Edward D. Kleinbard is a partner with Cleary Gottlieb Steen & Hamilton LLP. The author thanks his colleagues Lilian Faulhaber and Richard Sultman for their assistance in preparing this article for publication.
This report reviews the case for replacing the Internal Revenue Code's complex rules for taxing foreign direct investment with a territorial tax system. The report acknowledges that a territorial system offers one unambiguous advantage over current law, which is that it removes U.S. tax frictions on repatriating foreign profits. The report argues, however, that a territorial tax system would vastly exacerbate cross-border transfer pricing problems by rewarding successful transfer pricing gamers as "instant winners" of the tax lottery. In light of the overwhelming evidence of pervasive transfer pricing problems today, Kleinbard argues that this alone is sufficient reason not to move to a territorial tax system. Kleinbard also argues that other purported advantages of territorial systems, including simplicity and a more competitive tax environment for U.S. multinationals, are overstated.
Kleinbard believes a "full-inclusion" tax system also would eliminate the tax frictions on repatriating foreign earnings, and would genuinely be simpler than current law (in contrast to a territorial tax system). Importantly, he further argues, U.S.-based multinationals would have little reason to pursue aggressive transfer pricing tax strategies in a full-inclusion environment (again in contrast to a territorial tax system). Without more, however, a full- inclusion solution would be profoundly anti-competitive. Kleinbard shows how his business enterprise income tax proposal (first discussed in Tax Notes, Jan. 3, 2005, p. 97) addresses the competitiveness problems of a full-inclusion system, in large measure by enabling the tax rate imposed on U.S. firms to be substantially reduced and the foreign tax credit rules to be simplified.
Copyright 2007 Edward D. Kleinbard. All rights reserved.
Table of Contents
II. The BEIT in a Nutshell
B. Specific Rules
III. Why U.S. Should Reject Territorial Tax Solutions
A. Practical Implementations of Territoriality
B. The Critical Importance of Transfer Pricing
C. Territoriality and Transfer Pricing
D. Competitiveness and Economic Neutrality
E. Consequences of Territorial Systems
IV. 'Full-Inclusion' but Pro-Competitive
A. Transfer Pricing and Repatriation Neutrality
B. Application of BEIT to Outbound Investment
Territorial income tax systems are designed to exempt the "active" income of a U.S. firm's foreign branches or foreign subsidiaries from U.S. income tax when that income is repatriated to the United States. Territorial tax proposals are the current darling of many international tax reform recommendations, including those made in late 2005 by the President's Advisory Panel on Federal Tax Reform.1
This report advances three related arguments regarding the taxation of foreign direct investments by U.S. firms. The first is that territorial income tax proposals are a terrible idea. Once the actual implementing rules of a realistic territorial tax system are understood, territoriality cannot be recommended, even on the grounds advanced by that idea's proponents.
The report's second argument is that a properly constructed implementation of a "full-inclusion" income tax system for outbound investments (that is, an income tax that imposes current U.S. residual tax on income earned by a U.S. firm's foreign branches or subsidiaries, regardless of whether that income is "active" or "passive") can be pro-competitive, economically neutral in application, and infinitely more administrable than a territorial tax system. A properly constructed income tax satisfies four conditions: Its statutory tax rates are close to the low end of the range of rates employed by the major trading partners of the United States; it permits firms to claim foreign tax credits to the greatest extent compatible with protecting the fisc from the erosion of the domestic tax base; it permits the deductibility of foreign losses with no more restrictions than are imposed on the use of domestic losses; and it does not prefer outbound portfolio investment to outbound direct investment (or vice versa) by effectively imposing (for example) a "deferral regime" on one and a "full-inclusion system" on the other.
The third argument advanced by this report is that just such a properly constructed full-inclusion tax system has already been proposed. It is the business enterprise income tax, or BEIT -- a comprehensive business income tax reform proposal that I first presented in an article published in January 2005 and have continuously refined since then.2
To this author's infinite dismay, many readers of this report will not yet be familiar with the BEIT. Part II therefore takes matters out of order by quickly summarizing the BEIT's basic features. Part III then returns to the logical flow of the presentation by describing why the United States should not adopt a territorial tax system. Finally, Part IV demonstrates how the BEIT (in particular), or a more modestly overhauled version of the current income tax (as a second best), advances competitiveness, economic neutrality, and sound tax administration regarding foreign direct investment.
Two other introductory matters need to be addressed. First, for the avoidance of doubt, the arguments that follow are not a disguised attack on big business, multinational enterprises, or the pursuit of money. To the contrary, the BEIT is intended to advance the competitiveness of American businesses and the economic neutrality of the tax system, thereby eliminating many distortions that the current income tax system introduces into commercial and financial decisionmaking.3
Second, this report advocates a full-inclusion tax system for foreign direct investment by all U.S. firms, as part of the larger overall restructuring of the U.S. system for taxing business enterprises and business capital outlined below. In the absence of that sensible development, all active income of U.S. firms should be treated consistently, which is to say that income should be eligible for deferral. In particular, there is no justification for singling out the active international income of U.S.-based financial service firms for a more punitive tax regime than that enjoyed by the rest of the U.S. economy. Accordingly, while it is to be hoped that the BEIT becomes law, if fundamental business tax reform were not to occur, the case for making the "active financing exception" of section 954(h) and 954(c)(2)(ii) permanent would be persuasive, both as a matter of fairness and to prevent distortions in cross-industry investment over time.
The business enterprise income tax's individual proposals comprise an integrated package of reforms that rely on traditional income tax concepts but produce a more efficient and neutral system for taxing the returns to capital invested in private businesses. This part summarizes the BEIT's principal operating rules. The papers cited in note 2 describe the reasoning behind the rules and compare the BEIT with other income tax reform packages, particularly Treasury's 1992 "comprehensive business income tax" (CBIT) proposal.
The BEIT superficially resembles the current corporate income tax, but the underlying architecture has been completely overhauled. The result is a tax system that is economically neutral (returns to capital are burdened consistently) and that has much lower corporate (now "business enterprise") tax rates than current law's 35 percent corporate rate. The working hypothesis is that the new business enterprise tax rate can be in the range of 25 percent to 28 percent and that the system can remain revenue neutral compared with current law.
The BEIT abandons current law's multiple and frequently elective tax regimes (each turning on largely formal differences from the others) with a single set of tax rules for each stage of a business enterprise's life cycle:
1. Choosing the form of business enterprise.
2. Capitalizing the enterprise.
3. Selling or acquiring business assets or business enterprises.
As a result, under the BEIT, every form of business enterprise - - sole proprietorship, partnership, or corporation -- is taxed identically and every investor in a business enterprise is taxed identically on his investments, regardless of the label placed on an instrument as debt, or equity, or anything else. The BEIT thus moves the income taxation of business enterprises closer to the ideal of a featureless tax topography -- an environment in which there are as few special tax rules, exceptions to those rules, and antiavoidance glosses on the exceptions to the rules as is practical.
The centerpiece of the BEIT is a comprehensive and coordinated system for taxing time value of money returns, through the BEIT's cost of capital allowance (COCA) system. Under the COCA regime, a business enterprise deducts a time value of money return on all capital invested in its business (whether denominated as debt or equity) and investors include in income every year a time value of money return on their investments in financial capital (regardless of cash receipts). Investor-level calculations are based on an investor's cost basis in an instrument and thus do not require mark-to-market valuations or other financial information beyond simple arithmetic. The COCA system relies on the BEIT's other operating rules as a platform from which to apply the COCA calculations.
The fundamental theme of the COCA system (in conjunction with the BEIT's other rules) is to tax "economic rents" (the supersized returns attributable to unique commercial ideas or market positions) and risky returns entirely (or nearly so) at the business enterprise level and to tax time value of money returns once (and only once) at the investor level. The COCA system thus achieves both integration (that is, the elimination of double tax on corporate profits) and a consistent and accurate measure of income.
- Taxation of all business enterprises, regardless of form (for example, sole proprietorships, partnerships, or corporations), as separate taxable entities. Entrepreneurs thus are free to choose whatever form of business organization they wish, but that choice has no collateral tax consequences. The basic tax system looks much like today's corporate income tax, in that the entity tax roughly follows current rules for taxing corporations, subject to the major modifications described below. Also, investors are taxed under the new COCA system on their investment returns. The BEIT thus preserves a two-level tax system, which minimizes transition revaluations of financial assets. The two levels of tax, however, are for the first time coordinated and integrated.
- Substantially lower enterprise-level tax rates (working hypothesis: 25 percent to 28 percent) than the current corporate income tax rate (35 percent).
- Broadening of the business tax base by reforming some important but technical business tax accounting rules and industry-specific preferences (for example: last-in, first-out inventory accounting; like-kind exchanges; or percentage depletion). The largest base-broadening component, however, is the COCA system for taxing returns on investment, as described below.
- Adoption of the COCA system for taxing time value of money returns to investors and deducting the cost of capital by issuers. The basic theme of the COCA system (in conjunction with all the other rules described below) is to tax economic rents and risky returns at the business enterprise level and to tax time value of money returns on a current basis at the investor level. The critical difference between COCA and current law is that COCA taxes investors on a current basis on an expected time value of money return on all forms of financial capital invested in businesses, whether called debt or equity, without regard to cash receipts. That current income inclusion is determined by straightforward arithmetic, not observed market valuations for assets. The COCA system is described in a bit more detail a few paragraphs below.
- Mandatory "super tax consolidation" for affiliated enterprises. (All subsidiaries are treated as part of the parent company, as in financial accounting, rather than the hodgepodge consolidated return tax rules we have today.) Consolidation in general would be measured at the 50 percent level and would be measured by reference to all of a company's long-term financial instruments (with tiebreaker rules to prevent multiple consolidations). The rule both eliminates substantial complexity and serves as a foundation for the COCA system.
- As described in more detail in Part IV, the extension of the "super-consolidation" rules to international income. As a result, the BEIT eliminates the "deferral" of active foreign income from current U.S. tax.5 (The BEIT in this respect is the perfect mirror image of a territorial system.) At the same time, the BEIT contemplates (1) eliminating the allocation of U.S. interest expense (now COCA deductions) against foreign income -- the principal source of "excess FTC" problems for U.S. multinationals -- and (2) lowering the tax rate on global income. Finally, global super-consolidation also means that foreign losses will become currently deductible in the United States, thereby restoring neutrality to the U.S. tax analysis of foreign direct investments.
- Repeal of all tax-free organization/reorganization rules, and their replacement with a much simpler "tax-neutral" acquisition system in which all acquisitions of business assets or business enterprises --- basically, all incorporation transactions, or all entries to or exits from a super-consolidated group --- are treated as taxable asset acquisitions. The seller's tax rate, however, differs across the different asset classes that it transfers, depending on the present value to a taxpaying buyer of the step-up in the tax basis of the various assets acquired. The result, from the point of view of the tax system as a whole, is close to that of entirely tax-free transfers (at least at the business enterprise level), but with important technical and administrative advantages.
The COCA system is the centerpiece of the BEIT's ability to measure and tax returns to capital, but the COCA cannot be implemented in a logical fashion without the other reforms summarized above. Nonetheless, because of the COCA's central role, it is useful to outline how it would be implemented.
An investor's COCA income calculation for a year is simply the relevant rate of return for the year (as published by the IRS) multiplied by the taxpayer's tax basis in his financial investments.6 That amount -- termed the minimum inclusion -- is includable in income regardless of whether it is paid currently by the issuer to the investor. Cash received from the issuer is tax-free to the extent of current or prior minimum inclusion accruals. The COCA rate will be published regularly by the IRS (just as the applicable federal rate is today) and will be set by reference to the one-year Treasury note rate (for example, one-year Treasuries plus 1 percent).
In the COCA environment, issuers deduct each year, in lieu of current law's interest deductions, a uniform cost of capital allowance equal to the same COCA rate multiplied by the aggregate tax basis of their assets. Thus, an equity-funded issuer obtains exactly the same COCA deduction as does a debt-funded issuer, regardless of the coupons paid on its financial capital.
Under the COCA system, losses from sales of financial assets are currently deductible against ordinary income (to the extent of prior time value of money inclusions on those assets). The result is a more economically neutral investment environment than that provided by current law's capital loss limitation rules.
As currently contemplated, the COCA system would impose a small (10 percent to 15 percent) additional tax on an investor's gains beyond time value of money returns. That incremental tax is not compelled by the logic of the system, but rather is suggested in response to traditional fairness and ability-to-pay concerns.
Depreciation methods are unaffected by the COCA system, but the interaction of the COCA rules and depreciation at the business enterprise level has the effect of neutralizing the present value to the government of a firm's tax obligations regarding the capitalization/depreciation methods that it might employ: Faster depreciation means less remaining tax basis in business assets and smaller COCA deductions for the future.
While the COCA system does require some record keeping and arithmetic, it is feasible, in ways that "accruals" (universal mark-to-market) taxation and other ideal systems are not. The COCA's allocation of the incidence of tax between investors and issuers is technically superior to Treasury's 1992 CBIT proposal to tax all time value of money returns solely at the business enterprise level.7
The COCA system is intended to coexist with broad savings incentives similar to current law and the President's Advisory Panel on Federal Tax Reform's proposals. As a result, the COCA system adds progressivity to the tax code because its burden falls on only the wealthiest taxpayers (as the only taxpayers with significant financial investments not sheltered by tax-deferred savings plans).
A. Practical Implementations of Territoriality
Territorial tax systems seek to exempt from U.S. income tax the active foreign income of branches or subsidiaries of U.S. firms when that income is repatriated to the United States. Three principal reasons usually are advanced for preferring a territorial tax system as the basis for taxing the international income of U.S.-based multinationals. First, by exempting foreign income from any incremental U.S. taxation, territorial solutions are said to improve the international competitiveness of U.S. firms.8 Second, territorial systems are said to promote goals of economic neutrality, in particular by eliminating current law's bias in favor of keeping low-taxed foreign income offshore, rather than repatriating it, simply to avoid incremental U.S. repatriation tax costs.9 Third, territorial tax solutions are thought to be simpler than current law because, in particular, they do away with the FTC in respect of active foreign income.10
A practical territorial tax system requires several design elements that critically affect the validity of those claims. First, there appears to be a consensus among tax theorists that a territorial solution in practice would apply only to active foreign income; as the price for exemption from U.S. tax, that active income of course would not bring with it an FTC for any non-U.S. taxes that burdened that income.11 Active foreign losses would not offset domestic taxable income; that is, in effect, the mirror image of domestic exemption for active foreign income.12 Interest, royalties, or other deductible flows paid by a foreign affiliate to its U.S. parent would be fully taxable in the United States because that income would not have been subject to foreign tax. Further, current law's subpart F regime generally would be retained for passive/mobile income.13 The FTC system in turn would apply as it does today for any such nonexempt income.
One important implementation issue that is not explicitly discussed in most of the literature is what the treatment should be for "stripping" payments (deductible interest or royalties, for example) paid by one foreign affiliate of a U.S. firm to another foreign affiliate.14 Thus, if a German subsidiary pays royalties to an Irish sister company for the use of intangible assets owned by the Irish company, and those payments reduce German (high- tax) active foreign income, should the receipt of that deductible flow in (low-tax) Ireland be treated as active income or instead as passive/mobile income that is ineligible for the territorial regime? Until the adoption of section 954(c)(6) a few months ago, the answer under current law would have been that the Irish affiliate's income was subpart F income.15 Today, section 954(c)(6) would treat the income as retaining the active income character that it had in the hands of the German payor -- at least for the three years that section 954(c)(6) is scheduled to apply. So one could say that recent tax policy points in every possible direction on this critically important question that goes to the heart of the utility and fairness of a territorial tax system.
Finally, most territorial tax systems that have been seriously studied in the United States to date have included a provision to allocate interest expense incurred in the United States, and in some cases other classes of domestic expenses, against foreign "exempt" income (which, of course, is not necessarily exempt in a global sense and which may in fact have borne foreign tax at rates as high as or higher than the U.S. rate).16 Most commentators agree that some sort of sensible interest expense allocation rule, or some comparable provision (for example, an efficacious "thin capitalization rule" that would prevent the overleveraging of U.S. operations), unquestionably is required in the context of a territorial foreign tax system to protect the domestic tax base. In the absence of such a rule, analysts fear that U.S. firms would overleverage their U.S. operations to the point where they "zeroed out" their U.S. tax liability on their domestic operations and would service that debt with tax-exempt (from the perspective of the United States) foreign-source income.17
That last concern demonstrates in turn the critical importance of the treatment of interaffiliate stripping transactions, as described above. If one is confident that foreign income will bear a tax burden comparable to that of the United States, the case for domestic interest expense allocation rules becomes more attenuated. Conversely, if one believes that foreign-to-foreign income stripping to reduce foreign tax burdens is appropriate, the need to protect the domestic tax base becomes more urgent. (Of course, if one believes that foreign tax rates are highly likely to be comparable to those of the United States, one can fairly question the need to adopt a territorial tax system at all, as doing so would not reduce tax burdens or in practice significantly change repatriation policies.)
The remainder of this part demonstrates that practical implementations of territorial tax systems are anything but simple. For example, as described above, territorial tax systems in practice inescapably require two parallel tax regimes, one comprising current law (for passive/mobile income) and the other the territorial scheme. With those two parallel regimes come difficult coordination and line- drawing issues. Similarly, territorial tax systems are usually scored as revenue generators once ancillary expense allocation or comparable rules are considered. And while it is true that a territorial tax system removes current law's distorting effects on firm repatriation policies, the irony is that so too does a full-inclusion system. At the same time, territorial tax schemes introduce important new distortions, of which the most important by far is the pressure those schemes put on our transfer pricing systems. The next sections therefore turn to transfer pricing and its relationship to the choice of a foreign direct investment tax regime.
B. The Critical Importance of Transfer Pricing
Transfer pricing issues (that is, efforts by firms, whether U.S. or foreign-based, to reduce their U.S. tax liabilities by shifting U.S. profits to low-taxed non-U.S. affiliates) are the most important challenge today to the administration of the international tax provisions of the code. That observation is consistent as an anecdotal matter with the issues that many practitioners see in their practices.18 More usefully, that observation also is consistent with objective data.
The IRS now confronts transfer pricing cases involving staggering sums of money. For example, the IRS recently announced the settlement of a tax case against Glaxo- SmithKline in which the pharmaceutical company agreed to pay the IRS $3.4 billion (including interest) for tax deficiencies relating to a 12-year period (and concurrently agreed to abandon a $1.8 billion tax refund claim), all as a result of its transfer pricing practices.19 Similarly, Merck & Co. recently revealed that it is contesting similar transfer pricing (and other) cases, in which the tax claims against it by the IRS and the Canadian tax administration total some $5.6 billion.
In a recent and sophisticated paper, Dr. Harry Grubert of the Treasury Department and Prof. Rosanne Altshuler of Rutgers University (and formerly on the staff of the President's Advisory Panel on Federal Tax Reform) considered in detail the role of intangibles in cross-border transfer pricing.20 Paraphrasing the work of this academic study (hopefully without excessive violence to the authors' intent), Grubert and Altshuler concluded that:
- The exportation of intangible assets has been a "significant source" of foreign direct investment income; royalties and license fee income received by U.S. companies tripled from 1990 to 2004.21
- At the same time, royalties paid by foreign subsidiaries to U.S. parent companies "represent less than half of the contribution that parent R&D makes to subsidiary income."22
- The data suggest that low-tax countries "are becoming much more important destinations for U.S.-produced intangible assets"; in this connection, "the share of total affiliate royalties accounted for by Ireland and Singapore doubled between 1994 and 1999."23
- "Pre-tax profits in relation to sales are almost three times higher in Ireland on average than the group mean. These 'excess' profits presumably reflect the fact that very valuable intellectual property is located in Ireland and the royalties paid back to the United States, while significant, do not fully reflect its contribution."24
A recent economic analysis by Martin Sullivan reaches similar conclusions.25 Sullivan concludes, for example, that while foreign affiliates of U.S. multinationals in the aggregate earned a 7.2 percent return on sales in 2004, Irish subsidiaries had more than twice that profitability -- 14.8 percent. By contrast, the unweighted average of the returns on sales realized by subsidiaries in Europe's larger economies was much lower than the all-countries aggregate figure -- roughly 4.2 percent.26
An important Wall Street Journal article from November 2005 gives life to those dry statistics by describing in detail Microsoft's use of "cost-sharing agreements" with an Irish subsidiary to develop and exploit Microsoft's core intellectual property.27 According to that article, Round Island One, Microsoft's intellectual property holding company in Ireland, earned nearly $9 billion in gross profits in 2004, and roughly $2.4 billion in taxable income, by exploiting intangible assets to which it acquired ownership by virtue of its cost-sharing agreements with its U.S. parent.28
To be clear, I do not mean to suggest that Microsoft's arrangements with its Irish subsidiary violate the requirements of the extensive arm's-length transfer pricing regulations governing cost-sharing agreements. That is the purpose of the IRS examination process, to which I am a complete outsider. I do think it fair, however, to point to The Wall Street Journal article and the academic paper discussed above to illustrate the magnitude of the intangible property transfer pricing issue and its importance to tax administration.
I also believe it fair to draw from all of the above the inference that the IRS is shouldering a near-impossible burden in that area, for two reasons. First, the accurate valuation by outsiders of intangible assets like Microsoft's proprietary "crown jewel" software is nearly impossible, because the assets themselves are incredibly complex and because in practice genuinely comparable third-party transactions almost never exist. (That is, major software companies rarely enter into cost-sharing agreements with third parties to develop new versions of their crown jewel intangible assets.) Yet the arm's-length transfer pricing cost- sharing regulations require just such an inquiry, to measure "buy-in" payments for the existing intangible assets that form the basis for a cost-sharing agreement.29
Second, the entire premise of our transfer pricing rules -- that related parties should deal with each other for tax purposes at the prices and on the terms at which third parties would engage in comparable transactions -- is unachievable, particularly when applied to high-value intangible assets held by multinational enterprises. There is abundant literature to support the proposition that multinational enterprises thrive in the world economy precisely because the economy is increasingly global and because multinational enterprises can muster tightly integrated global resources to take advantage of that fact.30 The paradigmatic example of the integrated global strategies of modern multinational enterprises, of course, is the worldwide exploitation of a common pool of high- value intangible assets.
Arm's-length transfer pricing tends to deny (or perhaps misallocate) the synergies that flow directly from the globally integrated activities that explain the success of multinational enterprises in the first place. As applied to intangible assets, arm's-length transfer pricing requires us to pretend that a multinational group does not in practice control a single common pool of intangible assets with worldwide application, but rather comprises essentially independent enterprises negotiating with each other as if trade barriers to the direct global exploitation of those intangible assets still existed.31
As a result, the arm's-length transfer pricing principle at its core presupposes a business model that is fundamentally inconsistent with the business strategies of multinational enterprises that possess high-value and globally relevant intangible assets. When the tax model that we have created is so fundamentally agonistic to business realities, the administration of the tax system can never be wholly successful.
C. Territoriality and Transfer Pricing
Changing to a territorial tax system would greatly exacerbate the importance of transfer pricing issues. The reason is simple. Under current law, the principal "reward" for successfully gaming our transfer pricing rules is the accumulation of profits in a foreign subsidiary, presumably located in a low-tax jurisdiction.32 To collect that reward, however, a U.S. firm must keep those earnings offshore indefinitely. Territorial tax systems, by contrast, reward successful transfer pricing gamers as "instant winners" by enabling the successful U.S. firm to recycle immediately its offshore profits as tax-exempt dividends paid to the U.S. parent.33
That concern is widely shared, and has been identified as a topic of concern by the staff of the Joint Committee on Taxation and other authors who have described or proposed possible territorial tax systems.34 The principal difference between my views and the views of these other observers is that they typically conclude that the administration of our existing arm's-length transfer pricing rules simply will require greater vigilance in a territorial tax system.35 By contrast, I believe that it is unrealistic to expect that enhanced administration can ever adequately address the transfer pricing challenge that modern tightly integrated multinational enterprises possessing high-value intangible assets would pose to a territorial tax system.
D. Competitiveness and Economic Neutrality
In recent years, many observers have described how the rapid evolution of the global economy has compelled U.S. tax policymakers to become increasingly sensitive to issues of international competitiveness. For example, Glenn Hubbard, the dean of the Columbia Business School and former chair of the president's Council of Economic Advisers, recently testified before the House Ways and Means Committee on precisely that topic. Hubbard identified several important themes relating to the changing competitive landscape in his testimony, including the increasingly integrated nature of the global economy, the enormous rise in international capital flows (which include cross-border portfolio investments), and the shift over the last several decades from the United States' role as the world's largest exporter of capital to its current status as the world's largest capital importer.36
Hubbard rightly draws from these facts the conclusion that U.S. international tax policy norms from, say, 1962, do not necessarily serve the interests of the United States in 2006. The same underlying questions remain relevant, however: What principles should we in fact adopt as our international tax policy norms in the new world economy? And how can we measure different tax proposals against those norms?
It is the traditional practice in discussions of international tax policy choices to begin to address those questions by laying out the principle of "capital export neutrality" -- that a U.S. multinational firm should face the same tax burden on a new investment wherever in the world that investment might be made -- and the principle of "capital import neutrality" -- that a U.S. multinational firm should bear the same tax when competing in a foreign market as its local competitors face.37 To those can be added at least two other widely discussed "neutralities" - - "national neutrality" and "capital ownership neutrality."38
The traditional discussion then goes on to demonstrate that it is not fully possible to satisfy both capital export neutrality and capital import neutrality simultaneously in the real world.39 At the same time, most analysts acknowledge that, all other things being equal, maintaining capital export neutrality would be desirable, and, by the same token, so would maintaining capital import neutrality. Finally, every traditional discussion concludes by asserting that whatever policy is being proposed represents a fair balancing between those two irreconcilable objectives, in every case based largely on the author's preexisting intentions. No wonder our international tax policy is muddled!
In a refreshing break from that familiar presentation, Grubert and Altshuler implicitly conclude that the traditional "Battle of the Neutralities" (as I term the process) is an essentially sterile exercise that by itself cannot usefully guide tax policymakers in shaping the international tax policy norms of the United States.40 Instead, they urge policymakers to focus on the behavioral distortions among taxpayers (and, to a lesser extent, governments) that flow from current law and to evaluate reform proposals by reference to their success in mitigating the distortions:
What reform within an income tax can hope to accomplish is to eliminate unnecessary waste and the possibility of extremely high or low tax burdens that are not justified under any standard. Then we can at least be sure that we are moving toward the optimum without overshooting it and running the risk of making things worse.
International tax systems can act on many behavioral margins in addition to the choice of location. The current tax system induces a number of behavioral responses that both waste resources and lead to inappropriate incentives to invest tangible and intangible capital in various locations. These include strategies to avoid the U.S. repatriation tax on dividends, to shift debt from high-tax to low-tax locations, and to shift income to low-tax locations by distorting transfer prices or paying inadequate royalties. Besides directly wasting resources, these strategies can lead to inefficient choices between related party and arms' length transaction and a distribution of tangible and intangible assets that cannot be justified on any conceptual basis.
In our evaluation of the distortions that may be eliminated by some of the reform proposals, we focus on how the proposals affect (1) the location of tangible capital, (2) the location of intangible capital, (3) the repatriation decision, (4) financing decisions, (5) income shifting, (6) incentives to lower foreign tax burdens, (7) export decisions and (8) host government decisions regarding the taxation of U.S. companies.41
I submit that reviewing the effect of current law or any tax reform proposal on the eight criteria listed immediately above is a far more productive exercise than continuing the sterile "Battle of the Neutralities" that has dominated much of the policy discussion to date.
It also unfortunately follows from the above that it is absolutely necessary in evaluating any international tax reform proposal to wade into the technical details of how that proposal will be implemented. That is, it turns out that an international tax reform proposal must be specified and analyzed in detail, if one is to predict with any degree of accuracy how the behaviors of differently situated taxpayers will be affected by the proposal, and, therefore, what distortions in economic activity might follow.42
E. Consequences of Territorial Systems
This section considers the economic and competitiveness consequences of adopting a practical territorial tax system for taxing foreign direct investment. It turns out that when one applies the metrics proposed in the previous section to realistic implementations of territorial systems, the analysis becomes surprisingly complex and the answers not at all intuitive.
A territorial tax system unquestionably would reduce distortions inherent in the current code in one important respect, which is that it would eliminate the barriers to repatriation that current law imposes. As observed earlier, a U.S. firm today must "earn" the tax benefit of deferral through patiently deploying its active foreign profits outside the United States, even if the highest and best use of those funds would be in a domestic application.43 As a result, current law encourages the wasteful accumulation of profits abroad, and in some cases the wasteful investment of those profits in the expansion or acquisition of "active" businesses, solely to preserve the continuing benefits of deferral. A territorial tax system eliminates tax considerations from the repatriation decision and therefore removes this significant economic distortion of current law.
Many advocates of territorial tax systems also believe it to be self-evident that territoriality will enhance the competitiveness of U.S. firms by eliminating residual U.S. income tax. Those proponents view territorial tax systems as the paradigmatic implementation of capital import neutrality themes. The revenue implications of practical territorial tax systems, however, are more ambiguous than those advocates might expect.
In January 2005 the JCT staff proposed a comprehensive territorial tax system, described as a "dividend exemption system."44 The JCT staff estimated that its territorial system would raise $55 billion in tax revenue over 10 years. It is difficult to describe that proposal as self-evidently enhancing the competitiveness of U.S.-based multinational firms if by that phrase one means a reduction in total tax burden imposed on the income of U.S. multinationals.
Later in 2005, the President's Advisory Panel on Federal Tax Reform proposed a system similar in broad outline to the JCT staff proposal, although with some differences in detail (particularly regarding expense allocation rules).45 No official revenue estimate accompanied that proposal. Most recently, Grubert and Altshuler concluded that switching to a territorial system would generate a small revenue gain, but that the revenue estimate was critically sensitive to possible behavioral responses that are difficult to model.46 Their paper also summarizes earlier work that concluded that a territorial tax system would significantly increase the tax burden on investments in low-taxed foreign subsidiaries.47
There are two principal factors at work behind those surprisingly effective tax rate results. The first factor is the conclusion reached by the JCT staff and others that a territorial tax system must be accompanied by interest expense allocation rules modeled on current law, as described in Part III.A, with the result that interest expense allocated to tax-exempt income would not be deductible.
The second principal reason why a territorial tax system can raise effective tax rates in some cases is that it eliminates a taxpayer's ability under current law to average down high-taxed foreign income with zero-taxed foreign royalty income (or low-taxed affiliate income). I liken the process to a master distiller blending a perfect tax liqueur, in which the blended product bears tax at precisely 35 percent, so that no residual U.S. tax is due and no excess credits are generated.
More specifically, every territorial tax system that has been seriously studied in the United States would not exempt from tax royalty or interest income paid by a foreign subsidiary to its U.S. parent, on the theory that those amounts were deductible abroad and that exempting them from U.S. tax thus would result in those amounts bearing tax nowhere in the world. Under current law, a U.S. parent company's stream of royalty or interest receipts from its foreign subsidiaries nominally constitutes taxable income, but in fact the actual tax liability on those amounts is largely sheltered by the tax "master blender" at each company, who brings up sufficient high-taxed income from other foreign operations to shelter those income streams.
In a territorial system, by contrast, the royalty and interest income would be fully includable in income without offset for any tax credits attributable to exempt income. As a result, a firm's cask of exempt high-taxed income could not be blended with liqueur from a low-taxed cask in a way that would reduce the effective tax rate on the former.
It is for those sorts of reasons, I believe, that Stephen Shay, in his 2006 testimony before the Ways and Means Committee on the theme of international competitiveness, suggested that U.S. multinationals today actually enjoy the best of all worlds.48 In a similar vein, the National Foreign Trade Council in 2002 undertook a comprehensive review of territorial tax proposals on behalf of a wide range of U.S. multinational firms. That study concluded that the evidence did not unambiguously support the claim that a territorial tax system would enhance competitiveness:
While it is true that a territorial system could improve competitiveness and simplicity for some U.S.-based companies with substantial operations abroad, the accompanying reduction in foreign tax credits attributable to exempt income could more than offset that benefit for other such companies. Moreover, the benefit for any significant group of companies would be dependent on the adoption of a broad exemption, a cut back on the existing subpart F rules, and reform of the current expense allocation rules.49
It is ironic that some proponents of territoriality may be unaware that the current system often can be used to optimize a U.S. firm's global tax liabilities in ways that a territorial system cannot.50 Similarly, those proponents might not appreciate the complex and ambiguous effects of a well-designed territorial tax system (that is, one with proper expense allocations or other mechanisms to safeguard the domestic tax base) on a U.S. multinational firm's worldwide effective tax burden.
The previous paragraphs acknowledged that a territorial tax system would eliminate the behavioral distortions attendant on current law's repatriation tax burdens. The probable effect of a well-designed territorial tax system on effective tax rates, however, is not unambiguously pro-competitive, as that term ordinarily is employed. At the same time, a territorial tax system can exacerbate (or create novel) economic distortions, compared with those that exist under current law. Most importantly, a territorial tax system will encourage multinational firms to express increased enthusiasm for aggressive transfer pricing strategies (particularly relating to high-value intangibles), for the reasons described in Part II.C.51 Because that topic already has been addressed, the remainder of this section considers some other, less obvious, economic distortions that accompany practical territorial tax systems.
First, a territorial system can be expected to impose radically different tax burdens on the international income of different U.S. industries, largely as a result of different industry norms for debt- to-equity ratios,52 different levels of reliance on separately identifiable intangible assets (as opposed to goodwill and the like), and different rates of adopting tax-preferred methods of developing global intangibles. For example, if a territorial tax system includes an interest expense disallowance rule modeled on current law's FTC rules allocating domestic interest expense against foreign-source income, U.S. financial services firms (with their high debt-to-equity ratios) will be disadvantaged compared with other industries that are primarily equity funded. Similarly, territorial systems will reward those firms or industries that were early and aggressive adopters of cost-sharing agreements with their foreign subsidiaries, because they will be able to capture the returns of those non-U.S. intangibles as exempt income.
Second, an important potential source of economic distortion is that tax policy can distort investment by portfolio investors as well as direct investors. One example of that phenomenon is the tax-driven differences in the relative attractiveness for a U.S. investor of making a portfolio investment in a U.S. multinational firm (which in turn makes foreign direct investments), compared with making such investments in a foreign-domiciled multinational. The same issue can also arise from the perspective of a foreign portfolio investor considering the same two investments, or a U.S. multinational corporation considering a foreign portfolio as opposed to a foreign direct investment, or even a U.S. portfolio investor considering investing in U.S. multinational firms as opposed to U.S. domestically oriented businesses. In light of the enormous surge in global capital flows, the increased transparency and liquidity of many foreign capital markets, and the ease of global research through online tools, it is absolutely imperative that U.S. international tax policy consider any tax reform proposal's potential for distorting those portfolio investment decisions.53
As envisioned by the JCT staff, a territorial tax system would not directly distort portfolio investment decisions between U.S. and foreign portfolio investment opportunities, although of course the ultimate effective tax rates imposed on different firms or different industries in a particular implementation of territoriality might do so. Territoriality would, however, distort the decision to make a portfolio investment rather than a direct investment, because the former (at least in many proposed implementations) would be subject to full double taxation, while a direct investment would not.54
The territorial proposal made by the President's Advisory Panel on Federal Tax Reform would introduce still another particularly dramatic economic distortion for portfolio investors, because of the peculiar way in which the panel chose to combine its territorial tax proposal with domestic relief from the double taxation of dividends. Essentially, when viewed from the perspective of the ultimate owners of a business enterprise, the panel's proposal would have dramatically preferred portfolio investment in domestically oriented U.S. firms over portfolio investment in U.S.-based multinational enterprises that bore precisely the same effective global tax rate.
More specifically, the panel's "simplified income tax" (SIT) proposal, apparently following the (erroneous) logic of Treasury's 1992 CBIT proposal, would have imposed a sort of compensatory tax when a U.S. company paid dividends to its U.S. shareholders out of exempt foreign earnings.55 The result would have been a significantly anticompetitive step backwards for U.S. multinationals in respect of their cost of equity capital.56 In that respect, then, the panel's SIT proposal would have introduced a distortive double tax on foreign income.
For example, imagine two U.S. corporations, Domestico and Globalco. Domestico earns $100 pretax, entirely from U.S. operations; Globalco also earns $100 pretax, but entirely from operations in Freedonia. Both companies are entirely equity funded.
Under the panel's SIT, Domestico pays $31.50 in tax on its $100 income. Domestico then can distribute the remaining $69.50 to its shareholders as an exempt dividend.
Globalco, by serendipity, also pays $31.50 in income tax on its $100 income, but Globalco makes out the check for its tax payment to the Freedonia IRS. Globalco can repatriate its $69.50 of after- Freedonian-tax profits to the United States, but when it distributes that amount to its U.S. shareholders they will be subject to full ordinary income tax on the distribution, while their brethren who invested in Domestico keep the same $69.50 distribution free of any tax.
A third potential new distortion again relates to the role of income stripping transactions, in their broadest sense. At least some proponents of a territorial tax system use "competitiveness" as a code word for "the lowest possible tax on foreign income that can legally be devised." One can fairly ask, however, whether competitiveness in that sense is truly nondistortive or whether instead a less distorting goal might be to design a tax system that would enable a U.S. firm to compete against local firms in their domestic markets at an effective tax burden that is directly comparable to that faced by those local firms.
Those two thoughts are not identical. We all understand the importance of "check the box" disregarded entities, hybrid instruments, and hybrid entities in U.S. international tax planning today. The difficult question that deserves more debate is whether, if a U.S. firm can employ those arrangements to drive its effective tax rate on its Freedonian operations below the rate imposed in law and in practice by Freedonia on its domestic companies, we should applaud that result as enhancing competitiveness or instead decry the result as distorting investment decisions.
That point can be generalized by observing that territorial tax systems in practice inevitably bring with them the prospect of "stateless income" -- income that is taxed nowhere in the world (or, at least, taxed at extremely low rates in a country where the income is not earned). Stateless income is not simply an artifact of transfer pricing abuses, but also arises from decisions as to where to place financial capital within a multinational group (so as to generate interest expense in a high-tax country and offsetting income in a very-low-tax jurisdiction), differences in implementation of different tax systems, hybrid instruments, and hybrid entities. All territorial tax systems struggle with the issue of stateless income.57
For example, if a territorial system permits a deductible payment paid by one foreign affiliate out of its exempt income to retain its exempt character when paid to another foreign affiliate, that system will encourage -- indeed, impel -- taxpayers to use affiliate interest, rents, and royalties to strip out earnings from the countries in which that income economically is earned. That leads directly to the phenomenon of stateless income. Conversely, treating all such income as "passive" (and therefore as immediately taxable in the United States) will be criticized as undercutting the purpose of a territorial system. The conflict inevitably will lead both to difficult technical issues (for example, layering rules for determining from which income a deductible interaffiliate expense is paid) and to a political tug of war identical to that which has bedeviled subpart F of the code, as reflected in its various "same country" exceptions, the recent adoption of the temporary provisions of section 954(c)(6), and the even more recent passage by the House of Representatives of a bill to scale back some of the provisions of section 954(c)(6).58
The problem of stateless income is not an abstract academic concern. Recent European Court of Justice jurisprudence, for example, suggests that it is becoming difficult for one European Union member state to tax (through subpart F analogies or the like) the income of a subsidiary in another member state, or to protect its tax base from widespread income stripping within the EU (by imposing withholding tax on outbound deductible payments or imposing thin capitalization rules on foreign investments only).59 The effect of those developments, if combined with a U.S. territorial tax system that treats interaffiliate deductible payments made out of exempt income as retaining its exempt character, would be to ensure that a large fraction of the income earned by many U.S. multinational groups in the EU would be taxed at no greater rate than that imposed by whichever member state had the lowest rates.60
Finally, territorial tax systems are distortive in one unassailable respect, which is that they would bring with them substantial deadweight losses in the form of compliance and similar costs. A territorial tax system is simpler than current law only in the imaginations of those who have never immersed themselves in the detailed implementation of either.
More specifically, as described in Part III.A, every territorial tax system that has been seriously proposed in the United States would retain a subpart F construct for passive and mobile income.61 That subpart F income in turn would be entitled to FTCs, so that all the complexities of current law would be replicated, except that the new system would stimulate new taxpayer impulses, which in turn would require new antiabuse rules.62 In particular, because FTCs would be useless when attributable to exempt active income, but would remain valuable if allocated against subpart F income, elaborate policing mechanisms (which admittedly exist in more rudimentary form today) would be required to ensure that advanced tax planning tactics could not be used to cause tax credits to migrate (from a U.S. perspective) from active (exempt) income to subpart F income.
Today, subpart F income means the unavailability of deferral; tomorrow, categorizing revenue as subpart F income would mean that the revenue would move from wholly exempt to immediately taxable status. The result would be even greater stress on the divide between active (exempt) income and subpart F income than exists under current law.63 Similarly, the U.S. law on the "source" of income (and many losses or expenses) is relatively undeveloped, compared with other areas of the code. Those concepts would become critical, however, in defining and policing the scope of a territorial tax system.
A. Transfer Pricing and Repatriation Neutrality
In direct contrast to current law, or to a territorial tax system, a "full-inclusion" U.S. international tax system would greatly attenuate the role of transfer pricing strategies by U.S. multinationals as an affirmative taxpayer device to minimize global tax liability, because all income earned by a U.S. multinational group would be taxed by the United States on a current basis.64 As a result, any remaining transfer pricing issues for U.S. multinationals would relate primarily to conflicting positions that might be taken by different taxing jurisdictions. A U.S. multinational corporation ordinarily would be a disinterested bystander to any such disputes, except in the limited case in which the foreign jurisdiction's tax rates greatly exceed those of the United States.65
In practice, a full-inclusion U.S. international tax system will not eliminate transfer pricing cases involving U.S. multinationals, but it will elevate (or at least relocate) those cases to direct negotiations between affected tax administrations, rather then serial negotiations between a taxpayer and those tax administrations. It is my hypothesis that, with little or no money of its own at risk, a U.S.-based multinational will be both less ingenious in its internal transfer pricing strategies and more forthcoming in dealing with the IRS. By elevating the debate to one between tax administrations, a full-inclusion system also will increase the likelihood that all affected tax administrations will work from a common understanding of the facts and that 100 percent of the taxpayer's income -- neither more nor less -- will be accounted for.
Because a full-inclusion system would materially dampen current law's incentives for multinational corporations to embrace transfer pricing strategies with excessive enthusiasm, such a system would remove significant tax-induced distortions in corporate behavior attributable to transfer pricing gamesmanship. The data marshaled by Grubert and Altshuler and in other academic papers are just too powerful to ignore: It cannot simply be the luck of the Irish, for example, that explains the extraordinary and systematic profitability of Irish subsidiaries of U.S. firms. A full-inclusion tax model is the only approach that directly addresses this critical problem.
Of course, a full-inclusion U.S. tax system does not eliminate the incentives of foreign-owned multinationals to engage in U.S. tax transfer pricing planning, as the recent example of GlaxoSmithKline, described in Part III, illustrates. But, by enabling the IRS to concentrate nearly all of its transfer pricing resources on inbound investments by foreign multinationals, the full-inclusion system would indirectly improve compliance in that direction as well.
"Repealing deferral" also would enhance competitiveness directly in the same important respect that adopting a territorial tax system would, which is that without deferral, U.S. firms' repatriation policies would reflect the highest and best use of their cash surpluses, rather than tax rate arbitrage. Ironically, the most unambiguous economic argument for adopting a territorial tax system - - the elimination of tax considerations in firms' decisions whether to repatriate offshore profits -- is a feature that territoriality shares with its mirror image, a full-inclusion system.
A full-inclusion system also would eliminate the distortions attendant on policing the boundaries of a territorial tax policy. As described in Part III, the serious proposals for territorial tax systems for the United States suspend the availability of the FTC for exempt (active) income, but preserve the FTC, and all its attendant limitations, exceptions, and qualifications, for subpart F (passive) income. That requires drawing clear lines between the two categories of income, as well as even more elaborate mechanisms than exist under current law to ensure that uncreditable foreign taxes associated with active (exempt) income do not, through advanced tax planning, migrate over to a taxpayer's subpart F income (where those taxes would become valuable as credits). By dispensing with the sharp demarcation between exempt (active) and subpart F (passive) income, full- inclusion systems eliminate the need to police the border between uncreditable foreign taxes associated with exempt income and creditable foreign taxes associated with subpart F income.
Notwithstanding these attractive features of any full-inclusion system, simply "repealing deferral" by itself is likely to be profoundly noncompetitive. First, current U.S. corporate income tax rates are much too high, relative to those of our important trading partners.66 Second, without modification, our current FTC system, and in particular its interest expense allocation rules, would leave too many companies with "excess" FTCs, which in this context means that their global effective tax burden would be even higher than the (too high) nominal U.S. corporate tax rate. Third, most proposals to repeal deferral have been inconsistent with the economic neutrality that the proposal purports to espouse, in that the repeal of deferral is not accompanied by an ability on the part of the U.S. parent to deduct losses incurred by foreign operations.67 Fourth, proposals to end deferral for direct investments ordinarily drive a wedge between the tax burden imposed on direct investments and the burdens imposed on portfolio investment, because the latter means of employing capital in a foreign business would still enjoy the benefits of deferral.
While it follows from the above that simply repealing deferral would be anticompetitive, it remains the case that a full-inclusion system, like a territorial system, would eliminate current law's important distorting effects on firms' repatriation decisions. Full- inclusion systems also dampen the incentives found in current law (which would be exacerbated by territorial tax systems) for multinational corporations to engage in overenthusiastic transfer pricing strategies. And finally, the adoption of a full-inclusion system would eliminate current law's incentives for U.S. multinationals to game the boundary between exempt and subpart F income and to cause the migration of high effective foreign tax rates to subpart F income, all for the purpose of minimizing global tax liabilities.
In light of those attractive elements of a full-inclusion system, the intriguing question is, can a full-inclusion system be designed that retains those desirable features, but is pro- competitive as well? I believe that a review of how the BEIT would apply to outbound investments demonstrates that the question can be answered in the affirmative.
B. Application of BEIT to Outbound Investment
From an internationalist's perspective, the BEIT can be seen in large measure as the perfect mirror image of a territorial system. The international aspects of the BEIT begin with the "super tax" consolidation described in Part II, above. That idea is intended to apply globally. As a result, the BEIT treats foreign subsidiaries as if they were branches. The most obvious consequence of that, of course, is the end of deferral (and with it, the need to maintain rules to distinguish between active income and subpart F income). Another immediate consequence is to vastly attenuate the relevance to the United States of transfer pricing issues for outbound investments, for the reasons already described. Global consolidation also means that foreign losses will be deductible in the United States as those losses are incurred, thereby restoring true neutrality in application when compared to current law, and to many proposals over the years to "end deferral."
The BEIT divides all investments in business enterprises into two categories: controlling interests (which trigger the super- consolidation rules referenced earlier) and other interests (which give rise to current taxable income, through the minimum inclusion mechanism). As a result, current law's concept of a controlled foreign corporation that is controlled by, say, three unrelated U.S. investors in equal proportions would no longer exist. Similarly, the hope is that current law's passive foreign investment company rules also would no longer be required. In each case, investors will include annually their minimum inclusion amounts (without regard to cash distributions), just as they would with investments in U.S. firms.
Without more, the BEIT's international aspects could fairly be described as economically neutral regarding transfer pricing, repatriation decisions, and the location of risky investments, but probably on balance still anticompetitive. The BEIT contains two other critical design elements, however, that revise that calculus to yield a system that fair-minded business people should agree is pro- competitive. The first, and most important, is lower tax rates -- as mentioned above, 28 percent is the goal, but 25 percent (if affordable) would be even better -- financed through systematic base broadening.68 The second design element is the repeal of the allocation of domestic interest expense (now COCA) expense deductions against foreign income for purposes of calculating a U.S. business enterprise's allowable FTC for its international operations, for the reasons described below.
As previously described, sensible territorial tax proposals must incorporate an interest expense allocation system (or some equally painful alternative, such as an efficacious thin capitalization regime). The reason, of course, is that the failure to do so would mean that territoriality would quickly lead to a zeroing out of the U.S. domestic business tax base, by borrowing money (and deducting the resulting interest expense) domestically and supporting the attendant interest deductions with exempt cash flows from equity- financed foreign investments.
The BEIT abandons interest expense (now COCA expense) allocations for two reasons. First, by virtue of the "true" consolidation of foreign income, there is no income that is exempt or indefinitely deferred anywhere in the BEIT system. As a result, there is no urgent need to protect the U.S. tax base by ensuring that domestic interest expense is not ultimately serviced from deferred or exempt income.
The second, and ultimately more powerful, reason why domestic COCA expense need not be allocated against foreign income under the BEIT is that the purpose of the COCA deduction in the BEIT is different from today's interest expense deduction. In the BEIT, the COCA deduction exists to achieve a form of business enterprise- investor integration and applies across the board to all forms of financial capital invested in a business. As such, the COCA deduction is not an "expense"; it is an income allocation device.
If we were to imagine that all business enterprises were 100 percent equity funded, we would not spend much time worrying about allocating the (nonexistent) cost of capital deductions. The COCA result is the same in theory (but superior in many practical respects) to a world in which all interest expense is disallowed or in which (to put things in today's perspective) all firms are 100 percent equity funded. Accordingly, given that under the BEIT we have neither exempt nor deferred income and that we also have implemented an integrated tax system, there is no convincing reason to treat the device by which we achieve that integration as if it were an old- fashioned interest expense deduction.69
I previously observed that portfolio investments have taken on a larger role in cross-border financial flows in recent years. A tax system that produces radically different results for portfolio investments by U.S. households in foreign companies as compared with portfolio investments in U.S. business enterprises (which in turn make foreign direct investments) will prove not to be stable. One important question in that calculus is how to deal with foreign income when distributed by a U.S. business enterprise to its domestic investors.70
The BEIT addresses those issues differently than do other proposals. As described in Part II, full consolidation combined with the COCA deduction/inclusion system basically works to tax economic rents and risky returns at the business enterprise level, and time value returns at the investor level. The COCA component of the BEIT achieves neutrality between U.S. portfolio investors investing in either U.S.-based multinational firms or foreign-based firms -- between, say, investing in Exxon or investing in British Petroleum -- by the simple expedient of applying its investor minimum inclusion rules (current inclusion of time value of money returns, regardless of cash distributions) to portfolio investments in foreign companies, just as those new rules apply to domestic portfolio investments. Finally, the BEIT achieves source neutrality at the level of U.S. portfolio investors in U.S. firms with foreign income by not discriminating (through compensatory taxes or otherwise) against different source of enterprise-level earnings when ultimately received by investors.
This report does not generally address the BEIT's approach to taxing inbound investment into the United States, but the above discussion points to an advantage that the BEIT offers in dealing with inversion transactions, or more generally with the phenomenon of new business enterprises being organized as offshore companies for the purpose of shielding foreign direct investments from the reach of U.S. net income tax. Under the BEIT, U.S. portfolio investors will be taxed currently on time value of money returns on their investments through the minimum inclusion mechanism. As a result, organizing a new business enterprise outside the United States will not reduce the immediate U.S. tax burden on U.S. portfolio investors in that enterprise. Of course, the minimum inclusion device does not address the tax savings that might follow (and ultimately be enjoyed by U.S. investors) at the business enterprise level regarding the new enterprise's non-U.S. income if the average tax rate on that income is lower than the U.S. business enterprise rate. (By the same token, the BEIT does not create the problem either. It exists today in an even more dramatic form.) The answer here lies in rethinking the definition of a business enterprise's residence71 and in the withholding tax burdens that might be imposed under the BEIT for distributions from a U.S. subsidiary to a tax-haven parent company.
The BEIT also attempts to introduce some rough tax neutrality between majority and minority investments by U.S. multinationals in foreign joint ventures. The BEIT's super-consolidation rules are meant to apply to majority-owned affiliates, which would mean, for example, that the income derived by a 51 percent-owned foreign joint venture would be taxed in its entirety by the United States.72 By contrast, the income earned by a minority-owned foreign joint venture that did not conduct business in the United States would not be subject to U.S. net income tax. Under the BEIT, however, the U.S. multinational investor would be required to include in income each year its minimum inclusion (time value of money) amounts, regardless of cash distributions, as well as any excess distributions it might receive. That rule erodes, at least to some modest extent, current law's cliff effect, in which majority-owned joint ventures are subject to subpart F, and minority-owned ones are not.
Grubert and Altshuler review the economic theory and revenue effect of the international aspects of the BEIT (which their paper -- no doubt sensibly -- renames the "burden neutral" international proposal). They conclude that the BEIT's super- consolidation approach to taxing international investment (along with retention of the FTC system but abandonment of interest expense allocation) "seems to dominate" both current law and territorial tax proposals as a matter of theory.73 Moreover, they provide some encouraging news about tax rates. To be clear, Grubert and Altshuler do not offer a revenue estimate for the BEIT as a whole. But they do estimate that, if the super-consolidation/FTC provisions described above were grafted onto current law, the tax rate imposed on the international income of U.S. corporations could come down to 28 percent and the BEIT's international provisions would still be revenue neutral compared with current law.74
The principal criticism that can be leveled against the international provisions of the BEIT -- or, indeed, of any full- inclusion system -- is that the system can distort at the margin international investments by U.S. business enterprises. If foreign tax rates are materially lower than those of the United States, it is argued that U.S. firms would have no great incentive to minimize their foreign tax burden. Conversely, if tax rates are very high in a foreign jurisdiction, a U.S. firm at the margin would have an incentive to "average down" its effective foreign tax rate by making its next investment in a low-taxed jurisdiction.75
The first objection to a full-inclusion system -- the indifference to actual foreign tax liabilities, if the aggregate effective foreign tax rate is materially lower than that of the United States -- is substantially undercut in a world where the U.S. corporate tax rate has been repositioned at the low end of the rates imposed by the major world economies. That, of course, is a key component of the implementation of the BEIT. Moreover, we have today regulations in our FTC systems that prohibit the crediting of "voluntary" taxes, and, more importantly, so-called soak-up taxes.76 Those rules in fact work reasonably well. As a result, the United States is largely the beneficiary of a "free rider" phenomenon, in which local firms can be expected to lobby for lower local tax rates, which local subsidiaries of U.S. firms also will enjoy.
The BEIT responds to the second objection to any full-inclusion system -- that, at the margin, a U.S. firm might have an incentive to invest in a very low-tax jurisdiction to average down its overall foreign tax rate to the amount allowable as a credit in the United States -- by eliminating the allocation of U.S. interest expense (now COCA) deductions against foreign income for FTC purposes, for the reasons described above. Current law's interest expense allocation rules are necessary in our deferral system, but they also are the principal source of "excess" FTC problems, and, with them, the incentive for U.S. firms to average down their FTC systems.
Despite the above rebuttals, I acknowledge that even a well- implemented full-inclusion system brings with it the theoretical possibility of some distortions to investment behavior, particularly if U.S. tax rates are so low as to leave many U.S. firms in excess credit positions, even in a world without interest (COCA) expense allocation for FTC purposes.77 Ultimately, policymakers will not be able to choose a perfect international tax system -- that cannot exist in a world of many sovereign nations with different rates -- but they can endeavor to adopt the least distortive practical design. A territorial tax system brings with it two problems that, for all the reasons described above, are insuperable at a practical level: the policing of transfer pricing and the policing of the divide between active (exempt) and passive (currently taxable) income.78 Against those overwhelming problems, the objection to a well-designed full-inclusion system -- that it might encourage a firm to invest real capital in a location that makes little business sense, to average down its aggregate foreign tax rate to the U.S. rate -- seems, to this practitioner at least, a remote and speculative concern.
A further potential objection to a full-inclusion system is that it could raise complicated transition issues.79 That objection, however, could be leveled at any serious modification to the current regime, and many commentators have emphasized the need for altering the international tax rules.80 My proposal is thus premised on the growing consensus that change is necessary, and with change comes the cost of transition.81
Finally, the problem of stateless income (described above in Part III.E), which has become both more urgent and more obvious in recent years, explains my response to another criticism that might be leveled against the particular implementation of a full-inclusion system that I advocate, which is that it is different from the tax systems employed in the other major capital exporting countries. The major European capital exporting countries, in particular, can fairly be said to be in a state of crisis regarding their own territorial tax systems, as a result of the ECJ's approach to the intersection of EU member state cross-border investment rules and EU constitutional concerns.82 That is an area in which I believe the United States could lead by example. The result would be both conformity to a new norm and a sharp reduction in stateless income, which is another way to get to a playing field that is fair as well as level.
1 President's Advisory Panel on Federal Tax Reform, Simple, Fair, and Pro-Growth: Proposals to Fix America's Tax System (Nov. 2005), available at http://www.taxreformpanel.gov/final-report/ (hereinafter Tax Reform Panel Report).
2 I presented the bare-bones outline of the BEIT in "The Business Enterprise Income Tax: A Prospectus," Tax Notes, Jan. 3, 2005, p. 97 (hereinafter BEIT Prospectus). That outline was expanded in some respects in a presentation made to the President's Advisory Panel on Federal Tax Reform, available online at http://www.taxreformpanel.gov/meetings/meeting-05_11-12_2005.shtml (hereinafter BEIT Presentation). Finally, an explanation of the conceptual underpinnings of the BEIT, titled "Designing an Income Tax on Capital," was presented at the Brookings Institution in Sept. 2005 and is scheduled to appear in a volume to be published in 2007 containing the papers from that conference.
3 The BEIT is an income tax, and therefore by definition accepts one distortion that consumption taxes are designed to eliminate, which is the distortion attendant on taxing future consumption financed through savings more heavily than current consumption. In practice, that distortion will have no effect on the lives or savings of most Americans because the BEIT is intended to coexist with tax-deferred savings plans of the sort embodied in current law or in the recommendations of the President's Advisory Panel on Federal Tax Reform. As a result, under the BEIT the only savings that will be materially burdened by current taxation will be those of the wealthiest Americans because they are the only taxpayers with significant savings that exceed those sheltered by tax-deferred savings plans. The author, at least, believes that the resulting additional progressivity to the individual income tax is consistent with political ideals and practical revenue constraints.
4 There are also special rules for financial services firms, under which those institutions basically are taxed on a mark- to-market basis (for liabilities as well as assets). See BEIT Prospectus, supra note 2, at 103-105.
5 Repeal of the current deferral regime has been recommended before, for example, by Robert J. Peroni, J. Clifton Fleming Jr., and Stephen E. Shay. See Peroni, Fleming, and Shay, "Getting Serious About Curtailing Deferral of U.S. Tax on Foreign Source Income," 52 SMU L. Rev. 455, 507-519 (1999). They proposed a "passthrough" approach, under which each U.S. person owning stock in a foreign corporation would be required to include currently a pro rata share of the corporation's gross income and expenses in computing its own U.S. tax liability. A U.S. investor also would be permitted to deduct a pro rata share of the foreign corporation's losses, up to the amount of the shareholder's basis of its investment in the stock of the foreign corporation.
When applied to a wholly owned foreign subsidiary, the results reached under the Peroni-Fleming-Shay model would be roughly similar to those obtained under the BEIT. Even in this circumstance, however, there are important differences between the two recommendations. For example, the Peroni-Fleming-Shay approach limits loss use to a U.S. person's tax basis in its investment. Similarly, the Peroni-Fleming- Shay model does not contemplate revising the U.S. interest expense allocation rules for FTC purposes (as does the BEIT in respect of its replacement for interest expense deductions, the cost of capital allowance or COCA).
As applied to minority investments in a foreign corporation, the Peroni-Fleming-Shay model and the BEIT diverge more sharply, because the former applies its passthrough model to all U.S. investors, while the BEIT taxes noncontrolling U.S. shareholders in a foreign company in the same manner that they would be taxed regarding a noncontrolling domestic investment. The BEIT thus attempts to achieve neutrality in result across domestic and foreign minority investments. The BEIT also is more straightforward to apply for minority investors because it does not require a minority investor to have any knowledge concerning its pro rata share of the foreign company's results.
6 Special rules not described here ensure that COCA works seamlessly with financial derivatives. See BEIT Prospectus, supra note 2, at 105-106.
7 A potential political weakness of the BEIT system is that, at least in its idealized form, the BEIT would explicitly tax current tax-exempt investors on their time value of money returns, but not on excess returns. CBIT also would have currently taxed tax- exempt investors, but would have done so indirectly. A practical implementation of the BEIT is expected to modulate this ideal result.
8 See Testimony of Dean R. Glenn Hubbard, Impact of International Tax Reform on U.S. Competitiveness: Hearings Before the Subcommittee on Select Revenue Measures of the House Committee on Ways and Means, 109th Cong. (2006) (hereinafter Hubbard Testimony) (encouraging policymakers to review "fundamental reforms like a territorial system, with a view to removing biases [in the U.S. international tax system] against the ability of U.S. multinationals to compete globally").
Interestingly, however, the most sophisticated analyses have argued that, by virtue of (1) eliminating the FTC blending strategies described later in this report and (2) disallowing U.S. interest expense allocable to exempt income, a territorial tax system might actually raise total taxes on income derived from very-low- taxed jurisdictions. Rosanne Altshuler and Harry Grubert, "Where Will They Go If We Go Territorial? Dividend Exemption and the Location Decisions of U.S. Multinational Corporations," 54 Nat'l Tax J. 787 (2001). See also Part III.E.
9 See generally Mihir A. Desai, C. Fritz Foley, and James R. Hines Jr., "Repatriation Taxes and Dividend Distributions," 54 Nat'l Tax J. 829 (2001) (finding that repatriation taxes generate annual efficiency losses equal to 2.5 percent of dividends).
For an economic analysis of the changed incentives created through the elimination of a repatriation tax, see generally Rosanne Altshuler and Harry Grubert, "Repatriation Taxes, Repatriation Strategies and Multinational Financial Policy," 87 J. Pub. Econ. 73 (2003) (working with a model of a parent and its low-tax affiliate).
10 See staff of the Joint Committee on Taxation, "The Impact of International Tax Reform: Background and Selected Issues Relating to U.S. International Tax Rules and the Competitiveness of U.S. Businesses," JCX-22-06, Doc 2006- 12053, 2006 TNT 120-17, at 5 (June 21, 2006). See also American Bar Association, "Report of the Task Force on International Tax Reform," 59 Tax Law. 649, 786 (2006) (hereinafter ABA Report) ("Irrespective of one's views regarding the broader issues relating to deferral, there is a consensus regarding both the high cost of compliance with, and the ineffectiveness of many parts of, the subpart F rules. As such, the current rules may be viewed as the worst of all worlds: avoidable, but only with significant transaction costs.").
11 Michael J. Graetz and Paul W. Oosterhuis, "Structuring an Exemption System for Foreign Income of U.S. Corporations," 54 Nat'l Tax J. 771, 774 (2001).
12 Harry Grubert, "Enacting Dividend Exemption and Tax Revenue," 54 Nat'l Tax J. 811, 814 (2001).
13 Graetz and Oosterhuis, supra note 11, at 776- 778. See also ABA Report, supra note 10, at 673 (advocating the modernization of subpart F "if the exemption proposals of either the Joint Committee Staff or the President's Advisory Panel were adopted, since each would retain the subpart F regime").
14 For example, Harry Grubert and John Mutti contemplate that "the current anti-abuse regime that applies to controlled foreign corporations [that is, subpart F] . . . would also continue in force." Grubert and Mutti, Taxing International Business Income: Dividend Exemption Versus the Current System 9 (2001), The AEI Press available at http://www.aei.org/docLib/20021130_71546.pdf. This thought could be read as implicitly incorporating all of current law's treatment of related-party interest and royalties (ex-section 954(c)(6)), or it could be read as signaling that the authors simply did not expressly consider the issue.
15 The IRC contains rules that exempt from the reach of subpart F some "active" royalty income. Those rules require that the foreign affiliate that owns the intangibles in question have developed or added significant value to the intangible, or, in the case of marketing intangibles, have provided substantial services in connection with marketing the product in question. Section 954(c)(2)(A); reg. section 1.954-2(d).
Those exceptions from subpart F do not apply to royalties received from corporate affiliates, other than affiliates that employ the intangibles in question in the country in which the licensor is incorporated. However, royalties paid by a foreign affiliate that itself is treated as a disregarded entity owned by the licensor under the check-the-box rules of reg. section 301.7701-3 do not give rise to income in the United States sense at all, even though those payments are treated as real for foreign tax purposes.
Moreover, in some cases profits from high-value intangibles can readily be converted from royalty streams into operating income beyond the reach of subpart F. Thus, at the cost of building a highly automated manufacturing facility or CD-ROM pressing plant, a foreign subsidiary located in Ireland (for example) could exploit intangibles owned by it in respect of pharmaceuticals or computer software as sales of physical goods, which sales in turn would fall outside of subpart F.
Grubert, supra note 12, at 816, attempts to model the revenue impact of the expected migration of royalty streams paid to a U.S. parent as a result of a switch to a territorial system. He does not appear explicitly to break out the separate effect of a change in systems on the behavior of foreign subsidiaries holding high-value intangibles.
16 Grubert, supra note 12, at 814. The interest expense allocation proposals in particular typically apply "worldwide" fungibility principles (as developed in the American Jobs Creation Act of 2004), thereby avoiding the logical errors of prior law's "water's-edge approach," Daniel N. Shaviro, "Does More Sophisticated Mean Better? A Critique of Alternative Approaches to Sourcing the Interest Expense of U.S. Multinationals," 54 Tax L. Rev. 353 (2001), but even worldwide fungibility can be criticized as significantly imperfect because it does not treat foreign currency translation losses as, in effect, a component of worldwide interest expense.
17 Similar arguments have been made regarding other U.S. domestic expenses (for example, "head office" general and administrative expenses, or domestic research and development expenditures), but there is less of a consensus on how those expenses should be treated.
18 See ABA Report, supra note 10, at 716 (saying that the current U.S. tax rules encourage "using transfer pricing to shift additional income to foreign corporations subject to low effective tax rates").
19 IRS News Release IR-2006-142, Doc 2006-19012 , 2006 TNT 176-6 (Sept. 11, 2006).
20 Harry Grubert and Rosanne Altshuler, "Corporate Taxes in the World Economy: Reforming the Taxation of Cross-Border Income," presented at the Baker Institute for Public Policy on Apr. 27, 2006, available at http://bakerinstitute.org/Pubs/conferences/2006_tax_007.pdf (hereinafter Grubert and Altshuler).
21 Id. at 9.
22 Id. at 10 (emphasis added).
23 Id. at 18.
24 Id. at 26.
25 Martin A. Sullivan, "A Challenge to Conventional International Tax Wisdom," Tax Notes, Dec. 11, 2006, p. 951, Doc 2006-24455, 2006 TNT 238-6.
26 Id. The percentage figure in the text is the unweighted average of the returns on sales for subsidiaries located in France, Germany, Italy, the Netherlands, Spain, and the United Kingdom.
27 Glenn R. Simpson, "Irish Subsidiary Lets Microsoft Slash Taxes in U.S. and Europe," The Wall Street Journal, Nov. 7, 2005, at A1.
28 I derived the latter figure by grossing up Round Island's reported tax liability to Ireland of $300 million at the Irish tax rate of 12.5 percent. I ignored in this calculation the $17 million that The Wall Street Journal reported that Round Island paid in tax to other European countries (presumably through withholding taxes). If, as I believe to be the case, those payments were creditable in Ireland, Round Island's taxable income actually would have slightly exceeded $2.5 billion in 2004.
29 See reg. section 1.482-7(g)(1)-(7) (establishing rules for "buy-in" payments for preexisting intangible property). It is unusual for a cost-sharing agreement to be an entirely "greenfields" arrangement in which neither party contributes existing intangibles to the project.
Cost-sharing agreements can also be criticized as based on a false premise, which is that each party to the agreement bears the financial risk of the costs it has agreed to shoulder. That premise might be valid in the case of third-party arrangements, but is fundamentally not credible when applied to a U.S. parent and its foreign subsidiary: The subsidiary's "risk" is the parent's risk, and the former's capital comes from (or at the sufferance of) the latter. The combination of such an intragroup cost-sharing agreement and territorial tax systems in practice thus reduces to the U.S. parent's agreement to forgo U.S. tax deductions for a specified fraction of its global development costs, in exchange for obtaining an exemption from U.S. tax for the same percentage of its worldwide income derived from the intangibles covered by the agreement.
30 See, in this regard, Hubbard Testimony, supra note 8 ("Multinationals are an intrinsic part of global integration because they represent an alternative means by which nations conduct cross-border transactions. That is, the economic costs of production, transportation, distribution, and final sale may be lower [if] conducted within a single firm than via a series of market transactions. Accordingly, the rise in global integration carries along with [it] an increased volume of transactions for which multinationals have a particular advantage.").
31 See reg. section 1.482-1(b)(1) (stating that a "controlled transaction meets the arm's length standard if the results of the transaction are consistent with the results that would have been realized if uncontrolled taxpayers had engaged in the same transaction under the same circumstances (arm's length result)"). It is true, of course, that, whether under existing Treasury regulations or by virtue of multilateral advance pricing agreements, intercompany tax transfer pricing arrangements for intangibles or services often rely on various "profit split" methods. Those methods may have the indirect effect of allocating (or misallocating) the benefits of groupwide synergies, but do not do so explicitly, and in any event are not universally required under the "best method" approach. See reg. section 1.482-1(c)(1) ("The arm's length result of a controlled transaction must be determined under the method that, under the facts and circumstances, provides the most reliable measure of an arm's length result. Thus, there is no strict priority of methods, and no method will invariably be considered to be more reliable than others.").
32 Also, a U.S. parent company can employ a related strategy, under which it shelters from U.S. tax the zero-taxed royalty income from foreign subsidiaries paid to the U.S. parent company (and thereby not subject to deferral) with FTCs arising from repatriating very-high-taxed operating income from other foreign subsidiaries. Grubert and Altshuler describe that strategy in detail; their paper estimates that in 2000, royalties received by U.S. parent companies amounted to roughly $45 billion, but that roughly $30 billion of the amount was sheltered from tax by those FTC blending strategies. Grubert and Altshuler, supra note 20, at 9-10.
33 It is true, as Grubert and Altshuler point out, that territorial tax systems disable the popular current strategy of blending zero-taxed foreign-source royalties paid to the U.S. parent by foreign subsidiaries with high-taxed dividend income to shelter those royalties from tax. Grubert and Altshuler, supra note 20, at 28-30. Without considering any possible dynamic responses by U.S. multinational firms, the effect of a territorial tax system thus would be to raise the effective rate on the exploitation of intangible assets from low-taxed jurisdictions. Id. at 29.
One probable dynamic response by taxpayers to a territorial system would be to attempt to understate royalty payments owed to the U.S. parent by foreign subsidiaries. Id. at 30. In addition, cost-sharing agreements, in particular, do not ordinarily generate royalty payments to the U.S. parent company beyond any "buy- in" payments required from the foreign subsidiary. That means that, for companies that employ cost-sharing agreements, royalty payments to the United States should decline relative to the value of the intangible assets that the foreign subsidiary owns outright with the passage of time. As royalties paid to the United States decline (in absolute or relative terms), a foreign subsidiary will be able to capture more profits over time as exempt active foreign income.
34 Staff of the Joint Committee on Taxation, "Options to Improve Tax Compliance and Reform Tax Expenditures," JCS-02-05, Doc 2005-1714, 2005 TNT 18- 18, at 195 n.431 (Jan. 7, 2005) (hereinafter JCT Staff I) (noting that an "exemption system may place somewhat more pressure on [transfer pricing rules], thus making it somewhat more important to remedy existing defects in the design and administration of those rules."); Tax Reform Panel Report supra note 1, at 242 (stating that "because pressures" to use transfer pricing to minimize taxable income "are more pronounced in a territorial system, it would be necessary to continue to devote resources to transfer pricing enforcement."); Peter Merrill et al., "Restructuring Foreign-Source-Income Taxation: U.S. Territorial Tax Proposals and the International Experience," Tax Notes, May 15, 2006, p. 799, Doc 2006-7791, 2006 TNT 94-33 (arguing that the incentive for transfer pricing gaming will become greater under territoriality); Graetz and Oosterhuis, supra note 11, at 772, 775 ("A simpler system would no doubt result if the transfer pricing rules . . . rather than an exclusion from income, could be relied on to constrain tax avoidance [on passive/highly mobile income]"); Sullivan, supra note 25 ("The United States should beef up transfer pricing rules to prevent increasing the incentive effect of already favorable tax rates in production tax havens."); ABA Report, supra note 10, at 723 ("Transfer pricing would have higher stakes for the taxpayer and the Government and enforcement of the rules would have to be strengthened and, possibly, the rules reviewed.").
36 Hubbard Testimony, supra note 8. See also Testimony of Prof. Michael J. Graetz, Impact of International Tax Reform on U.S. Competitiveness: Hearings Before the Subcommittee on Select Revenue Measures of the House Committee on Ways and Means, 109th Cong. (2006) (hereinafter Graetz Testimony) (recognizing "integration of the world economy").
37 Staff of the JCT, supra note 10, at 3, 5, 57-61.
38 See, e.g., Mihir A. Desai and James R. Hines Jr., "Old Rules and New Realities: Corporate Tax Policy in a Global Setting," 57 Nat'l Tax J. 937 (2004).
39 See, e.g., Michael J. Graetz, "Taxing International Income: Inadequate Principles, Outdated Concepts and Unsatisfactory Policies," 54 Tax L. Rev. 261, 272 (2001).
40 The ABA's Task Force on International Tax Reform reaches a similar conclusion in its final report. In its discussion on the different forms of neutrality, the task force states:
The Task Force has not based its analysis on strict application of any one of [the neutrality] principles. None of the principles can be fully achieved by a country unilaterally, and no country applies any of the principles in a pure form. There is not sufficient evidence for the Task Force to conclude that any one of the principles should be determinative in the design of U.S. tax rules. Instead, the Task Force has taken a more pragmatic approach and attempted to evaluate how taxpayers would apply rules in practice and what the incentive effects of rules would be when analyzed in the context of the overall U.S. tax regime.
ABA Report, supra note 10, at 681.
41 Grubert and Altshuler, supra note 20, at 16 (enumeration in the last paragraph supplied by this author).
42 This, in effect, is one major theme of Grubert and Altshuler, supra note 20.
43 The 2005 experience with the one-year 5.25 percent repatriation tax afforded by section 965 illustrates the magnitude of the issue: One estimate put the size of the one-year repatriation flows triggered by that section as in the neighborhood of $200 billion. Grubert and Altshuler, supra note 20, at 19. Another $100 billion was expected to be repatriated by the end of 2006. American Shareholders Association, "ASA Repatriation Scorecard" (Mar. 20, 2006), available at http://www.americanshareholders.com/news/asa-repats-03-20-06.pdf.
44 JCT Staff I, supra note 34, at 189. The JCT staff proposal in turn was said to be modeled on that of Grubert and Mutti, supra note 14.
45 For a comparison of the two proposals, and a rough revenue estimate for the advisory panel's package, see Merrill, supra note 34, at 808-809.
46 Grubert and Altshuler, supra note 20, at 12.
47 Id. at 29. That observation leads to the conclusion, to paraphrase the dry humor of academic articles, that when applied to the lowest-taxed foreign affiliates, a territorial system actually is a step toward capital export neutrality.
48 Testimony of Stephen E. Shay, Impact of International Tax Reform on U.S. Competitiveness: Hearings Before the Subcommittee on Select Revenue Measures of the House Committee on Ways and Means, 109th Cong (2006). The JCT staff made a similar point in 2005: "In many cases, the present-law 'worldwide' system actually may yield results that are more favorable to the taxpayer than the results available in similar circumstances under the 'territorial' exemption systems used by many U.S. trading parties." JCT Staff I, supra note 34, at 189.
49 National Foreign Trade Council, "NFTC Territorial Tax Study Report," at 24 (2002), available at http://www.nftc.org/default/tax/Territorial%20Report.pdf (hereinafter NFTC Territorial Report).
50 For more on the effect of the current system, see ABA Report, supra note 10, at 689 ("The current U.S. international rules allow U.S. multinationals to achieve outcomes that are superior to exemption and therefore cannot be justified by reference to [neutrality principles.] These opportunities are a consequence of structural and technical rules that operate together to afford tax reduction opportunities that almost certainly are unintended.").
51 See supra note 34. See also ABA Report, supra note 10, at 730 ("The Joint Committee Staff and President's Advisory Panel exemption proposals are deficient on several grounds. The failure to include any requirement that the exempt income be subject to a foreign tax will invite substantial tax avoidance planning and place great pressure on transfer pricing rules.").
52 Traditional industrial firms, for example, might have debt-to-equity ratios of 1:1, while the financial services industries' debt-to-equity ratios might be on the order of 30:1.
53 Cf. National Foreign Trade Council, NFTC Foreign Income Project: International Tax Policy for the 21st Century, at 98-99 (2001), available at http://www.nftc.org/default.asp?Mode=DirectoryDisplay&id=162 (hereinafter NFTC Foreign Income Project); Grubert and Altshuler, supra note 20, at 6.
54 Grubert, supra note 12, at 813. Michael J. Graetz and Paul W. Oosterhuis observe that many OECD countries that employ exemption systems impose a 10 percent stock ownership threshold for qualification for that regime. Graetz and Oosterhuis, supra note 11, at 779. Those authors argue, to the contrary, that portfolio investments by U.S. corporations should be exempt from U.S. tax, without regard to active/passive distinctions. Id. They would not extend that result to portfolio investments by households. Id. at 780.
55 See Tax Reform Panel Report, supra note 1, at 243-244 (stating that under its proposal, "shareholders of U.S. corporations could exclude from income 100 percent of the dividends paid from income of the corporation reported as taxable in the United States," implying that the exclusion would be limited in the case of a corporation that is not taxable in the United States).
56 CBIT's designers apparently believed that a compensatory tax was appropriate in this case because the code as then drafted (and, indeed, today) did not grant an indirect FTC to individuals. The code does, however, grant the indirect credit to our principal vehicle for conducting business (the corporation). Because the whole purpose of CBIT and other integration proposals is to treat individual stakeholders as if they directly earned their share of business enterprise income, it is far more logical to assume in designing an integrated tax system that a tax credit that has always been available to prevent double taxation of business income should remain available when that business income is taxed only once, rather than twice. Otherwise, one simply substitutes one form of distortive double taxation for another.
The President's Advisory Panel on Federal Tax Reform followed the logic of CBIT in this respect in fashioning the international tax provisions of the panel's "simplified income tax." As a result, that proposal, like CBIT, would introduce a distortive double tax on foreign income.
57 OECD, Harmful Tax Competition: an Emerging Global Issue, at 43 (1998), available at http://www.oecd.org/dataoecd/33/0/1904176.pdf. One popular solution, rejected by most, but not all, U.S. proposals, is to limit the benefits of exempt income status in a territorial system to income earned in jurisdictions with specified minimum tax rates, or jurisdictions on a "good" list, or jurisdictions not on a blacklist. Id.
58 In light of the central importance of deductible interaffiliate payments in determining the consequences and scope of a territorial tax system, one would expect extensive discussion of the issue in the literature. Oddly, that does not appear to be the case.
59 See, e.g., Cadbury Schweppes PLC & Cadbury Schweppes Overseas Ltd v. Commissioners of Inland Revenue (Case C-196/04), Denkavit Internationaal BV, Denkavit France SARL v. Ministre de l'Économie, des Finances et de l'Industrie (Case C-170/05), and Lankhorst-Hohorst GmbH v. Finanzamt Steinfurt (Case C-324/00).
In Cadbury Schweppes (decided in Sept. 2006) the court held that differential treatment under "controlled foreign companies" legislation of companies resident in one member state on the basis of the level of taxation imposed on their subsidiaries in other member states is prohibited under European Community law, except to the extent the applicable legislation specifically counteracts wholly artificial arrangements aimed solely at escaping national tax normally due, when the legislation does not go beyond what is necessary to achieve that purpose.
Denkavit (decided in Dec. 2006) held that the imposition by a member state of withholding tax on dividends paid to a parent company in another member state is contrary to EC law when a dividend paid to a parent in the same country would not be subject to the tax. The court also held that the existence of a double tax convention that authorizes the withholding tax, and provides for an FTC for the withheld tax, does not alter that conclusion if the parent company is unable to take advantage of the credit.
In Lankhorst-Hohorst (decided in Dec. 2002), the court ruled that thin capitalization rules that apply only to cross-border loan finance without applying to comparable domestic loan finance are contrary to EC law. A new case on cross-border thin capitalization rules is now before the court (Test Claimants in the Thin Cap Group Litigation v. Commissioners of Inland Revenue (C-524/04)). The court has not yet issued its ruling but, interestingly, the advocate general's opinion (issued in June 2006) suggested that such rules may in fact be in conformity with EC law in circumstances in which they can be justified on antiabuse grounds and they do not go beyond what is necessary to attain that objective.
Partly in recognition of the increasing number of infringement proceedings being brought, the European Commission has recently announced a series of initiatives to promote coordination between member states in parallel with litigation. The stated objectives of the initiatives are to provide short- to medium-term targeted measures to remove discrimination and double taxation within the Community, to prevent unintended nontaxation and abuse, and to reduce compliance costs associated with taxpayers being subject to more than one tax system. The first two targeted areas identified for coordination are cross-border loss relief and exit taxation. In the long term, the commission believes that a common consolidated corporate tax base is the solution to removing underlying tax obstacles for corporate taxpayers operating in more than one member state. In his announcement of those initiatives, EU Taxation Commissioner Laszlo Kovacs emphasized the problems currently facing member states when he said, "There is an urgent need to improve coordination of national tax rules to allow them to interact more coherently . . . I am convinced that coordination would help member states to prevent unintended non-taxation or abuse and hence avoid undue erosion of their tax base." "Europe Outlines Coordination Plans for Exit Taxes, Cross-Border Relief," BNA Daily Tax Report No. 244 at G-4 (Dec. 20, 2006). For more on this recent commission proposal, see http://ec.europa.eu/taxation_customs/resources/documents/taxation/COM(2006)825_en.pdf.
See also Harry Huizinga and Luc Laeven, International Profit Shifting Within Multinationals: A Multi-Country Perspective, presented at the European Commission General Directorate Economic and Financial Affairs Workshop on Corporate Tax Competition and Coordination in Europe (Sept. 25, 2006), available at http://ec.europa.eu/economy_finance/events/2006/events_workshop_250906_en.htm (modeling income stripping within the EU).
60 See, e.g., Harry Huizinga, Luc Laevan, and Gaëtan Nicodème, "Capital Structure and International Debt Shifting," presented at the European Commission General Directorate Economic and Financial Affairs Workshop on Corporate Tax Competition and Coordination in Europe (Sept. 25, 2006), available at http://ec.europa.eu/economy_finance/events/2006/workshop250906/tax_conf_nicodeme_en.pdf (EU-based multinational groups' capitalizations of subsidiaries reflect international differences in corporate tax rates).
61 See, e.g., JCT Staff I, supra note 34, at 191.
62 For example, under a territorial tax system a U.S. parent company might try to convert high-taxed exempt income into subpart F income, so that those high FTCs could be used to shelter low-taxed subpart F income elsewhere in the system.
63 See NFTC Territorial Report, supra note 49, at 19 ("in light of the higher stakes presented by a territorial exemption . . . even greater pressure would be placed on the issues of whether and to what extent types of active business income now subject to subpart F (e.g., foreign base company sales and services income) would be eligible for exemption.").
64 Cf. Peroni, Fleming, and Shay, supra note 5, at 514 (under the authors' proposal, "the number of outbound pricing disputes under section 482 should be significantly reduced, thereby lowering taxpayer compliance costs and IRS administration costs. The deferral subsidy encourages U.S. multinational corporations to use intercompany pricing to shift profits to their CFCs operating in tax haven jurisdictions. This passthrough proposal would make such shifts an ineffective tax planning strategy since the profits would be subject to a current U.S. tax in the hands of the U.S. multinational owning stock in the foreign corporation.").
65 The U.S. firm might hope to either maximize low- taxed foreign-source income, or minimize high-taxed foreign income, but only for the purpose of averaging down that very-high-taxed income to the U.S. rate, to be able to use all of its FTCs.
67 See Peroni, Fleming, and Shay, supra note 5, at 501-507 (critiquing two such proposals for curtailing deferral).
68 The COCA system, in particular, is carefully designed, based on 30 years of practice in the area, to be a robust system to capture the time value of money component of financial investments -- the hallmark of an income tax -- on a current basis. As quickly summarized in Part II, the BEIT includes other significant base-broadening components as well. In some cases, that base- broadening flows from the imposition of the "super" consolidation and acquisition rules described earlier. In other cases, it is attributable to the reform of tax accounting rules (for example, the repeal of LIFO inventory accounting and percentage depletion).
69 The absence of a COCA expense allocation deduction can create the misimpression that FTCs are sheltering U.S. domestic business income, but that result is one of cosmetics, not substance. For example, assume that a company has $100 of invested capital (that is, tax basis in its assets), and that the COCA rate (the company's deduction for its cost of capital) is 5 percent. Further assume that the company earns $12 before its COCA deduction, that half of that amount ($6) is treated by both the United States and Freedonia as income arising in Freedonia, and that this $6 accordingly is taxed in Freedonia. Finally, assume that both the Freedonian and the U.S. tax rate is 30 percent.
The company will pay $1.80 in Freedonian income tax. All of that foreign tax will be creditable in the United States, because the company's pre-COCA foreign income is equal to $6, and the Freedonian tax is no greater than the U.S. tax on that income. The net result will be that the company will have $7 of taxable income and a tentative tax liability of $2.10, but will pay only $0.30 to the U.S. government -- or will it? The "missing" U.S. tax liability has not disappeared at all, but rather has migrated to investors, who will have minimum inclusions equals the COCA rate multiplied by their aggregate tax bases in their investment. Assuming for convenience that their bases also equal $100 (in fact of course, this will not be true, but it is a useful simplifying assumption), they will include $5 of income in respect of their investments, and pay $1.50 in tax. So in total the U.S. fisc collects $1.80, and Freedonia collects $1.80, on the company's pre-COCA income of $12, which reflects a tax split that precisely mirrors the relative domestic and foreign pre- COCA taxable incomes of the company.
70 For example, Treasury's 1992 CBIT proposal contemplated imposing a compensatory tax on foreign-source income earned by a U.S. firm when that income was distributed as a dividend to its domestic portfolio investors. See supra note 56.
71 See JCT Staff I, supra note 34, at 178-181 (proposing changes to the current law for determining corporate residency because the law as it now stands "is artificial, and allows certain foreign corporations that are economically similar or identical to U.S. corporations to avoid being taxed like U.S. corporations").
72 One can imagine special rules to deal with this case if the results reached under the general rule were thought inappropriate. For example, one could have a special rule that raised the affiliation test for foreign entities to 60 percent or 65 percent, provided that the minority interests were themselves not publicly traded and were foreign-owned.
73 Grubert and Altshuler, supra note 20, at 31.
74 Id. at 33 ("the burden neutral rate based on 'static' calculations is about 28%").
75 See, e.g., Testimony of Prof. James R. Hines Jr., Impact of International Tax Reform on U.S. Competitiveness: Hearings Before the Subcommittee on Select Revenue Measures of the House Committee on Ways and Means, 109th Cong. (2006).
76 Reg. section 1.901-2(e)(5) ("noncompulsory" taxes); reg. section 1.901-2(c) ("soak up" taxes).
77 In practice, U.S.-based multinationals are likely to deal with the incentive to "average down" in a much more straightforward manner than by locating physical capital in a low-tax jurisdiction. Instead, U.S. firms will average down by financing high-taxed operations with deductible financial capital (in the form of loans paying deductible interest) provided by low-taxed affiliates. Complex "solutions" to that sort of taxpayer behavior can be devised -- for example, by imposing special FTC limitations on interaffiliate interest payments, to discourage such cross-crediting. The text, however, points in a different direction, by arguing that the problem is too remote to require a "solution."
78 To this can be added the practical and political problems in designing a satisfactory interest expense allocation system (or an alternative, like an efficacious thin capitalization solution) to protect the domestic tax base -- and the associated problems of protecting that solution from erosion through years of taxpayer lobbying.
79 Cf. Peroni, Fleming, and Shay, supra note 5, at 519-523 (outlining potential transition relief for the authors' proposal for changing the current deferral rules).
80 See, e.g., JCT Staff I, supra note 34, at 189 ("The present-law system thus creates a sort of paradox of defects: on the one hand, the system allows tax results so favorable to taxpayers in many instances as to call into question whether it adequately serves the purposes of promoting capital export neutrality or raising revenue; on the other hand, even as it allows these results, the system arguably imposes on taxpayers a greater degree of complexity and distortion of economic decision making than that faced by taxpayers based in countries with exemption systems, arguably impairing capital import neutrality in some cases.").
81 For more on transition to the BEIT, see BEIT Presentation, supra note 2, at 15.
82 See supra note 59.
About Tax Analysts
Tax Analysts is an influential provider of tax news and analysis for the global community. Over 150,000 tax professionals in law and accounting firms, corporations, and government agencies rely on Tax Analysts' federal, state, and international content daily. Key products include Tax Notes, Tax Notes Today, State Tax Notes, State Tax Today, Tax Notes International, and Worldwide Tax Daily. Founded in 1970 as a nonprofit organization, Tax Analysts has the industry's largest tax-dedicated correspondent staff, with more than 250 domestic and international correspondents. For more information, visit our home page.
For reprint permission or other information, contact email@example.com