67 F.3d 1445
76 A.F.T.R.2d 95-6825, 95-2 USTC P 50,581,
95 Cal. Daily Op. Serv. 8089,
95 Daily Journal D.A.R. 13,903
INTEL CORPORATION AND CONSOLIDATED SUBSIDIARIES,
Petitioner-Appellee-Cross-Appellant,
v.
COMMISSIONER, INTERNAL REVENUE SERVICE,
Respondent-Appellant-Cross-Appellee.
Nos. 94-70105, 94-70158.
United States Court of Appeals,
Ninth Circuit.
Argued and Submitted Aug. 15, 1995.
Decided Oct. 16, 1995.
Wayne S. Kaplan argued the case. With him on brief were Joel V. Williamson, Roger J. Jones, and Patricia Anne Flaming, Mayer, Brown & Platt, Chicago, Illinois, and James P. Fuller and Catherine L. Curtiss, Fenwick & West, Palo Alto, California, for petitioner-appellee-cross-appellant.
Frank P. Cihlar, Gary R. Allen, Tax Division, United States Department of Justice, Washington, D.C., for the respondent-appellant-cross-appellee.
On Appeal from the Decision of the United States Tax Court.
Before: SNEED, KOZINSKI, and NOONAN, Circuit Judges.
SNEED, Circuit Judge:
The Commissioner of Internal Revenue sought recovery of taxes from Intel Corporation and Consolidated Subsidiaries ("Intel") for the years 1978 through 1980 in the respective amounts of $4,660,900, $6,539,152, and $24,676,625. Intel timely filed its petition in the Tax Court on September 19, 1989.
Settlement negotiations disposed of some issues, but others were unresolved. After a recomputation of the deficiencies, the Tax Court on December 9, 1993 determined that Intel overpaid taxes in the amount of $87,476, $10,294,785, and $2,524,388 for the years 1978, 1979 and 1980 respectively. The Commissioner appeals from the decision of the Tax Court, under its heading "Export Source Issue," while Intel appeals from the decision ordered under the heading "R & E Allocation Moratorium Issue."
We affirm each of the Tax Court's decisions.
I.
AN OVERVIEW
Few areas of federal income tax law rival in complexity the taxation of foreign source income. Not only must the indices of domestic and foreign source income be established, but its treatment by both the United States and the source nation must be determined.
The United States does not cede to the source nation the exclusive right to tax such income although it does provide a credit against the taxes it imposes for foreign taxes paid with respect to such income. To this extent and in this manner the United States does surrender a portion of its tax base to the foreign source. This "sharing of the tax base" with a foreign nation may be governed by tax treaties which supplant the tax regimes of both nations.
Moreover, deductions must also be apportioned between domestic and foreign sources in a rational manner. This too is not always a simple matter.
The dominant incentives applicable to a dispute between the Commissioner and a domestic taxpayer, such as Intel, having domestic and foreign income and deductions attributable to both sources are reasonably clear. The Commissioner ordinarily will seek to diminish the amount of foreign source income and allocate to that same source as large an amount of the deductions as feasible. In this manner the extent of cession of the United States tax base to a foreign nation is limited and the tax revenue to the United States is increased. The domestic taxpayer, on the other hand, reacts in the reverse fashion. Deductions should be brought home and income sent abroad.
These are the incentives that drive this case.
II.
SOURCING OF INTEL'S INCOME ON ITS FOREIGN SALES
Intel sells its products to customers abroad by arranging for title to pass within the foreign nation in which the customer takes delivery. The Commissioner seeks to enlarge the amount of income to be treated as having its source in the United States by insisting on applying what is known as the "Independent Factory or Production Price Method."
Under this approach, Commissioner argues that despite the absence of a foreign branch in the country to which Intel's products are shipped, and in which the title of such products passes, Intel should be compelled to treat as an amount realized on the sale equal to the "Independent Factory Price" and that any such gain should be United States income, not income having a source in the foreign country in which the product was sold.
Intel, on the other hand, contends that the point at which the title passes determines the source of at least a portion of the income. Intel rejects the Commissioner's use of the "Independent Factory Price" method and argues for an apportionment method that allocates more income to the foreign nation in which the title to the property passed.
Both the Commissioner and Intel focus on Treasury Regulation Sec. 1.863-3(b) (1981), in which three allocation methods of income having both domestic and foreign sources are set forth. The dispute centers on Examples (sometimes designated as Cases) 1 and 2, it being agreed the third example is not relevant. Example 1, the text of which is set out below, requires the use of the Independent Factory Price allocation method, while Example 2, the text of which is also set out below, does not.
The Commissioner contends that Intel must allocate its income from foreign sales pursuant to Example 1 even though it has no "independent distributor or other selling concerns" located in a foreign country. Were this proper, Intel's independent factory or production price would be determined, and the property destined for foreign customers would be treated "as sold by the factory ... at the independent factory price so established."
In rejecting the position of the Commissioner the Tax Court pointed out that section 863 of the Code has its origins in section 217(e) of the Revenue Act of 1921, ch. 136, 42 Stat. 227, 244-45 (1923), the pertinent provisions of which have been retained without material alteration. Also, Examples (1) and (2) of the present Treas.Reg. Sec. 1.863-3(b)(2) (1994) are substantially "Case 1 and Case 2" of Regs. 62, art. 327 (1922) promulgated pursuant to the Revenue Act of 1921.
Under these early regulations in a Case 2 situation the net income from cross-border sales was divided into two equal parts, one to reflect the gain from manufacturing and the other from marketing. The manufacturing portion was determined by multiplying one half the net income by a fraction: the numerator of which was the value of the taxpayer's property, held or used to produce cross-border income within the United States, and the denominator of which was the value of such taxpayer's property both within and without the United States. The marketing portion was determined by multiplying the remaining half of the net income by another fraction.
Free access — add to your briefcase to read the full text and ask questions with AI
67 F.3d 1445
76 A.F.T.R.2d 95-6825, 95-2 USTC P 50,581,
95 Cal. Daily Op. Serv. 8089,
95 Daily Journal D.A.R. 13,903
INTEL CORPORATION AND CONSOLIDATED SUBSIDIARIES,
Petitioner-Appellee-Cross-Appellant,
v.
COMMISSIONER, INTERNAL REVENUE SERVICE,
Respondent-Appellant-Cross-Appellee.
Nos. 94-70105, 94-70158.
United States Court of Appeals,
Ninth Circuit.
Argued and Submitted Aug. 15, 1995.
Decided Oct. 16, 1995.
Wayne S. Kaplan argued the case. With him on brief were Joel V. Williamson, Roger J. Jones, and Patricia Anne Flaming, Mayer, Brown & Platt, Chicago, Illinois, and James P. Fuller and Catherine L. Curtiss, Fenwick & West, Palo Alto, California, for petitioner-appellee-cross-appellant.
Frank P. Cihlar, Gary R. Allen, Tax Division, United States Department of Justice, Washington, D.C., for the respondent-appellant-cross-appellee.
On Appeal from the Decision of the United States Tax Court.
Before: SNEED, KOZINSKI, and NOONAN, Circuit Judges.
SNEED, Circuit Judge:
The Commissioner of Internal Revenue sought recovery of taxes from Intel Corporation and Consolidated Subsidiaries ("Intel") for the years 1978 through 1980 in the respective amounts of $4,660,900, $6,539,152, and $24,676,625. Intel timely filed its petition in the Tax Court on September 19, 1989.
Settlement negotiations disposed of some issues, but others were unresolved. After a recomputation of the deficiencies, the Tax Court on December 9, 1993 determined that Intel overpaid taxes in the amount of $87,476, $10,294,785, and $2,524,388 for the years 1978, 1979 and 1980 respectively. The Commissioner appeals from the decision of the Tax Court, under its heading "Export Source Issue," while Intel appeals from the decision ordered under the heading "R & E Allocation Moratorium Issue."
We affirm each of the Tax Court's decisions.
I.
AN OVERVIEW
Few areas of federal income tax law rival in complexity the taxation of foreign source income. Not only must the indices of domestic and foreign source income be established, but its treatment by both the United States and the source nation must be determined.
The United States does not cede to the source nation the exclusive right to tax such income although it does provide a credit against the taxes it imposes for foreign taxes paid with respect to such income. To this extent and in this manner the United States does surrender a portion of its tax base to the foreign source. This "sharing of the tax base" with a foreign nation may be governed by tax treaties which supplant the tax regimes of both nations.
Moreover, deductions must also be apportioned between domestic and foreign sources in a rational manner. This too is not always a simple matter.
The dominant incentives applicable to a dispute between the Commissioner and a domestic taxpayer, such as Intel, having domestic and foreign income and deductions attributable to both sources are reasonably clear. The Commissioner ordinarily will seek to diminish the amount of foreign source income and allocate to that same source as large an amount of the deductions as feasible. In this manner the extent of cession of the United States tax base to a foreign nation is limited and the tax revenue to the United States is increased. The domestic taxpayer, on the other hand, reacts in the reverse fashion. Deductions should be brought home and income sent abroad.
These are the incentives that drive this case.
II.
SOURCING OF INTEL'S INCOME ON ITS FOREIGN SALES
Intel sells its products to customers abroad by arranging for title to pass within the foreign nation in which the customer takes delivery. The Commissioner seeks to enlarge the amount of income to be treated as having its source in the United States by insisting on applying what is known as the "Independent Factory or Production Price Method."
Under this approach, Commissioner argues that despite the absence of a foreign branch in the country to which Intel's products are shipped, and in which the title of such products passes, Intel should be compelled to treat as an amount realized on the sale equal to the "Independent Factory Price" and that any such gain should be United States income, not income having a source in the foreign country in which the product was sold.
Intel, on the other hand, contends that the point at which the title passes determines the source of at least a portion of the income. Intel rejects the Commissioner's use of the "Independent Factory Price" method and argues for an apportionment method that allocates more income to the foreign nation in which the title to the property passed.
Both the Commissioner and Intel focus on Treasury Regulation Sec. 1.863-3(b) (1981), in which three allocation methods of income having both domestic and foreign sources are set forth. The dispute centers on Examples (sometimes designated as Cases) 1 and 2, it being agreed the third example is not relevant. Example 1, the text of which is set out below, requires the use of the Independent Factory Price allocation method, while Example 2, the text of which is also set out below, does not.
The Commissioner contends that Intel must allocate its income from foreign sales pursuant to Example 1 even though it has no "independent distributor or other selling concerns" located in a foreign country. Were this proper, Intel's independent factory or production price would be determined, and the property destined for foreign customers would be treated "as sold by the factory ... at the independent factory price so established."
In rejecting the position of the Commissioner the Tax Court pointed out that section 863 of the Code has its origins in section 217(e) of the Revenue Act of 1921, ch. 136, 42 Stat. 227, 244-45 (1923), the pertinent provisions of which have been retained without material alteration. Also, Examples (1) and (2) of the present Treas.Reg. Sec. 1.863-3(b)(2) (1994) are substantially "Case 1 and Case 2" of Regs. 62, art. 327 (1922) promulgated pursuant to the Revenue Act of 1921.
Under these early regulations in a Case 2 situation the net income from cross-border sales was divided into two equal parts, one to reflect the gain from manufacturing and the other from marketing. The manufacturing portion was determined by multiplying one half the net income by a fraction: the numerator of which was the value of the taxpayer's property, held or used to produce cross-border income within the United States, and the denominator of which was the value of such taxpayer's property both within and without the United States. The marketing portion was determined by multiplying the remaining half of the net income by another fraction. This one had a numerator consisting of the gross sales of the taxable period within the United States and the denominator being the gross sales both within and without the United States. "Gross sales within the United States" were defined to mean "the aggregate amount of all sales made during the taxable year which were principally secured, negotiated or effected by employees, agents, offices, or branches of the taxpayer's business resident or located in the United States." Case 2, Regs. 62, art. 327 (1922). While "gross sales without the United States" was not explicitly defined, it might suggest that, to source income in a foreign nation, some significant contact with that nation was necessary.
In this fashion these regulations implemented the purpose of Congress in enacting section 217(e) of the Revenue Act of 1921 to secure a mixed source treatment of income derived from cross-border sales.
Intel's contention that its cross border sales should be allocated in the manner set forth in Example 2 in the absence of any "independent distributors or other selling concerns" would appear to be even more unassailable after the adoption in 1957 of the so-called "title passage" rule in Treas.Reg. Sec. 1.861-7 (1981). That regulation in subpart (c) states in relevant part:
For the purposes of part I (section 861 and following), subchapter N, chapter 1 of the Code, and the regulations thereunder, a sale of personal property is consummated at the time when, and the place where, the rights, title, and interest of the seller in the property are transferred to the buyer.
Intel's foreign sales, to repeat, are fashioned to have title and risk of loss to pass in a foreign jurisdiction, and thus were consummated abroad and must produce at least some foreign-source income.
Although not applicable to the taxpayers involved here, in 1988 the Regulations with respect to Example 2 were changed significantly. However, its opening sentence explicitly made it applicable, as before, only "(1) [w]here an independent factory or production price has not been established as provided under Example (1)." Thus, Example (1), with its requirement of sales by a taxpayer "of part of his output to wholly independent distributors or other selling concerns ...," remains an obstacle to requiring Intel to apportion its sales according to the "independent factory price" method.
What the Commissioner overlooks in seeking adjustments of $42 million of allowable foreign tax credits for the years 1978 through 1981 is that the remedy it invokes is authorized by neither the statutes nor the regulations thereunder. We, like the Tax Court before us, are asked to justify an amendment either to Example (1) or (2) in Treas.Reg. Sec. 1.863-3(b)(2) (1981) which, under the facts of this case, would source most of Intel's gross income on its export sales in the United States despite the absence of "wholly independent distributors or other selling concerns." Example (1), Treas.Reg. Sec. 1.863-3(b)(2) (1981). Moreover, we agree with the Tax Court that Congress, by enacting the Revenue Act of 1921, gave the Treasury authority to promulgate regulations thereunder of which Example 1 was a part. We decline to alter its requirements.
We recognize that our holding will reduce the taxes Intel pays to the United States. No doubt Intel fashioned a method of doing business that reduced its taxes. It is for Congress to determine whether this favorable treatment of export sales should continue. The third branch of government is not equipped to decide issues embedded not only in the nation's arcane tax law of longstanding but also in its foreign trade policies and balance of payment concerns.
III.
THE ALLOCATION AND DEDUCTIBILITY OF R & E EXPENSES UNDER ERTA
The heart of the issue in dispute under this portion of this case is whether, under the facts of this case, a portion of Intel's research and experimental expenditures must be taken into account in calculating the "combined taxable income" of Intel and its "domestic international sales corporation" (DISC), under I.R.C. Sec. 994 (1982). The positions of both Intel and the Commissioner reflect their usual perspectives, although certainly in a different and unusual situation.
Although the fundamental issue is easy to state, its context is laboriously complex. Because we agree with the Eighth Circuit's analysis of this issue in a similar case, St. Jude Medical, Inc. v. C.I.R., 34 F.3d 1394 (1994), our analysis will be somewhat abbreviated.
We begin with a brief description of a domestic international sales corporation, commonly referred to as a DISC, authorized by I.R.C. Secs. 991-997 (1982). This legislation was enacted in 1971 to enable domestic exporters to enjoy a deferral of tax on their export earnings somewhat similar to that available to exporters who utilized foreign subsidiaries. It authorized the formation of a domestic subsidiary which, when designed to comply with the authorizing legislation, would be exempt from taxation. To qualify as a DISC, I.R.C. Sec. 992(a)(1)(A) (1982) required that "95 percent or more of the gross receipts of such corporation consist of qualified export receipts." I.R.C. Sec. 995(b)(1), (2) (1982) taxed the DISC's shareholders (principally its parent corporation) on a percentage of the DISC's profits as a deemed dividend, and deferred taxation on the remainder of those profits until either dissolution or disqualification of the DISC. The parent corporation of the DISC could either sell its product to the DISC for export or resale or retain title to the goods and pay a commission to the DISC for selling the goods abroad.
Obviously it was necessary to control the amount of export income that the parent could defer by use of a DISC. This was accomplished by means of I.R.C. Sec. 994 (1982) (original version at Revenue Act of 1971, Pub.L. No. 92-178, 1971 U.S.C.C.A.N. (85 Stat.) 545, 606), and the regulations thereunder. Section 994 provides three methods for determining the "transfer price" between the parent corporation and its DISC, the most favorable to Intel being an amount which does not exceed 50 percent of the "combined taxable income" of the DISC and its parent attributable to export sales. I.R.C. Sec. 994(a)(2) (1982). It is here the term "combined taxable income" appears. Its function is to fix a "transfer price" between the parent and the DISC which, in turn, limits the amount of income of which the parent can defer realization.
Parent corporations who do not sell to their DISC, but rather employ a commission payment for the services rendered by their DISC, are given no specific guidance by section 994(b) as to the limit of the income they can defer. Rather, the Treasury was directed to promulgate regulations addressing this mode of operation "which are consistent with the rules set forth in subsection (a)." That subsection, of course, sets forth the pricing rules applicable to sales by the parent to its DISC. The regulatory response appears in Treas.Reg. Sec. 1.994-1(d)(2) (1981). Under its provisions, the "combined taxable income" method of pricing, when selected by the taxpayer, also limits the commissions attributable to a DISC which sells the parent's product abroad on a commission basis.
The term "combined taxable income," which section 994(a) of the Code employs to limit a DISC's income deferral utility, is defined by Treas.Reg. Sec. 1.994-1(c)(6)(iii) (1981) to include a ratable share of overhead costs, apportioned according to the income allocation rules in Treas.Reg. Sec. 1.861-8 (1981). Under the original version of section 1.861-8, taxpayers could allocate to domestic sources income that the Treasury thought more properly were allocable to foreign operations, and domestic source income improperly to foreign sources. Again, there emerges the fundamental inclinations of both Treasury and taxpayers engaged in export trade. To repeat, the former leans toward "exporting" deductions and "importing" income, while the latter's inclinations are the reverse. The Treasury seeks to protect its tax base against erosion by the foreign tax credit, while the exporters seek to reduce current domestic liability and defer taxes on export income as long as possible.
This period of confusion was brought to an end in 1977 by the promulgation of a revised Treas.Reg. Sec. 1.861-8 (1981), which the United States asserts was designed to assure that both domestic and foreign operations of corporate taxpayers were assigned their proper share deductions. Under these regulations, research and experimental costs were recognized as speculative and it was provided "that the gross income derived from successful research and development must bear the costs of unsuccessful research and development."
The requirement that the expenses of research and experimentation be applied against the income it generated quickly encountered the problem that the tax law of some nations, in which the income was generated, would not permit the deduction of such costs when incurred in another nation. The result was that two countries might each assume the other was giving the taxpayer a deduction when neither was. This displeased exporting taxpayers. It also displeased Congress, which feared the rules would encourage companies to relocate their research facilities abroad, so that the foreign countries in which the United States sourced their expenses would actually allow deductions for them.
As a consequence, the Economic Recovery Tax Act of 1981 (ERTA), Pub.L. No. 97-34, 1981 U.S.C.C.A.N. (95 Stat.) 172, 249 (codified at 26 U.S.C. Sec. 174 note (1982)) was enacted, which, in section 223(a) provided:
(a) 2-YEAR SUSPENSION--In the case of the taxpayer's first two taxable years beginning within two years after the date of the enactment of this Act, all research and experimental expenditures ... which are paid or incurred in such year for research activities conducted in the United States shall be allocated or apportioned to sources within the United States. (Emphasis added.)
Thus, at long last, we come to the issue before us in this part of the appeal, viz., does ERTA Sec. 223(a) bar the allocation of research and experimental expenses in computing the combined taxable income of a DISC? We hold that the answer is "No."
The principal reason for that answer is that the combined taxable income calculation is a step in a process that allocates income between the DISC and its supplier parent, both of whom are not foreign organizations. The function combined taxable income performs is to fix the amount of income from export sales that can be immunized from current taxation by the United States. ERTA responded to a different need, viz., the loss of a deduction from domestically taxable income for research and experimental costs allocated to foreign sources whose tax laws often allowed no deduction for these costs when not incurred within their jurisdiction. ERTA's purpose is unrelated to the combined taxable income computation. Its purpose is clearly revealed by the following legislative history:
Taxpayers that allocate research and development expenses for purposes of the foreign tax credit claim that the allocation results in more deductions being allocated overseas than is allowed as a deduction by the foreign country. Thus, taxpayers claim that their foreign tax credit limitation is lower than the foreign taxes paid and that they will lose foreign tax credits.
Taxpayers argue that because of the application of the regulation, they must transfer research and development activities to the foreign country in order to get a deduction in that country and, thus, get a full foreign tax credit on the income earned in that country. The committee believes that the transfer of research and development activities overseas would not be in the best interest of the United States. Therefore, the committee has concluded that the Treasury should study the impact of the allocation of research and development expenses under Treas.Reg. Sec. 1.861-8 on U.S.-based research and development activities.
H.R.Rep. No. 201, 97th Cong., 1st Sess. 131 (1981).
To accede to Intel's argument in this branch of the case would increase the DISC deferral of income and the deemed dividend of the DISC which, when foreign source income, would not be diminished by research and experimental costs. That was not the intent of ERTA. The Commissioner points out that this would entitle Intel to benefits in addition to its own ability to deduct research and experimental costs responsible for foreign income against United States source income. No such double benefit was intended by Congress, asserts the Commissioner. We agree.
The court in St. Jude Medical, Inc. v. Commissioner, 34 F.3d 1394, 1403 (8th Cir.1994), explained its rejection of the position Intel now takes in these words:
The CTI computation rules do not require use of geographic sourcing rules. The sourcing rules are relevant when St. Jude is apportioning its R & D expenditures between its U.S. source income and its foreign source income for foreign tax credit purposes. See, e.g., Treas.Reg. Sec. 1.861-8(f)(1)(i), (ii) (1994). For those purposes, St. Jude gets a distinct benefit from the ERTA Moratorium: it need not apportion any of its R & D expenditures to the dividend it is deemed to receive from its DISC. All DISC dividends resulting from qualified export receipts are foreign source income. See I.R.C. Sec. 861(a)(2)(D) (1995).
"Combined Taxable Income" is merely a computation required by U.S. tax laws to control the allocation between Intel and its domestic DISC subsidiary of its export sales income in a manner that resembles the allocation that would prevail were Intel doing business through a foreign subsidiary. It is an artificial construct relevant only to the fashioning of a tax-proper relationship between a domestic corporate parent and its subsidiary which, for tax purposes, masquerades as a foreign subsidiary.
To repeat, foreign source income, including that from "deemed dividends" by DISC bears no part of the research and experimental costs when "received" by the parent corporation. Intel, the parent corporation in this case, is required by ERTA to allocate all such costs "to sources within the United States." The Commissioner demands no more.
AFFIRMED.