Opinion

We granted review to decide whether the use by the state Franchise Tax Board of a three-factor formula to apportion the income of a foreign-parent multicorporate unitary enterprise for state tax purposes violates the foreign commerce clause of the federal Constitution (art. I, § 8, cl. 3). We conclude that relevant treaty and other materials manifest a federal intent not to prohibit the states from employing formula apportionment in taxing the income of such a multinational unitary business. We therefore reverse the judgment of the Court of Appeal.

I

Plaintiff taxpayers, Barclays Bank of California and Barclays Bank International, Ltd. (collectively, the Bank), brought this refund action to recover assessments of $152,420 and $1,678 levied against them by the Franchise Tax Board (Board) for the 1977 tax year. The basis for the assessments was a finding by the Board that, together with their United Kingdom-based corporate parent and related worldwide subsidiaries, the Bank comprised a unitary enterprise, thereby subjecting it to the three-factor mathematical formula used by the Board to apportion the interjurisdictional income of a unitary business for state corporate income tax purposes. Although the Bank did not contest the Board’s predicate finding of corporate unity, it did claim that application of California’s apportionment formula to such a unitary group—that is, one whose corporate parent is a foreign domiciliary-—violates the foreign commerce clause of the federal Constitution.

Following a bench trial, the superior court ruled in favor of the Bank; the Court of Appeal affirmed, holding that California’s formula apportionment method was unconstitutional as applied to foreign-based unitary groups. In the view of the Court of Appeal, use of the Board’s method in such a case violated the foreign commerce clause in two respects. First, its application to a so-called “foreign parent” unitary business implicated foreign policy issues that were constitutionally required to be left to the federal government. Second, use of the formula apportionment method in the Bank’s case was at odds with a clear federal directive embodied in presidential and cabinet-level statements, letters, and press releases, task force reports and the congressional testimony of senior executive officials to the effect that American foreign commercial policy supported the use of an alternative accounting method to determine the taxable income of foreign-based corporations, one that is incompatible with formula apportionment.

As the Court of Appeal recognized, this suit is not the first litigation challenging on foreign commerce clause grounds the Board’s use of a three-factor formula to apportion the worldwide income of a multinational enterprise. In Container Corp. v. Franchise Tax Bd. (1983) 463 U.S. 159 [ 77 L.Ed.2d 545 , 103 S.Ct. 2933 ] (Container), the United States Supreme Court sustained against foreign commerce clause challenge the Board’s use of formula apportionment to determine the taxable income of a domestic-based unitary business group with foreign-domiciled subsidiaries. Despite the outcome in Container, the Bank successfully contended before the Court of Appeal that issues implicated by its foreign parentage are dispositive of the constitutional question and compel the opposite result in this case.

Although presented with a question left open in Container (supra, 463 U.S. at p. 189 , fns. 26 & 32 [ 77 L.Ed.2d at p. 568 ]), we approach its resolution along a path illuminated by the high court’s analysis in a series of recent opinions, Container among them, in the contemporary evolution of the “dormant foreign commerce clause” doctrine. Our opinion has two parts. As a prelude to the constitutional question, we first examine the extralegal issues raised by competing methodologies used to distribute multijurisdictional corporate income for state tax purposes; we then address the Bank’s central contention that a dormant foreign commerce clause analysis is appropriate here and that the Board’s application of formula apportionment to the Bank’s unitary business does not survive that analysis.

As we explain, neither of the two competing models used to allocate interjurisdictional income for state tax purposes is demonstrably superior to the other, even in an international multicorporate setting. Both methods meet the constitutional standard of avoiding “unreasonably” attributing extrastate value to the taxing jurisdiction. Moreover, the high court’s recent foreign commerce clause jurisprudence reflects a diminution in the reach of dormant foreign commerce clause analysis in favor of an expanded recognition that, under circumscribed conditions, governmental silence may constitute a ratification of state taxation of foreign commerce, rendering a dormant analysis inapposite. In our view, this is such a case.

II

Background: State Taxation of International Income

A

Limitations on taxation by the states of the income of corporations doing business in more than one jurisdiction inevitably implicate the sufficiency of quantitative measures used to identify that portion of taxable value reasonably attributable to the taxpayer’s intrastate activities, This pivotal role of technique arises from the stricture of the commerce and due process clauses of the federal Constitution that “a State may not tax value earned outside its borders.” (ASARCO Inc. v. Idaho State Tax Comm’n (1982) 458 U.S. 307, 315 [ 73 L.Ed.2d 787, 794 , 102 S.Ct. 3103 ] (ASARCO).) To meet this limitation, state tax schemes must comply with multiple criteria designed to produce a substantially accurate distribution of income so that only “values created by business within its borders” are taxed. (Butler Bros. v. McColgan (1942) 315 U.S. 501, 507 [ 85 L.Ed. 991 , 996, 62 S.Ct. 701 ] (Butler Bros.).)

Two distinct models have long competed for supremacy in identifying the required division of multijurisdictional income. One model, known as the “arm’s length/separate accounting” or “AL/SA” method, calculates income on a discrete and circumscribed basis, whether geographical, transactional, or functional. In a multicorporate interjurisdictional setting, the AL/SA method allocates income to a single taxing “sovereign” rather than apportioning it among jurisdictions, and treats intercorporate transfers of value between commonly held or related entities as if they were “arm’s length” transactions between unaffiliated businesses. There seems little reason to doubt that, as an operational matter, the AL/SA model is the dominant method employed by corporations both in the United States and internationally; that is, a majority of businesses use the AL/SA or a variant method for their own internal accounting purposes.

The competing model for taxation purposes is the “unitary business/ formula apportionment” method. Founded on the perception that “[i]n the case of a more-or-less integrated business enterprise operating in more than one State . . . arriving at precise territorial allocations of ‘value’ is often an elusive goal, both in theory and in practice” (Container, supra, 463 U.S. at p. 164 [ 77 L.Ed.2d at p. 552 ]), formula apportionment relies on mathematical generalization to distribute an aliquot share of income or taxable value among taxing jurisdictions. The dominant variation of formula apportionment—the so-called “three-factor” model employed by the Board in this case—defines the multijurisdictional scope of the unitary enterprise of which the taxable intrastate activities are a part, calculates the combined income of the components of the unitary group, and distributes a portion of that result to the taxing state using a mathematical formula based on an averaged ratio of property, payroll, and sales in the taxing jurisdiction to that of the unitary enterprise overall. (See Container, supra, 463 U.S. at p. 165 .)

So far as taxation of United States-derived income is concerned, the use of formula apportionment is both established and noncontroversial, being the preferred method of a majority of the states; a substantially smaller number of American jurisdictions—California among them—combine and apportion the worldwide income of multinational corporate taxpayers, a variant sometimes referred to as the “worldwide combined reporting” or “WWCR” method.

Although AL/SA is the method of choice for corporate operational and internal accounting purposes, its recognized deficiencies in purporting to locate and assign taxable value to a particular jurisdiction reduce its appeal among state tax administrators, the chief proponents of competing apportionment methods. These critics cite several shortcomings in the AL/SA method: comparative distortions in measuring income, and a resulting overtaxation or undertaxation; administrative complexity generated by the need to analyze thousands of intercorporate transactions; and the common absence of uncontrolled comparable prices by which to verify the value of intercorporate “arm’s length” transactions.

More fundamentally, critics of the AL/SA method have pointed to a theoretical failure of the model to account for income created by the effects of unitary interdependency. As the high court has stated in the case of a unitary business enterprise, “separate [geographical] accounting, while it purports to isolate portions of income received in various States, may fail to account for contributions to income resulting from functional integration, centralization of management, and economies of scale. [Citation.] Because these factors of profitability arise from the operation of the business as a whole, it becomes misleading to characterize the income of the business as having a single identifiable ‘source.’ ” (Mobil Oil Corp. v. Commissioner of Taxes (1980) 445 U.S. 425, 438 [ 63 L.Ed.2d 510, 521-522 , 100 S.Ct. 1223 ] (Mobil Oil).)

Of course, formula apportionment has its critics, as well. They too point to distortions in the measurement of taxable income, especially in a multinational setting where a relatively larger proportion of foreign to United States activities may result in overtaxation of the income of foreign-based unitary businesses; the substantial administrative burden of complying with income reporting requirements in United States dollars and accessing financial information that in some cases may be in the hands of literally hundreds of worldwide corporate affiliates; and the absence of uniform standards among the states for defining a unitary business. Not surprisingly, partisans on both sides of the issue contend that their method is the accepted standard—in the international arena in the case of AL/SA, and in the taxation of multijurisdictional unitary groups in the case of formula apportionment.

B

The United States Supreme Court first considered the sufficiency of formula apportionment in the constitutional sense over 70 years ago, upholding Connecticut’s use of a single-factor formula to apportion the income of a Connecticut business machine manufacturer whose products were marketed nationally. The high court rejected the taxpayer’s due process claim that the formula taxed business conducted “beyond title boundaries of the State.” As the Supreme Court explained, “[t]he profits of the [company are] largely earned by a series of transactions beginning with manufacture in Connecticut and ending with the sale in other states. . . . The legislature in attempting to put on this business its fair share of the burden of taxation was faced with the impossibility of allocating specifically the profits earned by the processes conducted within its borders.” (Underwood T’writer Co. v. Chamberlain (1920) 254 U.S. 113, 120-121 [ 65 L.Ed. 165, 169-170 , 41 S.Ct. 45 ] (Underwood).)

Since Underwood, supra, 254 U.S. 113 , the Supreme Court has confronted recurrent claims of the comparative superiority of each of these two theoretically irreconcilable techniques in responding to the distributive imperatives of the commerce clause. Despite claims of the surpassing merit of the separate accounting method, the court has refused to erect a “theoretical constitutional preference for one method of taxation over another” by mandating its use. (Mobil Oil, supra, 445 U.S. at p. 444 [ 63 L.Ed.2d at p. 525 ].) Likewise, it has rejected appeals to prescribe a variant of the formula apportionment method as a uniform national standard for interstate taxation. (Moorman Mfg. Co. v. Bair, supra, 437 U.S. 267, 278 [ 57 L.Ed.2d 197, 207-208 ] (Moorman).)

Although the high court has refused to impose “national uniform rules for the division of income” rooted in the commerce clause (Moorman, supra, 437 U.S. at p. 279 ), it has repeatedly validated the comparative empirical accuracy and constitutional adequacy of formula apportionment. Thus, not long after upholding the due process sufficiency of formula apportionment in Underwood, supra, 254 U.S. 113 , the court authorized its use in a multinational setting to apportion the combined income of a United Kingdom-headquartered brewer whose separate accounting showed no United States net income for the tax years at issue. Contending that the state was in effect taxing foreign income, the taxpayer asserted violations of the foreign commerce and due process clauses. (Bass, Etc., Ltd. v. Tax Comm. (1924) 266 U.S. 271 [ 69 L.Ed. 282 , 45 S.Ct. 82 ] (Bass).)

The high court rejected the challenge. The taxpayer’s business, the court explained, was a unitary enterprise conducted transnationally, “in which its profits were earned by a series of transactions beginning with the manufacture in England and ending in sales in New York and other places—the process of manufacturing resulting in no profits until it ends in sales . . . .” The state was thus justified “in attributing to [itself] a just proportion of the profits earned by the Company from such [a] unitary business.” Moreover, since the taxpayer had failed, as in Underwood, supra, 254 U.S. 113 , to show that formula apportionment “produced an unreasonable result,” its use was not unconstitutional. (Bass, supra, 266 U.S. at pp. 282, 283 [ 69 L.Ed. at p. 287 ].)

In Butler Bros., supra, 315 U.S. 501 , the taxpayer also challenged the state’s use of formula apportionment on the ground that separate accounting showed its sales office in the taxing jurisdiction had no net income for the tax year at issue and that apportionment thus resulted in the taxation of extraterritorial value. The court rejected this showing as insufficient, explaining that it “need not impeach the integrity of [the separate] accounting system to say that it does not prove [taxpayer’s] assertion that extraterritorial values are being taxed .... A particular accounting system, though useful or necessary as a business aid, may not fit the different requirements when a State seeks to tax values created by business within its borders.” (Id., at p. 507.)

And in Exxon Corp. v. Wisconsin Dept. of Revenue (1980) 447 U.S. 207 [ 65 L.Ed.2d 66 , 100 S.Ct. 2109 ], the court rejected the attempt of a vertically integrated, multistate petroleum company to demonstrate, based on its internal use of the separate accounting method applied to distinct operating divisions, that income was allocable to extrastate components and thus constitutionally was not subject to apportionment. “[A] company’s internal accounting techniques are not binding on a State for tax purposes,” the court wrote. “Exxon’s use of separate functional accounting . . . does not defeat the clear and sufficient nexus between [its] interstate activities and the taxing State” upon which the finding of corporate unity was based. (Id., at pp. 221 [65 L.Ed.2d at pp. 80, 82], 225.)

Again, in Mobil Oil, supra, 445 U.S. 425 , a nondomiciliary corporate taxpayer challenged the state’s inclusion in its apportionment formula of foreign-source dividend income received by the taxpayer from subsidiaries and affiliates, on the ground that its foreign origin made it constitutionally unapportionable. Although the taxpayer was able to isolate its foreign dividend income using separate accounting, the court observed that “the linchpin of apportionability in the field of state income taxation is the unitary business principle.” (Id., at p. 439 [ 63 L.Ed.2d at p. 522 ].) The divisibility of income produced by a separate accounting treatment, the court said, “may fail to account for contributions to income resulting from functional integration, centralization of management, and economies of scale .... Although separate geographical accounting may be useful for internal auditing, for purposes of state taxation it is not constitutionally required.” (Id., at p. 438 [ 63 L.Ed.2d at p. 521-522 ].) (See also ASARCO, supra, 458 U.S. 307 ; F. W. Woolworth Co. v. Taxation & Revenue Dept. (1982) 458 U.S. 354 [ 73 L.Ed.2d 819 , 102 S.Ct. 3128 ]; and Amerada Hess Corp. v. N. J. Taxation Div. (1989) 490 U.S. 66, 74 [ 104 L.Ed.2d 58, 67-68 , 109 S.Ct. 1617 ].)

In a slightly different context, the high court recently reaffirmed its views of both the theoretical problems inherent in locating the “source” of multijurisdictional income, and the validity for commerce clause purposes of the formula apportionment method. Last term, in upholding Michigan’s “value added” tax against commerce clause challenge, the court wrote that “the discrete components of a state income tax may appear in isolation susceptible of geographic designation. Nevertheless, since Underwood... we have recognized the impracticability of assuming that all income can be assigned to a single source.” (Trinova Corp. v. Michigan Dept. of Treasury, supra, 498 U.S. 358 , _ [ 112 L.Ed.2d 884, 907 , 111 S.Ct. 818, 831 ].)

The court went on to reiterate its statement in Container, supra, 463 U.S. at page 170 [ 77 L.Ed.2d at page 556 ], that the three-factor formula “has become something of a benchmark against which other apportionment formulas are judged,” noted its incorporation into UDITPA, adopted by almost half of the states, and acknowledged its accuracy in reflecting “the activities by which [taxable] value is generated.” “The same factors,” the court wrote, “that prevent determination of the geographic location where income is generated, factors such as functional integration, centralization of management, and economies of scale, make it impossible to determine the location of value added with exact precision.” (Trinova Corp. v. Michigan Dept. of Treasury, supra, 498 U.S. at p. _ [ 112 L.Ed.2d at p. 908 , 111 S.Ct. at p. 832 ].)

Thus, the rule that has emerged from this series of high court encounters—spanning over 70 years—with the contending merits of two “theoretically incommensurate” systems (Mobil Oil, supra, 445 U.S. at p. 444 [ 63 L.Ed.2d at p. 525 ]) is that, for state tax purposes, neither the commerce clause nor the due process clause of the federal Constitution mandates the use of a particular methodology to allocate or distribute multijurisdictional income. Either of the two principal methods and their variants are constitutionally permissible, as long as the one chosen does not operate “unreasonably and arbitrarily” to attribute to the taxing state a percentage of total income “out of all appropriate proportion to the business transacted by the [taxpayer] in that State.” (Hans Rees’ Sons v. No. Carolina (1931) 283 U.S. 123, 135 [ 75 L.Ed. 879, 907-908 , 51 S.Ct. 385 ].) Or, as the high court stated in Moorman, supra, 437 U.S. 267, 274 , “the States have wide latitude in the selection of apportionment formulas and ... a formula-produced assessment will only be disturbed when the taxpayer has proved by clear and cogent evidence that the income attributed to the State is in fact out of all appropriate proportion to the business transacted ... in that State.” (Internal quotation marks omitted.)

We present this extended account of the competing division-of-income methods and their treatment at the hands of the high court in order to meet at the outset the Bank’s unpersuasive contention that formula apportionment is an inherently inequitable method, at least when applied to foreign-based unitary groups such as the Bank and its affiliates. In substance, this argument is no different from the one explicitly rejected by the court in Container, supra, 463 U.S. 159 . There the taxpayer presented related challenges to California’s use of the three-factor formula to apportion the global income of a domestic-based unitary enterprise. Specifically, the taxpayer claimed that its foreign affiliates were significantly more profitable, and that by ignoring underlying economic realities such as lower wage and production costs among its foreign subsidiaries, three-factor formula apportionment systematically distorted the “true” allocation of income between unitary components, unfairly inflating the income apportioned to California.

“The problem with this argument,” the court said, “is . . . [that] the profit figures relied on by [the taxpayer] are based on precisely the sort of formal geographical accounting whose basic theoretical weaknesses justify resort to formula apportionment in the first place.” And the difficulty with the taxpayer’s evidence of differing costs, the court said, “is that it does not by itself come close to impeaching the basic rationale behind the three-factor formula. . . . [f] Both geographical accounting and formula apportionment are imperfect proxies for an ideal which is not only difficult to achieve in practice, but also difficult to describe in theory. . . .” ( Container, supra, 463 U.S. at pp. 181, 182 [77 L.Ed. 2d at pp. 563-564].) “Of course, even the three-factor formula is necessarily imperfect,” the court continued, “[b]ut we have seen no evidence demonstrating that the margin of error (systematic or not) inherent in the three-factor formula is greater than the margin of error (systematic or not) inherent in the sort of separate accounting urged upon us by [taxpayer].” (Id., at pp. 183-184.)

In light of this analysis and the precedents summarized above, we conclude that, in considering the Bank’s case for relief, neither of the two methods can lay claim to a decisive technical superiority or greater constitutional stature than the other. We turn, then, to the merits of the Bank’s contention that a dormant foreign commerce clause analysis renders unconstitutional the Board’s application of the three-factor apportionment formula to the Bank’s unitary enterprise.

III

The Dormant Foreign Commerce Clause Doctrine

Despite the framers’ explicit commitment to Congress of the power to “regulate commerce with foreign nations, and among the several states” (U.S. Const., art. I, § 8, cl. 3), by far the bulk of commerce clause jurisprudence has been developed by the high court itself under the judicially created doctrine of the “unexercised,” “negative,” or “dormant” commerce clause. From its origin early in the nation’s constitutional history with the opinion of Chief Justice Marshall in Gibbons v. Ogden (1824) 22 U.S. (9 Wheat.) 1, 209 [ 6 L.Ed. 23, 73 ], through its reformulation at the hands of Justice Curtis in Cooley v. Board of Wardens of Port of Philadelphia et al. (1852) 53 U.S. (12 How.) 299 [ 13 L.Ed. 996 ], the high court has posited irreducible and self-executing constitutional minima that limit state action affecting interstate and foreign commerce, even where Congress has failed to exert its plenary commerce clause power. “[T]he Commerce Clause was not merely an authorization to Congress to enact laws for the protection and encouragement of commerce among the States, but by its own force created an area of trade free from interference by the States. In short, the Commerce Clause even without implementing legislation by Congress is a limitation upon the power of the States. . . .” (Freeman v. Hewit (1946) 329 U.S. 249, 252 [ 91 L.Ed. 265 , 67 S.Ct. 274 ]; see also Southern Pacific Co. v. Arizona (1945) 325 U.S. 761 , 769 [ 89 L.Ed. 1915, 1924-1925 , 65 S.Ct. 1515 ] [“For a hundred years it has been accepted constitutional doctrine that the commerce clause, without the aid of Congressional legislation . . . affords some protection from state legislation inimical to the national commerce, and that in such cases, where Congress has not acted, this Court, and not the state legislature, is under the commerce clause the final arbiter of the competing demands of state and national interests. [Citations.]”]; Hood & Sons v. Du Mond (1949) 336 U.S. 525, 534 [ 93 L.Ed. 865 , 872, 69 S.Ct. 657 ]; Northwestern Cement Co. v. Minn. (1959) 358 U.S. 450, 458 [ 3 L.Ed.2d 421, 427 , 79 S.Ct. 357 , 67 A.L.R.2d 1292 ]; Hughes v. Oklahoma (1979) 441 U.S. 322, 326 , fns. 2 & 3 [ 60 L.Ed.2d 250 , 99 S.Ct. 1727 ]; Wyoming v. Oklahoma (1992) _ U.S. _, _ [ 117 L.Ed.2d 1, 21-24 , 112 S.Ct. 789, 799-801 ]

In the modern era, the consolidative work of the court led it to recast the interstate dimension of the dormant commerce clause doctrine as it applies to state taxation. In Complete Auto Transit, Inc. v. Brady, supra, 430 U.S. 274 , the court adopted a four-part test to evaluate state tax schemes to ensure compliance with the inherent demands of the “unexercised” commerce clause. If a tax “is applied to an activity with a substantial nexus with the taxing State, is fairly apportioned, does not discriminate against interstate commerce, and is fairly related to the services provided by the State,” it does not burden impermissibly interstate commerce. (Id., at p. 279 [ 51 L.Ed.2d at p. 331 ].)

This four-part test is not adequate, however, to subserve the additional policies underlying the foreign commerce clause. In Japan Line, Ltd. v. County of Los Angeles, supra, 441 U.S. 434 (Japan Line), the court held that “[w]hen construing Congress’ power to ‘regulate Commerce with foreign Nations,’ a more extensive constitutional inquiry is required.” (Id., at p. 446 [60 L.Ed.2d at p. 346].) Specifically, “two additional considerations, beyond those articulated in Complete Auto, come into play.” (Ibid.) “In addition to answering the nexus, apportionment, and nondiscrimination questions posed in Complete Auto, a court must also inquire, first, whether the tax, notwithstanding apportionment, creates a substantial risk of international multiple taxation, and second, whether the tax prevents the Federal Government from ‘speaking with one voice when regulating commercial relations with foreign governments.’ If a state tax contravenes either of these precepts, it is unconstitutional under the [foreign] Commerce Clause.” (Id., at p. 451, italics added.)

Although the court’s opinion in Japan Line, supra, 441 U.S. 434 , stressed the paramount national need for and plenary nature of congressional power over foreign commerce—greater, perhaps, than Congress’s textually parallel power over interstate commerce (id., at p. 448 & fns. 12-13 [60 L.Ed.2d at pp. 347-348])—as one of the fundamental policies animating the framers, it had little to say concerning the preemptive role of Congress in defining the contours of permissible state taxation of foreign commerce. The court illuminated that question a year later when it decided Mobil Oil, supra, 445 U.S. 425 , a challenge by Mobil to Vermont’s tax on foreign dividend income. Rejecting what it termed Mobil’s “forced” analogy to California’s tax on Japanese shipping containers at issue in Japan Line , the court said that the real issue before it was not one of multiple taxation at the international level, as in Japan Line , but of multiple taxation at the state level.

“Concurrent federal and state taxation of income, of course, is a well-established norm,” the court wrote. “Absent some explicit directive from Congress, we cannot infer that treatment of foreign income at the federal level mandates identical treatment by the States. The absence of any explicit directive to that effect is attested by the fact that Congress has long debated, but has not enacted, legislation designed to regulate state taxation of income. [Citations.]” (Mobil Oil, supra, 445 U.S. at p. 448 [63 L.Ed.2d at pp. 527-528], italics added.)

The view taken in Mobil Oil, supra, 445 U.S. 425 , that Congress, under its “ ‘exclusive and absolute’ power . . . over foreign commerce” (Japan Line, supra, 441 U.S. at p. 448, fn. 13 [60 L.Ed.2d at p. 348], quoting Buttfield v. Stranahan (1904) 192 U.S. 470, 492 [ 48 L.Ed. 525, 534 , 24 S.Ct. 349 ]), is the source, of “explicit directive[s]” preempting state taxation of foreign commerce, was amplified in the court’s treatment of the issue in Container, supra, 463 U.S. 159 . The court first noted that “allocating income among various jurisdictions bears some resemblance ... to slicing a shadow,” and concluded that “it would be perverse, simply for the sake of avoiding double taxation, to require California to give up one allocation method that sometimes results in double taxation [i.e., formula apportionment] in favor of another allocation method [i.e., AL/SA] that also sometimes results in double taxation.” (Id., at pp. 192-193 [77 L.Ed.2d at pp. 570-571].) It then turned to the “second inquiry suggested by Japan Line,” namely, whether California’s use of formula apportionment in an international context might “ ‘impair federal uniformity in an area where federal uniformity is essential,”’ and “‘prevent the Federal Government from “speaking with one voice” in international trade.’ ” (Ibid.)

In examining the “one voice” branch of the dormant commerce clause doctrine, the court in Container was careful to distinguish between state tax schemes that are unconstitutional because they implicate foreign affairs and those that, although they may have “foreign resonances,” are void because they “vio!ate[] a clear federal directive.” While either infirmity will offend the “one voice” standard, the latter does so, the court said, not as a result of a dormant commerce clause analysis, but because of “some explicit directive from Congress.” This latter analysis, the court noted, “is, of course, essentially a species of pre-emption . . . .” (Container, supra, 463 U.S. at p. 194 [77 L.Ed.2d at pp. 571-572], italics added.)

Applying the “explicit directive from Congress” language of Mobil Oil, supra, 445 U.S. 425 , and canvassing “specific indications of congressional intent,” the Container opinion found neither statutory preemption by Congress nor any requirement that the AL/SA method be applied to the states in any of the many bilateral tax treaties to which the United States was a signatory. Indeed, the court pointed out, “the Senate has on at least one occasion, in considering a proposed treaty, attached a reservation declining to give its consent to a provision in the treaty that would have extended [an AL/SA] restriction to the States.” (Container, supra, 463 U.S. at p. 196 [ 77 L.Ed.2d at p. 573 ].) “[I]t remains true,” the court concluded, “as we said in Mobil, that ‘Congress has long debated, but has not enacted, legislation designed to regulate state taxation of income.’ ” (Id., at p. 197 [ 77 L.Ed.2d at p. 573 ].)

Although the court concluded in Container that the Board’s use of formula apportionment did not violate the foreign commerce clause under a dormant analysis, it is important for our purposes to observe that the opinion carefully separates the analytical thread of congressional preemption from what the court termed a “more relaxed standard which takes into account our residual concern about the foreign policy implications of California’s tax.” (Container, supra, 463 U.S. at p. 197 [77 L.Ed.2d at pp. 573-574].) In the former case, there obviously is no need for resort to a dormant analysis since Congress has explicitly exerted its plenary authority under the commerce power to preempt a state tax scheme that, far from “implicating foreign affairs,” has only “foreign resonances.” As the Container opinion makes clear, however, the court will not lightly overturn the historic prerogatives of the states to administer their own tax systems. Before being declared nullified by Congress under this “species of pre-emption,” the “federal directive” that a state tax scheme allegedly violates must be “clear.” ( 463 U.S. at p. 194 [77 L.Ed.2d at pp. 571-572].)

The “more relaxed standard” of a dormant analysis—based on “residual concem[s]” about the “foreign policy implications” of a given state tax scheme—is appropriate only “in the absence of explicit action by Congress” preempting a given state taxation method. It is undertaken by a judiciary with “little competence in determining precisely when foreign nations will be offended by particular acts, and even less competence in deciding how to balance a particular risk of retaliation against the sovereign right of the United States as a whole to let the States tax as they please.” (Container, supra, 463 U.S. at p. 194 [77 L.Ed.2d at pp. 571-572].) Thus, it is for reasons of its limited foreign policy expertise that, in conducting a dormant “one voice” inquiry, the “best that [a court] can do, in the absence of explicit action by Congress [preempting a challenged state tax scheme], is to attempt to develop objective standards that reflect very general observations about the imperatives of international trade and international relations.” (Ibid.) Where they apply, the development of such standards is informed significantly by the views of executive officials charged with implementing the nation’s foreign policy—including its foreign commercial policy—“whose nuances ... are much more the province of the Executive Branch and Congress than of this Court.” (Id., at p. 196 [ 77 L.Ed.2d at p. 573 ].)

As might be expected, the Bank argues at length that the Board’s application of formula apportionment to the unitary business of which it is a part offends the foreign commerce clause precisely because it violates a “clear federal directive.” That directive, the Bank tells us, is embodied in an impressive array of federal executive communications—presidential statements and press releases; official letters from cabinet officers charged with overseeing national monetary, trade, and foreign policies; the testimony of senior Treasury Department officials; and the “white paper” of an executive task force established to study and present recommendations on the issues surrounding state use of formula apportionment to foreign-based unitary groups. Without exception, the Bank argues, these executive sources vigorously and explicitly support the exclusive use of the AL/SA method in such circumstances and condemn the use of formula apportionment by the states.

The Court of Appeal essentially agreed with the Bank’s underlying premise that executive pronouncements of what national foreign commercial policy should be qualifies as a source of the “clear federal directive.” It reasoned that because the Container opinion (supra, 463 U.S. 159 ) subsumed the search for specific indications of congressional intent within the “one voice” standard, congressional intent must therefore be an integral component of the dormant commerce clause test. From that analytical keystone, it had little difficulty in concluding on the basis of a substantial executive-compiled record that the federal directive was clear and explicit: according to executive branch officials, formula apportionment, in the words of the Court of Appeal, “is not to be applied to foreign-based corporate groups— those groups are to be taxed by the states only on income derived from the United States.”

Were we confronted with the Bank’s argument in the immediate aftermath of the Container decision, it might carry some force, although given the absence of textual support for the claim—both in the Container opinion and the commerce clause itself—it is by no means facially convincing. The fact is, however, that dormant foreign commerce clause jurisprudence has evolved in the nine years since Container, supra, 463 U.S. 159 , was decided, an evolution that, as we parse the cases, has reoriented the doctrine. That development has reduced the scope for a dormant analysis and makes its invocation here particularly inappropriate.

IV

Wardair: A “Governmental Silence" of a Different Kind

Three years after its opinion in Container, supra, 463 U.S. 159 , upholding California’s use of formula apportionment in calculating the intrastate tax liability of a domestic-based unitary business, the high court decided Wardair Canada v. Florida Dept. of Revenue (1986) 477 U.S. 1 [ 91 L.Ed.2d 1 , 106 S.Ct. 2369 ] (Wardair). At issue was a Canadian-based international air carrier’s challenge to an excise tax assessed by Florida on intrastate fuel purchases by common carriers, including airlines. The tax was levied at the rate of 5 percent on a deemed fuel price of $1,148 per gallon; air carriers were liable for the full amount of the tax whether the fuel was consumed in flights within or outside the state, regardless of the amount of intrastate business transacted by the taxpayer. (Id., at p. 4 [ 91 L.Ed.2d at p. 7 ].)

In attacking application of the tax to its Florida fuel purchases on foreign commerce clause grounds, the carrier—joined by the United States as amicus curiae—conceded that the tax, being assessed on discrete intrastate transactions, presented no risk of multiple international taxation. It relied instead entirely on the “one voice” element of the dormant foreign commerce clause analysis, the last of the six factors identified by the court in Japan Line, supra, 441 U.S. 434 .

Specifically, the carrier contended that a patchwork of reciprocal tax exemptions, embodied in a network of multilateral agreements and conventions to which the United States was a party, manifested a national policy to exempt from state taxation the instrumentalities of international air commerce, including aviation gasoline. Florida’s excise on aviation fuel purchased by a foreign carrier engaged in international air commerce, the carrier claimed, was inconsistent with that univocal national policy, thus threatening “the ability of the Federal Government to ‘speak with one voice.’ ” (Wardair, supra, 477 U.S. at p. 9 [ 91 L.Ed.2d at p. 10 ].)

The high court rejected the claim. Not only did the matrix of international conventions, resolutions and air commerce agreements relied on by the carrier and the United States fail to sustain a federal policy pretermitting Florida’s excise on aviation fuel, the high court said, “but, even more fundamentally, [it] shows also that in the context of this case we do not confront federal governmental silence of the sort that triggers dormant Commerce Clause analysis.” In point of fact, the court continued, “the international agreements cited demonstrate that the Federal Government has affirmatively acted, rather than remained silent, with respect to the power of the States to tax aviation fuel, and thus that the case does not call for dormant Commerce Clause analysis at all.” (Wardair, supra, 477 U.S. at p. 9 [ 91 L.Ed.2d at p. 10 ].)

As we explain below, the court’s opinion in Wardair, supra, 477 U.S. 1, establishes an interpretive framework for educing from a compilation of legislative materials a species of governmental silence that forecloses resort to a dormant foreign commerce clause analysis. In some cases where Congress affirmatively declines to adopt certain measures, the resulting governmental “silence” is not the sort that triggers use of a dormant foreign commerce clause analysis. Properly applied under the appropriate conditions, the Wardair methodology interdicts judicial resort to executive branch opinions as to the international commercial effect of a challenged state taxation practice because Congress has “acquiesced in” the contested practice, thereby validating it. (Id., at p. 12 [ 91 L.Ed.2d at p. 12 ].) Where appropriate, Wardair supplants what the court has termed the “quagmire” of dormant commerce clause analysis (Northwestern Cement Co. v. Minn., supra, 358 U.S. 450, 458 [ 3 L.Ed.2d 421, 427 ]) with a heightened judicial attentiveness to expressions of congressional foreign commerce policy. Because it delimits use of dormant foreign commerce clause analysis in important ways, it is useful to lay out the analytical lines of the Wardair paradigm in some detail before applying it to the case before us.

The foreign carrier in Wardair, supra, 477 U.S. 1, relied on a hierarchy of multinational agreements to support its thesis that federal policy precluded state taxation of intrastate aviation fuel purchases by international carriers: the Chicago Convention on International Civil Aviation (Convention), signed in 1944 by the United States, Canada, and 155 other nations; a resolution adopted in 1966 by the International Civil Aviation Organization, of which the United States was a member by virtue of being a signatory to the Convention; and more than 70 bilateral international aviation agreements between the United States and foreign nations, including a United States-Canadian aviation agreement. (Wardair, supra, at p. 10 [91 L.Ed.2d at p. 11].) But it was these very texts, the high court concluded, that in combination impeached the existence of a national policy to exempt international air commerce from state taxation and, by their “negative implications,” supported an inference that “the United States has at least acquiesced in state taxation of fuel used by foreign carriers in international travel." (Id., at p. 12 [ 91 L.Ed.2d at p. 12 ].)

First, a provision of the Convention explicitly prohibited taxation by both national and subnational governmental units of fuel “on board” arriving international aircraft. The Convention failed, however, to reach the issue of taxing intrastate fuel purchases by foreign aircraft following their arrival. This omission, the court reasoned, demonstrated by “negative implication” the “international community’s awareness of the problem of state and local taxation of international air travel . . . and represented] a decision by the parties to [the] Convention to address the problem by curtailing . . . only some of the localities’ power to tax, while implicitly preserving other aspects of that authority.” (Wardair, supra, 477 U.S. at p. 10 [ 91 L.Ed.2d at p. 11 ].)

Second, the resolution relied on by the carrier, although endorsing an international regime prohibiting duties of any kind by any taxing authority on international air travel, had “not been specifically endorsed, let alone signed, entered into, agreed upon, approved, or passed by either the Executive or Legislative Branch of the Federal Government. In other words, no action has been taken to give the Resolution the force of law.” (Wardair, supra, 477 U.S. at p. 11 [ 91 L.Ed.2d at p. 11 ].) It thus could not tenably represent, the court concluded, “a policy of the United States, as opposed to a policy of an organization of which the United States is one of many members.” (Ibid., italics in original.)

Third, in the years following the Convention, the United States entered into more than 70 bilateral international civil aviation agreements, “in not one of [which] has the United States agreed to deny the States the power asserted by Florida in this case.” Significantly, most of these agreements “explicitly commit the United States to refrain from imposing national taxes on aviation fuel used by airlines of the other contracting party . . . but. . . none . . . explicitly interdicts state or local taxes on aviation fuel used by foreign airlines in international traffic.” (Wardair, supra, 477 U.S. at p. 11 [ 91 L.Ed.2d at p. 11 ], italics in original, internal quotation marks omitted.) Reenforcing this view, the United States-Canadian agreement also limited tax exemptions granted foreign air carriers to “national duties and charges,” an “omission [to reach subnational duties] which must be understood as representing a policy choice by the contracting parties . . . .” (Ibid.)

Summing up the implications of this mosaic of texts for a dormant commerce clause attack on Florida’s tax, the court concluded that “[w]hat all of this makes abundantly clear is that the Federal Government has not remained silent with regard to the question whether States should have the power to impose taxes on aviation fuel used by foreign carriers in international travel.” (Wardair, supra, 477 U.S. at p. 12 [ 91 L.Ed.2d at p. 12 ].) “It would turn dormant Commerce Clause analysis entirely upside down,” the court continued, “to apply it where the Federal Government has acted, and to apply it in such a way as to reverse the policy that the Federal Government has elected to follow. For the dormant Commerce Clause, in both its interstate and foreign incarnations, only operates where the Federal Government has not spoken . . . .” (Ibid., first italics in original, second italics added.)

For our purposes, the high court’s analysis of the textual materials in Wardair, supra, 477 U.S. 1, can be abstracted into a kind of protocol for identifying those kinds of governmental silences that give rise to “negative implications” supporting an inference of federal acquiescence in the state tax under challenge. Thus, in Wardair the court found that bilateral recognition of an international taxation issue and its specific treatment at the national level (numerous international aviation agreements exempting foreign air carriers from national duties), impliedly supported a finding that the failure to address the correlative issue at the subnational level represented “a policy choice by the contracting parties.” (Id., at p. 11 [ 91 L.Ed.2d at p. 11 ].)

By a kind of parity of reasoning, Wardair found that the explicit treatment of some subnational aspects of an international taxation issue (Convention prohibition on state taxation of fuel aboard arriving foreign aircraft) supported an inference of international “awareness of the problem” at the state level and a “decision ... to address the problem by [limited curtailment of the subnational power] . . . while implicitly preserving other aspects of [subnational] authority.” ( 477 U.S. at p. 10 [ 91 L.Ed.2d at p. 11 ].) Last, the court found that notwithstanding an international aspiration to erect a particular tax regime (the resolution’s endorsement of the complete eradication of national and suhnational duties on international air travel), the fact that domestic “law as it presently stands acquiesces in taxation ... by political subdivisions” was decisive of the commerce clause issue. (Ibid., italics in original.)

Before considering the history of congressional consideration of curbs on the states’ application of formula apportionment to foreign-based multinationals in light of the court’s analysis in Wardair, we first consider the treatment of Wardair at the hands of the Court of Appeal.

V

The Wardair Canon and Congressional Refusal to Prohibit State Use of Formula Apportionment

A

As noted ante, the Court of Appeal declined to accept the view that Wardair, supra, 477 U.S. 1, represents, if not a change of course in the high court’s dormant foreign commerce clause jurisprudence, at least a retrenchment in its scope. It also concluded that statements of executive branch officials as to United States foreign commercial policy could constitute the “clear federal directive” component to the “one voice” analysis of Japan Line, supra, 441 U.S. 434 , Mobil Oil, supra, 445 U.S. 425 , and Container, supra, 463 U.S. 159 .

In our view, both of these conclusions are born of the root error of failing to grasp the conceptual impact of Wardair, supra, 477 U.S. 1, on dormant foreign commerce clause doctrine. As we explain, the failure of the Court of Appeal to appreciate Wardair's limitations on dormant commerce clause analysis is cognate to its erroneous view that executive branch aspirations as to what national foreign commercial policy ought to be can constitute a “clear federal directive,” at least where, under a Wardair analysis, Congress has decreed otherwise.

The Court of Appeal rejected the Board’s argument that Japan Une, Container, and Wardair demonstrate a trend in foreign commerce clause jurisprudence toward a heightened attention to governmental expressions of United States foreign commercial policy. Instead, it concluded that “the theoretical underpinning has remained intact through these cases.” What was different in them, it thought, “was the degree to which foreign affairs and international commercial relations were implicated” by the challenged state tax scheme.

It may be that the application of such a theme to these three cases would produce a coherent alternative explanation of the results reached by the high court. To contend, however, that Wardair, supra, 477 U.S. 1, turns on “the degree to which foreign affairs and international commercial relations were implicated,” is to misread fundamentally the court’s opinion. Such a view ignores the high court’s explicit statement that in Wardair it “[did] not confront federal governmental silence of the sort that triggers dormant Commerce Clause analysis,” that the case “does not call for dormant Commerce Clause analysis at all,” and that “[i]t would turn dormant Commerce Clause analysis entirely upside down to apply it where the Federal Government has acted.” (Id., at pp. 9, 12 [91 L.Ed.2d at pp. 10, 12].) We are confident that the overarching significance of Wardair lies in its explicit limitation on when a dormant foreign commerce clause analysis is appropriate, its affirmation that the analysis “only operates where the Federal Government has not spoken,” and its statement that the court “[has] never suggested . . . that the Foreign Commerce Clause insists that the Federal Government speak with any particular voice.” (Id., at pp. 12, 13 [91 L.Ed.2d at pp. 12, 13], italics in original.)

The Court of Appeal’s misapprehension of this central meaning of Ward-air, supra, 477 U.S. 1, led it to a related error—the conclusion, against the backdrop of an explicit congressional refusal to adopt curbs on state use of formula apportionment, that a trail of letters, press releases, task force reports, transcripts of congressional testimony of Treasury Department officials, and like communications orchestrated by the executive branch could constitute a “clear federal directive” condemning state use of formula apportionment in foreign parent cases. As we have indicated, however, the “clear federal directive” formulation in Container, supra, 463 U.S. 159 , has no role to play in a dormant foreign commerce clause analysis; rather, it confirms the preemptive power of Congress to interdict state tax schemes that would, had Congress not chosen to act, survive challenge under a dormant foreign commerce clause analysis because they present only “foreign resonances.” (Id., at p. 194 [77 L.Ed.2d at pp. 571-572].)

Whether, in the absence of a congressionally enacted “clear federal directive,” the executive branch can itself assume a preemptive role and in effect nullify state tax schemes as they affect foreign-based businesses is a distinctly different question. The Court of Appeal, relying principally on the concurring opinion of Justice Jackson in Youngstown Co. v. Sawyer (1952) 343 U.S. 579, 634 [ 96 L.Ed. 1153, 1198-1199 , 72 S.Ct. 863 , 26 A.L.R.2d 1378 ]—the famous “steel seizure case”—concluded that such a power inhered in the President’s authority to conduct foreign affairs and that the executive had spoken with sufficient clarity to prohibit state application of formula apportionment to foreign parent unitary groups such as the Bank.

In light of Wardair, supra, 477 U.S. 1, we need not and do not reach this issue. For in the debate over state use of worldwide formula apportionment—a controversy waged on multiple fronts by foreign governments, multinationals and their domestic and foreign affiliates, state tax authorities, senior Treasury and State Department officials, the White House and Congress—we hear the din of a “governmental silence” that cannot be ignored. In our view, Congress, after being repeatedly pushed and pulled in both directions, at least for the present has decided not to prohibit state use of formula apportionment in cases of this kind. To paraphrase the Wardair opinion, international agreements demonstrate that the federal government has affirmatively acted, rather than remained silent, with respect to the power of the states to employ formula apportionment in so-called foreign parent cases; as a result, this case does not call for dormant foreign commerce clause analysis at all. ( 477 U.S. at p. 9 [ 91 L.Ed.2d at p. 10 ].)

As we explain in some detail below, over the past 25 years, senior tax and foreign policy officials of the executive branch have sought to respond to the demands of foreign governments that the states be barred from applying formula apportionment to determine the tax liability of foreign-based multinationals. For much of this period, the chief forum for resulting executive branch initiatives was the Senate, where successive administrations sought to win ratification of a treaty provision barring state use of formula apportionment. But in 1978, with explicit recognition of the “negative implications” of its act, the Senate rejected an income tax convention negotiated by the executive branch with the United Kingdom whose centerpiece was a provision—article 9(4)—prohibiting the states from using formula apportionment in determining the intrastate tax liability of United Kingdom-based multinationals. Only after the ban on state use of formula apportionment was stricken was the administration able to muster the two-thirds majority required for Senate ratification under the treaty clause.

In addition, numerous bilateral tax treaties between the United States and other nations, although precluding use of formula apportionment by the signatory national governments, do not include within that prohibition political subdivisions such as the states. And while such tax treaties do include subnational governments within the scope of nondiscrimination provisions, they are not included within the prescription that the signatory governments employ an AL/SA methodology in taxing local branches of foreign corporations. Similarly, so-called “Friendship, Commerce and Navigation” treaties—a common form of commercial agreement between the United States and its trading partners—typically require both the federal and state governments to tax foreign enterprises within their jurisdictions on a “reasonably allocable or apportionable” basis, a standard which United States drafters viewed as “intended to cover all the various methods, proportionate or otherwise, by which a reasonable tax base might be determined.” (U.S. State Dept., Standard Draft Treaty of Friendship, Commerce and Navigation, prepared by Charles H. Sullivan (Aug. 1980) pp. 202, 203.)

Finally, in the wake of Senate rejection of article 9(4) and the decision in Container, supra, 463 U.S. 159 , upholding California’s use of worldwide unitary methods in taxing domestic-parent multinationals, the executive branch scuttled its effort to achieve treaty-imposed curbs on the states’ use of worldwide formula apportionment. Instead, the administration proceeded on other fronts, first seeking the voluntary cooperation of the states in mitigating the international effects of formula apportionment and, in a new tack, presenting its views in selected litigation through amicus curiae participation, before again renewing its call for Congress to enact restrictive legislation.

It is in the course of this undertaking that the executive branch has produced what the Bank insists is the “clear federal directive” that sustains its dormant foreign commerce clause case. We think, however, that rather than qualifying as a “clear federal directive,” these materials, in the context of Congress’s persistent refusal to regulate state taxation of multinationals and the Senate’s explicit rejection of article 9(4) of the United States-United Kingdom treaty, embody nothing more than executive aspirations of what the federal government’s policy in this area ought to be.

B

In 1975, the Ford administration concluded a bilateral income tax convention with the United Kingdom, a central feature of which was the following provision, article 9(4):

When the succeeding Carter administration sought Senate ratification of the convention, state tax authorities attacked article 9(4) as an infringement on state powers of taxation, reached by executive branch negotiators without prior consultation with the states. They vigorously lobbied for its excision. Following a series of committee and floor votes, the Senate ultimately ratified the treaty, subject to the telling reservation “that the provisions of paragraph (4) of Article 9 . . . shall not apply to any political subdivision or local authority of the United States.” (124 Cong. Rec. 18416 (1978).)

Explaining one of the motivations behind the reservation, which in effect struck article 9(4) from the treaty, its chief proponent objected to executive branch use of “the device of a tax treaty” to “impose major changes in internal tax policy” by circumventing congressional consideration of the states’ use of formula apportionment. (Remarks of Sen. Church, 124 Cong. Rec. 18416 (1978).) “For some 10 years,” the sponsor of the reservation stated, “Congress has been rejecting the type of limitation on the power of our State governments to tax which is incorporated in article 9(4) of the pending treaty.” (Ibid.) Immediately following ratification of the United States-United Kingdom tax convention as amended by the so-called “Church reservation,” Senator Javits, a chief sponsor of the treaty, including article 9(4), noted that “what we [the Senate] have done is very serious . . . We have for all practical purposes eliminated a very important provision of this treaty [article 9(4)] for which the United Kingdom believed it had entered into the treaty.” (Remarks of Sen. Javits, 124 Cong. Rec. 19077 (1978).)

The parallels between this evidence of “governmental silence” or refusal to act and that regarded as decisive in Wardair, supra, 477 U.S. 1 , seem to us both evident and compelling. As in Wardair, an international agreement (here the bilateral income tax treaty between the United States and the United Kingdom) demonstrates that while federal executive branch officials aspired to eliminate a state tax practice (here the use of formula apportionment to calculate the tax liability of foreign-based multinationals), “the law as it presently stands acquiesces” in the states’ continued use of that practice. As in Wardair, “the negative implications” of international agreements (here the tax treaty as ratified by the Senate) support recognition of a federal policy that acquiesces in the states’ tax practice. And certainly, in the circumstances of Senate consideration detailed above, the explicit removal of “political subdivisions” from the scope of article 9(4) effected by the Church reservation, like the omission of restrictions on taxation by political subdivisions in the international agreements considered in Wardair, “must be understood as representing a policy choice by the contracting parties.” (Wardair, supra, 477 U.S. at p. 11 [ 91 L.Ed.2d at p. 11 ].)

The Court of Appeal rejected the analogy to Wardair, supra, 477 U.S. 1 , on the ground that, in both a committee vote and on the Senate floor, the Church reservation failed to command a plurality, and the vote for the treaty with article 9(4) was only five votes short of the needed two-thirds majority. In light of these tallies, it reasoned that it was difficult to see a congressional policy permitting the states to use formula apportionment. In our view, however, the focus on preliminary votes misses the mark.

Preliminary voting tallies lack meaning precisely because they are not definitive, may be cast for any number of tactical parliamentary reasons, and thus do not reliably reflect legislative policy. The sole constitutional mechanism for congressional consideration of executive-negotiated treaties is Senate ratification by a two-thirds majority (U.S. Const., art. II, § 2); to that defining vote, institutional significance sensibly can and should be ascribed. In any case, the method enjoined upon us by Wardair, supra, 477 U.S. 1, requires that we ponder the significance of “the law as it stands,” not count noses.

The Senate’s action with respect to article 9(4) is only the most explicit example of a persistent congressional refusal to enact curbs on the states’ use of worldwide formula apportionment reaching back well before 1977, the tax year at issue in this case. The parties agreed in a pretrial stipulation that “various proposed Legislative bills have been introduced in the United States Congress that would, among other things, affect the states’ use of worldwide combined reporting.” The stipulation identifies twenty such House and Senate bills spanning twenty years. These range from House Resolution No. 11798 introduced in the House in 1965 (an ambitious “Interstate Taxation Act” that would have required the states to adopt a two-factor apportionment formula in taxing unitary groups) to 1985 legislation sponsored by the Treasury Department that would have limited state use of worldwide formula apportionment to members of foreign-based corporate groups actually doing business in the United States, that is, so-called “water’s edge” legislation. (See post, pp. 739-740.) None of these measures was enacted into law by Congress.

We likewise part company with the Court of Appeal in its view of the significance to be drawn from the web of bilateral tax conventions into which the United States has entered both before and after the Senate’s rejection of article 9(4). As noted, these conventions typically require use of the separate accounting method by the national signatory governments in their tax treatment of domestic branches of foreign-based businesses. This restriction, however, does not encompass division of income methods used by political subdivisions of the contracting states.

Again, the Court of Appeal found the analogy to Wardair, supra, 477 U.S. 1, inapt because, unlike that case—in which subsequent air commerce agreements rested on the foundational understanding of the Convention— many of the agreements relied on by the Board in this case predate an international awareness of formula apportionment issues that did not arise until the 1970’s. Thus, according to the Court of Appeal, affirmative prescriptions of national division of income methods in international treaties would not support Wardair's “negative implications” that the parties were aware of competing methods and made the conscious decision to acquiesce in their use by the states.

Despite the facial plausibility of this reasoning, we cannot accept the supposition underlying it that formula apportionment as a division of income alternative to separate accounting for taxation purposes did not penetrate the international financial and diplomatic consciousness until the multinational corporate boom of the 1970’s. The high court’s 1924 decision in Bass, supra, 266 U.S. 271 , upholding the constitutionality of state use of formula apportionment to a United Kingdom-based taxpayer, suggests at a minimum notice to the international business community of the valid use of an alternative to separate accounting by political subdivisions of the United States.

Although the Court of Appeal also rejected this argument on the ground that “awareness of a particular tax theory is one thing; to be subjected to that theory in practice is quite another,” we think that this reasoning requires greater proof than that demanded by Wardair, supra, 477 U.S. 1. The dramatic growth of multinationals during the 1960’s and 1970’s may well have served to sharpen corporate concern over the international effects of the states’ use of formula apportionment, but that increased apprehension does not demonstrate a prior lack of awareness of its potential applicability or negate the sensible decision of United States treaty negotiators—operating with an awareness of a federalist-based tax practice—to include a standard provision limiting only national governments to the AL/SA method.

There is evidence suggesting a basis for just such an international recognition of formula apportionment as a competing taxation method, at least from the mid-1950’s on. The Container case itself dealt with California’s use of formula apportionment for the 1963 to 1965 tax years ( 463 U.S. at p. 171 [ 77 L.Ed.2d at p. 557 ]) and a prior decision of our Court of Appeal (Anaconda Co. v. Franchise Tax Board (1982) 130 Cal.App.3d 15 [ 181 Cal.Rptr. 640 ]) dealt with the tax years 1955 through 1969. Moreover, in enacting the Internal Revenue Act of 1956, Congress authorized the Secretary of the Treasury to “distribute, apportion, or allocate gross income . . . between or among” multicorporate enterprises, including those with foreign domiciles, “in order . . . clearly to reflect [their] income” ( 26 U.S.C. § 482 , italics added), a formulation that reflects at least an awareness of apportionment methodologies. Similarly, as early as the late 1940’s, United States negotiators of “Freedom, Commerce and Navigation” agreements incorporated standards to preserve the states’ freedom to employ methods that produced a tax “reasonably allocable or apportionable” to the taxing jurisdiction, a formulation American drafters described as limiting state levies to “a fair portion of a global income derived from dual or multiple territorial sources.” (U.S. State Dept., Anno. Draft Treaty of Friendship, Commerce and Navigation for Portugal, prepared by Herman Walker (1947-1948) pp. 14a, 15.)

In addition, bilateral income tax treaties negotiated by the United States with many of its trading partners typically prescribe use of the AL/SA method by the signatory governments; they do not, however, impose such a requirement on taxation by subnational levels of government. Moreover, despite this methodological exemption, subnational organs of government are included for purposes of nondiscrimination treaty provisions. We think this latter evidence substantially parallels the Wardair paradigm, where the high court concluded that the “negative implications” arising from the Convention’s limited ban on state taxation of fuel “on board” arriving foreign aircraft demonstrated an awareness of subnational taxation issues and represented “a decision by the parties ... to address the problem by curtailing and limiting only some of the localities’ power to tax, while implicitly preserving other aspects of that authority.” (Wardair, supra, 477 U.S. at p. 10 [ 91 L.Ed.2d at p. 11 ].) In short, an extensive pattern of executive branch-negotiated diplomatic texts parallels, in our view, Congress’s own unwillingness to disallow legislatively the states’ application of formula apportionment methods to foreign-controlled multinational taxpayers, and bespeaks a coordinate “acquiescence.”

Finally, the strategy pursued by the executive branch in the wake of the high court’s opinion in Container, supra, 463 U.S. 159 , underlines the aspirational character of its insistence on an end to state use of formula apportionment in foreign-parent cases. Although in the immediate aftermath of the decision, the administration rejected appeals by the business community and major trading partners to seek amicus curiae status in Container and support a rehearing, the White House soon announced an alternative plan.

President Reagan directed the Secretary of the Treasury to establish a “working group” composed of representatives of the federal and state governments and the business community “charged with producing recommendations . . . that will be conducive to harmonious international economic relations, while also respecting the fiscal rights and privileges of the individual states.” (Chairman’s Rep., supra, atp. ii; see also 48 Fed.Reg. 208, p. 49570 (Oct. 26, 1983).) The final report of the working group, released in 1984 as the Chairman’s Report, ante, footnote 3, recommended efforts by the states to mitigate the international effects of formula apportionment, including limiting its use to the “water’s edge,” (that is, excluding from the unitary tax base members of foreign-controlled groups not doing business in the United States). Although these state efforts at reform were to be “voluntary,” in his transmittal letter to the President, Treasury Secretary Regan stated that “If there are not sufficient signs of appreciable progress by the states in this area by . . . next year, ... I will recommend to you that the Administration propose federal legislation that would give effect to a water’s edge limitation patterned after that in [the report].” (Chairman’s Rep., supra, at p. iii.)

Eighteen months later, the administration followed up on its threat to seek congressional action. In a November 1985 statement, the President called for federal legislation that would require the states to adopt “water’s edge” restrictions on the use of worldwide formula apportionment and directed the Treasury Secretary to “work with the Congress for [its] passage.” At the same time, the President directed the Attorney General “to ensure that the United States’ interests are represented in appropriate controversies and cases consistent with [the] approach” outlined in the President’s statement. (45 Weekly Compilation of Pres. Documents 1368.) In January 1986, Secretary of State Shultz wrote to then-Governor Deukmejian informing him that legislation had been introduced before Congress “at the express direction of the President” that “would prohibit states from taxing corporations under the worldwide unitary method.” (Letter of Jan. 30,1986, from Sect, of State Shultz to Governor Deukmejian.)

As a result of executive branch initiatives, some states agreed to cease using the formula apportionment method in the case of foreign-parent multinationals. Others, California among them, adopted meliorative measures designed to pacify critics. In 1986, our Legislature enacted “water’s edge” legislation, prompting senior Treasury Department officials to retreat from their previous insistence that Congress prohibit state use of worldwide formula apportionment. For the time being, the administration told Congress, neither prohibitory congressional legislation nor treaty restrictions on state use of worldwide unitary methods was appropriate.

As this account rather pointedly makes clear, the struggle for supremacy between the major interests with a stake in the fate of worldwide formula apportionment—states using that method and their federalist allies, on the one hand, and aggrieved multinationals and their supporters in the business community and the executive branch, on the other—has a long history. For much of this history, executive officials have sought unsuccessfully to impose their solution on the states through Senate ratification of a treaty provision embodying curbs and, failing that, legislation enacted under Congress’s foreign commerce clause power.

Only late in the campaign have the partisans of separate accounting changed the locus of attack by seeking to have the worldwide use of formula apportionment by the states judicially invalidated on dormant foreign commerce clause grounds. In pressing that challenge, opponents of the worldwide unitary method have marshalled executive branch arguments designed to persuade Congress and the states to legislate curbs, and have attempted to transmute them into the “clear federal directive” of Container, supra, 463 U.S. 159 . As explained above, however, that doctrine underlines the plenary power of Congress to preempt state taxation methods with “foreign resonances”; it does not give executive officials carte blanche to declare state tax methods null when they irritate our trading partners.

Conclusion

It is clear that federal limitations on the states’ use of worldwide formula apportionment is a controversial political and economic issue of which Congress has long been aware. In light of that history, we cannot turn away from the substantial evidence of Congress’s repeated refusal to intervene in the regulation of state division of income methods for tax purposes, even one that provokes continuing international complaint. Under the compulsions of established constitutional doctrine, the courts sometimes are required to divine what foreign commerce policy Congress would pursue in the absence of any indication that it has thought about the subject; it is a quite different matter, however, for a court to ignore a pattern of congressional action that evidences both an awareness of an issue and a refusal to adopt the remedy urged upon it by executive officials and resisted by its state constituencies. The latter, we believe, viewed in context alongside additional treaty materials, is a governmental silence that is eloquent.

In light of Congress’s awareness of antagonistic state taxation and international business interests, the path taken by the high court in Wardair, supra, 477 U.S. 1, seems the constitutionally correct one here. To invest a paper trail of executive aspiration with the dignity of a “clear federal directive” would, in the language of Wardair, “turn dormant Commerce Clause analysis entirely upside down.” (Id., at p. 12 [ 91 L.Ed.2d at p. 12 ].) Taking our lead from the high court, we decline to adjudicate on dormant foreign commerce clause grounds a debate between the political branches of the federal government over what is, in both its international and federalist dimensions, a sharply contested issue of national tax policy that has been repeatedly aired before Congress. We adhere to the central meaning of the high court’s opinion in Wardair in holding that Congress’s refusal to legislate restrictions on state use of worldwide formula apportionment is not the sort of governmental silence that triggers a dormant foreign commerce clause analysis.

Our holding does not end the matter, however. The trial court held that the cost to a foreign-based unitary enterprise of furnishing financial data required by the Board’s use of the worldwide formula apportionment method—the so-called “compliance burden”—violated due process. In addition, it held that same burden violated the nondiscrimination requirement of the four-part dormant interstate commerce clause analysis under Complete Auto Transit, Inc. v. Brady, supra, 430 U.S. 274 .

The Court of Appeal explicitly declined to decide the due process issue and does not appear to have passed directly on the nondiscrimination issue. The due process issue is a fact-dependent question that should be decided by the Court of Appeal in the first instance; moreover, we think its examination of the issue would profit from a consideration of its merit free of the view that a dormant foreign commerce clause analysis is appropriate in the circumstances present here.

Accordingly, the judgment of the Court of Appeal is reversed and the cause is remanded to that court for further proceedings consistent with this opinion.

Lucas, C. J., Mosk, J., Panelli, J., Kennard, J., Baxter, J., and George, J., concurred.

During 1977, the tax year at issue here, Revenue and Taxation Code section 25101 provided in relevant part: “When the income of a taxpayer subject to the tax imposed under this part is derived from or attributable to sources both within and without the state the tax shall be measured by the net income derived from or attributable to sources within this state in accordance with the provisions of Article 2 (commencing with Section 25120 of this chapter); . . ." Section 25120 et seq. of the Revenue and Taxation Code is California’s version of the Uniform Division of Income for Tax Purposes Act (UDITPA; see 7A West’s U. Laws Ann. (1985) p. 331); as explained more fully later in this opinion, the statute authorizes the use of a three-factor formula to apportion the net income from a taxpayer’s total business activities in order to determine the net income attributable to intrastate activities. (See post, p. 714 et seq.; Rev. & Tax. Code, §§ 25120-25140.) In this opinion, we sometimes use the term “formula apportionment” as a shorthand description of this three-factor mathematical formula employed by the Board to apportion for state tax purposes the interjurisdictional income of a worldwide or domestic unitary enterprise. Thus, the taxable income of a multijurisdictional unitary taxpayer in a state using the three-factor variant of formula apportionment would be calculated under the following equation: In-state In-state In-state Property + Payroll + Sales Total Total Total Total Income Property_Payroll_Sales x Corporate = Taxable by 3 Income the state (See Moorman Mfg. Co. v. Bair (1978) 437 U.S. 267, 283, fn. 1 [ 57 L.Ed.2d 197, 210-211 , 98 S.Ct. 2340 ] (Powell, J., dis.) [45 of 50 states use some form of income apportionment]; Trinova Corp. v. Michigan Dept. of Treasury (1991) 498 U.S. 358 , _ [ 112 L.Ed.2d 884, 906-909 , 111 S.Ct. 818, 831-832 ]; see also Chairman’s Rep. and Supplemental Views, Final Rep. of the Worldwide Unitary Taxation Working Group (Aug. 1984) (hereafter Chairman’s Report) p. 1 [all 45 states that levy a corporate income tax use formula apportionment to distribute taxable income of single multijurisdictional corporations]; U.S. General Accounting Office, Key Issues Affecting State Taxation of Multijurisdictional Corporate Income Need Resolving (July 1, 1982) GAO/GGD-82-38 (hereafter GAO Report), at appen. II, pp. 58-67 [tabular breakdown in use of apportionment variants among the states].) The critical literature assessing both methods is extensive. (For a sampling, see Chairman’s Rep., supra; GAO Rep., supra; Note, State Worldwide Unitary Taxation: The Foreign Parent Case (1985) 23 Colum. J. Transnat'l L. 445 ; Comment, California’s Corporate Franchise Tax: Taxation of Foreign Source Income? (1980) 20 Santa Clara L.Rev. 123; Rudy, The California Unitary Tax Concept as Applied to the Worldwide Activities of Foreign Corporations: A Modem Commerce Clause Analysis (1980-81) 15 U.S.F. L.Rev. 371; Note, Multinational Corporations and Income Allocation under Section 482 of the Internal Revenue Code (1975-1976) 89 Harv.L.Rev. 1202; Hellerstein, State Income Taxation of Multijurisdictional Corporations: Reflections on Mobil, Exxon, and H.R. 5076 (1978) 79 Mich. L.Rev. 113; Hellerstein, State Income Taxation of Multijurisdictional Corporations, Part II: Reflections on ASARCO and Woolworth (1982) 81 Mich. L.Rev. 157; Hellerstein, State Taxation Under the Commerce Clause: An Historical Perspective (1976) 29 Vand. L.Rev. 335; Corrigan, Interstate Corporate Income Taxation—Recent Revolutions and a Modem Response (1976) 29 Vand. L.Rev. 423; Langbein, The Unitary Method and the Myth of Arm’s Length (1986) 30 Tax Notes 625 ; see also Hearings Before the House Com. on Ways and Means on H.R. No. 5076, 96th Cong., 2d Sess. (1980).) (See, e.g., Chairman’s Rep., supra, at pp. 1-8; GAO Rep., supra, at pp. 32-40; and materials cited ante, at fn. 4.) The trial court found as a matter of fact that the AL/SA method was the “international standard of accounting” and that “[n]o other system is used internationally." To forestall any conceptual misunderstanding, we note that one of the two elements of the dormant foreign commerce clause analysis developed by the high court in Japan Line, Ltd. v. County of Los Angeles (1979) 441 U.S. 434 [ 60 L.Ed.2d 336 , 99 S.Ct. 1813 ] and subsequent cases (see post, this page et seq.) is the risk of multiple taxation posed by the challenged state taxation method. Although the high court’s assessment of the principal division of income methods speaks to that risk, the foregoing summary is undertaken for the limited purpose of gauging the comparative technical or “accounting” merits of the two models, not as part of a dormant commerce clause analysis. (See post, p. 729 et seq.) The Bank argues that the conclusion in the Container opinion (supra, 463 U.S. at p. 184 [ 77 L.Ed.2d 565 ]) that three-factor formula apportionment is a “proper and fair method of taxation” does not mean that its application to foreign-based multinationals produces an accurate determination of their intrastate income. The Container characterizations, the Bank contends, are merely the outcome of the four-part dormant interstate commerce clause analysis developed in Complete Auto Transit, Inc. v. Brady (1977) 430 U.S. 274 [ 51 L.Ed.2d 326 , 97 S.Ct. 1076 ]. This argument strikes us as semantical at best. The court made clear in Trinova Corp. v. Michigan Dept. of Treasury, supra, 498 U.S. 358 , _ [ 112 L.Ed.2d 884, 912 , 111 S.Ct. 818, 835 ], that if a state tax complies with the requirement of fair apportionment, constitutional demands are exhausted. Moreover, even under a dormant foreign commerce clause analysis, both the trial court and the Court of Appeal concluded—largely under the compulsion of the identical holding in Container, supra, 463 U.S. 159—that the Board’s use of worldwide formula apportionment in this case did not offend the risk-of-multiple-taxation leg of the dormant analysis. Compare the court’s statement in Merrion v. Jicarilla Apache Tribe (1982) 455 U.S. 130, 154-155 [ 71 L.Ed.2d 21, 40-41 , 102 S.Ct. 894 ], that “we only engage in [dormant analysis] review when Congress has not acted or purported to act. [Citation.] Once Congress acts, courts are not free to review state taxes or other regulations under the dormant Commerce Clause. When Congress has struck the balance it deems appropriate, the courts are no longer needed to prevent States from burdening commerce, and it matters not that the courts would invalidate the state tax or regulation under the Commerce Clause in the absence of congressional action. [Citation.] Courts are final arbiters under the Commerce Clause only when Congress has not acted.” (Convention Between United States and United Kingdom for Avoidance of Double Taxation, Dec. 31, 1975, 31 U.S.T. 5670, 5677, T.I.A.S. 9682.) Parliament eventually acceded to Congress’s demands, ratifying the income tax convention without article 9(4). (See 31 U.S.T., supra, 5709-5710.) Representative treaty language appears, for example, in article VII, section 2 of the United States-Canadian treaty of 1984, providing for tax treatment of domestic branches of foreign-domiciled corporations as “if it were a distinct and separate person engaged in the same or similar activities under the same or similar conditions and dealing wholly independently” with the rest of the corporation. (Convention Between United States and Canada with Respect to Taxes on Income and Capital, Aug. 16, 1984.) Thus, one of the leading spokesmen of executive branch opposition to state use of the worldwide unitary method, Donald Lubick, Assistant Secretary of the Treasury for Tax Policy, explained why, except for the nondiscrimination clause, subnational taxes were not included in the United States Model Income Tax Treaty by noting that “local U.S. taxes are not covered because it is unlikely that the United States would consent to the ratification of any treaty provision that restricted the rights of the . . . states to impose their own taxes.” (See International Tax Treaties: Hearing Before the Sen. Com. on Foreign Relations, 96th Cong., 1st Sess., at p. 112.) The United States (together with two other federalist nations, Canada and Australia) likewise reserved its position “on that part of paragraph 1 [of article 1 of the Organization for Economic Co-operation and Development Model Taxation Convention] which states that the Convention [which requires use of a separate accounting method] should apply to taxes of political subdivisions or local authorities.” (Model Double Taxation Convention on Income and on Capital (1977), at p. 50.) A key 1984 Treasury Department document notes that debate on state use of the worldwide version of formula apportionment “spans at least two decades'' (Chairman’s Rep., supra, at p. 3, italics added.) The 1954 version, as well as the current version of 26 United States Code section 482 , derives from section 45 of the Revenue Act of 1928. (See H.R. Rep. No. 2, 70th Cong., 1st. Sess., p. 16 (1928).) Pleas for administration support for a rehearing and for congressional legislation and the administration’s reaction are detailed in the Chairman’s Report, supra, at pages 2-3. Neither the Senate nor the companion House versions of the Treasury Department-drafted legislation were enacted. (See Remarks of Sen. Wilson on Sen. No. 1974, 131 Cong. Rec. S17975, daily ed. Dec. 18, 1985; remarks of Representative Duncan on H.R. No. 3980, 131 Cong. Rec. E574, daily ed. Dec. 19, 1985.) On the legal front, the Solicitor General filed a brief amicus curiae in Wardair, supra, 477 U.S. 1, supporting the foreign air carrier in its unsuccessful contention that Florida’s tax on aviation fuel was unconstitutional under a dormant foreign commerce clause analysis. (Id., at p. 9 [ 91 L.Ed.2d at p. 10 ].) (See, e.g., Langbein, The Unitary Method and the Myth of Arm’s Length, supra, 30 Tax Notes at p. 674, fn. 1. [by early 1986, four of the twelve states using worldwide unitary methods, “including all of the commercially significant states except California,” had repealed them].) (See Rev. & Tax. Code, § 25110 et seq.) (Statement of J. Roger Mentz, Asst. Sect. of the Treas. for Tax Policy, before the Subcom. on Taxation and Debt Management, Sen. Finance Com. (Sept. 29, 1986), at pp. 2-3.) The Assistant Secretary’s comments, stating the administration’s position on Senate Bill No. 1974—the Treasury Department-sponsored legislation drawn up at President Reagan’s direction—included the view that “Congressional action on S. 1974 should be deferred until the remaining worldwide unitary states have a full opportunity to act . . .” and that “a treaty resolution of the unitary issue is [not] necessary or appropriate at this time.” (Ibid.) These materials consist largely of numerous diplomatic notes of complaint (“demarches”) filed with the United States by its trading partners objecting to the states’ use of worldwide formula apportionment, presidential statements, statements and letters from cabinet officers and other senior Treasury and State Department officials to the same effect, and official documents such as the Chairman’s Report, supra. As noted, there is no doubt that many foreign governments object strenuously to the practice and that executive branch officials charged with conducting American foreign commercial policy agree with them. We thus reject the claim of the United States Department of Justice, appearing as amicus curiae in support of the Bank, that California’s use of formula apportionment in this case is “an egregious interference with the Federal Executive’s conduct of foreign affairs and is thus patently unconstitutional.” Although we accept the Justice Department’s argument that the views of the executive branch on the international effect of state taxation practices are entitled to “great weight” under a foreign dormant commerce clause analysis (cf. Container, supra, 463 U.S. at p. 196 [ 77 L.Ed.2d at p. 573 ]), our conclusion that such an analysis is not triggered here forecloses resort to those views. (Compare Wardair, supra, 477 U.S. at p. 9 [ 91 L.Ed.2d at p. 10 ] [rejecting views of United States as amicus curiae].)