The Federalism Project

American Enterprise Institute


Submitted to the 




August 1, 2001


Michael S. Greve
John G. Searle Scholar
Director, Federalism Project
American Enterprise Institute


1. Introduction 

Mr. Chairman, members of the Committee: thank you very much for giving me the opportunity to testify on the vexing question of interstate sales taxation.

The thrust of my testimony is that the Congress can and should harmonize the contending principles of federalism, economic efficiency, and tax equity. At first sight, that may seem impossible. The Governors’ insistence on the states’ rights seems to run headlong into remote sellers’ strenuous objections to being regulated by states with which they have no tangible connection. The e-commerce camp’s insistence on that position, in turn, conflicts with the principle of tax equity, championed vigorously by mainstreet retailers. And so on. Federalism, equity, and efficiency can be reconciled, however, by providing that interstate sales through whatever channel (direct, catalogue, or Internet) are subject to sales taxation at their point of origin, meaning the seller’s home state—not, as is currently the case, on the basis of their destination, meaning the customer’s home state. As a practical matter, that solution requires federal legislation.

I harbor no illusions about the political viability of adopting an origin-based approach, at least in the current round of legislation. While the proposal has been floated in the tax literature,[1] in the publications of some think tanks (such as the Competitive Enterprise Institute, the Heritage Foundation, the Cato Institute, and—in an international context—the Progressive Policy Institute),[2] and in some versions and applications before the Advisory Commission on Electronic Commerce,[3] it has received little, if any, public and legislative debate and consideration, and time is running out. Alternative proposals under consideration, however, seem unattractive, unworkable, or both. Moreover, the choice between origin and destination as the basic regulatory principle has broad applications in other arenas—consumer privacy, for example, and especially the international trade matters under this Committee’s jurisdiction. In those arenas, the choice of the wrong jurisdictional principle—destination—would entail terrible consequences. In that light, I respectfully urge the Committee, and the Congress, to refrain from hasty e-tax legislation that might set a bad precedent and preclude a shift to origin-based taxation at a more opportune time.

2. Principles

In light of the intense controversy over e-commerce taxation, it may be helpful to begin with the principles that are (or at least should be) common ground.

Efficiency. State and local governments, the business community, customers, and tax economists all agree that sales taxes should be easy and cheap to administer. All agree, moreover, that the existing sales and use tax regime is absurdly cumbersome and expensive. Compliance costs are very high, especially for smaller and medium-sized firms.[4] Even with the best intentions (and the best tax software), companies find it inordinately difficult to determine their tax remittance obligations in thousands of jurisdictions with different, and constantly changing, tax rates, definitions, and reporting requirements.

Equity. Happily, “[v]irtually all concerned parties agree that state taxes on electronic commerce should be economically neutral.”[5] Less happily, not everyone agrees on the meaning of “neutrality” and its cousins, such as “equity” or “fairness.” Tax experts argue that taxes should not divert economic resources from one use to an alternative use—in this instance, from conventional sales to e-commerce sales. Naturally, “mainstreet” merchants and state and local government lobbies have made much of this argument—arguably, too much, since no conceivable state and local sales tax regime can be entirely neutral.[6]   Nonetheless, the argument that our sales tax regime should not artificially favor some sales channels over others possesses considerable force. The rejoinder that the existing sales and use tax regime is in fact neutral with respect to all forms of “remote” commerce does not fully meet legitimate concerns over tax neutrality and equity.

Federalism. Governors and state organizations argue with great conviction that Congress should not interfere with the states’ traditional authority and autonomy over their traditional sources of revenue. For reasons explained below, I am inclined to think that the states have in the e-commerce debate pushed that principle beyond its breaking point. The principle is eminently reasonable, however, in this crucial respect: over the entire course of American history, the states and the national government have remained very respectful of each other’s revenue sources and taxing authority.[7]

Early in the e-commerce debate, proposals surfaced to nationalize sales taxes and to distribute the proceeds to the states. Those proposals, mercifully, appear to have died a well-deserved death. Any form of joint state-federal taxation would eventually transform the states from autonomous actors into supplicants and administrators of federal largesse. That result cannot be in anyone’s interest. 

3. Compromise?

Current compromise efforts—reflected, for example, in various bills before the Senate Commerce Committee[8]--essentially attempt to reconcile the clashing interests and principles by granting state and local governments expanded tax authority over “remote” commerce in exchange for a credible commitment to “simplify” and “harmonize” sales and use taxes. The search for a compromise along these lines is in my judgment a fool’s errand. Simplification has proven elusive for decades, and not for lack of trying.[9] It will continue to prove elusive. For the sake of brevity, I mention only two of the many reasons:

  • “Simplification” must, at a bare minimum, mean a single tax rate per state. Even under that regime, sellers would be stuck with 46 different sales tax regimes, but—less is more, in this case—46 state taxing jurisdictions beat 7,500 local jurisdictions any day of the week. A single-rate regime, however, would effectively wipe out the tax autonomy of local jurisdictions. That should not happen, and it is not going to happen. For this reason alone, “simplification” is a chimera.

  • From a tax efficiency standpoint, what really needs simplification is not the tax rate but the tax base.[10] Questions as to whether Air Jordans constitute taxable sportswear or tax-exempt footwear are legendary, as are questions as to whether potato chips and French fries do or do not constitute food. (The National Governors Association has, without a sense of irony, issued a press release on the complexities of taxing a marshmallow.) For all the easy ridicule, however, most of the seemingly absurd complexities have plausible reasons. Variations in the tax base reflect political efforts to favor domestic over foreign industries; attempts to mitigate the regressive effects of sales taxes; differences in local customs and habits; varying social judgments concerning individual consumption patterns that ought to be discouraged or encouraged; and many other considerations. Even if all the variations and exemptions were wholly irrational and illegitimate, which they are not, the forces that produced them are not going to go away. Even if it were possible to harmonize the sales tax base once (an extremely doubtful proposition), variations would soon re-appear.

Simplification and efficiency, we are being reassured by state officials, are nonetheless achievable through centralized, computerized sales tax data collection and administration under the auspices of a “Streamlined Sales Tax Project” (SSTP), sponsored by the Multistate Tax Commission and the Federation of Tax Administrators. Comparable experiments to centralize the calculation of local taxes in a few states, however, provide little reason to believe that the proposed project is technically feasible, and a representative of Taxware, the producer of perhaps the most sophisticated tax compliance software on the market, has testified that existing technology simply cannot cope with the maze of definitions, exemptions, and reporting and remittance requirements.[11] Moreover, the SSTP proposal raises grave privacy concerns. Considering the pronounced public unease over personal data collection and use, a governmentally sponsored and administered data collection system on interstate sales is not an appealing idea.

In addition to these pragmatic considerations, an important point of principle is at stake. We do not really have a problem with sales taxes. Rather, we have a problem with use taxes, which are owed by customers to their various home states (and subordinate jurisdictions) on purchases from “remote” out-of-state sellers. For practical and political reasons, governments find it inconvenient to collect use taxes—except on big-ticket, easily traced items, such as boats and cars—from individual, in-state citizens. Thus, the common practice is to impose the tax collection and remittance obligations on out-of-state sellers.

States are of course free to levy use taxes on their own citizens. The question is whether they may impose the obligations to calculate, collect, and remit those taxes on out-of-state sellers. Under existing Supreme Court precedent, the states are forbidden from doing so unless the seller has a “nexus,” such as a store, in the taxing jurisdiction.[12] That is why state and local governments are asking the Congress, under its authority to regulate interstate commerce, to lift that restriction and to extend the states’ regulatory and tax authority to “remote” sellers.

Even the existing sales and use tax regime, however, poses a serious threat to our most elementary political principles. Its extension to remote sales, no matter how “simplified” and “harmonized,” would greatly exacerbate that threat.

First, the imposition of tax collection, reporting, and remittance obligations on out out-of-state parties severs (as the economists say) the political incidence of taxation from its economic incidence. That is a variation on a theme our ancestors called “taxation without representation.”

Second, the genius of American federalism, as Justice John Paul Stevens has put it, is that citizens choose their state, not the other way around.[13] That principle applies not only to permanent physical relocations but also to work, tourism, and other activities—including shopping. Under the existing tax regime and especially under its proposed extension to remote sales, however, a citizen’s home state tax law tags after him like a junkyard dog. Such a system negates the citizen choice that is federalism’s principal attraction.

It also, and third (and relatedly), constitutes an insult to state sovereignty. States’ rights, like individual rights, must end where another’s rights begin. Federalism means that states may regulate and regiment their own citizens—but not the citizens of other states. The imposition of use tax obligations by each state on foreign entities amounts to mutual regulatory aggression on other states’ corporate citizens. That is not federalism but very nearly its opposite.

In sum, the sales-tax-expansion-with-simplification agenda will not work. Even if it could be made to work, it would be a bad idea. In that light, the alternative of an origin-based sales tax system merits consideration. 

4. Origin-Based Taxation

The effect of the existing sales-and-use tax regime, as noted, is to tax interstate sales transactions on the basis of their destination—that is, the customer’s residence. All the seemingly intractable problems of the internet sales tax debate—in particular, the differential treatment of “mainstreet” and “remote” sales, the extravagant compliance and administrative costs—stem from the choice of destination as the regulatory principle, not from the existence of a decentralized sales tax system per se.

Under an origin-based system, in contrast, interstate sales of all descriptions, through all channels, would be based on the seller’s instead of the buyer’s domicile state—that is, the company’s principal place of business, as defined (for other purposes) by the Uniform Commercial Code.[14]

That one swift move from destination to origin would solve equity and efficiency problems.’s sales would be taxed in the same fashion, at the same rate, by the same entity, as would the sales of the local book store—that is, by the state of Washington. No discriminatory tax treatment would occur unless a particular state or local jurisdiction decided, for the sorts of industrial policy reasons that often induce jurisdictions to favor some industries over others, to extend tax advantages (or disadvantages) to some sales channel or other. Administrative and compliance costs would plummet. Regardless of how and where a company’s products are sold, each company will be subject to reporting and remittance obligations for interstate sales only in its domicile state (and perhaps its local jurisdiction), and nowhere else.

At the same time, an origin-based system is fully consistent with sound federalism principles. Each state would be free to tax and regulate its own businesses and citizens as it sees fit. Each state’s regulatory autonomy and authority, however, would stop at the border—which is precisely where they ought to stop.

The origin principle is a perfectly natural choice. We already follow it, in real life, with respect to local sales, even if the parties are from different states. If I, as a Virginia resident, buy a Lacrosse stick for my son on a business trip to North Carolina, I will be charged the North Carolina sales tax. The company that sells me the stick cares not one whit whence I came, where I reside, or where the stick is going to be used. If my son purchases the next stick from the same company, which has no store in Virginia, over the Internet or by phone or mail order, he (or more likely I) will not pay North Carolina’s tax. We will instead owe the Virginia use tax—technically speaking, since neither of us has ever paid or been asked to pay that tax. Under the existing and under the proposed, “simplified” system, it matters whether the stick came to me, or I came to the stick. An origin-based system would harmonize the tax treatment.

We also follow the origin principle, even in interstate transactions, with respect to (of all items) flowers[15] and, since last year, mobile telephone calls. One reason why the origin principle has proven so readily acceptable in those areas is an expected reciprocity of advantage. A few areas (such a college towns) may experience a net export of flowers, thus reaping a benefit from origin taxation; a few other areas (such as those with lots of retirement communities) may experience sizeable net imports. By and large, though, states are content to ignore the question (Where have all the flowers gone?) because the flows will average out. So, for that matter, will telephone calls.[16]

The expectation that such average reciprocity would not prevail is a central reason for the existing, destination-based sales and use tax system. In the 1930s, when that system came into being, “consumer” states feared that “producer” states would reap all the advantages from an origin-based system, thus leaving stranded the states that were most in need of revenues.[17] That concern, however, while understandable in an industrializing country with enormous economic heterogeneity among the various states, seems increasingly incongruous in a far more homogeneous, service-oriented and information-based economy.

It is true that a shift from destination-based taxation to origin-based taxation would tend to benefit the net exporters of taxable retail products and hurt net importers. That objection, however, gains its plausibility chiefly from an unstated baseline comparison with a destination regime under which all interstate sales will in fact be taxed. That assumption is manifestly unrealistic. (As noted, use taxes are rarely collected under the existing system, and even the proposed SSTP tax cartel would exempt many purchases.) And while producer states might gain from a shift to origin-based taxation, that advantage is easily swamped by tax advantages that the “losing” consumer states may possess.[18] 

5. Objections

The elegant, origin-based sales tax solution is vulnerable to two objections. The first of these is based on considerations of efficiency. The second, politically more potent objection arises from concerns over possibly excessive tax competition.

Efficiency. An origin-based system implies jurisdictional variations with respect to both the tax base and the tax rate. Those variations will at the margins shape and, relative to a completely neutral system, “distort” both purchasing decisions and firm location. That objection, however, applies also to the existing system and, for that matter, to the SSTP.[19] 

Competition.  Under an origin-based sales tax system, every jurisdiction will attempt to match the largest number of sellers and purchasers. All else being equal, one might think, the sales tax rate will be zero in every jurisdiction.[20]

The trouble with this “race-to-the-bottom” argument is that all else is not in fact equal. Sales taxes are merely one stick in a bundle of services and obligations that are being offered by each jurisdiction. Thus, a jurisdiction that provides an educated labor force, an excellent infrastructure, a favorable regulatory environment, a sensible and efficient judicial system, or “quality of life” attractions will be able to exact a sales tax. An unattractive jurisdiction that drives up the cost of doing business, meanwhile, will be unable to compensate those self-inflicted disadvantages by becoming a “sales tax haven.”

It is true, of course, that an origin-based tax system would, relative to the existing sales tax regime, exert downward pressure on sales taxes and, quite probably, change the mix of the tax burden in many and perhaps most jurisdictions. The magnitude of that effect cannot be forecast with any confidence; its direction, however, is certain.

The question, though, is not whether we do or not want sales tax competition. We have that competition in any event: the evidence is visible up and down I-95, from Delaware’s blaring “Tax Freedom” billboards to the busloads of New York shoppers at Potomac Mills, VA. The benefits of tax competition, however, are very unevenly distributed. They are more available to the residents of Philadelphia than to those in Salt Lake City; more available to prosperous citizens with cars than to less wealthy and mobile citizens.

I call that phenomenon choice and competition. State and local officials call it tax “evasion” or “flight” and the jurisdictions that offer it, tax “havens.” Under any name, though, sales tax choice or “evasion” will continue no matter what the United States Congress, or any individual state, may do. The only legislative choice is to contain choice and competition within its current, artificial and inequitable boundaries or else, to expand and democratize it. Origin-based taxation implies a choice for expanded competition.

6. Origin Versus Destination: Implications  

The choice between origin and destination as a regulatory principle reaches far beyond the e-commerce tax question—substantively, and geographically. As to substance, consider the intensely contested question of internet privacy and consumer protection.[21] Let the service provider’s home state govern transactions with customers, and market participants will, sooner rather than later, sort themselves into jurisdictions that match their privacy preferences.[22] Let the customers’ state govern the transactions, and providers must tailor their product to each jurisdiction’s specifications or, if tailoring proves impossible or excessively expensive, comply with the most restrictive jurisdiction, which will by definition reflect nobody else’s preference. Since either result is intolerable to business, customers, and most states, the destination principle will prompt centralized intervention. In the end, then, states do not really have a choice between regulatory autonomy with or without competition. They only have a choice between autonomy under the competitive conditions of origin-based regulation—or else, neither competition nor autonomy.

Domestically, we can always trump the ill effects of decentralized, destination-based regulation by means of (yet another) preemptive federal law. While such centralized interventions are not always wise or well-designed, protecting the national common market from regulatory obstacles is, in the end, a constitutional mandate, entrusted to the United States Congress. On an international scale, in contrast, efforts to trump parochial, destination-based regulation through “harmonization” entails the creation of anti-competitive, unaccountable bureaucracies.[23]

Democratic and Republican administrations alike have, by and large, wisely defended international competition against “harmonization.” We have likewise resisted the notion that French judges or Brussels bureaucrats have any business regulating the content and practices of American-based websites solely because those sites are accessible from European destinations.

We can in some sense afford to suppress regulatory and tax competition here at home and yet champion it in the international arena—simply by throwing our considerable weight around. We do so, however, at the peril of international resentment and recriminations. It is much better to practice at home the competition that we preach abroad—both because we should set an example for a more competitive world economy, and because our domestic practices are not easily contained. Origin-based e-commerce taxation, for a pristine example, is easily scaled internationally: let us tax our companies’ international sales here at home, let foreign countries do the same to their companies, and let that be the end of the matter. Destination-based taxation on an international scale, in contrast, means that we have no cause for complaint when German tax inspectors insist on auditing Apple Computer’s books, the better to verify the company’s compliance with tax obligations incurred on sales from Aachen to Aschaffenburg. It means, in the end, some sort of international SSTP. I take it that we would not consent to such a scheme. The best way to resist it is to resist it here at home.

7. Conclusion  

At the present stage of the legislative process, it would be best to refrain from legislation that would affirmatively foreclose future consideration of an origin-based sales tax regime. A congressional endorsement of the SSTP would, unfortunately, likely entail that result. For that reason, I believe that the proposal should be rejected.

Alternative courses of action would keep the origin-based option open. An extension of the existing moratorium is one such option; a codification of “nexus” standards may be another option, depending on the standards chosen.[24] Finally, doing nothing at all may well be an acceptable course of action. Inaction would mean a missed opportunity to reform a sales tax system that is uniformly, and rightly, viewed as an extravagant absurdity. It would, however, preserve the option of reforming that system at a more opportune time, after more systematic consideration. 

Respectfully submitted, 

Michael S. Greve
American Enterprise Institute
1150 17th Street, N.W.
Washington, D.C. 20036
(202) 862-4874
[email protected] 

[1]   See, e.g., Andrew Wagner & Wade Anderson, Origin-Based Taxation of Internet Commerce, State Tax Notes (July 19, 1999), 187; Terry Ryan & Eric Miethke, The Seller-State Option: Solving the Electronic Commerce Dilemma, State tax notes (Oct. 5, 1998), 881.

[2]  See, e.g., Jessica Melugin, Internet Sales Taxation: Beyond the Moratorium (Competitive Enterprise Institute Policy Brief, Mar. 28, 2000); Adam D. Thierer, E-Commerce: A Taxing Issue (Mar. 23, 2000) (; Aaron Lukas, Tax Bytes: A Primer on the Taxation of Electronic Commerce (Cato Institute, Dec. 17, 1999), 37-38; Shane Ham & Robert D. Atkinson, A Third Way Framework for Global E-Commerce (Progressive Policy Institute, March 2001).


[3]  See Andrew Wagner & Wade Anderson, Proposal of an Origin Based Tax Solution for the Possible Taxation of Digitized Products Sold Over the Internet (Nov. 8, 1999) (


[4]   Robert J. Cline & Thomas S. Neubig, Masters of Complexity and Bearers of Great Burden: The Sales Tax System and Compliance Costs for Multistate Retailers (Ernst &Young, Sept. 1999) (estimating compliance costs from 14 percent of taxes collected for large retailers to 87 percent for small retailers).


[5]  Walter Hellerstein, State Taxation of Economic Commerce: Preliminary Thoughts on Model Uniform Legislation, state tax notes (Apr. 28, 1997), 1315.


[6]  The only locationally neutral (and in that sense efficient) sales tax is a uniform national sales tax. However, even a uniform national sales tax on products, but not services, will at the margin induce an inefficient substitution of services for products. Thus, the only truly neutral tax is a tax on both products and services or else, a national prohibition on any sales tax. I take it that state and local governments do not advocate either of these options.


[7]  The segregation of revenue sources is a bit of a miracle, considering that Washington and the states (and for that matter local governments) have not been remotely so conscientious in separating funding and administrative responsibilities.


[8]  See, e.g., S-288 (Sen. Wyden); S-512 (Sen. Dorgan).


[9]  See, e.g., Kendall L. Houghton & Walter Hellerstein, State Taxation of Electronic Commerce: Perspectives on Proposals for Change and Their Constitutionality, 2000 B.Y.U. L. Rev. 9, 29-30 (describing unsuccessful simplification efforts).


[10]  See Daniel Shaviro, federalism in taxation: the case for greater uniformity (American Enterprise Institute, 1993).


[11]  Jon W. Abolins (Chief Tax Counsel and Vice President, TAXWARE International, Inc.), Testimony to the House Judiciary Subcommittee on Commercial and Administrative Law, July 18, 2001.


[12]  Quill v. North Dakota, 504 U.S. 298 (1992).


[13]  Saenz v. Roe, 526 U.S. 489, 501-11 (1999).


[14]  Local, in-state sales would be taxed, as they are now, at the locally applicable rate.


[15]   Ryan & Miethke, The Seller State Option at 883 n. 11.


[16]  Cf. Goldberg v. Sweet, 488 U.S. 252 (1989) (sustaining state excise tax on interstate telecommunications on, inter alia, reciprocity grounds).


[17]  Ryan & Miethke, The Seller State Option at 888.


[18]  I have attached a table of commodity flows among the states for the most recent year (1997) for which that data is available. For any number of reasons, the table provides no insight at all as to whether any given state would gain or lose from any given sales tax regime. (E.g., the big exporters tend to be producers of agricultural and food stuffs, which are rarely subject to sales taxes.) For what it is worth, though, note that most of the big import states have access to revenue sources that are unavailable to other states. Florida and Utah have the luxury of taxing tourists, whereas “producer” states like Iowa and Nebraska do not. An arguable origin-tax “loser” such as Montana can charge severance taxes for raw materials; a “winner” like New Hampshire cannot. Each state has competitive and comparative advantages—some deserved, others not. It is a grave mistake to look at one dimension in isolation—or to think that national policy could or should equalize the conditions among the states.


[19]  By technical economic criteria of locational efficiency, any form of tax competition is ipso facto inefficient. The question of whether the gains from competition exceed its efficiency costs is the subject of a voluminous literature. One’s confidence in the centralized, supposedly efficiency-enhancing approach will stand in direct proportion to one’s confidence in the ability of centralized political institutions to generate tax regimes that conform to the optimality criteria on a blackboard at the University of Chicago or MIT. Since my own confidence on that score is nil, I opt for real competition over theoretical efficiency. 


[20]  That result might in fact obtain if retailers were entirely free to designate their home state, or to designate their place of incorporation as their home state. The principal-place-of-business rule, in contrast, disciplines the sellers’ choices.


[21]  For the sake of brevity, I limit the discussion to a single example. Many other examples come readily to mind, and some of these overlap substantially with the e-tax debate.  The contemplated extension of state e-tax authority, for instance, has generated controversy as to whether that extension is to apply only of sales and use taxes or also to the allocation of interstate business income and its taxation. For yet another example, economists who otherwise oppose origin taxation for tangible products have tentatively endorsed it for the taxation of digitized products in e-commerce, a closely related problem. See, e.g., Houghton & Hellerstein, State Taxation of Electronic Commerce at 54 and sources cited ibid.


[22]  See Bruce H. Kobayashi & Larry E. Ribstein, “A Recipe for Cookies: State Regulation of Consumer Marketing Information” (American Enterprise Institute, May 2001) (available at


[23]   For a very thoughtful discussion of international jurisdiction and the unworkability of the destination principle in that context see Ham & Atkinson, A Third Way Framework for Global E-Commerce.


[24]   A codified “nexus” standard can in principle generate any jurisdictional regime, from an extension to remote sales (under an extremely loose nexus test) to a de facto origin system (under a standard requiring a very tight nexus, such as the actual operation of an in-state store).