delivered the opinion of the court:
The plaintiff, American Standard, Inc., seeks to recover overpayments of federal income taxes and interest for the taxable year ending December 31, 1966, and the taxable period ending May 31, 1968. The case concerns the proper method for computing the deduction for Western Hemisphere trade corporations (WHTC), I.R.C. §§ 921-922, when such corporations are members of an affiliated group which files a consolidated income tax return, I.R.C. §§ 1501-1563. This case is before the court on the parties’ stipulations of facts. We hold for plaintiff.
Plaintiff is the successor by merger to Westinghouse Air Brake Company (WABCO). During the tax years in issue, WABCO and its affiliates filed consolidated income tax returns. The group claimed a deduction based on the portion of "consolidated taxable income” (as defined in Treas. Reg. § 1.1502-11 (1966)) allocable to the WHTCs multiplied by the fraction provided in I.R.C. § 922. [Hereinafter all Treas. Regs, cited are dated 1966.] The group calculated the amount of consolidated taxable income allocable to the WHTCs by the fractional method with losses.1 In determining the fraction, the group took *415into consideration all members in the particular class including those whose taxable income was a net loss. This included one WHTC and several non-WHTCs which had net losses. Upon audit, the Internal Revenue Service (Service) determined, under its interpretation of Treas. Reg. § 1.1502-25, that the group’s calculation of the deduction was improper because corporations whose taxable incomes resulted in a loss were not to be considered in either the numerator or denominator of the fraction.
Though plaintiff has argued that the Service’s interpretation of its own regulation is incorrect, we find no merit to this contention.2 Thus, plaintiffs method is clearly incorrect and plaintiff is not entitled to recover if Treas. Reg. § 1.1502-25 is valid law controlling in this case.
Plaintiff, however, advances two major theories under which, it contends, the regulation is invalid. Plaintiffs primary theory, which is directed to the substance of the *416regulation, is that the regulation is invalid because it is an arbitrary and unreasonable exercise of the Secretary of the Treasury’s power to promulgate regulations governing consolidated income tax returns. Plaintiffs second theory, which is based on procedural grounds, is that the regulation is invalid because it was not promulgated in accordance with the Administrative Procedure Act (APA).We conclude that the regulation is invalid under both standards.
The concept of dealing with separate affiliated corporations on a consolidated return basis goes back as far as 1917.3 The process by which income tax liability could be determined on the basis of consolidated returns subject to legislative regulations was inaugurated by section 141 of the Revenue Act of 1928, 45 Stat. 791.
The rationale for Congress’ delegation of legislative rulemaking powers was expressed by the Senate Committee report to the 1928 Act as follows:
* * * The committee believes it to be impracticable to attempt by legislation to prescribe the various detailed and complicated rules necessary to meet the many differing and complicated situations. Accordingly, it has found it necessary to delegate power to the commissioner to prescribe regulations legislative in character covering them. The standard prescribed by the section keeps the delegation from being a delegation of pure legislative power, and is well within the rules established by the Supreme Court. * * * [S. Rep. No. 960, 70th Cong., 1st Sess. 15 (1928).]
As noted in the committee report, the promulgation of consolidated return regulations is a legislative function. This court recognized this in Union Elec. Co. of Mo. v. United States, 158 Ct. Cl. 479, 486, 305 F.2d 850, 854 (1962), where the court concluded that consolidated return regulations "unlike ordinary Treasury Regulations, are legislative in character and have the force and effect of law.” Though legislative regulations are law, they are good law only if enacted in accordance with the authority vested in *417the Treasury by the enabling Act. A consolidated return regulation is invalid if it be inconsistent with the Act. See, e.g., American Trans-Ocean Nav. Corp. v. Commissioner, 229 F.2d 97 (2d Cir. 1956); Commissioner v. General Mach. Corp., 95 F.2d 759 (6th Cir. 1938); Corner Broadway-Maiden Lane, Inc. v. Commissioner, 76 F.2d 106 (2d Cir. 1935).
The delegation of rulemaking power is presently found in I.R.C. § 1502, as follows:
The Secretary shall prescribe such regulations as he may deem necessary in order that the tax liability of any affiliated group of corporations making a consolidated return and of each corporation in the group, both during and after the period of affiliation, may be returned, determined, computed, assessed, collected, and adjusted, in such manner as clearly to reflect the income-tax liability and the various factors necessary for the determination of such liability, and in order to prevent avoidance of such tax liability.
Thus, the Code grants to the Secretary broad legislative authority governing the manner in which a group’s tax liability is determined when a consolidated return is filed. But this power must be construed in terms of Congress’ purpose that both the group’s and its individual member’s actual tax liability be found under the regulations "in such manner as clearly to reflect the income-tax liability * * * and in order to prevent avoidance of such tax liability.” I.R.C. § 1502. Income tax liability is not imposed by the Secretary’s regulations, but by the Internal Revenue Code. Thus, the purpose of the delegation of power to the Secretary can be stated more broadly as the power to conform the applicable income tax law of the Code to the special, myriad problems resulting from the filing of consolidated income tax returns. Though there may be many reasonable methods to determine a group’s tax liability and the Secretary’s authority is absolute when it represents a choice between such methods, the statute does not authorize the Secretary to choose a method that imposes a tax on income that would not otherwise be taxed. See Joseph Weidenhoff, Inc. v. Commissioner, 32 T.C. 1222, 1242 (1959).
We note that, as discussed in part II, infra, the Secretary did not provide any statement of the basis and purpose of *418the regulation. We are thus unaware of any special factual or legal problem caused by the filing of consolidated returns with which the method adopted by the regulation was meant to deal. Defense counsel has added nothing to enlighten us. All we have is the obvious inference from Treas. Reg. § 1.1502-25 and the statute under which it was promulgated that the regulation represents the Secretary’s choice of a reasonable method to determine a group’s tax liability or to prevent avoidance of that liability. We acknowledge that the Secretary is the expert in this complex field and that the regulation must be sustained unless it is clearly unreasonable. We must determine, however, the question of reasonableness for ourselves.
Our analysis must begin with an examination of the purposes and policy of the statute and regulations. The basic purpose behind allowing corporations to file consolidated returns is to permit affiliated corporations, which may be separately incorporated for various business reasons, to be treated as a single entity for income tax purposes as if they were, in fact, one corporation. Therefore, as provided by regulation, Treas. Reg. § 1.1502-2(a), the tax is computed solely on the basis of consolidated taxable income.
The single entity framework does not mean that all items of income, deductions, and credit for the affiliated corporations are combined into single accounts as if the corporations were one. The consolidated return regulations, in fact, primarily deal with the affiliated corporations as separate corporate entities. This treatment as separate entities under the regulations is shown by the first item in computing consolidated taxable income which is "separate taxable income.” Treas. Reg. § 1.1502-ll(a). Separate taxable income of a member is basically the taxable income of a member which would be reported if separate income tax returns were filed, with two types of adjustments. Treas. Reg. § 1.1502-12. The first type is for certain items that are taken into consolidated taxable income on a consolidated basis. This includes any net operating loss deduction, capital gains (and losses), I.R.C. § 1231 net losses, charitable contributions, I.R.C. § 922 WHTC deduction, I.R.C. §§ 243-245 dividends received deductions, and I.R.C. § 247 dividends paid deductions. *419Treas. Reg. § 1.1502-12(h), (jMn). These items are all taken into consideration on a consolidated basis along with members’ separate taxable income in order to determine consolidated taxable income. Treas. Reg. § 1.1502-11. The second type of adjustment includes various adjustments and rules which change the amount, allowability, or manner of taking certain items, permitted when filing separate returns, into separate taxable income.
It is, however, the first type of exception to normal, separate returns that we are particularly concerned with here because the WHTC deduction is a consolidated deduction. The consolidated items are shared or pooled items which reflect the single entity concept which is what consolidation is all about. There are few consolidated items, however, demonstrating the Treasury’s decided preference for the multiple entity concept over the single entity concept of consolidated returns.
For the most part, consolidated items are those categories of income and deductions which are common to all corporations. There are two exceptions to this general statement and these are the WHTC deduction (section 1.1502-ll(f)) and the special public utilities deduction (section 1.1502-11(h)). Both deductions represent special, favorable tax treatment afforded enterprises Congress concluded were somewhat unique, requiring tax reduction incentives. The deduction for public utility corporations, provided in I.R.C. § 247, is an amount represented by a fraction of the lesser of dividends paid on preferred stock of that corporation or the net taxable income of that corporation. The WHTC deduction, provided in I.R.C. § 922, is an amount represented by a fraction of the net taxable income of that corporation. For the tax period in issue in this case the fractions were the same in both cases.
In many cases, however, Congress excluded certain types of corporations from the privilege of filing consolidated returns.4 Congress decided either that it simply did not wish to confer the benefit of consolidated treatment on these organizations or that the special tax treatment was incompatible with the consolidated scheme.
*420Congress did not exclude, however, WHTCs or public utilities from consolidated treatment. In one case, Congress specifically showed its concern that WHTCs and public utilities would not be harmed by consolidated treatment by specifically exempting the portion of consolidated taxable income attributable to those corporations from the 2-percent surcharge on consolidated taxable income.5
WHTCs and public utilities are only unique, however, in that they receive a special deduction (which has the effect of a tax rate reduction) computed on the basis of their taxable incomes. Otherwise, the computation of their taxable income is subject to the same rules under the Code as normal corporations. To the extent of the initial calculation of taxable income, it is a clearly reasonable answer to the needs of consolidated returns to treat them in the same way as other corporations. Though such treatment can change the character and/or amount of income and deductions because of the consolidation of certain items,6 this is the treatment all corporations are subject to when they elect the privilege of filing consolidated returns.
The deduction for WHTCs and public utilities are concededly unique, however, being determined by the amount of taxable income attributable to the particular subgroup (either WHTC or public utility). What is determined by treating these corporations just like every other corporation before the deduction is the separate taxable income of each member and the consolidated taxable income (without the deduction) or the consolidated items of income and deductions of the entire group. Separate taxable income, which is calculated on an individual corporate basis, could not be used alone as the base for the deduction because it is only a component of the individual *421corporation’s taxable income. It excludes the consolidated items attributable to the individual corporations which must be taken into consideration in determining taxable income. Consolidated taxable income as it stands could not be used as the base for the deduction because it includes the income of all corporations in the group including those not members of the particular subgroup. The problem, which is the source of dispute in this case, is to find a reasonable method to "break out” taxable income properly attributable to the subgroup to form the base for the deduction.
Though the objective is the same, the methods by which the regulations propose to determine the taxable income of the particular subgroup of corporations for the section 922 and section 247 deductions are quite different. For purposes of the section 247 deduction, Treas. Reg. § 1.1502-27 adopts the aggregate method with losses.7 This is the aggregate of the separate taxable incomes (or losses) of each member of the public utility subgroup adjusted for the amount of consolidated items attributable to each member of the subgroup.8 For purposes of the section 922 deduction, Treas. Reg. § 1.1502-25 adopts the fractional method without losses.9 This is the aggregate of the taxable incomes of each member of the WHTC subgroup over the aggregate of the taxable incomes of each member of the entire group, calculated in the same way as under section 1.1502-27 but excepting any corporation with taxable income resulting in a loss, times consolidated taxable income.10
The language governing the allocation of consolidated items attributed to a particular member is the same under both regulations. Each consolidated item is calculated separately for each member. The principle behind the allocation is that the amounts actually contributed by the individual member and actually taken into account in computing consolidated taxable income are allocated to *422that member.11 As to the consolidated deductions, only those members which incurred the expenses have any deduction attributed to them, and then only to the extent that full deduction was allowed in computing consolidated taxable income.12 As to the capital and section 1231 gains or losses, the method takes the actual amount of net gain and/or loss incurred by the member reduced by the amount of consolidated capital loss attributable to that member not taken into account or consolidated capital loss carryover taken into account in that year.
Great effort has been expended in analyzing the method for the allocation of consolidated items to an individual member because it is the method of allocating the consolidated items of deduction that has been basically adopted by Treas. Reg. § 1.1502-25 for allocating "the portion of the consolidated taxable income” attributable to WHTCs to the subgroup of WHTCs. As the regulation allocates the consolidated net operating loss deduction, the consolidated charitable contributions deduction, and the consolidated dividends received deduction on the basis that the item was incurred or received by the member, the regulation allocates consolidated taxable income solely on the basis of the member’s contribution of positive net income.
The effect of this allocation is considerably different from the aggregate method with losses. This latter method treats the privileged type of corporations as a subgroup netting *423gains and losses only within the subgroup. The fractional method without losses treats the net losses of all loss corporations as the common property of the entire group, allocating them on the basis of the positive taxable income of member corporations. The effect of the disputed regulation is to decrease the income base for the WHTC deduction proportionately to the amount of the loss of non-WHTCs excluded from the fraction, and to increase the income base for the WHTC deduction proportionately to the loss of WHTCs excluded from the fraction. Net losses of corporations, though not earned or attributable to members with positive net income, are treated as a consolidated deduction and apportioned among the members of the group on the basis of their positive contribution of taxable income. Thus, the regulation has set up a direct relationship between the computation of the WHTC deduction and the post-deduction consolidation of gain and loss.
The aggregate method with losses (or the fractional method with losses) which has the effect of netting income of the subgroup was adopted for purposes of the section 247 deduction and was the preferred method for the WHTC deduction under Treas. Reg. § 1.1502-31A(a)(15) for tax years prior to 1966.13 Defendant has not directly contested the reasonableness of the method employed in Treas. Reg. § 1.1502-27.14 Defendant does argue that there can be more than one reasonable method, and the fractional method without losses is one. Defendant’s theory is that the choice is reasonable if it operates in a neutral fashion; that is, it does not necessarily affect affiliated groups in an adverse manner. In filing consolidated returns, defendant argues, affiliated groups must "take the bitter with the sweet.” Georgia-Pacific Corp. v. Commissioner, 63 T.C. 790, 802 (1975).
Though we agree with defendant that the negative effect plaintiff complains of is solely a product of a taxpayer’s particular situation and that the regulation can just as easily provide an increased benefit from the deduction, we do not believe that defendant’s theory provides the easy solution to this problem. The fractional method without *424losses is a radical departure from past practice dealing with the special tax treatment of corporations within the consolidated framework. The "taxable income” of the subgroup of WHTCs under this method varies dramatically from that found under the aggregate method with losses.
It cannot seriously be contended that such a departure in methodology and resulting amount is merely a mechanical or technical method of adjusting individual returns to the consolidated return format.15 Critically, the regulation changes the conceptual basis upon which Congress permitted the deduction in section 922. We do not accept the view that this departure was within the Treasury’s power.
The deduction granted by section 922 was designed to foster American business within the Western Hemisphere. The purpose of the legislation "was to ameliorate somewhat the unfavorable situation in which domestic corporations had to compete with foreign corporations in the Western Hemisphere. S. Rep. No. 1631, 77th Cong., 2d Sess. 32 (1942).” Caterpillar Tractor Co. v. United States, 218 Ct. Cl. 517, 524, 589 F. 2d 1040, 1044 (1978). A corporation had to meet the rigid requirements of section 921 to qualify. Arm’s-length pricing was required between a related party and a WHTC in order that the benefit of the statute be strictly conferred on income earned by the WHTC. It is clear that the purpose of the deduction was to ensure that WHTCs’ products and services were competitively priced in Western Hemisphere trade.
Treas. Reg. § 1.1502-25 clearly defeats this intended inducement to American business. Sharing the losses of non-WHTC corporations with the profits of WHTC corporations for purposes of the deduction directly decreases the profitability of Western Hemisphere trade and concomitantly the price for which goods and services are provided. Congress specifically conferred both the privilege of filing consolidated returns and the special deduction on WHTCs, and reduced the onus on a taxpayer for having to incorporate its Western Hemisphere business separately. The regulation penalizes WHTCs which are affiliated with *425a group that desires to file consolidated returns. Congress granted a deduction to the earnings of WHTCs; the regulation changes this concept by reducing WHTC earnings by the losses of domestic or non-WHTC enterprises. In doing this, it is in direct conflict with sections 922 and 1502.16
The approach that permits, within the consolidated framework, the aggregation of profits and losses within the subgroup of corporations with a special benefit or incentive but does not permit the sharing of losses from without the group is aided by a prior decision of this court. In Sinclair Oil Corp. v. United States, 183 Ct. Cl. 92, 392 F.2d 249 (1968), the court dealt with the computation of the proper exclusion from consolidated taxable income of WHTC and utility corporation income for purposes of the 2-percent add-on tax applied solely to consolidated returns. In that case, the utility corporations had a net gain, whereas the WHTCs had a net loss. The question presented to the court was whether (1) each group (WHTCs and utilities) should be aggregated separately and separately offset against consolidated taxable income or (2) both groups should be aggregated or netted and the difference only offset against consolidated taxable income. The Government relied on a revenue ruling as supporting proposition (2). The court found, however, that proposition (1) was the correct result because the exemption was for each individual type of corporation. The court found that it was proper to *426aggregate within the type (WHTC or utility) but to aggregate between the types would effect a denial of the intended exemption.
Sinclair Oil Corp. did not present a question of conforming the Code to consolidated returns but instead presented a question unique to consolidated returns. Nevertheless, the rationale is equally applicable because the WHTC deduction (and the public utility deduction), like the exclusion, is reserved for and for the benefit solely of WHTCs. The fact that the profits of the public utilities are later offset against the losses of WHTCs made no difference. Further, the statute itself is a further indication that Congress did not intend that WHTCs or public utilities be penalized by being included in a consolidated return.
The net effect of what we have decided so far is that the part of Treas. Reg. § 1.1502-25(c) which provides that, in computing the "fraction” of consolidated taxable income which forms the base for the WHTC deduction, loss corporations will be deemed to have zero income is invalid. Plaintiff has stated that it would be satisfied with either the fractional method with losses or the aggregate method with losses. Plaintiff adopted on its income tax return the fractional method with losses. Defendant argúed, however, in response to plaintiffs attack on the reasonableness of the regulation’s method, that the fractional method with losses presents an unreasonable approach.17 The only argument offered by defendant to the effect that the fractional approach utilizing losses is unreasonable is that the possibility exists that the fraction could be greater than one (1), therefore resulting in more consolidated taxable income than actually exists being allocated to the WHTCs. Thus, defendant argues that "the portion of consolidated taxable income attributable to those members of the group which are Western Hemisphere trade corporations,” Treas. Reg. § 1.1502-25(c)(l) [emphasis added], cannot be larger than the whole.
*427Defendant’s argument is unfounded. The fractional method utilizing losses, with slight differences,18 gives the same result as the aggregate method with losses required by Treas. Reg. § 1.1502-27 for determining the "portion of consolidated taxable income” attributable to public utilities for purposes of the section 247 deduction. Under the aggregate method with losses, the "portion of the consolidated taxable income” attributable to the corporations to which the deduction is given by section 922 or section 247 is the actual net taxable income contributed by the group, and, therefore, can be greater than actual consolidated taxable income. Defendant’s argument questions, in effect, the reasonableness of Treas. Reg. § 1.1502-27. We conclude, however, that utilizing the actual net taxable income of the subgroup for which the deduction is intended, complies fully with section 922, section 247, and the consolidated return statutes.
II
Plaintiffs alternative argument is that the regulation was not promulgated in compliance with the Administrative Procedure Act (APA), and is, therefore, invalid. Plaintiff alleges that the requirement of notice 30 days before the promulgation of the regulation and a statement of the basis and purpose of the regulation were not provided, in violation of the APA.
5 U.S.C. § 553(b), (c) and (d) requires that notice of proposed rulemaking be published in the Federal Register 30 days prior to the adoption of a regulation. Such notice includes "either the terms or substance of the proposed rule or a description of the subjects and issues involved.” 5 U.S.C. § 553(b)(3). The purpose of this procedure is "to assure fairness and mature consideration of rules of general application,” NLRB v. Wyman-Gordon Co., 394 U.S. 759, 764 (1969), and to give affected members of the public an opportunity to comment. See Texaco, Inc. v. FPC, 412 F.2d 740, 744 (3d Cir. 1969). A substantive rule is *428invalid if it is not promulgated in accordance with the notice requirements of the APA. See Chrysler Corp. v. Brown, 441 U.S. 281 (1979).
On September 8, 1966, notice of proposed rulemaking and a regulation covering the consolidated WHTC deduction were published in the Federal Register. 31 Fed. Reg. 11845, 11848 (1966). The problem presented for resolution is whether this proposed regulation gave the public adequate notice of the regulation which was adopted on December 30, 1966. To analyze this problem properly, some background information is necessary.
Prior to January 1, 1966, the section 922 deduction for WHTCs’ filing consolidated returns was provided by Treas. Reg. § 1.1502-31A(a)(15). This regulation provided for the computation of the deduction as follows:
(15) Consolidated section 922 deduction. The consolidated section 922 deduction, relating to Western Hemisphere trade corporations, shall be that portion of the consolidated taxable income attributable to those members of the affiliated group which are Western Hemisphere trade corporations (computed without regard to the consolidated section 922 deduction) multiplied by the fraction specified in section 922(2).
On October 1, 1965, the Treasury gave notice in the Federal Register of a proposal to revamp generally the consolidated return regulations which included a proposed regulation dealing with the consolidated WHTC deduction. 30 Fed. Reg. 12564, 12581 (1965). Proposed Treas. Reg. § 1.1502-25 provided for the computation of the deduction as follows:
(a) In general. The consolidated section 922 deduction for the taxable year shall be determined by multiplying the fraction specified in section 922(2) by the aggregate of the taxable income (or loss), as determined under paragraph (c) of this section, of those members of the group which are Western Hemisphere trade corporations for such year.
Subsection (c) provided:
(c) For purposes of paragraph (a) of this section, the taxable income (or loss) of a member which is a Western Hemisphere trade corporation shall be determined under § 1.1502-12, adjusted for the following items taken into *429account in the computation of consolidated taxable income: * * *.19
Thus, the initial proposal for the WHTC deduction adopted the aggregate method with losses.
On September 8, 1966, the first proposed regulation (Oct. 1, 1965) was withdrawn and a second proposed regulation (Sept. 8, 1966) was promulgated. This second version of section 1.1502-25 provided that the base for the deduction was "that portion of the consolidated taxable income attributable to those members of the group” which are WHTCs. Such portion was defined as:
* * * an amount equal to the consolidated taxable income multiplied by a fraction, the numerator of which is the sum of the separate taxable incomes of those Western Hemisphere trade corporations having separate taxable income, and the denominator of which is the sum of the separate taxable incomes of all of the members having separate taxable income. For purposes of this paragraph, the separate taxable income of a member shall be determined under § 1.1502-12, adjusted for the following items taken into account in the computation of consolidated taxable income: * * *.20 [Proposed Treas. Reg. § 1.1502-25 (Sept. 8,1966).]
Treas. Reg. § 1.1502-12(o), adopted on September 8, 1966, provided that "[t]he term 'separate taxable income’ shall include a case in which the determination under this section results in an excess of deductions over gross income.” Thus, in plain English, the regulation includes a case where a corporation has a net loss.
It is defendant’s contention that the language in the second proposal "the separate taxable income [of the members] * * * having separate taxable income” presages the Treasury’s position as found in the adopted version that— *430Defendant asserts that the doubling of the terms leads to the conclusion that the words "having separate taxable income” mean only positive taxable income (i.e., gain situations).
*429[i]f the computation of the taxable income of a member under this subparagraph results in an excess of deductions over gross income, then for purposes of subpara-graph (1) of this paragraph such member’s taxable income shall be zero. [Treas. Reg. § 1.1502-25(c)(2).]
*430We cannot say that defendant’s interpretation is groundless, for there is a slight ambiguity in the language. The proposal’s direct reference to section 1.1502-12, however, strongly supports plaintiffs allegation that not only was the public unaware that losses would not be taken into the computation, but, in fact, that the proposal affirmatively represented that they would be taken into consideration. In light of the fact that the first proposal (Oct. 1,1965) clearly took losses into account, it is unlikely that the public would have been put on notice by the second proposal that the Treasury intended to change diametrically its substantive approach. Granting the confusion that the doubling of the terms results in, defendant’s position would result in either two different meanings being given to the same words (one in compliance with section 1.1502-12(o)’s definition and one not) or one meaning which is not in compliance with section 1.1502-12(o). Defendant’s interpretation, while not ridiculous, does stretch the imagination, and defendant’s argument does not convince us that the proposal did give the public fair notice of the intendment of the final form of the regulation. Therefore, we hold that the notice provision of the APA was violated in this case and the regulation is invalid.
The APA also requires that when a rule is adopted, a statement of its basis and purpose shall accompany its publication. 5 U.S.C. § 553(c). No such statement accompanied the promulgation of Treas. Reg. § 1.1502-25. Plaintiff contends that the regulation is thus invalid. See National Welfare Rights Organization v. Mathews, 533 F.2d 637 (D.C. Cir. 1976).
The purpose of requiring a statement of the basis and purpose is to enable courts, which have the duty to exercise review, to be aware of the legal and factual framework underlying the agency’s action. See SEC v. Chenery Corp., 318 U.S. 80, 94 (1943); Wright, The Courts and the Rulemaking Process: The Limits of Judicial Review, 59 Cornell L. Rev. 375, 395 (1974), quoted in National Welfare Rights Organization v. Mathews, supra, 533 F.2d at *431649. Derived from the very rationale of the requirement is an exception to its overtechnical application, and that is, when the basis and purpose of the rule is inherent in the rule and the enabling statute, then no separate statement is required. See Alabama Ass’n of Ins. Agents v. Board of Gov. of Fed. Reserve System, 533 F.2d 224, 236-37 (5th Cir. 1976); New York Foreign Frgt. F. & B. Ass’n v. Federal Maritime Comm’n, 337 F.2d 289, 296 (2d Cir. 1964). Defendant contends it is obvious why the regulation was promulgated; it was promulgated to adjust the Code provisions to the special problems incurred with consolidated returns.
The reason the lack of a rationale renders a regulation invalid is that a court cannot evaluate the reasonableness of a regulation without a statement of the purpose and basis. See National Welfare Rights Organization v. Mathews, supra. Consolidated return regulations differ from some legislative rules in that they clarify and establish a uniform system for reporting tax liability on a consolidated basis. Unlike some legislative rules that are promulgated pursuant to vague statutory commands and within a vague statutory framework which make review difficult, if not impossible, consolidated return regulations are reviewed in terms of a large, comprehensive code and a sizeable body of decisional law.
Though determining the reasonableness of the regulation is difficult in this context, and a challenge to its reasonableness by plaintiff must overcome significant obstacles, it is not impossible in our situation, as demonstrated by part I of our opinion. Therefore, we conclude that the purpose and basis requirements of the APA were not violated in this case but the notice requirement was.
CONCLUSION
In view of all the circumstances, we hold that part of Treas. Reg. § 1.1502-25(c) which excludes all corporations with net losses from the fraction determining the portion of consolidated taxable income is invalid in violation of the delegation of rulemaking power and of the notice requirement of the Administrative Procedure Act. Accordingly, judgment is entered for plaintiff with interest as provided *432by law and the case is remanded to the trial division for proceedings under Rule 131(c) to determine the amount of recovery.
At this point it will be helpful to set out the various methods that will be discussed in this opinion and the manner of their calculation. All methods have a common purpose which is to determine the amount of income properly allocated to the members of the affiliated group which are WHTCs for the taxable year. This amount represents the "base” to which the fraction specified in section 922 is applied to determine the WHTC deduction.
There are two basic methods for determining this base — the aggregate method and the fractional method. The fractional method begins at the same point as the aggregate method and includes as a component of that method the base achieved by use of the aggregate method.
(I) The aggregate method provides that the base is the aggregate (or net) of the taxable incomes of each member of the affiliated group which is a WHTC. It disregards those members of the group which are not WHTCs. The taxable income of a member is computed on the basis of the member’s "separate taxable income” (as defined in Treas. Reg. § 1.1502-12) to which adjustment is made for those consolidated items of income and deductions which are actually attributable to it. (Consolidated items are defined in Treas. Reg. § 1.1502-11 (1966)). The aggregate of the taxable incomes of all WHTCs can take one of two forms:
(A) Under one variation, the aggregate of the taxable incomes of all WHTCs, including both those with net gain (gross income exceeding deductions) and those with net losses (deductions exceeding gross income). [Hereinafter the aggregate method with losses.]
(B) Under a second variation, the aggregate of only those WHTCs with net gain, *415excluding those with a net loss, is computed. [Hereinafter the aggregate method without Zosses.]
(ED The fractional method ascertains the base by taking a fraction of the actual consolidated taxable income (without regard to the WHTC deduction). The fraction’s purpose is to determine the percentage of the consolidated taxable income properly allocated to the subgroup of WHTCs. The computation is illustrated as follows:
Aggregate of WHTC taxable income
Aggregate of all members1 taxable income
Consolidated taxable income (without WHTC deduction)
The numerator of the fraction represents the same figure (or base) reached by the aggregate method. The denominator of the fraction is the aggregate or net of all members of the affiliated group’s taxable income. The taxable income of a member is computed on the basis of the member’s separate taxable income to which adjustment is made for those consolidated items of income and deduction which are actually attributable to it. [As noted later, the fractional method adopted by the Secretary is not an entirely pure form of this method.] The computation of the aggregates in both the numerator and denominator can take one of two forms:
(A) Under one variation, the aggregate in both the numerator and denominator of the fraction take both those corporations with net gain and those with net loss in the particular corporate category into account. [Hereinafter the fractional method with losses.]
(B) Under a second variation, those members in the particular corporate category with net losses are excluded in both the numerator and denominator of the fraction. [Hereinafter the fractional method without losses.]
Plaintiff basically argued that under the fractional method without losses, losses were not to be included in consolidated taxable income. "Consolidated taxable income” is a term of art defined in Treas. Reg. § 1.1502-11 and there is no indication that this definition was varied by the regulation.
The practice was first instituted by regulation, Treas. Reg. 41, art. 78 (1917), only for purposes of the excess profits tax.
In two cases, Congress permitted corporations with the same special tax treatment to file consolidated returns. I.R.C. § 1504(c) (insurance corporations); I.R.C. § 1504(e) (tax-exempt organizations).
The 2-percent surcharge, originally imposed by section 1503(a) of the Internal Revenue Code of 1954, has since been repealed. In subsection (b), a full exemption from the tax was provided for WHTCs and public utilities.
One example of this can be seen in consolidating the treatment of I.R.C. § 1231, which requires the netting of gains and losses from the sale of certain property and confers capital asset treatment to all items if the net results in a gain and ordinary income treatment if the net results in a loss. Such determines the character and therefore the amount of gain or loss, thus affecting the taxable income of a member. Likewise, consolidation of the charitable deduction can either increase or decrease the amount allowed.
See note 1 supra.
One consolidated item is not included in the aggregate taxable income of the subgroup, and that is the consolidated WHTC deduction. Treas. Reg. § 1.1502-27(b). It is possible that a public utility can be a WHTC. See I.R.C. § 1503(b).
See note 1 supra.
One consolidated item is not included in the taxable income of any member and that is the consolidated section 247 deduction. See note 8 supra.
The amount of a consolidated item attributable to a member is the amount actually contributed limited by the amount actually allowed as a deduction from consolidated taxable income in that taxable year. This is probably calculated by the same method provided by Treas. Reg. § 1.1502-79 for determining the amount of such items that can be carried over by a member to separate taxable years. The basic method provided for the consolidated net operating loss deduction, charitable contributions deduction, and consolidated dividends received deduction is a fraction, the member’s actual contribution over the entire group’s contribution, times the consolidated amount taken into account in computing consolidated taxable income in that year. Capital and section 1231 gains or losses, however, appear to be calculated on the basis of the actual gains or losses incurred by the member in that year reduced by that year’s consolidated net capital loss not taken into account or consolidated net capital loss carryover which is taken into account in that year. See Treas. Reg. §§ 1.1502-25(c)(2)(ii), (iii), (iv), 1.1502-79(b)(2).
Limitations on the amount of deduction are set at the consolidated level so that the deduction allowed for that year can be less than the total amount of the members’ actual expenditures or more than the amount that could otherwise be deducted if separate returns were filed.
See Rev. Rul. 58-618,1958-2 C.B. 430, Rev. Rul. 56-316, 1956-2 C.B. 597. Contra, Rev. Rul. 68-529, 1968-2 C.B. 395.
See text at note 16 infra.
An example of this is found in Salem Packing Co. v. Commissioner, 56 T.C. 131 (1971), where a regulation requiring consistent accounting methods be used by all members of the group was upheld even though a Code provision specifically allowed a farming corporation to use the cash method.
The only distortion caused by consolidation on the WHTC deduction which we perceive is that a greater benefit can be obtained by WHTCs’ having alternative profit and loss years, and the losses of WHTCs are offset against the profits of all other corporations in that year and are not carried over to years in which there is WHTC income except to the extent of consolidated net operating losses attributable to the WHTCs. If separate returns were filed, such losses would be carried over to other tax years causing the deduction to be recomputed on the basis of the recomputed taxable income. Cf. Motors Ins. Corp. v. United States, 208 Ct. Cl. 571, 530 F.2d 864 (1976) (recomputation of the foreign tax credit).
The tax avoidance incentive is reduced, however, by the very fact of consolidation and because WHTCs are required to qualify as such for 3 years prior to obtaining the deduction. I.R.C. § 921(1). Those methods utilizing WHTC losses to offset WHTC gain in the taxable year partially ameliorates this undesirable effect and restricts the possibilities of tax avoidance. In any event, the fractional method without losses, by sharing WHTC losses with non-WHTCs, increases the tax avoidance possibilities. The penalty imposed by the sharing of non-WHTC losses with WHTCs bears no relation to the problem outlined above and could only fortuitously result in prevention.
This argument is, in effect, defendant’s last line of defense for the regulation. It appears that defendant’s view is that if plaintiffs method is unreasonable, then the reasonableness of the Treasury’s method cannot be considered. We do not, of course, agree with defendant that plaintiffs approach is unreasonable. But, even if we did, we do not believe that this theory alone could sustain the regulation if some reasonable approach could be found. Upon the finding, however, that there were imperfections or problems with all the methods proposed, the regulation might be sustained as a reasonable selection among concededly imperfect approaches.
The only significant difference is the treatment of the consolidated section 247 deduction in the fraction. Plaintiff does not appear to include any public utilities, so the fractional method with losses and the aggregate method with losses reaches the same result.
The method for computation of a member’s taxable income in the proposed regulation, i.e., separate taxable income adjusted for certain consolidated items, was employed in all proposals by the Treasury and was adopted in Treas. Reg. § 1.1502-25.
See note 19 supra.