Evaluating the Oxford Proposal for a Corporate Cash Flow Tax

After almost a century, the international system for taxing cross-border business income has entered a period of instability and change.¹ Following several years of study, the distinguished Oxford International Tax Group recently proposed a fundamental reform: The traditional corporate income tax would be replaced by taxation of a corporation’s domestic cash flows.² The rationales for the proposal include both efficiency (only pure profits above the normal rate of return would be taxed) and administrability (shifting accounting profits abroad would not reduce taxes).

Another, perhaps surprising, argument for this proposal is that its incidence would fall primarily on a country’s residents who own shares in domestic and foreign companies. For example, equilibrating changes in floating exchange rates would transform a U.S. tax on U.S. sales by U.S. and foreign corporations into a tax on U.S. shareholders of U.S. and foreign corporations, wherever their sales occurred. The Oxford group indicates that the resulting tax burden is very likely to be progressive.³

If, however, a progressive tax borne by individual owners of corporate stock is desirable, why not tax shareholders directly? The structure of such a tax would be similar to familiar U.S. provisions for retirement savings: Investments by individuals in domestic and foreign companies would be deductible, while disinvestments would be taxable. The Oxford group notes this possibility⁴ but does not examine such a tax to compare its strengths and weaknesses with those of its own proposal.

The purpose of this article is to illustrate the equivalence of the two taxes with a simple numerical example and to stimulate their comparison. If the two taxes are equivalent in effect, there might be reasons to favor one method of implementation over another. We begin with a brief review of the structure of the Oxford group’s proposal. In the interest of simplicity, we follow the authors’ decision to set aside questions of scope, such as the types of business organizations that would be subject to the tax.⁵ Accordingly, this discussion focuses on corporations and individual shareholders, but a broader implementation could include all businesses and their owners. We also put aside transition issues, such as the effects on different forms of existing capital, in order to focus on the equivalence argument.

Flow Tax Without Cross-Border Transactions

Consider first a corporation with no cross-border transactions. Under a comprehensive cash flow tax, the corporation would include all receipts and deduct all payments, other than the tax itself and payments to and from shareholders. The tax base would include real (R) and financial (F) flows, so the Meade Committee’s classic exposition deemed it an R + F base.⁶ The two most important differences between an R + F base and a traditional corporate income tax are that (1) capital expenditures would be immediately deductible, and (2) borrowings and principal repayments would be included and deducted, respectively.

The Meade Committee also discussed an R-base version of the cash flow tax, under which financial flows would be ignored.⁷ The two most important differences between an R base and a traditional corporate income tax are that (1) capital expenditures would be immediately deductible, and (2) interest payments and receipts would not be taken into account. In the years since the Meade Committee, there have been many proposals for company cash flow taxes, including a recommendation of the 2005 President’s Advisory Panel on Federal Tax Reform for an R base.⁸

With Cross-Border Transactions

Now consider international business transactions, which are much more important today than at the time of the Meade Committee. Following the previous work of some of its members,⁹ the Oxford group proposes a destination-based cash flow tax, which could be levied using either an R + F base or an R base. Under the simplest version of this approach, cross-border payments and receipts would be ignored. The United States would accordingly tax domestic and foreign companies on their U.S. receipts minus their U.S. costs. Value added taxes typically operate on a destination basis, so the major difference between the Oxford group’s R base and a subtraction-method VAT is that the former would allow the deduction of wages.¹⁰ As in discussions of VATs, the nontaxation of foreign sales and the nondeductibility of foreign costs are deemed “border tax adjustments.”¹¹

A destination-based tax (whether valued added or cash flow) is sometimes supported as a means to encourage exports (because foreign sales are not taxed) and discourage imports (because foreign costs are not deductible). Economists generally reject this argument on the grounds that floating exchange rates (or other prices) would adjust to offset the claimed effects on exports and imports.¹² This adjustment, which is central to the Oxford group’s analysis of incidence, is illustrated in the following example.

Example 1 (adjustment in exchange rates)¹³: Suppose that the EU sells one unit of a good in the United States at a price of $100 when there is no taxation in either jurisdiction. Similarly, the United States sells one unit of a good in the EU at a price of €100. In the interest of simplicity, there are no production costs. The exchange rate is $1 = €1.

What will happen if a destination-based cash flow tax is introduced in the United States with a tax rate of 20 percent? EU producers will see their revenues drop to $80 if they sell in the United States rather than the EU. U.S. producers will have nontaxable income of €100 if they export rather than sell domestically. Imports into the United States would therefore seem to be discouraged, while exports from the United States would seem to be encouraged.

If, however, U.S. imports decline while U.S. exports increase, there would be less demand for the euro by U.S. nationals (who are reducing their imports of the EU product) and more demand for dollars by EU nationals (who are increasing their imports of the U.S. product). As a result, the value of the dollar should rise relative to the euro until there is no longer any advantage to switching from EU products to U.S. products.

Suppose the value of the dollar rose until $1 could be exchanged for €1.25. The EU exporter would then receive €100 for its after-tax revenues of $80, eliminating the disincentive to sell into the United States. At the same time, U.S. exporters would see their revenues diminish, as €100 in receipts would produce only $80, eliminating the incentive to export because $100 of domestic sales would also produce $80 after taxes.

Although economists emphasize that floating exchange rates should adjust to eliminate any advantage to exports and detriment to imports, the business community has not always been convinced. The failure in 2017 of a proposed destination-based cash flow tax in Congress was due in part to opposition from major retailers, which were not persuaded that the loss of a tax deduction for imported goods would be fully offset by a reduction in the price they would pay for those goods caused by an increase in the value of the dollar.¹⁴

Equivalent Taxes

By definition, the sources of a corporation’s net funds (business inflows minus business outflows) must equal the uses of those funds (tax payments and net transfers to shareholders). The Meade Committee accordingly concluded that in a domestic context, an R + F cash flow company tax is identical to a tax on the net cash flows to shareholders (the excess of dividends, redemptions, and liquidations over share purchases), which it deemed an S base.¹⁵ Under such a tax, which could be levied on either the corporation or its shareholders, investments by individuals in company shares would be deductible, while disinvestments would be taxable. The result is a regime that is very similar to the taxation of traditional (non-Roth) retirement savings in the United States.

In the international context, the Oxford group states that its proposed company tax on domestic cash flows of domestic and foreign firms is equivalent to a tax on the net cash flows of domestic owners of firms worldwide.¹⁶ A destination-based cash flow company tax would therefore be equivalent to an S-base tax levied on investments by a country’s residents in both domestic and foreign firms.

From a U.S. perspective, a cash flow tax levied on U.S. sales by U.S. and foreign firms would be equivalent to a cash flow tax on U.S. owners of U.S. and foreign firms, wherever the companies’ sales occurred. This equivalence is central to the Oxford group’s conclusion that a destination-based corporate cash flow tax would be progressive.

The equivalence of the two taxes is not obvious. Consider two cases that would seem to refute the equivalence. In the first, a foreign individual owns a foreign firm that exports to the United States. A U.S. destination-based corporate cash flow tax would apply to the U.S. import, but an S-base tax on distributions to U.S. shareholders would not be applicable because the owner of the foreign firm is foreign. In the second case, a U.S. individual owns a U.S. firm that exports to a foreign country. A U.S. S-base tax would apply to distributions from the U.S. firm to the U.S. owner, but a U.S. destination-based corporate cash flow tax would not be applicable, as exports are exempt. While accurate, these initial observations are incomplete because the cross-border trade in each case would in principle be balanced by other transactions with U.S. tax consequences that must be taken into account.

The Oxford group offers two ways to understand the equivalence of the taxes. The first, which is primarily for economists, refers the reader to a paper by two of its members.¹⁷ That paper models cash flow company taxes and invokes the household budget constraint to conclude that the destination-based version is equivalent to a tax on the pure profits (above-normal returns) received by domestic residents. The second is a comparison with value added taxation. Putting aside preexisting capital, the incidence of a destination-based VAT is on consumption by a country’s residents out of labor income and economic rents (above-normal returns).¹⁸ Because the deduction of wages under a destination-based cash flow company tax offsets the burden on consumption out of labor income, the remaining burden must be on consumption out of economic rents of the residents.

While helpful, neither explanation provides a numerical example that illustrates the equivalence in operation. It may be useful to provide a schematic illustration by expanding an example discussed by the Oxford group to include the two cases above that seemed to refute the equivalence.¹⁹

Example 2 (equivalence of destination-based cash flow tax and S-base taxes): A foreign firm owned by a foreign individual exports a consumer product to the United States, where it has no costs, and the product sells for $100. There is no relevant tax in either country. The exchange rate is $1 = 1F, with F being the currency of the foreign country. The proceeds of $100 are distributed to the foreign owner, who uses the distribution to import a consumer product sold by a U.S. company owned by a U.S. individual. The U.S. company has no expenses in the current year. Its proceeds of $100 are distributed to the U.S. owner, who uses the distribution for consumption in the United States. There are no other international trade transactions, so imports and exports balance in the current year.

In the absence of taxation, the foreign owner of the foreign company in Example 2 has 100F of foreign consumption, while the U.S. owner of the U.S. company has $100 of U.S. consumption.

Suppose that the United States adopts an S-base tax at a rate of 20 percent.²⁰ That tax would not apply to the foreign owner who bought the U.S. consumer product. On the other hand, the distribution of $100 by the U.S. company to its U.S. owner would trigger a U.S. tax of $20, reducing U.S. consumption of the U.S. owner to $80.

Now suppose instead that the United States adopts a destination-based company cash flow tax at the same rate. As in Example 1, the exchange rate would in principle adjust to $1 = 1.25F. The foreign-owned firm would remit $20 to the U.S. treasury on its U.S. sales of $100, leaving $80 for the foreign owner. As the Oxford group emphasizes,²¹ the foreign owner would not bear the burden of the U.S. tax because the $80 distribution would still be worth 100F. Moreover, with 100F, the foreign owner could still purchase a product selling for $100 in the United States because exports are exempt from the destination-based cash flow tax.²²

Who then bears the burden of the $20 in tax that the foreign-owned company remits to the U.S. treasury? The Oxford group indicates that it would be U.S. residents,²³ but which residents? Consider the position of the U.S. firm. It does not pay tax on its exports but collects only $80 from the foreign purchaser because of the exchange-rate adjustment. The U.S. consumption of the U.S. owner is accordingly reduced from $100 (as in the no-tax case) to $80 (as under the S-base tax). The U.S. owner of the U.S. firm bears the burden of the U.S. tax paid by the foreign firm, confirming the equivalence of the two taxes, given the exchange-rate adjustment.

As noted above, this example is schematic rather than realistic. Exports and imports of goods and services are unlikely to balance in any given year. Corporate capital structures are certainly likely to be more complex than in the example. Given modern supply chains, diverse capital flows, and plausible capital structures, tracing the effects on U.S. residents of a destination-based cash flow tax remitted to the U.S. treasury by a foreign company in a more realistic example would be impossibly labyrinthine.

On the other hand, the example illustrates the logic behind the Oxford group’s conclusion that the two taxes are equivalent. In particular, it highlights the importance of two standard assumptions for the equivalence: the adjustment in exchange rates (or other prices) and the constraint that a country’s exports and imports must be equal in present value. Although not illustrated in the example, it is worth noting that the incidence of the company tax on U.S. shareholders in foreign companies would also be achieved by the exchange-rate adjustment. As Alan Auerbach, a member of the Oxford group, has put it, an R + F destination-based company cash flow tax accomplishes the same outcome as a tax on distributions received by domestic shareholders by “working through different markets (trade in goods and services versus capital and income flows), via the border adjustment.”²⁴

Shareholder Cash Flow Taxation

If taxing individual shareholders on the cash flows produced by their worldwide stock holdings is desirable, why not do it directly? As indicated earlier,²⁵ the Oxford group notes the possibility of an S-base tax but does not examine such a tax to compare its strengths and weaknesses with those of its own proposal. If one accepts the Oxford group’s argument that the incidence and overall economic effects of the ideal versions of the two taxes are the same, additional factors, such as administrability, taxpayer understanding, and consistency with other components of the tax system would be relevant to such a comparison. Only a few of the pertinent factors can be sketched here, with emphasis on some that support serious consideration of the shareholder taxation alternative.

Regarding administrability, the Meade Committee thought an S base would be simpler than an R + F base in a domestic context because there would be fewer relevant transactions to track.²⁶ Given increased investment abroad, shareholder cash flow taxation would today require Treasury to keep track of distributions from foreign companies to U.S. shareholders.²⁷ That burden would, however, be mitigated by the incentive for U.S. investors to register their foreign shares at the time of purchase to obtain the benefits of deductibility. Also, the United States has been willing in recent years to expand foreign reporting.²⁸

Regarding taxpayer understanding, a significant weakness of the destination-based corporate cash flow tax is surely its reliance on an adjustment in exchange rates (or other prices) to accomplish the desired incidence. Given the 2017 congressional experience, the difficulty of convincing affected taxpayers that the adjustment will operate in practice as well as it does in theory cannot be ignored.²⁹ A direct shareholder tax would avoid those issues because it would not depend on a change in prices to accomplish the desired incidence. One drawback of the S base for the Meade Committee was its unfamiliarity.³⁰ Today, a tax providing that stock investments are fully deductible, while disinvestments are fully taxable, is unlikely to surprise Americans familiar with qualified pension plans, traditional IRAs, and section 401(k) investments.

Finally, consider the need to integrate the proposals with other parts of the tax system, such as individual taxation of dividends, capital gains, and interest. Focusing on shareholder-level taxation, as under an S base, would make the need to address those issues obvious. The Oxford group takes the position that its entity-level proposal is desirable on its own merits, so the appropriate interaction with individual taxation of capital income need not be specified.³¹ Once the equivalence of the proposal with shareholder cash flow taxation is accepted, this position seems difficult to maintain.

Needless to say, these sorts of issues cannot be resolved or even fully identified in a short article such as this. At this stage, it is simply not possible to say whether a destination-based corporate cash flow tax or an individual shareholder cash flow tax would be the preferable means of implementing the taxes’ common goals. Under the rubric of “economically equivalent reforms,” the Oxford group provides a comparison of its proposal with a VAT accompanied by a tax subsidy for labor income.³² Given the importance accorded the progressive incidence of the proposal, a detailed comparison of a destination-based cash flow tax with the equivalent shareholder cash flow tax would also be helpful.