State tax shelters and state taxation of capital.

I. INTRODUCTION

State income tax is a most demanding area of practice. It requires knowledge of federal income tax, constitutional issues that play no role in federal income tax, and the substitutes at the state level for the income tax. I have not specialized in this field and will not address many of the important doctrinal distinctions or policy options that occupy the time of the practitioner or state official. However, I have had long experience with the larger topic of this issue: tax shelters and the taxation of capital. I have also had some limited experience with real life issues in state tax planning. In any event, I am prepared to venture an overview of where I think we are with respect to the tax shelter problem and the greater issue surrounding state taxation of income from capital.

To anticipate, I think we have had some signal victories in the fight against combined federal and state tax shelters. The shelter market as it existed a decade ago is pretty much gone. There is more to be done, at the state as well as federal level; but I think there is the will to do it. The only cloud on the horizon is the prospect of a Supreme Court decision (or a series of appellate court decisions) that could pump new life into the shelter market.

I am less optimistic as to whether we can prevent aggressive tax planning aimed at reducing state (but not federal and state) revenue from capital. The reason for this is the growing ascendancy of intellectual over physical capital. Traditional tax concepts, such as sourcing, are difficult to apply to intellectual capital. Intellectual capital is also more mobile, and because of that, tax competition for intellectual capital is more powerful. Unfortunately, some of the same reasons that make it hard to tax income from capital also threaten the state's ability to tap other revenue sources, such as sales and payroll. State tax officials will have to work hard to come up with the revenues needed to fund the typical functions of state and local government.

II. DEFINITIONS

One difficulty in discussing the tax shelter problem is that there is no uniform definition of the term tax shelter. Here, I will use the definition I have put forth in earlier writings. A tax shelter is a transaction that (a) reduces the tax on capital; (b) in a manner that is consistent (or purports to be consistent) with the literal language of a statute (regulation); (c) but is strongly inconsistent with any notion of legislative (administrative) intent or purpose; and (d) that has little or no non-tax business purpose. (1)

Shelters often exhibit "secondary" characteristics. These include the presence of non-taxable entities and the role of financial intermediaries as promoters. In a typical shelter, a revenue stream is bifurcated for tax purposes, with items of income shunted to the nontaxable entity and items of expense shunted to the shelter purchaser. (2)

I will use the term "quasi-tax shelter" or "aggressive jurisdictional tax planning" to define transactions that reduce state taxes but do not fall within this definition of shelter. A transaction that transfers intellectual property of a related corporate group to an intangible holding company is an example of a quasi-tax shelter or aggressive jurisdictional tax planning. I will use the term jurisdictional tax planning to describe less aggressive means to reduce state taxes, including but not limited to the practice of placing property in low tax states.

I differentiate between tax shelters, quasi-tax shelters, and jurisdictional tax planning because I believe transactions that fall into these categories raise different issues. To take but one example, tax shelters can be attacked through the economic substance and related doctrines; the same is not true of jurisdictional tax planning.

Other commentators have used other terms to describe the transactions discussed in this article. The State of California, for example, uses the term "abusive tax shelter" to describe what I call a tax shelter. (3) The Multistate Tax Commission includes most of what I call jurisdictional tax planning in the definition of tax shelter. (4) In general, my definition of tax shelter is narrower than other definitions. This is due to the prong of my definition that deals with legislative (administrative) intent. Under my definition, a transaction is not a shelter unless it is strongly inconsistent with legislative intent or purpose. Many aggressive transactions can fit within some notion of legislative intent or rely on statutes for which no meaningful notion of legislative intent can be gleaned. These transactions do not fall within my definition of tax shelter.

In the end, no definition can adequately encompass all transactions, and nothing much hinges on whether one adopts one definition over another. What is important is that we recognize that taxpayers use many different methods to reduce the tax on capital and that we separately discuss the proper response for each of these methods.

III. STATE AND FEDERAL TAX SHELTERS

Most of the transactions I describe as tax shelters are used, or have been used, to reduce taxable income on both the federal and state levels. The federal government has had considerable success in limiting these transactions and the states have benefited from that federal success. There have been significant victories at the state level, too, with corresponding benefit to the federal government.

A decade ago, the shelter market was in full flower. Some of the nation's largest companies had entered into transactions without (nontax) business purpose in order to generate a tax loss. Some of these transactions--such those involving contingent installment sale or stepped-down preferred--had been uncovered by the Internal Revenue Service (Service) or stopped by quick administrative action. The Service had not detected most shelters (such as lease strips or loss import shelters). (5) As a result, shelters had proven an attractive investment for "early bird" purchasers. Shelters were sold by leading financial intermediaries who had access to corporate America and who could amortize costs over many purchasers. Those intermediaries were just gearing up to "down" market shelters to individual entrepreneurs and executives.

These corporate and individual shelters were primarily aimed at reducing federal taxes but substantially reduced state taxes as well. Some idea of the effect of shelters on state revenues can be gleaned from the 2003 anti-shelter legislation enacted in California. That legislation did not change the substantive law but it did significantly increase penalties on certain transactions that were held invalid due to lack of economic substance. The increase was effective prospectively, and for adjustments made to past years still subject to audit. The increase was accompanied by an amnesty provision; taxpayers could avoid the new penalties by filing amended returns, giving up shelter losses and paying the full amount due, plus interest and any penalties under prior law, on the recomputed income. (6)

The amnesty provision raised approximately $1.4 billion dollars. (7) This reflected a "give back" of approximately $15 billion dollars of shelter losses claimed by California taxpayers or corporate taxpayers on income subject to California tax. (8) Some of the amounts received were paid by taxpayers who are contesting their liability and will be returned to taxpayers if their positions are upheld. In that respect, the amnesty overstates wrongful shelter deductions. In general, though, the amnesty figures undoubtedly understated shelter deductions. Many taxpayers did not qualify for amnesty and others no doubt decided to forgo the amnesty and bank instead on escaping detection on audit, or successfully defend their positions on audit, administrative hearings, or court.

Most of the readers of this article will be familiar with many of the state and federal responses to the shelter problem. A primary weapon against shelters has been stepped-up enforcement, due to better and more thorough audits, a series of victories giving the government the right to follow the paper trail of promoters, and, perhaps most importantly of all, the promulgation of disclosure rules. (9) Statutory civil penalties have increased, at the federal and state level. There also seems to be a noticeable increase in the number of cases in which the government asserts penalties. The government has substantially strengthened Circular 230, which governs practice before the Service. Circular 230 now imposes stringent requirements on attorneys who write shelter opinions or otherwise aide shelter promoters. (10) As noted above, there is no agreed-upon definition of a tax shelter and increased shelter-related penalties inevitably spill over and affect non-shelter related transactions. The most obvious consequence of new Circular 230 is the disclaimer that accompanies e-mails and other correspondence from most law firms. Of greatest interest, perhaps, is the criminal case now pending against promoters associated with the accounting firm of KMPG and the criminal charges settled with that firm. (11) Whatever the outcome (or even merits) of that case, the fact that the government is even willing to bring criminal charges against a leading financial intermediary and its employees is certain to have a chilling effect on the shelter market.

At its core, however, the battle against shelters has rested upon modern variants of old common law doctrines, such as business purpose and economic substance. As noted above, most shelters comply, or purport to comply, with the literal language of a governing regulation or statute. One might imagine the government might challenge shelters with an intention or purpose-based reading of the statute and, in fact, the government often does make such arguments. In tax, however, the economic substance and other doctrines do most of the work. The government argues that, notwithstanding compliance with the literal language of a statute, the transaction fails for want of economic substance. As I have argued elsewhere, the economic substance doctrine cannot (or at least should not) be used in cases in which the transaction complies with the spirit, as well as the letter, of the law. (12) It is for this reason that the doctrine does not apply to transactions such as the purchase of tax-exempt, rather than taxable, bonds. The economic substance doctrine, therefore, incorporates a non-literal reading of the statute.

Based on the economic substance and other doctrines, the government has had a remarkable string of judicial victories against tax shelters. Without these doctrines, or other doctrines or interpretive arguments that can be used in like fashion, tax shelters would work. They would be upheld against government challenge, and the other anti-tax shelter measures, such as increased penalties or enforcement, would be irrelevant.

The very good news is that the shelter market of the 1990s is pretty much dead. Leading financial intermediaries--who were terrifically efficient shelter producers and marketers and who gave legitimacy to the flimsiest of schemes--have for the most part gotten out of the shelter business. Law firms have made a similar exit. Today's shelters are not publicly marketed, designed to avoid the reporting requirements, and (from what I can glean) less egregious/aggressive. They are harder to detect, but there are fewer of them. Perhaps that is as it should be. A "no shelter world" would require otherwise undesirable changes to substantive law and enormous enforcement costs. Limiting tax shelters to a smaller, underground market is as close to victory as we will get--at least without wholesale tax reform. To use a homely metaphor: We have cleared the rats from the house and can live with a few of them scurrying around the basement. (Though if the opportunity presents itself, we should kill them, too.)

There are some shelter-like schemes now being marketed to professionals and small business owners. These typically require the taxpayer to send or shift money to an off-shore bank account. The shelters are often marketed both as a way to save taxes and protect assets from creditors. In many of these shelters, business owners set up an offshore corporation and funnel profits to that company, either through related party transactions with domestic corporations or simply by assigning consulting or other income to that company. Profits are repatriated through the use of offshore credit cards. In one shelter, participants shunted taxable income to an offshore bank and then had that same income repatriated in the form of a home mortgage. (13) Participants deducted the mortgage interest, thereby turning taxable income into a deduction.

As the above description suggests, these "downmarket" schemes make little effort to fall within even the literal letter of the law and are probably better described as frauds rather than tax shelters. The real issue here is enforcement priorities. As measured on a per participant basis, the taxes evaded are small, particularly when compared to the shelters of the 1990s. In the aggregate, however, the costs of evasion may be great. Moreover, these schemes are widely marketed on the web and elsewhere. Allowing promoters of these schemes to operate with impunity might dispirit more honest taxpayers and lead to a gradual erosion of tax compliance among that group.

A more general enforcement issue is whether the states ought to devote resources to shelters that prey on both federal and state revenues or that simply free ride on federal efforts. My own sense is that the larger states, at least, will find it worth their while to pursue these shelters. The stakes are too large to ignore the issue and experience has shown that states can find transactions that elude the Service. In this respect, shelters can simply be viewed as a subject of underreporting. Many states profitably attack the problem of the cash economy, and I believe the same is and will be true with respect to shelters.

This still leaves the problem of intergovernmental coordination. In theory, one might imagine that smaller states would contract out with the federal government or larger states. My chauvinistic belief is that my home state, California, would offer an attractive bundle of enforcement services and could pick up extra revenue in this manner--doing well while doing good. In practice, I suppose the most that could be hoped for is that the federal government and states share information and try to coordinate enforcement strategies on a case-by-case basis. The data base maintained by New York (but initially stocked with data from California) is a step in the right direction.

IV. QUASI-TAX SHELTERS AND JURISDICTIONAL TAX PLANNING

A. Mobility of Capital

Perhaps for good reason, the kinds of shelters discussed above receive relatively little attention in journals and conferences devoted to state taxation. To some extent, this is because shelters are primarily seen as a federal problem. More significantly, the transactions I describe as tax shelters comprise only a small part of the reason state corporate tax collections are down. Even at the height of the tax shelter boom, that decline was generally thought caused by tax competition among states, quasi-tax shelters in the form of intangible holding companies, or forms of international tax competition or tax planning that reduced the federal tax base and in so doing, reduced the state tax base as well. (14)

All these latter sources of decline can be traced to a common cause: the increased mobility of capital. A hundred years ago, wealth from capital was comprised of natural resources or tangible property in the form of utilities, factories, and the like. Due to high transportation costs, physical capital was located near population centers. In the short and mid-term, at least, capital was immobile. (15)

Today, wealth from capital is increasingly comprised of intellectual property--patents, copyrights, or trade secrets; or human capital that is entwined with such property. Intangible property in the form of trademarks or brand names comprises another significant source of capital. Both forms of intangible capital can be easily moved. Moreover, it is sometimes hard to know where intangible property is created, located, or used. As a result, traditional tax concepts are difficult to apply to income from such property. Examples of this difficulty abound in the tax literature. (16) Consider, for example, Silicon Valley engineers and venture capitalists who transfer software-related intellectual property and cash to a Cayman Islands corporation in exchange for the stock of that company. The company has development teams in California, Texas, and Shanghai and licenses software to multinational telecommunication companies. Accurate attribution of income is nigh impossible.

The combination of easily-moved property and difficult attribution of income issues fuels tax competition. States compete for situs of the company's jobs and physical capital with traditional tax incentives. States can also compete for the situs of intangible property, which is similar to the international competition for tax havens. Even states that eschew straightforward tax competition or tax haven-like incentives and adopt formulary apportionment can compete--witness the growing number of states whose formulas now overweight sales. (17)

Tax competition obviously reduces aggregate state tax revenues and disadvantages more states than it advantages. There are well known remedies or at least palliatives to tax competition. The problems of "tax havens" in the form of states that levy no tax on intangible-source income can be addressed through forms of combined reporting, or loss disallowance rules. (18) States could work, on an informal level at least, with neighboring states to turn deaf ears to more straightforward forms of competition that promise in short order to reduce revenue all around. Federal funding of traditional state provided services might supplant or at least supplement state funding--something that has already happened to some extent in education.(19) More ambitiously, some have proposed federalizing corporate tax collection and apportioning the funds under a uniform formula. (20) These proposals would reduce transaction costs and possibly eliminate some forms of tax competition. (21) More far-reaching solutions, such as binding multi-state compacts on tax rates as well as an apportionment formula, are too improbable to warrant discussion.

Whatever the causes or solutions, tax competition seems too strong for states to ignore for at least three reasons. First, the competitive equilibrium does not disadvantage all states. Smaller states that have attained tax haven status, such as Delaware, are obviously better off. States desperate for jobs may be better off, too. They may wind up with manufacturing plants other states lose, albeit plants that do not pay much tax. A second point is that there are those who would make a normative case for tax competition. From an academic perspective, tax competition under certain assumptions leads to efficient provision of goods and services. This point is explored in detail below. Some of the academic arguments in favor of tax competition, and many arguments that have no currency in the academy, are made in the political arena. Business lobbyists play an important role in state government, and when the political wind is right, it is their position that carries the day. Finally, coordinated action is difficult (in part for the first two reasons), and even a state that might wish to eschew tax competition worries about losing its tax base to states that feel otherwise. California is now considering single-factor sales formula primarily for this last reason.

B. Aggressive Tax Planning and Quasi-Tax Shelters

One might hope that, whatever its drawbacks, tax competition might replace the need for aggressive tax planning. That is not the case, however. States must contend with both sources of revenue loss. Probably the most well known aggressive tax planning technique is the use of the intangible holding company to shift income to a no-tax state. The transitions described in the Massachusetts (and New York) Sherwin-Williams case is typical. (22) In simplified form, Sherwin-Williams transferred trademarks to newly-formed Delaware subsidiaries. The company valued this property at over $300 million and paid approximately $50 million in royalties to the subsidiaries for the use of the trademarks. Sherwin-Williams treated the royalties as a (deductible) cost of business. The royalties were recognized as income by the subsidiary which took the position that it was taxable only in Delaware. Delaware, of course, does not tax holding companies whose income comes from royalties and other intangible assets. (23)

From one perspective, at least, the transaction resembles the tax shelters described above. Presumably, the Massachusetts legislature believed and intended its corporate tax to raise revenue from companies such as Sherwin-Williams, with the amount of tax a function of underlying business conditions and operations. The transfer of property to the Delaware holding company was a tax-motivated transaction without significant non-tax purpose. (24) If respected, however, the transfer would defeat the legislature's expectations and reduce the tax to near zero.

The transaction resembled a typical tax shelter in shunting income to a tax indifferent party. Here the income goes to a subsidiary organized in what amounts to a tax haven--a state in which the income would not be taxed.

Seen from another perspective, the taxpayer's position is consistent with not only the literal language, but the structure and intended operation, of the Massachusetts tax law. The amount of income subject to tax is a function of the value of the property placed in the state. This basic fact can be assumed to be known to the legislature. A newly formed company can organize itself so as to place valuable property in low tax states; there is no logical reason why that same option cannot be open to companies with ongoing operations. (25)

It might be useful to compare the intangible holding company structure to a simplified description of some of the more egregious tax shelters of the late 1990s. In one set of "Son of Boss" shelters, the taxpayer contributed cash and the liability for a short position to a partnership and took the position that the cash increased basis but the liability did not decrease basis. The result was an artificially high basis that produced a loss on liquidation. (26) In another set of shelters, the taxpayer formed a corporation, and entered into other economic relationships, with a tax indifferent party. The taxpayer then caused the corporation to redeem all of the tax indifferent party's stock. The redemption was structured so that it would be treated as a dividend under the section 302 regulations, with the result that, under the literal language of the regulations, the tax indifferent party's basis would flow through to the taxpayer. (27)

Legislative intent cannot be known with any certainty, and to the extent that intent can be inferred or assumed, both the intangible holding company technique and Son-of-Boss shelters can best be thought of as points on a continuum, rather than occupying opposite sides of a divide. That said, it is easy to see that basis shifting and Son of Boss shelters were particularly distant from any notion of legislative intent or purpose. One can imagine that a few Massachusetts legislators might think that Sherwin-Williams got it right: the law as written was intended to allow taxpayers to shift income by shifting the situs of intangible property and that the law ought to continue to give taxpayers that option. One cannot imagine any sane legislator coming to the same conclusion with respect to the Son of Boss shelters.

There is another obvious distinction between the intangible holding company technique and the Son of Boss and other tax shelters. There are known legislative fixes to the former tax minimization device: combined reporting, or rules disallowing deductions between related parties for payments relating to intangibles. The situation with respect to the Son of Boss or other more typical tax shelters is more complicated. There is always a legislative fix for the particular shelter but there is always another shelter just around the corner. I have argued elsewhere that it is impossible to prospectively fix all the leaks in our tax system and that in equilibrium, we can always expect a handful of shelters that seem to "work" if existing law is interpreted literally. (28) This may be apparent to judges who decide shelter cases, since the government often introduces into the record the taxpayer's search for the right tax minimization technique. Judges can therefore get a retrospective look at the panoply of shelters available in the market at one time.

The above analysis suggests that judges are more likely to accept the taxpayer's literalist interpretation of the tax law when the intangible holding company technique is litigated than in the case of the Son of Boss or other more typical shelters. The intangible holding companies can arguably fit within the spirit, as well as the letter, of the law; and since there is a legislative remedy, the judge need not resort to expansive interpretive doctrines such as economic substance.

I believe the actual case law supports this conclusion. The government has won nearly all of the shelter cases. The fact that courts have continually supported the government's position has made legislators willing to increase penalties. The victories in civil litigation have made it possible for the government to assert criminal penalties in at least one set of related cases. (29) Government victories have also led taxpayers caught in shelters to sue promoters. These results, together with other changes described earlier, have caused leading law and accounting professionals to drop most shelter work. The government's won-lost record on intangible holding companies is more mixed. Massachusetts lost the Sherwin-Williams case but New York, litigating the same facts under a slightly different statute, won. (30) However, it is possible to find somewhat similar cases Massachusetts has won and New York has lost. (31) In general, states have done well in litigating the intangible holding company cases, (32) but they have not done so well as to make government victory certain, and the government has not been able to leverage its victories with higher penalties. States have not been able to win the battle against intangible holding companies through litigation.

What the states have done, of course, is to attack the intangible holding company through legislation. States have eschewed the more general approach preferred by academics (combined reporting) and instead adopted loss disallowance rules, preventing taxpayers from deducting royalties and other expenses paid to related parties. (33)

Intangible holding companies probably represent the form of tax minimization that is most egregious and most easily challenged in court. Other forms of tax minimization can be attacked, if at all, only through legislation and interstate cooperation and uniformity of law. As noted earlier, there are structural reasons why (for the most part) that has not occurred. It is not surprising that state corporate tax revenues appear to be on a long term decline, both as a percent of federal corporate tax revenues or a percent of corporate profits. (34)

V. TAX COMPETITION

A. Theoretical Models

I have suggested above that while the battle against tax shelters is going well, tax competition appears to be gradually, but inexorably, reducing state corporate tax revenues. If this latter observation is correct, a key question is where all of this will lead. What are the implications of unrestrained tax competition?

Academics tend to love theory and be good at it, but are not so interested or knowledgeable about real world complexities. Academics tend to approach questions first by applying a theoretical construct to a world shorn of known complexity, and then to gradually introduce some (but perhaps not enough) of those complexities. I will follow that practice here, starting with Charles Tiebout's simple, but powerful, insight into the product of tax competition. (35)

In Tiebout's model, individuals can costlessly move between local municipalities and choose the municipality that offers the combination of services that they most prefer at the given price, or tax rate. Some residents might prefer a well-funded school system and be willing to pay a high property tax; others might prefer and pay for parks or activities for the elderly; still others might opt for fewer services at a lower cost. The market for municipal services resembles any other market, such as the market for cars where consumers choose the car with the right features at the right price. (36)

There are a number of attractive features to the world depicted in the Tiebout model. Citizens get the government they want at a price they are willing to pay. A municipality that does not offer good value will lose its residents. The linked relationship between cost and services eliminates the negative welfare effects of tax. Taxpayers respond to the typical tax by avoiding substituting away from the taxed activity. To the extent that occurs, the government gains no revenue and the taxpayers are worse off (because they have forgone an activity they otherwise would have chosen to engage in). A tax on wages leads to the substitution of leisure for labor; a tax on savings leads to the substitution of current consumption for savings and future consumption; and so on. (37) Not all taxpayers substitute away from the taxed activity and taxpayers who substitute may do so only a little bit. Taxes on labor and capital continue to raise money. But some substitution occurs, and to the extent it does, welfare is reduced.

The reason why a typical tax causes substitution away from the taxed activity is that there is no relationship, at the margin, between the tax paid and government services received. A taxpayer does not get any additional government services from working an additional hour and paying the tax on her labor or from saving an additional dollar and paying tax on her interest income. In the Tiebout model, however, the relationship between municipality and taxpayer can be thought of as that between a seller and purchaser. A resident purchases schools for her children with her property tax. The tax does not rise with income, and unless she pays the tax, she cannot live in the city and use its schools. There is a direct relationship between the tax paid and the services received. There is no inefficiency if a taxpayer works to purchase schooling for her children, just as there is no efficiency if the taxpayer works to buy a car.

The same reason why the typical inefficiencies of tax do not exist in the Tiebout model--residents get the bundle of services they want--precludes forced redistribution from the wealthy to the poor. A resident who is forced to pay tax that subsidizes a service for the poor always has the option to leave the jurisdiction and find another municipality that does not require that subsidy. Many citizens will want to live in a municipality that offers those subsidies and can elect officials or otherwise vote for that subsidy. For those individuals, the tax will continue to be efficient, since they will value the subsidy for the poor. But the majority cannot bind the minority who do not value the subsidies as the minority has the option of costless exit. In that sense, payment into redistributive government programs becomes voluntary. Which redistributive programs would continue to exist, and to what extent, is an open question.

In the real world, of course, we lack many of the conditions necessary to produce the result predicted by the Tiebout model. The closest to benefit taxation is probably the local property tax, as adjusted by zoning regulations. (38) We do see suburbs with quarter-acre or larger lots and high property taxes that go to support local schools and retirement communities with low taxes geared toward providing services for the elderly. But even there, the tax only loosely approximates services. It is not clear how one could achieve benefit taxation in more heterogeneous cities. Externalities imposed by one jurisdiction over another--in the form of traffic, pollution, and so on--also interfere with the neat result produced by the Tiebout model. (39) Some have questioned whether there could be an efficient municipal market in services for which there is not a private market, since the absence of a private market indicates some form of market failure. (40)

Moreover, since Tiebout first set forth his model in 1954, events have lessened the relationship between tax and benefit even in the paradigmatic case of property tax and schools. In many states, courts have required that property and other taxes be pooled so as to equalize school funding across districts. (41) The federal government now supplies a significant portion of local school funding, further lessening the relationship between local tax and local services. (42) It is not surprising that empirical support in favor of the Tiebout hypothesis is mixed. (43)

There would appear to be no place for the corporate tax in a Tiebout world, since the relationship between tax paid and services received in so tenuous. Indeed, it is not even clear to whom one should look for benefit: the company paying the tax, its shareholders, employees, or the purchasers of the company's goods or services. As a result, the corporate tax, even if levied in what might pass for a "pure" form, does not lead to anything like a Tiebout equilibrium. Instead, those who have modeled the effects of interstate competition on the corporate tax have come to more pessimistic conclusions. In one leading set of models, high corporate taxes cause companies to locate out of state, which in turn reduces in-state employment and wages. (44) States respond by reducing the corporate tax rate. In equilibrium, for states that must rely on the corporate tax, services are underfunded. There are services that would improve welfare (even after allowing for their tax cost) in a world without competition that cannot be provided in a world with tax competition. The models do not state how much services will be underfunded. In general, the models are consistent with (but do not themselves tell) a story of tax competition in which the corporate tax becomes such an expensive and poor source of revenue that it is dropped altogether.

B. Where Might This All Lead?

In the field of tax (as in all other fields) there have been countless predictions, based on theory, that have not come true. In the 1980s, corporations began replacing equity with debt, and non-deductible dividends with deductible interest. This threatened to seriously reduce the corporate tax. (45) Policymakers, who had long disfavored the tax, saw this as an opportunity to eliminate it through various integration proposals. (46) If, as academics had long suspected, the corporate tax could be and would be eroded though interest deductions, integration could be achieved without losing too much revenue. Of course, none of that occurred. Instead, the junk bond market collapsed and companies no longer replaced equity with debt. A short while later, the technology boom hit and investors flocked to buy new equity in private companies and initial public offerings.

Many of us have made false predictions on a subject closer to that of this article: international tax competition. Federal corporate tax revenues, relative to profits, declined throughout the 1990s. (47) Tax shelters were one source of that decline. But it did not seem credible that the decline was caused wholly, or even mostly, by tax shelters. Many observers, including myself, have speculated that the decline was in significant part due to international tax competition and it would be difficult to put the brakes on that decline. (48) We may yet be correct in that assumption. In the past few years, however, federal corporate tax collections are not down but up, and up considerably. (49)

Predictions of ruinous state tax competition may be similarly overstated. The forces keeping companies in a high tax state may be more powerful than imagined. We have seen some evidence of this in California, whose high technology sector seems to have survived not only high taxes, but the much higher implicit tax in the form of astronomical housing prices. The intangible holding company structure may be responsible for more of the revenue loss than for which it is credited, and recent state legislation may greatly reduce that loss. Erosion of the tax base may stiffen the resolve of state legislatures, which may yet adopt some of the measures discussed earlier. Whatever the long-term outlook, state tax officials will have to vigorously contest corporate attempts to shift income out of state and to urge their legislators to give them additional vehicles (such as combined reporting) to use in that pursuit. At the same time, of course, state and federal officials will need to diligently protect the hard-won gains against the transactions I have defined here as tax shelters.

I am nonetheless stuck with my belief that, despite the best efforts of state tax officials, competition will make the corporate tax harder and harder to implement. Unfortunately, the tax that I and others have urged we adopt to replace the corporate tax at the federal level--a version of the flat tax--would be difficult to apportion among the states and might well exacerbate state tax competition. (50) The most likely real-world candidates to replace the state corporate income tax--individual income taxes and sales taxes--also suffer from tax competition, although I believe less so than the corporate income tax. There is one tax that is (relatively) immune to tax competition. That is the real property tax, which, once levied, is impounded in property value and cannot be avoided by moving out of state. I do not need to tell readers how unpopular this revenue source is. This tax will nonetheless play an important role in state taxation, if for no other reason than that it will continue to be available. I believe this tax may also prove to be less redistributive in future years than it is now. The reason for this is that decisions requiring the equalization of school expenditures, always controversial, may eventually be overturned or (more likely) not enforced by today's more conservative judiciary. If that happens, the tax might become more popular and its role expand still further. The end result? A state and local sector many would describe as underfunded and others simply as smaller and less redistributive; and one that is a bit closer to that described by Charles Tiebout.

Joseph Bankman**

* This article originally appeared in The State and Local Tax Lawyer. See Joseph Bankman, State Tax Shelters and State Taxation of Capital, 12 ST. & LOC. TAX LAW. (forthcoming 2007).

** Ralph M. Parsons Professor of Law and Business, Stanford Law School; University of California at Berkeley, A.B., 1777; Yale Law School, J.D., 1980.

(1) See Joseph Bankman, The Tax Shelter Problem, 57 NAT'L TAX J. 925 (2004).

(2) See Joseph Bankman, An Academic's View of the Tax Shelter Battle, in CRISIS IN TAX ADMINISTRATION (Henry J. Aaron & Joel Slemrod eds., 2004).

(3) See, e.g., Mark Ibele & John Vashe, Abusive Tax Shelters in California, 39 STATE TAX NOTES 781 (Mar. 13, 2006).

(4) See MULTISTATE TAX COMM., CORPORATE TAX SHELTERING AND THE IMPACT ON STATE CORPORATE INCOME TAX REVENUE COLLECTIONS (July 15, 2003), available at http://www.mtc.gov/uploadedFiles/Multistate_Tax_Commission/Resources/Studies_and_Reports/Corporate_Tax_Sheltering/Tax%20Shelter%20Report.pdf.

(5) See Joseph Bankman, The New Market in Corporate Tax Shelters, 83 TAX NOTES 1775 (June 12,1999).

(6) The legislation is described in Joseph Bankman & Daniel Simmons, Terminating Tax Shelters: Has California Broken the Legislative Logjam?, 101 TAX NOTES 1111 (Dec. 1, 2003).

(7) The amnesty program is described in Joseph Bankman, Tax Enforcement: Tax Shelters, the Cash Economy, and Compliance Costs, 31 OHIO N.U.L. REV. 1 (2005).

(8) Id.; see also Ibele & Vashe, supra note 3.

(9) These and the other steps described in this paragraph are described in Bankman, supra note 1; see also JOSEPH BANKMAN, WILLIAM A. KLEIN & DANIEL N. SHAVIRO, FEDERAL INCOME TAXATION 576-85 (14th ed. 2006).

(10) See discussion in BANKMAN, KLEIN & SHAVIRO, supra note 9; see also Rachel Alexander, Randall Hanson & James Smith, Tax Shelter Opinion Letters Have Come Under Attack, PRAC. TAX STRATEGIES (Nov. 2004).

(11) See Sheryl Stratton, Tax Professionals Indicted: KPMG Says Shelters were Fraudulent, 108 TAX NOTES 1985 (Sept. 5, 2005); see also Resolution of Board of Directors of KPMG LLP: Deferred Prosecution Agreement (Aug. 26, 2005).

(12) Joseph Bankman, The Economic Substance Doctrine, 74 S. CAL. L. REV. 5 (2000).

(13) There is a voluminous literature on the use of offshore credit cards and bank accounts. See, e.g., Amy Hamilton, A Glimpse Inside Offshore Greed, 98 TAX NOTES 297 (Jan. 20, 2003).

(14) See supra note 4.

(15) In the long run, of course, population centers and mineral exploration could move toward low tax jurisdictions.

(16) See, e.g., JEROME R. HELLERSTEIN & WALTER HELLERSTEIN, STATE TAXATION [paragraph] 9.15 (3rd ed. 2000).

(17) See, e.g., Eric Parker, Single Sales Factor: Trend Among States or Just an Aberration?, 2005 STATE TAX TODAY 143-1 (July 27, 2005).

(18) See Harley Duncan & William F. Fox, State Strategies for Dealing with Tax Sheltering and Planning, ST. & LOC. TAX LAW. (forthcoming).

(19) Between 1965 and 2004, on-budget federal funding for K-12 education increased from about $2 billion to about $66 billion. See NAT'L CTR. FOR EDUC. STATISTICS, DIGEST OF EDUCATION STATISTICS (2004), http://nces.ed.gov/programs/digest/d04/ch_4.asp. In California, federal funding now provides about 12% of the total K-12 educational budget. See EDSOURCE, 2005-06 K-12 EDUCATIION FUNDING COMES FROM SEVERAL SOURCES (Oct. 2005), http://www.edsource.org/sch_fin_fund0506.cfm.

(20) See Eugene F. Corrigan, Taming the Beast--The State and Local Tax System, 40 STATE TAX NOTES 415 (May 1, 2006) (discussing apportionment in proportion to population); Carolyn Joy Lee, State Corporate Income Taxes No Longer Make Sense, 40 STATE TAX NOTES 197 (Apr. 17, 2006) (discussing an unspecified uniform apportionment formula).

(21) The proposals would not eliminate straightforward competition in the form of lower rates on income, but they would eliminate some forms of competition based on the apportionment formula.

(22) Sherwin-Williams Co. v. Commissioner of Revenue, 438 Mass. 71 (Mass. 2002); Sherwin-Williams Co. v. Tax Appeals Tribunal, 784 N.Y.S.2d 178 (N.Y. App. Div. 2004).

(23) See DEL. CODE ANN. tit. 30, [section] 1902(b)(8); see also DEL. CODE ANN. tit. 30, [section] 6401 (imposing a minimum tax on Delaware Headquarters Management Corporations).

(24) The board resolution authorizing the creation of the subsidiaries listed a number of business purposes other than taxes; however, I believe virtually all commentators find these purposes either pre-textual or of exceedingly modest importance. The Massachusetts court that held in the company's favor did so on the basis that the subsidiaries performed business operations, not that there was a non-tax advantage to organizing the subsidiaries.

(25) The commentary on the case is considerable. Compare James H. Peters, Sherwin-Williams v. Commissioner: A Flawed Decision, 2003 STATE TAX TODAY 86-30 (May 5, 2003) and James H. Peters, Sherwin-Williams: A Reprise, 2003 STATE TAX TODAY 125-18 (June 30, 2003), with Peter L. Faber, Margaret M. Duncan & James E. Griffith, "State of Practice": Sherwin-Williams Represents Sound Intellectual Property Law--Not a Flawed Decision, 2003 STATE TAX TODAY 105-15 (June 2, 2003).

(26) See Lee A. Sheppard, Treasury Shuts Down Baby Boss Deals, 88 TAX NOTES 850 (Aug. 14, 2000).

(27) See, e.g., Lee A. Sheppard, Dissecting the Basis-Shifting Transaction, 92 TAX NOTES 870 (Aug. 13, 2001).

(28) Bankman, supra note 2, at 19-20.

(29) See TAN, supra note 11.

(30) See supra note 22. The Massachusetts court held that in a context of a reorganization, subsidiaries need only conduct actual business operations in order to have economic substance and that the Sherwin-Williams subsidiaries met that test. New York law gave the Commissioner the right to combine unitary companies where certain ownership requirements were met and where combination was necessary to avoid distortion. The Tax Appeals Tribune found, and the Supreme Court, Appellate Division of New York affirmed, that the royalty arrangement distorted income. The holding was predicated upon finding that the royalty arrangement served no non-tax business function.

(31) See, e.g., In the Matter of Toys "R" US-NYTEX, Inc, TAT(E) 93-1039 (N.Y.C. Tax App. Trib. Jan. 14, 2004); Syms Corp. v. Commissioner of Revenue, 436 Mass. 505 (Mass. 2002).

(32) In the cases mentioned supra note 31 and in the text above, states have attempted to deny a deduction for the royalty payment or to combine the parent and subsidiary companies. States have also tried to tax the holding company directly. See, e.g., Kmart Props., Inc. v. Taxation & Revenue Dep't, 131 P.3d 27 (N.M. Ct. App. 2001); Geoffrey Inc. v. Okla. Tax Comm'n, 132 P.3d 632 (Okla. Ct. App. 2006); Geoffrey Inc. v. S.C. Tax Comm'n, 437 S.E.2d 13 (S.C. 1993), cert. denied, 510 U.S. 992 (1993). For a discussion of the difficulties of this latter approach, see Walter Hellerstein, Green Light, Red Light, or Blue Light? New Mexico Supreme Court Sends Mixed Signals with K-Mart Decision, 39 STATE TAX NOTES 141 (Jan. 16, 2006).

(33) Some of these royalty add-back provisions and other state responses to aggressive tax planning are described in David J. Shipley, Scott D. Smith, & Brad A. Bauer, A New Paradigm for State Corporate Income Tax Planning: Part I--The Changing State Environment, 2004 STATE TAX TODAY 173-2 (Sept. 7, 2004). Since the publication of that piece, other states have enacted similar legislation. See e.g., Gen. Assembly, Georgia Final HB 191 Phases in Single Sales Factor, 2005 STATE TAX TODAY 94-6 (May 17, 2005) (discussing add-back provisions in George legislation).

(34) See William F. Fox & LeAnn Luna, State Corporate Tax Revenue Trends: Causes and Possible Solutions, 55 NAT'L TAX J. 925 (2002).

(35) See Charles M. Tiebout, A Pure Theory of Local Expenditures, 64 J. POL. ECON. 416 (1956).

(36) In Tiebout's original article, only the consumers were actual agents; those who ran the municipalities were passive. The municipalities offered services but did not change their services to attract more residents or obtain a surplus. Subsequent scholars have expanded the Tiebout model to incorporate an active role for municipal managers. There is a vast literature on the Tiebout model. A good summary of some of the more important extensions can be found in William A. Fischel, Footloose at Fifty: An Introduction to the Tiebout Anniversary Essays, in THE TIEBOUT MODEL AT FIFTY: ESSAYS IN PUBLIC ECONOMICS IN HONOR OF WALLACE OATES (William A. Fischel ed., 2006), available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=895609&high=%20fischel.

(37) A tax might have a wealth effect, as well as a substitution effect. The tax makes the taxpayer poorer, which may also affect behavior. The wealth effect might, for example, reduce someone's reservation price so that instead of requiring $25 an hour to work, she will work at $20. It will still be the case, however, that given that new reservation price, the tax will interfere with otherwise optimal labor bargains. To see how this might occur, imagine that an employer offers this taxpayer $22 an hour ($2 above her reservation price), and the tax rate is 40%. The taxpayer will turn down the job offer. She and the employer will both be worse off (because they have lost the benefit of the bargain), and the government will be no better off.

(38) The application of Tiebout to property taxation is generally credited to Bruce W. Hamilton, Zoning and Property Taxation in a System of Local Governments, 12 URB. STUD. 205 (1975); see William A. Fischel, Property Taxation and the Tiebout Model: Evidence for the Benefit View from Zoning and Voting, 30 J. ECON. LITERATURE 171 (1992).

(39) The externalities can be positive, as well as negative. A tax-supported university may raise nearby educational levels, which might raise average income, and the benefits may spill over to a neighboring low tax state. One of the assumptions in the Tiebout model was that there were no spill-over effects.

(40) See Hans-Werner Sinn, The Selection Principle and Market Failure in Systems Competition, 66 J. PUB. ECON. 247 (1997).

(41) For a discussion of this development and a case study of one particular instance of court-mandated sharing, see Eric J. Bruner & Jon Sonstelie, California's School Finance Reform: An Experiment in Fiscal Federalism (Univ. of Conn. Dep't of Econ. Working Paper No. 2006-09), available at http://www.econ.uconn.edu/working/2006-09.pdf

(42) See supra note 20.

(43) One reason why the evidence is mixed is that researchers have tested different aspects of the Tiebout model. Compare Kenneth N. Bickers, Lapo Salucci & Robert M. Stein, Assessing the Micro-Foundations of the Tiebout Model, 42 URB. AFF. REV. 57 (2006) (finding that people are able to identify and move to jurisdictions with services aligned to personal preferences), with W. Norton Grubb, The Dynamic Implications of the Tiebout Model: The Changing Composition of Boston Communities, 1960-1970, 10 PUB. FIN. Q. 17 (1982).

(44) John Douglas Wilson, Theories of Tax Competition, 52 NAT'L TAX J. 269 (1999).

(45) For a contemporaneous account of this issue, see Alvin C. Warren, Jr., Recent Corporate Restructuring and the Corporate Tax System, 42 TAX NOTES 715 (Feb. 6, 1989).

(46) See, e.g., Gene Steuerle, Is Integration of Corporate and Individual Taxes Still Possible?, 44 TAX NOTES 229 (July 10, 1989).

(47) See Gil B. Manzon, Jr. & George A. Plesko, The Relationship Between Financial and Tax Reporting Measures of Income, 55 TAX L. REV. 175-214.

(48) This position is ably presented in Reuven S. Avi-Yonah, Globalization, Tax Competition, and the Fiscal Crisis of the Welfare State, 113 HARV. L. REV. 1573 (2000).

(49) TAX FOUND., CORPORATE INCOME TAXES, available at http://www.taxfoundation.org/taxdata/topic/91.html (last visited Apr. 10, 2007) (providing various tax data).

(50) Under the flat tax, corporations pay tax on net distributions. It will still be necessary, therefore, to apportion that tax among the states. The apportionment will be harder since the distributions might represent income earned over many years from many different states in different proportion. The alternative rule--tax due only to the state from which the distribution is made--would lead companies to shift income to a no-tax state before distribution.