Electronic copy available at: http://ssrn.com/abstract=1303717
Legal Studies Paper No. 2008-36
Four Core Principles of Tax System Design:
Introduced and Applied to the
Taxation of Multinationals
Professor Theodore P. Seto
Electronic copy available at: http://ssrn.com/abstract=1303717
Four Core Principles of Tax System Design:
Introduced and Applied to the Taxation of Multinationals
Theodore P. Seto∗
received a Nobel Prize in 1996,3 has had remarkably little impact on tax
policy analysis in the legal academy. This is due in part to the tools it asks
practitioners to wield – mathematics beyond the comfort level of most
lawyers. More fundamentally, however, the theory as originally
implemented made relatively limited claims. For any given utility
functions, the Vickery-Mirrlees methodology permitted computation of the
earned-income tax rate structure that would maximize aggregate social
welfare.4 Many of the best-known law review articles invoking the theory
have therefore focused on rates to be applied to income from labor
supplied by individuals5 – a relatively small part of the tax policy universe.
∗ Professor of Law, Loyola Law School, Los Angeles.
1 William Vickrey, Measuring Marginal Utility by Reactions to Risk, 13
ECONOMETRICA 319 (1945).
2 James A. Mirrlees, An Exploration in the Theory of Optimal Income
Taxation, 38 REV. ECON. STUD. 175 (1971).
See Peter Passell, 2 Theorists of Real-Life Problems Get Nobel, NY TIMES
(Oct. 9, 1996). Technically, the Nobel Committee cited their work on the
economics of asymmetric information; optimal tax theory was merely one of its
principal applications. Id.
See, e.g., Mirrlees, supra note 2, at 207 (“The examples discussed confirm,
as one would expect, that the shape of the optimum earned-income tax schedule is
rather sensitive to the distribution of skills within the population, and to the
income-leisure preferences postulated. … I had expected the rigorous analysis of
income-taxation in the utilitarian manner to provide an argument for high tax
rates. It has not done so. I had also been expected to be able to show that there was
no great need to strive for low marginal tax rates on low incomes when
constructing negative income tax proposals. This feeling has to some extend been
See, e.g., Joseph Bankman & Thomas Griffith, Social Welfare and the Rate
Structure: A New Look at Progressive Taxation, 75 CAL. L. REV. 1905 (1987);
Edward J. McCaffery, Taxation and the Family: A Fresh Look at Behavioral
Gender Biases in the Code, 40 UCLA L. REV. 983 (1993); Edward J. McCaffery,
Slouching Towards Equality: Gender Discrimination, Market Efficiency, and
Social Change, 103 YALE L.J. 595 (1993); Lawrence Zelenak & Kemper
Moreland, Can the Graduated Income Tax Survive Optimal Tax Analysis?, 52
TAX L. REV. 51 (1999); Lee Anne Fennell & Kirk J. Stark, Taxation Over Time,
Optimal tax theory, for which William Vickery1 and James Mirrlees2
Electronic copy available at: http://ssrn.com/abstract=1303717
The phrase “optimal taxation” is also used more broadly, however, to
refer to tax rules structured to minimize changes in taxpayer behavior – one
of Vickery’s and Mirrlees’ implicit goals. A tax that minimizes behavioral
change also minimizes “deadweight loss” – the difference between the loss
to the economy by reason of the behavioral change (measured in dollars
without taking into account macroeconomic adjustments) and the revenues
raised by the tax. A tax that minimizes deadweight loss is “optimal” if
taxpayers would have behaved optimally in the absence of the tax.
This Article explores how these broader principles of optimal taxation
might be operationalized nonmathematically and applied to problems of
system design beyond rates and the taxation of income from labor. It
focuses on two normative goals: economic efficiency and voluntary
compliance. By thinking more explicitly about taxpayer motivation, it
asserts, we may be able to maximize both efficiency and compliance
simultaneously. Part I begins by defining “avoidance” as either
(1) behavioral change for the purpose of reducing taxation or (2) willful
violation of the tax law without accompanying economic behavioral
change. Part I breaks avoidance into four categories, the most problematic
of which – lawful sheltering – results in both behavioral distortion and
revenue loss. Part II then proposes four nonmathematical principles which,
if applied at the design stage, should minimize both behavioral distortion
and revenue loss. Finally, to illustrate how these principles might operate
in practice, Part III applies them to the design of a tax system for
multinational corporations. Behavioral distortion and revenue loss will be
minimized, it concludes, if multinationals are taxed on book income,
computed on a consolidated basis and allocated to the jurisdiction imposing
the tax using a single factor formula based on adjusted sales – that is, sales
adjusted by industry average profit-to-sales ratios. Any deviation from this
base is likely to increase both economic distortion and compliance
problems. This base is therefore “optimal.”
I. The Core Design Problem: Lawful Sheltering
Economic nondistortion and voluntary compliance are more closely
related than is commonly recognized. A taxpayer predisposed to comply
voluntarily is less likely to alter his behaviors to avoid taxes. When Judge
Learned Hand wrote that “[a]ny one may so arrange his affairs that his
taxes shall be as low as possible; he is not bound to choose that pattern
59 Tax L. Rev. 1 (2005). But see Rebecca S. Rudnick, Enforcing the Fundamental
Premises of Partnership Taxation, 22 HOFSTRA L. REV. 229, 307 (1993)
(“Optimal tax theory can be applied to savings decisions.”).
which will best pay the Treasury; there is not even a patriotic duty to
increase one’s taxes,”6 he was reflecting on the appropriate role of judges
in enforcing existing law, not on principles of sound tax design. As
Professor David Weisbach has more recently written: “[A]ll tax planning is
inefficient, not just planning associated with transactions traditionally
thought of as shelters. There is no a priori way to distinguish shelters from
any other tax planning. Instead, we should be suspicious of all tax
It may be useful to begin, therefore, with a taxonomy of tax
avoidance. I use the term “avoidance” to include all behaviors intended to
reduce taxes, regardless of whether we deem them good or bad. Such
behaviors, in turn, can be divided into four mutually exclusive categories:
Desired behavioral responses: changes in economic
behavior affirmatively encouraged by the tax system – for
example, reduced cigarette consumption in response to
tobacco taxes or increased business investment in response
to bonus depreciation (“Type I avoidance”).
Lawful sheltering: changes in economic behavior to reduce
taxes lawfully in response to rules adopted for reasons other
than to encourage such behaviors – for example, lock-in in
response to the realization requirement or offshoring to take
advantage of international deferral (“Type II avoidance”).
Unlawful sheltering: changes in economic behavior to
reduce taxes unlawfully – for example, the infamous Bond
and Option Sales Strategy, also known as “BOSS”8 (“Type
Defiance: willful violations of tax law without changes in
economic behavior – failures to file, refusals to pay,
intentional misstatements, and the like (“Type IV
The first two are lawful; we commonly refer to them as “minimization.”
The third and fourth are unlawful; they constitute “evasion.”
Helvering v. Gregory, 69 F.2d 809, 810 (2d Cir. 1934), aff’d, 293 U.S. 465
(1935) (Hand, J.).
7 David A. Weisbach, Ten Truths About Tax Shelters, 55 TAX L. REV. 215,
See, e.g., Ben Wang, Supplying the Tax Shelter Industry: Contingent Fee
Compensation for Accountants Spurs Production, 76 S. CAL. L. REV. 1237, 1243
(2003) (explaining the complexities of BOSS). The IRS held that BOSS would not
be respected for tax purposes in Notice 99-59, 1999-2 C.B. 761 (1999).
Note that the first three all involve economic behaviors that would not
be undertaken in the absence of a tax system. Unless the relevant tax rules
are Pigouvian9 – intended to rectify market failures – they are all therefore
inefficient in the formal economic sense. None is warranted as a matter of
tax policy. Type I avoidance is the only type thought to be good, and then
only for non-tax reasons.10 We condone Type II avoidance – lawful
sheltering – only because we believe we have no choice.
Type II avoidance may actually be the most problematic of the four.
Defiance merely reduces collections. Lawful sheltering, by contrast, both
reduces collections and distorts economic decision-making, sometimes
quite seriously. In addition, the line between lawful and unlawful
sheltering is extremely difficult to draw.11 As a result, taxpayers seeking to
expand opportunities to engage in lawful sheltering often cross the line –
sometimes aggressively, sometimes unwittingly – forcing the Service and
the courts to struggle with the difference, often unsuccessfully. Finally,
lawful sheltering, although commonly perceived as illegitimate, is lawful
and therefore undermines the legitimacy of the system as a whole.
It should therefore come as no surprise that a significant part of
compliance doctrine and literature focuses on the difference between
lawful and unlawful sheltering. The “business purpose” and “economic
substance” doctrines, for example, tell us that if a transaction has an
9 Pigouvian taxes are taxes imposed to correct market failures – for example,
to internalize externalities. See, e.g., Maureen B. Cavanaugh, On the Road to
Incoherence: Congress, Economics, and Taxes, 49 UCLA L. REV. 685, 715-721
(2002); McCaffery, Taxation and the Family, supra note 5, at 1046-1053.
10 We commonly refer to provisions encouraging Type I avoidance as “tax
expenditures.” See, e.g., OFFICE OF MANAGEMENT AND BUDGET, ANALYTIC
PERSPECTIVES, BUDGET OF THE UNITED STATES GOVERNMENT, FISCAL YEAR
2009, at 287. One current controversy in the computation of tax expenditures is
the proper choice of baseline. The frame set forth in this Article suggests at least
three possible such baselines: (1) application of effective rate structures to
taxpayers’ maximands, (2) application of effective rate structures to taxpayers’
maximands as modified by Pigouvian taxes or subsidies, or (3) application of
effective rate structures to taxpayers’ maximands as modified by Pigouvian taxes
or subsidies and Type II avoidance.
See David Hariton, Kafka and the Tax Shelter, 57 TAX L. REV. 1, 2-3 (2003)
(“For almost two years now, congressional staffs have had as one of their
principal agendas the drafting of a provision that could help to define a tax shelter
and distinguish it from other financial transactions that are structured to minimize
their tax consequences. For almost four years now, Treasury has been trying to do
the same thing by regulation. The courts likewise have been struggling to apply a
coherent standard of scrutiny to tax benefits arising from tax-motivated
articulable business purpose or economic substance, it will more likely be
classified as lawful.12 This makes intuitive sense: on average, a transaction
with an articulable business purpose or economic substance probably
involves less economic distortion than a transaction without such purpose
In any event, the approach to tax design suggested in this Article
assumes that lawful sheltering is a key part of the compliance problem. In
particular, by taxing whatever it is that taxpayer otherwise seeks to
maximize, the approach strives to make it more difficult for taxpayers,
without compromising their core non-tax goals, to undertake behavioral
change in pursuit of tax reduction. In short, it seeks to minimize
opportunities for lawful sheltering in the design of the system itself.
II. Principles of Optimal Tax Design
Economists model anticipated behavioral responses by computing or
making assumptions about the elasticity of relevant supply and demand
curves. Non-economists, however, may find it easier to predict behavioral
responses by considering taxpayer’s motivations directly. What do
taxpayers want? A clear answer, which may well vary from taxpayer to
taxpayer and context to context, may allow us to structure a tax system that
See, e.g., Joseph Bankman, The Economic Substance Doctrine, 74 S. CAL. L.
REV. 5 (2000); Robert Thornton Smith, Business Purpose: The Assault Upon the
Citadel, 53 Tax Law. 1 (1999).
13 A few scholars have recently argued that we should focus on Congress’s
purpose in enacting relevant statutes; if a transaction is consistent with that
purpose, they urge, the transaction should be given tax effect; if not, not. See, e.g.,
Michael L. Schler, Ten More Truths About Tax Shelters: The Problem, Possible
Solutions, and a Reply to Professor Weisbach, 55 TAX L. REV. 325 (2002) (“[A]
tax shelter should be defined as a transaction, or portion of a transaction, that
(1) arguably complies as a literal matter with the Code and regulations, (2) is
accompanied by some level of tax motivation, and (3) reaches a tax result
unintended by Congress or the regulations.”); Shannon Weeks McCormack, Tax
Shelters and Statutory Interpretation: A Much Needed Purposive Approach, U.
ILL. L. REV. (forthcoming), available at http://papers.ssrn.com/sol3/results.cfm?
abstract_id=1119984, at 3 (“[C]ourts should ask whether the results of the
transactions (meaning the tax savings claimed) are within the purposes of the
Code and Regulatory sections they utilize. If the results are not, taxpayers should
be denied the attendant benefits. If the results are within the purposes of the law,
taxpayers are entitled to the benefits.”). The proposed “purposive” approach does
a good job distinguishing Type I avoidance (desired behavioral responses) from
Type III avoidance (unlawful sheltering); unfortunately, it does little to distinguish
between lawful and unlawful sheltering – the core of the compliance problem.
minimizes the avoidance behaviors that produce economic distortion and
revenue loss. To this end, I offer four principles of optimal tax design.
Principle 1 (“relative indifference”): If tax liability turns on factors as
to which taxpayer is relatively indifferent, the system will likely distort
behavior while inviting revenue loss. Thus, for example, if wealthy
taxpayers are relatively indifferent as between luxury yachts and homes on
the Riviera, a tax on luxury yachts will probably lead such taxpayers to buy
fewer yachts and more homes; such a tax distorts behavior and is not likely
to raise significant revenue.14 By contrast, most patients care a lot about the
difference between a liver transplant and cosmetic surgery; a tax-relevant
line between such procedures15 is therefore unlikely to result in behavioral
distortion or concomitant revenue loss. Taxpayers are not relatively
indifferent to whatever it is they seek to maximize. This first principle – the
principle of relative indifference – can therefore be generalized to specify
the optimal tax base. Hence Principle 2.
Principle 2 (“taxing the maximand”): Economic distortion and
revenue loss can be minimized by taxing whatever it is that taxpayer
otherwise seeks to maximize. Assume, for example, that taxpayer’s
overriding motivation is to maximize book income. If a tax is imposed on
book income, no action taxpayer can take will reduce tax liability more
than it reduces book income.16 A tax on book income should therefore
minimize both economic distortion and revenue loss. If we choose to tax
anything else – say, a differently-defined “taxable income” – actions will
likely exist that can save taxes in an amount greater than they reduce book
income. If so, taxpayer will likely take such actions. This will have several
adverse consequences: (1) the system will distort economic decision-
making, (2) tax revenues will decline, and (3) at some point, we will be
required to articulate and enforce a line between “minimization” and
“evasion” – with all the many problems such a line entails. If taxpayer
seeks to maximize a number already computed for non-tax purposes,
applying this second principle by reference to that previously computed
number will also minimize tax administrative costs to both government and
Principle 3 (“taxing a reported maximand”): If what taxpayer seeks to
maximize is a reported number, tax administrative costs will be minimized
if we tax that reported number. If, for example, taxpayer is motivated to
See 39 Authors, The Jurisprudence of Yogi Berra, 46 EMORY L.J. 697, 756-
59 (1997) (describing history of 1990 “luxury tax”).
15 IRC §213(d)(9)(A) (“The term ‘medical care’ does not include cosmetic
surgery or other similar procedures …”).
16 This assumes, of course, that tax is imposed at a rate of less than 100%.
maximize reported book income, a tax on reported book income should be
largely self-enforcing; violations will generally be limited to failures to file
or refusals to pay. Any deviation of the tax base from the relevant reported
number, however, should induce taxpayer to try to maximize the reported
number while minimizing its tax base. Such deviations will generally make
enforcement more expensive and less effective.
Principle 4 (“classification by maximand”): Taxpayers should be
classified for tax purposes by reference to whatever it is they would seek to
maximize in the absence of tax. A majority of wage-earning single
individuals arguably seek to maximize individual realized net income.
Wage-earning married individuals are more likely to seek to maximize the
couple’s realized net income instead, without regard to which partner
formally owns that income. This suggests that single and married
individuals should be taxed differently and that the line between the two
classes of taxpayers should be drawn by reference to their different
anticipated motivations.17 High wealth individuals and couples, by contrast,
are more likely to seek to maximize net worth plus consumption –
something more nearly resembling Haig-Simons income. Economic
distortion and revenue loss may therefore be minimized by requiring such
taxpayers to mark-to-market.
In significant regards, our current system is consistent with these four
principles; Haig-Simons income approximates most individual taxpayers’
maximand. More explicit consideration of taxpayer motivation, however,
could improve both efficiency and compliance. I mean to assert this
premise generally. The remainder of this Article, however, explores the
premise primarily in the context of multinationals.
III. The Case of Multinationals
In general, publicly reporting multinational businesses seek to
maximize reported book income. The principle that we should tax
taxpayer’s maximand therefore suggests that we should use reported book
income, computed on the same consolidated basis used for non-tax
reporting purposes, as such taxpayers’ primary tax base. This cannot be the
end of the story, however, since a taxing jurisdiction can only properly tax
an entity’s income to the extent such income is properly allocable to that
17 I have made this argument in greater detail elsewhere. See Theodore Seto,
The Unintended Tax Consequences of Gay Marriage, 65 WASH. & LEE L. REV. __
Multinationals have no obvious incentive to maximize “U.S. source”
character of their income – source being the concept used by our current
system to determine this country’s proper share. Nor, all else being equal,
do they have any incentive to maximize U.S. plant or payroll – two other
factors sometimes used for allocation purposes. What do they most
plausibly seek to maximize? Market share. The rule that we should tax
taxpayer’s maximand therefore suggests that we should allocate taxpayer’s
consolidated book income between the United States and the rest of the
world using a single-factor apportionment formula based on sales. A
multinational cannot maximize market share while minimizing sales.
But there is a problem.18 Different industries operate at different profit
margins – that is, different industries have different profit-to-sales ratios.
To the extent profits and sales diverge, use of unadjusted sales for
allocation purposes will create an incentive for multinationals to purchase
offshore operations with low profit-to-sales ratios (or high sales-to-profit
ratios), thereby artificially boosting apparent offshore sales and reducing
the portion of the multinational’s consolidated book income allocated to
the United States. Another way of stating the problem is that to the extent
profits and sales diverge, multinationals are relatively indifferent to sales.
The solution is to adjust each taxpayer’s sales, industry-by-industry or
product-by-product, by such industry’s or product’s worldwide average
profit-to-sales ratios. A numerical example may usefully illustrate both the
problem and its solution. Assume that Company A’s U.S. widget and
foreign wodget operations and sales are as follows:
Assume further that both operations are equally relatively profitable within
their respective industries – in other words, that, all else being equal,
Company A would be relatively indifferent as to which operation to buy or
sell at the same given price. Note that half of taxpayer’s pre-tax profits of
200 are attributable to its U.S. operations and sales and half to its foreign
operations and sales. Nevertheless, if we allocate consolidated book
income on the basis of unadjusted sales, we will allocate only one-third to
the United States. This will give multinationals an incentive to purchase
18 I am indebted to Prof. Jim Hines for identification of this problem.
offshore operations with low profit-to-sales (or high sales-to-profit) ratios,
all else being equal.
Now let us instead allocate the same consolidated book income on the
basis of sales adjusted for their respective industry average profit-to-sales
ratios. Assume, for simplicity’s sake, that both of Company A’s operations
are average in this regard, so that Company A’s numbers are representative
of the two industries’. The industry average profit-to-sales ratio for widget
manufacturing and sales is therefore 100 to 1000 (or 10%); the same ratio
for wodget manufacturing and sales is 100 to 2000 (or 5%). Adjusting
taxpayer’s sales by each industry’s average profit-to-sales ratios produces
the following adjusted sales numbers::
Average profit/sales 10%
Adjusted sales 100
If we allocate taxpayer’s consolidated book income on the basis of sales
thus adjusted, we correctly attribute half of that income to each operation
Much has already been written about the possibility of moving to a
formulary apportionment tax system.19 A separate and substantial literature
See, e.g., Julie Roin, Can the Income Tax Be Saved? The Promise and
Pitfalls of Adopting Worldwide Formulary Apportionment, 61 TAX L. REV. 169
(2008); Reuven S. Avi-Yonah & Kimberly A. Clausing, A Proposal to Adopt
Formulary Apportionment for Corporate Income Taxation: The Hamilton Project
(2007), available at http://papers.ssrn.com/sol3/papers.cfm?abstract_ id=995202;
Joann Martens Weiner, Practical Aspects of Implementing Formulary
Apportionment in the European Union, 8 FLA. TAX REV. 629 (2007); Reuven S.
Avi-Yonah, The Rise and Fall of Arm’s Length: A Study in the Evolution of U.S.
International Taxation, 15 VA. TAX REV. 89 (1995); Richard D. Pomp, Issues in
the Design of Formulary Apportionment in the Context of NAFTA, 49 TAX L. REV.
795 (1994); Michael J. McIntyre, The Design of Tax Rules for the North American
Free Trade Alliance, 49 TAX L. REV. 769 (1994); John S. Brown, Formulary
Apportionment and NAFTA, 49 TAX L. REV. 759 (1994); Paul R. McDaniel,
Formulary Taxation in the North American Free Trade Zone, 49 TAX L. REV. 691
(1994); Robert S. McIntyre & Michael J. McIntyre, Using NAFTA to Introduce
Formulary Apportionment, 6 TAX NOTES INT’L 851 (Apr. 5, 1993); Michael J.
McIntyre, Design of a National Formulary Apportionment Tax System, 1991
TH PROCEEDINGS 118 (1992).
exists on the pros and cons of book/tax consistency.20 It is not my purpose
here to enter generally into either debate. My purpose here is rather to
make a more narrow point: All else being equal, taxation of multinationals
in a manner consistent with the principles articulated in Part II –
specifically, taxation based their reported consolidated book income,
allocated using a single-factor apportionment formula based on adjusted
sales – is likely to be less economically distortive and subject to Type II
revenue loss than any alternative. In other words, it is likely to be optimal.
With this advance disclaimer, Subpart A identifies some of the most
salient problems currently besetting U.S. taxation of multinationals, both
domestic and foreign. Subpart B then concludes that the principles of Part
II, consistently applied, would significantly resolve many of those
A. Current Problems in the U.S. Income Taxation of Multinational
The U.S. international corporate income tax system is both opaque
and dysfunctional. Ostensibly, we tax U.S. corporations on their worldwide
income. Because our system respects corporate form, however, well-
advised U.S.-parented multinationals can use wholly-owned foreign
subsidiaries to defer U.S. taxation on most income from foreign activities.
Such income is taxed only when dividended by such subsidiaries to their
U.S. parents and then again when paid out to individual shareholders,
either U.S. or foreign.
See, e.g., David I. Walker, Financial Accounting and Corporate Behavior,
64 WASH. & LEE L. REV. 927 (2007): Celia Whitaker, How to Build a Bridge:
Eliminating the Book-Tax Accounting Gap, 59 TAX LAW. 981 (2006); Celia
Whitaker, Note, Bridging the Book-Tax Accounting Gap, 115 YALE L.J. 680
(2005); Wolfgang Schön, The David S. Tillinghast Lecture, The Odd Couple: A
Common Future for Financial and Tax Accounting?, 58 Tax. L. Rev. 111 (2005);
Linda M. Beale, Book-Tax Conformity and the Corporate Tax Shelter Debate:
Assessing the Proposed Section 475 Mark-to-Market Safe Harbor, 24 VA. TAX
REV. 301 (2004); Yoram Keinan, Book Tax Conformity for Financial Instruments,
6 FLA. TAX REV. 676 (2004); Anthony J. Luppino, Stopping the Enron End-Runs
and Other Trick Plays: The Book-Tax Accounting Conformity Defense, 2003
COLUM. BUS. L. REV. 35 (2003); Gil B. Manzon, Jr. & George A. Plesko, The
Relation Between Financial and Tax Reporting Measures of Income, 55 TAX L.
REV. 175 (2002); Terry Shevlin, Corporate Tax Shelters and Book-Tax
Differences, 55 TAX L. REV. 427 (2002); Mitchell L. Engler, Corporate Tax
Shelters and Narrowing the Book/Tax “GAAP”, 2001 COLUM. BUS. L. REV. 539
Foreign-parented multinationals can take advantage of similar structuring
techniques; indeed, the foreign-activity income of a well-advised foreign-
parented multinational should never be subject to U.S. tax until it is
actually paid out to a U.S. shareholder and should never be subject to U.S.
tax if paid to a foreign shareholder.
taxed not taxed
This basic schema is problematic for many reasons. First, it
effectively disfavors U.S.-parented multinationals with respect to non-U.S.
activities – that is, with respect to 75% of the world’s economic activity.21
In addition, U.S.-parented multinationals are subject to significantly more
restrictive U.S. tax limitations on the movement of earnings from
subsidiary to parent and from parent to shareholders.22 Indeed, if tax were
all that mattered, it is unlikely that most multinationals would now choose
to locate their parent corporations in the United States. As a result, in
recent years “corporate inversions” – transactions in which parentage is
See Part I.1, infra. See also GENERAL EXPLANATION OF TAX LEGISLATION
ENACTED IN THE 108
that corporate inversion transactions were a symptom of larger problems with our
current system for taxing U.S.-based global businesses and were also indicative of
the unfair advantages that our tax laws conveyed to foreign ownership”).
According to the International Monetary Fund, 2007 world GDP was $54.3
trillion, of which the United States accounted for $13.8 trillion. See Wikipedia,
List of Countries by GDP, http://en.wikipedia.org/wiki/List_of_countries_by_
GDP_%28nominal%29 (visited June 27, 2008).
See GENERAL EXPLANATION, supra note 21. Techniques exist for stripping
earnings out of a U.S. subsidiary owned by a foreign parent. Id.; Jim A. Seida &
William F. Wempe, Effective Tax Rate Changes and Earnings Stripping
Following Corporate Inversion, 57 NAT’L TAX J. 805 (2004) (“a large portion of
the … increase … in … foreign pre-tax income shares … is attributable to the
stripping of U.S. earnings via intercompany interest payments”).
TH CONGRESS (JCS-5-05, May 2005) (“The Congress believed
taxed not taxed
converted from U.S. to foreign – have become increasingly common.23 In
2004 Congress attempted to inhibit such inversions;24 these new rules,
however, have been only partially successful.25
Second, unless another country chooses to tax the multinational’s
foreign-activity income, the schema creates a strong incentive to locate
profit-making capacity outside the United States – colloquially, to
“offshore” that capacity.26 All else being equal, other countries can induce
multinationals to move profit-making capacity out of the United States by
taxing such activities at rates lower than effective U.S. rates. As will be
seen, they may do so simply by substituting a water’s-edge value added tax
for a corporate income tax without cutting rates or perhaps even overall
Third, even when a presence in the United States is required for
business reasons, related-party intercompany transactions can often be
priced to leave reported profits offshore.27 In theory, the IRS has the power
See generally Orsolya Kun, Corporate Inversions: The Interplay of Tax,
Corporate, and Economic Implications, 29 DEL. J. CORP. L. 313, 314 (2004)
(“Corporate inversions became a noticeable phenomenon between 1998 and 2002,
when a number of large U.S.-based multinational corporations elected to
expatriate.”); Mitchell A. Kane & Edward B. Rock, Corporate Taxation and
International Charter Competition, 106 MICH. L. REV. 1229 (2008).
24 IRC §7874. Among other things, IRC §7874 treats a foreign parent as
domestic if a multinational undertakes a corporate inversion under specified
25 One of the most widely-publicized examples of a still-anticipated post-§7874
inversion is the transfer of Halliburton Company’s headquarters to Dubai. See
http://abcnews.go.com/GMA/Politics/story?id=2943017. The move may permit
Halliburton to become foreign-parented without running afoul of the new anti-
inversion rules by locating significant management and administrative activities
offshore. Legislation has been introduced that would eliminate management and
administrative activities from the substantial business activities exclusion of IRC
§7874. H.R. 2937 (introduced June 28, 2007). Such legislation has not, however,
26 Offshoring has become a major problem for the U.S. economy. See, e.g.,
Ashok D. Bardhan & Cynthia Kroll, The New Wave of Outsourcing, Fisher Center
Research Reports Paper 1103 (2003), available at http://repositories.cdlib.org/
iber/fcreue/reports/1103 (visited June 27, 2008); Alan E. Blinder, Offshoring: The
Next Industrial Revolution?, 85 Foreign Affairs 113-128 (March/April 2006); Ron
Hira & Anil Hira, OUTSOURCING AMERICA: WHAT'S BEHIND OUR NATIONAL
CRISIS AND HOW WE CAN RECLAIM AMERICAN JOBS. (May 2005). See Part I.2,
See, e.g., Thomas C. Pearson, Proposed International Legal Reforms for
Reducing Transfer Pricing Manipulation of Intellectual Property, 40 N.Y.U. J.
to police such pricing.28 Indeed, one widely-used practice guide states
“nearly 75% of [IRS] adjustments to multinational companies’ incomes in
recent years have been related to transfer pricing issues.”29 In practice,
transfer pricing rules afford considerable leeway to taxpayers and are
costly to enforce;30 as a result, it is unclear how large a fraction of the
resulting abuses are corrected.
Fourth, the U.S. tax system then penalizes the repatriation and U.S.
reinvestment of offshore profits earned by U.S.-parented multinationals. In
2004, Congress authorized a one-time tax holiday with respect to such
repatriations, reducing the effective rate on such repatriations to 5.25
percent for a limited period.31 843 U.S. corporations took advantage of the
opportunity, repatriating some $362 billion in foreign profits.32 $442 billion
in profits, however, remained offshore when the holiday expired.33
Finally, to negotiate the maze that constitutes the U.S. international
tax system, expert tax advice is generally required for even the simplest
These problems are compounded by problems in the definition of the
corporate tax base and the treatment of tax liabilities under generally
accepted accounting principles (“GAAP”). Corporate tax shelters remain a
major concern.34 Although reliable estimates of the volume of corporate tax
INT'L L. & POL. 541, 542 (2008) (“A major type of abusive tax avoidance is
manipulation of transfer prices, which allows MNE Groups to shift income from
higher-tax countries to lower-tax countries.”); See Part I.3, infra.
28 IRC §482 permits the IRS to adjust the incomes of related parties accurately
to reflect income.
See Marc M. Levey et al., Practical Applications of Transfer Pricing: A
Case Study Approach, in MARC M. LEVEY ET AL., TRANSFER PRICING:
ALTERNATIVE PRACTICAL STRATEGIES, 890 BNA TAX MANAGEMENT PORTFOLIO
A-1, at A-1.
See, e.g., Michael C. Durst & Robert E. Culbertson, Clearing Away the
Sand: Retrospective Methods and Prospective Documentation in Transfer Pricing
Today, 57 TAX L. REV. 37 (2003) (“a large-volume exercise, to which virtually all
multinational companies today must devote substantial resources”); Eduardo
Baistrocchi, The Transfer Pricing Problem: A Global Proposal for Simplification,
59 TAX LAW. 941 (2006).
31 IRC §965.
See Lynnley Browning, A One-Time Tax Break Saved 843 U.S.
Corporations $265 Billion, N.Y. TIMES (June 24, 2008).
See, e.g., Lawrence H. Summers, Tackling The Growth of Corporate Tax
Shelters, http://www.treas.gov/press/releases/ls421.htm (Feb. 28, 2000) (“the most
serious compliance issue threatening the American tax system today: the rapid
growth of abusive corporate tax shelters”); Peter Cannellos, A Tax Practitioner's
shelter activity are hard to come by,35 Fortune 500 companies often report
significantly higher profits to their shareholders than they report to the
IRS.36 For reporting purposes, offshoring with a credible commitment
never to repatriate the resulting profits allows a U.S.-parented
multinational to avoid booking even a future tax liability.37 The effect is a
significant tax/GAAP incentive to commit never to return profits to the
United States. Nor does tax theory offer any realistic solution. A partial
consensus appears to exist that double taxation of corporate income – once
at the corporate level, once at the shareholder level – is undesirable and
unwarranted.38 Unfortunately, no politically palatable system of corporate
Perspective on Substance, Form and Business Purpose in Structuring Business
Transactions and in Tax Shelters, 54 SMU L. REV. 47, 47 (“It requires no
citations to establish what is obvious to tax professionals within and without the
government: corporate tax shelters are proliferating, in both type and number, and
their terms are becoming ever more audacious.”).
35 Joel Friedman, The Decline of Corporate Income Tax Revenues 3 (Oct. 24,
2003), available at http://www.cbpp.org/10-16-03tax.htm (“No precise estimates
of the extent of [corporate] shelter activities exist.”).
See, e.g, Robert S. McIntyre & T.D. Coo Nguyen, Inst. on Tax’n & Econ.
Pol’y, Corporate Income Taxes in the 1990s 2 (2000), available at http://
www.itepnet.org/corp00an.pdf. (study of 1996-1998 income statements of 250
largest U.S. companies, all of which were profitable for all years studied,
concluded that 41 paid less than zero in at least one year and 11 faced negative
federal income tax rates for every year studied).
See Part I.4, infra.
See Jennifer Arlen & Deborah M. Weiss, A Political Theory of Corporate
Taxation, 105 YALE L.J. 325, 325 (1995) (“The American tax system imposes a
double tax on the profits of corporations. This two-tier taxation is unusual, unfair,
and inefficient. The ill-effects of the double tax are well known in Washington.
Congress regularly considers legislation to eliminate the double tax by integrating
the personal and corporate taxes into a single system. These initiatives have had
the support of tax scholars, the Treasury, the public, and several presidents. Yet
proposals to integrate the tax system invariably die a quiet death.”). See also
TREASURY DEP’T, REPORT ON THE INTEGRATION OF THE INDIVIDUAL AND
CORPORATE TAX SYSTEMS: TAXING BUSINESS INCOME ONCE (1992); Katherine
Pratt, Deficits and the Dividend Tax Cut: Tax Policy as the Handmaiden of Budget
Policy, 41 GA. L. REV. 503 (2007); Joshua Mishkin, The State of Integration in a
Partial Integration State, 59 TAX LAW. 1047 (2006); Yariv Brauner, Integration
in an Integrating World, 2 N.Y.U. J. L. & BUS. 51 (2005); Katherine Pratt, The
Debt-Equity Distinction in a Second-Best World, 53 VAND. L. REV. 1055 (2000);
Terrence R. Chorvat, Taxing International Corporate Income Efficiently, 53 TAX
L. REV. 255 (2000); Deborah H. Schenk, Foreword, Colloquium on Corporate
Integration, 47 TAX L. REV. 427 (1992). But see Reuven Avi-Yonah,
Corporations, Society, and the State: A Defense of the Corporate Tax, 90 VA. L.
tax integration has yet been proposed. The U.S. corporate tax system
therefore remains in limbo, caught between theory and reality. Finally, the
IRS is badly outgunned on the corporate tax battlefield.39 Indeed, to
maximize the impact of its enforcement resources, the Service has recently
announced that it will focus its corporate tax audits on a limited subset of
How U.S. international tax rules favor foreign-parented
multinationals.—In theory, under current U.S. law, domestic U.S.
corporations are taxable on their worldwide income.41 Foreign
corporations, by contrast, are taxable only on income with a U.S.
connection,42 specifically: (1) income effectively connected with a U.S.
trade or business (“ECI”)43 and (2) specified types of U.S.-source income,
for the most part income other than income from sales (“U.S.-source
FDAP”).44 Whether a corporation is treated as domestic or foreign is
determined primarily by looking at place of incorporation – a purely formal
REV. 1193, 1196 (2004) (“integration of the corporate and shareholder taxes … is
not necessary to prevent ‘double taxation’”); Steven A. Bank, Is Double Taxation
a Scapegoat for Declining Dividends? Evidence from History, 56 TAX L. REV.
463, 467 (2003) (“integration may not be an effective solution to the retained
See Linda M. Beale, Putting SEC Heat on Audit Firms and Corporate Tax
Shelters: Responding to Tax Risk With Sunshine, Shame and Strict Liability, 29 J.
CORP. L. 219,220 (2004) (“The IRS is underfunded and understaffed, resulting in
inadequate enforcement. As in the case of the internal finance and accounting
departments that have moved from gatekeepers to profit centers, corporate tax
departments have expanded beyond their primary service role – ensuring
compliance with tax rules – to become profit centers that design novel ways to
circumvent those rules. Corporations can hide aggressive tax planning in
aggregate numbers on tax returns and financial statements. Company directors
either are uninformed about, or acquiesce in, potentially abusive tax planning.
Audit firms do not highlight these issues for directors or investors and, in some
cases, actively assist in designing aggressive tax shelters for lucrative fees.”).
40 IRS, Industry Issue Focus, http://www.irs.gov/businesses/article/0,,id=
167377,00.html (visited June 11, 2008).
41 Internal Revenue Code (“IRC”) §11.
42 IRC §§11(d), 881, & 882.
43 IRC §882. Rule governing the determining a corporation’s effectively
connected income are set forth in IRC §864(c).
44 IRC §881. The acronym “FDAP” is commonly used to refer to items listed
in IRC §881(a).
45 IRC §7701(4) & (5). An exception is created by the new corporate inversion
rules of IRC §7874. See note __, infra.
tax currently on their worldwide income. Operations outside the United
States are typically conducted through wholly-owned foreign subsidiaries.
Because the U.S. tax system respects the corporate form, such foreign
subsidiaries are subject to U.S. income taxation only if they receive or
accrue income with a U.S. connection – that is, ECI or U.S.-source FDAP.
Limited types of foreign activities undertaken by foreign subsidiaries are
taxed currently to the U.S. parent under Subpart F – the “controlled foreign
corporation” rules.46 The bulk of all business activities, however, are
exempt from Subpart F.
Foreign-parented multinationals typically conduct U.S. operations
through wholly-owned U.S. subsidiaries. The earnings of these subsidiaries
are subject to current U.S. income taxation at the subsidiary level. Foreign
operations are typically conducted through wholly-owned foreign
subsidiaries, which are exempt from U.S. income taxation unless they
engage in a U.S. trade or business or receive U.S.-source FDAP. Thus,
apart from Subpart F (which effectively applies only to U.S.-parented
multinationals), U.S.-parented and foreign-parented multinationals are
taxed similarly on their U.S. and foreign operations.
Significant differences arise, however, when the multinational
attempts to move earnings upstream. A U.S. subsidiary can generally pay
dividends to its U.S. parent without any U.S. tax consequences.47 Similarly,
dividends paid by a foreign subsidiary to a foreign parent are generally not
subject to U.S. tax.48 But dividends paid by a foreign subsidiary to a U.S.
parent49 or by a U.S. subsidiary to a foreign parent50 are taxable. Although
In practice, well-advised U.S.-parented multinationals never pay U.S.
46 IRC §§951-965.
47 A 100% dividends-received deduction is allowed by IRC §243 for qualifying
dividends – essentially, dividends paid by other members of the same “affiliated
group.” For purposes of IRC §243, only domestic corporations qualify as
members of an affiliated group. See IRC §1504(b)(3) (excluding foreign
corporations). If the affiliated group elects to file consolidated returns, see IRC
§1501, intercompany dividends are eliminated in consolidation. See Treas. Reg.
48 Dividends paid by a foreign subsidiary to a foreign parent are generally
subject to U.S. income tax only to the extent those dividends are U.S. source. IRC
§§881(a)(1) & 882(b).
49 Dividends paid by a foreign subsidiary to a U.S. parent are generally not
eligible for the IRC §243 dividends-received deduction, see IRC §§243(a). In
addition, foreign subsidiaries generally may not file consolidated returns with
their U.S. parents. See IRC §1502(b)(3).
50 Dividends paid by U.S. subsidiaries generally constitute U.S.-source FDAP.
IRC §861(a)(2)(A). The tax on U.S.-source FDAP is imposed on gross income.
these disadvantages are formally symmetrical, as a practical matter the
problem is obviated for foreign-parented multinationals by the fact that
they can generally sell the stock of their U.S. subsidiaries and thereby
redeploy capital off-shore without incurring any U.S. tax.51 In addition,
income-stripping techniques are commonly employed to minimize U.S. tax
on distributions from U.S. subsidiaries to their foreign parents.52 It is
substantially more difficult for U.S.-parented multinationals to avoid U.S.
taxation when they repatriate capital back into the United States.53 A U.S.-
parented multinational that seeks to participate fully and flexibly in the
world economy is therefore often disadvantaged vis-à-vis its foreign-
parented competitors. In addition, as will be discussed below, U.S. tax
rules often create incentives to locate productive capacity outside the
United States, even for products ultimately intended for sale here. U.S.-
parented multinationals benefit significantly from such incentives, but then
face severe tax and book costs on repatriation; foreign-parented
multinationals face no such disadvantages.
A second difference arises when the parent pays dividends to its
shareholders. Dividends paid by a U.S.-parented multinational are taxable
The dividends-received deduction is therefore unavailable with respect to such
51 Gain from the sale of a U.S. subsidiary’s stock by a foreign parent is
generally foreign source. IRC §865(a), in which case it will not be subject to U.S.
tax unless it constitutes ECI, which is highly unlikely. Gain from the sale of stock
in a U.S. real property holding corporation is treated as per se ECI and therefore
taxable to the foreign parent. IRC §897(a)(1). A well-advised foreign
multinational, however, will generally be able to avoid having its U.S. subsidiaries
characterized as U.S. real property holding corporations by debt-leveraging the
subsidiaries’ real estate acquisitions and including other substantial business
assets in those subsidiaries.
See note 22, supra.
53 In general, gain from the sale of a subsidiary’s stock by a U.S. corporation is
taxable. To avoid such taxation temporarily, a U.S. parent may interpose a foreign
holding company between the parent and its operating foreign subsidiaries. The
foreign holding company can then sell the stock of one its subsidiaries without
incurring U.S. tax. The problem still remains how to get the resulting capital back
into the United States. It is certainly possible for the foreign holding company to
create a new U.S. subsidiary without immediate tax consequences; the cost,
however, is an unwieldy structure with multiple layers of built-in future tax
liability if and when the structure is ever unwound, which future tax liabilities
would probably have to be reported for financial purposes currently anyhow.
Simply dividending the capital back will result in immediate U.S. tax liability.
regardless of whether paid to U.S. or foreign shareholders,54 unless a
shareholder can claim the benefit of a treaty exemption. Dividends paid by
well-advised foreign-parented multinationals are only subject to U.S. tax if
paid to U.S. shareholders.55 Dividends paid by foreign-parented
multinationals to foreign shareholders are subject to U.S. tax only if the
multinational is poorly advised.56
How U.S. international tax rules create incentives to move
productive capacity offshore.—As has been noted, foreign operations are
generally effectively exempt from current U.S. taxation, regardless of
whether the parent corporation is U.S. or foreign. The implicit assumption
underlying this schema is that foreign operations will be taxed by the host
jurisdiction. In fact, however, profits from such foreign operations are often
either not taxed at all or taxed under rules that preserve the incentive,
implicit in U.S. rules, to move productive capacity offshore.
Assume, for example, that a U.S. bank seeks to reduce its overall tax
liabilities. Assume further that its credit card collection operations will cost
the same to operate in the United States or developing Country X. If
Country X is willing to exempt the bank from tax on those operations, the
bank can reduce its overall tax liabilities by relocating those operations to
Country X.57 Mechanically, it creates a wholly-owned foreign subsidiary in
Country X and then contracts with that subsidiary for credit card collection
services. The fees it pays for the subsidiary’s services reduce the parent’s
U.S. taxable income (and therefore its U.S. tax liability) while increasing
the subsidiary’s income in identical amounts. The foreign subsidiary’s
income, however, is not subject to current U.S. or Country X taxation.
Country X, in the meantime, attracts well-paying jobs, generating income
at the employee level itself potentially subject to tax.
54 Dividends paid by domestic parents to U.S. shareholders are includible in
gross income. IRC §61. Dividends paid by domestic parents to foreign
shareholders generally constitute U.S.-source FDAP. See note 50, supra.
See note 54, supra.
56 Dividends paid by foreign parents will be U.S. source, and therefore subject
to the IRC §§871(a) tax on gross U.S.-source FDAP, only if 25% or more of the
foreign parent’s gross income during a three-year testing period was ECI. A
foreign-parented multinational can ensure that none of its gross income constitutes
ECI by dropping any U.S. trade or business into a subsidiary; only poorly-advised
foreign multinationals will therefore face this problem.
57 Tax Holiday, http://en.wikipedia.org/wiki/Tax_holiday (“In developing
countries, governments sometimes reduce or eliminate corporate taxes for the
purpose of attracting Foreign Direct Investment or stimulating growth in selected
industries.”) (visited June 11, 2008).
can game the U.S. tax system in much the same way, however, by adopting
a water’s-edge value-added tax (“VAT”).58 A water’s-edge VAT is
refunded when goods are exported to another country, such as the United
States. Conversely, it is imposed in full on imports. VAT taxation is
therefore effectively based on the location of sales, not on the location of
productive capacity. Consider the choice faced by a multinational in
deciding where to locate capacity. If it locates that capacity in the United
States, it will be subject to U.S. income tax on (1) its production and sales
profits from U.S. sales and (2) production profits from foreign sales. It will
also be subject to VAT on both production and sales activities on its sales
into a country with a water’s-edge VAT. In other words, if it locates its
productive capacity in the United States, its export production activities
will be subject to two levels of tax – U.S. income tax and foreign VAT. By
contrast, if it locates its production capacity in that same foreign country, it
will be entirely exempt from VAT for product sold into the United States,
and such product will be subject to U.S. taxation only on the portion of the
company’s profits attributable to sales. Its export production activities will
not be taxed at all.
In light of the foregoing, it should come as no surprise that large
numbers of U.S. jobs have moved offshore in recent decades. Some have
moved because non-tax costs are lower elsewhere; the theory of
competitive advantage asserts that this is a good thing. But the uneven
playing field created by the interplay of U.S. tax rules and the tax rules of
competitor countries has probably played a significant role as well; the
theory of competitive advantage offers no defense whatever to any such
uneven playing field.
How multinationals minimize reported profits from U.S.
operations.—All multinationals, regardless of parentage, generally
undertake operations in different countries through separate subsidiaries. A
subsidiary in Country X, for example, might manufacture the
multinational’s product, while a subsidiary in the United States might be
responsible for sales of that same product in the United States. Or the
manufacturing subsidiary might be in the United States and the sales
subsidiary in Country X. In either event, the product has to get from one
subsidiary to the other before it can be sold. The pricing of the transfer
transaction itself then creates a further opportunity for the multinational to
reduce its U.S. tax liabilities.
Developed countries are less likely to grant such tax holidays. They
58 A value-added tax is a tax imposed on the value added by businesses at each
stage of production, regardless of profitability.
sold at retail for $50. If the manufacturing subsidiary is located in Country
X, which uses a water’s-edge VAT and/or low corporate income taxes, and
the sales subsidiary is located in the United States, the multinational will
cause widgets to be sold by its manufacturing subsidiary to its sales
subsidiary at a relatively high price, say $45 per widget. This leaves $25
per widget of profit in Country X, subject to low Country X income
taxation and no VAT (since the widgets are produced for export). If the
sales subsidiary incurs $5 per widget of sales costs, the pricing also leaves
no net profit in the U.S. sales subsidiary. The same technique works in the
other direction as well. If production occurs in the United States, the
multinational causes each widget to be sold to its Country X sales
subsidiary for $20 per widget, leaving no profit in the U.S. production
subsidiary. By manipulating intercompany pricing within the corporate
group, the multinational can move profits from one jurisdiction to another
relatively arbitrarily. If both production and sales occur in high-tax
countries, the multinational can even interpose a third subsidiary in a low-
tax jurisdiction to absorb the profit.
In theory, the IRS has the power to adjust all such prices to market
levels.59 Penalties are also imposed if a taxpayer is caught using
sufficiently egregious pricing. Unfortunately, these rules are not self-
enforcing. An unknown but presumably significant number of companies
use aggressive intercompany pricing to reduce their overall tax liabilities
and get away with doing so. In addition, U.S. transfer pricing regulations
permit multinationals, without risk, to use prices anywhere within an
“interquartile range.”60 When it is to their advantage to do so,
multinationals presumably use prices at the high end of the permitted
range; when their interests are to the contrary, they presumably use prices
at the low end of the permitted range. Either way, significant profits can be
moved offshore and out of the reach of current U.S. taxation.
In 2005, reported profits (receipts less deductions) of large foreign-
controlled domestic corporations comprised only 4.8% of total receipts of
Assume, for example, that a widget costs $20 to produce and can be
59 IRC §482 authorizes the IRS to restate the income of related businesses to
“clearly reflect the income” of such businesses. See, e.g., Michael C. Durst &
Robert E. Culbertson, Clearing Away the Sand: Retrospective Methods and
Prospective Documentation in Transfer Pricing Today, 57 TAX L. REV. 37
See Treas. Reg. §1.482-1(e)(2)(iii)(C) (defining “interquartile range as “the
range from the 25th to the 75th percentile of the results derived from the
uncontrolled comparables”); Treas. Reg. §1.482-1(e)(1) (“A taxpayer will not be
subject to adjustment if its results fall within such range (arm’s length range).”).
such corporations.61 By contrast, reported profits of large domestic
corporations that were not foreign-controlled comprised 8.8% of total
receipts.62 This suggests that foreign-parented multinationals were able to
use intercompany pricing to cut the reported profits of their U.S.
subsidiaries to about half those reported by comparable domestic-parented
corporations. Even the latter were probably significantly understated,
however, since domestic-parented multinationals engaged in international
trade undoubtedly used intercompany pricing techniques as well.
How U.S. tax and accounting systems penalize repatriation and
U.S. reinvestment of offshore profits earned by U.S.-parented
multinationals.—Dividends paid by foreign subsidiaries to their U.S.
parents are fully taxable.63 Even if a foreign subsidiary retains all its
earnings and pays no such dividends, however, a financial accounting
charge for the future tax must generally be reported currently.64 In such
event, the multinational gets the benefit of deferral, but must still take a
current hit to reported earnings.
There is an exception to this general rule, however, “if sufficient
evidence shows that the subsidiary has invested or will invest the
undistributed earnings indefinitely”65 – in other words, if the multinational
makes a credible commitment never to repatriate. In such event, the
multinational not only avoids having to pay current U.S. taxes on those
earnings, it also avoids having to report any such liability on the income
See Table 4.—“Large” Foreign-Controlled Domestic Corporations
Compared to Other “Large” Domestic Corporations: Selected Items and
Percentages, by Industrial Sector,
See note 49, supra. In general, such dividends are not eligible for the
temporary rate cut on dividends enacted as part of the Jobs and Growth Tax Relief
Reconciliation Act of 2003, P.L. 108-27, 117 Stat. 752, unless the corporation is
eligible for the benefits of a treaty identified by the Treasury as satisfactory. IRC
§1(h)11)(C)(i)(II). To qualify, a foreign corporation must be a resident within the
meaning of such term under the relevant treaty and must satisfy any other
requirements of that treaty, including the requirements under any applicable
limitation on benefits provision. Notice 2003-69, United States Income Tax
Treaties That Meet the Requirements of Section 1(h)(11)(C)(i)(II), IRB 2003-
42 (Oct. 20, 2003).
64 FASB Statement No. 109, Accounting for Income Taxes, left intact the
provisions of APB Opinion No. 23, Accounting for Income Taxes – Special Areas,
that relate to the accounting treatment for unremitted earnings in a foreign
consolidated subsidiary. See FSP FAS 109-2.
65 APB Opinion 23, para. 12.
Tax Year 2005, available at
statements by which its performance is judged. In other words, U.S.-
parented multinationals face strong tax/GAAP incentives to make credible
commitments never to repatriate foreign earnings.
Recognizing this to be a problem, the American Jobs Creation Act of
2004 provided a partial tax amnesty for extraordinary repatriations of
foreign earnings.66 Code Section 965 authorized a one-year67 85%
dividends-received deduction68 for repatriations in excess of taxpayer’s
recent average repatriations.69 To qualify, such dividends must have been
used to fund “worker hiring and training, infrastructure, research and
development, capital investments, or the financial stabilization of the
corporation for the purposes of job retention or creation” in the United
States.70 The Congressional conferees emphasized “that this is a temporary
economic stimulus measure, and that there is no intent to make this
measure permanent, or to ‘extend’ or enact it again in the future.”71 The
Financial Accounting Standards Board then allowed corporations to make
extraordinary repatriations in response to the amnesty offer without
jeopardizing ongoing commitments not to repatriate.72 Not surprisingly,
“[t]he temporary tax break turns out to have been particularly beneficial to
U.S. companies that had significant amount of permanently reinvested
earnings in subsidiaries located in low-tax countries.”73
B. A Motivation-Based Alternative
My premise, the reader will recall, is that our tax system can be made
both less distortive and more likely to elicit voluntary compliance if we
focus more closely on why taxpayers behave as they do. In particular, I
See generally Philip A. Stoffregen, Bringing it home: summary and analysis
of the repatriation provision of the American Jobs Creation Act, TAX EXEC. (Jan.-
67 IRC §965(f).
68 IRC §965(a)(1).
69 IRC §965(b)(2).
70 IRC §965(b)(4)(B).
71 House Conference Report No. 108-755 at 302, P.L. 108-357 (Oct. 7, 2004).
See FSP FAS 109-2.
Id. (“The list of the top 10 locations for subsidiaries that made qualifying
dividend repatriations looks remarkably similar to a list of top low-tax locations:
Singapore, Malaysia, the Netherlands, Hong Kong, Ireland, the Cayman Islands,
Luxembourg, Bermuda, Switzerland, and Sweden. By contrast, just 10 percent of
the dividends repatriated from Japan – the only industrialized country with a
statutory rate higher than the U.S. rate – qualified for the dividends received
assert that (1) if tax liability turns on factors as to which taxpayer is
relatively indifferent, the system will likely distort behavior while inviting
revenue loss, (2) economic distortion can be minimized by taxing whatever
it is that taxpayer otherwise seeks to maximize, and (3) if what taxpayer
seeks to maximize is a reported number, then compliance costs can be
minimized by taxing that reported number.
Part III.A’s recital of problems in the U.S. taxation of multinationals, I
suggest, strongly supports the proposition that a tax system based on
factors to which taxpayer is relatively indifferent does in fact distort
behavior while inviting revenue loss. Multinationals are relatively
indifferent to whether to undertake operations directly or through U.S. or
foreign subsidiaries. As predicted, we find that their corporate structures
are highly complex. Multinationals are becoming relatively indifferent to
whether their parent corporations are U.S. or foreign. As predicted, we find
major U.S. corporations attempting to expatriate; we do not, conversely,
find major foreign corporations attempting to move their headquarters to
the United States. All else being equal, multinationals are relatively
indifferent to the location of their productive capacity. As predicted, we
find multinationals locating productive capacity outside the United States,
often establishing U.S. presences only to further U.S. sales. Multinationals
are also relatively indifferent to intragroup pricing. As predicted, distorted
intragroup pricing remains a major compliance problem. Similarly, in the
face of significant tax incentives, multinationals become relatively
indifferent to the purely economic profitability of their investment
opportunities. As predicted, corporate tax sheltering persists. In sum, under
current U.S. international tax rules, multinationals’ liability turns
significantly on factors to which they are likely to be relatively indifferent.
We should therefore not be surprised that the behaviors of multinationals
are distorted, resulting in significant losses to the U.S. and world
economies and serious revenue losses to the U.S. tax system.
The solution is to make tax liability turn on factors about which
multinationals care deeply for non-tax reasons. If multinationals are
motivated to maximize reported consolidated book income, we should use
reported consolidated pre-U.S.-tax book income as the starting point for
our tax base. We will then need to determine the portion of that income
properly allocable to the United States. The solution, again, is to make sure
such allocation depends on factors about which multinationals care deeply
for non-tax reasons. All else being equal, multinationals are relatively
indifferent to the location of productive capacity. They do appear to care
deeply, however, about market share – that is, about sales. If so, we should
tax multinationals on reported consolidated pre-U.S.-tax book income,
computed on a unitary basis and allocated using a single-factor
apportionment formula based on sales, adjusted as I have already described
to reflect differing industry average profit-to-sales ratios.
If my assumptions are correct, a tax on reported consolidated pre-
U.S.-tax book income, computed on a unitary basis and allocated using a
single-factor apportionment formula based on adjusted sales, should
significantly reduce economic distortion and increase voluntary
compliance. (1) Such a tax would apply equally to U.S.- and foreign-
parented multinationals. It would therefore eliminate U.S. tax incentives to
expatriate. (2) Such a tax would apply without regard to where the
multinational’s productive capacity was located. It would therefore
eliminate U.S. tax incentives to offshore. (3) Such a tax would ignore
intragroup transactions completely. It would therefore make transfer
pricing irrelevant to U.S. tax liability and transfer pricing enforcement
unnecessary. (4) Such a tax would ignore capital flows. It would therefore
permanently eliminate U.S. tax inhibitions to the repatriation of profits.
Because it would eliminate deferral, it would also eliminate GAAP
inhibitions to such repatriation. (5) If multinationals are motivated to
maximize book income, then corporate tax shelter activity must be driven
by book-tax differences. If so, a tax based on book income would eliminate
U.S. corporate tax shelters. Indeed, even if GAAP itself is flexible,
manipulable, or idiosyncratic, multinationals should be motivated to
maximize reported consolidated pre-U.S.-tax book income – and therefore
their U.S. tax base.
Such a tax would have important collateral benefits as well. (1) The
corporate tax creates a double tax problem only if ultimately borne by
shareholders. A tax allocated in accordance with sales would more likely
be borne by consumers. If so, a tax on book income allocated in
accordance with sales could by itself at least partially resolve the
corporate-individual tax integration issue. (2) A tax based on book income
would sharply inhibit Enron-style overstatements of book income. (3) Such
a tax would also make intervention in U.S. tax policy less attractive to
multinationals. Their reduced participation in U.S. tax policy debates might
give Congress greater freedom to legislate based on policy, rather than
political pressure. (4) Studies suggest that different industries face vastly
different effective U.S. tax rates.74 A tax based on book income would
equalize effective rates across industries, thereby eliminating the economic
distortions of effective rate differentials. (5) Such a tax would likely
generate significant amounts of additional revenue, much of which would
likely be collected from foreign-parented multinationals now largely
exempt from U.S. tax. These revenues could be used to lower corporate
See, e.g., McIntyre & Nguyen, supra note 38.
rates or make other adjustments to federal tax or spending rules or
programs.75 (6) Such a tax would dramatically reduce the value added by
tax advice. For those of us who make a living giving such advice, this
would be profoundly unpleasant, at least in the short run; the rest of the
world, however, would be substantially better off. (7) Finally, such a tax
would reduce the advantage taxpayers often have on the corporate tax
battlefield. Instead of having to devote extensive resources to policing lines
inherently difficult to draw, the IRS could focus on the basics. I suspect it
would then be able to do a better job.
Congress often reaches for sticks whenever compliance is in question.
This paper suggests an alternative – a tax system designed around
taxpayers’ pre-existing motivations. Structural solutions may not solve all
distortion or compliance problems. Design principles that take taxpayer
motivation into account may, however, move us significantly toward a
more efficient, less costly system.
75 For example, some portion of such additional revenues might be sued to
fund individual AMT repeal.