Cline and Neubig (1999) found revenue losses in 1998 from the failure to tax electronic commerce to be only one-tenth of one percent of total sales tax revenue. Goolsbee and Zittrain (1999) estimated that revenue losses in 1998 were less than one-quarter of one percent of sales tax revenue and that in 2003 losses would be less than two percent of total sales tax revenue. Bruce and Fox (2001b) estimated losses in 2003 to be about 1.5 percent of total state and local tax revenue. The GAO (2000) estimated that revenue losses for 2000 would be less than two percent of total sales tax revenue.
The most significant ongoing effort to quantify the economic losses from e-commerce has been provided by Bruce and Fox (2001a, 2001b).
They believe e-commerce is likely to cause total state and local government revenue loss of $13.3 billion. They estimate that by 2006 the loss will be more than triple to $45.2 billion and in 2011 the loss will be $54.8 billion (Table 7.2). Part of the loss would have occurred anyway even without e-commerce on sales, for example, which might have otherwise been made by purchasers using the telephone and catalog sales. In 2011 states will lose anywhere from 2.6 percent to 9.92 percent of their total state tax collections to e-commerce losses. In 2011 rates will have to rise between 0.83 to 1.72 percentage points to replace the total electronic commerce losses (Bruce and Fox, 2001b).
E-commerce and the Future of the State Sales Tax System g 195
State 2001 2006 2011
Alabama $177.40 $604.30 $734.40
Arkansas $143.80 $488.00 $590.90
Arizona $231.10 $799.20 $982.50
California $1,750.00 $5,952.00 $7,225.00
Colorado $200.70 $686.40 $836.20
Connecticut $190.50 $648.90 $788.20
Florida $932.20 $3,214.00 $3,944.40
Georgia $439.00 $1,517.80 $1,865.60
Hawaii $105.10 $359.20 $438.30
Iowa $111.80 $372.30 $443.70
Idaho $44.40 $151.50 $184.60
Illinois $532.90 $1,795.30 $2,161.70
Indiana $215.50 $728.50 $879.80
Kansas $134.40 $451.50 $542.20
Kentucky $158.70 $535.50 $645.80
Louisiana $302.60 $1,008.10 $1,202.50
Massachusetts $200.60 $683.00 $828.60
Maryland $194.40 $664.30 $809.20
Maine $43.10 $146.40 $177.50
Michigan $502.90 $1,696.20 $2,043.60
Minnesota $270.60 $920.60 $1,117.20
Missouri $261.60 $884.10 $1,066.70
Mississippi $136.50 $462.80 $560.00
North Carolina $293.40 $1,010.90 $1,239.40
North Dakota $26.40 $87.60 $103.90
Nebraska $70.90 $238.70 $287.30
New Jersey $337.80 $1,150.00 $1,396.10
New Mexico $129.10 $440.20 $535.40
Nevada $126.30 $441.70 $549.00
New York $1,052.90 $3,569.20 $4,318.40
Ohio $446.70 $1,502.20 $1,805.90
Oklahoma $202.80 $670.60 $794.50
Pennsylvania $446.40 $1,503.40 $1,811.00
Rhode Island $36.80 $124.50 $150.40
South Carolina $153.40 $525.00 $640.50
South Dakota $39.40 $133.40 $161.30
Tennessee $362.30 $1,242.80 $1,518.70
Texas $1,162.10 $3,957.00 $4,805.60
Utah $104.50 $359.00 $439.20
Virginia $238.50 $817.00 $997.20
Vermont $21.00 $71.70 $87.20
Washington $416.50 $1,427.30 $1,745.30
Wisconsin $213.50 $721.50 $871.00
West Virginia $70.10 $232.40 $276.20
Wyoming $26.10 $85.20 $100.00
Total $13,293.10 $45,204.30 $54,849.50
Source: Bruce and Fox (2001b).
A recent estimate by the Congressional Budget Office (2003) indicates that estimates of uncollected use taxes from all remote sales in 2003 range from $2.5 billion to $20.4 billion. Projections for 2011 of uncollected taxes from Internet commerce also vary widely, ranging from $4.5 billion to $54.8 billion.
Evidence suggests that the cost of complying with that multiplicity of tax systems, particularly for smaller firms, will exceed compliance costs for local sellers dealing with a single sales tax system. In addition, requiring remote sellers to collect and remit use taxes would have unclear effects on social costs: distortions would probably be reduced, but compliance costs would probably rise. The decision by policy makers to either grant or withhold from states the authority to collect use taxes on remote sales involves a trade-off between those two costs.
In terms of the revenue at stake an estimated 80 percent of sales in electronic commerce are from one business to another; many of these transactions are explicitly exempt, and use tax is currently being collected on the remainder. Second, a substantial share of electronic commerce to households involves services, intangibles, or goods that are not subject to sales and use taxes. Finally, some electronic commerce involves sales to households diverted from other remote vendors that lack a duty to collect use tax. In short, failure to tax electronic commerce may not be as critical because a substantial amount of revenue comes from sales to businesses, which would be taxed and most services would not be taxed. For example, suppose that households in New York order furniture from stores in New Jersey and vice versa for delivery by common carrier on ‘‘big ticket’’
items, the savings from evading the tax would be significant. Both states would be deprived of the revenue they would receive if the sales were made by local merchants. Therefore, there would be a lot of unproductive cross hauling across states (McLure, 2000).
A critical issue that has potentially important consequences for state and local revenues concerns the propensity for tangible goods to be converted into digitalized goods. In some states, sales of certain tangible property are taxable but sales of digital counterpart are not. The revenue loss estimates are overstated to the extent that this shift reduces the tax base, but most states could be expected to react quickly to such base erosion and redefine the base to include many digitalized sales. State and local governments will be confronted with several choices in the face of these revenue losses: they must either cut expenditures, increase existing sales tax rates, or shift to another tax source, such as the income tax (Bruce and Fox, 2001).
Controlling for a variety of conventional demographic characteristics such as income, education, and age, Goolsbee (2000) found that the proba- bility of buying something online grows as the local sales tax rate rises.
This author found that the coefficient on local tax rate is positive and
E-commerce and the Future of the State Sales Tax System g 197
significant, which implies that the higher the local sales tax rate, the greater the amount of money the average consumer spends online. Applying existing tax rates to the Internet, Goolsbee concluded that this would reduce the number of buyers online from 20 to 25 percent and reduce total sales from 25 to 30 percent. Furthermore, when controlling for demogra- phic similarities across generations, his results suggest that, as consumers become more aware of the tax code, they become more experienced with the Internet and hence become more sensitive to local sales taxes in their purchase decisions. There seems to be a tax sensitivity on the part of consumers that could have a negative impact on electronic commerce were sales taxes to be instituted.
Several important reforms have impacted the future of taxing e-commerce. These reforms are discussed in the following section.