Table of Contents
Beware of geeks bearing gifts.
On June 7 Christopher Hanna, senior policy adviser for tax reform to the Finance Committee majority and a professor at Southern Methodist University's Dedman School of Law, gave an important opening address concerning the recent corporate tax integration work of the Finance Committee to a conference organized by the American Enterprise Institute.1 The efficiency case for corporate tax integration has been made many times over the last several decades, and the technical problems associated with various implementation proposals have also been explored.2 What made this talk particularly interesting was not new empirical data or breakthroughs in corporate integration system design, but rather Hanna's straightforward explanation of the political economy considerations underlying the sudden revival of enthusiasm within the Finance Committee majority for corporate tax integration generally, and an unusual form of corporate tax integration -- a dividends paid deduction -- in particular.
It is now apparent that policymakers' agenda underlying their pursuit of dividends paid corporate tax integration is not economic efficiency or aesthetic purity, but rather quick, dirty, and painless corporate tax reform of a sort that leaves the unincorporated sector unscathed and that gives large public corporations most of what they want -- lower effective corporate tax rates on their public financial statements, along with a new way to repatriate low-taxed offshore cash and redistribute the money to shareholders without a U.S. generally accepted accounting principles earnings hit to the corporation.3
More specifically, the expectations are that, in conjuction with the dividends paid deduction, the headline corporate tax rate would remain essentially unchanged and that a withholding tax on corporate dividends (and, as discussed below, possibly interest payments) would be levied on all holders at the same corporate tax rate. This combination of a dividends paid deduction and a withholding tax on dividend distributions would enable corporations to claim much lower effective tax rates for financial accounting purposes, even though their cash outlays do not change, because the withholding taxes that corporations pay to Treasury will be characterized for financial accounting purposes as tax liabilities of the investors in those companies, rather than of the corporations that pay the tax on the investors' behalf.
At the same time, and without any change in law on international taxation, the dividends paid deduction would enable U.S. multinational corporations to rid themselves of their excess liquidity held in offshore tax havens without incurring a tax charge for financial statement purposes, provided only that they in turn redistribute those amounts to their shareholders -- which is exactly what they want to do in any event. Stateless income planners thus would be able to repatriate their offshore cash without financial accounting costs to themselves.4 Shareholders, however, would not be so lucky. Their all-in tax cost on repatriated and redistributed earnings most likely would be higher than what reasonably could have been expected under broad-scale tax reform.
The political economy intuition is that the passthrough business sector, which to date has resolutely resisted business tax reform (under the justifiable suspicion that their tax preferences would be taken away to pay down the corporate tax rate), would be on board with narrow corporate tax reform that takes the form of a dividends paid deduction, because it would leave passthroughs' tax liabilities unchanged and would not systematically favor the corporate form of business organization. Further, the intuition is that the corporate community, which in years past has been at best diffident about corporate tax integration (in part on the ground that integration would force corporations to increase dividends), would enthusiastically support corporate tax integration that takes the form of a dividends paid deduction. They would do so because they have no better alternative in light of the pathological inability of Congress to address tax reform in a serious and comprehensive manner; because this one narrow change would open the floodgates to the repatriation of offshore permanently reinvested earnings without incremental financial accounting tax charges; and because corporations could present themselves to their principal audience, the consumers of financial accounting statements, as enjoying a sharp cut in their U.S. income tax costs.
In reality, a dividends paid form of corporate integration would create more problems than it solves. It would not necessarily create parity between corporations and passthrough forms of business organizations regarding business tax preferences and incentives. It also by its terms would not create parity between distributed and retained earnings or between debt and equity. By the time optional bolt-on modules were added to deal with those important issues, the dividends paid proposal would increasingly resemble a Rube Goldberg construction, dominated both in theory and in elegance of implementation by any of several clean-sheet comprehensive capital income tax reform ideas. The withholding tax that is central to the plan might violate U.S. tax treaties regarding dividends paid to foreign investors, and would certainly violate tax treaties if it is extended to interest payments (as some proponents wish).
Further, the standard economic efficiency case for dividend integration is much weaker than usually is asserted, after taking into account the new tax-induced distortions in behavior that such a system would introduce. Dividend integration is an awkward way of getting at the real efficiency issue, which is the taxation of capital income. Moreover, its underlying premise that one level of tax on business income necessarily is the correct model is largely unexamined and uncompelled by logic. And finally, by offering a quick and dirty form of corporate tax reform now, dividends paid integration would remove much of the urgency for genuine corporate tax reform that would lower the corporate tax rate and make the United States a more attractive environment for inbound foreign direct investment.
Yet Finance Committee Chair Orrin G. Hatch, R-Utah, is vigorously pushing ahead with the idea.5 The Finance Committee held three hearings during April and May on the subject of business tax reform generally and the virtues of corporate tax integration in particular. While recognizing the many past failures to implement corporate integration, Hatch was outspoken in his enthusiasm for the concept, both at the hearings and elsewhere.6 Further, several prominent academics testified favorably about corporate integration at the recent hearings, primarily for well-rehearsed economic efficiency reasons. Similarly, the 2015 report of the Finance Committee's Bipartisan Business Income Tax Working Group also recommended further work on the idea.7
Nonetheless, until the political economy agenda became clear, many academics (including me) had been puzzled by the surge of interest within the Finance Committee for an idea last seriously mooted roughly 15 years ago. Moreover, in light of the improbability of substantive tax legislation in the remaining months of the 114th Congress, those who did notice the development were disinclined to parse closely the evolution of the Finance Committee majority's thinking. Even Hatch noted at a May 24 hearing that "there's a graveyard near the White House filled with corporate integration proposals."8
The real political economy story has not been completely hidden. For example, professors Michael J. Graetz and Alvin C. Warren Jr. in May posted online a working paper, "Integration of Corporate and Shareholder Taxes" (Graetz and Warren 2016), that will appear later this year in National Tax Journal.9 In that paper, the authors add to traditional efficiency rationales for corporate tax integration the claims that a dividends paid deduction coupled with a withholding tax is responsive to cross-border income shifting concerns, earnings stripping, and inversions. In passing, they also mention that "a dividend deduction could have the effect of reducing effective corporate tax rates and thereby increasing a company's earnings per share" and "would also permit U.S. multinationals to repatriate foreign earnings to the United States free of any residual U.S. corporate tax when those earnings were distributed as dividends to shareholders."10 But these are not ancillary observations; they are the principal motivation for dividends paid integration.
The remainder of this report develops the claim that financial accounting for corporate income taxes and the tax-free repatriation of offshore earnings would be the principal immediate consequences of corporate tax integration in the form of a dividends paid deduction, and that those consequences in turn explain the interest shown by both the noncorporate and corporate sector in the idea. By understanding those consequences, observers will be better situated to understand the motivations of policymakers and taxpayers alike if and when this proposal advances, and to evaluate the larger implications of such a reform.
Section II puts dividends paid integration into context and hazards predictions about how the mechanics of the Finance Committee's proposal might operate. Section III briefly reviews the reasons that might explain the Finance Committee's sudden surge of interest in dividends paid integration, looking to Hanna's recent speech for guidance.
Section IV examines the underlying technical issues in more detail. Section IV.A considers some general arguments for corporate integration. Section IV.B addresses the critically important question of how the Finance Committee's dividends paid integration plan might address excess distributions -- distributions out of corporate earnings and profits that have not borne the 35 percent corporate income tax. (This problem often is couched as whether corporate tax preferences should be "passed through" to shareholders.) In the context of a dividends paid deduction in which all distributions out of E&P are taxable to shareholders at their respective tax rates, the question boils down to whether excess distributions should give rise to a corresponding corporate benefit, such as a net operating loss carryover or an immediate refund of any NOL generated by the application of the dividends paid deduction.
A shareholder often recognizes capital gains attributable to a corporation's retained earnings. Section IV.C addresses this coordination issue. Finally, dividends paid integration does not address the disparity between equity and debt investments in a corporation; without more, current law's debt bias would remain because tax-exempt and foreign investors effectively would still earn business net profits unburdened by tax at any level. Section IV.D addresses this. Section V concludes.
The report considers federal income tax consequences only. For convenience, I round off the top individual tax bracket to 40 percent, and I ignore section 1411's 3.8 percent tax on net investment income, on the theory that the underlying (albeit imperfectly implemented) idea is that all personal income above the relevant threshold should be subject either to it or to payroll or self-employment taxes.
To emphasize, the report's focus is on the actual consequences of a dividends paid deduction form of corporate tax integration, not who is or is not in on the joke. The report does not maintain that Hanna or any other academic who has expressed enthusiasm for corporate tax integration in the form of a dividends paid deduction has done so for any reason other than the usual economic efficiency reasons.
A. Prior Work
Academics and policymakers alike have explored the theme of corporate tax integration for many decades. The existing literature is voluminous, and it would serve no purpose to reprise it, particularly given the number of recent works adverted to in this report that have ably done so.
A classical corporate income tax's double taxation of corporate dividends and capital gains on corporate stock (discussed below) stands in contrast to a single level of taxation on most other forms of capital income (for example, interest, rents, royalties, and the net business income of unincorporated companies). Experts within and outside government have internalized this aesthetic discontinuity and apparent economic inefficiency to design alternative regimes that are said not to penalize investment in corporate stock, through different mechanisms to eliminate the double taxation of corporate earnings when distributed to shareholders.
Early corporate integration proposals were tied directly to the taxation of dividends, and possibly corporate capital gains. Dividend imputation credit systems are the most common example. A dividends paid deduction is another possible mechanism; when coupled with a shareholder withholding tax on dividends, it has essentially the same economics as a dividend imputation credit system. I am not familiar, however, with any jurisdiction that to date has implemented a dividends paid form of corporate integration. For convenience, these traditional responses can be lumped together as "dividend integration proposals."
Within the United States legal academy, the analyses of corporate integration prepared by the Treasury Department in 1992 (Treasury 1992) and the American Law Institute in 1993 (ALI 1993) are particularly important dividend integration studies.11 These publications have been kept alive and up-to-date particularly through the continuing work of Graetz and Warren, in Tax Notes and elsewhere.12 In December 2014 the Finance Committee's Republican staff released a long report on tax reform generally (SFC 2014 staff report), which includes a helpful chapter on corporate integration that summarizes the many earlier false starts in implementing dividend integration in the United States.13 A few months later, the Finance Committee published a bipartisan report of the Business Income Tax Working Group (2015 working group report), which also offers important insights into the thinking of the Finance Committee and its staff regarding corporate integration.14
Most recently, Graetz and Warren 2016 demonstrates again the economic equivalence of imputation credit integration to a dividends paid deduction coupled with withholding, updates the authors' earlier efficiency claims, and adds to them the arguments that a dividends paid form of corporate integration would mitigate the infirmities of current U.S. law's international tax system (inversions, income shifting, earnings stripping, and lockout). Graetz and Warren 2016 describes without qualification many of the design elements of the dividends paid integration proposal that the authors anticipate emerging from the Finance Committee staff in 2016. This report takes those authors at their word and assumes that their article accurately reflects the thinking of the Finance Committee staff.
Simplifying a bit, in recent years the focus within academia has shifted from dividend integration to the taxation of capital income more broadly. Business net profits, dividends, interest, capital gains and so on are simply different instances of capital income. The common idea behind these broader approaches is to begin with a general theory of capital income taxation (for example, that it is a bad idea, and therefore normal returns on investment should not be taxed at all, or alternatively that capital income should be taxed at flat rates lower than the maximum marginal rate on labor income) and to build out a system that taxes all instances of capital income similarly, regardless of labels, assigning the imposition of tax to the taxpayer from which it is most easily collected or measured. Dividend taxation as such therefore typically is not the principal focus of those efforts, although the resolution of double taxation concerns is one important byproduct.
Comprehensive mark-to-market solutions, business cash flow taxes, Treasury's comprehensive business income tax, my own work on the dual business enterprise income tax, and the recent proposal developed by Eric Toder and Alan D. Viard15 are all examples of this more comprehensive approach to capital income taxation -- although they adopt different starting points for the wisdom of taxing capital income at all, and different mechanisms for achieving their aims. These comprehensive capital income taxation proposals all strive to impose the same tax burden (which in some cases might be zero) on business net income, regardless of the form of business entity used, the mode of financing that entity, or the mode of realizing returns from that business activity (that is, as distributions or market gains). In doing so, they are exciting and offer the prospects of much larger efficiency gains than do dividend integration proposals.
Nonetheless, this report focuses on dividend integration ideas, because a dividends paid deduction best fits in this category. Current law's preferential tax rates applicable to corporate dividends and capital gains can be understood as easily implemented partial solutions to dividend double taxation, although each in turn has been criticized as poorly targeted -- the capital gain preference because it is not tied solely to gains on corporate stock, and the preferential tax rate on dividend income because it does not distinguish between corporate income that has borne corporate tax and income that was tax-exempt at the corporate level.
Every approach to capital income taxation, whether of comprehensive ambitions or of the dividend integration variety, will need to grapple with a recent article by Steven M. Rosenthal and Lydia Austin.16 That article concludes that only about 24 percent of the stock issued by U.S. corporations is held by individuals in taxable accounts and that this number has been in steep decline for many years.17 (In 1965, by contrast, taxable accounts held 84 percent of U.S. equities, and even at the time of Treasury's 1992 analysis, the figure was 52 percent.) These conclusions are important for revenue considerations and for measuring the deadweight losses of current law or the efficiency gains of any new system.
As the SFC 2014 staff report and the many contributions of Graetz and Warren have demonstrated, there are many different paths through which dividend integration can be achieved. For example, dividends and capital gain on stock could be tax-exempt, all corporate interest expense could be disallowed, and the corporate income tax could serve as the final tax on corporate income. (This is the basic intuition behind Treasury's 1992 comprehensive business income tax.) Alternatively, the corporation could be treated as a passthrough vehicle, although most analysts dismiss this solution as unworkable in a world of rapid share turnover. A mark-to-market system limited to corporate stock only, coupled with the elimination of corporate-level tax, could also be understood as a form of dividend integration; here the most frequently observed problem is the administrative issue of extending the system to include private firms.18
Outside the laboratory or the classroom, the most common form of full dividend integration today is an imputation credit system.19 (Many countries, including the United States, can be said to have some form of partial integration through preferential tax rates on dividend income or, alternatively, a dividends received deduction.) Nonetheless, full dividend integration is a rarer feature of tax systems within the OECD than might be expected from all the academic study of the issue: Australia's imputation system is usually proffered as the paradigmatic example of a well-implemented one.20 According to the SFC 2014 staff report, Canada, Chile, Mexico, and New Zealand are the only other members of the OECD to use full dividend integration through an imputation system.21 It appears that no OECD member country implements full dividend integration through a dividends paid deduction today, and as noted earlier, I am unaware of a large economy that has ever implemented this approach for corporations generally.
An imputation system is easily understood; the papers referenced earlier describe the design details. A corporation continues to pay corporate-level tax on its income, just as in a classical tax system. As is true today for the indirect foreign tax credit, however, when the corporation distributes a dividend, a shareholder includes in income both the net dividend received and the corporate-level tax attributable to the dividend (the gross-up, in foreign tax credit terminology); the shareholder then claims a credit on his tax return for the corporate tax imputed to him.
In most cases, an imputation system is designed to be plugged into individual-level progressive tax rate structures. Thus, if the corporate tax rate is lower than the top individual rate, a high-bracket investor would have to pay the difference to bring the total tax burden up to the individual's tax rate, while a low-bracket investor would get a refund of the overpayment. The underlying premise is that capital income should be taxed identically to labor income, at progressive rates. The dual business enterprise income tax and some other comprehensive capital income tax proposals take exception to this unstated premise.
Because an imputation system pushes up to shareholders the corporate-level taxes attributable to a dividend, most analysts agree that this approach requires a corporation to maintain a "taxes paid" or "franked income" account from which the imputed tax can be drawn. This in turn raises difficult questions for the treatment of distributions of corporate income not subject to tax. For example, should foreign-source income that is sheltered from home country tax by foreign tax credits, or by virtue of a territorial tax system, be taxed when distributed to shareholders, without the benefit of an offsetting imputation credit? Section IV.B returns to this question.
The imputation credit often is said to enable a country to tax foreign and tax-exempt investors by making imputation credits nonrefundable to them. This is true when compared with a dividend exclusion system, but it is not really a property of imputation systems as such. One can better explain the result by saying that the introduction of an imputation system is often the occasion for countries to rethink the tax status enjoyed by tax-exempt and foreign investors. That is, the United States today could impose on tax-exempt and foreign investors the same shareholder-level tax on dividends that it imposes on individual investors: One does not need a particular form of integration to do that. One does, however, as a practical matter need taxes to be collected at the source, that is, at the corporate level, whether phrased as corporate income taxes that are then imputed to investors, or as withholding taxes on distributions.
Dividend imputation systems were popular in Europe several decades ago but are no longer.22 Their disappearance frequently is ascribed to the jurisprudence of the Court Justice of the European Union, whose decisions as to the demands of European Union law on the refundability of cross-border dividends were incompatible with the separate taxing jurisdictions of each member state of the EU, but in fact the retreat began earlier. As Malcolm Gammie described in a contemporary paper, the United Kingdom vacillated for years between classical and imputation systems, based on the prevailing view of different governments as to whether corporate retained earnings or dividends should be encouraged, before adopting legislation in 1997 that began the process of unraveling its integration schemes.23 Moreover, under the U.K. system, every dividend was accompanied by a corporate tax payment (advance corporation tax); the ACT could be applied against the corporation's "mainstream" tax, but a U.K. multinational company with substantial foreign operations would find that its mainstream U.K. tax liability was much smaller than the ACT it shelled out with every dividend. The result was known as the "ACT mountain" problem. The United Kingdom eventually responded with a separate parallel system for "foreign income dividends," with its own complexities and distortions. U.K. ACT was refundable to tax-exempts, but the equivalent amount of foreign tax paid by a U.K. corporation was not, thereby making foreign income dividends less attractive to tax-exempt institutions than were ACT-franked dividends.
Finally, as I had occasion many years ago to see up close, some European imputation systems were subject to the depredations of "dividend washing" schemes, in which shares held by tax-disadvantaged holders (for example, foreign investors ineligible for imputation benefits) would at dividend season mysteriously migrate into the hands of tax-qualifying holders (for example, domestic taxable financial institutions) and then fly out again to the hands of their original owners. Readers convinced that these schemes are easily stopped through antiabuse rules underestimate the creativity and aggressiveness of some financial institutions -- as demonstrated by a May investigative report revealing that contemporary iterations on dividend-washing trades are costing Germany about $1 billion per year.24 My experience was that the usual commercial deal was to share the benefits of the trade 50-50 between the long-term beneficial owner and the dividend-washing institution.
U.S. commentators sometimes invoke the current Australian dividend imputation system to prove that such a regime is feasible and to imply that it is desirable.25 The Australian experience is probably inapposite, however, not simply because Australia is a much smaller economy than is the United States (about one-tenth the size, as it happens), but because its economy is dominated by natural resources rents and because it is and has always been a large-scale net capital importer:
- Foreign direct investment has played a critical role in Australia's economic development. Throughout our history, domestic investment opportunities have exceeded domestic saving. Foreign capital inflows, including FDI [foreign direct investment], have been an essential source of the new capital formation that drives long-term growth in productivity and real per capita incomes. Foreign investment accounts for around half of Australia's overall capital stock.26
These distinctions mean that in the design of its tax system, Australia can focus on taxing inbound foreign direct investment and not worry nearly so much about the double taxation that results when non-Australian-source income is distributed to domestic shareholders (because no Australian tax has been paid on that income).27
By contrast, the United States is simultaneously the world's largest locus for inbound foreign direct investment and the largest outbound foreign direct investor. As a result, a tax system in which foreign-source income derived from outbound foreign direct investment is subject to double taxation (at the foreign source and in the hands of shareholders) does not achieve complete corporate integration as that term might be colloquially understood.
There are no examples of major economies that today implement comprehensive dividend integration through a dividends paid deduction. To the contrary, the massive 1993 ALI study of corporate integration, whose principal author was Warren, rejected a dividends paid deduction form of dividend integration because it would have enabled corporations to pass through tax-favored income to shareholders.28 Even Graetz and Warren 2014 pointed in a different direction; the authors there recommended a lower statutory corporate tax rate combined with a partial dividend imputation system.
Since then, however, a new form of dividends paid integration has emerged, in which a dividends paid deduction would be accompanied by a universal dividend withholding tax imposed at the corporate income tax rate and collected by the corporation. As applied to fully taxable domestic earnings distributed to shareholders, an imputation credit system is identical in all respects, including cash flows, to a dividends paid deduction accompanied by a withholding tax on distributions.29 Graetz and Warren 2016, at Table 2, provides the calculations, but the point can be intuited by focusing on the economic equivalence of a "corporate income tax" out of which distributions are made, on one hand, and a "withholding tax on distributions," on the other. In each case the corporation lays out the same amount of cash, which cash expense translates into a reduction in shareholder-level tax. Moreover, the labels are irrelevant to the question of the incidence of the tax (that is, the party on whom the burden comes to rest), just as the half of employment tax withheld from an employee's paycheck and the half "paid" by the employer have the same incidence (in that case, the employee).
Given this functional equivalence, one might well ask why the Finance Committee has gone down the path of a dividends paid deduction rather than the better explored path of dividend imputation. It is odd that proponents invoke the example of Australia to justify the feasibility of the idea but then do not suggest following Australia's example (and the example of most other major economies that have implemented dividend integration at one point or another) by enacting an imputation credit. This is beyond puzzling -- to an old and cynical observer of Washington tax politics, it is strong evidence that the source of the enthusiasm must lie elsewhere.
And as summarized in Section I, that other attraction is not difficult to find. As Hatch and Graetz and Warren 2016 have implied, and as explicitly developed by Hanna in his talk, the game here is not integration as much as it is a nominal corporate tax rate reduction for financial accounting purposes, along with a new mechanism to enable the tax-free repatriation and redistribution to shareholders of the fruits of corporations' offshore stateless income planning. Corporate cash flows would not change relative to current law or to an imputation system, and all-in investor taxes are more likely to go up than down. Efficiency here is the beard, not the genuine object of the heart's desire.
Hanna's speech, Graetz and Warren 2016, and various news articles, all as informed by the many prior dividend integration studies referenced earlier, permit one to infer the rough contours of the Finance Committee's dividends paid deduction before its publication. Assume for purposes of the examples that follow that the top tax rate on personal income is 40 percent; doing so just keeps the arithmetic neater.
The core idea is simple. As noted earlier, Graetz and Warren 2016 has a helpful, comprehensive table. The corporate tax rate would remain 35 percent. A corporation paying a dividend out of taxable income would receive a tax deduction equal to the amount of that dividend. (Dividends paid out of tax-preferred income are discussed below.)
Corporate dividends would be fully taxable to individual stockholders at their respective tax rates; current law's preferential rate would be repealed. Tax-exempt institutions and foreign investors would be subject to tax on their dividend income at the corporate tax rate of 35 percent. Corporations would collect withholding tax on behalf of all stockholders out of every dividend at a flat 35 percent rate.
The withholding tax would be refundable to individual shareholders whose personal tax rates were lower than 35 percent. Top-bracket individual shareholders would owe an additional 5 percentage points of tax. Tax-exempt and foreign investors would not receive any refund (because the dividend tax rate applicable to them would be 35 percent), and as in other cases of withholding on U.S.-source portfolio income, foreign investors that have suffered the withholding tax would not be required to file U.S. tax returns.
Imagine a corporation that earns domestic taxable income of $100 and distributes all $100 as dividends to individual investors. The company would have a tax liability of zero after the dividends paid deduction. Investors would receive $65 after withholding taxes were subtracted from their dividends and would be eligible for $35 in tax credits on their returns. If this same arrangement instead were called an imputation credit system, the company would face a nominal corporate 35 percent tax rate and distribute $65 as a dividend; investors would gross up the $65 to $100 of income on their personal returns, along with a $35 imputation credit.
Net, for every $100 received, top-bracket individual investors would put $60 in their pockets after paying the incremental $5 in personal income tax on the $100 dividend. Lower-bracket shareholders would credit the overwithholding against other income tax liabilities or obtain a refund. Tax-exempt and foreign investors would pocket $65, just as if the corporation had paid corporate income tax of $65 and distributed its after-tax profits.
The dividends paid deduction and the withholding tax would presumably be coordinated across tax years through a "taxes paid account" or the like. That account would comprise prior years' corporate income tax payments since the effective date of the legislation. In my understanding, at least, the dividends paid deduction would be applied first against current-year taxable income; to the extent it was so used, it would reduce a company's corporate income tax liability for the year but increase, dollar for dollar, its withholding tax liability for the year. Any remaining dividends paid deduction would then be applied against the taxes paid account, such that $100 of dividends paid deduction would be credited against $35 of prior years' corporate income taxes in the taxes paid account. Those freed-up prior years' corporate income taxes would be credited against the corporation's current-year withholding tax obligations. The net effect would be equivalent to an NOL carryback when the freed-up taxes were automatically applied to current-year withholding obligations rather than refunded in cash.30
The expectation is that dividends paid integration would not change the definition of "dividend" for tax purposes. It thus would remain the case that corporations frequently will have more E&P than after-tax income. How will the dividends paid integration system operate in this case?
First, there appears to be a universal assumption that distributions out of E&P will be taxable to shareholders, and subject to withholding tax, regardless of whether a distribution comes out of after-tax income or untaxed E&P.
Second, distributions will apparently be treated as coming entirely out of taxable income to the extent thereof.31 Imagine that a corporation earns $1,000 in taxable income and $1,000 in tax-exempt income in a year, and it distributes $1,000 to shareholders. The corporation will apparently be permitted to deduct the entirety of the distribution against its taxable income, leaving it with no tax liability and $1,000 in E&P that have not borne any explicit corporate tax. This is a very taxpayer-friendly rule, and one that gives the greatest possible returns to the ultimate financial statement earnings game. In this example, the anticipation would be that the corporation's effective tax rate for financial statement purposes would be zero.
Dividend income received by a U.S. corporation from foreign subsidiaries out of existing permanently reinvested earnings would apparently be fully subject to corporate tax and eligible for direct and indirect FTCs, just as is true under current law.32 (It might well be the intent that this rule would apply only to pre-enactment earnings of foreign subsidiaries and that post-enactment earnings would be subject to a territorial tax regime, but this is guesswork on my part.) That taxable income could then be redistributed to shareholders free of corporate income tax, after taking account of the dividends paid deduction, although stockholders would suffer the 35 percent withholding tax, collected and paid over by the corporation, of course.
Finally, a critical question in the design of a dividends paid integration system is what should happen when a corporation makes "excess distributions" -- distributions out of E&P when those E&P are not themselves subject to corporate tax. Bonus depreciation is a simple example; the result of first-year bonus depreciation is to depress taxable income but not E&P. If taxable dividend distributions are made that do not come from after-tax income (that is, corporate E&P arising from taxable income that has borne corporate tax), whether in the current year or prior years, through the taxes-paid account, those dividends would still be taxable to shareholders. But the issue becomes, should there be some concomitant tax relief to the corporation that shells out the 35 percent withholding tax? (This is the dividends paid analogue to the issue of phantom imputation credits -- shareholder credits for corporate taxes not actually paid.) If the corporation obtains some sort of benefit, in effect its business tax preferences will have been passed through to passive shareholders. If it does not, shareholders and their corporations will be subject in effect to a flat 35 percent tax on pretax E&P, which destroys any parity with unincorporated business vehicles.
Section IV.B considers this nettlesome question in a little more detail. The leading academic proponents of dividend integration -- Graetz and Warren -- have been adamant that corporate business tax preferences should not generally pass through to shareholders, and one can find hints of similar views in some of the Finance Committee's work product to date. But this is not an issue specifically addressed by Hanna in his speech, and so when considering the benefits and costs of dividends paid integration, one must keep in mind that there is some uncertainty surrounding how this important question would be resolved.
When the dust settles, what changes would follow for cash tax liabilities from dividends paid integration that followed the summary above? As applied to distributions out of taxable income, almost none. A corporation would incur $1 in withholding tax obligation for every dollar in reduced corporate tax liability that the new deduction provided, so that corporate cash flows would be wholly unaffected. Looking for simplicity at statutory federal rates, top-bracket individuals in theory would enjoy a rate cut on distributed profits, from 48 percent33 to 40 percent. Whether this is an improvement over current law depends on the effective tax rate actually imposed on corporations today. If in this case it were 20 percent, which is perfectly plausible in some cases, top-bracket individual investors would actually be worse off.34
At first glance, tax-exempt and foreign investors seem worse off under a dividends paid integration scheme, but of course the integration system simply would substitute a formal shareholder withholding tax for current law's corporate tax, dollar for dollar -- at least for corporate taxable income. (As mentioned at the end of the preceding section, taxable income generally is smaller than pretax E&P, and distributions out of untaxed earnings raise very difficult issues.35) Therefore, ignoring the exotica of ultimate corporate tax incidence (which should not be affected by this largely optical change in law), nothing should change in economic or cash flow terms for tax-exempt or foreign investors regarding distributions out of corporate taxable income.
The shadow of untaxed corporate income hangs over this simple picture. As developed in more detail below, if corporate tax preferences are not passed through to shareholders, parity with unincorporated business organizations is not achieved, and shareholders essentially would be subject to a flat 35 percent tax on pretax E&P. This would be a substantial tax increase relative to current law.
There is also an important question whether the substitution of what is in form a 35 percent withholding tax imposed on shareholders for a 35 percent corporate income tax is consistent with the many tax treaties to which the United States is a party, under which the treaty withholding tax rate on dividends paid to qualifying foreign portfolio investors is far below 35 percent. In economic substance (and again abstracting from ultimate tax incidence questions, which should not be affected by this reformulation of liabilities), a foreign shareholder's tax burden would be unaffected by the adoption of a dividends paid integration system (at least for dividends paid out of after-tax income), but tax treaty issues often turn on formalisms such as this. This report leaves that interesting question for another day.
A. GAAP Accounting -- How Stupid Are We?
A dividends paid deduction coupled with a withholding tax would not increase corporate after-tax cash flows, and indeed might actually reduce them, if an effective corporate offset is not implemented for withholding taxes imposed on the distribution of corporate-level untaxed economic income. Nonetheless, the expectation is that financial accountants will look to the party against whom the tax nominally is assessed and treat reductions in "corporate income tax" as genuine, even when offset dollar for dollar by withholding taxes for which the corporation is the paying agent.36 The earnings-per-share pickup would be apparent in all cases in which a corporation distributes dividends out of taxable income. If the corporation used a dividend reinvestment plan (DRIP) in the manner discussed in Section IV, apparently it would declare that its corporate income tax liabilities had disappeared entirely. These reductions in GAAP effective tax rates are one of the two plausible explanations for policymakers' sudden fascination with dividends paid integration.
I accept that financial accounting will follow the nominal statutory incidence of the tax. What other practical choice is there? We cannot and do not expect financial accounting to search for the ultimate economic incidence of a tax; if it did, corporations would not be shown as bearing any tax at all. So a dividends paid form of dividend integration, combined with a dividend withholding tax at the corporate income tax rate and with a mandatory DRIP, would have the marvelous property of raising as much or more tax revenue than does current law, while slashing the statutory and financial accounting effective tax rate on corporations. What is more, the unincorporated sector would simply be bemused bystanders, completely unaffected by the relabeling of corporate taxes.
The real issue will be whether consumers of financial information will be so dim as to accept all this at face value, given that the total corporate tax outlay for corporate revenue will not have gone down and that for many shareholders, the all-in tax cost may go up. At best, shareholders net would receive the same all-in returns per dollar of pretax corporate income for actual or constructively distributed pretax profits that they do under current law. More plausibly, unless corporations were fully compensated for paying withholding tax on tax-preferred corporate income (a design issue discussed at length in Section IV), shareholders actually would be worse off regarding their all-in burdens. One would like to believe that this would lead them to be unhappy, in ways that are reflected in stock prices and therefore in the cost of capital, rather than pleased by the nominal corporate rate cut. In any event, it would be a fascinating natural experiment, in which rational expectations and efficient markets hypotheses could be put to the test.
The second reason why policymakers are promoting a dividends paid deduction might actually be the driver of this bus. We all know that U.S. multinational corporations have $2 trillion or thereabouts of offshore earnings, of which a substantial fraction is held in cash. Their dearest hope is to repatriate those earnings (or at a minimum the cash portion) at little or no U.S. tax cost, whether in cash tax or in financial accounting terms, and then use the cash to buy back stock or pay dividends. Their greatest fear is that large-scale tax reform would include a mandatory deemed repatriation of existing offshore earnings to clean up the past, which deemed repatriation would be subject to a high rate of tax. Since the deemed repatriation tax would be a one-time event relating entirely to past behavior, this fear is well grounded, because the tax would be not only fair but efficient. And corporations do not record a tax charge for financial statement purposes for their permanently reinvested earnings, even though they include those earnings on their consolidated financial statements, so any repatriation tax would be a pure hit to financial statement effective tax rates, without any offsetting earnings.
The great irony of the dividends paid integration proposal being mooted is that without any other change in law, the dividends paid system would give U.S. multinationals exactly what they want regarding their hoards of offshore earnings. Dividends from foreign subsidiaries tentatively would be fully subject to U.S. tax and eligible for the FTC, just as they are today. However, any U.S. corporate tax would disappear for GAAP purposes on redistribution of the proceeds to shareholders, by virtue of the dividends paid deduction.
Because corporations do not show a tax provision on their GAAP financial statements for permanently reinvested earnings, this ability to move offshore cash through the corporation to shareholders without corporate income tax avoids what today is seen as the cataclysmic possibility of a large hit to financial statement earnings if, as a component of large-scale tax reform, a one-time transition tax on corporations' offshore earnings were imposed. As a result, companies could put behind them the overhang of offshore permanently reinvested earnings that in reality are not so permanently reinvested at all.
Investors would suffer the 35 percent withholding tax. When compared with the probable shape of large-scale tax reform, this would likely be a bad trade for investors, at least as a cash tax matter. Most observers anticipate that large-scale tax reform would in fact include a mandatory corporate repatriation tax on offshore earnings, but even the most progressive policymakers would be thrilled if, when the dust settled, that tax were in the neighborhood of 15 percent. President Obama's recent "Framework for Business Tax Reform," for example, advocated a repatriation tax rate of 14 percent.37
Under current law, a hypothetical 15 percent corporate repatriation tax would be a final tax for tax-exempt holders and would offer a combined tax rate of 32 percent (plus any NII tax) to top-bracket individuals who receive distributions out of the repatriated cash, after taking account of the 20 percent personal income tax on dividends. If a mandatory transition tax in the neighborhood of 15 percent is the worst-case scenario for multinationals in large-scale tax reform, corporate managers who push for dividends paid integration as a means to repatriate and redistribute their offshore cash hoards are avoiding a financial accounting hit to their corporations at the cost of much higher all-in cash costs to investors. In a conjoining of odd bedfellows, tax policy wonks should be thrilled to see the government receiving the augmented revenue attributable to corporate managers' cynical behavior.
The dividends paid integration idea seems to contemplate that any repatriation of existing offshore earnings would be subject to corporate income tax and therefore fully eligible for the dividends paid deduction -- although, again, it is entirely possible that a distinction could be drawn between pre- and post-enactment earnings.38 Making any repatriation taxable accomplishes the objective of forcing corporations to redistribute the money to shareholders and also vitiates what otherwise would be an immediate and very large example of the problem of how to handle the distribution of exempt income.
The corporate integration literature is filled to the brim with conflicting opinions over whether the FTC should be treated in effect as taxes paid for purposes of distinguishing after-tax from tax-exempt E&P. My intuition is that foreign taxes will be treated as taxes paid. Otherwise, and assuming that dividends paid integration does not offer a fully offsetting benefit to corporations that distribute tax-exempt income, companies would need domestic taxable income to plug the hole left by the FTC; that is, to the extent corporations relied on the FTC, their taxes paid account would be that much smaller.39 In that case, if one further assumes that dividends from post-enactment foreign earnings will remain taxable (and eligible for FTCs), the system will reward the most successful stateless income tax planners; the lower the aggregate foreign tax burden on repatriated income, the smaller the hole that would need plugging with domestic income. Indeed, and contrary to Graetz and Warren 2016, one can imagine some corporations doubling down on stateless income planning to create a stream of purely taxable U.S. income on repatriation, which in turn would be used to fund the company's aggregate cash dividends to shareholders.40
This form of corporate integration takes all the pressure off reaching consensus on the design of a stable tax regime for outbound foreign direct investment. Like Br'er Rabbit, corporations would beg to be thrown into the briar patch of continued taxation of repatriated dividends to create the stream of nominally taxable income out of which then to pay deductible dividends. The worst threat a senator could make to them would be that their permanently reinvested income would not be taxable in the United States.
Graetz and Warren 2016, the SFC 2014 staff report, Hanna's recent talk, and testimony at Finance Committee hearings have all tacked from the usual efficiency arguments for dividend integration to repurpose the idea as a cure-all for the design flaws in the current U.S. tax system applicable to corporate outbound foreign direct investment. Thus Graetz and Warren 2016 argues that dividend integration will dampen U.S. corporations' enthusiasm for stateless income planning because they will need a taxes paid account from which to obtain the benefits of integration. (The authors' imputation and dividends paid proposals are generally premised on the idea that only dividends paid out of taxes paid (or in the current year, payable) accounts bring credits to an investor or a deduction to the corporation.) Graetz and Warren 2016 therefore sees the threat of double taxation of foreign earnings as a desideratum in that it would reduce income shifting out of the United States. In doing so, the authors invoke Australia as an exemplar, but as already noted, Australia does not have the same magnitude of outbound foreign direct investment as does the United States.
Graetz and Warren 2016 predicates its incentive effect analysis on the idea that the benefits of the dividends paid deduction would be available only to after-tax earnings. But at the risk of being tiresome, a dividends paid integration system that does not allow the passthrough of corporate business tax incentives to shareholders introduces a profound new disadvantage -- a deadweight loss -- to the corporate form of business organization.
It is difficult to see why real investments in real foreign jurisdictions should be systematically disfavored relative to domestic investments. Indeed, this is a form of what in other contexts is called "national neutrality," in which only taxes paid to the United States are treated as important to U.S. welfare concerns. National neutrality stands in sharp contrast to the environment in which corporations operate in the global economy, and it is inconsistent with both an FTC system and with a territorial approach to international taxation, which again is plainly the direction Congress is pointed.
A system that introduces a new and powerful disincentive to outbound foreign direct investment is inconsistent with most economists' views of the welfare interests of the United States. But if the arguments of Graetz and Warren 2016 persuade Congress, that would be the result -- foreign income subject to foreign income tax for real foreign activity and paid to the appropriate foreign country (it does happen, after all) would become fully double taxed. This is just substituting a new inefficiency for an old one.
There is an unstated premise at work here, which is that an FTC or territorial system is a corporate preference item, as opposed to a fundamental design decision limiting the scope of a country's taxation of net business income. It is odd that some scholars who find a territorial tax system logical and natural for corporate foreign direct investment should then express qualms about "passing through" that exempt income to the owners who have acted collectively through the agency of the corporation to earn that income.
In a new paper, Harry Grubert and Rosanne Altshuler consider how a dividends paid deduction would affect corporate behavior, under the assumptions that the new dividends paid form of integration would be married to a territorial tax system under which dividends from controlled foreign corporations would be exempt from tax, and that the benefits of the territorial system are not passed through to shareholders.41 (Their premise can be reconciled to this report's earlier suggestion, which hews to Hanna's speech, that permanently reinvested earnings would be subject to U.S. corporate income tax on repatriation, by imagining that different rules would be adopted for pre- and post-enactment foreign earnings.) Grubert and Altshuler find that about half of U.S. corporations today have sufficient domestic taxable earnings from which to fund dividends under an integration system and therefore would not be constrained by the implementation of the idea.42 Moreover, it is not clear that corporations would react by clawing back profits to the United States, or instead letting their existing structures continue and investing in a domestic cash cow to generate sufficient earnings in the tax paid account, if that is how dividends paid integration is in fact implemented.
Proponents also argue that dividend integration would solve outbound earnings stripping, but the only reason it would do so is if withholding tax were imposed on all interest paid to foreign lenders, with all the market chaos and deadweight losses noted in below in Section IV.D. Earnings stripping can be dealt with surgically, through toughening the existing earnings stripping rules. To argue that the nuclear option of dividend integration coupled with withholding taxes on both dividends and interest is desirable because it resolves this narrow problem is an implicit acknowledgment that the base case for dividend integration simply does not hold up and that new makeweight arguments must be thrown into the fray.
The same points can be made about the claim that dividend integration removes the pressure that some corporations feel to engage in inversion transactions. The idea is that lower corporate tax rates will make the United States sufficiently attractive as to remove the incentive to invert. But dividends paid integration does not offer lower corporate tax rates in any real sense, because the withholding tax outlays offset dollar for dollar any decrease in the nominal corporate income tax. Recent work by Kimberly A. Clausing demonstrates that corporations today book most of their offshore income in havens with tax rates under 5 percent,43 dividends paid integration achieves this (at least as a cash tax matter) only if one imagines the proposal coupled with a territorial tax system, and at the same time some form of corporate compensation for the withholding tax costs of distributing exempt income. As in the case of earnings stripping and the debt bias, surgical revisions to the relevant code provisions that directly target inversions (already drafted, as it happens, by House Ways and Means Committee member Lloyd Doggett, D-Texas, in his perennial tax bill on the subject) can address the issue without inducing major new tax-driven behavioral distortions.
A. Efficiency Arguments for Integration
It is worth considering just how incomplete or downright threadbare the standard efficiency case is for dividend integration, however implemented. Dividend integration is an awkward way of getting at the real efficiency issue, which is the taxation of capital income. That requires a more aggressive clean-sheet approach. Ironically, the actual implementation of some of these more conceptually radical ideas would probably be much simpler than the dividends paid integration scheme under consideration.44
The intuitive appeal of eliminating the double taxation of corporations is obvious. Indeed, it is so intuitively appealing that in the view of the Finance Committee's Republican staff, it goes almost without saying: "The difficult decision is not whether business income should be subject to more than one level of tax -- it should not -- but whether the business income should be taxed at the entity level or at the owner level."45 But this plainly cannot be a universal truth. Two taxes of $50 each total the same as one tax of $100, and if there are other reasons to prefer the first arrangement -- as we do with employment taxes, for example -- that approach cannot be dismissed simply by saying that one level of tax is naturally superior.
The dual business enterprise income tax, for example, deliberately taxes business enterprises on their returns to rents, and investors on normal returns, because in each case that is the most robust place to assign the taxation of that category of capital income. David M. Schizer and David Gamage have further developed this theme that two well-targeted taxes are often better than one larger tax; Schizer's paper in particular is a direct challenge to the Republican staff's premise about the design of a corporate income tax.46
One meaning of the phrase "double taxation" or "two levels of tax" is that the tax burden on distributed dividends today is simply too high, but, of course, that requires a comparison to some benchmark rate. What is often meant is that the total tax burden on corporate profits distributed as dividends is greater than if the same revenue was earned through an unincorporated business vehicle, but that in turn simply begs the question whether dividend taxes are too high, or instead whether unincorporated entities should all be treated as associations taxable as corporations.
Assume, nonetheless, that if the aggregate tax burden on distributed corporate income today is greater than that enjoyed by an investor in an unincorporated business entity, that difference is evidence that the tax burden on distributed dividends is too high. The predicate to this claim is largely empirical. It is not difficult to imagine contrary cases. For convenience we can assume that dividends held in taxable accounts of individuals are taxed at a flat rate of 20 percent because individual owners of U.S. equities who hold those equities in taxable accounts are disproportionately at the very top of the income distribution. It is also the case that most business tax preferences are available to unincorporated firms as well as corporations. But international income is an exception;47 for a great many reasons, that is an attribute of large corporations in particular, and under current law, those companies can distribute low-taxed foreign income to shareholders today through techniques as simple as borrowing in the United States.48 We know that many of the largest U.S. corporations enjoy foreign, and even worldwide, effective tax rates in single digits.49
Consider the case of Apple Inc., which in recent years has funded dividend distributions to shareholders by borrowing in the United States while allowing its very-low-taxed foreign income to compound offshore. In this way, Apple achieves a de facto tax-free repatriation of its foreign earnings, and the all-in tax cost to individual shareholders holding Apple stock in taxable accounts is only a little higher than the individual dividends received tax rate of 20 percent.50 From a tax policy perspective, eliminating this gambit is a good idea, but from a political economy point of view, one wonders whether members of the Finance Committee fully appreciate that as applied here, dividends paid integration would increase the overall tax charge on Apple's constructive distributions of its offshore income.
In the best case, then, dividend integration would directly benefit only individuals holding stock in taxable accounts (about one-quarter of all stock held by investors), and even they might be surprised to find that the effective tax rate on dividends attributable to a corporation's earnings on outbound foreign direct investment have increased. As Rosenthal pointed out in his May testimony before the Finance Committee:
- Because of the trends in stock ownership, and the reduction in tax rates, corporate earnings now face a very low effective tax rate at the shareholder level. Three-fourths goes untaxed, by our estimate, and much of the rest faces low rates. Further, the tax on any gain can be deferred or eliminated if the stock is held until death or donated to charity. So, the United States effectively tries to collect the bulk of the tax on corporate earnings at the corporate level.
. . . .
The United States could allow corporations to deduct dividends paid to shareholders. That would reduce the taxable income of corporations and increase that of shareholders. By our calculations, however, only about a quarter of dividends are paid to taxable accounts. So, the shift might generate relatively little revenue. To keep reform revenue neutral, Congress would need to substantially increase the tax rate on dividends and capital gains -- perhaps both to individuals and tax-exempt accounts and institutions.
At the same time, and depending in particular on how corporate tax preferences are handled, it might operate more systematically to raise taxes on tax-exempt and foreign investors. As the next section in this report develops, without some form of relief for corporate tax preferences, dividends paid integration would turn today's corporate income tax into a flat 35 percent income tax on an approximation of economic income (pretax E&P). Given that effective corporate tax rates are generally far below this level, tax-exempt and foreign investors would find their all-in tax burdens to have increased substantially.
Proponents of dividends paid integration are quick to claim that it reduces the deadweight losses inherent in current law's bias against dividends (which bias, to emphasize, is relevant principally to the extent that top-bracket individual investors are the marginal investors in corporate equities), but they typically do not consider the deadweight losses that would result from the clientele effects of their proposal (for example, portfolio realignments to avoid the tax or disinvestment in retirement accounts and a concomitant increase in current consumption) or from dividend washing schemes. Dividends paid integration by itself does not address existing law's bias in favor of debt financing. See Section IV.D., below.
It might be the case that the taxable accounts of U.S. individuals are the marginal investors in U.S. corporate equity and therefore drive prices, but I am unaware of empirical work that reflects the Rosenthal and Austin parsing of the ownership data. Moreover, if this were true, it is unclear why conveying a windfall to existing taxable investors (through higher equity prices following the implementation of dividend integration) is desirable, and none of the plans currently being mooted appear to be limited to new equity.
Forty years ago, U.S. corporations faced higher statutory rates than today, derived most of their income from domestic sources, paid higher effective rates, and were overwhelmingly owned by U.S. taxable accounts. Taxable investors in turn faced tax rates as high as 70 percent on dividend income. In that environment, dividend integration had a persuasive story to tell. Today, by contrast, many major U.S. corporations enjoy very low effective tax rates, the share of U.S. corporate equities held by U.S. taxable accounts has declined precipitously, significant dislocations would follow from extending tax liability on equities to currently tax-exempt investors, and windfalls would inefficiently be conveyed to taxable accounts. The case for dividend integration has become attenuated for these reasons.
Finally, quick and dirty tax reform is not harmless, even if tax revenues are unaffected. If the result is to remove pressure to work to genuine corporate reform, there are real welfare losses from this gambit. The U.S. statutory corporate tax rate is too high, the tax base too narrow, and the returns to different forms of investment and modes of financing hopelessly dispersed. Too much attention has been paid to the demands of U.S. multinationals interested in the taxation of their foreign direct investments, and more particularly in preserving the fruits of their stateless income labors, while too little attention has been paid to the United States itself as an environment in which to conduct business in the corporate form.51 Viewed from this perspective, dividends paid integration is a frolic and detour -- a distraction from what should be the challenging work of the Finance Committee in forging consensus, rather than trying to please everyone.
A perennial puzzler in dividend integration design is how to handle excess distributions -- distributions out of untaxed earnings. If the outline of terms presented in Section II.D is correct, shareholders will face the same withholding tax in this case as in the case of distributions out of taxable income, and the issue therefore becomes the handling of the corporation's dividends paid deduction when it exceeds taxable income. For the sake of consistency with prior work, however, this section sometimes describes the issue in the language used in the dividend imputation credit literature as the problem of passing through corporate preferences to shareholders.
Observers view the analogous question of shareholder credits for phantom corporate taxes as a very important problem for imputation credit integration regimes because it is asserted that the system would not produce the intended results if shareholders received credits for taxes not paid by the distributing corporation -- although, as developed below, the case can be made that if parity with unincorporated firms is the principal agenda, the concern is misplaced.52 In one sense, a dividends paid deduction makes the issue more tractable because, in contrast to a (potentially refundable) shareholder credit, a corporate deduction is valuable only to the extent the corporation has taxable income.53 Nonetheless, Graetz and Warren 2016 explicitly asserts that under the dividends paid system, a dividends paid deduction should never give rise to an NOL -- as would be the case if a corporation distributed more than its taxable income and the deduction were not explicitly constrained in some fashion.54 The issue therefore requires some closer examination.
Even without recourse to deep policy, full deductibility (and the concomitant possibility of an NOL) does achieve a useful averaging function. Imagine that a corporation in year 1 earns $1,000 in taxable income and $1,000 in untaxed income and pays out $2,000 as a dividend. In year 2 it earns $1,000 in taxable income and pays no dividend. An NOL carryover from year 1 to year 2 would put the corporation in the same position as if it had earned the year 1 and year 2 taxable income in the same year, just as an NOL deduction does today for corporate tax liabilities. This on its face seems more appropriate than allowing the accidents of annual accounting to drive corporate tax rates over longer periods of time.
One argument for not passing through untaxed corporate income in dividends paid integration (again, more accurately, offering the corporation an offsetting benefit) is that corporate tax preferences are meant to motivate managers, not owners. But that argument is belied by the fact that most incentives are available to both corporate and passthrough vehicles.55 The analysis might be different if there were corporate tax preferences designed to more closely align the interests of owners and managers by overcoming the classic agency problem in corporate governance, but I do not know of a preference that might fit that description.
The more general claim in the dividend integration literature is that integration systems should not pass through any corporate tax preferences to individual shareholders without some explicit congressional direction to do so. I would turn this on its head and suggest that a more useful heuristic here is that if Congress thinks a business tax preference is worth having, it is worth passing through -- at least in the absence of some specific reason not to do so.
If the goal is parity with unincorporated forms of business enterprises, a general prohibition on passing through corporate tax preferences is a very peculiar starting place. With the principal exception of the indirect FTC, corporate tax preferences are business tax incentives, freely available to unincorporated and corporate enterprises alike. A world in which integration leads to the recapture of corporate tax incentives when the fruits of those incentives are distributed to shareholders drives an enormous tax wedge between corporate and unincorporated forms of business organization, probably dwarfing in importance the relatively modest tax rate differential imposed under integration, on the one hand, and current law, on the other, on dividends distributed to the taxable accounts of individuals.56
ALI 1993 agonized over the issue.57 It suggested that one argument for not passing through corporate preferences to shareholders is that many of those preferences are designed to mitigate the high statutory tax burden on distributed corporate income. It also suggested that by not passing through some preferences (taken in the broadest sense), such as the ability to borrow nonrecourse against appreciated real estate without triggering a realization event, the resulting system would undo some of the mischief caused by the realization doctrine. But again, business tax preferences are freely available to unincorporated businesses. Those businesses now earn roughly half of all domestic business income, and they are heavy users of nonrecourse real estate financing. What is more, one might rationally assume that poorly targeted business tax preferences would disappear to pay for any business tax overhaul. In light of the rapid growth in the domestic unincorporated business sector over the last few decades, it makes no sense to purport to mitigate the consequences of double taxation through a dividends paid deduction while simultaneously raising the effective tax rate on distributed income by disallowing the passthrough of corporate preferences.
In the end, ALI 1993 left room for some unspecified passthroughs of corporate preferences to shareholders.58 It also adopted "stacking" rules designed to mitigate the problem by treating all distributions as coming out of tax-paid income first. That study was firm, however, that the returns to foreign direct investment should be fully taxed to shareholders when ultimately distributed to them, without the benefit of the FTC (or presumably, the benefits of a dividend exemption territorial tax system).
Corporate tax-preferred income is wholly tax-free to both tax-exempt and foreign investors today when realized through capital gains, and dividend distributions out of tax-preferred income are taxed today to those investors only to the extent of (treaty-reduced) withholding tax on dividends received from foreign investors. Top-bracket individual investors today face a 20 percent dividend or capital gains tax rate on corporate-level untaxed profits. Imposing a 35 percent withholding tax without an offsetting benefit to the corporation would materially raise the all-in tax cost of earning that income through the corporate form.
For example, imagine that a corporation distributes to shareholders $1,000 in E&P attributable to tax-exempt bonds held by the company. If the Graetz and Warren argument is followed (and ignoring any specific congressional relief for tax-exempt bond income), all shareholders would suffer a 35 percent withholding tax on the dividend -- a tax that would not have been imposed on earning the income directly. The result would be an enormous increase in the total all-in tax burden imposed on distributed corporate profits. That burden would extend not only to taxable individual accounts but also to tax-exempt institutions and foreign investors through the withholding tax mechanism. If shareholders instead were to invest through a limited liability company, any enterprise-level preference income would pass through to them -- although tax-exempt institutions, for example, would be required to overcome their collective discomfort with the unrelated business income tax.
In other words, a 35 percent withholding tax on distributions out of tax-favored income without offsetting corporate relief would turn the integrated tax system into a flat 35 percent income tax on an approximation of economic income (pretax E&P). But we know that passthroughs today enjoy much lower average effective rates, and fully distributed corporate income also bears a lower average effective tax rate.59
As a result, without offsetting corporate relief, corporate equity investors in a company that enjoys any significant tax preferences (including accelerated depreciation or the expensing of research and development costs), whether individuals' taxable accounts, tax-exempt institutions, or foreign investors, would find that their effective all-in tax costs would be greater than those of investors in comparable passthrough vehicles, and higher than under current law. When compared with corporate equity investments today, tax-exempts (and foreign shareholders who capture their returns as capital gains) would find their all-in tax cost on corporate economic income that is not subject to corporate income tax rising from 0 to 35 percent; taxable individual accounts' all-in tax cost would rise from 20 percent to 35 percent. In sum, the elegantly integrated system would drive a deeper wedge than exists today between corporate and unincorporated forms of business organizations and increase tax burdens on corporate income compared with current law.
Moreover, if a dividends paid deduction were not allowed for untaxed corporate income, companies might be induced to buy low-risk taxable income-generating operations to create a stream of income to fund dividends that would not add to the tax burdens of either the company or shareholders but would add to the GAAP earnings of the company. The income would be taxable to the company, but that tax would be offset by the dividends paid deduction for tax but not GAAP purposes. Shareholders would be in the same position as if they had invested in that safe asset directly, but the company's GAAP earnings would increase.
Graetz and Warren 2016 acknowledges these issues and essentially argues that the revenue pickup that would follow from recapturing corporate business tax incentives when the resulting E&P are distributed to investors could be deployed to reduce the tax-exempt institution tax rate. But the discontinuities extend to all investors, including individual taxable accounts. The larger question should be why exactly are we twisting ourselves into pretzels here for a modest tax rate cut for holders of less than one-quarter of outstanding U.S. corporate equity when the net effect is to create major new differences in tax outcomes between different forms of business organization.
It can further be argued that different forms of tax-favored corporate income should be handled differently. For example -- in theory, at least -- tax-exempt bond income is subject to tax in the hands of holders in the form of the lower coupon rate they accept. This is the well-known phenomenon of implicit taxation, in which tax-preferred returns on assets are capitalized into the prices for those assets.60 Returns bearing a clear implicit tax arguably should be considered after-tax income in a dividends paid system that extends only to after-tax income. But implementation might be maddening. The phenomenon is salient (if imperfectly captured) in tax-exempt bond yields, but it is not always clear how far it extends. For example, do market prices of equipment go up when bonus depreciation is introduced?
A second cluster of tax-preferred income items are those equally available to corporations and unincorporated businesses. As just discussed, an "integration" system that treats those items very differently depending on the form of business organization would seem to be moving rapidly in exactly the opposite direction from that which it claims to be its principal objective.
Finally, there are indirect FTCs, which, under current law are available only to corporations. But this fact is not necessarily grounded in any coherent foundation, because unincorporated businesses could, but for the nuisance aspect, operate in unincorporated form internationally, thereby capturing direct tax credits. Perhaps the original theory rested on the idea that double taxation of international income through the disallowance of indirect tax credits in a worldwide corporate tax system would distort corporations' capital allocations (as indeed it would) but that less mobile investors did not need to be placed in the same position as if they in fact conducted international operations directly. But the rise of the passthrough business implies that it is the limitation on indirect credits that is now the policy outlier.
Some observers also suggest that limitations within a dividends paid deduction regime on passing through the benefits of the indirect tax credit would be a valuable cure-all for various international tax policy design conundrums. Section III.E, above, addressed this analytically separate question.
On balance, the dividend integration literature here seems to be stuck in a bit of a time warp, when business income meant corporate income and when a radical distinction could be drawn between passive investors and active business corporate enterprises. The SFC 2014 staff report, the 2015 working group report, congressional testimony, and news reports all pitch the dividends paid deduction as an exercise in bringing the corporate form of business organization into parity with unincorporated business enterprises, but that would seem to require at a minimum a corporate NOL deduction for excess distributions. In fact, true parity would require more -- either an immediate corporate refund of an NOL attributable to an excess distribution or augmenting unused carryovers each year by a time value of money rate of return.
A dividends paid deduction form of integration does not necessarily create parity between corporations and passthrough business organizations in the handling of business tax preferences. Without further tweaking, it also does not create parity between distributed and retained earnings.
To promote efficiency goals, any new system that removes a structural bias against corporate dividends must avoid injecting a new structural bias that works against investors who realize their returns through sales of stock in the secondary markets -- capital gains. The problem was vividly on display in the tortured history of the U.K. integration system. Over the years, the United Kingdom vacillated from one regime to another, in each case consciously aware that it was favoring either corporate distributions or corporate retained earnings, and never achieving neutrality between the two.61
When a shareholder recognizes capital gain on the sale of stock, that gain might reflect the value of corporate retained earnings, unrealized changes in the value of corporate assets, larger revaluations of returns to capital more generally, or inflation. Presumably to reflect the second through fourth sources of nominal gain and to preserve tax distributional results, it is anticipated that the Finance Committee will retain current law's capital gain regime for sales of corporate stock. The problem then becomes how to coordinate an integration system for dividends with an unintegrated capital gains regime. Without that coordination, taxpaying shareholder returns attributable to the distribution of taxable corporate income would bear one level of tax, but returns captured through sales of stock whose value reflects the retention of those same earnings would be subject to double tax.
Graetz and Warren 2016 observes that a constructive DRIP can equalize the taxation of distributed and retained earnings and remove as well the incentive for what I call "capital stuffing," in which the corporation becomes a tax shelter to the extent its profits can compound indefinitely inside the corporation at higher after-tax rates than those available to investments outside the corporate sector.62 The DRIP would convert retained earnings into taxable distributions, thereby equating the two within the dividends paid regime.
A DRIP limited to a corporation's taxable income, combined with the withholding tax contemplated by the dividends paid integration proposal, would not impose current-year out-of-pocket cost on investors other than top-bracket individual investors. These individuals would be required to fund the 5 percentage point differential between the corporate tax rate and their individual tax rate (plus the 3.8 percent NII tax) out of their own pockets.
To accomplish the efficiency goals ascribed to dividends paid integration, a DRIP component would need to be mandatory. I have no other way of making sense of a dividends received integration proposal; to do otherwise would be to introduce one large disparity or another between distributed and retained earnings that no single capital gains tax rate could bridge. A mandatory DRIP also would obviate the differences between dividend distributions and redemptions treated as exchanges and would mitigate the importance of the tax-free step-up at death. Finally, a mandatory DRIP would enable the United States to collect some tax on what otherwise would be capital gains recognized by foreign investors; through the DRIP mechanism the United States would have collected taxes all along on a corporation's earnings and faute de mieux would forgo taxation of any gains attributable to unrealized corporate income.
In short, a mandatory DRIP is one of those bad ideas that might nonetheless be absolutely necessary in a dividends paid form of dividend integration. But this is a far more radical suggestion than the SFC 2014 staff report and the 2015 working group report contemplate.
A mandatory DRIP that extended to all E&P would exacerbate all the issues discussed in Section IV.B. concerning the treatment of corporate preference items. In particular, it would force to the foreground the treatment of shareholders on excess distributions -- or more accurately, the treatment of the corporation in respect of the NOL thereby generated. A compromise position might be a mandatory DRIP limited to a company's after-tax income (taxable income less taxes paid) rather than the entirety of its E&P, but this solution still would bias taxable investors in favor of capital gains to the extent a company earned untaxed profits and an effective mechanism to preserve the benefits of those preferences was not implemented.
A mandatory DRIP of all E&P would mean that every corporation would function like a mutual fund does today, in that 100 percent of corporate earnings would be actually or constructively distributed every year. Astute investors avoid buying into mutual funds immediately before a large annual dividend distribution because they will end up with current-year income offset only by higher basis (the cum-dividend price paid), which might not be useful for many years. The dividend paid deduction, and in particular a mandatory DRIP, might have some of the same distortive effects on trading close to actual or constructive dividend dates.
Dividend integration is also presented as reducing current law's bias in favor of debt financing. A dividends paid deduction has a nice optical symmetry about it in this regard, but the identical substantive analysis would apply to a dividend imputation approach.
The debt bias is real and important.63 My own work on the dual business enterprise income tax was originally motivated in large measure by this concern, and business cash flow taxes, allowances for corporate equity, and similar measures all address the issue (in general by exempting normal returns from tax at the corporate level). If a dividends paid deduction coupled with a withholding tax on investors is presented as an efficiency gain along this margin, it needs to be compared with these other proposals.
A dividends paid deduction coupled with a universal withholding tax might ensure one level of taxation for returns to equity, but the debt bias would be retained, in that pretax net business income paid to tax-exempt institutions and foreign investors as interest expense would still be untaxed at any level. Clientele effects suggest that the problem might become much worse with a universal withholding tax on dividends, since foreign portfolio holders of corporate equities, for example, might migrate to heavily equity-flavored debt instruments, which the financial markets would be only too happy to construct for them.64
Proponents of dividends paid integration have proposed to solve the problem by extending the universal dividend withholding tax to apply as well to interest paid by U.S. corporations to investors, including tax-exempt and foreign investors. But this is a truly bad idea, attractive only to those too young or too cloistered to remember the operation of the debt capital markets before the introduction of the portfolio interest rules in 1984.65 Since then, most countries have abandoned withholding taxes on portfolio interest.
Investors and corporations alike today operate in a liquid global market in debt instruments denominated in all the major tradable currencies, without local withholding taxes. Withholding taxes on interest do not collect tax but rather operate as on-off switches for portfolio investors, as we saw in the days before the adoption of the portfolio interest rules. Must we dust off our ancient Netherlands Antilles finance subsidiary technologies to preserve the global debt markets?
In short, the introduction of a withholding tax on interest would have enormous clientele effects, as also observed in Grubert and Altshuler 2016. Withholding on portfolio interest paid by U.S. corporations would simply send investors to dollar-denominated debt obligations of foreign issuers, because U.S. dollar debt instruments are not uniquely offered by U.S. corporations. London bankers and the currency swap markets would profit amid the chaos and deadweight losses that would follow from massive portfolio readjustments by both U.S. tax-exempt and foreign investors in response to interest withholding tax imposed on them.
Finally, by imposing withholding tax on interest at rates at or about the maximum tax rate on personal income, the United States would on its face violate virtually every one of its tax treaties. While an argument might be constructed that a 35 percent withholding tax on dividends is the economic equivalent of the existing corporate income tax, no such rationalization is available for interest income, which would go from wholly untaxed by the United States to fully taxed. Even in respect to dividend withholding tax, economic equivalence is usually not the standard applied in treaty interpretations. It is bizarre to think that the Finance Committee would think that a good idea. Broader thin capitalization rules that apply domestically are a more targeted response to the debt bias within the confines of the current tax system. This is not a problem whose solution requires turning the global debt markets upside down and violating our tax treaties.
Both the SFC 2014 staff report and the 2015 working group report understand and consider these issues, and in the end both punt. But one cannot punt forever, and if in fact one is going to hold out dividend integration as having desirable efficiency properties, one must specify exactly what iteration of dividend integration one has in mind and then analyze all the tax-induced responses that the proposal would trigger.
The efficiency case for dividend integration is much weaker today than it was a few decades ago. Individuals hold much less corporate equity in taxable accounts than they previously did, and the effective tax rates of the largest U.S. companies in many cases are very low (so that the all-in tax cost of dividends is lower than reference to statutory rates might suggest). A rigorous application of dividends paid integration in which corporate tax preferences are not passed through to shareholders would drive a much bigger wedge between doing business in corporate and unincorporated form than exists today. Moreover, it could introduce profound dislocations in corporate behavior by systematically preferring domestic to foreign investment, notwithstanding that a substantial fraction of large U.S. public corporations' incomes today are earned outside the United States.
Dividend integration schemes that attempt to do more than simply lower the all-in tax rate on dividend income introduce enormous technical difficulties, override treaties, and inject their own new deadweight losses whose welfare costs must be weighed against any perceived benefits. Moreover, the repurposing of dividend integration as a solution to the design flaws of the current U.S. international tax regime is bizarre, given that the problems identified can in fact be solved through relatively easy updates to existing code sections.
Dividends paid integration has additional pernicious political economy implications in dissipating the interest of a large component of the business community in achieving genuine business tax reform. Thoughtful policymakers from both parties understand that the business tax system of the United States is in need of more than fresh wallpaper and curtains.66 Both from the vantage point of legislative time (a very scarce commodity) and business self-interest, dividends paid integration will make fundamental business tax reform more difficult.
At a deeper level, dividend integration begs the question, how should capital income be taxed? At the same progressive income tax rates that apply to labor income, notwithstanding abundant economic evidence that capital and labor have different tax elasticities? At a lower flat rate, as in Nordic dual income taxes and my work on the dual business enterprise tax? Or not at all, in the technical sense that normal returns (the core meaning of capital income to an economist) should be exempt from tax, while rents should fall within the system? These are the right questions with which to begin, and when one does, other solutions emerge that are arguably easier to implement and have larger efficiency gains than does dividend integration.
The dividends paid iteration of dividend integration solves far fewer problems than it creates. It does, however, offer U.S. corporations a quick and dirty form of costless corporate tax reform in which their financial accounting effective tax rate decreases, but for entirely artificial reasons. Its use as a tax-free offshore cash repatriation technique might actually increase U.S. welfare by raising more total tax revenues than currently expected from broad corporate tax reform, but it does so by setting corporate managers at odds with their own shareholders. The dividends paid form of dividend integration has been wheeled forward in the manner of a true Trojan horse, seemingly offering a free gift of the end of double taxation but all the while containing in its belly the agenda of U.S. multinationals desperate to escape from under the mountain of offshore earnings that are the result of their own aggressive stateless income gaming.
1 The talk can be viewed at https://youtu.be/a50SKpTepZs. By way of disclosure, I participated in a panel at the same conference, immediately following Hanna's talk.
2 For example, many years ago, Michael L. Schler published an article outlining some of the technical problems with dividend integration proposals. Schler, "Taxing Corporate Income Once (or Hopefully Not at All): A Practitioner's Comparison of the Treasury and ALI Integration Models," 47 Tax L. Rev. 509 (1992). See also Schler, "Two Reasons Full Corporate Integration May Never Happen," Tax Notes, Feb. 3, 2003, p. 776.
3 Finance Committee Chair Orrin G. Hatch, R-Utah, remarked in February, "Corporate integration, once again depending on how it is designed, could significantly reduce effective corporate tax rates without all the difficult and highly politicized tradeoffs that will accompany a reduction in the statutory corporate tax rate." Hatch, speech at Bloomberg BNA tax policy event, "The Politics of Tax: Making Sense of Uncertainty" (Feb. 24, 2016). That observation was puzzling at the time, but its import now is clear.
In that same speech, Hatch emphasized the ad hoc nature of his integration idea:
- I get asked a lot about how I plan to move this proposal forward. One of the best features of a potential integration proposal is that, while it would have significant positive effects on its own, it does not prejudice or limit our ability to enact other reforms. We could pass a stand-alone integration bill to quickly address immediate problems. Or, as Chairman Brady recently said, it could be enacted to complement a broader international tax reform package. It could also eventually be included as part of a comprehensive tax reform bill, once we get to that point.
4 Edward D. Kleinbard, "Stateless Income," 11 Fla. Tax Rev. 699 (2011); Kleinbard, "The Lessons of Stateless Income," 65 Tax Law Rev. 99 (2011).
5 Hatch speech, supra note 3 ("I'll just say that I believe corporate integration is a viable and workable option, one that I think members of both parties should eventually be able to get behind. It may not be a silver bullet, but it would help address many of the persistent problems caused by our current tax system."); Bernie Becker, "Business Skeptics on Corporate Integration," Politico, June 17, 2016 ("Senate Finance Chairman Orrin Hatch has suggested this time might be different precisely because the debate over tax reform -- and math surrounding it -- is so hard. We have seen numerous corporations coming in -- corporate America coming in and saying, 'We're very much interested in this,' a top Hatch aide said recently.").
6 Hatch speech, supra note 3; Finance Committee hearing "Integrating the Corporate and Individual Tax Systems: The Dividends Paid Deduction Considered" (May 17, 2016) (Hatch's opening statement).
7 Finance Committee, "The Business Income Tax Bipartisan Tax Working Group Report," at 33-38 (July 8, 2015) (2015 working group report).
8 Andrew Velarde, "Hearing Weighs Distortion Elimination, Capital Expenditures," Tax Notes, May 30, 2016, p. 1157.
9 Graetz and Warren, "Integration of Corporate and Shareholder Tax" (May 5, 2016), available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2780490 (Graetz and Warren 2016). See also the passing references by Hatch in the Hatch speech, supra note 3.
See also 2015 working group report, supra note 7, at 60 n.32 ("Part of the attractiveness of the [dividends paid deduction] coupled with a withholding tax from the standpoint of corporations is that such an approach could have the effect of increasing the corporation's earnings per share by reducing its tax expense for financial accounting purposes, depending on how the Financial Accounting Standards Board (FASB) treats the withholding tax.").
10 Graetz and Warren 2016, supra note 9, at 11-12. Graetz also made the first point, using essentially the same words, in his statement before the Finance Committee at its May 17 corporate integration hearing.
11 Tax Analysts has republished as an eBook, available on Amazon (http://www.amazon.com/dp/B00PCRPKJW), the 1998 volume of Integration of the U.S. Corporate and Individual Income Taxes: The Treasury Department and American Law Institute Reports, along with an introduction by Graetz and Warren (hereinafter Treasury 1992 and ALI 1993, respectively).
12 In addition to the introduction to the volume referenced in the prior footnote, see Graetz and Warren, "Integration of Corporate and Individual Income Taxes: An Introduction," Tax Notes, Sept. 27, 1999, p. 1767 (Graetz and Warren 1999); and Graetz and Warren, "Unlocking Business Tax Reform," Tax Notes, Nov. 10, 2014, p. 707 (Graetz and Warren 2014).
13 Republican staff, Finance Committee, "Comprehensive Tax Reform for 2015 and Beyond" (Dec. 2014) (SFC 2014 staff report).
14 2015 working group report, supra note 7.
15 Toder and Viard, "A Proposal to Reform the Taxation of Corporate Income" (June 17, 2016). Howard Gleckman summarized the Toder-Viard proposal in "A New Way To Tax Corporations At The Shareholder Level, Forbes, June 17, 2016.
16 Rosenthal and Austin, "The Dwindling Taxable Share of U.S. Corporate Stock," Tax Notes, May 16, 2016, p. 923; Rosenthal, testimony before the Finance Committee, "Integrating the Corporate and Individual Tax Systems: The Dividends Paid Deduction Considered" (May 17, 2016) (Foreign investors hold about 26 percent of U.S. equities. The figures exclude inbound foreign direct investment.).
17See supra note 16.
18 For a recent proposal that would combine a mark-to-market system for publicly traded financial assets with a "complementary tax" comparable to the investor side of the dual business enterprise income tax, see Mark P. Gergen, "How to Tax Capital," Tax L. Rev. (coming), available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2749422.
19See the chart in SFC 2014 staff report, supra note 13, at 209-210.
20E.g., Graetz and Warren 2016, supra note 9, at 8.
21 The Tax Foundation adds Estonia to the list of countries to which the United States should look for dividend integration design ideas. Kyle Pomerleau, "Eliminating Double Taxation Through Integration," Tax Foundation Fiscal Fact No. 453 (Feb. 2015).
22 SFC 2014 staff report, supra note 13, at 192; Graetz and Warren 2016, supra note 9, at 4-5.
23 Gammie, "The End of Imputation: Changes in U.K. Dividend Taxation," 25 Intertax 333 (1997).
24 Cesare Podkul and Allan Sloan, "Wall Street Stock Loans Drain $1 Billion a Year From German Taxpayers" (May 3, 2016). A shorter version of the story appeared in The Washington Post, and a longer form in ProPublica.
25 Graetz and Warren 2016, supra note 9, at 8.
26 Financial Services Institute of Australasia, "Regulating Foreign Direct Investment in Australia," at 9 (Feb. 2014). By contrast, foreign portfolio investment represents about one-quarter of the equity of U.S. publicly traded firms. Rosenthal and Austin, supra note 16.
See also Rosanne Altshuler, Stephen E. Shay, and Eric Toder, "Lessons the United States Can Learn From Other Countries' Territorial Systems for Taxing Income of Multinational Corporations," Urban-Brookings Tax Policy Center, at 32 (Jan. 21, 2015):
- Australia differs from the United States in a number of dimensions. It is a substantially smaller economy and is a net importer of direct investment. It does not have a large pass- through sector and relies heavily on its corporate tax. Its tax rates are robust. Its imputation system of corporate taxation is increasingly unique and adversely affects untaxed foreign profits distributed by an Australian company to its resident shareholders.
27Cf. Altshuler, Shay, and Toder, supra note 26.
28 ALI 1993, supra note 11, at sections 2.1(b) and 2.4; see also Treasury 1992, supra note 11, at Ch. 12.A.
29 ALI 1993, supra note 11, noted the equivalence but considered a withholding tax as serving a purely compliance role:
- If, for enforcement purposes, a withholding tax on dividends were desirable under the dividend deduction method, that withholding tax would serve the same function as an equivalent rate corporate tax under the shareholder credit. The two integration methods would then be equivalent not only in their ultimate effect assuming perfect capital markets, but also in their immediate result.
Graetz and Warren introduction, supra note 11, at Kindle locations 9592-9595.
ALI 1993, supra note 11, also hinted at the possibility of collateral differences, as, for example, in financial accounting, although it did not explore the point in detail:
- Although the substantive results are the same under the two methods, the difference in labels might cause different legal, regulatory, or accounting consequences. For instance, the amount of the "dividend" paid for various purposes might differ. . . . [A portion] of the tax paid by the corporation might therefore be subject to different treatment by regulatory authorities or under tax treaties because it could be characterized as withholding (and therefore paid on behalf of shareholders) under the dividend deduction, but as a corporate tax under the shareholder credit mechanism.
Graetz and Warren, supra note 11, at Kindle locations 9601-9609.
30 It has been suggested that coordination across tax years would be handled differently, but this alternative reading breaks with the fundamental claim that dividends paid integration, when coupled with a withholding tax, has the same immediate cash tax consequences as an imputation system and would produce anomalous results.
Consider this example. For simplicity, assume that the shareholder tax rate, corporate income tax rate, and dividend withholding tax rate are all 35 percent. In year 1 Corp earns $200 (before any dividends paid deduction) and distributes $100 (before any corporate or withholding tax is considered) as a dividend. Year 2 is identical.
If this were an imputation system, Corp would pay $70 in corporate income tax each year ($140 total); no withholding tax would apply; Corp would distribute $65 net in cash each year as dividends; and shareholders would owe no incremental tax on the dividends they received. Shareholders instead would be viewed as receiving $100 in dividend income, on which they would be treated as having already paid $35 in tax.
The reading in the text produces the same result. In year 1 Corp would have $100 in taxable income after a $100 distribution, which would trigger a $100 dividends paid deduction. Corp would pay $35 in corporate income tax, and it would pay on behalf of shareholders $35 in withholding tax on a nominal $100 dividend, for a total outlay again of $70. Shareholders would receive $65 for the dividend, after withholding tax, just as in the imputation system. Year 2 would be identical. In particular, although Corp would open year 2 with a $35 taxes paid account, the $100 dividends received deduction would be "used up" against current-year income, and, as a result, the year 2 tentative $35 withholding tax obligation could not be satisfied by recourse to the taxes paid account. (If Corp distributed more than $200 in year 2, the taxes paid account would be relevant.)
The alternative reading yields the same result for year 1 but then effectively allows the year 2 distribution to do double duty, first by generating a dividends paid deduction that reduces Corp's corporate income to $100, and therefore its corporate income tax to $35, and second by allowing the withholding tax on the dividend distribution to be satisfied from the taxes paid account. The result would be a total all-in tax cost of $105 for the two years, rather than $140.
This alternative reading reduces total tax collections still further by simple timing adjustments to dividend flows. Imagine that the facts are the same as before, except Corp pays no dividend in year 1, and a $200 dividend in year 2. In year 1 Corp incurs $70 in corporate income tax and opens its books in year 2 with a $70 taxes paid account. In year 2 Corp reports zero taxable income ($200 operating income minus $200 dividends paid deduction) and, under the alternative reading, satisfies its $70 tentative withholding tax obligation out of its $70 taxes paid account. When the dust settles, tax collections have been slashed in half, relative to an imputation system or the reading in the text.
The source of the alternative reading's confusion can be attributed to an over-literal recollection of the dividend imputation system recommended in ALI 1993, section 2.4, Recommendation 1. That discussion does contemplate that the dividend withholding tax can always be satisfied out of an existing taxes paid account, but the reason is because it is describing the operation of an imputation system in which corporations pay corporate income tax on their taxable income without reduction for any dividends paid deduction and again owe dividend withholding tax on dividends. In that circumstance, the corporate income tax goes into the taxes paid account and acts as a credit against dividend withholding tax, so that the total tax is collected once, either on the corporate income if it is taxable income, or on the distribution if it is untaxed income that is not blessed through ALI 1993's separate exempt income account. In every case, however, tax is collected once by the corporation, as in the reading adopted in the text here.
31 ALI 1993, supra note 11, discusses stacking rules such as this in detail.
32 Hanna talk, supra note 1, at minute 26:04. The text's discussion focuses entirely on the repatriation of existing low-taxed offshore income held in foreign subsidiaries and follows Hanna's summary in this regard.
33 This, of course, is the combination of the 35 percent corporate rate and a 20 percent individual income tax on the $65 dividend per $100 pretax profit.
34 For example, the corporation might have benefitted from bonus depreciation. The analysis further depends on whether the dividends paid system would compensate corporations that distributed untaxed E&P.
Dividends are supported by all of a corporation's earnings, so current law's all-in costs should reflect the company's average, not marginal, tax rate.
35 Section IV.B expands further on this.
36 My intuition is that the argument that the dividends paid withholding tax is a liability of investors and not the corporate payer would be enhanced by making the tax credit refundable as a formal matter and simply providing that the tax rate applicable to tax-exempt institutions and foreign investors is the maximum individual rate.
37 White House, "The President's Framework for Business Tax Reform: An Update," at 24 (Apr. 1, 2016).
38 Hanna talk, supra note 32.
39 Harry Grubert and Altshuler, "Shifting the Burden of Taxation From the Corporate to the Personal Level and Getting the Corporate Tax Rate Down to 15 Percent," 69 Nat'l Tax J. (coming 2016), draft available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2802109. The text here assumes that corporations are not compensated for distribution of tax-favored income.
40 That is, there are many possible tax-induced responses to a dividends paid deduction, which responses in turn would vary depending on whether FTCs were treated as taxes paid. Some corporations might double down on stateless income planning to create revenues that compound at very low after-tax rates; others facing significant foreign taxes might buy U.S. cash cows to fund dividends, or alternatively consider major dispositions of foreign direct investments if a mandatory DRIP is part of the package, and still others might use streams of zero-taxed foreign income to fund dividend distributions.
41 Grubert and Altshuler, supra note 39.
43 Clausing, "The Effect of Profit Shifting on the Corporate Tax Base," Tax Notes, Jan. 25, 2016, p. 427. This is a shorter version of Clausing's paper, "The Effect of Profit Shifting on the Corporate Tax Base in the United States and Beyond," Nat'l Tax J. (coming 2016), available at.
44 I have in mind here by way of examples my dual business enterprise tax or Alan Auerbach's destination-based cash flow tax.
45 SFC 2014 staff report, supra note 13, at 122.
46 Schizer, "Between Scylla and Charybdis: Taxing Corporations or Shareholders (or Both)" (May 31, 2016), available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2788713; Gamage, "The Case for Taxing (All of) Labor Income, Consumption, Capital Income, and Wealth," 68 Tax L. Rev. 355 (2014).
47 So, too, as a practical matter is access to public debt markets.
48 Testimony of Kleinbard, "International Tax Reform Begins at Home," at 9 n.20 (Feb. 24, 2016).
49 For example, Pfizer Inc. has paid cash tax bills on its worldwide income at a rate on the order of 6 percent or 7 percent over the last several years. Frank Clemente, "Pfizer's Tax Dodging Rx: Stash Profits Offshore," Americans for Tax Fairness (Nov. 10, 2015). Essentially, Pfizer records each year for financial statement purposes a U.S. tax provision -- a reserve, if you will -- for the ultimate U.S. repatriation tax on a large portion of its low-tax foreign earnings, even though Pfizer devoutly hopes never to pay any such repatriation tax, at least at the statutory 35 percent rate. This explains the enormous difference between the 25 percent worldwide effective tax rate that Pfizer reports to shareholders and the cash taxes actually paid each year.
50 The analysis is developed in Kleinbard testimony, supra note 48.
52 ALI 1993, supra note 11, at section 2.2, adds to the mix the concept of "superintegration," in which the combined tax on distributed income is lower than would be the result if the preference were simply passed through in a manner akin to partnership taxation. The text does not mean to invoke that concept.
53 Reuven S. Avi-Yonah, "Is Corporate Integration a Good Idea?" Tax Notes, June 20, 2016, p. 1697.
54 Graetz and Warren 2016, supra note 9, at n.8.
55 Avi-Yonah, supra note 53.
56 Those accounts hold only about one-quarter of U.S. corporate equities. Rosenthal and Austin, supra note 16.
57 ALI 1993, supra note 11, at section 2.2.
58Id. at section 2.4, Recommendation 3 ("exempt income account" mechanism).
59 Michael Cooper et al., "Business in the United States: Who Owns It and How Much Tax They Pay," Treasury and National Bureau of Economic Research (Sept. 20, 2015).
60 Kleinbard, "The Lessons of Stateless Income," supra note 4, at 99, 118-134.
61 Gammie, supra note 23.
62 Kleinbard, "Corporate Capital and Labor Stuffing in the New Tax Rate Environment" (2013), available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2239360.
63 Ruud de Mooij, "Tax Biases to Debt Finance: Assessing the Problem, Finding Solutions," International Monetary Fund Staff Discussion Note SDN/11/11 (May 3, 2011), available at http://www.imf.org/external/pubs/ft/sdn/2011/sdn1111.pdf.
64 In theory, this should not be the case to the extent that the withholding tax perfectly mirrors current law's corporate income tax, but optics do matter at least to some extent, and, more important, the perfect mirroring requires an effective mechanism to pass through the benefits of corporate tax-preferred income. To emphasize again, that income is tax-free to both tax-exempt and foreign investors today when realized through capital gains and is taxed only to the extent of (treaty rate) withholding tax on dividend distributions today.
65 Sections 871(h) and 881(c).
66 Kleinbard, "Why Corporate Tax Reform Can Happen," Tax Notes, Apr. 6, 2015, p. 91.
END OF FOOTNOTES