September 14, 1999

Statement of the Center on Budget and Policy Priorities
To the Advisory Commission on Electronic Commerce

The Center on Budget and Policy Priorities, a non-profit research institute that studies government programs and public policy issues that have an impact on low-income Americans, appreciates the opportunity to submit comments to the Advisory Commission on Electronic Commerce.

The single most important decision confronting the Commission is whether to recommend federal legislation empowering states to require all large businesses to charge applicable state and local sales taxes on all Internet, mail-order and other "remote sales" and remit the taxes to the states in which the businesses' customers are located. Such an "expanded duty to collect" could be conditioned, of course, on significant simplification and standardization of sales taxes by the states.(1)

The enactment of such legislation would eliminate the current de facto sales tax exemption that applies to most Internet and mail-order purchases made by businesses and individuals.(2) Failure to close this obsolete loophole would have three profoundly negative impacts on low-income Americans:(3)

If the de facto sales tax exemption for Internet and mail-order sales is not eliminated, low-income consumers without the resources or desire to shop on-line or by mail will pay a disproportionate share of state and local sales taxes. Businesses and affluent households increasingly will avoid sales taxes that low-income households consigned to shopping in stores will continue to pay.

At present, the de facto sales tax exemption that applies to mail-order and Internet purchases provides disproportionate tax savings to the more affluent households and businesses most likely to shop through these mechanisms. Households with incomes above $80,000 are more than twice as likely to shop from catalogs as households with incomes below $25,000, while businesses account for approximately 40 percent of catalog purchases.(4) Internet purchasing is even more skewed toward relatively affluent households and businesses. Owning an Internet-equipped computer is the price of admission to the cyber-mall, and households with incomes over $75,000 remain more than eight times as likely to have home Internet access as households with incomes between $10,000 and $15,000.(5) Accordingly, e-commerce consultant Forrester Research estimates that while households with incomes below $25,000 made 25 percent of all retail purchases last year, they made only 6 percent of on-line purchases.(6) And much of the sales tax savings associated with Internet sales flows to businesses; business purchases on the Internet are five times larger than consumer purchases.(7)

The de facto sales tax exemption for Internet purchases is likely to be disproportionately beneficial to businesses and more affluent households for the foreseeable future. Business Internet purchases are expected by most electronic commerce industry analysts to remain many multiples of consumer purchases.(8) With respect to households, low incomes per se may prove less of a barrier to accessing the Internet than they are at present as the cost of computers and Internet access services continue to drop. Nonetheless, a substantial "digital divide" between more affluent and lower-income households in access to the Internet for shopping and other purposes seems likely to persist due to other factors — particularly gaps in education and training needed to use the Internet effectively.

The latest (July 1999) U.S. Commerce Department study of the "digital divide" found that 48.9 percent of households headed by someone who had completed college used the Internet at home, whereas only 16.3 percent of households headed by someone with a high school education did so. Only 6.3 percent of households headed by someone who started but failed to complete high school used the Internet.(9) The gap in at-home Internet access between more- and less-educated households grew significantly between 1997 and 1998; for example the rate of home Internet access by college-educated households exceeded that of high-school drop-out households by 35.3 percentage points in 1997 and 42.6 percentage points in 1998.(10) Educational attainment in the population changes only slowly. Because low educational attainment by household heads is closely associated with both low rates of Internet use and low household income, for the foreseeable future it seems likely that lower-income households will remain much less able to access the Internet for shopping than more affluent households.

The persistent income gap in catalog shopping is another piece of evidence suggesting that low- and moderate-income households will continue to be less likely to shop on the Internet. The barriers to catalog shopping are relatively low: the ability to read a free catalog, fill out an order form, and visit a Post office or store where cash can be converted into a money order. Despite these relatively low barriers, only 29.3 percent of households with incomes below $25,000 shop by catalog at present, more than 100 years after the Sears catalog came into being.(11) With the ability to shop on the Internet depending on the availability of a relatively complex, modem-equipped computer or Web TV device, an Internet access account subscription, some knowledge of how to find vendors on the Internet, and a general-purpose credit card, it seems likely that substantial gaps in Internet shopping between upper income, college-educated and lower-income, less-educated households will persist for a considerable period of time.

In sum, if states and localities remain effectively powerless to tax goods and services purchased by relatively affluent households and by businesses through the mail and on the Internet, lower-income households consigned to shopping in local stores are likely to continue to bear more than their fair share of sales taxes.

States and localities may seek to preserve a target level of sales tax revenue in the face of sales tax base erosion resulting from on-line purchasing by increasing sales tax rates. If state and local governments increase sales tax rates, low-income households would face higher absolute sales tax burdens in addition to paying more than their fair share of sales taxes.

The erosion of the sales tax base resulting from on-line purchasing by affluent consumers and businesses could compel states and localities to raise sales tax rates to preserve a desired level of sales tax revenues. A vicious cycle leading to ever-higher sales tax burdens on the poor could ensue: tax rate increases could encourage more on-line buying by businesses and the affluent, forcing further rounds of rate increases. Ultimately, low- and moderate-income households unable to buy on-line could be forced to devote more income to paying sales taxes — money that could be used instead for education, transportation, child care and other services the poor need to participate in the labor force.

State and local governments, which already may be losing on the order of $5 billion in sales tax revenues annually from their inability to tax most mail-order sales, could be losing an additional $10 billion annually in just a few years if Internet purchases remain effectively tax-exempt.(12) This loss of revenue could significantly impair the ability of some states and localities to meet demands for education funding and other services that enable the disadvantaged to get ahead in an increasingly technology-oriented economy.

Numerous studies project hundreds of billions of dollars in annual purchases by consumers and businesses over the Internet just three to four years from now. (13) Not all such purchases represent lost sales tax revenues to state and local governments, of course; some will be goods and services that are not taxed, and others represent purchases upon which the tax will be remitted by the seller or the purchaser. Nonetheless, given current growth projections for electronic commerce, state and local governments conservatively could be losing $10 billion in tax revenues annually in just a few years from untaxed Internet sales — this in addition to losses from traditional mail-order purchases.(14)

Even assuming no growth in traditional direct marketing sales between now and 2003, the combined annual revenue loss from untaxed Internet and mail-order sales in 2003 could be $15 billion ($5 billion in lost revenue from untaxed mail-order sales plus $10 billion from Internet sales). That should be regarded as significant by any objective standard; it is, for example, more than two and one-half times what state and local governments currently spend on public libraries.(15) It is the equivalent of $50,000 salaries for 300,000 teachers. The economic opportunities available to the poor could be reduced if tax base erosion impairs the ability of states and localities to finance public services, such as education, job training, library, child care, health, and similar services.

These adverse impacts of the de facto tax exemption for Internet and mail-order purchases can be avoided if Congress ensures equal tax treatment of the customers of Internet, mail-order, and brick-and-mortar retailers. Equal treatment can be achieved by empowering state and local governments to require all large businesses to charge and remit sales taxes whether or not they are physically present in the jurisdictions where their customers reside. The Advisory Commission has a crucial role to play in strongly recommending this equal treatment.


The Center on Budget and Policy Priorities submits this statement to the Advisory Commission on Electronic Commerce on its own behalf and not on behalf of any other clients, persons, or organizations.

Submitted by:
Michael Mazerov
Senior Policy Analyst
Center on Budget and Policy Priorities
820 First Street, NE, Suite 510
Washington, DC 20002
phone: 202-408-1080
e-mail:
[email protected]


End Notes

1. All of the major state and local government organizations are on record in support of a trade of such an expanded duty to collect for a commitment to standardize and simplify sales tax policies and rules. See, for example: National Governors' Association, Policy EC-12, Streamlining State Sales Tax Systems, ; National Conference of State Legislatures, resolution on Electronic Commerce and the State Sales and Use Tax, www.ncsl.org/statefed/ commerce.htm#ecommerce .

2. The de facto sales tax exemption for most Internet and mail-order sales flows from the U.S. Supreme Court decisions in National Bellas Hess vs. Illinois (1967) and Quill vs. North Dakota (1992). In both cases the Court held that a state could not require an out-of-state business to charge and remit sales tax (technically, use tax) on sales made to the state's residents if the company's contact with the state was limited to communicating with its customers and delivering goods to them by common carrier (e.g., United Parcel Service) and the U.S. mail. Although a state's residents are legally obligated to pay applicable use taxes directly to the state even if the seller does not charge them, compliance with this requirement by businesses is uneven and by individual consumers almost non-existent. Most experts on this issue agree that states cannot cost-effectively collect more than a small share of the use taxes that are legally due on interstate purchases by households unless the seller collects and remits them at the time of sale — at least not without detailed reporting of consumer purchases by sellers to state tax authorities that would raise significant privacy concerns. For a primer on sales and use taxes and the implications of the Bellas Hess and Quill decisions, see: Michael Mazerov and Iris J. Lav, A Federal "Moratorium" on Internet Commerce Taxes Would Erode State and Local Revenues and Shift Burdens to Lower-Income Households, Appendix, www.cbpp.org/512webtax.pdf .

3. The Merriam-Webster dictionary defines "loophole" as "an ambiguity or omission in the text through which the intent of a statute, contract, or obligation may be evaded." This is a proper characterization of the de facto sales tax exemption that applies to the vast majority of Internet and mail-order sales as a result of the Bellas Hess and Quill decisions (see note 2). There are very few mail-order companies anymore whose contacts with states are limited to common carrier and U.S. mail communication and delivery; many mail-order companies place advertisements in local newspapers, deliver their wares using contractors that perform services at the time of delivery, accept credit cards issued by banks in their customer's states, etc.. By definition, Internet sellers (a category which now includes nearly all of the large, traditional mail-order companies) have contact with their customers via the Internet, whose categorization as a common carrier network they vociferously resist in other venues because it would instantly subject them to a vast array of federal telecommunications regulatory provisions. Notwithstanding these contacts beyond "common carrier and the U.S. mail," the vast majority of Internet merchants and mail-order companies refuse to collect state and local sales taxes on the grounds they are immunized by the Bellas Hess/Quill decisions. Until further litigation provides the Supreme Court with an opportunity to clarify the intended scope of its safe-harbor for interstate sales, the tax immunity claims of most remote sellers can reasonably be characterized as taking advantage of a loophole. Of course, careful consideration and pro-active resolution of this issue by the Advisory Commission and Congress is far preferable to litigation. In the absence of such action, however, further litigation seems all but inevitable.

4. Catalog purchases by income level: Catalog Age 1999 Consumer Catalog Shopping Survey, May 1999. Business catalog purchases as share of total: "Study: Catalog Sales Will Grow at Steady Pace," DM News, June 2, 1998. Of course, not all catalog or Internet purchases by consumers or businesses are subject to sales tax.

5. U.S. Department of Commerce, Falling Through the Net: Defining the Digital Divide, A Report on the Telecommunications and Information Technology Gap in America, July 1999, Chart I-21, p. 25. Available at www.ntia.doc.gov/ntiahome/digitaldivide . The ability to access the Internet at home rather than at work or at public libraries seems the relevant determinant of how likely households are to engage in on-line shopping routinely. Employers and libraries are likely to discourage the use of computers for shopping, and it also seems unlikely that individuals would make frequent trips to libraries to use computers for shopping.

6. November 1998 Forrester Research, Inc. study summarized in "Digital Commerce: The Issues," CQ Outlook (Congressional Quarterly), February 20, 1999, p. 14.

7. "Net Business Trade Soaring," CNET News.com, December 21, 1998, citing Forrester Research, Inc. study.

8. See source cited in note 7. As of last December, Forrester was forecasting that 2002 business-to-business Internet sales would be 11 times larger than sales to consumers.

9. Falling Through the Net, Table I-4(d), p. 32.

10. This 7.3 percentage point increase in the "digital divide" based on education between 1997 and 1998 represented a 20.7 percent growth rate. In contrast, the rate of home Internet use by households with incomes above $75,000 exceeded the rate of use by households with $15,000-$20,000 in income by 44.3 percentage points in 1997 and by 52.9 percentage points in 1998 — a 19.4 percent growth rate in the "digital divide" based on income. That the growth rate in the "digital divide" based on education actually slightly exceeded the rate of growth in the "digital divide" based on income between 1997 and 1998 is a further, small piece of evidence that with prices of computers falling, a low level of education may remain more of a barrier than a low income to routine at-home use of the Internet.

11. See Catalog Age survey, note 4.

12. The U.S. Advisory Commission on Intergovernmental Relations estimated the nationwide state and local government revenue loss from untaxed mail order sales in 1994 at $3.3 billion. (See: Taxation of Interstate Mail Order Sales, 1994 Revenue Estimates.) According to the Direct Marketing Association, catalog sales grew 8.6 percent annually between 1994 and 1999.   If all forms of direct marketing matched the growth in catalog sales during the 1994-9 period and the share of all such sales going untaxed remained constant, a rough estimate of untaxed mail order sales in 1999 would be $5 billion. See note 14 for an estimate of 2003 revenue loss from untaxed Internet sales.

13. For forecasts of business-to-business Internet sales, see: Reuters, "E-Commerce Revenue Seen at $1.1 Trillion by 2002," June 23, 1999 (Deloitte & Touche estimate, "The Online Revolution," Wall Street Journal, July 12, 1999, p. R6 (Forrester Research, Inc. estimate); "B2B Commerce Boom Expected," For forecasts of business-to-consumer Internet sales, see: "The Online Revolution, Wall Street Journal, July 12, 1999, p. R6 (Jupiter Communications); "Internet Retail Sales in '99 Are Expected to Double," Wall Street Journal, May 18, 1999, p. A4 (Direct Marketing Association estimate); Seema Williams, Forrester Research, "Synchronous Selling," Presentation to the Federation of Tax Administrators' Annual Meeting, June 7, 1999.

14. Harley Duncan, Executive Director of the Federation of Tax Administrators, has taken projections of consumer Internet purchases in 2003 prepared by Forrester Research, Inc. and shown that they imply as much as $4 billion in lost state and local sales tax revenues in that year. (See: "State Revenue Losses from E-Commerce Underestimated," State Tax Notes, July 26, 1999, p. 245. Forrester recently predicted that consumer Internet purchases in 2004 would be $184 billion, 70 percent greater than the $108 billion forecast for 2003 that formed the basis of Duncan's estimate. See: Seema Williams, "Post-Web Retail," September 1999.

Like most industry analysts, Forrester expects business-to-business sales over the Internet to remain many times larger than consumer purchases. Forrester currently projects $1.3 trillion in U.S. business-to-business Internet sales in 2003. (See: "The Online Revolution," Wall Street Journal, July 12, 1999, p. R6.) If just one-third of these sales represent items that would be subject to state and local sales taxes, and if it is assumed that just one-third of the taxes due on this one-third of sales would go uncollected, then the 2003 revenue loss from untaxed business-to-business Internet sales (at the average 6.5 percent tax rate used by Duncan) would be $9.6 billion. (Assuming that one-third of the $1.3 trillion would be subject to sales tax seems reasonable. Computers, industrial equipment, office supplies, and shipping supplies alone account for 40 percent of the total; such items generally do not qualify for sales tax exemptions). Combining the $9.6 billion in lost revenues on business-to-business sales with the $4 billion in lost revenues on consumer purchases estimated by Duncan yields $13.6 billion. Thus, an estimated $10 billion total revenue loss from untaxed Internet sales in 2003 appears reasonable, if not conservative.

15. According to the U.S. Census Bureau, state and local governments combined spent $5.7 billion on libraries in FY95-96, the most current fiscal year for which data are available. See: www.census.gov/ftp/pub/govs/estimate/96stlus.txt .