With the election of Donald Trump and the Republican Party’s domination of Congress, House Speaker Paul Ryan’s blueprint for fundamental tax reform requires more careful analysis. The Ryan blueprint combines reduced individual rates with a destination-based cash flow type business tax applicable to all businesses. The destination-based business tax at the center of the blueprint has several major problems: It is incompatible with our WTO obligations, it is incompatible with our tax treaties, and it will not eliminate the problems of income shifting and inversions it is designed to address. In addition, these proposals generate vexing technical problems that are not easily fixed as well as significant political problems. Finally, due to the tax rates that have been proposed, the plan is likely to generate large revenue losses and a less progressive tax system. We conclude by recommending better tax policy solutions to our current corporate tax problems.
- I. Introduction
- II. Is the “Better Way” Proposal Compatible with the WTO?
- III. What About Tax Treaties?
- IV. Tax Avoidance, Income Shifting and Inversions
- V. Vexing Technical Problems
- VI. Progressivity and Revenue Effects
- VII. Conclusion
This part describes the Ryan proposal in more detail, describing in particular the plan’s destination-basis corporate tax. Part II discusses problems of WTO compatibility and trade distortions under this plan. Part III discusses issues surrounding tax treaty compatibility, and Part IV discusses the lingering potential for profit shifting under the plan. Part V describes technical problems associated with implementing the plan. Part VI addresses effects on the progressivity of the tax system and on government revenues, and Part VII concludes and offers other suggestions for reform.
House Speaker Paul Ryan’s (R-WI) blueprint to reform the tax code is gaining new prominence because of the Republican ascendancy in Washington following the 2016 election. Since President Trump is likely to sign any tax reform passed by a Republican Congress, it is worth serious consideration.
The introduction to the Ryan proposal (the “Blueprint”) states that:
This Blueprint represents a dramatic reform of the current income tax system. This Blueprint does not include a value-added tax (VAT), a sales tax, or any other tax as an addition to the fundamental reforms of the current income tax system. The reforms reflected in this Blueprint will deliver a 21st century tax code that is built for growth and that puts America first.
This statement is important, because as will be discussed below, the business part of the proposal can be seen as a modified subtraction method VAT. If it were a VAT, it would not have problems with tax treaties or with the WTO rules. But since it declares itself not to be a VAT, and has at least one crucial feature that differs from a VAT, it may have problems with both.
The individual tax section of the Blueprint is not a structural change, although it is quite regressive and would lead to massive budget deficits. It envisages a lower rate structure for ordinary income (up to 33%), a capital gains and dividends and interest rate that is half the rate for ordinary income (up to 16.5%), and abolishing the individual AMT and estate tax. For pass-through businesses, the Blueprint envisages a rate of 25%, with special provisions to prevent shifting of wage income to pass-throughs.
A particularly radical portion of the Blueprint is the corporate section. In addition to cutting the corporate tax from 35% to 20%, the Blueprint envisages three major reforms. First, businesses will be allowed to expense capital expenditures, resulting in a zero rate for the marginal return on investment:
This Blueprint will provide businesses with the benefit of fully and immediately writing off (or “expensing”) the cost of investments. This represents a 0 percent marginal effective tax rate on new investment.
Second, businesses will not be able to deduct net interest expense:
Under this Blueprint, job creators will be allowed to deduct interest expense against any interest income, but no current deduction will be allowed for net interest expense. Any net interest expense may be carried forward indefinitely and allowed as a deduction against net interest income in future years.
Third, the Blueprint will be destination-based, i.e., be fully imposed on imports (without any deductions) and not imposed at all on exports:
This Blueprint eliminates the existing self-imposed export penalty and import subsidy by moving to a destination-basis tax system. Under a destination-basis approach, tax jurisdiction follows the location of consumption rather than the location of production. This Blueprint achieves this by providing for border adjustments exempting exports and taxing imports, not through the addition of a new tax but within the context of the transformed business tax system. The Blueprint also ends the uncompetitive worldwide tax approach of the United States, replacing it with a territorial tax system that is consistent with the approach used by our major trading partners.
This means that imports will be taxed and exports exempted. In addition, the Blueprint will enable dividends from foreign subsidiaries of U.S.-based multinationals to be fully exempt, but will maintain the current Subpart F provisions for passive income, eliminating only the base company rule and section 956:
Today, all of our major trading partners raise a significant portion of their tax revenues through value-added taxes (VATs). These VATs include “border adjustability” as a key feature. This means that the tax is rebated when a product is exported to a foreign country and is imposed when a product is imported from a foreign country. These border adjustments reduce the costs borne by exported products and increase the costs borne by imported products. When the country is trading with another country that similarly imposes a border-adjustable VAT, the effects in both directions are offsetting and the tax costs borne by exports and imports are in relative balance. However, that balance does not exist when the trading partner is the United States. In the absence of border adjustments, exports from the United States implicitly bear the cost of the U.S. income tax while imports into the United States do not bear any U.S. income tax cost. This amounts to a self-imposed unilateral penalty on U.S. exports and a self-imposed unilateral subsidy for U.S. imports.
Because this Blueprint reflects a move toward a cash-flow tax approach for businesses, which reflects a consumption-based tax, the United States will be able to compete on a level playing field by applying border adjustments within the context of our transformed business and corporate tax system. For the first time ever, the United States will be able to counter the border adjustments that our trading partners apply in their VATs. The cash-flow based approach that will replace our current income-based approach for taxing both corporate and non-corporate businesses will be applied on a destination basis. This means that products, services and intangibles that are exported outside the United States will not be subject to U.S. tax regardless of where they are produced. It also means that products, services and intangibles that are imported into the United States will be subject to U.S. tax regardless of where they are produced. This will eliminate the incentives created by our current tax system to move or locate operations outside the United States. It also will allow U.S. products, services, and intangibles to compete on a more equal footing in both the U.S. market and the global market.
The Blueprint then addresses the potential WTO issue as follows:
The rules of the World Trade Organization (WTO) include longstanding provisions regarding the use of border adjustments. Under these rules, border adjustments upon export are permitted with respect to consumption-based taxes, which are referred to as indirect taxes. However, under these rules, border adjustments upon export are not permitted with respect to income taxes, which are referred to as direct taxes. This disparate treatment of different tax systems is what has created the historic imbalance between the United States, which has relied on an income tax – or direct tax in WTO parlance – for taxing business transactions, and our trading partners, which rely to a significant extent on a VAT – or indirect tax in WTO parlance – for taxing business transactions. Under WTO rules, the United States has been precluded from applying the border adjustments to U.S. exports and imports necessary to balance the treatment applied by our trading partners to their exports and imports. With this Blueprint’s move toward a consumption-based tax approach, in the form of a cash-flow focused approach for taxing business income, the United States now has the opportunity to incorporate border adjustments in the new tax system consistent with the WTO rules regarding indirect taxes.
This approach is similar to the one taken by the 2005 advisory panel on tax reform in the Growth and Investment Tax (GIT) proposal. Under the GIT, corporations were subject to a cash flow tax with expensing and no deduction for interest, but wages were deductible. The GIT was destination-based, but for revenue estimating purposes, the revenue associated with border adjustments was disregarded because of concerns about WTO compatibility. Since the U.S. has a large trade deficit, this represented a difference of $775 billion dollars in revenues over the ten-year budget window. According to the Tax Policy Center analysis (2017), the revenue effects of the border adjustment are even larger now, at about $1.2 billion dollars.
Under the WTO Subsidies and Countervailing Measures (SCM) Agreement, a tax may only be border adjustable if it is an “indirect” tax. A border adjustable “direct” tax is a prohibited export subsidy that can subject the U.S. to trade sanctions.
Annex I of the SCM includes as a prohibited export subsidy:
(e) The full or partial exemption remission, or deferral specifically related to exports, of direct taxes (58) or social welfare charges paid or payable by industrial or commercial enterprises (59).
Footnote 58 provides:
For the purpose of this Agreement:
The term “direct taxes” shall mean taxes on wages, profits, interests, rents, royalties, and all other forms of income, and taxes on the ownership of real property; …
The term “indirect taxes” shall mean sales, excise, turnover, value added, franchise, stamp, transfer, inventory and equipment taxes, border taxes and all taxes other than direct taxes and import charges.
Footnote 59 provides:
The Members recognize that deferral need not amount to an export subsidy where, for example, appropriate interest charges are collected. The Members reaffirm the principle that prices for goods in transactions between exporting enterprises and foreign buyers under their or under the same control should for tax purposes be the prices which would be charged between independent enterprises acting at arm’s length. Any Member may draw the attention of another Member to administrative or other practices which may contravene this principle and which result in a significant saving of direct taxes in export transactions. In such circumstances the Members shall normally attempt to resolve their differences using the facilities of existing bilateral tax treaties or other specific international mechanisms, without prejudice to the rights and obligations of Members under GATT 1994, including the right of consultation created in the preceding sentence.
Paragraph (e) is not intended to limit a Member from taking measures to avoid the double taxation of foreign-source income earned by its enterprises or the enterprises of another Member.
The business tax regime of the Blueprint can be seen as a modified version of a consumption tax–specifically, a subtraction method VAT (although the Blueprint explicitly denies that it is a VAT). Specifically, the Blueprint imposes tax on cash flow, allows expensing of capital expenditures, and disallows net interest expense. All of these are also features of a subtraction method VAT.
However, the Blueprint does allow a deduction for wages, while a subtraction method VAT would disallow them. This feature makes the Ryan tax not WTO compatible. Fundamentally, we need to consider the reason why a VAT, whether using a credit-invoice or subtraction method of calculating the tax, is border adjustable. Sales taxes, excises and VATs are border adjustable because there is no distortion introduced by the tax; goods receive like tax treatment in the domestic market irrespective of where they are produced. Both the tax component in exports and the price of imports are measurable, and the border adjustment does not exceed the tax that is levied because (in the case of import) the full tax is levied at the border, and (in the case of exports) the refunded amount in an invoice-credit VAT is only the amount that was levied at previous stages, as shown on the invoice. By so limiting border adjustments, the WTO reduces opportunities for countries to subsidize exports or overtax imports.
The Ryan Blueprint’s treatment of purchases (including capital and inventory) and labor highlights the difference between a tax on value added and Ryan’s tax on an income base.
If the factors of production employed at each stage of production and distribution of goods are totaled up, they should equal the retail sales price of the goods. A traditional VAT is imposed mainly on two factors of production, labor (about 2/3 of base) and income from capital or rents (extra profits above the normal return to capital). Under a sales-subtraction method VAT, taxes are collected and remitted to the government by business at each stage of production and distribution. The resulting tax should be equal to the tax imposed on the retail price of taxable goods under a single-stage retail sales tax. Purchases taxed at a prior stage of production or distribution are deductible, so that this value is not taxed again. Under that method of calculating VAT, the cost of labor is not deductible so that this factor of production can be included in the tax base. In contrast, under the Ryan Blueprint tax, a business can take an immediate deduction for its wage expense, leaving that factor of production out of the tax base. Workers bear tax at multiple rates on that labor income under the individual income tax. Even if the tax paid by the workers may be viewed as a surrogate for a business’s tax on labor, that surrogate tax cannot be accurately measured and that tax cost does not enter the tax-inclusive prices of the business’s outputs. Giving a full deduction for labor costs effectively subsidizes exports and overtaxes imports.
For example: Assume that a domestic grape grower has no business inputs. He has labor costs of 30 and profit of 10. He sells the grapes to a wine producer for 30 + 10 = 40. Since labor is deductible, the grape grower pays tax only on his profit. The tax is 10 x 20% = 2, so the tax inclusive price is 40 + 2 = 42. The wine producer buys the grapes for 42. She has labor costs of 45 and profit of 15. She sells the wine to a domestic consumer for 42 + 45 + 15 = 102, and pays tax only on the profits of 15 since the other elements are deductible. Total tax paid by the wine producer is 15 x 20% = 3, and the tax inclusive price to the consumer is 102 + 3 = 105.
If the wine producer instead exports the wine by selling it to a foreign customer, she has 100 in exempt income, or zero income (assuming no other income). She also has 40 + 45 = 85 in deductible costs, so in principle she should get a check from the Treasury of 85 x 20% = 17. The foreign customer, assuming that his country also charges 20% VAT on imports, will pay 100 plus VAT of 100 x 20% = 20, and the tax inclusive price will be 120. Note that this is a higher price than the price to the domestic wine consumer, because in the domestic sales the costs of goods sold and the labor are deductible whereas in the foreign sale they are not.
Now let us compare this to a normal invoice credit VAT of 20%. In the domestic case, the grape grower has 30 in labor costs and 10 of profit, and he will charge the wine producer a tax inclusive price of 40 + (40 x 20%) = 48. The wine producer will pay 48 to the grape grower and has 45 of labor costs and 15 of profits, so she will charge a tax inclusive price of 48 + 45 + 15 = 108 minus 8 refund of VAT paid on inputs, or 100 + (100 x 20%) = 120.
In the export case with an invoice-credit VAT, the grape grower still charges the wine producer 48. The wine producer adds labor costs of 45 and profits of 15 and since the wine is exported in a zero rated sale she receives a refund of 8 and the sale price to the foreign consumer is 48 + 45 + 15 – 8 = 100, plus 20% foreign VAT or 120.
If we compare the two cases, under the Ryan tax the domestic consumer pays 105 and the foreign consumer 120. The difference of 15 is the tax on the deductible U.S. labor costs (= (30 + 45) x 0.20). But if the wine producer wants to undercut wine produced in the foreign country, she can easily afford to sell for less than 100. Specifically, she could sell for as low as (100 – 17) + 20%, or $99.60 (tax inclusive). This demonstrates the export subsidy, which results from the ability to deduct labor costs in the U.S., whereas such costs are not deductible in the normal VAT in the foreign country. Under the normal VAT, the prices to the domestic and foreign customers are the same (120 domestic, 120 foreign) and there is no check from the Treasury other than the refund of VAT actually paid.
The reason for the export subsidy in the Ryan tax is that labor costs are deductible. In theory this should not make a difference if we could be sure that labor is subject to at least a 20% tax rate, since then the deduction and inclusion would offset each other. However, much labor income is taxed at lower rates due to the progressivity of the federal income tax as well as the earned income tax credit. Ryan also envisages a zero bracket of the first $24,000 of income and a 12% rate for those currently in the 10 or 15% brackets, so it is likely that many of the employees of the grape grower and the wine producer will be subject to individual tax at less than 20%.
Thus, the Ryan Blueprint should be classified as a modified consumption-style tax imposed on an income base. As such, it is not a border adjustable tax under the WTO rules, as currently interpreted. If the U.S. treated a Ryan-type tax as border adjustable, we can expect our international competitors to challenge the tax at the WTO before it takes effect.
Economists, however, argue that exchange rate changes may offset this, because U.S. dollar appreciation would undo the export subsidy. But the exchange rate offset will not be perfect since the tax treatment will depend on individual firm circumstances, and the exchange rate only affects the overall prices of imports relative to exports. In particular, different goods will receive the export subsidy to different extents, because not all goods have the same share of labor in their production costs, and different tax rates apply to corporate and pass-through business. Yet any exchange rate changes will affect all goods equally.
Even more important, the literature on exchange rate determination makes any exchange rate offset hardly predictable or clear cut. Empirical studies in international finance makes it quite clear that exchange rates movements are divorced from most coherent theories of exchange rate determination. As noted by Rogoff:
The extent to which monetary models, or indeed, any existing structural models of exchange rates, fail to explain even medium term volatility is difficult to overstate. The out-of-sample forecasting performance of the models is so mediocre that at horizons of one month to two years they fail to outperform a naïve random walk model (which says that the best forecast of any future exchange rate is today’s rate). Almost incredibly, this result holds even when the model forecasts are based on actual realized values of the explanatory variables.
This may be due in part to the huge speculative component of exchange rate trading. The foreign exchange market has transactions that exceed $5 trillion each day; the U.S. dollar is involved in 88% of these currency trades. Compare the size of the world economy, with an annual GDP of about $75 trillion. All of world GDP could be purchased with about 15 days of foreign exchange! Thus, the bulk of exchange rate trading is not related to the purchase of goods or even assets, but rather to financial market trading. This may help explain why exchange rate movements are difficult to predict with standard theories or macroeconomic models. Indeed, macroeconomists have a dismal record of predicting exchange rate movements based on any fundamental theories of exchange rate determination. Thus, there should be grave doubts that exchange rate changes will smoothly offset the effects of the border adjustment.
The exchange rate offset argument is sometimes made by noting that trade must balance in the long run, or by simply assuming balanced trade. Yet while trade must balance in the long run, there is no reason why countries can’t run persistent trade deficits and surpluses. Indeed, the United States has experienced a trade deficit for every year of the last 40 years. Our persistent trade deficit is due to macroeconomic considerations, and in particular, the fact that U.S. savings are low relative to our private investment desires and government borrowing. If nothing changes those macroeconomic variables, then our trade deficit should remain constant, so the exchange rate offset must offset any trade distortions introduced by the tax changes. Still, it is far from clear that a tax change of the magnitude imagined here would not affect macroeconomic variables such as savings, investment, tax revenues, and government spending.
In addition, many countries do indeed fix their exchange rates, and this will also slow any adjustment to the introduction of the Ryan tax. Auerbach and Holtz-Eakin recognize that, but they note that most countries do this for reasons of “competitiveness” and therefore could be expected to adjust pegs accordingly. We disagree. Most countries peg to achieve other macroeconomic goals, and in particular to import creditability with respect to monetary policy, to target inflation, to enhance exchange rate stability, etc. It is far from clear that competitiveness is the determinative motive in most cases. (And often pegs will have the opposite effect, when countries intervene to support overvalued currencies.)
Further, trade contracts are often set in advance in dollar terms, so even if exchange rates were to adjust immediately and fully, there would still be a disruptive lag in terms of effects on those engaged in international trade. This shock could be quite damaging to retailers in the short run. Also, if lags in exchange rate adjustment convince trading partners to undertake protectionist trade measures in response, those measures are likely to prove more long-lasting.
In addition, one shouldn’t be sanguine about the effects of a large dollar appreciation, as this redistributes wealth away from U.S. owners of foreign assets (since their assets are now worth less in dollar terms) and toward foreign owners of U.S. assets. These wealth effects involve amounts in the trillions of dollars. Dollar appreciation can also have dire fiscal consequences for emerging economies that are borrowing in dollars; indeed U.S. dollar appreciation played a large contributing role in several past developing country debt crises, including the Latin American debt crises of the mid 1980s and the Argentine debt crisis and default of 2001.
We are not aware of any empirical evidence on the exchange rate mechanism, but that should be provided before adjustment is taken on faith. Indeed, it seems dangerous to “bet” entire sectors of the economy on such untested grounds, especially when no other major country has adopted this type of corporate tax. The only empirical study, by Desai and Hines, in fact suggests that trade effects may be counter to expectations. According to Desai and Hines, “[e]conomic theory implies that exchange rate adjustment prevents destination-based VATs from affecting exports and imports. Indeed, this proposition is so well accepted among economists that it has not been subjected to serious prior testing.” Still, Desai and Hines found that countries that relied on VATs actually had worse export performance (and also lower imports), and this finding typically (but not always) persists when control variables and country fixed effects are included.
Desai and Hines note that the real world features of VATs can explain their finding, since VATs tend to fall more heavily on traded goods than non-traded goods, and export rebates are often incomplete, thus discouraging trade. These two explanations also likely apply in the Ryan tax context. For reasons explained below, the unlikelihood of exporters getting full rebates for their export “losses” are even stronger in the Ryan tax context than in a traditional VAT, and there is reason to believe that the Ryan tax will be imposed differentially on tangible goods than on services and intangibles (discussed below).
There are other WTO related problems with the Blueprint as well. First, the Blueprint explicitly declares up front that it is not a VAT but a corporate income tax (“This Blueprint does not include a value-added tax (VAT), a sales tax, or any other tax as an addition to the fundamental reforms of the current income tax system”). Second, the retention of an exemption for dividends from controlled foreign subsidiaries (on top of the destination basis) and subpart F and the imposition of tax on some interest and dividends make the Blueprint look more like a corporate income tax. In contrast, VATs are purely destination-based and do not apply to any foreign source income, so territoriality is not needed, and financial flows are disregarded.
Auerbach and Holtz Eakin argue that:
There is an open question whether a destination-based cash flow tax (DBCFT) would be determined to be compliant with the rules of the World Trade Organization. There are two primary issues here. First, WTO rules currently limit border adjustments to “indirect” taxes – taxes on transactions (e.g., sales, payroll, etc.) rather than “direct” taxes on individuals or businesses. It is not clear that a DBCFT would be successfully characterized as an indirect tax, even though it is economically equivalent to a policy based on indirect taxes (a VAT and a reduction in payroll taxes), and even though the distinction between direct and indirect taxes has little meaning and no bearing on any economic outcomes. In addition, there might be concerns under existing WTO rules regarding the combination of border adjustments with a deduction for domestic labor costs, since the border adjustment assessed on imported goods applies to the entire cost of the imports, with no deduction for the labor costs that went into the production of these imported goods. Some might see this treatment as favoring domestically produced goods over imported ones. But such an inference makes little sense from an economic perspective. Again, consider the equivalent policy of introducing a VAT and reducing payroll taxes, both elements of which are compatible with WTO rules. A reduction in payroll taxes would indeed encourage domestic production and employment to the extent that it lowered domestic production costs. But this is true of any reduction in taxes on US production, and it is difficult to comprehend why international trade rules should dictate the tax rate a country applies uniformly to its own domestic economic production activities.
Given these arguments, one might legitimately query why proponents have not simply suggested replacing the corporate tax with the combination of a VAT and a cut in payroll taxes. (Though to achieve a 20% wage subsidy, one would have to provide more tax relief than a complete elimination of the 15% payroll tax.) Still, regardless of the merits of such equivalence arguments, which neglect real world features of modern payroll taxes, it is unlikely that they will sway the WTO. WTO decisions tend not to respect this type of argument even if economists find this “difficult to comprehend.” The whole point of introducing the Ryan tax, as Auerbach and Holtz-Eakin concede, is to make the United States into a giant tax haven from the perspective of our trading partners, and induce their multinationals to move operations into the United States. Given the likely harm to their tax revenues from such a shift following the initial introduction of the Ryan tax, our trading partners, and especially the EU, are likely to sue. The result would be years of litigation with an uncertain outcome and potentially very large trade sanctions. Recent estimates suggest that dispute could result in retaliatory tariffs sufficient to eliminate over $200 billion in U.S. exports.
Such an outcome would be very worrisome for several reasons. First, we are already in an environment where the gains from trade are being threatened by a President that frequently urges the imposition of tariffs. Adding protectionist features to the tax code, even if some economists are convinced that there would be no net effect on prices, risks misunderstanding and increases the probability of retaliatory tariffs. Indeed, some countries have already pledged tariff retaliation if the United States moves forward with this plan. Protracted and contentious litigation could also reduce the U.S. political backing for the WTO, harming both the long-run prospects for an open trading system and our international relations.
Second, the ambiguities of whether these tax provisions would pass muster with the WTO creates a far more uncertain investment climate, making it more difficult for companies to resolve investment and location decisions. Further, if there is no assurance that the tax will be retained, that could also hamper the process of exchange rate adjustment.
Finally, if the WTO authorizes trade sanctions in response, such sanctions may lead to an endgame result where the U.S. government complies with the WTO by turning the Ryan tax into a “normal” VAT by denying the deduction for labor. This would make the tax far more regressive than the proposed cash-flow corporate tax it replaces. Of course, the border adjustment feature could not be dropped without huge revenue losses as well as enormous tax avoidance and profit shifting problems. Absent the border adjustment, the whole structure of the destination-based tax breaks down.
There are three problems with tax treaties in the Blueprint, assuming that the proposed tax is an income tax subject to the treaties. The first problem is that if the business tax is an income tax covered by the treaties and we are serious about taxing on a destination basis goods and services imported into the U.S., we need to do away with the permanent establishment (PE) limitation in Article 7, because we need to be able to tax importers without a PE (or physical presence required under domestic law). While we believe that this is a long overdue reform, bringing the income tax treaty into the 21st century and the age of electronic commerce, it should be recognized that it involves a massive treaty override of a crucial aspect of the treaty bargain, which was considered and rejected by our treaty partners in the BEPS context.
The second problem is that if the business tax is an income tax, in order to levy it on a destination basis and include all imports, it must be imposed not just on goods and services (under Article 7) but also on intangibles that produce royalties (Article 12) and other types of deductible payments that can substitute for royalties (e.g., payments on derivatives, generally classified as Other Income under Article 21). While interest and dividends are not deductible, allowing royalties and derivatives to escape the tax on imports invites abuse (since there will always be lower tax jurisdictions). This requires another treaty override that can be avoided if the business tax is a VAT.
Finally, it could be argued that because the Ryan tax advantages domestic companies that export from the U.S. over similar foreign companies that import into the U.S., the Ryan tax is a violation of the non-discrimination provision of the treaties.
An important related question is how our treaty partners will react to such sweeping changes and treaty overrides (which they regard as violations of international law). Given that the new U.S. tax (20% rate with expensing, territoriality, border adjustments) will create a strong attraction for foreign-based multinationals to shift profits into the U.S., it is likely that they will (a) refuse to give credit for the U.S. tax under tax treaties because (given expensing) it is not an income tax, and (b) apply their CFC rules to U.S. affiliate operations by their multinationals, which cannot invert in response because of exit taxes. The possible end result could be a collapse of the treaty-based international tax regime, to the disadvantage of U.S. taxpayer who will face increased withholding taxes overseas as well as increased transfer pricing enforcement.
The Better Way proposal argues that:
Taken together, a 20 percent corporate rate, a switch to a territorial system, and border adjustments will cause the recent wave of inversions to come to a halt. American businesses invert for two reasons: to avail themselves of a jurisdiction with a lower rate, and to access “trapped cash” overseas. Those problems are solved by the lower corporate rate and the territorial system, respectively. In addition, border adjustments mean that it does not matter where a company is incorporated; sales to U.S. customers are taxed and sales to foreign customers are exempt, regardless of whether the taxpayer is foreign or domestic.
We do not believe the Blueprint proposal will completely stop the incentive for U.S. corporations to shift income overseas, because even with a 20% rate and expensing, rents (for example, from intangibles such as Apple’s “Irish” profits) can still be located in zero tax jurisdictions and then repatriated tax-free. While this problem can be minimized if it is limited to rents from exploiting foreign markets (which would be exempt even if carried out from the U.S.), we are doubtful that the line between U.S. and foreign markets can be drawn precisely where services and intangibles are concerned, where there can be no enforcement of the tax at the border. Even a normal (invoice credit) VAT has issues where imports of services and intangibles are concerned, since it is difficult to collect the tax from consumers who are not eligible for deductions or input credits.
Moreover, experienced tax practitioners have already suggested ways of gaming the Blueprint. For example:
- A U.S. pharmaceutical with foreign subsidiaries could develop its intellectual property in the United States (claiming deductions for wages, overhead and R&D), and then sell (i.e., export) the foreign rights to its Irish subsidiary (at the highest price possible). The proceeds would not be taxable. Ireland would allow that subsidiary to amortize its purchase price. This creates tax benefits in each jurisdiction by reason of the different regimes. If the Irish subsidiary manufactures drugs, the profits could be distributed up to the U.S. parent tax-free under a territorial system. If the Irish subsidiary is in danger of becoming profitable for Irish tax purposes, the U.S. parent would just sell it more IP.
- If an Irish parent owns a U.S. subsidiary, the Irish parent can issue debt to fund the purchases of the IP. The U.S. subsidiary then invests the cash to generate more IP (expensing all equipment and deducting all salaries) and sells the IP to its parent.
- If an Irish parent has purchased the U.S. IP rights, it would not want to license the rights to the U.S. subsidiary (income for Irish parent under Irish tax law and no deduction for U.S. subsidiary). So it just contributes the rights to another U.S. subsidiary. Could the U.S. subsidiary amortize the parent’s basis under the Blueprint? When one U.S. subsidiary licenses to another, no net tax would be paid. Any royalties would be taxable to the licensor but deductible for the payor.
- How does the Blueprint work for services? If a U.S. hedge fund manager provides services to an offshore hedge fund, is that considered an export that is tax exempt? What if the U.S. manager develops a trading algorithm and sells it (or licenses) it to an offshore hedge fund? Are the proceeds and royalties exempt? If so, then the hedge fund becomes a giant tax shelter to the manager, because he would not pay 25% on this income – he would pay zero, with no further tax. This is much better than the current carried interest provision, which has attracted bipartisan condemnation because it enables individuals with income of many millions to pay a reduced rate. The Blueprint result is much worse.
First, this tax system is very difficult to explain to the public or, even, experts. This creates a risk that loopholes will be easier to design due to the deliberate exploitation of the system’s complexity by savvy tax planners and lobbyists. Yet if the system is implemented in a more theoretically pure form, without opening the door to loopholes, it is not clear that the MNC business community would support the proposed changes. The net effect would be a tax increase for the intangible- intensive MNCs that had previously succeeded in achieving single-digit tax rates by gaming the old system (and shifting U.S. profits abroad). It is also a tax increase for highly-leveraged firms, since debt-financed investments would no longer be subsidized. Retailers that import into the U.S. and manufacturers that import parts are likely to object to a new tax system that means they cannot deduct their cost of goods sold.
Second, there is an increased likelihood that many profitable firms would show losses. This is especially the case for exporters, since they may have deductible expenses, but no taxable revenue. Exporting firms with persistent losses will find the credits do them no good, which would affect export incentives. While economists would support a refund system in order to keep tax neutral, there is a large potential for fraud, and politically it seems unlikely that the government could issue large checks to profitable corporations on a permanent basis. The alternative suggested by the Blueprint is unlimited carry-forwards, but this doesn’t solve the problem for businesses with losses that may not be offset. And this introduces new trade distortions since the border adjustment is not symmetric. Exporting companies could of course merge with non-exporters in order for the losses to be more useful, but inducing a slew of tax-motivated mergers would be inefficient.
Auerbach and Holtz-Eakin recognize that this would be a large problem for exporting firms. They suggest allowing firms to use credits to offset payroll taxes, or have a system of refundable border adjustments, but both of these solutions are problematic and difficult to implement.
Third, there are myriad technical problems that remain to be worked out. For example, financial institutions require separate treatment. The pure form of this tax leaves out financial flows entirely. An augmented form of the tax can capture financial transactions in the base, but this would introduce complexity as all companies would need to keep track of financial transactions, as well as whether the transactions occurred with foreign companies. There is also substantial ambiguity between what transactions are real and what are financial, and such ambiguity raises both technical considerations as well as opportunities for tax avoidance.
Fourth, there are likely to be important impacts on state government corporate tax systems, and these have also not been carefully considered. Fifth, there are large transition effects associated with moving to a destination-basis cash flow system that would need to be carefully considered.
An essential problem with the Ryan Blueprint concerns the tax rates that were chosen. These very low tax rates make the system likely to lose a large amount of revenue in a regressive manner.
Indeed, the corporate rate chosen is intellectually incoherent. One of the purported advantages of a destination-basis corporate cash flow tax is that it is supposed to curb profit shifting by removing the incentive for shifting profits and activities abroad. But, if that is the case, why is the rate cut needed? If tax burdens truly depend only on the location of immobile customers, why not keep the corporate rate at the same level as the top personal rate? The usual argument for the lower rate relies on the international mobility of income and competitiveness concerns. If such concerns are moot, then there is no reason to tax at a low rate.
Further, the discrepancy between the top personal rate and the business rate will create new avoidance opportunities as wealthy individuals seek to earn their income in tax-preferred ways, reducing their labor compensation in favor of business income. Companies would be inclined to tilt executive compensation toward stock options and away from salary income, and high-income earners would be inclined to earn income through their businesses in pass-through form.
The Ryan proposal exempts the normal return from capital, giving these returns zero-tax treatment. Further, excess returns (profits above the normal level) are taxed through the business tax system, but at rates far lower than the top personal income tax rate. The theoretical rationale for justifying such a favorable tax treatment for rents (excess profits) is simply absent. From an efficiency or an equity perspective, taxing rents at a higher rate makes sense.
Recent evidence from Treasury suggests that now about 75% of the corporate tax base is rents/extra-normal profits; this fraction has been steadily increasing. If destination-based taxes are meant to fall solely on rent, this implies a higher ideal optimal tax rate, since taxing rents is far more efficient than taxing labor or capital.
Further, the regressive nature of these tax changes is unjustifiable given the increases in economic inequality over the previous decades and the large surge in the share of income earned by the top 1% of the income distribution. Capital income, and rents, are far more concentrated than labor income. Cutting taxes on capital and rents so dramatically risks further exacerbating recent increases in income inequality.
The Tax Policy Center calculates the distributional effect of the Ryan plan, which benefits the wealthy disproportionately. The average federal tax rate falls by about 0.4 percentage points for the bottom 80% of the population, but it falls by 3.4 percentage points for the top quintile, and by 9 percentage points for the top 1%. The top 1% receive a tax cut that averages $213,000. The tax cut of the bottom 80% averages $210.
Finally, the Ryan proposal loses large amounts of tax revenue. The business tax features of the proposal are a large share of the ten-year, $3 trillion revenue loss, according to the Tax Policy Center. Prior research by Auerbach suggests that this type of corporate tax reform would not change revenue very much at the same corporate tax rate, and work by Devereux has suggested that the tax base would be smaller under a DBCT, but that this could be compensated for by higher rates. Under the Ryan plan, however, the rate is much lower, leading to large deficits.
The Ryan Blueprint destination-based cash-flow tax is not ready for prime-time. No other country had adopted a similar tax, and as the above analysis makes clear, there are myriad issues that would need to be worked through before any such tax were adopted. These issues are not small: the plan is incompatible with trade rules in a manner which harms our trading partners, it is incompatible with our treaty obligations, it is unlikely to end income shifting, it generates political problems due to large numbers of companies that would experience adverse tax treatment changes, it makes the tax system less progressive at the proposed tax rates, and it is likely to generate large revenue losses. In addition, there are important issues surrounding how exporters with losses would be handled (which could lead to inefficient mergers), how financial firms and financial transactions would be handled, how U.S. state corporate tax systems would be affected, and how the transition to the new tax system would be handled.
One pressing problem is that the Ryan blueprint is incompatible with WTO rules. And this incompatibility is no mere technicality. U.S. trading partners are likely to be hurt in several ways. The effects of the wage deduction render the corporate cash-flow tax different from a VAT, and these differences have the net effect of increasing the incentive to operate in the United States, as both proponents and economists recognize. In addition, such a tax system would exacerbate the profit-shifting problems of our trading partners, since the United States will appear like a tax haven from their perspective. If multinational firms shift profits to the United States on paper, this will reduce foreign revenues without affecting U.S. revenues.
While economists have argued that exchange rate changes may reduce trade-distorting effects of such tax law changes, there are several reasons to suspect that such exchange rate changes will not be sufficient to neutralize the effects of such a tax law change. First, exchange rate changes are uniform, yet the export subsidy component of the DBCT plan would treat different firms differently. Second, exchange rate markets are very large, exchange rate movements are not well predicted by economic fundamentals, and many countries fix their exchange rates, all factors that would reduce hopes of smooth countervailing exchange rate adjustment. Third, exporting firms may receive incomplete loss offsets, and that would cause trade distortions.
However, even if these economic effects were disregarded, it is clear that the DBCT is on shaky legal ground with respect to both WTO rules and our tax treaties. The WTO is likely to recognize that this DBCT is non-equivalent to a VAT, and thus a direct tax, where border adjustments are not allowed. This will likely lead to years of litigation and perhaps an endgame whereby the DBCT is simply jettisoned in favor of a VAT. This would convert one of the most progressive tax instruments in our tax system into a regressive consumption tax. In the meantime, we are likely to face the prospect of large tariff retaliations by our trading partners, in an environment where the incoming U.S. administration has already provided ample reason to fear trade wars.
Given these concerns, we would recommend that Congress reject the Ryan Blueprint. Instead, it should focus on a revenue-neutral tax reform that reduces the corporate tax rate and eliminates the major corporate tax expenditures including deferral, taxing accumulated offshore earnings in full. Eliminating deferral would eliminate the incentive to earn income in low-tax countries, by treating foreign and domestic income alike for tax purposes. Pairing that reform with a lower corporate tax rate need not raise tax burdens on average, although it would create winners and losers among corporate taxpayers. A more fundamental reform would require worldwide corporate tax consolidation; this would better align the tax system with the reality of globally-integrated corporations.
Taxing foreign income currently also eliminates the incentive to build up large stocks of unrepatriated foreign income, now estimated at $2.6 trillion. This income is often invested in U.S. capital markets, and it increases the creditworthiness of U.S. multinational corporations, who can easily finance worthy investments. But corporations are inhibited from repatriation by the prospect of more favorable tax treatment if they delay, so this makes it difficult for them to return profits to shareholders. Indeed these concerns about repatriation are likely to give the multinational business community a large interest in corporate tax reform. Settling the future tax treatment of foreign income should be a key goal of these efforts.
In terms of more incremental reforms, even a per-country minimum tax would be a big step toward reducing profit shifting toward tax havens and protecting the corporate tax base. A minimum tax would currently tax income earned in the lowest tax countries, and work by Clausing suggests that 98% of the profit shifting out of the United States is destined for countries with foreign tax rates below 15%. Other helpful incremental steps include stronger “earnings-stripping” rules and anti-corporate inversion measures such as an exit tax.
 A Better Way: Our Vision for a Confident America, GOP Tax Reform Task Force (June 24, 2016), http://abetterway.speaker.gov/_assets/pdf/ABetterWay-Tax-PolicyPaper.pdf [perma.cc/G6B3-YMT3].
 Id. at 15 (emphasis added).
 See Leonard E. Burman, James R. Nunns, Benjamin R. Page, Jeffrey Rohaly & Joseph Rosenberg, An Analysis of the House GOP Tax Plan, 8 Colum. J. Tax L. 257 (2017) (estimating that the Blueprint would decrease revenue and increase the debt by $3 trillion over the first decade).
 The Tax Policy Center analysis mentions that there would likely be large enforcement problems with these rules, especially given the large rate differential under the plan. See Burman et al., supra note 3. Nonetheless, they assume that the rules would be enforceable in their revenue analysis. The plan would lose even more revenue absent that assumption.
 As explained below, if the Blueprint proposal reduced profit shifting opportunities as its proponents believe, it is not clear why a rate cut is indicated since the main rationale to cut corporate tax rate is reducing base erosion and profit shifting (BEPS).
 A Better Way, supra note 1, at 25.
 Id. at 26.
 Id. at 27 (emphasis added).
 The base company rule, I.R.C. § 954 (2015), provides that selling goods or services through a “base company” in a low-tax jurisdiction triggers U.S. tax to the parent, and I.R.C. § 956 (2007) provides that using income otherwise eligible for deferral to invest in U.S. property (including a loan to the parent) triggers U.S. tax to the parent. The latter rule has been under pressure recently because of the $2.5 trillion in deferred income of foreign subsidiaries of U.S. parents located in low-tax jurisdictions.
 A Better Way, supra note 1 at 28 (emphasis added).
 Id. (emphasis added).
 Simple, Fair, and Pro-Growth: Proposals to Fix America’s Tax System, Report of the President’s Advisory Panel on Federal Tax Reform (Nov. 2005), http://www.treasury.gov/resource-center/tax-policy/Documents/Report-Fix-Tax-System-2005.pdf [perma.cc/V6HP-NXPW] at 172.
 Burman et al., supra note 3.
 Agreement on Subsidies and Countervailing Measures, 1869 U.N.T.S. 14 [hereinafter SCM].
 In addition, it is likely that the Blueprint would constitute prohibited discrimination against imports and in favour of domestic production under Article 3 of the GATT, because foreign businesses exporting to the U.S. would be pressed to move production to the U.S. in order to get a deduction for wages. This is particularly true for manufacturing units in developing countries, where you do not have sufficient local sales to compensate with. See General Agreement on Tariffs and Trade, Oct. 30, 1947, 61 Stat. A-11, 55 U.N.T.S. 194 [hereinafter GATT].
 SCM, supra note 16 (emphasis added).
 A subtraction method VAT is a cash-flow tax that includes all sales but allows a deduction for all outlays, except for interest and wages. In principle, it has the same tax base as the normal invoice-credit VAT, as adopted by most countries. In an invoice-credit VAT, tax is paid at each stage of production on the sale price of outputs, with a credit given for tax on inputs. Both methods can be origin or destination-based, but all existing VATs are destination-based (imports are taxed and exports are exempt). The main difference is administrability: In an invoice-credit VAT, no credit is given unless tax was paid on the input, as shown on an invoice. In a subtraction method VAT, care must be taken not to allow a deduction unless there is a corresponding inclusion by the provider of goods, services or intangibles. This difference explains why no country has adopted a subtraction method VAT. The Blueprint proposal is based on a subtraction method VAT, but with a deduction for wages.
 For similar conclusions, see Wolfgang Schön, Destination-Based Income Taxation and WTO Law: A Note (Max Planck Inst. for Tax Law and Pub. Fin., Working Paper No. 03, 2016), http://ssrn.com/abstract=2727628 [perma.cc/T5M3-TNDF]; Wei Cui, Destination-Based Cash-Flow Taxation: A Critical Appraisal (Sept. 30, 2015) (unpublished manuscript), http://ssrn.com/abstract=2614780 [perma.cc/J6GW-QNMQ]; Wei Cui, Destination-Based Taxation in the House Republican Blueprint, 173 Tax Notes Today 7 (2016); Lee A. Sheppard, News Analysis: Freedom Fries: The House Republicans’ Cash Flow Tax (Section 954 – Foreign Base Company Income), 157 Tax Notes Today 1 (2016).
 In this example, we assume that the tax gets passed on to consumers in the form of higher prices.
 Under the Blueprint, Net Operating Losses (NOLs) are carried forward with an interest charge, rather resulting in an actual refund, but the end result should be the same. Still, many exporters may never show positive income under this tax system, so they may not be able to use NOLs.
 See, e.g., Alan J. Auerbach & Douglas Holtz-Eakin, The Role of Border Adjustments in International Taxation, Am. Action Forum (Nov. 30, 2016), http://www.americanactionforum.org/research/14344/ [perma.cc/YY2W-GJ6F].
 Kenneth Rogoff, Perspectives on Exchange Rate Volatility, International Capital Flows 441-53 (Martin Feldstein ed.,1999).
 Id. at 444.
 See Triennial Central Bank Survey: Foreign Exchange Turnover in April 2016, Bank for Int’l Settlements (Sept. 2016), http://www.bis.org/publ/rpfx16fx.pdf [perma.cc/2NVK-UAVP].
 The borrowing that occurs from abroad is the “flip side” of the trade deficit. In particular, basic national income accounting indicates that EX-IM (the trade balance) must always equal the sum of the private savings/investment balance (S-I) and the government budget balance of tax revenues relative to government spending (T-G). In the case of the United States, our trade balance is often negative since our savings (S) fall short of demand for loanable funds due to private Investment (I) and government borrowing (G-T).
 In the IMF survey of exchange rate regimes, they classify only 29 of 191 countries as freely floating. Other countries have other arrangements, including fixed rates, currency boards, and more actively “managed” floats. See Annual Report on Exchange Arrangements and Exchange Restrictions, Int’l Monetary Fund (2014), http://www.imf.org/external/pubs/nft/2014/areaers/ar2014.pdf [perma.cc/L4H7-VXZG]
 Auerbach & Holtz-Eakin, supra note 22.
 See, e.g., Alan Viard, Border Tax Adjustments Won’t Stimulate Exports, Am. Enter. Inst. (Mar. 2, 2009), http://www.aei.org/publication/border-tax-adjustments-wont-stimulate-exports/ [perma.cc/65VB-695H].
 See Michael Graetz, The Known Unknowns of the Business Tax Reforms Proposed in the House Republican Blueprint, 8 Colum. J. Tax L. 117 (2017).
 Mihir Desai & James Hines, Value-Added Taxes and International Trade: The Evidence (Nov. 2002) (unpublished manuscript), available at http://pdfs.semanticscholar.org/0245/59563b9d1470c5932a0b858bb9153ab750df.pdf [perma.cc/5SCZ-E7U5].
 Under the Blueprint, NOLs are carried forward with an interest charge. This may or may not result in an eventual payment. Many exporting firms may not ever show a taxable profit under this system.
 A final bit of evidence comes from de Mooij & Keen, discussed within Auerbach, et al., Destination Based Cash-Flow Taxation 20-21 (Oxford U. Ctr. for Bus. Tax’n, Working Paper No. 17/01, 2017), http://eml.berkeley.edu/~auerbach/CBTWP1701.pdf [perma.cc/YVH9-QXL5]. There is evidence that Eurozone countries that pursue fiscal devaluations (a combination of increased consumption taxes and reduced business taxes on labor that is similar to a destination-based cash-flow tax) do see a noticeable short-run increase in net exports. See Ruud de Mooij & Michael Keen, ‘Fiscal Devaluation’ and Fiscal Consolidation: The VAT in Troubled Times (NBER, Working Paper No. 17913, 2012), http://www.nber.org/papers/w17913.pdf [perma.cc/V7K4-5L9G]. This working paper also appears in Fiscal Policy After the Crisis, U. of Chicago Press 443-85, (Alberto Alesina & Francesco Giavazzi eds., 2013).
 Auerbach & Holtz-Eakin, supra note 22.
 The real reason, one suspects, is the widely-held belief in the political implausibility of enacting a VAT in the United States. Given the WTO issue facing any border-adjusted tax that is not a VAT, this belief may be misguided. See Reuven Avi-Yonah, The Inexorable Rise of the VAT: Is the U.S. Next?, 150 Tax Notes 127 (2016).
 Auerbach & Holtz-Eakin, supra note 22 at 12-14.
 See, e.g., Chad P. Bown, WTO-consistency of the BTA and potential retaliation, Presentation at the Peterson Institute for International Economics (Feb. 1, 2017), in Peterson Inst. for Int’l Econ., http://piie.com/system/files/documents/bown20170201ppt.pdf [perma.cc/43ML-R27K].
 See Sheppard, supra note 19, at 914.
 Reuven Avi-Yonah, International Taxation of Electronic Commerce, 52:3 Tax L. Rev. 507 (1997).
 See Sheppard, supra note 19 at 909-10.
 See Reuven Avi-Yonah, The International Implications of Tax Reform, 69 Tax Notes 913 (1995); Reuven Avi-Yonah, From Income to Consumption Tax: Some International Implications, 33 San Diego L. Rev. 1329 (1996).
 A Better Way, supra note 1 at 26.
 For the serious problems raised by application of VAT to cross-border trade in services and intangibles, see International VAT/GST Guidelines, OECD (Nov. 2015), http://www.oecd.org/tax/consumption/international-vat-gst-guidelines.pdf [perma.cc/8TQB-885Y] (recommended by the Council in September 2016). In an invoice credit VAT, exports are zero-rated in the country of origin, so a business importer does not get a tax credit on the purchase. If there is an output tax to the final consumer, it is simply charged and paid (like a typical retail sale under the U.S. RST). This means that, unlike the typical VAT situation, the entire collection even in a B2B context depends on the final sale to the consumer, and experience with retail sales taxes has illustrated that at high rates this becomes an avoidance problem (as anyone living in states that border states that do not tax sales can attest). The real problem in the B2C domain is simply that there is no jurisdiction to enforce the B2C tax, because there is no jurisdiction over the remote supplier. In the B2B context, the answer is the reverse charge, where the business purchaser self-assesses the tax and therefore gets an input tax credit on any further sale. In the B2C context, relying on the consumer to self-assess the tax amounts to a tax on honesty (like the U.S. state use tax where there is no collection by the remote seller). In general, determining exports and imports and tracking purchases of those engaged in cross-border business is not trivial. It is difficult to judge where services are consumed and to trace location of downloaded services.
 Thanks to David Miller for suggesting these. For further elaboration, see David S. Miller, How Donald Trump Can Keep His Campaign Promises, Grow the Economy, Cut Tax Rates, Repatriate Offshore Earnings, Reduce Income Inequality, Keep Jobs in the United States, and Reduce the Deficit (Tax Forum Paper No. 680, 2017), available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2908207 [perma.cc/2K9H-5KDT].
 As Miller argues, this example suggests that inversions would in some cases still be valuable. Moreover, to the extent the Blueprint retains Subpart F, inversions can be helpful in avoiding it. For example, if an Irish subsidiary of a U.S. parent licenses intangibles to consumers in the U.S. and because it is difficult to enforce the tax on the consumers the IRS relies on Subpart F to tax the royalties, this rule (which is included in what remains of Subpart F in the Blueprint) can be avoided by an inversion.
 Auerbach & Holtz-Eakin, supra note 22.
 For a more detailed treatment of these complex issues, see David Weisbach, A Guide to the GOP Tax Plan – The Way to a Better Way, 8 Colum. J. Tax L. 171 (2017). See also Auerbach, et al., supra note 34.
 Absent relief, consumption taxes generate a tax on the initial capital stock; while this is an efficient tax (since it is an unexpected lump sum tax on the capital stock), it is arbitrary. However, attempts to provide relief would be expensive and would reduce the progressivity of the tax system, since the capital stock is concentrated in the upper part of the income distribution. See Weisbach, supra note 48.
 See Laura Power & Austin Frerick, Have Excess Returns to Corporations Been Increasing Over Time?, 69 Nat’l Tax J. 831, 831-46 (2016).
 Also note that double-taxation arguments are vastly overstated since about three-quarters of U.S. corporate equity income is not taxed at the individual level. See Steven M. Rosenthal & Lydia S. Austin, The Dwindling Taxable Share of U.S. Corporate Stock, 151 Tax Notes 923, 923-34 (2016); Leonard Burman & Kimberly Clausing, Is U.S. Corporate Income Double-Taxed? 70 Nat’l Tax J. (forthcoming, Sept. 2017) (on file with authors).
 The U.S. Treasury reports that the top 5% of tax units report 24% of income in 1986 (the earliest year available), increasing to 37% in 2012. See Internal Revenue Serv., SOI Tax Stats – Individual Statistical Tables by Tax Rate and Income Percentile (Aug. 31, 2016), http://www.irs.gov/uac/soi-tax-stats-individual-statistical-tables-by-tax-rate-and-income-percentile [perma.cc/G99F-W2E4]. Indeed, capital income is much more concentrated that labor income. Data from the Tax Policy Center for 2012 indicate that the top 5% of tax units report 68% of dividend income and 87% of long-term capital gains income. T09-0492 – Distribution of Long-Term Capital Gains and Qualified Dividends by Cash Income Percentile, 2012, Tax Policy Ctr. (Dec. 10, 2009), http://www.taxpolicycenter.org/model-estimates/distribution-capital-gains-and-qualified-dividends/distribution-long-term-capital-2 [perma.cc/Y4D8-FVSG]. The U.S. Treasury also reports data on the top 400 taxpayers. This particularly small group of taxpayers reports 1.48% of total income in 2012, but 0.16% of total wage and salary income, 8.3% of total dividend income, and 12.3% of total capital gain income. Internal Revenue Serv., The 400 Individual Income Tax Returns Reporting the Largest Adjusted Gross Incomes Each Year, 1992–2013 (Dec. 2015), http://www.irs.gov/pub/irs-soi/13intop400.pdf [perma.cc/ZA26-XM7X]. The overall share of this tiny group has more than doubled since 1992 (when the data series begins). The wage income share has been flat, while the capital gains share has more than doubled, and the dividends share has more than quadrupled. All of these trends occur within a context where the labor share of income is falling relative to the capital share of income. For more discussion of these trends, see Kimberly Clausing, Strengthening the Indispensable U.S. Corporate Tax,” Wash. Ctr. for Equitable Growth (Sept. 12, 2016), http://equitablegrowth.org/report/strengthening-the-indispensable-u-s-corporate-tax/ [perma.cc/DQE8-WUMJ].
 Burman et al., supra note 3, at 273 (Table 4).
 Some recent work by Treasury economists has a somewhat more optimistic view of the tax base under this type of tax system, but the additional size of the base comes entirely from the border adjustment. See Elena Patel & John McClelland, What Would a Cash Flow Tax Look Like for U.S. Companies? Lessons from a Historical Panel (Office of Tax Analysis, Working Paper No. 116, 2017). It is important to note that the revenue from the border adjustment is contingent on the U.S. being in a trade deficit position. Since trade deficits (and the associated financial borrowing) eventually have to be paid back in the form of trade surpluses, these revenue gains are really being borrowed from future U.S. taxpayers. See Alan Viard, The Border Tax Adjustment Can’t Make Mexico Pay for the Wall, AEI, (Jan. 27, 2017, 10:45am), http://www.aei.org/publication/the-border-tax-adjustment-cant-make-mexico-pay-for-the-wall/ [perma.cc/T8Q2-TEKD].
 Even if the dollar appreciates fully, there are still serious concerns. Exchange rate appreciation generates its own risks to the world economy, generating serious financial vulnerability in a large number of countries that have substantial dollar liabilities. Dollar appreciation will also result in a multi trillion-dollar deterioration in the U.S. net international investment position, a large wealth redistribution away from U.S. holders of foreign assets and toward foreign holders of U.S. assets.
 Toward this end, the U.S. Congress did a great disservice when they enacted a one-time holiday on dividend repatriation as part of the American Jobs Creation Act of 2004. American Jobs Creation Act of 2004, Pub. L. No. 108-357, 118 Stat. 1418 (2004). Ever since, companies have been more likely to delay repatriation in the hope of future holidays (or permanently more favorable treatment).
 See Kimberly Clausing, The Effect of Profit Shifting on the Corporate Tax Base in the United States and Beyond, 69 Nat’l Tax J. 905, 905-34 (2016).