FROM:   Kim Alexander, CVF President
DATE:   March 13, 2001
RE:   NSF report, summary, and Aschroft's remarks

Hi Folks,

Thanks to all of you who sent me the url for the National Science Foundation's Internet voting report -- much appreciated!

Those who are interested in reading the NSF report can find a PDF version online at Additional info about this report is available at, where an html version is going to be published soon as well.

You may also be interested in reading about U.S. Attorney General John Ashcroft's new Voting Rights Initiative, launched last week. A transcript of his speech making this announcement is available from the Department of
Justice's web site at The transcript includes an interesting Q&A session with reporters at the end.

Today I'm distributing a summary of the NSF report, prepared by David Jefferson, a longtime member of the California Voter Foundation's board of directors who served on the NSF study's executive committee as well as the California Internet Voting Task Force. Comments or questions about his remarks can be sent via email to

-- Kim Alexander, California Voter Foundation, 916-325-2120,

David Jefferson's Summary of the National Science Foundation report on Internet voting:

In early 2000 then-President Bill Clinton asked the National Science Foundation to conduct a study of the prospects for voting over the Internet in public elections. The primary event in the study was a 2-day workshop held in Rosslyn, Virginia in October, 2000 and organized by the Internet Policy Institute, a small Washington think tank. The IPI assembled a distinguished panel of technical experts, political scientists, and election officials from academia, industry, government, and from the nonprofit sector from all over the country to address all of the issues and make recommendations. The report was issued on March 6, 2001, and can be found online at Although the workshop was conducted before the November presidential election that focused so much attention on our electoral systems, the subsequent report was, of course, informed by those events.

The primary conclusions of the NSF report strongly echoed those of an earlier, January 2000 report issued by the California Secretary of State's Task Force on Internet Voting (

Here are the NSF report highlights:

* Remote Internet voting, i.e. voting from home or office PCs, or any other private platforms not under the control of elections officials, presents profound security and privacy problems for which there are no good solutions on the horizon, and hence the panel recommends against fielding such systems for the foreseeable future.

Among the security problems with remote Internet voting are: the possibility of massive vote stealing through spoofing (fake voting sites) or virus and Trojan horse attacks on voters' computers; the likelihood of denial of service attacks against vote servers; the unequal distribution of Internet access (the "digital divide"); the loss of voter privacy in family and workplace situations; and the technical complexity of supporting the enormous number of combinations of platforms and software that people might wish to vote from (e.g. PCs, Macs, or various handhelds, with or without color, running Windows, MacOS, PalmOS, WinCE, Linux, or BeOS, etc., using Internet Explorer, Netscape, or other browsers, with or without Java, Javascript, etc.)

* Poll site Internet voting (in which voters come to precinct sites as they do now in order to vote, but they vote on touchscreen-equipped computers connected to the Internet) also has serious security problems but these problems are much more manageable than for remote voting. It therefore makes sense to begin certifying and deploying such systems over the next few election cycles.

* Kiosk voting systems are intermediate between remote and poll site voting systems. People would vote on machines that are similar in appearance to ATM machines; the machines would be located at convenient places around the county, and unlike poll site voting systems, they would not have to be staffed by elections officials. Such systems make a lot of sense, but before they can be certified there has to be a satisfactory political and technical solution to the problem of reliable voter authentication.

* Internet voter registration is fraught with potential for massive fraud and cannot be recommended, although Internet-based systems for modifying registration information, such as change of address, are worth study.

* Internet voting--even remote Internet voting--should not be expected to have much of an impact on voter participation, in spite of the widespread publicity behind this idea. Social scientists agree that the reasons for low voter turnout are many and complex, but they have little to do with voting convenience or distaste for the traditional methods of voting. Voting participation has consistently declined for the last several decades in the U.S. in spite of ever wider use of the more convenient mail-in ballot option that many states permit.

Back to CVF-NEWS index
Subscribe to CVF-NEWS
Contact Kim Alexander

Main Page

What's New




Contact Us

Support CVF

This page was first published on March 13, 2001 | Last updated on March 13, 2001
copyright 1994 - 2001, California Voter Foundation. All rights reserved.