Equal Vote
6/30/2004
Key Critic of Electronic Voting Says Paper Ballot Mandate Impossible for 2004
That's the headline from today's BNA Money & Politics Report story on the electronic voting report, issued yesterday by the Leadership Conference on Civil Rights and Brennan Center and discussed in the post immediately below. BNA reports that Avi Rubin -- one of the most prominent critics of electronic voting (see here) -- now says that it's "too late to impose a new requirement for such systems to have a voter-verified paper trail in time for this November's elections." The BNA Report continues:
Instead, computer scientist Avi Rubin of Johns Hopkins University endorsed a voting security plan proposed by the Leadership Conference on Civil Rights (LCCR) and the Brennan Center for Justice at New York University Law School. The new plan recommends that state and local election officials hire independent security experts to assess and test their electronic voting machines before the election.In addition to the report itself, it's worth reading Leadership Conference Executive Director Wade Henderson's statement announcing the report. As usual, Mr. Henderson provides a cogent analysis of the issue, emphasizing that about 1.5 million Americans' votes weren't counted in 2000 due to faulty equipment. He notes that:
Rubin acknowledged that he has changed his position since his high-profile testimony before the U.S. Election Assistance Commission hearing in May. At that time, he told the commission that requiring a paper trail for ballots cast on electronic voting systems was necessary to ensure public confidence in the integrity of the 2004 elections.
Rubin said he still supports the paper trail as a "medium term" solution that should be in place by 2006 to guard against possible tampering or computer bugs altering the results on an election. However, he and others said the four months before the next general election is simply not enough time to retrofit all electronic voting systems not now equipped to print paper ballots.
Touch-screen machines, or DREs, have a great deal of potential in many respects. They reduce ballot errors that result in overvotes and undervotes, they can make the jobs of pollworkers and election administrators easier so they can devote more time to dealing with other issues at polling places like voter registration problems, and they make independent voting far easier for people with disabilities and language minorities. Yet at the same time, even many of the leading experts in the fields of computer technology and security have argued that it is just too difficult to protect these machines from malfunctions or security breaches that could affect the integrity of an election – and, frankly, some of the manufacturers themselves, particularly Diebold, have been going out of their way to help convince the public that these experts are right.Mr. Henderson's remarks remind us that technology can't be assessed in a vacuum. The realities of election administration and the legal protections for the right to vote must also be given consideration. While we can all agree that security is vital, it's equally important that we not lose sight of laws protecting the rights of people of color, people with disabilities, and non-English proficient voters.
The concerns that have been raised about touch-screen systems are serious, and it's clear that we need to do what we realistically can between now and November to help build the public's confidence in the final vote count. Assuring the security of everyone's votes is no less important than having the right to vote itself.
6/29/2004
Civil Rights Groups Issue Report on Electronic Voting Security
The Leadership Conference on Civil Rights and the Brennan Center for Justice today released this report on improving the security of electronic voting systems for the 2004 elections. The report neither condones nor condemns the use of electronic voting. Instead, it wisely focuses on what can realistically be done between now and November to make electronic voting safer. Among the report's recommendations, as set forth in an accompanying press release, are:
-Retaining independent security experts to assess potential vulnerabilities through such tools as "red team" exercises to test electronic voting machines under election-day conditions. The experts also will recommend specific protocols and practical steps for improving security. The assessment should cover, at a minimum, hardware design, hardware and firmware configuration, software design and configuration, election procedures and physical security.The report recognizes that "all voting systems stand to benefit from an assessment of the sort provided for DREs here," but states that such an assessment is beyond the scope of the present report. The LCCR/Brennan Center report has garnered support from respected electronic voting skeptics, including Avi Rubin and David Dill.
-Providing a thorough training program for all election officials and workers on security procedures to ensure that the protocols recommended are not disregarded under the pressure of Election Day.
-Esablishing a permanent independent technology panel at the state and local levels that includes both experts in voting systems and computer security, and citizens representing the diverse constituencies in the region, to serve as a public monitor over the entire process and to perform post-election security and performance assessment.
-Establishing standard procedures for regular reviews of audit facilities and operating logs for voting terminals and canvassing system to verify correct operation and to uncover any evidence of potential security breaches.
-Preparing and following standardized procedures for responses to alleged or actual security incidents. This includes a standardized system for reporting and publication of incidents.
My take: The Leadership Conference has been a consistent voice of reason in the electronic voting debate. Its recommendations should be taken seriously. Let's hope that this report can move us beyond the simplistic paper-is-the-one-and-only-answer rhetoric that has dominated so much of the public discourse.
Studying How Voters Interact with Machines
From my mouth to God's ear ... well, maybe not but it's still a good idea.
A few days ago (see here) blogged the need for greater examination of how real-live human beings interact with voting technology. In this story, electionline.org reports that researchers at the University of Maryland are doing just that:
To better understand how people interact with voting systems, University of Maryland researchers this week conducted a round of field tests videotaping would-be voters casting simulated ballots, gauging their reactions and requesting a running commentary on usability. The tests were conducted at the University’s College Park campus as well as at the Leisure World retirement community in suburban Maryland.One of the machines being tested is a Sequoia DRE that generates a contemporaneous paper replica (aka, "voter verified paper trail"). Results are to be released in November or December.
“It is important to understand the machines’ design principles and determine which is the best,” said Paul Herrnson, director of the Center for American Politics and Citizenship, a University of Maryland-based group that researches issues related to the nation's political institutions, processes, and policies.
Herrnson said the group wants to make the public aware of the machines’ strengths and weaknesses as well as to help election officials understand the technology available to them and how voters respond to it.
Disenfranchised by Design
While there has been plenty of debate over voting technology since 2000, there's been far too little attention devoted to studying how ballot design affects access to the vote. "Disenfranchised by Design" is a 1998 paper by Susan King Roth, formerly of Ohio State and now associate dean at Virginia Commonwealth University. That paper can be found here. It examines how the height of text, organization of the ballot, and other factors affect equality of access. Two years before the Florida fiasco, Roth concluded that there were serious problems with the punch card system. She emphasized the need to consider "human factors" -- more specifically how the user interacts with the voting system -- when making decisions about ballot design.
My take: We need more research of this nature. Roth's comment that we need to consider human factors in designing voting systems may seem obvious. But it's surprising how little attention is being focused on this question.
Among the most pressing subjects of inquiry is the contemporaneous paper replica (or "voter verified paper audit trail"), which some argue is needed to ensure the integrity of electronic voting. While we can all agree that secure voting is essential, there's been little attention focused on how the CPR/VVPAT would affect usability. Would it confuse voters? Would it complicate the voting process for voters with disabilities or those with limited English speaking abilities? Might it actually result in more lost or mistaken votes than it saves? Until these questions are answered, it would be a serious error to mandate any particular piece of equipment -- be it a contemporaneous paper replica or some other device.
Digging in the Dirt
Wired.com offers this story on the present and forthcoming efforts of electronic voting critics. It reports on their scorched-earth tactics aimed at discrediting electronic voting ... and election officials who are trying to implement it.
"Best Practices" from K-School Symposium
The conveners of a symposium on voting technology, which was held at Harvard's John F. Kennedy School of Government on June 1, 2004, have released a list of recommended "best practices." It can be found here.
6/25/2004
Quarter of Voters Will Use Unreliable Machines
That's the headline to this story from Scripps News Service, on the slow progress in replacing punch card voting machines. It reports that:
Flaws in counting the 2000 general election were much more widespread than was generally known at the time. At least 1.6 million ballots cast in 38 states did not register a vote for president, according to a study of officially certified election returns by Scripps Howard News Service. Nearly 1.1 million of these so-called "undervotes" were cast on easily damaged cardboard punch cards or mechanical counting devices, some of which date to the early 1960s.This story provides an interesting counterpoint to those who claim that we've "rushed" to implement new voting equipment since 2000.
Fewer than half of the 105.4 million voters in the last presidential election used these aging machines, but they accounted for two-thirds of the known undervote.
Several state and local election officials said they are anxious to upgrade to more state-of-the-art electronic touch-screen voting machines or to optically scanned ballot systems.
"These delays are killing us," said Cook County, Ill., Clerk David Orr. "The people want change. But Congress and the White House moved so darned slow on this."
Testing of Voting Machines
The A.P. features this story on testimony before a subcommittee of the House Science Committee regarding the testing of voting systems. It reports that:
Much of the responsibility so far has fallen to a volunteer organization of retired and active election officials, which has certified three little-known testing companies to verify the integrity of every machine and every line of code in e-voting equipment nationwide.My take: The testing of voting systems -- including both electronic and paper-based equipment -- warrants more intensive scrutiny than it has to this point received. Nipping potential problems in the bud is much more likely to promote election integrity than the contemporaneous paper replica (aka, voter-verified paper audit trail) that some are advocating. Rather than insisting on the CPR or any other specific piece of hardware, we should be focusing on improved testing and other procedures, to ensure that voting equipment functions as it's supposed to.
After the testing companies qualify the machines and software, it's up to state and local election officials to certify them and accept them for use in their jurisdictions. These officials may not have the expertise necessary, and in many states there's nothing stopping an equipment vendor from sending an uncertified software patch directly to a local registrar to install in a machine, experts said.
6/24/2004
Back and Ready for More
I'm now back from vacation and find my inbox filled with news on the voting technology debates. Catching up will probably take a few days, but here are some noteworthy stories that have appeared over the past week:
- The Boston Globe features this story on how opposition to paperless electronic voting has delayed replacement of punch card systems.
There is no chance the issue will be resolved before the Nov. 2 election, and some states are sticking with punch cards until 2006, when new electronic-voting standards that include some form of paper record are supposed to be in place. That will result in continued use of punch cards in precincts that are home to about 19 percent of registered voters this fall, Election Data Services projects, down from about 31 percent four years ago.I'm quoted on the unfortunate civil rights consequences of failing to replace punch cards.
- Greg Palast has this report/commentary on the approximately one million black votes lost in the 2000 election. Palast is right to argue that punch cards and other forms of paper ballots result in a wildly disproportionate loss of African Americans' votes. He's incorrect, however, to suggest that implementation of electronic voting machines won't deal with the problem. While it's certainly true that problems can occur with electronic voting, the evidence shows that technology can reduce the racial gap that tends to occur with paper-based technology.
- The Desert Sun has this story on the retirement of Mischelle Townsend, Riverside's registrar and one of the most intelligent and effective election officials in the country. She's been a pioneer in effectively implementing electronic voting -- a distinction that has won her the wrath of the black box crowd and forced her to endure unfounded attacks on her integrity and reputation. After 34 years of public service, she's retiring to be nearer her elderly father in law.
6/16/2004
This Time I Mean It
As noted here, I'm on vacation this week. Regular blogging will resume next week.
6/15/2004
The Real Story on the League of Women Voters
At its meeting yesterday, the League of Women Voters adopted a resolution supporting "voting systems and procedures that are secure, accurate, recountable and accessible." The League's action prompted this A.P. story which appears to have been picked up by numerous media outlets.
The A.P. story characterizes the League as having "rescinded its support of paperless voting machines on Monday, after hundreds of angry members voiced concern that paper ballots were the only way to safeguard elections from fraud, hackers or computer malfunctions." It proceeds to quote anti-electronic voting crusader Barbara Simons (who was nominated to the LWV board but defeated by a two-to-one margin) as expressing her "joy and relief" at the new resolution.
Today, the League of Women Voters issued the following statement:
STATEMENT BY LEAGUE OF WOMEN VOTERS OF THE U.S. ON VOTING MACHINESAs I've mentioned before (see here), the League has been a leader in providing balanced and truthful information on voting technology. There's good reason to hope that it will continue to do so.
Today's Associated Press story entitled "League of Women Voters Drops
Support of Paperless Voting Machines" is misleading.
The LWVUS has just concluded its 46th biennial national convention. The
delegate body in attendance, representing 47 states, the District of
Columbia and the Virgin Islands, adopted a resolution that revises the
LWVUS stance on voting machines.
The new resolution reads, "In order to ensure integrity and voter
confidence in elections, the LWVUS supports the implementation of voting
systems and procedures that are: secure, accurate, recountable, and
accessible."
The League continues to support voting systems that are well-managed and
meet the above four criteria, including electronic voting systems. Each
voting system should be looked at on a case-by-case basis to ensure that
it meets each of these four criteria and that the operational and
management systems supporting it will be well-run.
6/14/2004
Punch-Card Ballots Live, So Poke Hard
That's the headline to this story in the Dayton Daily News, reporting on the decision of most Ohio counties to stick with "hanging chad" punch card machines despite their demonstrated defects. (Disclosure: I'm co-counsel in the ACLU of Ohio lawsuit challenging the continued use of punch card voting.) According to the DDN:
Some people wanted to make the change. Some didn't. At the last minute, the Ohio General Assembly insisted that the new high-tech machines must have a "paper trail" that would show voters that their votes had been recognized by the machine.See here for more on the state of play in Ohio. Lots of states seem to be vying for the dubious honor of being the "next Florida." Let's hope there isn't one ... but if the election is close, the continuing use of punch cards in this key swing state could spell trouble.
Lawmakers made this requirement effective in 2006. But counties that were thinking about going ahead with the new ATM-like machines in 2004 (without the paper trail, which isn't available yet) were left wondering if they'd be better off waiting to buy the whole package. Retrofitting new machines could be expensive, and no one knew where that money would come from.
Actually, even before the legislature stepped in, there were problems slowing the process down in a lot of counties (including that the counties were waiting to see what the legislature would do).
At any rate, not much attention has focused yet on the fact that Ohio will be doing things the old way. But a lot will, you can be sure, if the national campaign matures along its current lines. Many see Ohio as perhaps the pivotal state in a close election.
In 2000, almost 2 percent of the ballots cast in Ohio were not counted in the presidential tally, either because voters cast no recognizable vote for president, or because they voted for more than one candidate.
That fact wasn't much noted at the time, because President George W. Bush's margin of victory was more than 2 percent. But it will get noted now.
The record of the ATM-like machines in other states suggests that Ohio probably could have improved its performance by going ahead with the new machines, with or without paper trails. But it wouldn't have been a slam-dunk.
6/13/2004
Napa County Hopes to Continue Electronic Voting
The Napa Valley Register reports here that the county hopes to continue using its electronic voting system, despite Secretary of State Shelley's decertification order. That's no surprise, in light of the ballot-tampering accusations with the paper-based optical scan system that Napa uses for absentee voting -- see here for details. Secretary of State Shelley has now recertifed the systems used in two counties (Merced and Orange, as discussed in this story from the L.A. Times). Four other counties (Kern, Plumas, Riverside and San Bernardino) have joined disabled citizens in a legal challenge to Shelley's decertification order. A hearing is set for July 2 in that case.
On the Road . . .
I'm going to be travelling for the next week, first to a racial justice conference in Portland and then to Glacier National Park, so blogging will be sporadic. A few days with the grizzlies -- far away from newspapers and email -- should provide some welcome relief from the electronic voting controversy. Regular blogging will resume the week of June 21st.
The NYT's Vegas Road Trip
In its latest effort to discredit electronic voting, the N.Y. Times editorial page today reports on its trip to Las Vegas. Not to study electronic voting machines, but to study electronic slots.
Today's editorial -- the second this week on electronic voting (see here) -- asserts that the testing requirements for sin city's electronic vote machines are stricter than those that exist for electronic voting machines. It therefore concludes that we are "Gambling on Voting" (clever, eh?) by allowing the use of electronic voting machines. It recommends that the state have access to all software used in voting machines, that there be "spot-checking" of those machines, that standards be constantly updated, and that there be an immediate investigation if there's foul play.
My take: Comparing electronic voting to a Vegas wager is a clever bit of rhetorical gamesmanship, but doesn't hold up to careful scrutiny. Nevertheless, the NYT has a point in contending that the procedures that we use for testing and implementation of electronic voting could be improved.
I've been arguing for weeks that we should be more focused on procedures rather than hardware, when it comes to electronic voting. See here.In particular, the software used should be held in escrow by state and/or local election officials, which is already done in some places. There should also be random spot checking of voting machines to make sure they're working the way they're supposed to -- what's known as "parallel monitoring" in the world of elections. This too is already being implemented in California and some other states.
To the extent the gaming machine parallel is an appropriate one, the NYT would have done well to take that parallel to its logical conclusion. For months now, the NYT has asserted that a contemporaneous paper replica (aka, "voter verified" paper trail) is the solution to the security vulnerabilities of electronic voting. But if today's editorial is to be believed, then gaming equipment functions well without a comparable paper trail. By emphasizing (properly in my view) the importance of procedures, the NYT has unwittingly exposed the flaw in its argument that electronic voting can't be secure without a contemporaneous paper replica.
The fact that there are procedures that could enhance the security of electronic voting doesn't mean that the use of electronic voting is tantamount to a roll of the roulette wheel. What's missing from the NYT's analysis is any discussion of the comparative risks of electronic voting, in comparison with the paper-based alternatives. While there's a long history of fraud and error with paper-based voting systems -- such as the infamous punch card -- there's no documented instance of the sort of election-stealing that has some technologists so worried. That's not to say we should be unconcerned with such a possibility. But we should also keep such risks in perspective, balanced against the considerable civil rights advantages that come with electronic voting.
Those advantages are something that the NYT has almost completely overlooked. Social science evidence demonstrates that these systems result in substantial numbers of lost votes. See this page of analyses from U.C. Berkeley political scientist for a sampling. Punch cards and optical scan systems (at least some of them) also result in a substantial racial gap in uncounted votes, see here, which gap is eliminated by moving to touchscreens. If the use of electronic machines is a gamble, then the use of punch cards is a sure loser, particularly for people of color. Unfortunately, the consequence of the touchscreen-phobia that has gripped the NYT editorial page and others has been for many jurisdictions to stick with punch cards, as discussed here.
Finally, the NYT doesn't mention that at least some jurisdictions are already implementing better procedures, including some of the ones that it mentions. To be sure, there's still room for improvement. But instead of a Vegas junket, the NYT would have been better off spending its "research" budget on a visit to some less glamorous locations -- like New Mexico, Georgia, or Riverside, California -- where election officials are doing their best to make secure and accessible electronic voting a reality.
Former DNC Chair Speaks Out on Electronic Voting
Joe Andrew, the former National Chair of the Democratic National Committee, delivered a speech on electronic voting to the Maryland Association of Election Officials on June 8. Andrew identifies four goals that should guide our consideration of voting technology: accessibility, accuracy, usability, and security. He explains how electronic voting advances the first three goals, and then goes on to debunk some of the myths and conspiracy theories that have been propogated about security, relying in part on the analysis of Carnegie-Mellon computer scientist Prof. Michael Shamos (see here). Here's a sample:
I am a former election official, a lawyer who makes a career of corporate audits, and a democrat, first with a small ‘d’ and last with a capital D. As a small ‘d’ democrat, I think voting is important. Really Important—as in, to state the obvious, it is the building block of democracy. But more importantly, I believe allowing everyone to vote is the most important goal of any democracy— a goal that should trump other, competing goals.The full text of Andrew's speech can be found here, under "The Perils and Promise of Electronic Voting."
What are these competing goals?
The Four Competing Goals
When it comes to the process of voting, I believe we should focus on four things:
1) Accessibility who gets to vote and how easy is it for them to vote?
2) Accuracy: is each voter’s intention correctly recorded and tabulated?
3) Usability: can voters easily use the machine, and often more importantly, can Election Day volunteers and election officials use the machine?
4) Security: can the vote totals or the system of voting be corrupted any time during the process?
Now, let me repeat the obvious, since it is the obvious things that don’t get said. These four goals often compete with each other. We could all build a perfectly secure machine, but it would be impossible to get within ten feet of it, it would weigh a thousand pounds, and it would only be accurate in that it would record no votes since no one could vote on it....
Accessibility needs to be for everyone. As the Leadership Conference on Civil Rights has reported, older voting lever machines, punch cards, and many printed ballots, such as the butterfly ballot, have varying rates of error depending on the characteristics of voters, including their socioeconomic status and education levels. Persons with disabilities, well represented by the American Association of People with Disabilities and other groups, have reported extensively on how persons with disabilities have historically been forced to vote separately, but never equally, with voting systems that don't allow them to cast a Secret ballot or vote independently. Persons with limited English proficiency have also been prevented from having equal access to voting by machines and systems that don't recognize their needs.
The simple fact is that many forms of voting, like old fashioned paper ballots, result in discrimination, at least as it is defined by common sense, if not in the law.
Accuracy is the obvious goal of any election. People forget that. In 2000, as many as 2 million voters in America left the polling place believing that their votes would be counted, but because of obsolete and inaccurate voting systems, they were not counted at all. Those systems can hardly be described as accurate.
In California, during the Gubernatorial Recall, punch card systems failed to record a valid vote on 6.3 percent of all ballots cast. For Optical scan systems, the rate was 2.7 percent and for DREs the rate was only 1.5 percent. Electronic voting, as I will discuss, isn't perfect, but if accuracy was the only goal, then it is twice as accurate as the next available technology.
Accuracy, and therefore democracy, is not measured only by who votes, but to quote the least democratic person I can think of, Joseph Stalin, democracy is judged by who counts the votes. Clearly Stalin did a much better job of counting than Democrats have done recently, at least since the first Mayor Daley and Lyndon Johnson both left us.
Unfortunately, nearly 70 percent of Americans will cast their votes in the 2004 election on the same kind of voting systems that produced those significant error rates in 2000. But now there will be much more scrutiny because the 2000 election taught the media that balloting is never perfect and close elections are often in doubt. This additional scrutiny has already led to a clear renewal of concern about the accuracy of voting in America.
Accessibility and Accuracy are closely tied to the third goal, Usability. Voting systems must be as easy to use as possible not only for voters, but also for election officials and the legions of volunteers we all rely on to make elections work in America. Machines have to be light enough to be moved by volunteers, easy enough to be used by volunteers, and hardy enough to be stored, opened, closed, dropped and roughly handled by volunteers who only use them on average one day every two years. Studies done at MIT by Professor Ted Selker on the ergonomics of voting, and the polling done here in Maryland, Georgia, New Mexico and elsewhere, all show that voters, volunteers and election officials, usually prefer electronic voting machines to any other method of voting. Ironically, touchscreen systems are easier to use for most voters than even old fashion paper ballots.
So if the electronic voting machines are more accessible, more accurate, easier to use, why all the fuss?
Ah, here's the rub, because of the fourth competing goal-- security. Because people believe that the election can be stolen. Who are these people? Well, they tend to fall into two groups, computer experts with impressive technical experience but little practical voting experience and progressive computer users who are conspiratorial. The first group has made some excellent points that will lead to better voting systems and better machines, but they have clearly overstated their case. The second group has made some excellent partisan political points which they wrongly believe will benefit their side in the next election but again, they have mistaken their case....
The Seven Supposed Sins of Electronic Voting
Let's quickly address what I believe are the seven basic objections, The Seven Sins of Electronic Voting:
1. The Vast Right Wing Conspiracy Theory: The DRE industry only has a few real players, most of who are close to one political party -- the Republicans – and therefore Democrats and all voters shouldn’t trust them. The fact that Rush Limbaugh has been attacking DREs apparently only proves that his brain was totally fried by all those drugs.
2. The Luddite View; or the Jessica Simpson Perspective: It is a Black Box where the public cannot see how it works, and therefore we will never have any confidence in it, and confidence is everything in the voting process. There is a blonde joke in there somewhere.
3. The “Hal from 2001 a Space Odyssey” View: There is no way we can make a computer secure. So don’t trust the computer. They are all vulnerable and untrustworthy so we are all vulnerable.
4. The “Matrix View:” Computers are taking over the world: Everything can be hacked into, even the FBI’s own website— so how can these machines be safe?
5. The Machines Just Don’t Work View: DRE’s have failed all over the country, so even if they are a good thing in the abstract, they simply aren’t ready for primetime. This is the perfect, make pursuit of perfection the enemy of the good argument.
6. The Trust The Professor With The Pocket Protector Perspective: Many well-known computer scientists from prestigious universities have said that DREs should not be trusted. Since most people cannot even program their VCRs, we should defer to the authority of those with pocket protectors.
7. The Paper Messiah View: A Voter-Verified Paper trail allows the voter to see that their preferences have been recognized and this solves all problems -- including making all election officials thinner, taller and better looking.
Time may not allow me to talk in length about all seven, so I will be happy to give you the full text of this speech so you can put it on your web site and see my thoughts on all seven supposed sins of electronic voting, but I want to dwell a while on my unique perspective as a Democratic political activist....
The Vast Right Wing Conspiracy Theory
So let me first spend some time on the vast right wing conspiracy theories, because being the kind of guy who thinks that there is a little bit of politics in everything, I think that this theory is underlying much of the hype and heat that is being generated around electronic voting today. Having quoted Joseph Stalin earlier, let me also paraphrase Karl or Groucho Marx—if you understand the politics behind something you understand the something.
Remember the phrase, only Nixon could go to China? Well, it wasn’t because only Nixon had Henry Kissinger as a travel agent. Only President Nixon could normalize relations with China because he was such a staunch anti-communist. Well, as the former National Chair of the Democratic National Committee, I have spent my life in the trenches with progressive and liberal groups, fighting arm and arm for the causes we believe in. But let me say this as clearly as possible, when it comes to electronic voting, most liberals are just plain old fashion nuts.
Any one who has read any of David Brock’s books, like his classic, “Blinded by the Right,” or goes to the great new web site “Media Matters”, will see that there is a vast conservative network of organizations that promote and give ideas, money, influence and media attention to conservative candidates. Yes, they are great at coordinating message and goals for conservative politicians. But that does not mean that there is a vast right wing conspiracy trying to steal votes in America, as the loudest voices on the left are saying today....
Now this is the only partisan point I will make in this speech. Any Democrat who thinks that getting rid of direct electronic voting machines will help the Democrats win is simply out to lunch. Study after study shows that the voters who are most likely to be helped by DREs are the disabled (they are not intimidated and vote independently), the less educated, (the machines are easier to read and are more like atms, which almost everyone uses) lower socioeconomic groups(who trust machines more than they trust people—it is people who are always taking advantage of them), the truly elderly (bigger print, no pressure sensitive punching or marking on small circles), and US citizens who are more comfortable voting in a language other than English. (Older Hispanics and Asian Americans)
Now it doesn't take a genius, or even a former Chair of the Democratic Party, to tell you that those are all Democratic voters....
But what about the ultimate conspiracy theory where the vast right wing conspiracy of corporate America, with help from the evil military-industrial complex and three guys holed up in an outhouse in Idaho with automatic weapons, all get together to put malicious code made by the conservative machine manufacturer into each electronic voting machine? The secret code would that would secretly, and without leaving any detectable path, shift votes from John Kerry to George Bush.
Cue the spooky music....
This is where we must separate fact from fiction. Is there zero possibility that this could not happen? No, but we never require perfect systems in the real world. We balance the risks against the other benefits, as outlined by Professor Shamos and others who have spent a lifetime in the voting world, and make a reasoned determination of what is best under the circumstances we find ourselves in. Again, we don't let perfection become the enemy of the good.
This should give pause to some of the well-meaning progressives who have supported the call for a contemporaneous paper replica. In the forthcoming election, a majority of voters in some critical swing states will still be using punch cards, a system we know to lose lots of votes. These include approximately 70% of Ohio's voters, 65% of those in Missouri and 59% of those in Illinois. While anxiety over DREs isn't the whole reason for this, it's a big part of the reason. Democrats or Republicans, we can only hope that the election isn't close . . . at least in these key states.
6/11/2004
The N.Y. Times' Attack on the Disability Community
Strong words, but there's no other way to describe the latest New York Times editorial entitled "The Disability Lobby and Voting." Up until now, the NYT editorial page has virtually ignored the civil rights advantages of electronic voting technology, which was used by over 10% of voters in 2000. Foremost among these advantages is that electronic voting systems allow independent voting for people with disabilities. They have an audio capacity through which people unable to see or read can vote -- and verify their votes. They also allow easier independent voting for non-English proficient voters. Just as importantly, electronic voting machines -- because they are more user-friendly and allow error correction -- close the racial gap that tends to exist with paper-based systems like the notorious "hanging chad" punch card.
Given it's reputation as a liberal newspaper, one would expect that the N.Y. Times would enthusiastically support the replacement of punch cards with electronic voting systems. But instead, the NYT has chosen to take a position that -- if followed -- would effectively require many jurisdictions to retain demonstrably unreliable and discriminatory punch card voting. To make matters worse, the NYT has refused even to air the views of knowledgeable people who disagree with its alarmist position, as described here.
The current debate focuses on whether electronic voting systems should be required to generate a contemporaneous paper replica (CPR). The NYT editorial continues to insist that the CPR, more euphemistically known as the "voter verifiable paper audit trail" is the one and only solution to the problem of electronic voting security. See here.This device would require an attached printer, to print out a contemporaneous paper replica of the electronic voter at the time the voter votes. To protect the integrity of the paper replica, it would have be printed behind a glass or screen, so that the voter can see but not touch it.
In the few jurisdictions that have tried to implement the CPR on a limited basis, it has caused serious problems. In Sacramento, California, printer jams had to be removed with such creative devices as windshield wipers and back scratchers. And in Wilton, Connecticut, the deputy registrar labeled the user interface "appalling." That's because the voter has to shift his or her eyes back and forth between the screen and the printer in order to "verify" the vote. And in places with lengthier ballots, the paper replica could wind up being 57 inches long. Verifying one's votes under such conditions is a dubious proposition at best, yet the NYT steadfastly ignores these inconvenient facts.
Moreover, the contemporaneous paper records will not do any good unless they're actually counted. Contrary to what some may believe after Florida 2000, a full manual recount of ballots is not routine, but is extremely rare. Thus, even aside from the practical problems in implementing the CPR, there's no reason to believe that this device will do much to increase the integrity of elections. It's fine to experiment with the CPR. But there may be much more effective ways than paper to ensure that electronic voting machines function as they're supposed to -- as discussed here.
Blithely ignoring the practical problems with the CPR, today's NYT editorial harshly criticizes disability rights advocates for daring to oppose requiring this as-yet unproven device. Even the title of the editorial, referring to the "disability lobby" suggests the NYT's bias -- creating the false impression that disability rights activists wield great power on Capitol Hill. The editorial itself is littered with half-truths and outright misstatements that demand a response. Here are a few of them:
- "The issue is whether electronic voting machines should provide a 'paper trail' -- receipts that could be checked by voters and used in recounts."
Well, actually, that's not really the issue. The Help America Vote Act of 2002 already requires that electronic voting machines produce a paper trail, which could be used in manual recounts, by no later than 2006. Disability rights groups do not object to this kind of paper trail. The debate is instead over whether all electronic voting machines should be required to produce a particular type of paper trail, one that has yet to prove workable or effective anywhere -- more specifically, a contemporaneous paper replica of the electronic ballot, or "voter verified" paper trail. As noted above, this device has caused serious problems in the few jurisdictions that have tried to implement them.
It's therefore no surprise that few jurisdictions are rushing to implement the CPR, but are instead sticking with punch cards and other unreliable voting systems. Here in Ohio, for example, two-thirds of counties will be using the "hanging chad" punch card, as a direct result of the legislature's decision to require the CPR effective 2006. See here.
The NYT is also wrong to characterize the type of paper trail at issue as a "receipt." Under the proposed CPR system, voters shouldn't receive or even touch the paper replica of the ballot -- the whole purpose of which is to provide election officials with an audit trail. To provide a receipt that the voter could take with him or her would compromise the anonymity on which our voting system depends and can be expected to lead to vote-buying and selling.
- "Disability-rights groups have had an outsized influence on this debate despite their general lack of background on security issues. The League of Women Voters has been a leading opponent of voter-verifiable paper trails, in part because it has accepted the disability groups' arguments."
Excuse me? Where's the evidence that disability rights groups have had an "outsized" influence on the debate. One could far more plausibly argue that a small group of technologists and conspiracy theorists -- virtually none of whom have any experience in elections -- have had an impact on the debate that is grossly disproportionate to their experience in the field of election administration and voting rights.
By contrast, the League of Women Voters and the American Association of People with Disabilities are both groups that have taken the time to study the realities of election administration inside out, and recognize the practical problems with the CPR that the NYT refuses even to acknowledge. The League of Women Voters has extensively explained its reasons for opposing a CPR requirement -- see here and here. Curiously, the NYT has failed to address the practical problems that the League and others, such as the Leadership Conference on Civil Rights have repeatedly raised.
What's particularly ironic about the NYT's position is that they criticize civil rights groups for their lack of background on security, while uncritically accepting the breathless speculation of computer scientists who know little or nothing about election administration and voting rights. As I've said before, intelligent analysis of this controversy requires knowledge of 1) voting technology, 2) election administration, and 3) voting rights. Civil rights advocates (like me) and the many election officials who oppose the CPR have spent a great deal of time learning about voting technology. It sure would be nice if the most prominent DRE critics -- including the NYT editorial page -- were half as interested in learning about election administration or voting rights.
- "Some supporters of the voter-verifiable paper trail question whether disability rights groups have gotten too close to voting machine manufacturers."
This is the kind of baseless innuendo that we've come to expect from Bev Harris and other members of the "black box" crowd, whose mode of argument is to smear those with whom they disagree. Such tactics are given a false sense of dignity when their innuendos are repeated on the editorial page of the New York Times. The implication is that disability rights groups have been bought out by the voting machine manufacturers. In reality, disability rights groups have opposed the CPR for months, for good reason. It's become abundantly clear that imposing a CPR requirement will lead -- and in fact already has led -- counties to abandon their plans to switch to accessible voting machines. As a direct result of the move to require the CPR, many counties will stick with paper-based systems like the "hanging chad" punch card, equipment that is bad for everyone but especially bad for racial minorities and people with disabilities.
- "The real issue, though, is that disability-rights groups have been clouding the voting machine debate by suggesting that the nation must choose between accessible voting and verifiable voting."
Hmmm, let's take a close look at who's really clouding the debate here. It's certainly true that the choice between accessibility and verifiability is a false one ... but for a reason quite different than the one the NYT is pushing. The flaw in the NYT's argument is that they assume that "verifiable" voting is synonymous with paper. But in fact, contemporary electronic voting machines allow all voters to verify their votes before they're cast -- something paper-based systems like punch cards don't do. On the other hand, if a CPR is required, then counties will shy away from accessible voting systems. Even the most fervent proponents of the "voter verified" paper trail acknowledge that this device is still at the experimental stage. See here. Yet the mere specter that this experimental device will be required has discouraged counties from moving to electronic voting at all.
It's not at all clear that voting machine vendors -- the big bad guys in this story as the NYT would have it -- are leading the charge to stop the CPR requirement. It's possible that some may be lobbying against it. But think about this for a second. Voting machine manufacturers (who generally make all types of systems, paper-based and electroinic) will make a buck regardless of whether the CPR is required. Under these circumstances, they'd be perfectly happy to tack on a few hundred dollars to the price of each machine for an attached printer, whether or not it will actually enhance the integrity of the voting process.
- Finally, the punchline: "Disabled people have historically faced great obstacles at the polls, and disability-rights groups are right to work zealously for accessible voting. But they should not overlook the fact that the disabled, like all Americans, also have an interest in ensuring their elections are not stolen."
Well, thanks. For nothing. After paying lip-service to the enormous barriers to full and equal voting that people with disabilities still face, the NYT proceeds to kick them in the teeth -- suggesting that they don't care, and maybe are to dumb to notice, that the election will be "stolen" from under them if we implement paperless electronic voting. Let's put aside the fact that there's no evidence, empirical or otherwise, showing the CPR upon which the NYT insists is workable or effective. Where does the NYT get the idea that the election will be stolen unless we rely on paper?
I know there are many people who believe the 2000 Florida election was stolen -- and in my opinion, they've gotten a reasonable argument. But let's not forget that, to the extent that election was stolen, it was stolen with paper ballots. Where does the NYT, or for that matter anyone who wasn't in a coma during the last two months of 2000, get the idea that paper will prevent an election from being stolen? In fact, if electronic voting had been in place in those Florida jurisdictions where punch cards were used in 2000, the result would very likely have been different. Fewer votes would have been lost, especially in minority communities that tend to benefit most from voting technology that provides error notification and "second chance" voting. And you don't have to take this disgruntled liberal's word for it either. Conservative jurist Richard Posner has acknowledged that Gore might have won if more user-friendly technology had been in place in Florida.
As a proud and unabashed liberal, I find today's editorial to be a disgraceful effort, especially coming from the supposed bastion of liberal journalism. To me, liberalism means having a commitment to individual rights and liberties. It also means being committed to equality, especially in areas where rights of democratic participation are at stake. Finally, liberalism entails a willingness to carefully and reasonably examine the evidence, including facts that are not favorable to your viewpoint. Instead of abiding by these principles, the NYT has embraced a position that distorts the facts, falls prey to the worst forms of fear-mongering, and manifests disdain for the rights of our most vulnerable citizens.
6/10/2004
California Recertifies One Touchscreen System
California Secretary of State Kevin Shelley has now recertified one of the electronic touchscreen systems that he decertified in April, according to this press release. Shelley commends Merced County and the system's manufacturer ES&S, for meeting all 23 conditions required for recertification. Among those conditions are:
· Voters must have the ability to cast a paper ballot;Litigation is currently pending in federal court over Shelley's decertification order, as discussed here. Visually impaired citizens charge that Shelley's decertification of touchscreens inhibits disability access required under the Americans with Disabilities Act and other laws.
· The computer software, firmware, or hardware must be made available to the Secretary of State’s office;
· No telephone, wireless, or Internet connections are permitted on the machines; and
· The county is required to provide the Secretary of State with a comprehensive poll worker training program.
Sacramento County to Use Optical Scans in November
The Sacramento Bee offers this story on the county's plan to continue using a paper-based optical scan system in November. The story notes that: "In March, elections officials had to recreate four times the normal number of ballots - 20,000 - because many voters had not clearly indicated their choices." For the coming election, Sacramento County says it will make improvements to the system -- among which is to provide precinct-counting machines, that allow voters to check for overvotes before their ballots are actually cast.
The story also notes that, by November, the county plans "to have touch-screen machines in each polling place to mark paper ballots for people with disabilities." Sacramento is among the counties that has been sued by disabled citizens for failing to comply with the Americans with Disabilities Act and other civil rights statutes. It will be interesting to see whether this accommodation -- which to my knowledge hasn't been implemented anywhere else to date -- works as intended.
6/09/2004
EAC Chair Speaks Out on Electronic Voting
DeForest "Buster" Soaries, Chair of the Election Assistance Commission, is quoted here on the electronic voting debate. While Soaries believes that additional safeguards are needed, he doesn't support requiring that electronic voting machines produce a "contemporaneous paper replica" (aka, voter verified paper audit trail) at this time, saying instead that the issue needs more study:
"The increased use of electronic voting devices has created security concerns that the U.S. Election Assistance Commission must address," he said in remarks prepared for delivery at a Maryland conference of election officials.Among the recommendations Soaries makes is the implementation of parallel monitoring, in which randomly selected machines are tested to make sure they perform the way they're supposed to.
In an interview before the speech, Soaries said the issue of paper ballots that voters can verify - perhaps the most-debated aspect of the controversy over electronic voting - requires more study and that calling for such receipts by November would be unrealistic. He said it was possible the panel would recommend paper ballots in the future.
"If there was unanimity among scholars and scientists on the paper issue it would be a more compelling case," Soaries said. "All of the research, all of the testimony we've received, all the writings that I've read argue for more research."
Update: For the New York Times' spin on Soaries' remarks, see here.
Reassuring Voters in Napa County
Napa County Registrar John Tuteur is doing his best to reassure voters, after the conclusion of the lawsuit alleging fraud with the paper-based system used by the county's absentee voters. The Napa Valley Register reports here that:
Napa County's Registrar of Voters John Tuteur is already taking measures to prevent another election like the one in March that had him, Supervisor Mike Rippey and Supervisor-elect Harold Moskowite in court for two weeks.The reference is to the lawsuit by Rippey alleging ballot tampering in his race against Moskowite. As reported here the results of the election were upheld by the court, despite the fact that Rippey did present some evidence of tampering, as I mentioned in this post.
"I think I have a couple of tasks I need to do to make sure the public has confidence in our office," he said.
While serious problems occurred with the paper-based optical scan system used for absentee ballots, the electronic voting system used for in-precinct voting performed well, as discussed The Napa County Register also provides this editorial, noting that "there are reforms or new procedures Tuteur can implement to restore faith in the registrar's work."
My take: The morals of this story? First, paper is no guarantee of election integrity, a lesson we should have learned in 2000. That doesn't mean that paper-based voting systems are inherently insecure, any more than the problems that have occurred with electronic voting systems means that that paperless systems are inherently insecure.
What it does mean is (and this is moral #2) whatever form of voting technology is used, we should focus on having procedures in place to make sure that votes are counted as voters intended to cast them, to the extent humanly possible. To their credit, Napa County officials seem intent on getting such procedures in place by November.
Of course, with some systems -- most notably the "hanging chad" punch card -- this is extremely difficult due to the nature of the medium. See the post immediately below for more on that.
Two-Thirds of Ohioans to Vote by Punch Card in November
That according to the Columbus Dispatch, which reports:
Despite fears that Ohio could face a Florida-like election debacle this year if the state didn’t replace its antiquated voting equipment, no more than six counties will have new devices for the November presidential election.While a fee is required to view the Dispatch story, the Toledo Blade features this article on the latest news from the Buckeye state. That story reports that the four counties planning to upgrade their voting equipment this year is a "far cry from Secretary of State J. Kenneth Blackwell's original goal of deploying modern machines in all 88 counties by the time the first presidential ballot is cast Nov. 2." (Disclosure: I'm cocounsel with the ACLU in the lawsuit challenging Ohio's continuing use of punch card voting.)
In fact, an upgrade from the 2000 election is guaranteed only in the two counties that already have modernized their voting technology.
Sunday was the deadline for county elections officials to declare whether they wanted to obtain new equipment in time for the fall vote. Only those from four counties — of 31 eligible — said they wanted to change.
Even in those four, an improvement is far from certain, as the voting devices must win federal and state approval by the end of the month.
The bottom line: At least twothirds of Ohio voters will again use the punch-card ballots that proved controversial during the Florida recount of 2000. The Buckeye State, widely viewed as ground zero in the presidential contest, will have one of the highest rates of punch-card voting in the nation.
My take: Ironically, Democratic legislators continue to lead the charge against electronic "touchscreen" voting machines, with State Senator Teresa Fodor (D - Toledo) quoted as saying that counties made the "best choice" by sticking with punch cards. She's one of foremost legislative proponents of the "voter verified" paper trail -- or contemporaneous paper replica (CPR) -- that will be required effective 2006. She asks how anyone could be against the "voter verified" paper replica she endorses. Gee, I dunno Maybe because it has yet to be proven workable or effective in a real-life election, because it's not even certified in the state, and because the effect of requiring this unproven device has been to discourage Ohio counties from dumping their punch cards?
The Dem's actions are ironic because the punch cards they now tout has the "best choice" have been demonstrated to lose many more votes than other systems -- and to hit African American voters especially hard. See here for a study of the racial gap arising from punch cards and other paper-based voting systems. Are they unaware of this evidence? Or do they simply not care? The Dem's actions are doubly ironic, given that the CPR requirement that Fedor and others have championed will result in very few counties providing accessible electronic voting systems for their disabled voters.
Maybe I'm old-fashioned, but I had thought that the Democrats were the party of traditionally disenfranchised voters, such as people of color and people with disabilities. Apparently not here in Ohio . . .
6/08/2004
Election Officials Clash in California
Today's Los Angeles Times offers this story on the acrimonious debate that has erupted between county election officials and Secretary of State Kevin Shelley over electronic voting. It reports that some county officials believe that Shelley's "jeopardizing their ability to run a smooth and accurate election in November," by decertifying electronic voting systems.
"California was considered a good example of cooperation between local elections officials and the secretary of state up until January 2003," when Shelley took office, said John Tuteur, registrar of voters in Napa County. "The deterioration of the relationship of the local election community and the secretary of state's office is solely due to the actions and even personality of the secretary of state."On the other side of the debate, Shelley maintains that his only goal is to have an election that runs smoothly in November, even if that means that he must endure the criticism of local election officials along the way. See here for links to other coverage of the California controversy.
The story doesn't mention that some civil rights advocates -- particularly those representing people with disabilities and non-English proficient voters -- are also miffed at Shelley for decetifying electronic voting systems, despite their potential to allow secret and verifiable voting for these groups of citizens. In fact, disabled citizens were the ones to originally file the lawsuit (Benavidez v. Shelley) challenging Shelley's decertification order. See here. Unfortunately, the L.A. Times story doesn't address the disability rights angle of this debate.
The Electronic Voting Debate
Yesterday's Government Computer News features this story on the electronic voting debate. Here's a sample:
As the administrator of elections for Maryland, Linda Lamone believes electronic voting machines are safe and secure, even without a paper trail.The article includes a quotation from me on the problems with paper-based systems relative to those which have occurred with electronic voting systems.
“Most folks do not understand the tremendous amount of testing and security procedures that we do in Maryland,” Lamone said. “We have used electronic voting equipment for 30 elections, and we haven’t had a single problem with the equipment. People don’t like change, and when we switched from paper ballots to lever machines or from lever machines to punch card machines, we went through the same hysteria.”
David Dill, a computer science professor at Stanford University and founder of VerifiedVoting.org, said direct-recording electronic voting machines are not reliable enough.
“In 50 years of computer science, we haven’t figured out how to deal with software bugs,” Dill said. “Electronic voting is reasonable as long as there is a paper trail.”
Lamone and Dill are on opposite sides of the e-voting debate about whether electronic machines should have paper verification of votes cast. And with less than six months before November’s general election, Congress and many state legislatures are debating the issue.
While Dill and others doubt the security of e-voting machines, Maryland’s Lamone and other e-voting advocates say the technology works, and the potential for problems derives more from human error. Other voting methods are far less reliable, they say.
For example, 33 out of 16,000 voters in a recent Orange County, Calif., election received the wrong electronic ballot. “In every election where punch card machines are used, you lose at least that many votes,” said Daniel Tokaji, an assistant professor at Moritz College of Law in Columbus, Ohio, at a recent e-voting discussion hosted by the IT Association of America in Washington.
6/07/2004
Another Purge in Florida?
A purge of voters believed to be felons from Florida's rolls may in fact prevent many eligible people from casting votes, according to this story from the Tampa Bay Tribune. The story reports that the state is attempting to remove as many felons from voting rolls as possible, pursuant to a law "grounded in Old South racism" that forbids them from voting. Some of those purged, however, may actually be entitled to vote. According to the story:
* More than 43,000 Floridians are on the waiting list to have their rights restored, some of whom first learned in 2000, after voting for years, that they weren't legally entitled to vote. The restoration process can take years, and the list is growing, not shrinking.
* Hundreds of people wrongly removed from voter rolls in 2000, who never committed felonies or whose rights had been restored, may not yet have been put back on the rolls.
* A lawsuit charges that Florida's felon disenfranchisement is unconstitutional and affects up to 600,000 people.
Despite all this, state officials have just sent elections supervisors in Florida's 67 counties another list of 47,000 names of individuals who may have committed felonies in the past, telling the supervisors to purge their rolls again.
California Panel to Meet to Discuss Voting Systems
The California Secretary of State's office has published this notice that its Voting Systems & Procedures Panel will meet in July (on the 12th, 19th, and 26th) to discuss several voting systems. Among them are systems that were provisionally decertified last month, but can be used if 23 security measures are taken. See here, here, and here for prior discussions of this issue. The notice is less than perfectly clear about the subject matter, but presumably recertification will be on the table.
Groups Working to Increase Disability Vote
A group called Iowans with Disabilities in Action seeks to boost voter turnout among people with disabilities, according to this report.
Iowa’s disabled population votes at lower rates than the general population and ID Action reported that if people with disabilities voted at the same rate as other Americans, an additional 46,000 Iowa votes could be cast. The trend of low voting rates among disabled people is consistent throughout the country.The group is working on a statewide voter education campaign designed to educate voters with disabilities of their rights. It's also identified polling place access as a continuing problem.
Iowa’s disabled population encounters many obstacles other than a lack of confidence and education, including inadequate polling place accessibility and transportation. In Iowa, 250 polling places are not considered fully accessible to people with disabilities.In a related story, the St. Cloud Times features this story on Minnesota counties that have received funding to increase disability access.
Ohio Secretary of State Chides County for Inaction
A frustrated Ohio Secretary of State Ken Blackwell has written a strongly worded letter to Lucas County, challenging its board of elections for failing to purchase new voting machines. The Toledo Blade reports that two members of the board were unwilling to purchase a touchscreen system, due to the requirement that any such system produce a contemporaneous paper replica (aka, "voter verified paper trail") by 2006.
The Blade further reports that, in a two-paged letter, Blackwell wrote:
"The Lucas County board of elections has developed a disturbing pattern of not being able to make decisions on these important matters," Mr. Blackwell wrote in his decision. "I remind the members of the Lucas County Board of Elections that it is their responsibility to make election-related decisions on behalf of the citizens of Lucas County. I am once again disappointed in the lack of unity and cooperation among the members of the board."As a result, Blackwell has barred the county from buying new machines for November to replace their lever machines. The county will instead be forced to lease the same optical scan equipment used in its troubled March election. According to the Blade, the rental market for such systems has become expensive -- apparently due to the large number of counties that have decided to forego purchasing new electronic voting systems, at least for the time being.
My take: It's become increasingly clear that requiring electronic voting machines to have a CPR has, for better or for worse, stymied the replacement of antiquated paper-based systems. In Ohio, 64 of 88 counties will stick with the notorious punch card, which results in more lost votes than other systems and has an especially negative impact on minority voters. See here. At the same time, one can understand why the Lucas County Board of Elections would opt to lease equipment rather than buy a new system, given the requirement the uncertainty as to whether electronic voting machines they buy now can be retrofitted. It's not at all clear, however, that the optical scan system they'll be using is more reliable than old-fashioned lever machines. In fact, lever machines in Ohio had a lower over/undervote rate in the 2000 presidential elections than did optical scan voting systems.
To the extent that punchcards and other unreliable paper-based systems are used, the voters who can expect to suffer most are people of color, non-English speakers, and voters with disabilities. It's those voters who stand to benefit most from the conversion to electronic voting. Yet ironically, it's largely been Democrats who (as in Lucas County) have opposed implementation of paperless electronic voting. One wonders how carefully they've thought through the consequences of their position.
6/04/2004
Ballot Tampering Trial in Napa County
The Napa Valley Register reports here on the trial regarding alleged ballot tampering in a Napa County board of supervisors race in March. The challenger, Harold Moskowite, won by only 108 votes over incumbent Mike Rippey -- who is now trying to prove that ballots were tampered with. Rippey's expert testified that, on numerous ballots, more than one type of ink was found.
Moore admitted that on their own, individual paper ballots are not conclusive. "I can't exclude the possibility a voter may have picked up a different pen, but there's no logical doubt in my mind (there was tampering)," he said.For more stories on the Napa County trial, see this post in Rick Hasen's Election Law blog.
But Moore has contended there were more than 100 such ballots. On Tuesday, he said, "If I'd found one or two, it's not a problem. Ultimately it became cumulative."
My take: With all the attention being devoted to the prospect of fraud with electronic voting systems, it's worth remembering that fraud can also happen -- sometimes more easily -- with paper ballots. As I've described here the electronic voting system used in Napa precincts performed well in March. It was the paper ballots used by absentee voters that have resulted in controversy. And contrary to what some may wish to believe, those paper ballots don't always provide an unambiguous record of what really happened. Paper is no panacea.
San Diego to Put Touchscreens on Hold
The County of San Diego plans to use an optical scan voting system provided by Diebold instead of its touchscreen voting system in the November election, according to this story in the San Diego Union-Tribune. See here for my earlier discussion of San Diego problems in the March election. This is happening despite the fact that the optical scan system used by absentee voters also had problems.
In a related story last week, the S.D. Union-Tribune reported on a San Diego grand jury's finding that the problems with touchscreens that occurred in March weren't likely to recur in November. That report criticized California Secretary of State Kevin Shelley for his decision to decertify the electronic voting systems used in San Diego and other counties.
The 19-member grand jury panel criticized the decision, stating that it would undermine voter confidence in using electronic voting machines in the future.The grand jury report also recommended several measures to promote secure electronic voting. Among them is to require that provisional ballots be available, in case the electronic voting machines malfunction.
"We believe this decision was not completely based on analysis and sound logic," the grand jury stated.
About 10,000 touchscreen voting machines were in use countywide for the first time during the primary election. The county switched to the machines in 2003 after the state decertified the use of punch card ballots because of the 2000 presidential election in Florida, where hanging chads on paper ballots put the outcome in doubt.
The grand jury reported that going back to paper ballots read by optical scanners for the November election would leave the voting process "more open to irregularities" than moving forward with touchscreen voting.
"It is no small irony that the optical scan option the secretary of state recommends for use in November is the only component that actually had a failure in the vote tabulation for the 2004 primary election," the grand jury report stated.
My take: The concerns regarding electronic voting are certainly understandable, but the cure that the California Secretary of State's action has forced may well be worse than the disease. Touchscreens allow people to verify their votes before their cast, thereby minimizing the risk of mistaken votes or unintentional undervotes. They also prevent overvoting, and provide greater accessibility for people with disabilities.
Instead of using accessible touchscreens, it appears that San Diego will be using an optical scan system. While this is better with the punch card, it's unclear whether San Diego will be using a precinct-based system that allows voters to check their work before voting, to ensure that they've not mistakenly overvoting. This can help reduce the racial gap in lost votes. The San Diego Union-Tribune story doesn't make clear whether the optical scan system to be used in November provides such notice, commonly known as "second chance" voting. If it doesn't, then the county's decision is likely to have a negative impact on racial minorities.
6/03/2004
Florida's Voter Registration and Education Efforts
Hoping to avoid the problems that plagued its 2000 election, the State of Florida is undergoing an extensive voter registration and education effort, according to this report in the Daytona Beach News Journal:
The $3.2 million campaign, which will coordinate voter registration and education drives in 67 counties, is the first of its kind in the state, Volusia County elections supervisor Deanie Lowe said.Secretary of State Glenda Hood expresses confidence that the state will do better in 2004 than it did in 2000.
Funded by $3 million in federal money from the 2002 Help America Vote Act, the campaign will take voter registration into businesses, churches and schools over the summer. The remaining funds were provided by the state's association of election supervisors' Get Out the Vote Foundation and will allow elections officials to target specific populations.
"With this colossal, tireless effort being extended statewide," she said, "I have tremendous confidence that Florida will shine in November."Let's hope so.
Pennsylvania Hones in on Election Procedures
Pennsylvania Governor Ed Rendell met privately with local officials to discuss the new procedures prescribed by the Help America Vote Act, according to this A.P. story.
"We want to have as much care as possible that Pennsylvania doesn't become this year's Florida," Rendell said. "This is mostly 'an ounce of prevention equals a pound of cure.'"Among the areas of concern are complying with HAVA's requirement that first-time voters who registered by mail show identification. If they don't have such identification, then they should be permitted to cast provisional ballots.
Pennsylvania, which Democrat Al Gore won in 2000, is considered a key battleground state in the 2004 presidential race.
The story also discusses Pennsylvania's voting technology, reporting that 11 of the state's 67 counties still use punch cards.
CIO Magazine on Voting Technology Debate
The June 1 issue of CIO Magazine has this story on the electronic voting debate. It provides a useful summary of the positions expressed by various participants in the debate. Here's a sample:
The mess in Florida following the 2000 presidential election shone an unforgiving light on the problems with existing voting systems, down to the last hanging chad. In Palm Beach County alone, more than 29,000 punch-card ballots were thrown out; 19,000 from citizens who unwittingly selected more than one presidential candidate, and another 10,000 from people who either didn't select a candidate or didn't clearly punch out a vote. That was a monumental failure in a race that was ultimately decided by a margin of 537 votes out of nearly 6 million cast. Congress responded with the Help America Vote Act of 2002 (HAVA), legislation that authorizes up to $3.9 billion ($2.3 billion of which has already been appropriated) for overhauling the federal election system (see "The Help America Vote Act"). Concerned computer scientists say that pot of money is fueling a rush of vendors to offer e-voting technology without paying sufficient heed to principles of creating secure software systems.The article goes on to summarize three challenges facing electronic voting and, at the end, includes a chart summarizing the risks and benefits of different forms of voting technology. One can quibble with CIO's conclusions as to what the "best" fix is for the problems it identifies, but its encouraging to see some attention given to the comparative advantages and disadvantages of different voting technologies.
6/02/2004
Catching Up: Electronic Voting News
I've returned to the midwest from Boston, where I attended a symposium on voting technology at Harvard's Kennedy School of Government yesterday. Below are some of the stories regarding electronic voting that have appeared since Friday:
The transition to electronic voting continues to generate controversy throughout the country. For discussion of Luzerne County, PA's decision to replace lever machines with electronic voting machines, see this story. See here for news on the transition to DRE's in Montgomery County, TX and here for discussion of the debate in Douglas County, KS.
The New York Times offers this editorial on the certification process for voting machines and this story in the Sunday magazine advocating open source. The NYT continues its mantra that "voter verified paper trail" should be required -- while also continuing to ignore the problems and dubious advantages of such a requirement, discussed here and here. Apparently, the NYT seems to believe that emphatic insistence on this unproven device is a substitute for evidence and reason. On a more constructive note, the NYT suggests some changes to the certification process that are worthy of consideration. Open source is also worth considering, although it's unlikely to be the panacea that the NYT seems to suppose.
The Herkimer Evening Telegram offers this story on voting for people with physical disabilities. It states that the voting system used in New York isn't accessible to people with mobility, visual, and dexterity impairments and that, despite the Help America Vote Act, a change in state law will be required before accessible machines can be purchased.
Finally, on the humorous side, see this demo of touchscreen voting and this op-ed from Howard Dean, who appears to know about as much about voting technology as he does about running a successful presidential campaign. (Apologies to any Dean fans who happen to be reading.) Dean rails against the unreliability of DREs, overlooking the fact that the paper-based systems they're replacing generate substantial racial disparities and aren't accessible to disabled voters. He also seems to have forgotten that it was paper-based voting systems -- like the "hanging chad" punch card -- that cost his endorser Al Gore the election in 2000. Those are the very systems we'll end up using if we "shelve" paperless DRE's, as Dr. Dean prescribes.
