In defense of requiring back-flips.

In his article, In Defense of Tax Shelters, (1) Leo Katz lives up to his usual standard of being iconoclastic, provocative, illuminating, and fun to read. The article challenges the conventional wisdom with respect to tax shelters, and I believe such challenges are worth making. Whether or not they prove persuasive, they can help to clarify everyone's thinking. That being said, I am not personally persuaded by the article's analysis. In particular, I believe it mistakenly assumes that anti-tax shelter rules must take the approach of identifying disallowed tax planning that is somehow worse than permissible tax planning. To the contrary, I believe that socially optimal rules (given all of the underlying constraints) might actually do the opposite, by using filters that end up catching what are in some sense the less bad transactions. This reflects that the rules' proper aim is not to make ethical judgments about or between tax-motivated transactions, but to minimize overall social harm from deadweight loss.

Professor Katz starts his article with a quotation from Myron Scholes that comes within shouting distance of stating the purpose of anti-tax shelter rules, but then stops short. Specifically, as quoted by Katz, Scholes says:

Although the deadweight costs associated with time spent in tax

planning may seem socially wasteful, the relevant question is how much

waste would exist using alternative means to achieve the same social

goals. In other words, how does the net benefit of the altered

economic activity brought about by the tax system compare with the

net benefits of the next best alternative? Obviously, if we could

implement social policy through a mechanism that would result in zero

waste, we would do so, but such a goal is not realistic. (2)

Up to a point, Scholes shows genuine insight here. He understands the importance and social undesirability of deadweight loss from tax planning. The only problem is that he takes a straw man approach, implicitly treating the loss as fixed so he can imply the further statement: "Wouldn't it be nice if we could reduce it to zero, but we can't, so tough luck."

One more step and Scholes would have been there. He could have asked: Can tax rules reasonably be designed to reduce the deadweight loss from tax planning? This is what I believe anti-tax shelter rules are and should be all about.

Professor Katz appears to assume a different objective function for anti-tax shelter rules. He describes the consensus view about tax shelters, which he is criticizing, as holding that an economic substance test should be used to strike down "at least the most egregious tax maneuvers." (3)

A search for the worst or "most egregious" transactions is quite distinct from the aim of seeking to reduce overall social harm. The notion of egregiousness brings to mind malum in se, or acts that are inherently wrong, such as using force or fraud against other individuals. With tax shelters, however, the issue is one of malum prohibitum, or acts that are wrong simply in a compliance sense once they have been identified as legally impermissible or ineffective.

To illustrate, consider a transaction described some years ago by Joseph Bankman, in which a U.S. company induces a foreign accommodation party to create a currency straddle and contribute the losing leg to a new subsidiary, with the end result that the U.S. company gets huge ordinary losses through pure, unadulterated paper shuffling. Bankman notes:

It is easy to see that if this shelter were respected (say, approved

in a Revenue Ruling), it would reduce corporate taxable income to near

zero. Each corporation would place profit-producing assets in a

subsidiary and have that subsidiary purchase loss positions from a

foreign taxpayer. The only limit on the use of the shelter would be

the transaction costs involved in purchasing the loss position, and

locating foreign persons who would be willing to serve as

accommodation parties. (4)

I have argued that this transaction should not be allowed to work, and that under current law it generally does not work, due to the standards of economic substance and business purpose. (5) If, however, such transactions actually were blessed by the Internal Revenue Service, and practitioner friends asked me if I believed that they could ethically engage in it, I would have to say yes. By contrast, if the U.S. government announced that hired killers would not be prosecuted in any way so long as they reported their hit fees on their income tax returns, I would still regard the murders as wrong.

As this example illustrates, the ethical issue taxpayers face in planning tax shelter transactions that they will report favorably to themselves is purely one of malum prohibitum. (6) The question for policymakers is how to set optimal rules under which some subset of the potentially objectionable transactions will be banned, not what are the worst transactions wholly apart from issues of rule design. In practice, this sometimes means that a potentially banned transaction will actually become legally permissible and tax-effective if the taxpayer tweaks it to be socially worse.

Thus, suppose a taxpayer, realizing that a proposed transaction would be ineffective due to its lack of economic substance and business purpose, finds a way to satisfy those requirements by altering the transaction. For example, suppose the taxpayer exposes itself to significant business risks of some kind--risks that it would rather not bear, but that its advisors determine are necessary if the transaction is to be respected for tax purposes. At some point, the transaction might have enough substance that everyone would agree that under current law it works and is not even required to be disclosed. Would this, however, really mean that the transaction was "better" than in its tax-ineffective form, other than in the sense that claiming the tax benefits (without full disclosure) was no longer malum prohibitun?

I would say not. All the taxpayer has done, given that it did not want the business risk, is increase the deadweight loss associated with the transaction to include not only paper-shuffling costs but also those of having a sub-optimal risk position. Taking the decision to proceed as fixed, aggregate social welfare is presumably lower in this scenario than in that where only paper-shuffling was involved.

Why do we tax rules of economic substance and business purpose that induce taxpayers to do this? The best rationale relates to the fact that, as Edward Kleinbard put it:

[A transactional income tax such as ours] works by describing a finite

number of idealized transactions and attaching to each a set of

operative rules--what might be termed a set of tax cubbyholes. Tax

professionals spend a modest amount of time learning to identify these

tax cubbyholes and their consequences, and a great deal of time

massaging reality to fit within the desired cubbyhole. (7)

Kleinbard was speaking of financial instrument labels such as debt and equity, but the same point applies throughout the income tax law. For example, it applies to defining a realization event or a tax-free section 351 exchange, to determining who is the owner of a share of stock that pays a dividend, to examining whether the taxpayer "really" created a Bermuda partnership as claimed, and so on ad infinitum as the issues arise.

If we did not use economic substance tests to challenge the reality of the cubbyholes taxpayers try to exploit, we would in effect have created a regime of pure electivity to claim whatever losses and ignore whatever gains one likes. The frictions imposed by an economic substance rule burden taxpayer electivity. If tax rules are sufficiently well-designed, they can reduce overall deadweight loss even though they induce particular taxpayers to waste even more resources.

Elsewhere, as Professor Katz notes, (8) I have put this argument a bit more colorfully by saying the following:

From this perspective, economic substance is just a tool for

accomplishing aims that have little to do with how one might define it

as a matter of internal logic. Leaving aside the institutional reasons

why (for courts in particular) economic substance is a particularly

suitable tool for deterring undesirable transactions, one might as

well condition favorable tax consequences on whether the taxpayer's

chief financial officer can execute [twenty] back-somersaults in the

IRS National Office at midnight on April Fool's Day, if such a

requirement turns out to achieve a better ratio of successful

deterrence to inducing wasteful effort in meeting requirements

that are pointless in themselves. (9)

This, I gather from conference remarks, has come to be known as the "back-flips" explanation of economic substance rules.

In sum, anti-tax shelter rules that reduce the overall deadweight loss from tax planning need not have anything to do with singling out transactions that are "worse" in some abstract sense. In practice, "better" and "worse" must be determined by reference to what the rules actually allow. Thus, using those concepts to determine what the rules should allow would be circular.

This suggests that one can respond to Professor Katz's analysis by simply demurring. While he is correct about the difficulty of drawing lines here based on better and worse, I do not see this as weakening the case for back-flip rules that impede purely elective tax sheltering.

Daniel N. Shaviro*

* Wayne Perry Professor of Taxation, NYU Law School.