These writs of certiorari were granted to examine the deductibility as interest of certain payments which the taxpayer corporations made to holders of their corporate obligations. Although the obligations of the two taxpayers had only one striking difference, the noncumulative in one and the cumulative quality in the other of the payments reserved under the characterization of interest, the Tax Court (formerly the Board of Tax Appeals, 56 Stat. 957, 26 U.S.C.A.Int.Rev.Code, § 1100; only its present name will be used herein) held that the payments under the former, the Kelley Company case, were interest and under the Talbot Mills were dividends. The Circuit Court of Appeals reversed the Tax Court in the Kelley case and another circuit affirmed the Talbot Mills decision.1 On account of the diversity of approach in the Tax Court and the reviewing courts, we granted certiorari.

In the Kelley case, a corporation, all of whose common and preferred stock was owned directly or as trustee by members of a family group, was reorganized by authorizing the issue of $250,000 income debenture bearer bonds, issued under a trust indenture, calling for 8% interest, non-cumulative. They were offered only to shareholders of the taxpayer but were assignable. The debentures were payable in twenty years, December 31, 1956, with payment of general interest conditioned upon the sufficiency of the net income to meet the obligation. The debenture holders had priority of payment over stockholders but were subordinated to all other creditors. The debentures were redeemable at the taxpayer's option and carried the usual acceleration provisions for specific defaults. The debenture holders had no right to participate in management. Other changes not material here were made in the corporate structure. Debentures were issued to the amount of $150,000 face value. The greater part, $114,648, was issued in exchange for the original preferred, with six per cent cumulative guaranteed dividends, at its retirement price and the balance sold to stockholders at par, which was eventually paid with sums obtained by the purchasers from common stock dividends. Common stock was owned in the same proportions by the same stockholders before and after the reorganization.

In the Talbot Mills case the taxpayer was a corporation which, prior to its recapitalization, had a capital stock of five thousand shares of the par value of $100 or $500,000. All of the stock with the exception of some qualifying shares was held by members, through blood or marriage, of the Talbot family. In an effort to adjust the capital structure to the advantage of the taxpayer, the company was recapitalized just prior to the beginning of the fiscal year in question, by each stockholder surrendering four-fifths of his stock and taking in lieu thereof registered notes in aggregate face value equal to the aggregate par value of the stock retired. This amounted to an issue of $400,000 in notes to the then stockholders. These notes were dated October 2, 1939, and were payable to a specific payee or his assignees on December 1, 1964. They bore annual interest at a rate not to exceed 10% nor less than 2%, subject to a computation that took into consideration the net earnings of the corporation for the fiscal year ended last previous to the annual interest paying date. There was, therefore, a minimum amount of 2% and a maximum of 10% due annually and between these limits the interest payable varied in accordance with company earnings. The notes were transferable only by the owner's endorsement and the notation of the transfer by the company. The interest was cumulative and payment might be deferred until the note's maturity when 'necessary by reason of the condition of the corporation.' Dividends could not be paid until all then due interest on the notes was satisfied. The notes limited the corporation's right to mortgage its real assets. The notes could be subordinated by action of the Board of Directors to any obligation maturing not later than the maturity of the notes. For the fiscal year in question the maximum payment of 10% was made on the notes.

The payments in question on corporate obligations were for the years in the Kelley case, 1937, 1938 and 1939; in the Talbot Mills case for the year 1940. Both corporations deducted the payments as interest from their reports of gross income under statutory sections and regulations set out in the footnote.2 The applicable statutes and regulations were identical for all periods. The Commissioner asserted deficiencies because the payments were considered dividends and not interest.

There is not present in either situation the wholly useless temporary compliance with statutory literalness which this Court condemned as futile, as a matter of law, in Gregory v. Helvering, 293 U.S. 465, 55 S.Ct. 266, 79 L.Ed. 596, 97 A.L.R. 1355. The demonstrated possibility of sales by the holders of the obligations to persons other than stockholders alone proves the differentiation. As material amounts of capital were invested in sock, we need not consider the effect of extreme situations such as nominal stock investments and an obviously excessive debt structure.

From the foregoing statements of facts, if appears that the characteristics of all the obligations in question and the surrounding circumstances were of such a nature that it is reasonably possible for determiners to reach the conclusion that the secured annual payments were interest to creditors in one case and dividends to stockholders in the other case. In the Kelley case there were sales of the debentures as well as exchanges of preferred stock for debentures, a promise to pay a certain annual amount, if earned, a priority for the debentures over common stock, the debentures were assignable without regard to any transfer of stock, and a definite maturity date in the reasonable future. These indicia of indebtedness support the Tax Court conclusion that the annual payments were interest on indebtedness. On the other hand, in the Talbot Mills case, the Tax Court found the factors there present of fluctuating annual payments with a two per cent minimum, the limitation of the issue of notes to stockholders in exchange only for stock, to be characteristics which distinguish the Talbot Mills notes from the Kelley Compoany debentures. Upon an appraisal of all the facts, the Tax Court reached the conclusion that the annual payments by Talbot Mills were in reality dividends and not interest.

We think these conclusions should be accepted by the Circuit Courts of Appeals and by ourselves. Judicial review of Tax Court decisions depends upon the Internal Revenue Code, Section 1141(c) Powers (1), 26 U.S.C.A.Int.Rev.Code § 1141(c)(1). It reads:

'To affirm, modify, or reverse. Upon such review, such courts shall have power to affirm or, if the decision of the Board is not in accordance with law, to modify or to reverse the decision of the Board, with or without remanding the case for a rehearing, as justice may require.'

It is only recently that we gave careful consideration to the problems of review of Tax Court decisions. Dobson v. Commissioner, 320 U.S. 489, 64 S.Ct. 239, 88 L.Ed. 248. That opinion emphasized that our interpretation of Congressional purpose, in enacting the statute, just quoted for judicial review of Tax Court decisions, was that Congress intended to leave to the final determination of the Tax Court all issues which were not clear-cut questions of law.

The provisions for review are the same now as they were when enacted in 1926. Congress, and all others interested, were then well aware of the difficulties in drawing a line between questions of fact and questions of law.3 The legislation was upon a subject, the collection of the revenue, in which federal administrative finality had been given wide scope.4 The Tax Court was originally established to 'secure an impartial and disinterested determination of the issues involved,'5 so that the taxpayer and the Government would have an independent review of the position of either on tax demands before payment of the tax or foreclosure of an asserted deficiency. Two years later its success was recognized by committee commendation and the enlargement of the finality of its decisions from 'prima facie evidence of the facts contained therein' to reviewability only 'if the decision of the oard is not in accordance with law'6 As to the mischief which the limitation of the scope of judicial review was to cure, we find only the words of the committee reports.7 Without a clearer description by Congress of the intended line to separate reviewability of the Tax Court decisions from non-reviewability, courts must interpret the review statute, as best they can, to accomplish the declared Congressional purpose of adequate control of administrative action without substituting judicial opinion for that of the Tax Court upon the evidence. Note 7, supra.

The illustrations in the report, note 7, supra, are legal questions without doubt, except the possibility that the words 'application of the statute or any regulatio having the force of law' may be thought to give a reviewing court power to pass upon the Tax Court's conclusion from the primary or evidential facts. So that in the present cases, it might be said to be a question of law as to whether the primary facts adduced made the payments under consideration dividends or interest. But we think such conclusion gives inadequate weight to the purpose of the Tax Court. The finality of the Tax Court's rulings was being enlarged by the 1926 Act. The then Board was spoken of as an impartial and independent tribunal of experts 'for the determination of tax liabilities as between the Government and the taxpayer.' H.Rep. No. 1, 69th Cong., 1st Sess., p. 17. There would hardly need to be experts in tax affairs to decide questions of dates or amounts or values or to calculate rates. Their usefulness lies primarily in their ability to examine relevant facts of business to determine whether or not they come under statutory language. Adequate reason for the use of the word 'application' of course exists in situations where true legal questions arise, as in whether an act applies to transfers antecedent to its enactment or to income or estate taxes from trusts or to situations which involve conflicts of law. There is nothing in the context in which the word 'application' is used which suggests to us that it should be given its widest connotation.

These cases now under consideration deal with well understood words as used in the tax statutes—'interest' and 'dividends.' They need no further definition. Equiable Life Assurance Society v. Commissioner, 321 U.S. 560, 64 S.Ct. 722, 88 L.Ed. 927; Deputy v. Du Pont, 308 U.S. 488, 498, 60 S.Ct. 363, 368, 84 L.Ed. 416. The Tax Court is fitted to decide whether the annual payments under these corporate obligations are to be classified as interest or dividends. The Tax Court decisions merely declare that the undisputed facts do or do not bring the payments under the definition of interest or dividends.8 The documents under consideration embody elements of obligations and elements of stock. There is no one characteristic, not even exclusion from management, which can be said to be decisive in the determination of whether the obligations are risk investments in the corporations or debts. So called stock certificates may be authorized by corporations which are really debts and promises to pay may be executed which have incidents of stock. Such situations seem to us to fall within the Dobson rule.9

This leads us to affirm the Talbot Mills decree and to reverse the Kelley judgment. It is so ordered.

Judgment in No. 36 reversed; judgment in No. 47 affirmed.

Mr. Justice BLACK concurs in the result in No. 47. He is of the opinion that No. 36 should be affirmed for the reasons given by the Circuit Court of Appeals, 146 F.2d 466.

Mr. Justice BURTON concurs in the result in the Kelley case but dissents from the result in the Talbot Mills case on the grounds stated in the dissenting opinion of Magruder, J., in the Circuit Court of Appeals. CJ Mr. Justice JACKSON took no part in the consideration or decision of these cases.

Mr. Justice RUTLEDGE.

I think the judgments in both cases should be affirmed. On the records presented, I can see no satisfactory basis for deciding one case one way and the other differently. An I agree with the Court of Appeals that, on the substantially identical facts, the payments were dividends and not interest.

In the first place, I do not believe that Congress has authorized the Tax Court to make or the reviewing courts to sustain directly conflicting determinations of tax liability in identical fact situations. Nor, in my opinion, was this the purpose or effect of the Dobson decisions, 320 U.S. 489, 64 S.Ct. 239, 88 L.Ed. 248. So to regard them or the statute nullifies the right to review expressly given by Congress. Moreover that view destroys the very uniformity which Dobson sought, transferring the conflict of decision from the Courts of Appeals back to the Tax Court, by making the conflicting decisions of its sixteen divisions final.1 This affords relief to the taxpayer from judicial review and to the courts from judicially reviewing. But it defies Congress' mandate for review and, what is more, perpetuates chaos in the law.

All this presupposes, of course, that the records now here present fact situations identical in all material respects. That is true in my judgment. It is hardly necessary to attempt demonstration. But, besides referring to the opinions of the Courts of Appeals for the small details of the facts and their minute differences,2 it may be noted that there was no question of credibility. Substantially all of the evidentiary facts were stipulated in both cases. Nor is there any finding in either case that the arrangements were a sham. Cf Gregory v. Helvering, 293 U.S. 465, 55 S.Ct. 266, 79 L.Ed. 596, 97 A.L.R. 1355. Apart from such considerations, the material facts in my opinion were not substantially different in any respect sufficient to support one ultimate conclusion, whether labelled of 'law,' of 'fact,' or 'mixed,' for one case and the opposite conclusion for the other.

That is true whether the final conclusion of 'interest' or 'dividend' is to be drawn from a minute comparison of, and effort to differentiate, the multitudinous miscroscopic details by which in both cases it was sought to convert stock into 'debentures' or 'registered notes,' without losing any of the stock's substantial advantages; or, on the other hand, the final plunge of judgment is to be made from wholesale weighing of the evidentiary facts. Neither approach discloses factors of substantial difference in what was done sufficient to sustain contrary judgments.

There were some highly technical differences in the two types of 'security' which were devised to replace the preexisting preferred stock issues. But in both instances the original stock and the replacing security were closely held. There was no substantial change in the distribution after the 'reorganization.' The difference between the stock and the substituted security was so small, in its effect upon the holders' substantial rights, that for all practical purposes it was negligible. For example, a remote right to sue to enforce the obligation, deferred in one case for 25 years, took the place of the holder's right to share in the corporation's assets on dissolution or winding up. Meanwhile 'interest' was hooked in large part to net annual earnings and was made entirely ubject to the directors' power to suspend payment until the ultimate maturity date. The shortened story is that the preferred shareholders who went into the wash came out substantially, for all purposes material to any tax determination and it may be for practically all others, just about what they were when they went in.

The Court indeed does not attempt to find a substantial differentiating factor other than in the Tax Court's 'appraisal of all the facts,' in other words its ultimate conclusion. That is true as between the two cases and also as affects the positions of the respective shareholders before and after the wash. Rather the opinion concedes that in each case the circumstances were such that determiners reasonably could conclude that the so-called annual payments were either interest or dividends. Hence, it seems to follow, the conclusion may be drawn in squarely conflicting ways, if the Tax Court sees fit so to draw it; and it is immaterial that no factor of substantial difference is or can be pointed out.

One might entertain the view that in a close situation the Tax Court's judgment should be accepted whatever way the die were cast, although reviewing courts might differ on the direction. But it would not follow, and in my judgment should not, that they are powerless when the throw is in opposite directions at the same time. When this occurs, in my opinion a 'clearcut' question of law is presented, rising above the rubric of 'expert administrative determination.' The more apt characterization would be 'expert administrative fog.'

I think the Court's of Appeals and we are bound to review such cases; they by plain mandate of § 1141(c)(1) of the Code, 26 U.S.C.A.Int.Rev.Code, § 1141(c) (1), we by that section (see Bingham's Trust v. Commissioner, 325 U.S. 365, 65 S.Ct. 1232) and the provision of our rules making conflict between circuits 'special and important reasons' for granting certiorari. Rule 38, subd. 5(b), 28 U.S.C.A. following section 354. Conflict is not removed simply because judgments of the Court of Appeals judicially formalize the contrary ultimate, but nevertheless administrative, conclusions of the Tax Court. When no facts can be pointed to which are sufficient to distinguish Tax Court decisions in legal effect, except that the Tax Court has decided differently in two cases, the Courts of Appeals and we are bound by law and by our duty to exercise a sound discretion in review to resolve the conflict.

Another reason convinces me that both judgments should be affirmed. What has been said applies to conflicting determinations of the Tax Court, whatever the particular line which is to be drawn and regardless of its general location. But in these cases I think that as a matter of law the line should not be located where the Tax Court has placed it.

Tax liability should depend upon the subtle refinements of corporate finance no more than it does upon the niceties of conveyancing.3 Sheer technicalities should have no more weight to control federal tax consequences in one instance than in the other. The taxing statute draws the line broadly between 'interest' and 'dividend.' This requires one who would claim the interest deduction to bring himself clearly within the class for which it was intended.4 That is not done when the usual signposts between bonds and stock are so obliterated that they become invisible or point equally in both directions at the same time.

'Dividend' and 'interest,' 'stock and 'bond,' 'debenture' or 'note,' are correlative and clearly identifiable conceptions in their simpler and more traditional exemplifications. But their distinguishing features vanish when astute manipulation of the broad permissions of modern incorporation acts results in a 'security device' which is in truth neither stock nor bond, but the half-breed offspring of both. At times only the label enables one to ascertain what the manipulator intended to bring forth. But intention clarified by label alone is not always legally effective for the purpose in mind.5 And there is scarcely any limit to the extent or variety to which this kind of intermingling of the traditional features of stock and bonds or other forms of debt may go, as the books abundantly testify.6 The taxpayer should show more than a label or a hybrid security to escape his liability. He should show at the least a substantial preponderance of facts pointing to 'interest' rather than 'dividends.'

Something more is at stake in these cases than nice distinctions between 'stock' and 'bonds' on the one hand or between ultimate conclusions of 'fact' and 'law' or 'mixed fact and law,' on the other, just as was true in the conveyancing cases. The border cutting across one set of normally opposing conceptions may be deliberately obscured and made into a no man's land as readily as that involved in the other. When this happens, the final link in the chain of judgment is decisive whatever its label.7 If the ultimate conclusion of the Tax Court or its divisions can be made in exactly opposing ways, and must be left undisturbed, without substantial differentiating facts, or when hybrid arrangements bear tax indicia equally with marks of nontaxability, not only is the statutory review nullified. The right of taxpayers to be treated with equal justice before the law is denied.