Opinion

I.

INTRODUCTION

Amdahl Corporation (Amdahl), a multinational business, sought a refund of $3,390,388 in taxes arising from assessments by the Franchise Tax Board (FTB) for the tax years 1988, 1989, 1991 and 1992. In the underlying tax refund action, Amdahl alleged that the FTB improperly assessed taxes against it for these years based on its erroneous treatment of dividends distributed by Amdahl’s first-tier and second-tier subsidiaries. Specifically, Amdahl claimed that the FTB: 1) incorrectly treated tax credit payments from Amdahl’s United Kingdom subsidiaries as nondividend income; 2) incorrectly computed the inclusion ratio used to determine how much of the income of Amdahl’s foreign subsidiaries should be included in the combined income of the “water’s-edge” group; and 3) incorrectly applied Revenue and Taxation Code sections 25106 and 24411 to dividends received from Amdahl’s foreign subsidiaries. Amdahl also alleged that the FTB erred in concluding that its tax refund action, as it relates to tax year 1988, was not timely filed. Finally, Amdahl claimed that California’s “water’s-edge” method of apportioning the combined income of a unitary business group for tax purposes improperly discriminates against foreign subsidiaries in favor of domestic subsidiaries, in violation of the commerce clause of the United States Constitution (U.S. Const., art. I, § 8).

After the State Board of Equalization (the SBE) rejected Amdahl’s arguments, it filed the underlying action in the superior court. The matter was tried to the court largely on stipulated facts. Amdahl prevailed on each issue except for its constitutional claim.

We now consider three consolidated appeals. In appeal No. A101101, the FTB appeals from the superior court judgment in the underlying action; in appeal No. A102558, the FTB appeals from a postjudgment order granting Amdahl attorney fees; and in appeal No. A101203, Amdahl has cross-appealed the constitutional issue. We affirm in all respects.

II.

Facts and Procedural History

Amdahl, headquartered in Sunnyvale, California, is a Delaware corporation engaged in the business of providing integrated computer solutions to meet the needs of many of the largest users of information technology in the world. Amdahl operates extensively throughout the United States, Europe, and Asia, often through various subsidiaries and holding companies.

The FTB is the state agency empowered to determine the California tax liability of multistate or multinational corporations, such as Amdahl. (§ 23001 et seq.) The FTB has the authority to audit the operations of such corporations. (§ 26423; Franchise Tax Board v. Firestone Tire & Rubber Co. (1978) 87 Cal.App.3d 878 [ 151 Cal.Rptr. 460 ].)

The issues in this case may be more easily understood if Amdahl’s corporate structure and several rather esoteric tax terms are first explained. During 1988 through 1992, the tax years at issue, Amdahl and its subsidiaries, including Amdahl International Corporation (AIC); Amdahl (U.K.) Ltd.; Amdahl International Management Services (AIMS); Amdahl Ireland, AOCC; Amdahl Lease BY; and Amdahl Netherlands BY, were treated as engaged in a single unitary business. (See §§25101, 25102; Edison California Stores v. McColgan (1947) 30 Cal.2d 472, 479 [ 183 P.2d 16 ].) A unitary business has been judicially defined as one in which the following factors are present: (1) unity of ownership; (2) unity of operations, as evidenced by central accounting, purchasing, advertising, and management divisions; and (3) unity of use in a centralized executive force and general system of operation. (D ental Ins. Consultants, Inc. v. Franchise Tax Bd. (1991) 1 Cal.App.4th 343 [ 1 Cal.Rptr.2d 757 ]; Butler Brothers v. McColgan (1941) 17 Cal.2d 664 [ 111 P.2d 334 ].) A unitary business is one that receives income “from or attributable to sources both within and without the state . . . .” (§ 25101.) If a unitary business exists, taxes are apportioned based on property, payroll, and sales to allocate to California for taxation “its fair share of the taxable values of the taxpayer . . . .” (Butler Brothers v. McColgan, supra, 17 Cal.2d at pp. 667-668.)

In 1986, California passed legislation permitting taxpayers to make a “water’s-edge” election. Under the water’s-edge method, qualified taxpayers determine their income derived from or attributable to California by including only a formula-based allocation of the income from California and United States (U.S.)-based affiliated entities. Essentially, California’s water’sedge method is an accepted accounting method using the United States as the jurisdictional boundary. Thus, generally speaking, the effect of a water’s-edge election is for the taxpayer to account only for the income and apportionment factors of affiliates incorporated in the United States, subject to a number of exceptions for certain types of income produced by foreign affiliates, one of which is at issue in this case. The relevant exception, section 25110, subdivision (a)(6), adds to the water’s-edge group a portion of the income and apportionment factors of affiliates that are controlled by foreign corporations (CFC’s) if all or part of their income is “Subpart F” income.

Subpart F income gets its name from Subpart F of the Internal Revenue Code (26 U.S.C.), as defined in Internal Revenue Code section 952. It includes certain forms of passive income earned by CFC’s—for example, dividends, income from bank accounts, and stock investments. “Subpart F of the Internal Revenue Code (sections 951-964) was enacted to deter taxpayers from using foreign subsidiary corporations to accumulate earnings in countries that impose no taxes on accumulated earnings. [Citations.]” (R.E. Dietz Corp. v. U.S. (2nd Cir. 1991) 939 F.2d 1, 6 .) Thus, as discussed more fully in a later section of this opinion, under the water’s-edge method of taxation, a portion of the income of CFC’s that have Subpart F income (that which is not taxed in the foreign countries in which it is earned) is included in the water’s-edge group’s combined income.

Amdahl, as the parent company of its unitary group, made a water’s-edge election effective for each of the tax years at issue, and signed an agreement consenting to taxation under the water’s-edge regime. Accordingly, Amdahl filed a water’s-edge combined income tax return for each of the relevant tax years that included the combined income of its unitary group members incorporated in the U.S.—Amdahl Corporation and AIC—as well as income of its controlled foreign subsidiaries that earned Subpart F income.

Amdahl objected to certain tax assessments by the FTB for the tax years 1988, 1989, 1991 and 1992. The FTB rejected those objections. In administrative proceedings, the SBE determined all issues in the FTB’s favor. Having paid the taxes in question, Amdahl filed the underlying tax refund action. The action was tried to the court largely on stipulated facts, supplemented by the testimony of witnesses and documentary evidence.

The trial court ruled in Amdahl’s favor on all of the issues in this action, except Amdahl’s constitutional challenge to California’s treatment of dividends from foreign subsidiaries. Consequently, Amdahl was awarded tax refunds in the following amounts: $1.26 million for tax year 1988; $1,396 million for tax year 1989; and $254,000 for tax year 1992, for a total judgment of $2,676 million. This appeal and cross-appeal followed.

III.

Discussion

A. Standard of Review

On appeal, we apply the substantial evidence test to the trial court’s factual findings, but review legal determinations independently. (Metropolitan Life Ins. Co. v. State Bd. of Equalization (1982) 32 Cal.3d 649, 658 [ 186 Cal.Rptr. 578 , 652 P.2d 426 ]; Parsons v. Bristol Development Co. (1965) 62 Cal.2d 861, 865 [ 44 Cal.Rptr. 767 , 402 P.2d 839 ]; Southern Pacific Pipe Lines, Inc. v. State Bd. of Equalization (1993) 14 Cal.App.4th 42, 54 [ 17 Cal.Rptr.2d 345 ].) In our review, we are mindful of our Supreme Court’s declaration that ambiguities in the governing statutes are resolved in favor of the taxpayer. (Agnew v. State Bd. of Equalization (1999) 21 Cal.4th 310, 326 [ 87 Cal.Rptr.2d 423 , 981 P.2d 52 ].)

As noted, in the earlier administrative proceedings, the SBE determined all issues in the FTB’s favor. The parties dispute the degree of judicial deference owed to the SBE’s decision in the underlying litigation between the FTB and Amdahl. The Legislature has delegated to the SBE the duty of hearing and determining appeals from actions of the FTB. (§§ 19045-19048.) It has been judicially recognized that the SBE has accumulated a “ ‘body of experience and informed judgment’ in the administration of the business tax law ‘to which the courts and litigants may properly resort for guidance.’ [Citation.]” (Yamaha Corp. of America v. State Bd. of Equalization (1998) 19 Cal.4th 1, 14 [ 78 Cal.Rptr.2d 1 , 960 P.2d 1031 ] (Yamaha Corp.).) Accordingly, the FTB claims we must accord “great weight” to the SBE’s interpretation of the statutes and regulations at issue in this litigation. (Id. at pp. 12-13.)

Amdahl claims that the authoritative strength of the SBE’s decision in this litigation is severely weakened by the fact that in several key issues in this case, the SBE has subsequently reevaluated its position and changed its mind. We agree with Amdahl on this point.

The level of deference due to an agency’s statutory and regulatory interpretation turns on a legally informed, commonsense assessment of its merit in the context presented. (Yamaha Corp., supra, 19 Cal.4th at p. 14 .) An agency’s consistent maintenance of the interpretation under scrutiny, “ ‘especially if [it] is long-standing,’ ” is a circumstance which weighs in favor of judicial deference. (Id. at p. 13, quoting Culligan Water Conditioning v. State Bd. of Equalization (1976) 17 Cal.3d 86, 93 [ 130 Cal.Rptr. 321 , 550 P.2d 593 ].) This rule is supported by practical considerations. “When an administrative interpretation is of long standing and has remained uniform, it is likely that numerous transactions have been entered into in reliance thereon, and it could be invalidated only at the cost of major readjustments and extensive litigation. [Citations.]” (Whitcomb Hotel, Inc. v. Cal. Emp. Com. (1944) 24 Cal.2d 753, 757 [ 151 P.2d 233 ].)

As the FTB emphasizes, the fact that an agency changes its interpretation of a statute is not evidence that either interpretation was legally impermissible. (Henning v. Industrial Welfare Com. (1988) 46 Cal.3d 1262, 1269-1270 [ 252 Cal.Rptr. 278 , 762 P.2d 442 ].) “ ‘In the general case, of course, an administrative agency may change its interpretation of a statute, rejecting an old construction and adopting a new. [Citations.] Put simply, “[a]n administrative agency is not disqualified from changing its mind . . . .” [Citation.]’ ” (Californians for Political Reform Foundation v. Fair Political Practices Com. (1998) 61 Cal.App.4th 472, 488 [ 71 Cal.Rptr.2d 606 ].) However, the fact that the SBE has vacillated in its decision on several key points entitles us to give its administrative decision only limited deference in deciding this case.

B. Characterization of ACT Refund for California Tax Purposes

Amdahl claims the FTB improperly assessed tax liability arising from refunds of a United Kingdom (U.K.) tax called the Advance Corporation Tax (ACT). The FTB characterized the ACT refunds as “nondividend gross income.” Amdahl’s position on this issue, which was accepted by trial court, was that the ACT refunds received by the U.S. parent of a U.K. subsidiary are “dividends” for California tax purposes, and are therefore subject to elimination under section 25106 or deduction under section 24411, subdivision (a). This issue concerns only the Amdahl subsidiaries incorporated in the U.K.—Amdahl International Management Services (AIMS) and Amdahl (U.K) Ltd.

For the tax years relevant here, under U.K. tax law, a U.K. corporation that paid a dividend to its shareholders was required to pay the ACT to the U.K.’s taxing authority, U.K. Inland Revenue (Inland Revenue). Also under U.K. law, the ACT was deemed an advance payment in partial or full satisfaction of the paying corporation’s general U.K. corporate income tax, and the paying corporation used the ACT to reduce its corporate tax liability on its taxable profits. The ACT was also deemed a payment of tax if the recipient of the dividend was a U.K. resident. Thus, the U.K. resident recipient of such a dividend received a tax credit from the U.K. taxing authority for the amount of the ACT payment made by the corporation that related to the dividend received by the resident. (The tax credit was refundable to the dividend recipient if the recipient’s tax owed was less than the credit.)

However, under U.K. domestic law, the tax credit attached to a dividend paid by a U.K. company is not generally available to a shareholder who is not a U.K. resident. Thus, in the absence of an income tax treaty, a nonresident shareholder of a U.K. company receiving a dividend would suffer double taxation—once in the U.K. at the corporate level (the ACT payment), and once in his or her home country at the shareholder level.

Effective in 1980, the U.S. and the U.K. entered into a treaty commonly referred to as the “Income Tax Treaty Between the United Kingdom and the United States” (United States-United Kingdom Income Tax Convention, Dec. 31, 1975, as amended by an Exchange of Notes, signed on Apr. 13, 1976, and Protocols, signed on Aug. 26, 1976, Mar. 31, 1977, and March 15, 1979, eff. April 25, 1980) (the Tax Treaty), which imputes some of the benefits of the U.K. system to U.S. shareholders. Under the Tax Treaty, the U.S. parent corporation will generally be entitled (assuming it owns at least 10 percent of the voting stock of the U.K. company) to a payment from the Inland Revenue of a tax refund (not a tax credit) equal to one-half the tax credit which would be received by a U.K. individual shareholder, less an amount not exceeding 5 percent of the aggregate of the dividend and the tax credit.

Inland Revenue has allowed many U.K. corporations with U.S. shareholders to pay the additional tax refund directly to their U.S. shareholders, thereby avoiding the need for the U.S. shareholders to claim a refund from Inland Revenue. Under this arrangement, the U.K. company also pays a correspondingly lesser amount of the ACT on the dividend to Inland Revenue (the amount Inland Revenue would refund directly to the nonresident shareholder under the Tax Treaty), although the U.K. corporation is given credit for the full amount of the ACT. Overall, through the mechanism of the Tax Treaty, a U.S. shareholder of a U.K. corporation gains some relief by receiving a tax refund (either from Inland Revenue or the corporation directly) for a portion of the ACT payable by the U.K. corporation to Inland Revenue.

As shown in documentary evidence submitted to the trial court, dividends declared and paid by Amdahl’s U.K. subsidiaries, AIMS and Amdahl (U.K.) Ltd., triggered an ACT liability, one-half of which was refunded to AIMS’s sole shareholder, AIC, a California corporation, pursuant to the terms of the U.S.-U.K. Tax Treaty. For example, Amdahl’s U.K. subsidiaries paid $41,104 million in dividends to AIC in 1988. The ACT on the $41,104 million in dividends was $14,380,018. Pursuant to established procedures of Inland Revenue with respect to ACT refunds, AIMS actually paid only one-half of the ACT, or $7,190,009, directly to Inland Revenue. The other one-half of the ACT was paid directly by AIMS to AIC, less the 5 percent dividend withholding tax.

Amdahl treated the ACT refund as a dividend from AIMS to AIC on its federal tax returns for 1988 and 1991, and the IRS accepted such treatment. However, when Amdahl treated this ACT refund as a further dividend from AIMS to AIC in its California water’s-edge combined returns, the FTB took the position that the refunded portion of the ACT received by AIC was nondividend gross income of AIC from the U.K. government. Amdahl asserts that treating the ACT refund as additional income is inconsistent with the language and purpose of the Tax Treaty, which clearly mandates that ACT refunds are to be treated as dividends to the U.S. recipient.

The SBE rejected Amdahl’s appeal of the FTB’s denial of its protest on the ACT refund issue in April 2000 (Appeal of Amdahl Corp. 9A-0054 (Apr. 6, 2000) SBE Nos. 89002459110, 29780 [ 2000 WL 781986 ]). Nevertheless, in September 2000, less than six months later, the SBE reconsidered the identical issue of the treatment of ACT refunds for California franchise tax purposes in conjunction with another taxpayer’s appeal and held that such refunds should be characterized as dividends (Appeal of Thomas & Betts Corporation (Sept. 15, 2000) SBE No. 32822. This decision certainly represents a reversal by the SBE on this issue.

On balance, we find Amdahl’s argument on this issue is more persuasive. Treating the ACT credit refunds as dividends, as Amdahl urges, will effect the purpose of the Tax Treaty. The salient Tax Treaty provision, as stipulated by the parties, reads: “the aggregate of the amount or value of the dividend and the [ACT refund] paid by the United Kingdom to the United States corporation or other resident (without reduction for the 5 or 15 percent deduction, as the case may be, by the United Kingdom) shall be treated as a dividend for United States tax credit purposes.” (Tax Treaty, art. 10(2)(a)(iii).) As the italicized language indicates, the Tax Treaty envisions that the ACT part of the U.K. corporate tax would be refunded directly to the U.S. shareholder and, for U.S. tax purposes, be treated as an additional dividend distribution to be added to the shareholder’s dividend income.

The FTB emphasizes that California is not bound by the Tax Treaty’s pronouncements. (See Container Corp. v. Franchise Tax Board (1983) 463 U.S. 159, 196 [ 77 L.Ed.2d 545 , 103 S.Ct. 2933 ] [“[T]he tax treaties into which the United States has entered do not generally cover the taxing activities of subnational governmental units such as States . . . .” (Fn. omitted.)].) However, the Tax Treaty’s characterization of the ACT refund as additional dividend income to the taxpayer appears to be the result most supported by both the mechanics of the ACT refund system and California’s definition of dividend income, which tracks the federal definition. (See former § 24495 and current § 24451, both of which directly incorporate the federal definition of dividends from Int. Rev. Code § 316.)

We reject the FTB’s argument that the ACT refund cannot be a California dividend because the refund is a payment from the U.K. government and Amdahl is not a shareholder in the U.K. government. This argument does not withstand close scrutiny because it ignores the U.K. government’s role as that of a pass-through or agent that is legally obligated to forward the payment to a third party. In fact, in this case, consistent with its role as an intermediary or agent, the U.K. government did not even go through the formality of collecting the full amount of the ACT from the U.K. company, but permitted the ACT refund payment to be made directly to the U.S. parent by the U.K. subsidiary. Accordingly, we agree with the trial court that the ACT refunds must be treated as dividends for California tax purposes.

C. Computation of the Inclusion Ratio

To determine the includable portion of a CFC’s Subpart F foreign source income in the water’s-edge report, section 25110, subdivision (a)(6) sets out a computation formula, or what the trial court called “the inclusion ratio.” In the case of Amdahl’s first-tier CFC’s (Amdahl Ireland, ANBV, and AIMS), the parties are in dispute as to whether or how dividends received by each of these first-tier subsidiaries from the corresponding second-tier subsidiary (AOCC, Amdahl Lease, and Amdahl (U.K.) Ltd.) are taken into account in the determination of the inclusion ratio of the first-tier subsidiary. The trial court found that in its water’s-edge combined report, Amdahl could completely exclude from this ratio the dividends paid out of income already included in the combined income of the group. The FTB challenges this conclusion on appeal.

As previously noted, the income of CFC’s in a water’s-edge group that has Subpart F income is potentially subject to California tax. The portion of the CFCs’ income to be included in the group’s combined income is determined by the inclusion ratio set forth in section 25110, subdivision (a)(6). The inclusion ratio is defined as the following fraction:

CFC’s Subpart F Income CFC’s Earnings and Profits

Section 25110, subdivision (a)(6) defines Subpart F income as “income . . . defined in Section 952 of Subpart F of the Internal Revenue Code” and earnings and profits “as defined in Section 964 of the Internal Revenue Code.” The resulting fraction may not be less than zero nor more than one. The CFC subsidiary’s inclusion ratio is then multiplied by its net income to obtain the amount of the CFC’s income to be included in the water’s-edge group’s combined income. As the trial court noted, “this statutory formulation results in the inclusion of Subpart F income, increased (or decreased, as the case may be) by a pro-rata share of California adjustments.”

The court went on to note that “[t]he issue upon which the parties are in dispute is how to compute the inclusion ratio in the case of a [CFC] that receives dividends from an affiliate that is also a [CFC] (a ‘lower-tier subsidiary’).” The court found that when a CFC receives a dividend that was paid by a lower-tier CFC out of earnings that were wholly included in the combined income of the water’s-edge group, such dividend is excluded and is not taken into account in applying the inclusion ratio of section 25110, subdivision (a)(6) to the recipient CFC. While we agree with the trial court’s result, we do not altogether adopt its reasoning.

As noted, with certain exceptions not relevant here, California incorporates the federal definition of Subpart F income through section 25110, subdivision (a)(6) and California Code of Regulations, title 18, section 25110, subdivision (d)(2)(F)(l). Additionally, in determining whether a corporation has Subpart F income for a given year, certain federal exclusions and special rules apply (Cal. Code Regs., tit. 18, § 25110, subd. (d)(2)(F)(3).) In the case of dividends that are received by foreign subsidiaries from lower tier foreign subsidiaries, Internal Revenue Code section 959(b) excludes from gross income such dividends to the extent that they “are, or have been” included in the gross income of a U.S. shareholder under Subpart F. Under U.S. Treasury Regulations, Subpart F income excludes “distributions of previously taxed income under [Internal Revenue Code section] 959(b).” ( 26 CFR § 1.954-2 (b)(l)(i) (2002); later renumbered § 4.954-2(b)(l)(i) (2004).)

Significantly, both Amdahl and the FTB agree that under the foregoing statutory and regulatory scheme, the dividends at issue here are excluded from Subpart F income under the federal definition. Nevertheless, the FTB argues that “[t]he fact that the dividends would be excluded for federal purposes as a result of the operation of [Int. Rev. Code] § 959(b) does not remove them from the ‘Subpart F income’ used to compute the inclusion ratio of the payee under the California Revenue and Taxation Code.” The trial court agreed with this portion of the FTB’s argument, finding that “in using the federal definition of Subpart F income in [Int. Rev. Code] § 952, California did not adopt [Int. Rev. Code] § 959 or its principles, and the exclusion in that Treasury regulation therefore has no application for California tax purposes.”

We disagree with this conclusion. It is clear that California has chosen to measure Subpart F income by incorporating the federal definition— a standard that implies California’s willingness to follow the federal lead. In defining Subpart F income for purpose of calculating the inclusion ratio defined in section 25110, subdivision (a)(6), absent clear language in the statute or in administrative regulations refusing to do so, we may assume California has adopted into its definition of Subpart F income the federal exclusions, including “distributions of previously taxed income under [IRC §] 959(b).” ( 26 CFR § 1.954-2 (b)(l)(i); later renumbered § 4.954-2(b)(l)(i).)

Nevertheless, there is a separate and distinct reason why the second-tier dividends at issue here must not be included in the inclusion ratio—section 25106 forbids it. Section 25106 provided the following for the years at issue in this case: “In any case in which the tax of a corporation is or has been determined under this chapter with reference to the income and apportionment factors of another corporation with which it is doing or has done a unitary business, all dividends paid by one to another of such corporations shall, to the extent such dividends are paid out of such income of such unitary business, be eliminated from the income of the recipient and shall not be taken into account under Section 24344 or in any other manner in determining the tax of any such corporation.” (Italics added.)

The Legislature could hardly have chosen words with a clearer meaning. Simply put, section 25106 ensures that amounts included in the combined income of a unitary group can be moved (in the form of dividends) among members of the unitary group without tax consequence. The reason for this is also clear. In a combined unitary group, the subsidiaries’ apportioned earnings are taxed as income of the unitary business. Because the state has already taxed the earnings out of which dividends are paid, the dividends themselves are not subject to taxation. This prevents dividends from subsidiaries from being taxed twice—once as earnings of the issuing subsidiary, and once as separate income to the unitary business from receipt of the dividend.

The FTB acknowledges “there is no regulation which squarely addresses the question of whether [Revenue and Taxation Code] section 25106 applies in computing the ‘inclusion ratio’ required by [Revenue and Taxation Code] section 25110(a)(6).” Nevertheless, by delving into the legislative history and purpose of section 25106, the FTB argues that section 25106 was never intended to be applied to the computation of the inclusion ratio, which is merely a measure of how much income of the CFC is included (subject to apportionment) in the water’s-edge return. In making its point, the FTB emphasizes that “section 25106 was enacted in 1967, when Subpart F income and water’sedge tax reporting were unknown to the Revenue and Taxation Code.”

The FTB has not provided us with a compelling reason to disregard the clear statutory language of section 25106. It is elementary that the objective of statutory interpretation is to ascertain and effectuate legislative intent. The first step in determining that intent is to scrutinize the actual words of the statute, giving them a plain and common sense meaning. (Hughes v. Board of Architectural Examiners (1998) 17 Cal.4th 763, 775 [ 72 Cal.Rptr.2d 624 , 952 P.2d 641 ].) If there is no ambiguity in the statutory language, a court must presume that the Legislature meant what it said, and the plain meaning of the statute governs. (Lennane v. Franchise Tax Bd. (1994) 9 Cal.4th 263, 268 [ 36 Cal.Rptr.2d 563 , 885 P.2d 976 ].) Because the language of section 25106 is clear and unambiguous, it would be improper for us to refer to extrinsic evidence in an attempt to create an ambiguity from which we could construe the statute to mean something other than what it says. (See Hartford Fire Ins. Co. v. Macri (1992) 4 Cal.4th 318, 326 [ 14 Cal.Rptr.2d 813 , 842 P.2d 112 ]; Solberg v. Superior Court (1977) 19 Cal.3d 182, 198 [ 137 Cal.Rptr. 460 , 561 P.2d 1148 ]; Farnow v. Superior Court (1990) 226 Cal.App.3d 481, 486 [ 276 Cal.Rptr. 275 ] [“a court may not rewrite a law, supply an omission or give words an effect different from the plain and direct import of the terms used”].)

We must also presume that the Legislature was well aware of the rules governing intercompany dividends, including section 25106, when it enacted section 25110, subdivision (a)(6). (See Fermino v. Fedco, Inc. (1994) 7 Cal.4th 701, 720 [ 30 Cal.Rptr.2d 18 , 872 P.2d 559 ]; Building Industry Assn. v. City of Livermore (1996) 45 Cal.App.4th 719, 730 [ 52 Cal.Rptr.2d 902 ]; Bailey v. Superior Court (1977) 19 Cal.3d 970, 977-978 [ 140 Cal.Rptr. 669 , 568 P.2d 394 ] [Legislature is presumed to have enacted legislation with existing law in mind].) Consequently, we assume that at the time it enacted section 25110, subdivision (a)(6), the Legislature was aware that section 25106 made intercompany dividends paid from unitary income nontaxable and provided such dividends “shall not be taken into account... in any . . . manner in determining the tax of any member of the group.” (Italics added.)

For the foregoing reasons, we endorse the trial court’s conclusion that the legislative scheme contemplates that “dividends paid out of unitary income of lower-tier subsidiaries should be excluded from all the factors used in the computation of the amount included under [Revenue and Taxation Code section] 25110(a)(6); that is, such dividends should be excluded from the numerator (Subpart F income), the denominator (earnings and profits) and the amount to which the inclusion ratio is applied (the income of the controlled foreign corporation).” Like the trial court, we are persuaded that is the only conclusion possible from the plain and unambiguous language of section 25106.

D. Ordering of Distributions

A question remains, however, as to how dividends received by the unitary group from a CFC should be treated where part of the CFC’s income is Subpart F income and thus included in the unitary group’s tax return, and some is not. Amdahl argued, and the trial court adopted as correct, the view that such dividends should be deemed paid first out of included income. The FTB, on the other hand, claims that the dividend should be prorated between earnings that have been included in the combined income of the water’s-edge group and excluded income.

The importance of this distinction stems from the fact that, generally speaking, section 24411, subdivision (a) provides that 75 percent of dividends received by the water’s-edge group, and not eliminated by section 25106, can be deductible for purposes of computing the taxable income for the combined report. The ordering determines whether the dividend elimination or dividend deduction provision applies, i.e., section 25106 (100 percent deduction for earnings previously included in a California combined return) or section 24411(a) (75 percent dividends received deduction).

The superior court’s decision directs that: “[Revenue and Taxation Code section] 25106 should be applied to dividends from [CFC’s] that are partially included in the Water’s Edge group under [Revenue and Taxation Code section] 25110(a)(6) in a manner that deems dividends to be distributed first from income that has already been included in the unitary group, to the extent thereof, and then from the non-unitary income.” Consequently, under the superior court’s ruling, such dividends would be deemed to have been paid first out of already taxed, unitary group income (subject to elimination under section 25106), and only after the section 25106 income had been exhausted would they be taxed at the 25 percent rate remaining after application of section 24411, subdivision (a)’s 75 percent “dividends received” deduction.

The superior court reached its result in order to “harmonize[] [the statutes] and avoid[] constitutional infirmities.” The court came to the conclusion that the FTB’s pro rata ordering of such dividends might raise a constitutional concern about section 24411, subdivision (a) because “the burden on foreign commerce that Amdahl alleges is lesser or greater depending on whether dividends are treated as coming first or last from income of the unitary group.”

No statute, regulation or other administrative pronouncement provides clear guidance on this question. For 1989 and later years, regulations under section 24411, subdivision (a) provided that dividends paid by partially included corporations would be treated as prorated between amounts eligible for section 25106 elimination and amounts eligible for partial deduction under section 24411. However, commencing in 2001, the FTB’s new unitary combined reporting intercompany transaction regulations seem to indicate that when a dividend is paid out of a mix of previously included and not previously included income, any earnings previously included in the unitary group are deemed to be distributed first, dollar for dollar. (Cal. Code Regs., tit. 18, § 25106.5-l(f)(2).)

In the absence of any clear and controlling guidance on this question, our “construction is to favor the taxpayer rather than the government.” (Edison California Stores v. McColgan (1947) 30 Cal.2d 472, 476 [ 183 P.2d 16 ].) Furthermore, we “ ‘ “must select the construction that comports most closely with the apparent intent of the Legislature, with a view to promoting rather than defeating the general purpose of the statute, and avoid an interpretation that would lead to absurd consequences.” [Citation.]’ [Citation.]” (Torres v. Parkhouse Tire Service, Inc. (2001) 26 Cal.4th 995, 1003 [ 111 Cal.Rptr.2d 564 , 30 P.3d 57 ].) “And, wherever possible, ‘we will interpret a statute as consistent with applicable constitutional provisions, seeking to harmonize Constitution and statute.’ [Citation.]” (People v. Superior Court (Zamudio) (2000) 23 Cal.4th 183, 193 [ 96 Cal.Rptr.2d 463 , 999 P.2d 686 ].)

For the reasons indicated above, including those relied on by the trial court, we conclude that dividends paid by first-tier subsidiaries from current year earnings should be treated as paid (1) first out of earnings eligible for elimination under section 25106, with 2) any excess paid out of earnings eligible for partial deduction under section 24411. In the case of a CFC that is partially included in a unitary group, the CFC will be able to move amounts that have been included in the combined income of the unitary group without tax incident only by adopting the ordering rule described above.

E. Discriminatory Treatment of Foreign Dividends

Turning to Amdahl’s cross-appeal, appeal No. A101203, Amdahl argues that section 24411, subdivision (a)’s deduction limitation for foreign source dividends unconstitutionally discriminates against foreign commerce in violation of the United States Constitution’s foreign commerce clause. As seen, section 24411, subdivision (a) provides that dividends received by the water’s-edge group from a foreign subsidiary, to the extent not eliminated by some other provision such as section 25106, are only 75 percent deductible. In its cross-appeal, Amdahl claims that similar dividends received from a U.S. subsidiary are, through various provisions, 100 percent eliminated or deductible. Thus, Amdahl alleges that section 24411, subdivision (a), to the extent that it taxes foreign subsidiary dividends more heavily than domestic subsidiary dividends, discriminates against foreign commerce in violation of the commerce clause of the United States Constitution.

The trial court rejected Amdahl’s constitutional challenge. Among other things, the court pointed out that section 25106, acting in conjunction with section 24411, posits its different treatment of dividends not on whether the dividends are paid from a foreign or domestic subsidiary, but on whether or not the income from which the dividends are paid has been included in the water’s-edge combined report. Thus, if a subsidiary’s dividend has been fully included in the combined report, it is eliminated pursuant to section 25106, whether the subsidiary is foreign or domestic. If the subsidiary’s dividends are paid out of earnings and profits that have not been included on the combined report, it is nevertheless eligible for the 75 percent dividends received deduction found in section 24411, subdivision (a).

Consequently, the trial court found that California’s water’s-edge system actually favors foreign commerce, rather than discriminating against it, because it subjects less of foreign subsidiaries’ income to tax when compared to domestic subsidiaries. The court reasoned: “Under California law, 100 [percent] of the income of domestic unitary subsidiaries is included in the combined report, and is subject to interstate and intercorporate apportionment. Thus, while the domestic dividends are eliminated [by section 25106], the income from which they are paid is included 100 [percent] on the combined report, which renders that income subject to apportionment and taxation. Similarly, foreign source dividends paid from income included on the combined report are eliminated in exactly the same manner as domestic dividends. It is only when the income of a foreign subsidiary has been excluded from the combined report by Amdahl’s water’s-edge election under [Revenue and Taxation Code section] 25110 that dividends paid by a foreign subsidiary are not eliminated by [Revenue and Taxation Code section] 25106.”

The trial court went on to explain, “For those dividends not eliminated by § 25106, California provides a 75% ‘dividends received’ deduction under [Revenue and Taxation Code section] 24411.” Consequently, the trial court found that California’s water’s-edge system actually subjected less of a foreign subsidiary’s income to taxation when compared to that of domestic subsidiaries, which is 100 percent included on the combined report. Relying on this reasoning, the FTB argues that section 24411 does not discriminate against foreign commerce. We agree.

The United States Constitution’s foreign commerce clause provides that “Congress shall have Power ... to regulate Commerce with foreign Nations.” (U.S. Const., art. I, § 8, cl. 3.) The term “commerce” includes the flow of dividends from a foreign subsidiary to its parent company. (Kraft Gen. Foods, Inc. v. Iowa Dept. of Revenue and Finance (1992) 505 U.S. 71, 76 [ 120 L.Ed.2d 59 , 112 S.Ct. 2365 ] (Kraft).)

The foreign commerce clause not only grants Congress the authority to regulate commerce between the United States and foreign nations, it also directly limits the power of the states to discriminate against foreign commerce. (Wardair Canada v. Florida Dept. of Revenue (1986) 477 U.S. 1, 7-8 [ 91 L.Ed.2d 1 , 106 S.Ct. 2369 ].) This is commonly referred to as the “dormant” or “negative” aspect of the foreign commerce clause. The dormant aspect of the foreign commerce clause serves two related purposes. First, it prevents states from promulgating protectionist policies. Second, it restrains the states from excessive interference in foreign affairs, which are the domain of the federal government. (Japan Line, Ltd. v. County of Los Angeles (1979) 441 U.S. 434, 448-451 [ 60 L.Ed.2d 336 , 99 S.Ct. 1813 ]; National Foreign Trade Council v. Natsios (1999) 181 F.3d 38, 66 .) Because matters of concern to the entire nation are implicated, “the constitutional prohibition against state taxation of foreign commerce is broader than the protection afforded to interstate commerce . . . .” (Kraft, supra, 505 U.S. at p. 79 .)

The United States Supreme Court applied these principles in Kraft, a case relied upon by Amdahl. In Kraft, Iowa allowed a deduction from base taxable income for dividends paid to a parent company by a domestic subsidiary not doing business in Iowa, while it did not allow a deduction from base income for dividends paid to a parent company by a foreign subsidiary not doing business in Iowa. The Supreme Court held that the fact that dividends received from a unitary business’ foreign subsidiaries were always treated less favorably than dividends received from its domestic subsidiaries constituted an unconstitutional discrimination under the foreign commerce clause.

Kraft involved a separate entity tax return of a parent company. The court pointed out that Iowa was not applying unitary combination. The court wrote an important footnote, as follows: “If one were to compare the aggregate tax imposed by Iowa on a unitary business which included a subsidiary doing business throughout the United States (including Iowa) with the aggregate tax imposed by Iowa on a unitary business which included a foreign subsidiary doing business abroad, it would be difficult to say that Iowa discriminates against the business with the foreign subsidiary. Iowa would tax an apportioned share of the domestic subsidiary’s entire earnings, but would tax only the amount of the foreign subsidiary’s earnings paid as a dividend to the parent. . . .” (Kraft, supra, 505 U.S. at p. 80, fn. 23 .)

Relying on this rationale, the FTB argues that the single-entity reporting system involved in Kraft raises constitutional concerns that are not present under California’s combined water’s-edge method of apportioning the combined income of a unitary business group for tax purposes. The FTB claims Kraft should have no application within a combined unitary group, because each member’s income and apportionment factors are included in the return equally, regardless of place of incorporation or country of operation. The FTB’s argument finds substantial support in the case law.

Since Kraft was issued, several other courts have been asked to determine whether a given state’s tax system discriminates against foreign commerce in a manner prohibited by Kraft. In In re Morton Thiokol, Inc. (1993) 254 Kan. 23 [ 864 P.2d 1175 ] (Thiokol), the court limited the holding in Kraft to states that do not use a combined water’s-edge or domestic combination reporting method. The Thiokol court reasoned that a combined reporting state (i.e., water’s-edge) does not discriminate against foreign subsidiaries. While the foreign subsidiary’s dividend payments to the unitary business are taxed, its total income is not included in the unitary business overall income. Conversely, while a domestic subsidiary’s dividend payments to the unitary business is not taxed, its total income is included in the unitary business overall income. Thus, no discrimination against foreign commerce occurs.

Following the lead of Thiokol, the Supreme Court of Maine in E.I. Du Pont de Nemours & Co. v. State Tax Assessor (Me. 1996) 675 A.2d 82 (Du Pont), held that combined water’s-edge reporting saved the Maine income tax statute from the fate of the Iowa statute in Kraft. The Du Pont court reasoned: “Far from discriminating against foreign commerce, Maine’s water’s edge combined reporting method provides a type of ‘taxing symmetry’ that is not present under the single entity system. . . . Because the income of the unitary domestic affiliates is included, apportioned, and ultimately directly taxed by Maine as part of the parent company’s income, the inclusion of dividends paid by foreign subsidiaries does not constitute the kind of facial discrimination against foreign commerce that caused the Supreme Court to invalidate Iowa’s tax scheme in Kraft.'” (Id. at pp. 87-88.)

Other courts have found the rationale of the Thiokol and Du Pont courts to be persuasive and have determined that the taxation of foreign-source income is not invalid under Kraft where the consolidated or combined methodology is used. (See Caterpillar, Inc. v. Commissioner of Revenue (Minn. 1997) 568 N.W.2d 695 [interest and royalty payments by foreign subsidiary]; Caterpillar, Inc. v. Dept. of Rev. Admin. (1999) 144 N.H. 253 [ 741 A.2d 56 ] [interest and royalty payments by foreign subsidiary]; see also Emerson Elec. Co. v. Tracy (2000) 90 OhioSt.3d 157 [ 735 N.E.2d 445, 448-449 ]; Caterpillar Financial Services Corp. v. Whitley (1997) 288 Ill.App.3d 389 [ 680 N.E.2d 1082, 1086-1089 , 223 Ill.Dec. 879 ]; Dart Industries, Inc. v. Clark (R.I. 1995) 657 A.2d 1062, 1065 .)

We find the rationale of these courts to be persuasive, and hold that California’s water’s-edge method of apportionment of income does not facially discriminate against foreign commerce. As explained by the trial court, “Like Du Pont, in the case of California’s water’s-edge reporting method, foreign subsidiaries’ dividends are partially included, while the entirety of domestic subsidiaries’ income is included, in the water’s-edge combined report. ... In addition, California gives those included foreign subsidiary dividends a 75% ‘dividends received’ deduction from the water’sedge group’s combined income pursuant to [Revenue and Taxation Code section] 24411. Thus, the same kind of ‘taxing symmetry’ present in Du Pont is present here. Therefore, the holding in Kraft does not apply to these facts.” We affirm the trial court ruling that California’s water’s-edge reporting method does not unconstitutionally discriminate against foreign commerce on this basis.

F. Timeliness of Action with Respect to 1988 Tax Year

The FTB claims the trial court erred in finding Amdahl’s action seeking a refund for the 1988 tax year was timely filed. The statute of limitations applicable to tax refund suits is section 19384, which provides: “The action . . . shall be filed within four years from the last date prescribed for filing the return or within one year from the date the tax was paid, or within 90 days after (a) notice of action by the Franchise Tax Board upon any claim for refimd, or (b) final notice of action by the State Board of Equalization on an appeal from the action of the Franchise Tax Board on a claim for refund, whichever period expires later.” Relying on section 19384, the FTB argues “[S]ince this action was not filed within 90 days of the SBE’s April 6, 2000 denial of Amdahl’s appeal, as it relates to 1988 taxes, this action is time barred.”

In rebuttal, Amdahl argues that its claim for refund did not ripen until July 31, 2000, when the FTB formally notified Amdahl how it would apply various payments and credits for various tax years to the 1988 tax liability. Having paid the tax, Amdahl filed its administrative claims for refund in a timely manner in two parts on August 30, 2000.

In the administrative proceedings below, the FTB rejected Amdahl’s claim for refund of the 1988 tax year based on the assertion that Amdahl’s earlier protest had been “ ‘converted’ to a claim for refund three years earlier.” The FTB took the position that because the earlier protest had actually become a claim for refund, Amdahl’s time for filing suit for refund had already expired (90 days after the April 6, 2000 SBE decision). In challenging this ruling below, Amdahl claimed that prior to the FTB’s rejection of its administrative protest, there was no evidence that Amdahl’s tax protest was being treated as a claim for refund nor was Amdahl ever notified that the FTB was treating its tax protest as a claim for refund, or that it considered the 1988 protested taxes to have been paid.

At the center of this controversy is the legal effect of Amdahl’s wire-transferred payment of $2 million to the FTB in January 1998, during the pendency of its administrative protest against the FTB’s proposed assessment for the 1988 tax year. By this payment, the FTB asserts that Amdahl’s earlier protest was converted “by operation of law” to a claim for refund by section 19335. Section 19335 provides that a taxpayer may pay a tax under protest, before the FTB acts on a claim or the SBE acts on an appeal, in which case the protest is treated as a claim for a refund or an appeal from the denial of a claim for refund. Relying on section 19335, the FTB claims that, having paid the taxes in question during pendency of the administrative proceedings, Amdahl missed the 90-day statutory deadline to file a tax refund action.

Amdahl does not dispute the fact that it made a $2 million payment to the FTB in January 1998, although it adamantly denies the payment was made to satisfy its 1988 tax obligation. Instead, Amdahl claims that by making this payment, it was paying unprotested taxes and potential but unassessed 1988 taxes. Furthermore, as of January 1998, with ongoing federal and state audits and likely proposed adjustments and assessments coming, Amdahl wanted to stop interest from accruing on unpaid amounts.

The trial court concluded that Amdahl’s action with respect to its 1988 taxes was timely. After considering documentary and testimonial evidence, the trial court made the following findings of fact: 1) the disputed payment by Amdahl in January 1998 was not applied to the liability for the protested tax until after the SBE’s denial of Amdahl’s appeal in April 2000; 2) at no time during the pendency of Amdahl’s protest did the FTB treat the protest as a claim for refund; and 3) at no time during the pendency of Amdahl’s appeal from the denial of the protest did the SBE or the FTB treat the appeal as the appeal from the denial of a refund. The trial court reasoned: “Had the FTB acted in the manner which the Statute mandates, Amdahl would have had ample notice of the FTB’s position and could have commenced this action in what the FTB would necessarily concede to be a timely fashion. The FTB cannot now take advantage of its own failure to follow the statute, that is, the FTB cannot now invoke [Revenue and Taxation Code section] 19335 to divest Amdahl of its right to due process on its refund. The Court holds, therefore, that Amdahl’s refund claim with respect to the 1988 year is not time-barred.”

We conclude Amdahl introduced evidence sufficient to sustain the findings in its favor on the theory enunciated by the trial court. The record in this case shows that Amdahl sent a letter dated January 23, 1998, accompanying the $2 million payment, stating: “The payment represents a combination of additional franchise tax owed on certain audit adjustments made during the franchise tax field audit and additional franchise tax owed due to potential audit adjustments for 1988. . . . Certain other proposed FTB audit adjustments for 1988 are currently being protested by Amdahl Corp. To prevent additional interest from accruing Amdahl decided to make a payment at this time.” (Italics added.) The FTB never treated the protest as a claim for refund and never informed Amdahl that its protest had been converted into a claim for refund until February 2001, over a year after the 90-day statute of limitations had allegedly run. On the contrary, the FTB consistently treated the protest, logically enough, as a protest.

The FTB argues that to endorse the trial court’s factual findings in this case would be opening the door to both the piecemeal litigation of tax claims, and tolerating the evasion of the Legislature’s strict rules for the filing of tax refund actions. However, we note that in 1999 the Legislature enacted section 19041.5, which expressly authorizes the procedure employed by Amdahl herein. Section 19041.5 allows the taxpayer to make a payment to stop the accrual of interest or to cover nonprotested tax without having those funds considered a payment of taxes that would trigger the section 19384 statute of limitations regarding claims for refund. As the Legislature recognized, section 19041.5 simply “[c]odifies existing . . . [FTB] practice . . . .”

G. Award of Attorney Fees with Respect to Time-bar Issue

In appeal No. A102558, the FTB appeals from the postjudgment order awarding Amdahl $20,000 in attorney fees under section 19717. Invoking that section, the trial court found that the FTB’s position on the time-bar issue was not “substantially justified,” thus entitling Amdahl to an award of attorney fees for its defense of that issue.

Section 19717 provides that a party who brings a civil proceeding against the state to recover franchise taxes may recover reasonable litigation costs, including attorney fees, if: (1) the suit is brought in a California court; (2) the party has exhausted its administrative remedies under the applicable tax laws; (3) the party establishes that the position of the state was not substantially justified; and (4) the party substantially prevails. (See Lennane v. Franchise Tax Bd. (1996) 51 Cal.App.4th 1180, 1183-1184 [ 59 Cal.Rptr.2d 602 ] [statute quoted at fn. 1]; McDonnell Douglas Corp. v. Franchise Tax Bd. (1994) 26 Cal.App.4th 1789, 1797 [ 33 Cal.Rptr.2d 129 ].) The FTB urges this court to reverse the award because its position on the time-bar issue was “substantially justified” and Amdahl did not “substantially prevail.”

“California cases have defined a ‘substantially justified’ position to mean one which is justified to a degree that would satisfy a reasonable person, or ‘ “has a ‘ “reasonable basis both in law and fact.” (Wertin v. Franchise Tax Bd. (1998) 68 Cal.App.4th 961, 977 [ 80 Cal.Rptr.2d 644 ], citing Lennane v. Franchise Tax Bd., supra, 51 Cal.App.4th at pp. 1188-1189.) “[T]he use of the word ‘reasonable’ in explaining ‘substantially justified’ implies an objective standard that does not depend on an analysis of the subjective motivations of the government in taking the position it did.” (Wertin, supra, 68 Cal.App.4th at p. 978 .) In this regard, we stress that the FTB’s position need not be the one accepted by the trier of fact. So long as the position is one that a reasonable person could think is correct, it may be substantially justified even in the face of conflicting evidence. Finally, the burden of showing substantial justification is on the FTB, not the taxpayer. (§ 19717, subd. (c)(2)(B)(ii).)

Applying these principles to the case at bar leads to the conclusion that the FTB’s position, when viewed from the totality of the circumstances, was not substantially justified. While section 19335—which transforms a protest by law into a claim for refund upon the payment of the tax—provided the FTB with a reasonable legal basis for the theory it propounded, there was virtually no factual support for the legal theory advanced. On appeal, the FTB does not challenge the superior court’s factual findings or even attempt to meet the requirements of the substantial evidence standard of review. Nor does it explain how Amdahl could have paid the protested tax during the pendency of the protest, given the documentary evidence attesting to the fact that the FTB did not apply Amdahl’s $2 million payment to the 1988 tax liability until after the SBE’s rejection of Amdahl’s appeal in April 2000. Nevertheless, adopting its untenable position that Amdahl’s claims with respect to the 1988 tax year were time-barred, the FTB forced Amdahl into lengthy administrative proceedings to develop the record on this point before final vindication of its right to bring an action with respect to the 1988 tax year. The FTB has simply not shown how it was substantially justified in advancing this argument.

Additionally, the FTB claims Amdahl cannot be considered a prevailing party because the single issue for which attorney fees were awarded—the time-bar issue for the 1988 tax year—was not the most significant issue in the case. In making its point, the FTB emphasizes that the amount of the refund attributable for the 1988 tax year was only 30 percent of the total refund sought by Amdahl.

Regardless of the relative importance of the time-bar issue, Amdahl unquestionably was the prevailing party below, having prevailed on almost every significant issue at trial and having been awarded $2,676 million, which was all but $714,000 out of the refunds sought.

Where a lawsuit consists of related claims and the taxpayer has won substantial relief, we believe a trial court has discretion to award the taxpayer attorney fees for discrete issues under section 19717, even if the issues for which fees are awarded do not represent the bulk of the amount in controversy or the most significant issues in the case. To hold otherwise, as pointed out by Amdahl, would allow “the government free rein to adopt positions and argue issues that are not substantially justified so long as the issues are less significant than other issues in the case.” To the extent that an award of attorney fees will act as a disincentive to the FTB to take positions that it cannot substantially justify, we believe such an award is well within the court’s discretion.

IV.

Disposition

The judgment is affirmed.

Kline, P. J., and Haerle, J., concurred.

A petition for a rehearing was denied July 27, 2004, and the petition of appellant Franchise Tax Board for review by the Supreme Court was denied October 20, 2004. Werdegar, J., did not participate therein.

Amdahl Corporation has changed its name to Fujitsu IT Holdings, Inc. For convenience, however, we will continue to refer to Amdahl, as it was known throughout the tax years at issue. All further undesignated statutory references are to the Revenue and Taxation Code. The parties, for purposes of this case, have stipulated to the dollar amounts involved in each of these disputed issues and have provided the court with detailed calculations to illustrate the differences between each party’s position in this case. We will not add to the length of this opinion by reprinting these calculations, but assure the parties that they have been considered in resolving the issues presented. A CFC, generally, is organized in a foreign country and is more than 50 percent owned by U.S. shareholders. AIC is a California corporation that is a wholly owned subsidiary of Amdahl operating as a U.S. holding company for foreign subsidiaries of Amdahl. AIMS, a U.K. corporation, is a wholly owned subsidiary of AIC. AIMS owns all of the shares of Amdahl (U.K.) Ltd., a U.K. corporation. Current section 25110, subdivision (a)(6) was originally numbered section 25110, subdivision (a)(8), and then changed to section 25110, subdivision (a)(7), before receiving its current numbering. In a subsequent ruling involving another taxpayer, the SEE reversed the position it took in Amdahl’s administrative appeal, and interpreted section 25110, subdivision (a)(6) in the same manner as the trial court. In other words, the SEE held that dividends paid out of included income of a lower-tier CPC should not taken into account in the determination of the inclusion ratio. (See Appeal of Baxter Healthcare Corp. (Aug. 1, 2002, SEE No. 150881, rehg. denied Dec. 19, 2002.) Section 25110, subdivision (a)(6) provides, in pertinent part, “Any affiliated corporation which is a ‘controlled foreign corporation,’ as defined in Section 957 of the Internal Revenue Code, if all or part of the income of that affiliate is defined in Section 952 of Subpart F of the Internal Revenue Code (‘Subpart F income’). The income and apportionment factors of any affiliate to be included under this paragraph shall be determined by multiplying the income and apportionment factors of that affiliate without application of this paragraph by a fraction (not to exceed one), the numerator of which is the ‘Subpart F income’ of that corporation for that taxable year and the denominator of which is the ‘earnings and profits’ of that corporation for that taxable year, as defined in Section 964 of the Internal Revenue Code.” A taxpayer may also be eligible for a deduction under section 24402. Section 24402 provides that even in the absence of a unitary business, where the payor corporation was subject to California tax, the recipient corporation may deduct from its gross income dividends that were declared from income already included in the measure of California franchise tax imposed upon the payor corporation. The purpose of this deduction is to avoid double taxation. (Safeway Stores, Inc. v. Franchise Tax Board (1970) 3 Cal.3d 745, 749-750 [ 91 Cal.Rptr. 616 , 478 P.2d 48 ].) In order for the recipient corporation to claim the California deduction, however, the payor corporation must have had income from sources in California so that the payor corporation was subject to California tax. As the term “separate entity” implies, states using that method of reporting income treat the various subsidiaries of a multijurisdictional enterprise as separate from one another and the income of those entities not doing business in the state are not considered in the income of the single entity. The court’s disposition of Amdahl’s constitutional challenge to section 24411, subdivision (a) makes it unnecessary to address the FTB’s additional argument that Amdahl’s voluntary water’s-edge election prevents it from raising a constitutional challenge to the application of section 24411, subdivision (a). Section 19041.5, subdivision (a) provides, in pertinent part, that “any amount paid as a tax or in respect of a tax that is paid after the mailing of a notice of proposed deficiency assessment and designated by the taxpayer as a deposit in the nature of a cash bond made to stop the running of interest, shall not be considered a payment of tax for purposes of filing a claim for refund pursuant to Section 19306 or an action pursuant to Section 19384 until either of the following occurs: [f] (1) The taxpayer provides a written statement to the Franchise Tax Board specifying that the deposit shall be a payment of tax for purposes of Section 19306, 19335, or 19384. []Q (2) The deficiency assessed becomes due and payable in accordance with Section 19049.” While Amdahl prevailed on all of the primary issues in this case, Amdahl sought and was granted attorney fees only in connection with the statute of limitations issue for which it believed the FTB’s position had no substantial justification.