Update on State and Local Revenue Loss From Internet Sales	 
(06-NOV-01, GAO-02-83R).					 
								 
GAO examined the Bureau of the Census' estimates of e-commerce in
1999 to determine if they provided a basis for GAO to revise the 
results of its 2000 report on sales tax issues from e-commerce	 
(GGD/OCE-00-165). In this report, GAO used a range of available  
private-sector forecasts of Internet and total remote sales as	 
the basis for several scenarios illustrating the impact of such  
sales on state and local sales and use tax revenues. The	 
scenarios showed that there is considerable uncertainty about the
size of the impacts and how various assumptions about sales,	 
compliance, and other factors contribute to that uncertainty.	 
Census' new e-commerce estimates do not provide a basis for	 
revising the results GAO found in its 2000 report. Census'	 
definition of e-commerce is broader than Internet sales. Because 
of the way firms provide information on their e-commerce sales to
Census, Internet sales cannot be separated from the broader	 
e-commerce sales at a level of detail that allows GAO to compare 
Census' results with the Internet sales estimates used in its	 
2000 report.							 
-------------------------Indexing Terms------------------------- 
REPORTNUM:   GAO-02-83R 					        
    ACCNO:   A02430						        
  TITLE:     Update on State and Local Revenue Loss From Internet     
Sales								 
     DATE:   11/06/2001 
  SUBJECT:   Economic analysis					 
	     Electronic commerce				 
	     Internet						 
	     Sales taxes					 
	     Electronic data interchange			 

******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO Testimony.                                               **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
******************************************************************
GAO-02-83R
     
GAO- 02- 83R Update on e- Commerce Revenue Losses United States General
Accounting Office

Washington, DC 20548

November 6, 2001 The Honorable Max Baucus, Chairman The Honorable Charles E.
Grassley Ranking Minority Member Committee on Finance United States Senate

Subject: Update on State and Local Revenue Loss From Internet Sales As you
requested, we have examined the Census Bureau?s estimates of e- commerce in
1999 1 to determine whether they would provide a basis for us to revise any
of the results we presented in our 2000 report on sales tax losses from e-
commerce. 2 In that report, we used a range of available private- sector
forecasts of Internet and total remote sales as the basis for several
scenarios illustrating the impact of such sales on state and local sales and
use tax revenues. The scenarios showed that there is considerable
uncertainty about the size of the impacts and how various assumptions about
sales, compliance, and other factors contribute to that uncertainty. The
purpose of this letter is to present our conclusions from a review of the
Census estimates.

In summary, we found that Census? new e- commerce estimates do not provide a
basis for revising the results in our 2000 report. Census? definition of e-
commerce is broader than Internet sales. 3 Because of the way firms provided
information on their e- commerce sales to Census, Internet sales cannot be
separated from the broader ecommerce sales at a level of detail that allows
us to compare Census? results with the Internet sales estimates we used in
our report.

In our 2000 report, we made a distinction between sales by businesses to
other businesses and sales by businesses to individual consumers. Private-
sector forecasters typically estimate these two categories of sales
separately. The Census Bureau also makes separate estimates for broad
categories of e- commerce. Census? estimate for manufacturers? value of e-
commerce shipments is its estimate that is closest in concept to the
estimates of business- to- business Internet sales that we used

1 See ?1999 Multi- sector report,? E- Stats, (Mar. 7, 2001), www. census.
gov/ estats. 2 See Sales Taxes: Electronic Commerce Growth Presents
Challenges; Revenue Losses Are Uncertain

(GAO/ GGD/ OCE- 00- 165, June 30, 2000). 3 ?The Census Bureau?s e- commerce
measures report the value of goods and services sold online

whether over open networks such as the Internet, or over proprietary
networks running systems such as Electronic Data Interchange (EDI).? (See E-
Stats (Mar. 7, 2001) www. census. gov/ estats.) The Internet sales estimates
we used in our 2000 report covered only sales transacted over the Internet.

GAO- 02- 83R Update on e- Commerce Revenue Losses Page 2 in our report.
Census? estimate of retail e- commerce is closest in concept to the

estimates of business- to- consumer Internet sales that we used. We cannot
use Census? estimate for manufacturers? value of e- commerce shipments in
1999 to revise the business- to- business Internet sales estimates that we
used for the scenarios in our report. The Census estimate includes
transactions completed over Electronic Data Interchange networks, which have
been widely used for years and were in use before the recent growth in
Internet commerce. EDI can involve transactions over the Internet, but a
large volume of EDI takes place over proprietary networks. Firms that
responded to a Census questionnaire concerning e- commerce in 1999 did not
report separate values for EDI transactions and Internet transactions. The
respondents reported total on- line sales and indicated which type of
computer network they used most frequently, but they did not provide a
percentage distribution of sales that corresponded to the various networks.
Moreover, Census? e- commerce estimates are disaggregated by broad
manufacturing sectors. The estimates of Internet sales that we used in our
2000 study were reported at a different level of detail. We are unable to
aggregate our 2000 estimates in a manner that would permit a comprehensive
and accurate comparison to the Census estimates.

Census? new estimate of retail e- commerce does not provide a reason for us
to revise the range of revenue loss scenarios that we presented in our 2000
report. The difference between Census? estimate and the high- end estimate
that we used in our report is small relative to the margins of error
associated with many of the assumptions that we needed to use in that study.
4 Census estimates that U. S. retailers sold about $27.3 billion in goods
and services on- line in 2000. This number is higher than the $24.8 billion
figure that we used for the higher scenario in our study. However, the
Census estimate includes export sales and sales by retailers to other
businesses, which were not included in the sales estimates that we used and
which would not be part of the U. S. business- to- consumer sales tax base
that we were estimating. If the exports and sales to businesses were
excluded from the Census estimate, it is likely that the estimate would not
be significantly above the value that we used.

We based our conclusions in this letter on a comparison of the documentation
for Census? e- commerce estimates and for Forrester Research?s estimates of
business- toconsumer and business- to- business Internet sales that we used
in our 2000 report. We discussed the differences between Census? and
Forrester?s estimates with individuals who were involved in preparing both
sets of estimates.

The Department of Commerce reviewed a draft of the correspondence and had no
changes to suggest. We conducted our audit work in accordance with generally
accepted government auditing standards between August and September 2001.

As arranged with your offices, unless you publicly announce its contents
earlier, we plan no further distribution of this report until 30 days after
the date of this letter. At

4 The fact that Census includes EDI in its definition of e- commerce does
not have significance for this comparison of business- to- consumer sales
estimates because EDI occurs between businesses, not between businesses and
individual consumers.

GAO- 02- 83R Update on e- Commerce Revenue Losses Page 3 that time we will
be sending copies of this letter to interested congressional

Committees and the Director, Bureau of the Census. The letter will also be
available on GAO?s home page at www. gao. gov.

If you have any questions, you may contact Jim Wozny or me at (202) 512-
9110. In addition, Demian Moore and Tara Carter made key contributions to
this letter.

Sincerely yours, James R. White Director, Tax Policy

and Administration Issues (440082)
*** End of document. ***