The issue presents still another phase of a controversy much considered with respect to the justiciability of tax disputes wherein there has been resort by the Commissioner of Internal Revenue to permitted or required special treatment under so-called special assessment provisions. Beginning with the case of Williamsport Wire Rope Co. v. United States, 277 U.S. 551, 48 S.Ct. 587, 72 L.Ed. 985, and continuing through United States v. Memphis Cotton Oil Co., 288 U.S. 62, 53 S.Ct. 278, 77 L.Ed. 619; United States v. Prentiss & Company, 288 U.S. 73, 53 S.Ct. 283, 77 L.Ed. 626; United States v. Factors & Finance Co., 288 U.S. 89, 53 S.Ct. 287, 77 L.Ed. 633; Bemis Bros. Bag Co. v. United States, 289 U.S. 28, 53 S.Ct. 454, 77 L.Ed. 1011, and Welch v. Obispo Oil Co., 301 U.S. 190, 57 S.Ct. 684, 81 L.Ed. 1033, and in this court in Cleveland Co. v. United States, 6 Cir., 70 F.2d 365, and Joseph & Bros. Co. v. United States, 6 Cir., 71 F.2d 389, the doctrine has been developed that no court has power to review the grant or denial of a special assessment or the correctness of the computation made thereon when an essential factor in the determination of tax liability has been committed to the discretion of the taxing authority.
The present taxpayer, conceding that essential factors in determining tax liability are committed by §§ 327 and 328 of the *230Revenue Act of 1918, 40 Stat. 1093, to the broad discretion of the Commissioner, and that computations of tax employing such factors may not be reviewed, challenges a determination of its 1917 income as one controlled by the provisions of the 1917 Revenue Act, 40 Stat. 300, whereby the only discretion committed to the Commissioner is to determine what portion of its income is normal, so that the higher rates may be applied to the excess; and asserts that it is not thereby attacking the ascertained deduction and so does not seek judicial review of any factor of computation committed to administrative discretion, but seeks only a correction in the determination of its net income, which in no way affects any determination which the Commissioner 'is permitted at his discretion to make.
The suit was against the Collector for the refund of a tax for the taxable year 1917 alleged to have been unlawfully collected. Upon a showing that a special determination had been made by the Commissioner under § 210 of the Revenue Act of 1917, the suit was upon motion dismissed for want of jurisdiction. It is from the order of dismissal that the appeal is taken.
The history of the controversy is involved. We shall try to simplify it. Prior to 1917 corporations were subjected only to a straight percentage income tax. The 1917 Revenue Law imposed upon them for the first time an excess profits tax, so that corporate profits in excess of what were deemed to be the normal profits in pre-war years were to be taxed at progressively higher rates. The normal profits were to be ascertained by determining the invested capital for the taxable year, and by assuming income up to a stated percentage thereof to be normal. The remainder was the excess. By § 210, however, printed in the margin,1 it was provided that if the Commissioner was unable “satisfactorily to determine the invested capital” he might resort to a comparison with representative corporations, partnerships and individuals, in a similar trade or business to ascertain the ratio of their average income to their average invested capital, and to apply that ratio to the taxpayer’s income.
The appellant, an Ohio corporation, though affiliated with other corporations, filed a separate income and profits tax return for the calendar year 1917, and paid the tax due thereon. Upon investigation the Field Examiner substantially increased net income and reduced invested capital. The taxpayer complained and requested relief under § 210 of the 1917 law. The Commissioner approved the action of the examiner, advised the taxpayer that "the tax had been computed under § 210, and that an additional tax of $25,438.48 had been determined and assessed. The appellant paid this amount on February 13th, 1922. This constitutes the first phase of the case.
The Act of 1917 had made no provision for a consolidated return by affiliated corporations, although by regulation the Commissioner had assumed power to- require it. § 240(a) of the Act of 1918, 40, Stat 1081, however, made consolidated returns obligatory, and § 240(e) of the Act of 1921, 42 Stat. 260, extended the requirement retroactively to include all years prior to 1922. In 1923 the Commissioner required a consolidated return from the taxpayer and affiliates, and accordingly such return was filed on October 7th of that year. It showed a consolidated net loss indicating no excess profits tax. The controversy thereafter was with respect to a sale by one of the affiliates of land, buildings and mining equipment in March, 1917, and whether by it loss or gain was realized. During its continuance three different determinations of excess profits tax liability involving differing determinations of consolidated invested capital and consolidated net income were made. The taxpayer excepted to each, and on January 30th, 1924, requested special consideration by the Special Assessment Section of the Bureau. The Commissioner thereupon reconsidered all questions involved, recomputed the excess profits tax upon the consolidated return, and allocated *231to the taxpayer a sum sufficient to avoid the necessity of making refund. No assessment was, however, made against the taxpayer, and so no appeal could therefore be taken to the Board of Tax Appeals. This is the second phase of the controversy.
On December 12th, 1925, the case entered its final phase by the taxpayer filing its claim for refund on the ground that the tax which had been paid had been assessed upon income shown by its separate return. Since this had been superseded by later proceedings it was contended that the basis for the assessment had disappeared. The Commissioner reaudited the entire consolidated return. He redetermined consolidated net income and consolidated invested capital. He advised the taxpayer by letter of September 24th, 1928, that its excess profits tax was based up on comparison with a group of representative corporations in a like trade or business under § 210, that he had redetermined the excess profits tax, that the amount allocated to it had already been assessed and collected, that a portion of the amount allocated to its affiliates had also been assessed and collected, and that the balance due from the latter was barred by the statute of limitations.
On January 7th, 1930, ’the taxpayer’s claim for refund was rejected, and on September 23d, 1930, the taxpayer began its suit. It renews the contention made in its claim for refund that since its separate return was superseded by the consolidated return, the former disappears as a basis for denying the refund, and asserts its status under § 210 need not be considered. The principle is now well established, however, that upon a claim for refund the Commissioner may redetermine the entire tax liability of the taxpayer, and even though no new assessment can be made because the bar of the statute has fallen, the taxpayer is nevertheless not entitled to a refund unless he has overpaid his tax, for that is the ultimate question presented for decision. Lewis v. Reynolds, 284 U.S. 281, 52 S.Ct. 145, 76 L.Ed. 293. It is, therefore, upon the question of jurisdiction immaterial whether the original assessment upon the separate return was superseded, and likewise immaterial that no assessment was made in respect to the Commissioner’s final determination of the tax. Determination of factors employed in tax computation precede and control assessments, regular or special. Since the present inquiry is whether review is sought of factors or result within the discretion of the Commissioner, it is unimportant that determinations or computations failed to ripen into effective assessment. As the defense is equitable in diminution of the plaintiff’s right, with no counter demand upon him, it would not at the trial be precluded. Stone v. White, 301 U.S. 532, 538, 57 S.Ct. 851, 81 L.Ed. 1265. It must follow that a tax liability reached by a determination of discretionary factors by the Commissioner may be the basis for the denial of a refund, and even in the absence of assessment, control the question of justiciability.
More important are the contentions that the Commissioner had no jurisdiction to proceed under § 210, that if he had he did not do so, and that if he did so the court may nevertheless correct an erroneous finding of net income since such finding in no way affects the exercise of the limited discretion which the section requires or permits. In challenging the jurisdiction of the Commissioner to proceed with special assessment, it is pointed out that unlike §§ 327 and 328 of the 1918 Act, § 210 of the 1917 Act limits his jurisdiction to proceed to special assessment to those cases and those alone wherein he is unable “satisfactorily to determine the invested capital”, and then only to the extent of determining by a comparison with other taxpayers in the same trade or business “the amount of the deduction” which should be made from total net income to determine the portion thereof upon which to compute the excess taxes. The point is made that in various stages of the proceedings the Commissioner did determine invested capital, and though in original and subsequent determinations he later found error, yet he fails to declare his final determination to be unsatisfactory or erroneous. The only basis of jurisdiction, therefore, disappears, since the Commissioner could satisfactorily (to himself) determine invested capital.
The fact remains, however, that the Commissioner did make special assessment, and this at the instance of the taxpayer. We may concede without deciding that the taxpayer’s request for such assessment does not condition the Commissioner’s decision to make it, but doubtless such request may bring to the Commissioner’s attention some lack of certainty or definiteness in the factors upon which earlier determinations are based. While inability to satisfactorily ascertain invested capital is a fact, *232yet it is a fact to be established in the judgment of the Commissioner upon consideration, or even repeated reconsideration, of all' elements involved, and when he undertakes to act, a court should not hold except upon most clear and convincing evidence that the jurisdiction was lacking. In the light of the Commissioner’s repeated change of view and the taxpayer’s repeated protests, we are unable to conclude that the Commissioner could satisfactorily ascertain invested capital.
, ,, , „ . [6] Assuming a holding that the Commissioner had jurisdiction to proceed under § 210, the taxpayer endeavors to show that he failed to do so, and that this failure grew out of confusmg the proceedings required m the W17 ^ Wlth ,those permitted m the 1918 Act,_ since at the time the challenged tax liability .was determined the 1918 Act was already being administered and a special assessment division was thereunder functioning. That the distinction between the two acts had been obscured, if not destroyed, was pointed out in United States v. Factors & Finance Co., supra, 53 S.Ct. 289, where questions of the validity of regulations under the later act, if applied to proceedings under the earlier, were indicated but not decided, and where it was said, “section 210 of the Act 1917 is the precursor of section 327(a) of the Act of 1918, and is not at all the analogue of section 327(d).” Both the 1917 and the 1918 Acts provided, § 201 of the former, 40 Stat. 303, and § 301 of the latter, 40 Stat. 1088, for rates at which excess profits taxes should be assessed upon net income. These rates were to be adjusted upon a basis determined by the amount of the invested capital after deducting therefrom assumed normal profits determined by the average profits of pre-war years. It was apparent, however, that there would be cases where the application of this rule would be unjust, and which would require special relief. In the 1917 Act such cases were conceived to be those in which invested capital was too uncertain to be satisfactorily determined. Later, additional cases of injustice resulting from gtrict application were perceived, and ^ remedial sections of the 1918 Act pro. vided for ¡al treatment not only in cases wbere satisfact determination of invested ital could not be made but in otberS) ¡ncluding. cases where income had come from abnormal circumstances and where ex s bf operation had been abnormajiv |ow
Not only was the scope of .special treatment enlarged in the later statute, but there was likewise an enlargement of the circumstances under which and the manner in which it was to be granted. §§ 327 and 328 (a) are printed in the margin.2 It will be observed [§, 327(d)], that where relief is indicated by abnormal conditions affecting capital or income, the discretion of the Commissioner is invoked upon application of the taxpayer and his decision thereon is held not subject to review by a court, Welch v. Obispo Oil Co., supra. It will *233also be noted [§ 328(a)], that the relief to be granted is the application of a tax rate to net income that is determined by the ratio of average tax to average net income of representative corporations engaged in like or similar trade or business. The relief granted under § 210 of the 1917 Act involves no change of tax rate, but only a determination of the deduction for assumed normal profits ascertained by comparison. Thereafter there is applied to excess profits the statutory rates. It is urgently insisted that what the Commissioner did in the instant case was not to ascertain the deduction by comparison with representative corporations, but to find and apply to the present taxpayer’s income an average tax rate, and that this he was not empowered to do under the 1917 Act.
That the Commissioner laclced authority to find and apply an average rate to the taxpayer’s income is clear; that he did so is not demonstrated. The fact that the Commissioner in his letter of March 24tli, 1924, advised the taxpayer that its “excess profits tax” had been computed under § 210, and in 1928 that its “excess profits tax was based” upon a comparison with a group of representative concerns, is not, it seems to us, important. The confusion already noted may have led to the use of phrasing employed in the 1918 Act rather than to that of 1917. Even so, a statement that the tax had been computed under § 210, or that it was based upon comparison, is not inaccurate. It is, of course, the deduction that is determined by comparison and not the tax rate, but the deduction controls the amount of excess income, and consequently the amount of the tax, and so the tax is in a very true sense ultimately, if not immediately, based upon the comparison. A presumption of regularity no doubt attaches to the acts of a public officer in the usual case, but whether it here applies or has weight we need not determine. It was asserted by counsel for the collector both in brief and argument that personal examination of the Commissioner’s files made since the point was raised discloses that the Commissioner followed the procedure outlined in § 210, that he determined the deduction for the taxpayer by the use of data on comparable corporations to arrive at assumed normal income, that he determined thereafter a constructive invested capital for the taxpayer and its affiliates for 1917, and that lie computed the profits tax liability of the taxpayer by the statutory sliding scale rates specified in § 201 of the Act, using the deduction thus determined and the invested capital so constructed. To this declaration there is no answer, and we must assume there can be none.
We come finally to the contention that has given us greatest difficulty. It is urged that even where § 210 has been invoked and a determination made thereunder, the court may still review and correct an erroneous finding of net income because that finding in no way or degree affects the exercise of any discretionary step taken by the Commissioner. We have already seen that under § 210 the decision to make or not to make a deduction for normal income by resort to comparatives may involve no exercise , of discretion. If the Commissioner can satisfactorily determine invested capital there may- be no resort to § 210. If he cannot, we assume it to be his duty under the law to proceed whether requested so to do or not. Having determined to go forward under § 210, he selects the representative taxpayers in a similar trade or business with which to make comparison. Here is involved an exercise of discretion. So much the taxpayer concedes. It is a more limited field for discretionary decision than is permitted under corresponding sections of the 1918 Act, but within it the Commissioner has broader leeway. Under § 210 he is merely required to select representative concerns in a like or similar trade or business. Under § 328(a) he is confined to corporations “similarly circumstanced with respect to gross income, net income, profits per unit of business transacted and capital employed, the amount and rate of war profits or excess profits, and all other relevant facts and circumstances.” It may be granted that in practical application little difference will be perceived in the requirements of the two sections, and that in selecting representative corporations under the 1917 Act all relevant circumstances will bear upon choice, but if the Commissioner’s selection of comparatives is more circumscribed in one act than in the other, the more precise limitation is to be found in that of 1918. We do not assume it to be the province of the court, however, to measure the relative degrees of discretion committed to the Commissioner by the Acts of Congress, but rather to determine whether the exercise of a discretionary judgment is to such extent a factor in the result sought to be judicially set aside, that adjudication is in effect a review by the court of a non-justiciable administrative decision.
*234It is true that the Commissioner having invoked § 210, and having selected his comparable taxpayers, determines only the ratio between their average profits for the current year and their average pre-war profits; that this involves only arithmetic, that he uses the ratio not to determine the rate of tax, but only to determine the amount of income which is arbitrarily assumed to be the taxpayer’s normal or prewar profits, and this again is only arithmetic; So it is urged that by no possibility can a different computation of net income make any difference in fixing a rate for the profits tax, for that rate is fixed by statute and the Commissioner may not change it.
We must not forget, however, that what is in issue is a tax that is claimed to have been illegally collected, and that the suit is for refund of that tax. While a determination of the deduction is of no effect-and is not a factor in the ascertainment of net income, the deduction is a factor in the determination of the amount of income subjected to excess profits taxes and so is a factor in the determination of the tax or of the tax liability, as is the case here. The greater the deduction for normal income the less the amount of excess income, and the smaller the deduction the greater the amount of excess and the higher the tax. The tax rate must not be confused with the tax. The deduction may not affect the rate, insofar as we have now considered it, but id a true sense it bears upon the amount of the tax. Furthermore, excess income when determined is not, however, under § 201 of, the 1917 Act taxed at a uniform rate. The rates themselves are adjusted upon a basis determined by the amount of the invested capital, for after deducting from the net income the assumed normal profits, th,at portion of the excess income which is not more than 15% of the invested capital is taxed at the rate of 20%, that portion which is between 15 and 20% of invested capital is taxed at the rate of 25%, and so on up until all over 33% is taxed at 60%. Since invested capital cannot be satisfactorily determined when resort is had to § 210, a constructive invested capital is built up by using the ratio of average net income of representative corporations to their average invested capital. So is not only the amount of the tax- but so likewise are the zones within which the statutory graduated rates *are to be applied, and perhaps even the rates themselves in the adjustment of their impact on the several zones, dependent in a very true sense upon the deduction arrived at through the exercise of an administrative discretion.
Thus are we finally brought to the narrow question whether the collection of a tax may be judicially set aside, when in its ascertainment a factor of computation is determined by the exercise of~a discretion exclusively committed to the Commissioner, though other factors are subject to review when none are within administrative discretion. In Cleveland Automobile Co. v. United States, supra, 70 F.2d 368, we said that “If the courts cannot review the Commissioner’s determination of a rate or ratio, nor apply a rate used by the Commissioner in computing taxes upon one income base to a base substantially different, the result must be * * * a denial of judicial review * * * ”, and that “Neither the determination, nor the factors used in computation, nor the result itself, is open to review.” No tax rate resulting from the exercise of the Commissioner’s discretion is here as it was there involved, but a result is challenged which is materially affected by a determination committed to his discretion, and it would seem that the rationalization there made applies here. In Welch v. Obispo Oil Co., supra, the question for decision was whether, when a special assessment is made of the profits tax, a court may entertain an action for refund of the accompanying income tax on the ground that income was erroneously determined. Since profits taxes are deducted from income before income tax is computed, the taxpayer contended that the profits tax credit was not so essential an element in computation as to deprive it of all other adjustments. The court said, “By the statute the amount of the income tax payable is dependent upon the amount of the profits tax; and the amount of the profits tax is dependent upon the amount of the income.” 57 S.Ct. 686. ' While here there is no such interdependence of the two factors — the amount of the deduction for assumed normal income not being dependent on total net income, though the amount of the excess profits tax liability is affected by it, yet the court’s rejection of the argument by the Obispo Co. that it would be entitled to a substantial refund even if it were denied deduction of the whole of the profits tax from net income, and the fact that there, as here, the factor was one used as a deduction, affecting the amount of the challenged tax, leads us to conclude that the Obispo Case decides that if one factor in the computation of a tax is committed to the discretionary deter*235mination of the Commissioner, then neither it nor the result in tax determined or collected may be judicially reviewed.
If we are wrong in this, there still remains the question whether if a different computation of net income had been made it would, have been proper to compare the taxpayer with a wholly different group of representative corporations, in which case there might have resulted a wholly different determination of ratio between income and invested capital, and so a different deduction. The taxpayer vigorously asserts that by no possibility could a different computation of net income have made any difference because the comparison would have been with the same representative corporations. The contention is neither demonstrated nor supported, and the concession of discretion in the selection of comparatives would seem to repel it. Within the limited field of the Commissioner’s discretion, his selection of representative corporations is unrestricted, and it would seem that in selecting them he could no more ignore considerations of net income when his search was for a ratio between income and invested capital than he could ignore such considerations when his search was for a proper rate to apply to the taxpayer’s excess profits. So here as in the Obispo Case, and in Heiner v. Diamond Co., 288 U.S. 502, 53 S.Ct. 413, 77 L.Ed. 921, the taxpayer’s net income is an essential factor in determination of liability, and our decision in Joseph & Bros. Co. v. United States, supra, reached the right result, even though present contentions were not there presented to the court.
We do not overlook the fact that under § 210 no element of waiver or the exercise of election is involved such as under the 1918 Act added persuasiveness to the results reached in the cited references. This would perhaps call for further consideration were such elements here wanting. The taxpayer asked for special consideration and was given it. Granted lack of precision in the requests, the Commissioner was nevertheless warranted in interpreting both as a desire for special consideration under § 210. Nor are we unmindful of the fact that the determination here made was in’ such form and under such circumstances as precluded appeal to the Board of Tax Appeals. The granting of a review of administrative decisions, either by the courts or within the taxing organization itself, has always been part of the settled policy of Congress in tax legislation, and failure to provide it in every case is repugnant to our concept of safe-guards against arbitrary action. Some crudeness and inconsistency, leading in exceptional cases even to unfair results, are perhaps to be expected in the painful evolution of a wholly new scheme of taxation.
The judgment below is affirmed.
“Sec. 210. That if the Secretary of the Treasury is unable in any ease satisfactorily to determine the invested capital, the amount of the deduction shall be the sum of (1) ah amount equal to the same proportion of the net income of the trade or business received during the taxable year as the proportion which the ' average deduction (determined in the same manner as provided in section two hundred and three, without including the $3,000 or $6,000 therein referred to) for the same calendar year of representative corporations, partnerships, and individuals, engaged in a like or similar trade or business, bears to the total net income of the trade or business received by such corporations, partnerships, and individuals, plus (2) in the case of a domestic corporation $3,000, and in the case of a domestic partnership or a citizen or resident of the United States $6,000.”
“Sec. 327. That in the following cases- the tax shall be determined as provided in section 328:
“(a) Where the Commissioner is unable to determine the invested capital as provided in section 326;
“(b) * * *
te) * * *
“(d) Where upon application by the corporation the Commissioner finds and so declares of record that the tax if determined without benefit of this section would, owing to abnormal conditions affecting the capital or income of the corporation, work upon the corporation an exceptional hardship evidenced by gross disproportion between the tax computed without benefit of this section and the tax computed by reference to the representative corporations specified in section 328. * * *
“Sec. 328. (a) In the cases specified .in section 327 the tax shall be the amount which bears the same ratio to the net income of the taxpayer (in excess of the specific exemption of $3,000) for the taxable year, as the average tax of representative corporations engaged in a like or similar trade or business, bears to their average net income (in excess of the specific exemption of $3,000) for such year. In the ease of a foreign corporation the tax shall be computed without deducting the specific exemption of $3,000 either for the taxpayer or the representative corporations.
“In computing the tax under this section the Commissioner shall compare the taxpayer only with representative corporations whose invested capital can be satisfactorily determined under section 326 and which are, as nearly as may be, similarly circumstanced with respect to gross income, net income, profits per unit of business transacted and capital employed, the amount and rate of war profits or excess profits, and all other relevant facts and circumstances.” 40 Stat. 1093.