58 F.2d 463, 74 Ct. Cl. 512 (1932). The earnings and profits were then reduced as of that date by the amount of the distribution. The result of this was that, for example, a corporation which made a dividend distribution in January 1917, and which had sufficient earnings and profits in 1917 to cover the distribution, was nevertheless deemed to have made the distribution out of its 1916 earnings and profits because the 1917 profits had not yet accumulated; thus the dividend tax was at the lower 1916 rates. Mason v. Routzahn, supra.

This accounting procedure was altered by the 1936 amendment to the dividend definition (now Sec. 316(a) (2)) which provided that current earnings and profits were to be computed as of the end of the taxable year without diminution by reason of any distributions made during the year and without regard to the amount of earnings and profits at the time the distribution was made. Since, as demonstrated above, the new dividend definition meant that current earnings and profits were to be diminished by neither ordinary nor redemption distributions, the deletion of the language in what was then Sec. 115(c) (now Sec. 312(c)) becomes explainable. This provision, in contrast to the dividend definition, does not determine the tax treatment to be given to a redemption distribution, but merely determines the tax accounting consequences of a redemption distribution at the corporate level. Whereas, on the one hand, if an ordinary distribution is determined to be a dividend by virtue of the dividend definition, it is then automatically given ordinary income treatment at the distributee level, on the other hand, the tax treatment to be given a redemption distribution at the distributee level does not automatically follow its designation as a redemption, but depends on whether certain other conditions specified in the revenue act are met.

Moreover, the tax treatment of a Sec. 302 redemption distribution is wholly independent of the earnings and profits account; it receives exchange treatment regardless of whether the source is earnings and profits or capital. Thus, since the purpose of the new dividend definition was to free current earnings and profits for availability as taxable dividends, this purpose could be easily met by providing that all current earnings and profits were to be applied first to ordinary distributions, as the new definition seems to imply.

Since the dates of all distributions became irrelevant in determining the extent of current earnings and profits available for use as dividend distributions, the language pertaining to the accounting treatment to be given liquidating distribution as affecting "the taxability of subsequent distributions" also became irrelevant. Current earnings and profits were to be computed at the end of the year without diminution by any distributions, ordinary or redemptive. Then to the extent of this source, all ordinary distributions were deemed dividends and taxed as such. The redemption distributions could not affect the taxability of subsequent distributions out of current earnings and profits, and thus the language was rendered superfluous.

The taxpayer makes two arguments relative to this analysis. First of all, it argues that Congress was utterly indifferent to the order in which redemptive and ordinary distributions were to be applied against current earnings, because a "dividends paid" credit was allowed under Sec. 27(f) for redemptive distributions, and thus the most that can be concluded is that redemption and ordinary distributions reduce current earnings in a pro rata amount. Secondly, it argues that to conclude that ordinary distributions take priority out of current earnings is inconsistent with the admitted fact that they do not take priority out of accumulated earnings.

In considering both of these arguments, it is to be kept in mind that in enacting the new dividend definition, Congress did not merely intend to give relief to deficit corporations from the undivided profits tax. It could have accomplished this purpose rather simply by giving, in addition to the dividends paid credit, a credit for any other distributions up to the amount of the current earnings, regardless of the fact that they were not taxable dividends because of impaired capital. However, Congress did not take this approach, but instead mandated that the distribution of current earnings was a taxable dividend in the hands of the shareholder, regardless of impaired capital. Thus Congressional intent was not merely to give relief to deficit corporations, but to tax distributions of current earnings as dividends in the hands of shareholders. Effect should be given to this intent unless a contrary result is clearly manifested by another section of the statute.

As to the dividends paid credit of Sec. 27(f) with respect to liquidating distributions, it is clear that that credit would not have been available in the instant case because this redemption was not pro rata among the shareholders as was required by Sec. 27(g). See, in this regard, 118 Regs. Sec. 39.27(h)-1, preferential distributions, and Example (b) 2. (Sec. 27 (g) became Sec. 27(h) in the 1939 Code.) Thus were the undistributed profits tax still in existence, the corporation would not be entitled to a dividends paid credit for the redemption in question. If we were to accept the taxpayers' alternative argument that dividend and redemption distributions are to be allocated pro rata to current earnings and profits, the result in the instant case would be that slightly more than half of the earnings and profits would be allocated to the dividend distribution, with the remaining amount allocated to the redemption distribution. Since the redemption distribution would not be entitled to the dividends paid credit, its allocated share of the earnings and profits (almost half) would be subject to the undistributed profits tax. I do not think that this result would be in accordance with Congressional intent in imposing the undistributed profits tax, since the corporation in fact distributed an amount equal to its earnings and profits in ordinary distributions, and thus I must reject the taxpayers' suggestion that redemption and ordinary distributions are entitled to pro rata allocations of current earnings.

Moreover, even in the case where a liquidating distribution was made pro rata, and therefore theoretically available for the dividends paid credit, such a credit would not have been available to a deficit corporation whose deficit exceeded its current income. See, St. Louis Co., supra, and cases cited therein. And this result would not be altered even if we adopted the taxpayers' argument that current earnings are to be prorated between ordinary and liquidating distributions. Take for example the case of a corporation with a deficit of $150,000 and current earnings of $100,000 (an example given in the House reports in discussing the new dividend definition). Assume that it makes ordinary distributions of $100,000 and liquidating distributions of $50,000. Under the taxpayers' suggestion that these distributions should be prorated against current earnings and profits, the ordinary distributions would be credited with only $66,667 of the current earnings. Since there was no excess of current earnings over the existing deficit, no dividends paid credit would have been allowed for the liquidating distribution, and therefore the corporation would be liable for an undistributed profits surtax on the amount of $33,333. I simply refuse to believe that this was a result intended by Congress, and the only way to avoid it is to give the ordinary distributions a priority claim on the current earnings. I think this is the proper analysis of Sec. 316(a) (2).

As to the argument that this results in dissimilar treatment being given to current earnings and accumulated earnings, because dividend distributions do not have a priority claim on accumulated earnings, I can only answer that this is a result of the Congressional scheme, I note, however, that this is not a new result under the existing statute, for even now current earnings and accumulated earnings do not receive the same treatment as to ordinary distributions. Whereas, current earnings are prorated over ordinary distributions of the current year, accumulated earnings are allocated to ordinary distributions on a first-in-time basis, that is, giving priority to those distributions in order of time paid.³  See, Regs., Sec. 1.316-2(c). Thus, this result is not unique.

In conclusion, I recognize that under the present state of the statutes, a completely satisfactory answer to the problems of the claims of liquidating distributions on earnings and profits is not possible. This is clearly an area which is ripe for Congressional action. However, as to the issue before us, I think that the District Court correctly determined that ordinary distributions have a priority claim over liquidating distributions on current earnings and profits.

BRIGHT, Circuit Judge (concurring):

I agree with an affirmance in this case and, generally, Judge Gibson's analysis. I am reluctant, however, to suggest that redemptive distributions may not be charged against current earnings and profits for tax accounting purposes under Sec. 312(a).¹  I am satisfied to rest an affirmance on the theory that redemptive distributions which are not taxable under Sec. 301 are not taken into consideration in determining a corporation's current earnings and profits available for dividends under Sec. 316(a).

Section 316(a)²  provides that a dividend is a distribution of property made: 1) out of "earnings and profits accumulated after February 28, 1913" (earnings and profits which are not relevant here); or 2) out of "earnings and profits of the taxable year."

Under Sec. 316(a) (2), earnings and profits of the taxable year are computed at the close of the taxable year "without diminution by reason of any distributions made during the taxable year" and "without regard to the amount of * * earnings and profits at the time the distribution was made." Accordingly, to determine whether an ordinary distribution, such as the distribution in question here, is taxable as a dividend or as a return of capital, the recipient must look to the corporation's year-end earnings and profits fund. If that fund, computed "without diminution by reason of any distributions made during the taxable year," is sufficient to cover ordinary distributions, the ordinary distribution received is taxable as a dividend.

Section 316(a) sets out a hypothetical, year-end accounting method for determining the taxability of an ordinary distribution. Section 312(a) provides a day-to-day accounting procedure for calculating the current status of the earnings and profits account; it speaks in terms of the effect on earnings and profits "on the distribution." A distribution of property by a corporation, whether ordinary or redemptive, serves to decrease the corporation's earnings and profits account under the accounting procedure set forth in Sec. 312(a). This decrease, however, has no effect upon the taxability of an ordinary distribution made during the same tax year by a corporation which had no accumulated earnings and profits. Under Sec. 316(a), the recipient of an ordinary distribution, in calculating the corporation's year-end earnings and profits fund available for distribution as dividends, must disregard any decrease caused by ordinary or redemptive distributions. As a corollary to this express accounting directive, the total of all ordinary distributions must then be compared with this hypothetical fund in order to determine the percentage of ordinary distributions which will be deemed dividends.

Redemptive distributions which are not taxable under Sec. 301 are not taken into account in this comparison because the last part of Sec. 316(a) provides:

Except as otherwise provided in this subtitle, every distribution is made out of earnings and profits to the extent thereof, and from the most recently accumulated earnings and profits. To the extent that any distribution is, under any provision of this subchapter, treated as a distribution of property to which section 301 applies, such distribution shall be treated as a distribution of property for purposes of this subsection.

I interpret this last section to mean that Sec. 316(a) does not encompass distributions which do not fall within Sec. 301. The redemptive distributions here do not fall within Sec. 301; they qualify as Sec. 302 distributions. See Judge Gibson's opinion ante, n. 1.

Construing Sec. 316(a) in this way, I find no conflict between its provisions and the provisions of Sec. 312(a). The sections set out separate accounting concepts which are relevant to the separate purposes served by each section. In the case of a corporation with no accumulated earnings, ordinary or redemptive distributions may be charged against current earnings and profits "on the distribution" under Sec. 312(a), but distributions which do not fall within Sec. 301 are not taken into consideration in determining the year-end earnings and profits fund available for dividends under Sec. 316 (a), In short, since the redemptive distributions in this case are not taxable as dividends, these distributions do not diminish the current earnings fund available for dividends.

LAY, Circuit Judge (concurring in part and dissenting in part).

I join the per curiam analysis that the words "any distribution" within the parenthetical clause of Sec. 316(a) (2) include redemptive distributions but I cannot agree that this interpretation accords a priority in the reduction of current earnings and profits for ordinary dividends. I cannot accept the positions taken by Judge Gibson and Judge Bright which for differing reasons accord a priority to dividends in reducing current earnings and profits.

Judge Gibson reasons that Sec. 312(a) "has no application to the problem before us." Thus, he limits the focus of his opinion to the interplay of Sec. 312(e) and Sec. 316(a) (2). By contrast, Judge Bright finds that Sec. 312(a) does govern the charges which are to be made against earnings for both dividends and redemptions, but he reasons that Sec. 316(a) (2) presents a wholly separate and independent accounting calculation which governs the instant problem without relationship to the operation of Sec. 312(a).

These separate holdings basically overlook that Subchapter C of the Internal Revenue Code presents a comprehensive statutory scheme for the taxation of corporate distributions and that each of the three above-mentioned sections were intended to play some operative part in the treatment to be given the dividend and redemptive distributions made in the year in question. By giving effect to all of the words of each statute and thus achieving harmony between these sections, I conclude that dividends can claim no priority in the reduction of current earnings and that each distribution (whether redemptive or dividend in nature) must play some part in reducing current earnings in a year when those earnings are insufficient to cover all distributions.

Section 312(a)

Judge Gibson holds that Sec. 312(a) has no application to the instant case. He reasons that by enacting subsection (e) of Sec. 312, Congress removed redemptions entirely from the purview of (a). I cannot agree. This interpretation lacks authoritative support,¹  is not ascribed to by the government, the taxpayers or Judge Bright, and is at variance with the very manner in which Sec. 312 was drafted. Congress has expressly recognized that adjustments to earnings for redemption distributions are to be made under subsection (a). In 1969 Congress amended subsection (c) of Sec. 312 to provide that the gain now recognized on certain redemption distributions is to be taken into account in adjusting earnings for such a transaction. As the amended statute specifically states, this gain is to be taken into account "in making the adjustments to the earnings . . . under subsection (a) . . .."²  If, as Judge Gibson contends, subsection (e) controlled the adjustments which are made to earnings for redemption distributions, the above quoted language would have necessarily made reference to (e) instead of (a).

* * *

". . . The purpose of the third sentence of section 115(c) of the 1939 Code was to make clear that a liquidating distribution which was properly chargeable to the capital account should not be considered a distribution of earnings or profits insofar as subsequent distributions were concerned. There is no reason to suppose that when the sentence in question was reenacted as section 312(e) of the 1954 Code it was intended that it should have any different effect than that which it had ever since 1924."

I am in agreement with this analysis. Hence, since the earnings portion of a redemption distribution is intended to have an effect on the taxability of "subsequent" distributions, it must do so by reducing earnings before the subsequent distribution is made. This, of course, occurs by reason of the application of the "time rule" of Sec. 312(a) which causes earnings to be reduced at the time of distribution. I cannot accept Judge Gibson's interpretation that this language was deleted in recognition of a legislated priority for dividends.

Section 316(a) (2)

I am in respectful disagreement with the analysis of Judge Gibson and Judge Bright that the language of the parenthetical clause of Sec. 316(a) (2)¹³  is intended to mean that these earnings are so computed for the purpose of making them first available for dividends.¹⁴  If this interpretation is correct then the obvious result is a complete negation of the operations of Sec. 312(a) and Sec. 312 (e). Clearly, if redemptions which are made earlier in time than dividend distributions are forced to wait in abeyance until the subsequent dividends have been given the first opportunity to absorb earnings, then the redemptions cannot affect the taxability of subsequent distributions as Sec. 312(e) indicates, nor can they decrease earnings at the time of their distribution as Sec. 312(a) indicates. I do not believe that Congress intended Sec. 316(a) (2) to be interpreted in a manner which would create such a manifest disharmony between these statutes.¹⁵ 

The operative effect of Sec. 316 is to define those distributions which shall be considered to be "dividends." The title of this section explicitly states that purpose. The manner of definition is one which is important to the resolution of our problem. Dividends are there defined as "any distribution" which comes "out of" earnings (current or accumulated). Thus, the dividend label attaches only after it is found that a distribution has reduced the earnings account. Nothing in Sec. 316 governs the actual reduction of earnings. In fact Sec. 316 does not become operative until the earnings account has been reduced. Hence, the rules of Sec. 312(a) must be first applied before Sec. 316's definition of a dividend can function. As a result I fail to grasp the approach suggested by Judge Bright that Sec. 312(a) and Sec. 316(a) (2) provide independent accounting methods for a corporate taxpayer-one for purposes of corporate bookkeeping and the other for purposes of determining taxation. This suggestion seems patently inconsistent with the realities and purposes of tax legislation. What possible purpose could Congress have had in prescribing within Sec. 312(a) of the Internal Revenue Code a bookkeeping procedure to account for the reduction of corporate earnings caused by distributions if such a procedure is to have no relevancy in the determination of the taxability of such distributions? Even the Treasury regulations indicate that there must be some relationship between the methods used in maintaining the earnings account on the corporate books and the accounting method used for determining earnings available for taxable distributions. See Treas.Reg. 1.312-6(a). Clearly, the operations of Sec. 312(a) and Sec. 316(a) are interlocking steps in the statutory scheme of taxing corporate distributions.¹⁶ 

In my view every distribution, whether redemptive or dividend, reduces earnings (to the extent available) at the time of distribution. This occurs under Sec. 312 (a). Of course, in order to determine the "extent" of the earnings available on each distribution date some computation of earnings must be made.¹⁷  The parenthetical clause of Sec. 316(a) (2), in my judgment, merely sets forth the manner of computing current earnings when those earnings are a source of dividend distributions. The legislative history of Sec. 316(a) (2) bears this out.

Prior to the enactment of this section in 1936 "accumulated earnings" were the only available source of dividend distributions. Current earnings, however, could also serve as a source for dividends provided they qualified as "accumulated" -i. e., any existing deficit had to be first absorbed before any part of current earnings could be considered "accumulated." Unlike "true" accumulated earnings whose actual balance could be readily determined at any point during the year, practical difficulties inhibited the accurate computation of the actual amount of current earnings existing at any one point during the year.¹⁸  This information, of course, was necessary so that it could be known by how much a distribution reduced earnings on the date it was distributed. As a result a special method of accounting was devised by the Treasury which was intended to facilitate the computation of available current earnings so that the amount "accumulated" as of any one day during the year would be known. Under this procedure, if it was not possible to show the actual amount existing in the current earnings account on a distribution date, then earnings were to be computed as of the end of the year in toto (without regard to the effect of any distribution therefrom made during the year). These earnings were then deemed to have accumulated ratably over the 365 days of the year. See Edwards v. Douglas, 269 U.S. 204, 46 S. Ct. 85, 70 L. Ed. 235 (1925) and Mason v. Routzahn, 275 U.S. 175, 48 S. Ct. 50, 72 L. Ed. 223 (1927).¹⁹ 

In 1936 Congress for the first time made current earnings a dividend source in its own right by the enactment of the forerunner of Sec. 316(a) (2). The result of this enactment was to free current earnings from the necessity of qualifying as "accumulated." As part of this change the Treasury redesigned its earnings computation rules. The new statute (Sec. 115(a) (2) Rev.Act of 1936) enacted that part of the former computation which required current earnings to be computed as of the end of the year. The Treasury, through its regulations, added the innovation that these earnings, so computed, were to be allocated on a proportionate basis among the various distribution dates of the year rather than ratably to each and every day of the year.²⁰  The purpose of such a change was obviously to remove the undesirable result of having part of a distribution charged to capital in a year when the total current earnings were sufficient to absorb the entire charge. As with the prior computation, the earnings so computed are available for both dividends and redemptions. There is no authority for concluding that the result of making current earnings an independent source for dividend distributions was to preclude other distributions from also being charged against this source. The computation made under Sec. 316(a) (2) cannot control which distributions reduce earnings on their distribution dates.

The government argues that by these changes Congress removed the significance of both the dates of distribution and the order of distributions, and thus reserved current earnings for first reduction by dividends. According to the government the example contained in Treasury Reg. 1.316-2(c) which shows how to prorate earnings proportionately among dividend distributions demonstrates that only dividend distributions are to be treated as reducing earnings (in a year when total current earnings are insufficient to cover all distributions of the year). It is true this method of computation eliminates the significance of dates as they were used in making a daily computation of earnings. However, the regulation did not change the rule of Sec. 312(a) (which applies to both dividends and redemptions) which requires a reduction of earnings on the date of each distribution. The computations under Sec. 316(a) (2) and the regulation that prorates current earnings among the actual distribution dates are a clear indication that these dates are still significant as the points in time when earnings are reduced as Sec. 312(a) demands. Thus, if a distribution reduces current earnings on the date when made, the only purpose of the yearend computation is to facilitate a determination of the "extent" of the earnings existing on the distribution date. In a year when both dividends and redemptions are distributed, a year-end computation is necessary to determine how much of each distribution came "out of" earnings. Nor, can one reasonably infer that by means of this computation and the example contained in the regulations, earnings must be considered available on dividend distribution dates but nonexistent on redemption distribution dates. Under this strained interpretation a redemption which may have preceded a dividend distribution by only one day would have to be charged to capital because no earnings would be deemed to be in the current earnings account on the date of the redemption distribution, even though on the very next day a sufficient amount of earnings would be considered to be in the account to cover the dividend. I cannot believe that Congress intended such an artificial conceptualization of when earnings become earned and available.²¹  Nor do the regulations require such a result since they exemplify a year wherein only dividend distributions were made. Nothing in the regulations governs the computations to be made in a year when both dividends and redemptions (or other non-dividend distributions) are distributed.

The taxpayers' argument which results in priority treatment for redemptive distributions must be similarly rejected. The taxpayers contend that the effect which the redemption distributions had on earnings on each date of distribution should be determined by the daily computation of earnings formula. In making their computations the taxpayers have considered that 1/365th of the total current earnings entered the account each day. However, taxpayers reduce these earnings for each redemption distribution on its distribution date without giving effect to the reductions which would have occurred on the interim dividend dates. Their contention is, of course, that the dividends did not reduce earnings on their distribution dates because the year-end computation of Sec. 316(a) (2) suspends their effect on earnings until that time. Meanwhile, they contend the redemptions were diminishing the earnings account. This position is as contradictory of the "time rule" of Sec. 312(a) as the government's argument of priority for dividends.²² 

In my judgment the daily computation formula cannot be used in a year when both dividend and redemption distributions are made and current earnings are insufficient to cover both types. The regulations obviously require the proportionate prorata formula be used in such circumstances for dividends. Use of this formula for dividends while at the same time using another method of computing available earnings for redemptions (daily proration) results in a negation of the meaning of the parenthetical clause of Sec. 316(a) (2). There would be no purpose in computing earnings without regard to the diminutions caused by all distributions from the account during the year, if thereafter in determining the amount available for dividends the redemption distributions are treated as having already diminished the account. Hence, in a year when both types of distributions are made from current earnings and these earnings are insufficient to cover all distributions of the year, the only reasonable conclusion is that the total current earnings computed at the end of the year without regard to the diminishing effect of any distribution made during the year are to be prorated among all distributions (both dividend and redemption).

I conclude, therefore, that the parenthetical clause of Sec. 316(a) (2) merely provides a method of computing the total extent of available current earnings without preempting those earnings for dividends only. Thus, each distribution (dividend and redemption alike) has the effect of decreasing earnings "on the distribution" to the "extent" of that source as Sec. 312(a) requires. The requirements of Sec. 312(e) and the regulations thereunder are thereby given effect since the portion of the redemption which is allocable to earnings will diminish the earnings available for later distributions and thereby have an effect in "determining the taxability of subsequent distributions." And, Sec. 316(a) (2)'s method of computing current earnings without regard to "any distribution" made therefrom during the year will be given literal application. Only in this way can harmony be achieved among these statutes.

The principle of Foster v. United States, 303 U.S. 118, 58 S. Ct. 424, 82 L. Ed. 700 (1937), does not conflict with this result. Foster relates to redemptions which were free of tax consequences, whereas the redemptions here are subject to tax although at a preferred rate. In construing statutes levying a tax it is not our duty to impose the greatest tax possible, especially is this so when an apparent ambiguity exists. It is well settled " [I]f doubt exists as to the construction of a taxing statute, the doubt should be resolved in favor of the taxpayer." Hassett v. Welch, 303 U.S. 303, 314, 58 S. Ct. 559, 565, 82 L. Ed. 858 (1938); Gould v. Gould, 245 U.S. 151, 153, 38 S. Ct. 53, 62 L. Ed. 211 (1917).

For these reasons I would reverse and allow plaintiffs to recover a partial refund of tax paid.