Congressional Budget OfficeSkip Navigation
Home Red Bullet Publications Red Bullet Cost Estimates Red Bullet About CBO Red Bullet Press Red Bullet Careers Red Bullet Contact Us Red Bullet Director's Blog Red Bullet   RSS
PDF
CBO
TESTIMONY
 
Statement of
G. Thomas Woodward
Assistant Director for
Tax Analysis
 
Economic Analysis of Taxing Internet and Other Remote Sales
 
before the
Committee on Finance
United States Senate
 
August 1, 2001
 
This document is embargoed until 10:00 a.m. (EDT), Wednesday, August 1, 2001. The contents may not be published, transmitted, or otherwise communicated by any print, broadcast, or electronic media before that time.
 

Mr. Chairman, Senator Grassley, and Members of the Committee, I am pleased to be here today to talk about collecting sales and use taxes on remote purchases such as those made over the Internet. My remarks are limited to that issue; they do not address other taxes on Internet use.

I will make three central points in my statement.


THE EROSION OF THE SALES TAX BASE

Forty-five states and the District of Columbia impose sales taxes on purchases made within their borders, and those taxes account for about 33 percent of their total tax revenue. Among states with sales taxes, 33 of them also allow localities to impose such taxes; those receipts make up about 12 percent of total local tax revenue. All of those taxing jurisdictions impose an equivalent "use" tax on their residents' purchases of items out of state. Such purchases, called remote sales, include items that have been ordered over the Internet, by telephone, or by mail. It is the collecting of taxes on remote sales that is at issue.

States collect sales and use taxes in different ways. For in-state purchases, taxing jurisdictions require retailers to collect and remit sales taxes, an arrangement that reduces the state's administrative costs for collection from what they would be if the taxes had to be collected directly from consumers. That kind of collection mechanism is not in place for the identical use tax imposed on remote sales: states must largely collect that tax from individual purchasers. While states find it relatively easy to ensure that their residents comply with the sales tax--because it is collected and remitted by retailers--they find it much more difficult to make individuals remit use taxes on out-of-state purchases. For the most part, the use taxes that states and localities collect come from business-to-business purchases that are subject to tax, since those purchases are easier to monitor and audit.

Most taxing jurisdictions want a mechanism for collecting use taxes that parallels the one they have for collecting sales taxes. (For the remainder of my statement, I will use the term "sales taxes" generically to mean both sales and use taxes.) Under such an arrangement, a firm selling to an out-of-state customer would assess and collect the tax and remit it to the customer's home state. However, states are currently limited in their power to impose such a requirement on out-of-state establishments. Only if the firm has a physical presence in the taxing jurisdiction--a situation known as nexus--does that jurisdiction have the legal authority to require sellers to collect and remit the tax.

The Supreme Court has held that when nexus is absent, states' efforts to collect taxes on remote sales run afoul of the Constitution by placing an undue burden on interstate commerce. The Court found that requiring vendors to collect and remit sales taxes on out-of-state purchases would subject firms to an intolerable obligation: that of calculating taxes--for thousands of taxing jurisdictions--that differ in their rates, in the categories included in their tax bases, and in the definitions of goods within those categories. Added to the direct costs of complying with remote collection would be the indirect costs associated with being subject to audit by the administrative tax authorities of all those jurisdictions. The costs of that compliance would probably be much greater than those borne by an in-state retailer that must comply with only a single set of tax laws. Nonetheless, the Court concluded that if the Congress so wished, it could give states the authority to require vendors to collect and remit sales taxes on remote purchases.

Until recently, concerns about collecting such taxes centered on catalog and telephone sales. Now, the growth of electronic commerce (e-commerce) has shifted that focus and heightened those concerns. Indeed, remote sales of all kinds have shown strong growth: the Bureau of the Census estimates that remote retail sales in total (including catalog, telephone, and e-commerce) rose from $35 billion in 1992 to $92 billion in 1999. Of the latter amount, e-commerce is estimated to account for $15 billion, or 17 percent of the total.

Much of e-commerce is not retail in nature--most of it is business-to-business transactions that are exempt from sales tax. As a result, the amount of on-line purchases subject to sales taxes is much smaller than the amount of total sales over the Internet. Data on e-commerce, however, are incomplete. For example, the federal government has been publishing statistics on such sales only since 1998. As of 1999, the annual value of e-commerce totaled $660 billion, with more than 90 percent of that amount attributable to nonretail transactions. Private-sector forecasters expect e-commerce sales to rise to about $2 trillion by 2003.

Because of the paucity of good data and the fact that many remote sales would not be subject to sales taxes anyway, estimating how much revenue states lose by being unable to collect sales taxes is difficult. The General Accounting Office has projected that the revenue loss in 2003 will be between $2.5 billion and $20.4 billion. The mid-point of that range, $11.5 billion, is about 4 percent of the total revenue that states expect to receive from sales taxes. That estimate is not too different from other projections made by forecasters in the private sector.

Although imprecise, such estimates of the erosion of the sales tax base are large enough to generate debate within the states about what should be done to deal with their lost revenue. Of course, options that are more fundamental than remote collection--and that are beyond the scope of this testimony--are open to them, including reducing spending, raising sales tax rates on transactions that remain in the base, raising other existing taxes, or imposing new taxes. Generally, states have indicated that they want the Congress to grant them the authority to require remote sellers to collect sales taxes, and some observers of state tax policy suggest that in exchange for that right, states could change their tax systems to reduce remote sellers' compliance costs.
 

ECONOMIC EFFICIENCY AND REMOTE COLLECTION OF TAXES

Most taxes generate economic inefficiency because they interfere with the choices that producers and consumers would otherwise make. The more unevenly a tax is applied, the more producers and consumers waste resources in their efforts to avoid it. Effectively exempting remote purchases from sales taxation is one source of such unevenness. Consumers are willing to purchase a good remotely even if the total cost of production and delivery exceeds the comparable in-state cost because the money they save in taxes compensates for the money they pay in shipping costs. Similarly, producers are willing to construct facilities in locations where production and shipping costs are high to avoid nexus and the need to charge their customers sales taxes. And as more sales escape taxation over time, states and localities may seek to maintain the same level of receipts by increasing tax rates, which exacerbates the tax's inherent inefficiency.

Effectively exempting goods from taxation because they are purchased out of the taxing jurisdiction thus results in more resources being used to produce and deliver goods than would otherwise be the case. The economic logic of requiring remote collection by vendors is essentially the same as that underlying most proposals for tax reform: a broader tax base combined with a lower tax rate tends to result in less economic inefficiency.

In certain circumstances, a subsidy in the form of special tax treatment for a sector of the economy may not lead to more inefficiency. That may be the case when the sector is characterized by positive "externalities." Externalities are benefits that accrue to people other than the ones who make the economic decision to purchase or produce a good. The benefit that is sometimes cited for the Internet is "network externalities"--in which a person joining the network not only benefits him- or herself but, by adding to the total number of participants in the network, provides benefits to other Internet users who may interact with the new user. Network externalities arising from additional users, however, occur primarily in the early stages of a network's growth. At this point in the Internet's development, there appear to be few external network benefits to be garnered from additional users. Moreover, subsidies that help more people gain access to the Internet already exist. Effectively exempting remote purchases from sales taxes is an indirect and unevenly focused means of promoting the Internet's growth that is unlikely to bring significant benefits in terms of additional users or uses.
 

REDUCING COMPLIANCE COSTS

The compliance costs that retailers would face in collecting and remitting taxes on their remote sales could be reduced. The process by which merchants calculate and remit taxes could be streamlined by using computer technologies and by creating institutional structures that would centralize and simplify the process of directing collections to appropriate jurisdictions. The taxes themselves could be simplified through "harmonization," in which states with sales taxes adopt common definitions for categories of goods in their sales tax bases or adopt more-similar bases for their sales taxes.

Computer technology might reduce the costs that remote sellers incur in assessing sales taxes on purchasers from a variety of jurisdictions. For example, sellers could use shipping destinations to help identify the state or locality for which the tax must be collected and then directly access a database to determine whether the good being sold is taxed in that jurisdiction. The computer system would automatically calculate the appropriate rate and the division--if any--of the total tax between overlapping jurisdictions. Supplementing the software for computing taxes could be institutional arrangements that would provide a central entity to administer the system and remit the collected taxes, and legal protocols that would simplify the process of computing and remitting them. Those protocols might include automated tax compliance systems, simplified vendor registration, and so-called safe-harbor mechanisms designed to limit the potential for multiple audits.

Harmonization holds some promise of decreasing the compliance costs of remote collection because so much of that burden stems from the multiplicity of different sales tax regimes. The categories of purchases subject to taxation vary among jurisdictions since many kinds of products are exempt. In addition, the definitions of exempt categories--such as food, medicine, and clothing--vary from one jurisdiction to the next. Harmonization could range from agreed-upon definitions of categories of purchases, to more similar bases, to uniform rates among states that impose sales taxes. It could also limit how often rates and bases might be changed and provide uniform procedures by which purchasers might obtain tax exemptions.

A number of states that levy sales taxes have already begun to explore ways to lessen the costs of remote collection through an endeavor known as the Streamlined Sales Tax Project. Their efforts could result in a voluntary agreement to simplify the cost of collecting and remitting taxes through changes in the collection process and the tax structures themselves. Simplification or harmonization, or both, could also be a condition of conferring on states the right to require remote collection.
 

OTHER ISSUES

There are three additional issues relevant to the debate about remote sales tax collection that economic analysis can illuminate but not fully address. First, both proponents and opponents of remote collection argue that effectively exempting remote purchases from sales taxes raises issues of fairness. Second, opponents argue that states' inability to force their residents to comply with such taxes works as a desirable brake on the size of the public sector and that requiring remote collection would eliminate that restraint. Third, both sides raise concerns about state and local fiscal autonomy.

Fairness

One aspect of the issue of fairness relates to consumers. By effectively exempting remote sales from taxation, a jurisdiction's tax system violates the standard expectation that taxpayers in similar situations will be treated in the same way. A violation of that kind occurs when two people with equivalent purchases of goods can pay different taxes because one person buys remotely without paying the sales tax. In addition, because access to the Internet--as well as the use of credit cards and other instruments necessary to buy something on-line--tends to rise with income, effectively exempting e-purchases from taxation causes the states' sales tax systems to be more regressive than they would be without such an exemption.

The other issues related to fairness involve businesses. Proponents of collecting taxes on remote sales argue that in-state vendors are at a competitive disadvantage relative to out-of-state firms that do not collect the tax. That contention is really about whether the price of a good being sold reflects the cost of the resources used in producing it--which is essentially the argument about efficiency that I discussed earlier. Opponents of remote collection maintain that businesses that receive no benefits or services from a state or locality by virtue of being located elsewhere should not have to pay taxes to support those benefits and services. Sales taxes, however, are taxes on consumers that are collected by firms. Ending the effective exemption for remote sales would help ensure that they were borne by consumers in a jurisdiction and not by businesses or others involved in producing a good.

Constraining the Size of the Public Sector

Some opponents of remote collection regard the public sector as too large or as subject to an inherent bias toward a larger-than-optimal size. From their point of view, the difficulties that states face in collecting taxes on remote sales act as a brake on the growth of government. For example, a jurisdiction might find it difficult to raise sales tax rates or enact other taxes to compensate for the revenue lost on remote sales. Remote collection of sales taxes might mean collecting more taxes. Thus, lower collection costs might tip the balance toward more taxes and bigger government. If jurisdictions decided to reduce compliance costs by harmonizing their state sales taxes, interstate tax competition--another potential brake on the public sector's size--might also decline.

State and Local Fiscal Autonomy

The remaining issue looming over the debate about remote collection of sales taxes is that of state and local fiscal autonomy. Granting states the authority to require vendors to carry out those collections in exchange for harmonizing the various sales tax regimes would constrain the states' options for tailoring their sales taxes to their citizens' preferences. As a result, harmonization in some forms could reduce differences in sales taxes among jurisdictions. At the same time, formally exempting purchases over the Internet from sales taxes or weakening the nexus standard --proposals that have been advanced by some opponents of remote collection--would also undermine state and local fiscal autonomy by imposing new limitations on states' and localities' ability to collect taxes on some sales.