Click for next page ( 10


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 9
INTERNATIONAI TECHNOLOGY AX POLICY AND VESTMENTS

OCR for page 9

OCR for page 9
1 The Taxation of Foreign Direct Investment: Operational and Policy Perspectivesi JOEL SLEMROD University of Michigan This chapter has three objectives. First, it provides a brief overview of the U.S. system of taxing the income from foreign direct investment, with particular emphasis on the changes introduced by the last major overhaul of the tax system in the Tax Reform Act of 1986 (TRACK. Second, it reviews the literature assess- ing the quantitative impact of tax systems on the volume and pattern of foreign direct investment. Finally, the chapter puts international tax rules into a policy perspective by drawing analogies to trade policy. U.S. TAXATION OF INCOME FROM FOREIGN DIRECT INVESTMENT Every country asserts the right to tax the income that is generated within its borders, including the income earned by foreign multinational corporations. Countries differ widely, however, in the tax rates they apply, the definitions of the tax base, and the special incentives they offer for investment. Nevertheless, the first and most important tax burden on foreign direct investment (FDI) is imposed by the government of the "host" country where the investment is located. Many countries, including the United States, Japan, and the United King- dom, also assert the right to tax the worldwide income of their residents, includ- ing resident corporations. As a general rule, the income of foreign subsidiaries is recognized only on repatriation of earnings through dividends, interest, or royalty iParts of this chapter draw on "The Impact of the Tax Reform Act of 1986 on Foreign Direct Investment to and from the United States" (Slemrod, l990b); "Comments on Tax Policy and the Activities of Multinational Corporations by James R. Hines, Jr." (Slemrod, 1997); and on "Free Trade Taxation and Protectionist Taxation" (Slemrod, 1995). 1 7

OCR for page 9
2 BORDERLINE CASE payments.2 To avoid the potentially onerous burden of two layers of taxation, countries that tax on a worldwide basis also offer a credit for income and with- holding taxes paid to foreign governments. The total credit available in any given year is usually limited to the home country's tax liability on the foreign-source income, although credits earned in excess of the limitation may often be carried forward or backward to offset excess limitations for other years. Several other countries, including France and the Netherlands, operate a "territorial" system of taxing their resident corporations, under which foreign-source business income is completely exempt from home country taxation.3 This would be the end of the story if the geographical location of income were not a matter of dispute. In fact, even if all the information necessary to ascertain the location of income were readily available, the conceptual basis for locating income is controversial (Ault and Bradford, 1990~. In reality, corpora- tions have no incentive to reveal fully all the information on which to base a determination of the geographical source of income. For any pattern of real in- vestment decisions, a multinational seeks to shift the apparent source of income out of high-tax countries into low-tax countries. This can be accomplished through, for example, the pricing of intercompany transfers of goods and intan- gible assets or borrowing through subsidiaries in high-tax countries. Note that this incentive applies regardless of whether the home country operates a territo- rial or a worldwide system of taxation. Much of the complexity of the taxation of foreign-source income arises from countries' efforts to defend their revenue base against the fungibility of income tax bases. Complex rules cover standards for acceptable transfer pricing, alloca- tion rules for interest expense and intangibles, and taxing certain types of income on an accrual basis. It is impossible to summarize concisely the various rules that countries employ to determine the location of income. In some countries the statutes are not as important as the outcomes of case-by-case negotiations be- tween representatives of the multinational corporation and government officials. In other cases the source rules are governed by bilateral tax treaties. What is clear, however, is that the de facto rules that govern the sourcing of income are at least as important for understanding the effective taxation of foreign direct in- vestment as tax rates, depreciation rules, and tax credits. The United States operates a worldwide system of taxation. Thus, both the domestic-source income and the foreign-source income of U.S. multinationals are subject to U.S. taxation. For the most part, the income of foreign subsidiaries4 2Unrepatriated earnings may be taxed currently in the case of certain passive or related party in- come or income earned in "excessive" retention of monies abroad as defined by the so-called subpart F rules of the Internal Revenue Code. 3By statute, Canada and Germany have a worldwide system of taxation. However, their tax treaties with the United States provide that repatriated dividends are generally subject to no further tax liability. 4The income of foreign branches of U.S. corporations is taxed as accrued. Partly for tax reasons, most foreign activity of U.S. corporations is carried out by subsidiaries rather than branches.

OCR for page 9
TAXATION OF FOREIGN DIRECT INVESTMENT 13 is not taxed as accrued but instead enters the tax base of the U.S. parent upon repatriation of dividends, at which point it is "grossed up" by the average tax rate paid to foreign governments. The grossed-up dividends, minus certain expenses of the multinational allocated to foreign-source income, enter into the taxable income of the parent. Foreign-source income of the parent also includes interest and royalty payments from subsidiaries and certain types of "passive" income on an accrual basis, plus the foreign-source income of foreign branch operations. In general, income taxes paid by foreign affiliates to foreign governments can be credited against U.S. tax liability. This credit is limited, however, to the U.S. tax liability on the foreign-source income, which is approximately equal to the U.S. statutory corporation tax rate multiplied by the net foreign-source in- come of the subsidiary. Multinationals whose foreign taxes exceed the limitation on credits are said to be in an excess credit position. These excess credits may be carried forward for five years or backward for two years without interest, to be used if and when the parent's potentially creditable taxes fall short of the limita- tion. If the potentially creditable taxes are less than the limit on credits to be taken in a given year, the corporation is said to be in an excess limitation position. Distinguishing excess credit and excess limitation positions is critically impor- tant for the financial behavior of a corporation. Changed Incentives for Foreign Direct Investment: The Tax Reform Act of 1986 Outward Investment The three most significant aspects of TRA86 for outward investment, in or- der of importance, were as follows: (1) the reduction in the statutory corporate rate from 46 to 34 percents and the resulting increase in the number of firms in an excess credit situation; (2) the change in rules governing the sourcing of income and the allocation of expenses, especially interest expense, between domestic- and foreign-source income; and (3) the tightening of the foreign tax credit limit- ing the averaging of different types of income. It is well known that the net effect of the tax system on the incentive to invest depends not only on the statutory rate but also on, among other things, the sched- ule of depreciation allowances, the rate and scope of investment tax credits, the source of financing, and the rate of inflation. The Tax Reform Act of 1986 elimi- nated the investment tax credits that previously applied to equipment and ma- chinery and made depreciation allowances somewhat less generous. Both of the steps tended to offset the reduction in the statutory rate of corporate income tax. Most analysts concluded that the net effect of these provisions was to increase 5The corporate rate has since been increased to 35 percent. To avoid confusion the rest of this chapter refers to a 34 percent, rather than a 35 percent, rate.

OCR for page 9
4 BORDERLINE CASE slightly the effective corporate-level tax on new domestic investment, an impor- tant alternative to FDI. An analysis of how these same changes affected the effective tax rate on FDI must proceed quite differently because, with certain exceptions, the foreign- source income of foreign subsidiaries enters the parent's tax base only to the extent that dividends are actually, or are deemed to be, repatriated. There is thus no calculation of foreign-source taxable income from which depreciation allow- ances are deducted and against which investment tax credits can be offset. The tax base to which the corporation statutory tax rate is applied is simply dividends received minus allocable deductions, grossed up by the average rate of foreign taxation. This is calculated using an earnings and profits measure of taxable income, which is not sensitive to legislated changes in the tax depreciation sched- ules used for domestically located assets,6 investment credits, and so forth. Thus, if one ignores the source-of-income rules discussed below, the corpo- rate tax changes of TRA86 reduced the statutory rate from 46 to 34 percent but did not broaden the tax base, resulting in an unambiguous reduction in the tax rate on income from FDI. If the taxes imposed by foreign governments remained unchanged,7 it follows that the amount of additional taxation imposed by the U.S. upon repatriation either stayed the same or declined. It stayed at zero for multina- tionals whose averages tax rate paid to foreign governments exceeds 46 percent. Any multinational subject to an average tax rate by foreign governments between 34 and 46 percent had formerly been paying taxes upon repatriation but would no longer be liable for any additional taxes under the new rate. For firms paying less than a 34 percent average tax rate to foreign governments, the tax due on repatria- tion would fall substantially, although not to zero.9 Seen from the perspective of 1986, the other important implication of the reduction of the U.S. statutory rate from 46 percent to 34 percent was that a much greater fraction of U.S. multinationals were likely to be in an excess credit 6The depreciation rules used in the calculation of earnings and profits do, however, change. For example, since 1980 the depreciation rules that apply to property used overseas have been made less generous. These schedules have tax implications because they affect the calculation of tax deemed paid by subsidiaries to foreign governments and the amount of foreign tax credit available for any given amount of dividends remitted. 7Since the passage of TRA86, many other countries have enacted tax reforms that share some of the corporate rate-reducing, base-broadening aspects of TRA86. To the extent that TRA86 caused these reforms (or increased their likelihood), the host country effective tax rate was influenced by U.S. tax reform. The analysis that follows holds constant the foreign tax system. The average tax rate paid to foreign governments is subject to a degree of control by the multina- tional via its repatriation policy. By repatriating income primarily from high-tax countries, the aver- age tax rate on its foreign-source income is high and less likely to attract additional U.S. tax liability. 9Hartman (1985) has argued that regardless of the excess credit status of the U.S. parent, the level of repatriation tax is irrelevant for the incentive to undertake FDI financed by earnings of the foreign subsidiary. This is because the repatriation tax reduces equally both the return to investment and the opportunity cost of investment (reduced dividends). This argument would not apply to the infusion of new equity capital from the parent. See Jun (1989) for a critique of this view.

OCR for page 9
TAXATION OF FOREIGN DIRECT INVESTMENT 15 situation because the average tax paid to foreign governments exceeded 34 per- cent.~ For a firm in excess credit status, every additional dollar paid in tax to a foreign government generates a foreign tax credit that cannot be used immedi- ately and has some value to the multinational only if the firm will be in an excess limitation position either in the next five years (the carryforward limit) or had been in an excess limitation position in the previous two years (the carryback limit). Thus, a U.S. multinational in an excess credit position is likely to be much more sensitive to differences in foreign effective tax rates than a firm in an excess limitation situation. This increases the relative attractiveness of invest- ment in a low-rate foreign country such as Ireland, compared to a high-tax coun- try such as Germany. From a 1997 perspective, however, this increase in excess credit status did not materialize, as Grubert et al. (1996) document. By 1992, about the same fraction of foreign-source income in the general basket was in excess credit as had been the case in 1984. They conclude that the primary reason for this was a decline in average foreign tax rates between 1983 and 1992; it was not attribut- able in large part to changes in the income or dividend repatriation patterns, or location decisions, of U.S.-based multinational corporations. A firm in excess credit status can reduce the present value of its tax burden to the extent that it can increase the limit on foreign tax credits. This has increased the importance of rules determining the source, for U.S. tax purposes, of world- wide income. Holding worldwide income constant, if a dollar of income is shifted from domestic source to foreign source, it increases the foreign tax credit limita- tion by one dollar and allows 34 cents more of foreign taxes to be credited imme- diately against U.S. tax liability. Only to the extent that foreign governments enforce the same source rules will there be an offsetting increase in foreign tax liability. One existing source rule that becomes more important applies to production for export. According to current regulations, between 40 and 50 percent of the income from domestic U.S. production of export goods can effectively be allo- cated to foreign-source income. For a multinational in an excess credit position, this reduces the effective tax rate on domestic investment for export by as much iGrubert and Multi (1987) quote U.S. Treasury Department estimates that the fraction of manufac- turing multinationals (weighted by worldwide income) in excess credit would increase from 40 to 69 percent. Goodspeed and Frisch (1989), using updated corporate tax return information, estimate that the fraction of foreign-source income subject to excess credits would rise from 50 to 78 percent, and from 32 to 82 percent in manufacturing. These calculations, however, consider only the change in statutory rate and do not consider changes in the allocation rules or the separate baskets, discussed later. In addition, neither analysis considers changes, perhaps induced by the U.S. reform, in other countries' tax rates. Perhaps most importantly, the analyses do not take into account any behavioral response of the multinationals. roof course, Hartman's argument implies that for investment financed by retained earnings, only the host country's tax rate matters even for firms in an excess limitation position, so no post-TRA86 increased sensitivity to host country tax rates should be observed.

OCR for page 9
16 BORDERLINE CASE as half. Thus, if a contemplated FDI is to produce goods for sale outside the United States, the alternative of domestic U.S. production has become relatively tax favored for those firms that have shifted into excess credit status, in spite of the base-broadening aspect of TRA86. This reasoning would not, however, ap- ply to FDI designed to reexport to the United States, because the alternative of domestic production for internal consumption does not benefit from the export source rule. Interest expenses of the U.S. parent corporation must be allocated to either U.S.-or foreign-source income. The general rule is to allocate on the basis of the book value of assets, so that interest expenses deductible from foreign-source income are equal to total interest payments multiplied by the fraction of world- wide assets represented by assets expected to generate foreign-source income. Although TRA86 did not significantly alter this allocation formula, it did add a "one-taxpayer" rule under which corporations that are members of an affiliated group are consolidated for the purpose of allocating interest expenses between U.S. and foreign sources. In the absence of this rule a multinational could load its debt into a U.S. subsidiary with no foreign-source income and allocate the interest expense entirely to U.S.-source income, thus maximizing foreign-source income and the limitation on foreign tax credits. With the one-taxpayer rule, a fraction of these interest payments has to be allocated to foreign-source income regardless of the legal structure of the multinational. For multinationals in an excess credit position that are forced to reallocate interest payments, this provision increased the average cost of capital of domestic or foreign investment to the extent debt finance is used. It also increased the marginal cost of foreign investment, because foreign investment increases the amount of interest payments that must be allocated abroad, which decreases for- eign-source income and therefore the amount of foreign taxes that are immedi- ately creditable.~3 This provision is obviously most important for multinationals with a high debt-to-capital ratio. TRA86 also changed the operation of the foreign tax credit by creating sepa- rate ("basket") limitations for certain categories of income. Foreign taxes im- posed on taxable income in a particular basket can offset only U.S. taxes due on that category of income. There are nine separate baskets, including passive in- come, high withholding tax interest, and financial services income. In some cases (e.g., passive income), the objective was to prevent fungible income from being earned in low tax rate foreign jurisdictions and thus to increase the amount of available foreign tax credits that could offset taxes paid on other income to foreign governments. In other cases (e.g., high withholding tax interest), the objective was to prevent multinationals (often banks) in an excess limit position i2The one-taxpayer rule already effectively applied to the allocation of expenses on research and development. i3This analysis presumes that the interest allocation rules of foreign governments have not changed.

OCR for page 9
TAXATION OF FOREIGN DIRECT INVESTMENT 17 from paying effectively high withholding taxes (which, due to the excess limit, could be credited immediately against U.S. tax liability) in return for favorable pretax terms of exchange (i.e., higher-than-otherwise pretax interest rates on loans). These objectives share the common thread of limiting the revenue loss to the United States that can arise from manipulation of the foreign tax credit mechanism. In general, the creation of separate foreign tax credit baskets increases the effective taxation of foreign-source income, because it makes it more difficult in certain cases to credit foreign income taxes against U.S. tax liability. In addition, the baskets can add significant complexity to the typical multinational's compli- ance procedure, and to this extent the provisions add a hidden tax burden to mul- tinational operation. Inward Investment Foreign corporations, and U.S. corporations controlled by a foreign corpora- tion, that are engaged in a trade or business in the United States are subject to taxation according to rules that are roughly comparable to those that apply to U.S. corporations. Thus, the reduction of the statutory rate, elimination of the invest- ment tax credit, and changes in depreciation schedules apply directly to foreign subsidiaries. The United States also imposes a "withholding" tax of 30 percent, modified by treaty to a much lower figure for many countries, on payments from corporations within the United States to foreign corporations. These withholding tax rates were not affected by TRA86. TRA86 did introduce a new branch profits tax, which imposes a 30 percent tax (often reduced by treaty) on the repatriated profits and certain interest pay- ments of a U.S. branch of a foreign corporation. This tax, which affects primarily financial institutions, was designed to equalize the tax treatment of foreign corpo- rations operating through a U.S. branch and those operating through a wholly owned domestic subsidiary. IMPACT ON FOREIGN DIRECT INVESTMENT This is not the place for a comprehensive review of the impact of taxes on foreign direct investment. Fortunately, the interested reader can refer to an up-to- date, careful, and comprehensive review of this topic by Hines (1997~. Hines concludes that "taxation exerts a significant effect on the magnitude and location of foreign direct investment," of a magnitude "that is generally consistent with a unit elasticity with respect to after-tax returns." (p. 415) One aspect that Hines' (1997) paper does not explicitly address is the extent to which the tax impact on foreign investment can be broken down into the im- pact of the host country's tax system versus that of the home country's tax sys- tem. My own reading of the evidence suggests that the potentially significant

OCR for page 9
18 BORDERLINE CASE impact of the former is fairly well established, but the impact of the latter is subject to much more doubt. However, it is the home country tax effect that is critical for understanding how changes in U.S. tax policy affect outbound invest- ment by U.S.-based multinational corporations. The evidence on this subject is mixed. Slemrod (199Ob) fails to uncover any evidence of an effect of the home country tax system on FDI into the United States, in particular whether the home country offers tax credits against tax pay- ments to the United States. Hines (1996), on the other hand, finds strong evi- dence that the across-state pattern of FDI is consistent with the influence of the home country tax system, in particular that the FDI into high-tax states is more likely to come from home countries that offer tax credits for taxes paid in the United States. Most recently, Grub ert and Multi (1996) analyze tax return data from 500 U.S. multinational corporations to investigate how taxation affects their location decisions. Although they find that host country effective tax rates have a significant effect on both the probability that a location will be chosen and the amount of capital invested there, they find no consistent effect of the parent's excess credit position, which should affect whether it is the tax rate of the host country or the home country that is effective at the margin. In the face of accumulating evidence of the impact of (at least host country) taxation on FDI, we should be cautious about the policy implications of this re- search, for two reasons. The first is that the welfare economics of international taxation is very much an unsettled issue (Slemrod, 1995~. The second reason is that the estimated response elasticities do not provide adequate information about the structural parameters of the choice problems faced by multinational corpora- tions (MNCs) and, therefore, are not well suited to gauging the impact of particu- lar policy changes. In other areas of economics we adopt parsimonious enough models so that the estimated parameters map directly into particular structural parameters. For example, under certain assumptions, labor supply elasticities correspond to char- acteristics of utility functions, and investment demand elasticities inform us about capital-labor substitutability and adjustment cost functions. Yet what does an FDI elasticity inform us about? To answer this question, we need to refer back to the prevailing theory of the nature of a multinational enterprise, coined the "eclec- tic theory" by Dunning (1985~. According to this theory the firm controls certain intangible assets that have a public good aspect to them in that they can be ex- ploited extensively with low marginal cost. The intangible asset is often thought of as information about product or process. To take full (i.e., global) advantage of this intangible the firm could produce its product domestically and export goods embodying the intangible asset, could license the intangible asset to foreign firms and receive a royalty, or could create controlled foreign subsidiaries to produce and distribute the product abroad. As Dunning puts it, FDI by firms of country A in country B is more likely if A's firms (1) possess ownership-specific advan

OCR for page 9
TAXATION OF FOREIGN DIRECT INVESTMENT 19 sages relative to B's firms in sourcing markets, (2) find it profitable to use these advantages themselves rather than leave them to B's firms, and (3) find it profit- able to utilize these ownership-specific advantages in B rather than A. Which choice is taken depends on an array of factors including transportation costs, tariffs, the expropriability of intangibles via licensing arrangements, export sub- sidies, and the tax rules that apply to foreign direct investment. Leamer's (1996) terminology is evocative. He contrasts the "mutual fund" model of MNCs, in which the firm allows investors access to foreign equity mar- kets that are otherwise inaccessible, to the idea of an MNC as a kind of "safe- deposit box" that keeps trade secrets from being used by foreign businesses and facilitates the deployment of intangible assets in foreign locations of production. The latter is closer to Dunning's eclectic theory. Within this conceptual framework it is undoubtedly true that, ceteris pari- bus, a lower effective tax rate on FDI makes it more attractive and thereby should increase its volume. However, what does the measured elasticity of FDI to its after-tax rate of return inform us about? The responsiveness of cross-border capi- tal movements? Not likely, because FDI can be financed largely by local borrow- ing or local sales of shares, so that increased FDI need not mean increased capital movement. Capital-labor substitutability? Also unlikely, since FDI tells us about who controls a given investment. Certainly the elasticity of who controls the shoe industry in the United States is different from the elasticity of the size of the U.S. shoe industry, regardless of ownership. The distinction makes a difference not only for how we interpret empirically estimated elasticities but also for the evaluation of policy. Consider the question of optimal tax policy toward inward FDI of a small, open economy. The standard result, discussed in Gordon (1986), is that such a country should, ignoring foreign tax credits offered by the capital-exporting country, impose no distortionary tax on the income earned by capital imports: taxes such as the value-added tax that do not provide a disincentive to investment are allowable. The model generating this standard result conceives of inward FDI as a capital investment, but if the spread of intangibles is involved rather than capital movement, what policy is optimal? If pure profits are at issue, the rate of a tax that is nondistortionary on physical capital investment may matter a great deal. The real world is messier than Leamer's two-tiered classification suggests because some MNCs do not fit neatly into either of Leamer's categories they are neither safe-deposit boxes nor mutual funds. Consider the Cook Island corpo- rations set up to accept the funds of New Zealand residents to be invested back into New Zealand the so-called Cook Island runaround. Many such firms are simply conduits designed to suck in taxable profits from other relatively high-tax jurisdictions, and as such are tax shell games. There are tax haven countries that are, for a fee, in the business of facilitating MNC tax avoidance. The presence of tax havens and the opportunities for avoidance and evasion they provide are a key

OCR for page 9
28 BORDERLINE CASE import duty. It is not appropriate to levy an export subsidy high enough to offset the import tariff. The analogy to the taxation of cross-border flows of income from investment is as follows. If the country in which the investment is located levies a tax, the country in which the investor is resident should offer no credit for these taxes. Furthermore, if a country wishes to levy an income tax on its residents, the base for such a tax should be income net of taxes levied by the source country (this argument was made in Richman, 1963~. It is as if the source country is levying a tax on imports of capital. The capital-exporting country, in its own interest, ought not to offset it with an export subsidy. In practice, no country, and certainly not the United States, follows this policy. Instead all capital-exporting countries offer some form of offset to taxes imposed by the source country, either in the form of a limited tax credit for for- eign taxes paid or by exempting foreign-source income from the taxation that is applied to domestic-source income. However, as discussed above, some coun- tries with imputation systems of integration can effectively impose an additional level of taxes on foreign direct investment. The U.S. policy of providing foreign tax credits has been characterized as "mercantilist" by Schmidt (1975) because it favors foreign investment at the ex- pense of the national interest. This claim is correct from a unilateral perspective because it is in the interest of one country to ensure that at the margin, the return to the country of all investments be equal, and the return to the country includes taxes paid to the country and not taxes paid to the host country. Thus, full taxa- tion of foreign investment income with deductibility of foreign taxes paid is uni- laterally appropriate but not consistent with global free trade in the presence of ubiquitous source-based taxes. Is it appropriate to characterize full taxation with deductibility as beggar-thy-neighbor behavior? In a sense it is, because the loss from this policy to the host country due to lost investment would exceed the gain to the capital-exporting nation. This usage of the term is somewhat strained, though, as the trade analogy makes clear: if all importing countries impose a tariff on a particular good, then it is beggar-thy-neighbor for an exporting country not to impose an export subsidy. In practice, all nations forgo the unilaterally optimal but beggar-thy-neigh- bor policy and, in so doing, avoid one route to double taxation of foreign invest- ment that would be inimical to global free trade. It is as if, faced with tariffs imposed by all nations importing a certain good, all the exporting nations im- posed exactly offsetting export subsidies. This would eliminate any trade distor- tions and thus be optimal from a global perspective, but it would not be in the exporting countries' interest because it would essentially be a transfer payment to a foreign government. Thus, it is inevitable that the division of revenues be- comes an important and contentious element of the current international tax re- gime. Bilateral tax treaties generally feature a reciprocity clause, requiring equal withholding levies for capital flows in both directions. This is designed to main

OCR for page 9
TAXATION OF FOREIGN DIRECT INVESTMENT 29 lain an equitable distribution of tax revenues in the presence of two-way capital flows. Whether it in fact achieves this goal depends also on the corporate tax rates and the details of the integration systems in place (Ault, 1992~. Thus, a basic feature of U.S. taxation of foreign-source income, the allevia- tion of double taxation, cannot be defended on the ground of unilateral free trade but must be viewed as part of a system that could be consistent with global free trade. However, other aspects of international tax policy can be made consistent with unilateral free trade, in the sense of avoiding beggar-thy-neighbor policies. For example, because the U.S. domestic saving and investment is large relative to world markets, it likely has some monopoly power, namely, the ability to affect world interest rates. If the United States is a net capital exporter it can take advantage of its power by taxing capital exports, thus driving up the rate of return on its net exports. If it is a net capital importer, it should tax capital imports, thus driving down the rate of return it must pay on such imports. The spirit of unilat- eral free trade dictates that these opportunities not be exploited. The prescriptions for free trade include the possible use of strategic policies directed at countries that do not themselves play by the rules of global free trade. Are there any useful analogies of this idea to international tax policy? Can one identify tax actions that are protectionist in nature and could justify "counter- vailing" tax action? Discrimination against foreign-owned firms would probably qualify. How- ever, most developed countries pledge nondiscrimination of company taxation through the existing network of bilateral tax treaties. A standard feature of these treaties is a clause stipulating that tax treatment of a domestic company will not depend on whether the company is domestic or foreign owned. Thus, two of the fundamental features of the international tax structure- avoidance of double taxation and nondiscrimination are broadly consistent with, but by no means ensure, global free trade. Continued adherence to these prin- ciples is clearly desirable, but leaves many aspects of a country's international tax regime unspecified. It also leaves unclear what criterion ought to be used to evaluate these aspects. Should the criterion be unilateral national income maxi- mization, with the presumption that adherence to double taxation relief and non- discrimination fulfills the country's obligations to free trade? Alternatively, should the United States seek to extend the range of policies that if adopted mul- tilaterally, can, by enhancing free trade in capital, potentially lead to increased national income at home and abroad? This is a critical question underlying all policy analysis of international tax policy, but it is usually left implicit. Aspects of Tax Policy: Free Trade or Protectionist? Border Protectionism and Ownership Protectionism One important difference between trade policy and tax policy is that whereas

OCR for page 9
30 BORDERLINE CASE trade policy operates at the border and is blind to corporate residency, tax policy can operate at the margin of corporate residency. For example, U.S. tariffs are imposed on all imported products, regardless of whether the good is produced abroad by a foreign-owned company or an affiliate of a U.S.-owned company. Domestically produced goods are not subject to tariffs and benefit or suffer, if the imported goods are inputs, from the higher domestic prices caused by tariffs, regardless of whether the producer is U.S. or foreign owned. Thus, trade policy raises the issue of what might be called "border protectionism." Income taxation, because it can impose differential taxation depending on corporate residence, may also involve another kind of protectionism that I term "ownership protectionism," although this is something of a misnomer to the ex- tent of international diversification of share ownership. Whether it does or does not depends on the structure of the income tax in place. If, for example, all countries scrupulously practiced nondiscrimination of business enterprises, lev- ied no withholding taxes, and operated territorial systems of taxation, any two corporations with the same real operations and results spread over the world would pay the same total tax, regardless of the residency of the parent corporations and even in the face of varying tax rates across countries. A French company and a U.S. company would pay the same total tax if both companies operated exclu- sively in the United States, exclusively in France, exclusively in Singapore, or in some combination of these and/or other countries. Differences can arise, though, because the United States taxes its resident multinationals on a worldwide basis and France taxes on a territorial basis. In this case there is a potential tax penalty placed on a U.S. multinational versus a French multinational that depends on the locational pattern of activity. There would be no substantial difference if the two multinationals operated exclusively in countries of similar tax rates such as France and the United States. The differ- ence arises to the extent that operations are located in a low-tax country. The U.S. parent company, but not the French parent, could be subject to a residual tax upon repatriation of income from its affiliate in the low-tax country. The appar- ent difference is also mitigated if the U.S. multinational operates not only in low- tax countries but also in foreign countries with average tax rates that exceed the U.S. average rate. In this case the U.S. system allows repatriated income from a low-tax country such as Ireland to be "mixed" with repatriated income from a high-tax country such as Germany, with the result that no net tax need be paid to the U.S. government. From the standpoint of global efficiency, there is no reason that the total corporate plus individual tax burden on the income of a multinational enterprise should depend on the parent company's country of incorporation. It is no more efficient than, in a domestic context, taxing corporations with names beginning with the letters A through K at one rate, while taxing at a higher rate those with names beginning with L through Z and not allowing name changes. If enacted, Lollapollooza Corporation could not compete with Kennebunkport Corporation

OCR for page 9
TAXATION OF FOREIGN DIRECT INVESTMENT 31 if they produced exactly the same products. If they produced slightly different products, Lollapollooza might survive, but on a smaller scale and with a dimin- ished variety of output. As Frisch (1990) has argued, one efficiency cost of this discrimination could be a reduced variety of products available to the world market. Whether higher corporate taxes translate into higher total taxes depends on the tax system in place. This would not occur if all countries adopted a pure residence-based tax, where residence refers to the residence of individuals and not corporations. Corporations subject to more tax would necessarily have share- holders who had a lower personal tax burden, so their cost of capital would be low enough to offset the higher corporation tax payments. Under the current international system of taxation, which is not a pure residence-based system and not perfectly integrated, this offset will not occur. Some have claimed that a tax penalty on U.S. resident multinational enter- prises would be especially harmful to the U.S. national interest. One argument is that legal residence generally is associated with "headquarters" activities, such as research and development, that are especially beneficial to U.S. economic perfor- mance. If these benefits to the country cannot be captured by the firms them- selves, there is an economic argument for subsidizing, and certainly not penaliz- ing, such "externality"-producing activities. A relatively high tax rate on the worldwide activities of U.S.-based multinationals would, over the long run, di- vert economic activity to other multinationals, reducing the amount of headquar- ters activities carried out in this country. This argument ignores the availability of alternative policies that are better targeted to address the externalities issue. Any specific activities, associated with the headquarters of multinational enterprises or not, that produce positive exter- nalities should be subsidized and, for the most part, already are. Research and development expenditures, for example, may be expensed rather than amortized for tax purposes and furthermore are eligible for an incremental tax credit. It may be that the current effective rate of subsidy is too low, but this argues for raising it. Yet the rate of tax on all U.S.-headquartered multinational enterprises, regard- less of their externality-producing activities, is too blunt an instrument for this purpose. Another line of argument is based on the empirical fact that U.S. companies tend to be owned primarily by U.S. citizens, whereas foreign companies tend to be owned by foreigners. In this circumstance it would be in the national interest to have policies that shift profits from foreign to U.S. companies. In oligopolistic markets, tax breaks can work just as well as export subsidies in shifting profits toward domestic firms. However, as Levinsohn and Slemrod (1993) show, in the simplest case this profit shifting is best achieved by subsidizing the output of domestic firms, not by distinguishing between domestically located and foreign- located production and allowing foreign taxes as a deduction. There may, though, be a case for favoring foreign operations if some sectors are perfectly competitive and some are oligopolistic and if trade policy, but not tax policy, can be sector

OCR for page 9
32 BORDERLINE CASE specific. In this case a tariff may have to be used, and location nonneutrality tolerated, so as to target the subsidy to the oligopolistic industries. In this case the targeting advantage of trade policy over tax policy overrides the production inef- ficiency caused by the tariff. Income Shifting and Tax Havens Another important difference between tariff policy and tax policy is that the basis for duties is the value of transaction, whereas the basis for income tax policy is a measure of income. Income is a considerably more slippery concept to de- fine, and the location of the income of an integrated global enterprise is a concep- tual nightmare. Ault and Bradford (1990) have gone so far as to argue that it is not meaningful. Given differences in tax rates across countries and the fact that no country has a pure residence-based system of taxing corporations, there are incentives to take advantage of the difficulty of locating income to reduce an enterprise's world- wide tax burden. A multinational operating in two countries in which the mar- ginal tax rate on a dollar of income is different would, ceteris paribus, prefer to shift income from the high-tax country to the low-tax country. Such shifting can be accomplished by the judicious setting of prices of transactions between corpo- rate affiliates or by judicious international financial policy such as borrowing in high-tax countries. In holding the location of real activity constant, a country gains when a dollar of taxable income is shifted into it, while the country from which it is shifted loses. The world is currently populated by a set of countries, known loosely as tax havens, that set low marginal tax rates and look the other way or even en- courage the inward shifting of taxable income. To stanch the outward flow of taxable income, countries that have relatively high tax rates must establish an enforcement structure to monitor transfer pricing, earnings stripping, and other methods of income shifting. Tax havens can be classified into two types. In one type, the country levies a very low tax rate on the income from manufacturing operations located in its jurisdiction. In the second type, the country offers a low tax on the income of corporations whose legal domicile is that country. One motivation behind be- coming the first type of tax haven is to attract real investment and economic activity into the country. This is not a primary motivation behind the second kind of tax haven. In that case the country is essentially offering its services, for a fee, to individuals and corporations pursuing tax avoidance and evasion. Even the first type of tax haven opportunistically gains from income shifting. Consider the example of Ireland, which offers a preferential tax rate of 10 percent on the income reported due from manufacturing operations in that country. Hav- ing established an affiliate in Ireland, a multinational enterprise has the incentive to shift taxable profits to that country from higher-tax countries. Thus, it is no

OCR for page 9
TAXATION OF FOREIGN DIRECT INVESTMENT 33 coincidence that such countries implement a low marginal effective tax rate on investment via a low statutory tax rate strategy as opposed to a strategy of a high statutory rate combined with generous investment tax credits and/or depreciation allowances. Although any particular low marginal effective tax rate can be ob- tained with the latter strategy, it would not make the country a magnet for income shifting, only for real activity. Local content rules are a useful analogy to tax havens in the domain of inter- national trade. Imagine that the United States imposes quantity restrictions on the import of steel from Japan and Korea. To enforce such restrictions, there must be a way to identify imports from an unrestricted country, such as Mexico, as having originated in Mexico rather than in Japan or Korea. This is usually accomplished by attempting to measure the "local content" of the imports from Mexico and requiring it to be above a prespecified level to be imported without restriction. These rules are similar to the anti-treaty shopping provisions of in- come tax treaties, which seek to limit the rerouting of income through tax havens to minimize tax payments. A country that for some compensation, collaborates with the restricted countries to evade the U.S. local content rules is acting simi- larly to a tax haven. I refer to this behavior as "predatory tax protectionism." It is predatory because it is clearly a zero-sum or a negative-sum game in which the tax haven's gains are offset by losses to the rest of the world. From a global perspective, the presence of tax havens is costly for at least two reasons. First, there are substantial resource costs expended by the tax col- lection agencies of the rest of the world to minimize inappropriate income shift- ing and substantial resources costs expended by the multinationals themselves to accomplish such shifting. Second, there are distortions in the kind of real activity that the first type of tax haven attracts (i.e., high-margin production such as phar- maceuticals and electronics that facilitate income shifting). Absent income shift- ing considerations, there is no economic reason why such activities should be located in Ireland or Puerto Rico, which can offer income shifting advantages to U.S. corporations. I have argued elsewhere (Slemrod, 1988) that the costs due to tax havens and income shifting are appropriately dealt with via a multilateral agreement restrict- ing statutory corporate tax rates to a small band and imposing sanctions on coun- tries that choose not to comply. Countries would be permitted to be magnets for real investment, but they would have to do so by offering investment tax credits rather than low statutory tax rates. A minimum statutory corporate tax rate is the approach suggested as a first step toward more corporate tax harmonization by the Ruding committee, the experts' committee of the European Commission charged with recommending what, if any, tax harmonization should be adopted in concert with the 1992 curtailment of barriers to free trade in goods and services; this suggestion was not, however, embraced by the European Community. On a unilateral basis, it is imperative that the U.S. international tax system be designed to counteract predatory tax protectionism. It is clear that U.S. rules in

OCR for page 9
34 BORDERLINE CASE this regard impose large costs on multinationals operating here (see Blumenthal and Slemrod, 1995), but it is also evident that the potential stakes involved in income shifting are large (Harris et al., 1993~. In a domestic context it does not generally make economic sense to push tax enforcement until the point where, at the margin, revenue gained equals cost, because the revenue gained does not represent a benefit to the country but instead is a transfer of resources. In the context of international income shifting, revenue gained does represent a benefit to the country, although not to the world as a whole, because it comes at the expense of a foreign treasury. Thus, it does make sense from a unilateral perspec- tive to push enforcement much harder in an income shifting context than in a domestic context. From a global perspective, as suggested above, there will be an inefficiently large expenditure on this kind of enforcement. Worldwide or Territorial System? Because it involves most of the issues discussed above, let me address the choice between taxing worldwide income, with a limited foreign tax credit, and taxing only income earned within the United States. To be precise, the territorial alternative I consider would tax passive or portfolio income on a worldwide basis but tax active business income on a territorial basis. Because either system affords relief from double taxation, either is generally consistent with free trade. However, a territorial system allows and, relative to a worldwide system, encourages host countries to attract capital investment by of- fering a low marginal effective tax rate. This is inimical to free trade because it implies that the hurdle rate will be lower in these countries than elsewhere. It allows lower rates because there is no residual tax imposed by the U.S. govern- ment. It encourages low rates because high rates are less likely to be offset by credits from the home country government once the United States gets out of the business of offering foreign tax credits. Only Japan and the United Kingdom will remain, and these countries have tax sparing treaties with many developing coun- tries essentially exempting foreign-source income earned in the treaty partners' countries from residual taxation by the home country. How does the choice look from a unilateral perspective? First of all, the territorial system is simpler to administer and comply with than the worldwide system, so a switch will in the long run save on collection costs. However, as Tillinghast (1991) has noted, an exemption system would still be complex both because passive and active income would have to be distinguished and because a territorial system would increase the potential gains from income shifting and therefore put pressure on the transfer pricing rules and other enforcement tools. Another important criterion is whether a worldwide system of taxation offers a better defense against predatory tax protectionism in the form of tax havens. Does the potential residual tax imposed upon either accrual or repatriation pro- vide an important backstop to our attempt to tax domestic-source income in the

OCR for page 9
TAXATION OF FOREIGN DIRECT INVESTMENT 35 same way that the corporation tax can be justified as a backstop to the objective of taxing capital and labor income in general? This important question remains open. In this regard it would be worthwhile to compare the success of countries such as France and the Netherlands that operate territorial systems. How do the two systems stack up with regard to ownership protectionism? There is no compelling reason for the U.S. government either to penalize or to subsidize U.S.-parented multinational enterprises versus those of other countries. Under the current system there is a relative tax penalty to U.S. multinationals whose foreign operations are predominantly in low-tax countries; the United States will impose a residual tax on repatriated earnings, a tax not owed by non- U.S. multinationals. Note, however, that a careful recent study (Grubert and Multi, 1993) con- cludes that the tax shortfall from switching to an exemption system would be small, $0.2 billion compared to $97.9 billion of foreign-source income. This is because many of the U.S. multinationals that operate in low-tax countries also operate in high-tax countries and thereby can set their repatriation strategy to avoid any residual tax to the U.S. government. It is also because some U.S. multinationals now repatriate a mix of high-taxed dividends with low-taxed roy- alty and interest income to avoid substantial residual tax. Under an exemption system, the dividend income would be exempt, but the royalty income would still face worldwide taxation subject to a foreign tax credit. These revenue estimates suggest that when all U.S. multinationals are con- sidered as a group, whatever ownership tax penalty currently exists is small and would not be altered significantly by the change to a territorial system. A cau- tionary note about the revenue estimates is in order, though, because they depend critically on an assumption of minimal behavioral response to the change in tax regime. This is unlikely however. For example, repatriations now labeled as royalties could be reclassified as dividends, thus avoiding U.S. tax. The extent of this kind of "relabeling elasticity" is difficult to forecast accurately. CONCLUSIONS Compatibility with free trade is not the only standard against which to judge an international tax system. Its implications for equity, within and across coun- tries and its consistency with domestic tax regimes are two other important crite- ria. Nevertheless, as national economies become more integrated and as barriers to trade in goods and services fall, the importance of international taxation for the efficient functioning of capital markets will become a central concern. Free trade in capital is achieved only if the hurdle rate for investment is equal regardless of the location of the real activity, the nationality of the corporation doing the investing, and the nationality of the ultimate owner of the equity in- come. The existing international tax regime could be consistent with, although certainly does not achieve, this result because countries of residence adopt some

OCR for page 9
36 BORDERLINE CASE form of relief from double taxation. As in the case of a set of export subsidies offsetting import tanffs, this system reduces trade distortions although it creates nettlesome issues of transfers between the importing and exporting countnes. The fact that a central feature of U.S. international tax policy double tax relief is best viewed as part of a multilateral understanding that supports free trade makes it problematic to evaluate the ancillary characteristics of that policy. In cases where they conflict, what criterion should be used to evaluate tax policy? Should it be unilateral national income maximization, with the presumption that adherence to double taxation relief and nondiscnm~nation fulfills the country's obligations to free trade, or should the United States seek to extend the range of policies that, if adopted multilaterally, can by enhancing free trade in capital- potentially lead to increased prosperity in the United States and abroad? Income shifting and tax havens are critical problems for an international tax system that relies on source-based taxation. It is important for the U.S. tax system to defend its revenues with policy and enforcement measures that are more stnn- gent than those applied to domestic compliance. The United States should also pursue multilateral means to harmonize corporate tax rates so as to reduce the incentives for income shifting and the reward to tax havens that practice "preda- tory tax protectionism." There is no reason, from either a global or a national perspective, that U.S.-parented multinational enterprises should pay more tax than a foreign-parented group with similar operations, but neither is there a compel- ling argument for tax breaks on the grounds of "ownership protectionism." Fi- nally, although both a territorial and a worldwide system of taxation could be consistent with global free trade, there are important differences between the two systems. The territorial system is simpler and more likely to avoid a penalty for U.S. ownership but also more likely to encourage divergent worldwide tax rates and therefore deviations from free trade. Perhaps the most critical question is whether a territorial system can be as effective as a worldwide system in defend- ing against predatory tax protectionism in the form of income shifting. REFERENCES Ault, H. 1992. "Corporate integration and tax treaties: Where do we go from here?" Tax Notes International. Ault, H., and D. Bradford. 1990. "Taxing international income: An analysis of the U.S. system and its economic premises." In Taxation in the Global Economy, A. Razin and J. Slemrod, eds. Chicago: University of Chicago Press and National Bureau of Economic Research. Blumenthal, M., and J. Slemrod. 1995. "The compliance cost of taxing foreign-source income: Its magnitude, determinants and policy implications." Pp. 37-53 in International Tax and Public Finance; Also appears in The Taxation of Multinational Corporations, J. Slemrod, ed. Boston: Kluwer Academic Publishers, 1996. Deardorff, A.V., and R.M. Stern. 1987. "Current issues in trade policy: An overview." In U.S. Trade Policies in a Changing World Economy, R.M. Stern, ed. Cambridge, Mass.: MIT Press. Diamond, Peter, and James Mirrlees. 1971. "Optimal taxation and public production. I: Production efficiency." American Economic Review 8-27.

OCR for page 9
TAXATION OF FOREIGN DIRECT INVESTMENT 37 Dunning, J.H. 1985. Multinational Enterprises, Economic Structure, and International Competitive- ness, J. H. Dunning, ed. Clichester: Wiley. Frisch, D.J. 1990. "The economics of international tax policy: Some old and new approaches." Tax Notes (April):581-591. Goodspeed, T.J., and D.J. Frisch. 1989. "U.S. Tax Policy and the Overseas Activities of U.S. Multi national Corporations: A Quantitative Assessment." Washington, D.C.: U.S Department of the Treasury, Office of Tax Analysis. Gordon, R. 1986. "Taxation of investment and saving in a world economy." American Economic Review 1086-1102. Grubert, H., and J. Mutti. 1987. "The impact of the tax reform act of 1986 on trade and capital flows." In Compendium of Tax Research 1987. Washington, D.C.: Office of Tax Analysis, U.S. Department of the Treasury. Grubert, H., and J. Mutti. 1995. "Taxing Multinationals in a World with Portfolio Flows and R&D: Is Capital Export Neutrality Obsolete?" International Tax and Public Finance 293-317. Grubert, H., and J. Mutti. 1996. "Do taxes influence where U.S. corporations invest?" paper pre- sented at the Conference of the Trans-Atlantic Public Economics Seminar, Amsterdam, May 29-31, revised August. Grubert, H., and J. Slemrod. Forthcoming. "Tax effects on investment and income shifting to Puerto Rico." Review of Economics and Statistics. Grubert, H., W.C. Randolph, and D.J. Rousslang. 1996. "Country and multinational company re- sponses to the tax reform act of 1986." National Tax Journal (September):341-358. Harris, D., R. Morck, J. Slemrod, and B. Yeung. 1993. "Income shifting in U.S. multinational corpo- rations." In Studies in International Taxation, A. Giovannini, R.G. Hubbard, and J. Slemrod, eds. Chicago: University of Chicago Press. Hartman, D.E. 1985. "Tax policy and foreign direct investment." Journal of Public Economics 107- 121. Hines, J.R., Jr. 1996. "Altered states: Taxes and the location of foreign direct investment in America." American Economic Review 1076-1094. Hines, J.R., Jr. 1997. "Tax policy and the activities of multinational corporations." In Fiscal Policy: Lessons from Economic Research, A. Auerbach, ed. Cambridge, Mass.: MIT Press. Jun. J. 1989. What Is the Marginal Source of Funds for Foreign Investment? Working paper no. 3064. Cambridge, Mass.: National Bureau of Economic Research. August. Krugman, P.R., and M. Obstfeld. 1991. International Economics: Theory and Policy, 2nd ed. New York: Harper Collins. Leamer, E.E. 1996. "What do we know about the impact of offshore investment on the U.S. economy?" In The Taxation of Multinational Corporations, J. Slemrod, ed. New York: Kluwer Academic Press. Levinsohn, J., and J. Slemrod. 1993. "Taxes, tariffs, and the global corporation." Journal of Public Economics (May):97-116. Richman, P. 1963. Taxation of Foreign Investment: An Economic Analysis. Baltimore: Johns Hopkins Press. Robinson, J. 1947. Essays in the Theory of Employment. Oxford: Blackwell. Schmidt, W.E. 1975. "U.S. capital export policy: Backdoor mercantilism." U.S. Taxation of Ameri- can Business Abroad. Washington, D.C.: American Enterprise Institute. Slemrod, J. 1988. "Effect of taxation with international capital mobility." Uneasy Compromise: Problems of a Hybrid Income Consumption Tax. Washington, D.C.: Brookings Institution. Slemrod, J. 1990a. "Tax principles in an international economy." In World Tax Reform, M. Boskin and C.E. McLure, Jr., eds. San Francisco: ICS Press. Slemrod, J. 1990b. "The impact of the tax reform act of 1986 on foreign direct investment to and from the United States." In Do Taxes Matter? The Impact of the Tax Reform Act of 1986, J. Slemrod ed. Cambridge, Mass.: MIT Press.

OCR for page 9
38 BORDERLINE CASE Slemrod, J. 1994. "A General Model of the Behavioral Response to Taxation." Mimeograph. Uni- versity of Michigan. Slemrod, J. 1995. "Free trade taxation and protectionist taxation." International Tax and Public Finance 471-489. Slemrod, J. 1997. "Comments on tax policy and the activities of multinational corporations by James R. Hines, Jr." In Fiscal Policy: Lessons from Economic Research, A. Auerbach, ed. Cam- bridge, Mass.: MIT Press Tillinghast, D.R. 1991. "International tax simplification." American Journal of Tax Policy 8(2):187- 247.